Maria Knobelsdorf, University of Dortmund, Germany - Didaktik der ...
Maria Knobelsdorf, University of Dortmund, Germany - Didaktik der ...
Maria Knobelsdorf, University of Dortmund, Germany - Didaktik der ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Fakultät für Mathematik, Informatik<br />
und Naturwissenschaften<br />
<strong>Maria</strong> <strong>Knobelsdorf</strong>, Technische Universität <strong>Dortmund</strong>, <strong>Germany</strong><br />
Ralf Romeike, <strong>University</strong> <strong>of</strong> Potsdam, <strong>Germany</strong><br />
(editors)<br />
PRE-PROCEEDINGS<br />
7 th Workshop in Primary and Secondary<br />
Computing Education<br />
WiPSCE 2012<br />
November 7–9, 2012<br />
Hamburg<br />
Fachbereich Informatik
From the Conference Chairs<br />
Welcome to Hamburg and the 7th Workshop in Primary and Secondary Computing Education<br />
(WiPSCE) 2012.<br />
You are looking at the workshop’s preproceedings that contain all papers and posters being<br />
presented within the next two days. Because we give all authors the possibility to incorporate<br />
into their manuscripts feedback and comments that they will receive during the workshop, the<br />
overall final version <strong>of</strong> the proceedings will be <strong>of</strong>ficially published in the ACM digital library<br />
after the workshop (approximately in December).<br />
We hope you enjoy these contributions and their presentations during the workshop.<br />
<strong>Maria</strong> <strong>Knobelsdorf</strong> and Ralf Romeike<br />
WIPSCE 2012 Conference Chairs<br />
2
Program Committee<br />
Michal Armoni Weizmann Institute <strong>of</strong> Science, Israel<br />
Tim Bell <strong>University</strong> <strong>of</strong> Canterbury, New Zealand<br />
Yifat Ben-David Kolikant The Hebrew <strong>University</strong> <strong>of</strong> Jerusalem, Israel<br />
Roger Boyle <strong>University</strong> <strong>of</strong> Leeds, UK<br />
Torsten Brinda <strong>University</strong> <strong>of</strong> Duisburg-Essen, <strong>Germany</strong><br />
Michael E. Caspersen <strong>University</strong> <strong>of</strong> Aarhus, Denmark<br />
Paul Curzon Queen Mary <strong>University</strong> <strong>of</strong> London, UK<br />
Ira Diethelm <strong>University</strong> <strong>of</strong> Oldenburg, <strong>Germany</strong><br />
Judith Gal-Ezer The Open <strong>University</strong> <strong>of</strong> Israel, Israel<br />
Mark Guzdial Georgia Institute <strong>of</strong> Technology, USA<br />
Peter Hubwieser Technische Universität München, <strong>Germany</strong><br />
<strong>Maria</strong> <strong>Knobelsdorf</strong> (chair) Technische Universität <strong>Dortmund</strong>, <strong>Germany</strong><br />
Michael Kölling <strong>University</strong> <strong>of</strong> Kent, UK<br />
Johannes Magenheim <strong>University</strong> <strong>of</strong> Pa<strong>der</strong>born, <strong>Germany</strong><br />
Orni Meerbaum-Salant The Weizmann Institute <strong>of</strong> Science, Israel<br />
Ralf Romeike (chair) <strong>University</strong> <strong>of</strong> Potsdam, <strong>Germany</strong><br />
Ulrik Schroe<strong>der</strong> RWTH Aachen <strong>University</strong>, <strong>Germany</strong><br />
Carsten Schulte Freie Universität Berlin, <strong>Germany</strong><br />
Peer Stechert RBZ Technik Kiel, <strong>Germany</strong><br />
Chris Stephenson Computer Science Teachers Association (CSTA), USA<br />
Leigh Ann Sudol Carnegie Mellon <strong>University</strong>, USA<br />
Jan Vahrenhold <strong>University</strong> <strong>of</strong> Münster, <strong>Germany</strong><br />
Michael Weigend Fernuni Hagen, <strong>Germany</strong><br />
3
Table <strong>of</strong> Contents<br />
On the Importance <strong>of</strong> Being Earnest: Challenges in Computer Science Education . . . . . . . . . 6<br />
Jan Vahrenhold<br />
Grand Challenges in Computing Education. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7<br />
Judith Gal-Ezer<br />
The introduction <strong>of</strong> Computer Science to NZ High Schools — an analysis <strong>of</strong> student work 8<br />
Tim Bell, Heidi Newton, Peter Andreae and Anthony Robins<br />
Is Self-Efficacy in Programming Decreasing with the Level <strong>of</strong> Programming Skills?. . . . . . . . 20<br />
Michail N. Giannakos, Peter Hubwieser and Alexan<strong>der</strong> Ruf<br />
Concept <strong>of</strong> an Extracurricular Learning Environment for Computer Science . . . . . . . . . . . . . . 26<br />
Nadine Bergner, Jan Holz and Ulrik Schroe<strong>der</strong><br />
The school experiment InTech - How to influence interest, self-concept <strong>of</strong> ability in<br />
Informatics and vocational orientation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34<br />
Claudia Hildebrandt and Ira Diethelm<br />
Uncovering Structure behind Function the experiment as teaching method in computer<br />
science education. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44<br />
Carsten Schulte<br />
Agile Projects in High School Computing Education - Emphasizing a Learners’<br />
Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53<br />
Ralf Romeike and Timo Göttel<br />
Conceptual Change and Epistemological Belief Framework for Web Site Credibility<br />
Instruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63<br />
Marie Iding<br />
How Teachers in Different Educational Systems Value Central Concepts <strong>of</strong> Computer<br />
Science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67<br />
Peter Hubwieser and Andreas Zendler<br />
Ways <strong>of</strong> Planning Lessons on the Topic <strong>of</strong> Networks and the Internet . . . . . . . . . . . . . . . . . . . . . 75<br />
Ana-<strong>Maria</strong> Mesaros and Ira Diethelm<br />
Promoting Computational Thinking with Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79<br />
Cynthia Selby<br />
Preparing Teachers for Teaching Informatics: Theoretical Consi<strong>der</strong>ations and Practical<br />
Implications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83<br />
Vassilios Dagdilelis and Stelios Xinogalos<br />
Grand challenges for the UK : Upskilling teachers to teach Computer Science within the<br />
Secondary curriculum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87<br />
Sue Sentance, Adam McNicol, Mark Dorling and Tom Crick<br />
Articulating (Some) Grand Challenges <strong>of</strong> CSE at face <strong>of</strong> The Digital Age: A<br />
Socio-Cultural Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91<br />
Yifat Ben-David Kolikant<br />
4
Challenge and Creativity: Students experiences <strong>of</strong> .NET Gadgeteer . . . . . . . . . . . . . . . . . . . . . . . 96<br />
Sue Sentance and Scarlet Schwi<strong>der</strong>ski-Grosche<br />
eledSQL A New Web-Based Learning Environment for Teaching SQL at Secondary<br />
School Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107<br />
Andreas Grillenberger and Torsten Brinda<br />
Bringing Contexts into the Classroom A Design-Based Approach . . . . . . . . . . . . . . . . . . . . . . . . 111<br />
Detlef Rick, Marcel Morisse and Ingrid Schirmer<br />
E-mail for You (only?) - Design and Implementation <strong>of</strong> a Context-based Learning<br />
Process on Internetworking and Cryptography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122<br />
Andreas Gramm, Malte Hornung and Helmut Witten<br />
Comparing CSTA K-12 Computer Science Standards with Austrian Curricula . . . . . . . . . . . . 131<br />
Daniel Egger, Sabrina M. Elsenbaumer and Peter Hubwieser<br />
Information Theory on Czech Grammar Schools: First Findings . . . . . . . . . . . . . . . . . . . . . . . . . . 139<br />
Daniel Lessner<br />
Data modeling and database systems as part <strong>of</strong> general education in CSE . . . . . . . . . . . . . . . . 143<br />
Claudia Strödter<br />
The Mindstorm Effect: A Gen<strong>der</strong> Analysis on the Influence <strong>of</strong> Lego Mindstorms in<br />
Computer Science Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147<br />
Catherine Ball, Faron Moller and Reena Pau<br />
An experience in exploring the processing <strong>of</strong> formatted texts by a kynesthetic approach. . . 149<br />
Carlo Bellettini, Violetta Lonati, Dario Malchiodi, Mattia Monga, Anna Morpurgo<br />
and Mauro Torelli<br />
Teachers’ Perceptions <strong>of</strong> the Value <strong>of</strong> Research-Based School Lectures . . . . . . . . . . . . . . . . . . . . 151<br />
Jonathan Black, Paul Curzon, Chrystie Myketiak, Laura R. Meagher and Peter W<br />
McOwan<br />
Technocamps: Dod â gwyddoniaeth Gyfrifiadurol ir gorllewin pell (Bringing Computer<br />
Science to the far west) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153<br />
Roger D. Boyle, Hannah M. Dee and Fre<strong>der</strong>ic Labrosse<br />
Computational thinking in context <strong>of</strong> ”Abenteuer Informatik” . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155<br />
Jens Gallenbacher<br />
Gaming and Mathematics: A Cross Curricular Event . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157<br />
Sharon Jones, Renada Poteat and Beth Frierson<br />
Turi: Chatbot s<strong>of</strong>tware for schools in the Turing Centenary year. . . . . . . . . . . . . . . . . . . . . . . . . . 159<br />
Mathew J. Keegan, Hannah M. Dee and Roger D. Boyle<br />
Learning Fields in Vocational IT Education - Why Teachers Refrain From Taking an<br />
Opportunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161<br />
Simone Opel and Torsten Brinda<br />
Save our Turtle Robots? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163<br />
Emma Posey<br />
5
ABSTRACT<br />
On the Importance <strong>of</strong> Being Earnest:<br />
Challenges in Computer Science Education<br />
Despite an increasing number <strong>of</strong> success reports from several<br />
countries, establishing computer science as a subject<br />
worth full curriculum credit is one <strong>of</strong> the most frequently<br />
named goals in secondary computer science education. In<br />
my keynote address, I will first present a personal perspective<br />
on this issue and summarize challenges in research, recruitment,<br />
and curriculum design that have to be met before<br />
this goal can be reached in-breadth. Following up on the anticipated<br />
success <strong>of</strong> our ambitions, I will then comment on<br />
some probably even more pressing challenges in research, assessment,<br />
and teacher training that we need to be prepared<br />
to face once computer science has eventually been established<br />
as a subject worth full curriculum credit.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers and Education]: Computer and Information<br />
Science Education<br />
Keywords<br />
Secondary Computer Science Education<br />
Acknowledgements<br />
This work was partially funded by the Deutsche Telekom<br />
Stiftung un<strong>der</strong> the name dortMINT. Many <strong>of</strong> the thoughts<br />
and observations discussed are based on work done together<br />
with Holger Danielsiek, Rebecca Doherty, Arno Pasternak,<br />
Wolfgang Paul, and Renate Thies.<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WiPSCE 2012 Hamburg, <strong>Germany</strong><br />
Copyright 2012 ACM X-XXXXX-XX-X/XX/XX ...$15.00.<br />
Jan Vahrenhold<br />
Westfälische Wilhelms-Universität Münster<br />
Department <strong>of</strong> Computer Science<br />
48149 Münster, <strong>Germany</strong><br />
jan.vahrenhold@uni-muenster.de<br />
6
Challenges in Computer Science Education<br />
ABSTRACT<br />
Discussing challenges involves some knowledge <strong>of</strong> future trends.<br />
Can one say anything about the future <strong>of</strong> computer science? Can<br />
one predict the future <strong>of</strong> computing generally? In the September<br />
issue <strong>of</strong> the CACM (Communication <strong>of</strong> the ACM) in his column<br />
Peter Denning relates to this question and concludes by saying:<br />
"Don't feel bad if you can't predict the future".<br />
Thus, in my presentation I will not predict the future, but rather<br />
describe the current situation <strong>of</strong> computer science education, and<br />
the challenges we have to address at the moment, showing common<br />
concerns related to computer science education we share<br />
worldwide, and in particular Israel and the United States.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers and Education]: Computer Science Education<br />
Keywords<br />
Secondary computing education<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that<br />
copies bear this notice and the full citation on the first page. To copy<br />
otherwise, or republish, to post on servers or to redistribute to lists, requires<br />
prior specific permission and/or a fee.<br />
Conference’10, Month 1–2, 2010, City, State, Country.<br />
Copyright 2010 ACM 1-58113-000-0/00/0010…$10.00.<br />
Judith Gal-Ezer<br />
Computer Science Department<br />
Open <strong>University</strong> <strong>of</strong> Israel<br />
galezer@openu.ac.il<br />
7
The introduction <strong>of</strong> Computer Science to NZ High Schools<br />
— an analysis <strong>of</strong> student work<br />
Tim Bell<br />
Department <strong>of</strong> Computer<br />
Science and S<strong>of</strong>tware<br />
Engineering<br />
<strong>University</strong> <strong>of</strong> Canterbury<br />
Christchurch, New Zealand<br />
tim.bell@canterbury.ac.nz<br />
Heidi Newton<br />
School <strong>of</strong> Engineering and<br />
Computer Science<br />
Victoria <strong>University</strong> <strong>of</strong><br />
Wellington<br />
Wellington, New Zealand<br />
heidi.newton@ecs.vuw.ac.nz<br />
Peter Andreae<br />
Anthony Robins<br />
School <strong>of</strong> Engineering and Computer Science<br />
Computer Science<br />
Department<br />
Victoria <strong>University</strong> <strong>of</strong><br />
<strong>University</strong> <strong>of</strong> Otago<br />
Wellington<br />
Wellington, New Zealand<br />
Dunedin, New Zealand<br />
anthony@cs.otago.ac.nz<br />
peter.andreae@ecs.vuw.ac.nz<br />
ABSTRACT<br />
In 2011 New Zealand introduced computer science as a topic<br />
that students could take as part <strong>of</strong> their studies in the last<br />
three years <strong>of</strong> high school. The change was initiated in late<br />
2008, so the new material was introduced with barely two<br />
years <strong>of</strong> preparation and minimal teacher training. Despite<br />
this tight timeline, many schools adopted the new topic,<br />
and many students successfully completed assessment in it<br />
in 2011. The format <strong>of</strong> the assessment was required to be<br />
a report. In this paper we look carefully at the work that<br />
students submitted by examining publicly available information<br />
(statistics, markers’ comments and exemplars), and<br />
performing a detailed analysis <strong>of</strong> a sample <strong>of</strong> 151 student<br />
papers. We describe the nature <strong>of</strong> the assessment (which is<br />
report-based with very flexible criteria for how students can<br />
demonstrate their un<strong>der</strong>standing), and examine the kind <strong>of</strong><br />
work that students submitted to meet the criteria, drawing<br />
out good practices that enabled students to do well. A recurring<br />
theme is the importance <strong>of</strong> students being able to use<br />
personal authentic examples so that the examiner can hear<br />
the “student’s voice” in their report work, which provides<br />
evidence that the student has un<strong>der</strong>stood the topic rather<br />
than paraphrased descriptions. The analysis also reveals the<br />
value <strong>of</strong> prompting students effectively to get them engaged<br />
properly with the concepts, and identifies successful ways<br />
to achieve this in the three areas <strong>of</strong> the analysed standard<br />
(algorithms, programming languages and usability).<br />
8<br />
Categories and Subject Descriptors<br />
K.3.2 [Computer and Information Science Education]:<br />
Computer Science education<br />
General Terms<br />
Measurement, Experimentation<br />
Keywords<br />
Computer Science education, report-based assessment, high<br />
school, algorithms, programming languages, usability<br />
1. INTRODUCTION<br />
Over the period 2011 to 2013, new computer science standards<br />
are being introduced to New Zealand (NZ) schools<br />
that will give students the opportunity to explore topics<br />
such as algorithms, human-computer interaction, cryptography,<br />
AI, graphics, and other areas <strong>of</strong> computer science<br />
that previously were normally encountered for the first time<br />
in university courses. This change addresses a problem that<br />
is common to many countries, where computing in schools<br />
had become focussed on learning to use the computer as a<br />
tool rather than supporting students interested in developing<br />
new s<strong>of</strong>tware and systems. For example, a January 2012<br />
report from the UK Royal Society points out that “many<br />
pupils are not inspired by what they are taught and gain<br />
nothing beyond basic digital literacy skills such as how to use<br />
a word-processor or a database” [8]. In the US, similar complaints<br />
were being made; for example, Margolis and Goode<br />
describe having to address the issue that “computer science<br />
exists on the margins <strong>of</strong> many public school core requirements”<br />
[10], and a 2009 report on the state <strong>of</strong> computing<br />
in US high schools (surveyed in 2005 and 2007) points out<br />
that “many schools and countries, policy-makers and administrators<br />
are failing to provide students with access to the<br />
key academic discipline <strong>of</strong> computer science”, and concludes<br />
that “the results <strong>of</strong> these studies appear to indicate a continuing<br />
and troubling decline in student interest in computer<br />
science courses in high schools” [9].
These echo a 2008 report from the New Zealand Computer<br />
Society that noted: “We suspect that there is a huge number<br />
<strong>of</strong> potential computing pr<strong>of</strong>essionals who have already<br />
opted out <strong>of</strong> the discipline during secondary school, either<br />
because <strong>of</strong> the lack <strong>of</strong> relevant achievement standards, or<br />
because <strong>of</strong> the unpalatable <strong>of</strong>fering <strong>of</strong> what they are told is<br />
relevant for a future computing career” [11]. A report from a<br />
group <strong>of</strong> NZ teachers also raised serious concerns: “Computing<br />
is perceived to be second rate, is uncoordinated, un<strong>der</strong><br />
resourced, unsupported, lacking in a pr<strong>of</strong>essional body, and<br />
in dire straits” [7]. Computer science has long existed as a<br />
high school subject in Israel and South Korea [8], but the<br />
above quotes illustrate how in the early 21st century serious<br />
computing courses were essentially on the decline in the<br />
US, UK and New Zealand. These concerns triggered a rapid<br />
sequence <strong>of</strong> events beginning in 2008 that resulted in the<br />
introduction <strong>of</strong> computer science as a subject in NZ schools<br />
in 2011 [3, 4].<br />
Part <strong>of</strong> the purpose <strong>of</strong> the new topics being introduced to<br />
schools is to provide students with some grounding in CS<br />
concepts, but the primary goal is giving them exposure to<br />
the topic, providing the opportunity for them to find out<br />
if it is something they might be passionate about, without<br />
the confusion <strong>of</strong> it being presented as low-level digital literacy<br />
skills. There are several related initiatives internationally<br />
that have been started in recent years to provide<br />
formal computer science courses in high schools, including:<br />
(a) the “Georgia computes!” program “to increase interest<br />
in computing at the pre-teen level, improve quality <strong>of</strong><br />
computing education at the high school level, draw students<br />
into the un<strong>der</strong>graduate level, and make apparent to students<br />
the opportunities for graduate study in computing” [6] (announced<br />
in 2006); (b) the “Exploring Computer Science”<br />
program in the Los Angeles Unified School District, which<br />
aims to “make computer science knowledge more available<br />
to and engaging for a broa<strong>der</strong> segment <strong>of</strong> our student population”<br />
[10] (started in 2008); (c) the US “Computer Science:<br />
Principles”pilots to introduce CS as an Advanced Placement<br />
course [2] (started in 2010); (d) the new German compulsory<br />
standards introduced in some regions from 2008 [5]; and<br />
(e) the 2012 report by the Royal Society in the UK advocating<br />
sweeping changes to the ICT curriculum [8] that has<br />
resulted in planned changes. The more general CSTA K-12<br />
standards reflect a strong interest in changing how computing<br />
is taught in schools 1 .<br />
In the NZ high school context, computer science has been<br />
introduced through three new “achievement standards” that<br />
are typically taken in the last three years <strong>of</strong> high school.<br />
The focus <strong>of</strong> this paper is the first standard, which was first<br />
taught in January 2011 (the start <strong>of</strong> the school year in NZ),<br />
and is described in Section 2. The paper looks in detail at<br />
the actual work done by students un<strong>der</strong> the new standard. A<br />
sample <strong>of</strong> 151 student submissions were analysed in detail to<br />
identify both good practice and issues such as student misconceptions.<br />
From this we can learn better ways to deliver<br />
computer science topics to students in this age group.<br />
An important consi<strong>der</strong>ation for interpreting this analysis is<br />
the context and the fast timeline with which the new ma-<br />
1 http://csta.acm.org/Curriculum/sub/K12Standards.html<br />
9<br />
terial was introduced. The body <strong>of</strong> knowledge defining the<br />
area was published in August 2009, just 16 months before<br />
the standard was first taught. Between this point and the<br />
beginning <strong>of</strong> 2011 the standard had to be written, teachers<br />
prepared, and courses designed. This meant that the preparation<br />
wasn’t as thorough as it could be, but it was seen<br />
to be better to have a few schools adopt the new standards<br />
quickly (they are not compulsory) than to delay for a year.<br />
Effectively the first year was a pilot, although the scale <strong>of</strong> uptake<br />
was remarkably high: in 2011 students from 49 schools<br />
registered for the new computer science standard [4]. Adjusted<br />
by population, this would be equivalent to 3,500 high<br />
schools in the US, so the level <strong>of</strong> uptake would be a good<br />
start on the 10,000 high school target <strong>of</strong> the CS10K project,<br />
and was achieved with just two years lead time.<br />
The formal training provided to teachers was minimal; some<br />
had access to sessions at teacher events, and one-day workshops<br />
were run around the country, but there was no formal<br />
systematic training program and engagement was largely<br />
driven by individual teachers’ motivation. A key role was<br />
played by the newly-formed New Zealand Association <strong>of</strong><br />
Computing and Digital Technology Teachers (NZACDITT)<br />
in sharing information and initiating events, and a network<br />
<strong>of</strong> contacts was set up in the country’s seven universities that<br />
teach computer science. A lot <strong>of</strong> the lea<strong>der</strong>ship (including<br />
lea<strong>der</strong>s <strong>of</strong> the NZACDITT) were in Christchurch, and these<br />
people were impacted by the September 2010 and February<br />
2011 Canterbury earthquakes, which occurred just at the<br />
point that the standards were about to be taught, and had<br />
a further impact on providing support for teachers making<br />
the changes! An important component <strong>of</strong> teacher support in<br />
the New Zealand system is exemplars <strong>of</strong> student work that<br />
show what might be expected, and the computer science exemplar<br />
didn’t become available until June 2011, well into<br />
the year in which the new standard was being <strong>of</strong>fered.<br />
Despite the constraints on preparation time from introducing<br />
the new standards so quickly, and adoption <strong>of</strong> the standards<br />
being optional, the number <strong>of</strong> students attempting the<br />
computer science standard in its first year <strong>of</strong> introduction<br />
was still substantial (1,429 students in 49 schools registered<br />
for the new CS standard [4], and 654 were submitted by the<br />
deadline). To give a sense <strong>of</strong> scale, there are currently (in<br />
2012) about 485 schools teaching at high school level in NZ,<br />
although many <strong>of</strong> the schools are specialist or smaller rural<br />
schools. In NZ there are about 62,500 students at “Year 11”<br />
(the third to last year <strong>of</strong> high school, typically around 15–<br />
16 years old), which is the year that most students would<br />
attempt the new computer science standard. The first year<br />
was essentially a pilot, but nevertheless had sufficient uptake<br />
that it represents widespread national engagement.<br />
In this paper we analyse the student submissions in detail<br />
based on three sources: <strong>of</strong>ficially released statistics and commentary<br />
about the grading, the published exemplars <strong>of</strong> student<br />
work, and a detailed analysis <strong>of</strong> a sample <strong>of</strong> 151 student<br />
submissions. Section 2 explains the new computer science<br />
standard and how it is assessed; Section 3 reviews the publicly<br />
available information about the submissions, Section 4<br />
is a detailed analysis <strong>of</strong> the sample <strong>of</strong> student submissions,<br />
and Section 5 reports some lessons learned based on the observations.
2. THE AS91074 STANDARD<br />
During the final three years <strong>of</strong> high school in NZ students are<br />
focused on the “National Certificate <strong>of</strong> Educational Achievement”(NCEA),<br />
which involves completing a number <strong>of</strong>“standards”.<br />
Details <strong>of</strong> how computing appears in these standards<br />
are given in previous work [3, 4], and in this paper we focus<br />
on one particular standard which was introduced to schools<br />
in 2011, called“AS91074: Demonstrate un<strong>der</strong>standing <strong>of</strong> basic<br />
concepts from computer science” 2 . The standard is set<br />
and regulated by the New Zealand Qualifications Authority<br />
(NZQA).<br />
The AS91074 standard is worth 3 credits, which translates<br />
into approximately 30 hours <strong>of</strong> student work, so it can be<br />
used as a relatively small component <strong>of</strong> a larger course, typically<br />
combined with other standards on programming, web<br />
design, digital media, information systems, electronics and<br />
so on. It is a “level 1” standard, which means it is generally<br />
used in the third to last year <strong>of</strong> high school; levels 2 and 3<br />
apply to the final two years respectively.<br />
The standard covers three main topics from computer science,<br />
which can be summarised as:<br />
algorithms: the difference between algorithms, programs,<br />
and informal instructions; the kinds <strong>of</strong> steps that are<br />
used to make an algorithm; and comparing the “cost”<br />
<strong>of</strong> different algorithms for the same problem<br />
programming languages: the kinds <strong>of</strong> programming languages<br />
that exist (including high and low level languages)<br />
and the roles <strong>of</strong> compilers and interpreters<br />
usability: evaluating and comparing user interfaces in terms<br />
<strong>of</strong> usability<br />
These topics provide a broad exposure to ideas from computer<br />
science, ranging from the mathematical analysis <strong>of</strong><br />
algorithms to the human-factors aspects <strong>of</strong> usability evaluation.<br />
They are broad rather than deep — students don’t<br />
have to implement algorithms, but they can run programs<br />
that are provided; they don’t write compilers, but they need<br />
to document how they use them; and they don’t design interfaces,<br />
but instead learn to look at existing interfaces critically.<br />
AS91074 is the first in in a series <strong>of</strong> computer science standards<br />
(the level 2 and 3 standards are being phased in during<br />
2012 and 2013) that together cover a wide range <strong>of</strong> topics<br />
from the ACM curricula. The general area <strong>of</strong> “Programming<br />
and Computer Science” includes two other standards at level<br />
1, AS91075 (which is essentially about designing computer<br />
programs) and AS91076 (implementing programs), with further<br />
standards at levels 2 and 3. These are not analysed in<br />
detail in this paper, but come up occasionally because most<br />
students doing the CS standard are also attempting the programming<br />
ones.<br />
2 It is also known as “Digital Technologies 1.44”, which is<br />
the number it had when in draft. AS91074 is available<br />
publicly from http://www.nzqa.govt.nz/ncea/assessment/<br />
search.do?query=Digital+Technologies&view=all .<br />
10<br />
AS91074 is an “external” standard, which means that it is<br />
assessed at the national level; the majority <strong>of</strong> standards (including<br />
the programming ones) are “internal” and assessed<br />
at the school but mo<strong>der</strong>ated (i.e. audited) nationally. For<br />
various reasons, AS91074 is assessed using a student report,<br />
which means that the student submits a report on what is<br />
essentially project work. The report is limited to 14 pages.<br />
The “AS” in the title stands for “achievement standard”,<br />
which means that it can be awarded with four levels <strong>of</strong><br />
achievement: Not achieved, Achieved, Merit and Excellence,<br />
abbreviated as N/A/M/E respectively. Thus unlike many<br />
other grading systems, “A” means a minimal pass, and “E”<br />
is the best result; “N” means that the student gets no credit.<br />
The standard itself is very brief — just over 2 pages, almost<br />
half <strong>of</strong> which is boilerplate information such as review dates<br />
and standard text. The main content <strong>of</strong> the standard is the<br />
criteria for the A, M, and E levels <strong>of</strong> achievement, which are<br />
given in just a few “bullet points” — 4, 5, and 4 points for<br />
A, M, and E respectively. The bullet points for M and E<br />
generally describe the higher standard <strong>of</strong> work required for<br />
those levels rather than additional content. For example,<br />
one <strong>of</strong> the Achieved bullet points is:<br />
• describing an algorithm for a task, showing un<strong>der</strong>standing<br />
<strong>of</strong> the kinds <strong>of</strong> steps that can be in an algorithm,<br />
and determining the cost <strong>of</strong> an algorithm for<br />
a problem <strong>of</strong> a particular size.<br />
The equivalent point at Merit is:<br />
• showing un<strong>der</strong>standing <strong>of</strong> the way steps in an algorithm<br />
for a task can be combined in sequential, conditional,<br />
and iterative structures and determining the<br />
cost <strong>of</strong> an iterative algorithm for a problem <strong>of</strong> size n.<br />
And the one at Excellence is:<br />
• determining and comparing the costs <strong>of</strong> two different<br />
iterative algorithms for the same problem <strong>of</strong> size n.<br />
Each <strong>of</strong> the three criteria require that the student shows that<br />
they can determine the “cost” <strong>of</strong> an algorithm (typically the<br />
running time or number <strong>of</strong> steps/comparisons); if they simply<br />
report the time from a single experiment then they have<br />
reached the Achieved level; if they compare different values<br />
<strong>of</strong> n (i.e. show the shape <strong>of</strong> the cost function) that would<br />
reach the Merit level; and if they compare two algorithms<br />
(typically a graph with two curves) then that would be Excellence.<br />
Of course, the markers will be looking for more<br />
than just numbers: the student needs to present the data<br />
well and discuss the trends that it shows.<br />
The brevity <strong>of</strong> the standard gives great flexibility to teachers<br />
and students. In the algorithms example above, they can<br />
explore whatever algorithms they choose as long as the chosen<br />
algorithms can be used to demonstrate that the student<br />
has met the criteria. Of course, this flexibility also puts a<br />
lot <strong>of</strong> responsibility on teachers and students to choose good
examples, and in practice teachers share ideas and subject<br />
experts publish advice to guide these choices.<br />
Marking <strong>of</strong> the student submission is done “holistically”,<br />
which means that if a student does very well in most areas<br />
but has problems in just one point, they won’t get a<br />
low grade just because they didn’t meet one basic criterion.<br />
For example, in general a student will have met all 4 <strong>of</strong><br />
the Achieved bullet point criteria to be given the Achieved<br />
grade. It may be that their work was weak in one <strong>of</strong> the<br />
criteria, but if they had done particularly well in another<br />
(perhaps at Merit or Excellence level) then they would be<br />
given achieved overall.<br />
The use <strong>of</strong> a report for assessment is challenging. To“demonstrate<br />
un<strong>der</strong>standing <strong>of</strong> basic concepts from computer science”,<br />
a student report cannot just paraphrase definitions<br />
or on-line articles about the topic. For example, a report<br />
that describes selection sort and states that its cost is O(n 2 )<br />
does not constitute a demonstration <strong>of</strong> un<strong>der</strong>standing, since<br />
this could be simply parroting text from a teacher or a paraphrase<br />
<strong>of</strong> one <strong>of</strong> many descriptions available online. Similarly,<br />
just listing the properties <strong>of</strong> a good user interface could<br />
be paraphrased from a textbook, with no un<strong>der</strong>standing. Instead,<br />
students need to report on a personalised activity or<br />
investigations in or<strong>der</strong> to demonstrate that they have acquired<br />
their own un<strong>der</strong>standing <strong>of</strong> the concepts.<br />
There are many ways to personalise the reports to meet<br />
this expectation. For example, timing an implementation <strong>of</strong><br />
selection sort on their own computer using their own input<br />
data reflects a personal experience with the issue <strong>of</strong><br />
quadratic behaviour as n increases. Given the number <strong>of</strong><br />
possible permutations <strong>of</strong> even a small sequence <strong>of</strong> values,<br />
it is easy for students to example traces data that will be<br />
unique to themselves. For the programming language topic,<br />
a demonstration <strong>of</strong> a program written in some language,<br />
and compiling and running it can be done use the local environment<br />
in the student’s own account, using a program<br />
they have written themselves (perhaps for the programming<br />
standard, or even a simple modification <strong>of</strong> a “Hello World”<br />
program that uses their name). For the HCI topic, a student<br />
can apply the list <strong>of</strong> principles to the interface <strong>of</strong> a device<br />
that they personally own, such as an alarm clock or an mp3<br />
player.<br />
3. STUDENT SUBMISSIONS<br />
There is consi<strong>der</strong>able public information available about the<br />
student work for the AS91074 standard, which we have collated<br />
and analysed in this section. The three areas we look<br />
at are the student grade statistics from 2011, the published<br />
exemplars, and the markers’ comments on the work.<br />
3.1 Student grade statistics<br />
The New Zealand Qualifications Authority publishes statistics<br />
on student performance in each Achievement Standard 3 ,<br />
and in Table 1 we show the number <strong>of</strong> candidates for the<br />
three Programming and Computer Science standards, along<br />
with the success rates <strong>of</strong> students (Not achieved, Achieved,<br />
Merit and Excellence).<br />
3 http://www.nzqa.govt.nz/studying-in-new-zealand/<br />
secondary-school-and-ncea/secondary-school-statistics/<br />
11<br />
In 2011, the first year that the AS91074 achievement standard<br />
was available, 654 student reports were submitted.<br />
The programming standards were significantly more popular,<br />
with nearly five times as many students taking those<br />
standards. This isn’t surprising because the topic is better<br />
un<strong>der</strong>stood, and is one that many teachers are familiar with<br />
and would have been more confident to <strong>of</strong>fer it.<br />
To put the participation rates in perspective, there were<br />
about 62,500 Year 11 students in school in NZ in 2011 4 ,<br />
so participation in the new standards was relatively high<br />
consi<strong>der</strong>ing that at least 75% <strong>of</strong> schools didn’t <strong>of</strong>fer them [4].<br />
The student grade distribution for AS91074 is typical <strong>of</strong><br />
achievement standards, with relatively few students reaching<br />
the Excellence level, but the majority getting credit for the<br />
standard (Achieved or better). The programming standards<br />
have a higher proportions <strong>of</strong> Excellence grades — in fact, the<br />
results show similar numbers <strong>of</strong> students getting Excellence<br />
and Merit, and reflects the bimodal distribution <strong>of</strong> performance<br />
that has commonly been observed for programming<br />
courses [13]. We also note that there is a difference in grades<br />
between male and female students; fewer females attained<br />
the Excellence grades, with the difference being particularly<br />
noticeable for the programming standards.<br />
It would appear that <strong>of</strong>fering these standards early in high<br />
school has attracted a higher proportion <strong>of</strong> female students<br />
than advanced courses typically do. The proportion <strong>of</strong> female<br />
students taking the computer science standard (AS91074)<br />
was 27%, and the related programming standards (AS91075<br />
and AS91076) had 30% and 36% respectively. While low,<br />
this fraction, which is for students taking the standards<br />
in their third to last year before university, is significantly<br />
higher than the proportions who enrol in university computer<br />
science courses — universities in NZ typically have<br />
about 10 to 20% female students in first-year computer science<br />
classes — so there is some hope that <strong>of</strong>fering computer<br />
science earlier in the education system may reduce the gen<strong>der</strong><br />
imbalance.<br />
3.2 Public exemplars<br />
For privacy reasons our main analysis <strong>of</strong> student work in Section<br />
4 is limited to broad statistics and observations about<br />
the sample reports. However, there are some public samples<br />
<strong>of</strong> student work: after the 2011 reports were marked, a<br />
set <strong>of</strong> “exemplar” reports was selected (two or three at each<br />
level <strong>of</strong> A/M/E) for publication 5 . The rea<strong>der</strong> is encouraged<br />
to view these to get an idea <strong>of</strong> the kind <strong>of</strong> work students<br />
are submitting, and we will use some examples from these<br />
in our discussions. The exemplars include annotations from<br />
the markers to show how the criteria have been met.<br />
The exemplars are provided to guide teachers on the standard<br />
<strong>of</strong> work that is expected, but because they are public,<br />
it has to be assumed that students will have access to them.<br />
This has both benefits and problems; modelling good practice<br />
is clearly helpful pedagogically, but it also provides the<br />
4 http://www.nzqa.govt.nz/assets/About-us/Publications/<br />
ncea-annualreport-2011.pdf (page 13)<br />
5 These are available from http://www.nzqa.govt.nz/<br />
assets/qualifications-and-standards/qualifications/ncea/<br />
NCEA-subject-resources/Technology/91074-exp-2011.zip.
Table 1: Student success rates in Programming and Computer Science standards 2011<br />
Candidates Gen<strong>der</strong> N/A A M E<br />
AS91074 Demonstrate un<strong>der</strong>standing <strong>of</strong> basic<br />
concepts from computer science<br />
654 32.7% 37.2% 18.2% 11.9%<br />
Male 475 72.6% 33.3% 36.0% 18.3% 12.4%<br />
Female 179 27.4% 31.3% 40.2% 17.9% 10.6%<br />
AS91075 Construct an algorithmic structure<br />
for a basic task<br />
2131 34.3% 31.5% 16.1% 18.1%<br />
Male 1493 70.1% 34.3% 31.3% 15.0% 19.4%<br />
Female 638 29.9% 34.3% 31.8% 18.7% 15.2%<br />
AS91076 Construct a basic computer program<br />
for a specified task<br />
3245 31.8% 33.2% 18.0% 17.1%<br />
Male 2082 64.2% 29.9% 33.4% 17.6% 19.2%<br />
Female 1163 35.8% 35.3% 32.8% 18.7% 13.3%<br />
temptation for students to simply paraphrase the model answers<br />
in attempt to imitate the best practice, rather than<br />
generate their own authentic report from scratch.<br />
The first Excellence exemplar covers the criteria for the<br />
standard, although it isn’t perfect. It draws a sound conclusion<br />
about binary vs linear search, including analysing<br />
the time needed for linear and binary search <strong>of</strong> 100 items;<br />
linear search was estimated as “1-100” comparisons; binary<br />
search is estimated at 7, calculated by repeatedly dividing<br />
the number in half (i.e calculating log2100 without using logarithms).<br />
The material about programming languages compares<br />
Pascal and Scratch with binary code, which provide<br />
a good contrast. The HCI report looks critically at some<br />
aspects <strong>of</strong> two mobile touch-screen devices; it is a little adhoc,<br />
but does provide a balanced view based on personal<br />
experience. This report illustrates the holistic marking and<br />
the focus on the criteria: despite being rated an Excellent<br />
report, it does contain some confused reasoning; for example<br />
the report claims “A binary search uses the yes/no, on/<strong>of</strong>f,<br />
0/1 concept” (perhaps confusing binary search with binary<br />
numbers); and tells us that “[Pascal] cannot be used with a<br />
mouse, and can only be controlled by the keyboard”. These<br />
are remin<strong>der</strong>s that students are encountering the terminology<br />
and environments for the first time and are still making<br />
sense <strong>of</strong> them.<br />
The second Excellence example is longer and more carefully<br />
presented. It compares two sorting algorithms using the<br />
a balance scale activity from CS Unplugged (csunplugged.<br />
org/sorting-algorithms), and includes notes and photos <strong>of</strong><br />
the activities (Figure 1) that provide good evidence that<br />
the student has engaged in the process. The number <strong>of</strong><br />
comparisons used for n = 5, 10, 15 and 20 are given for a<br />
single manual simulation <strong>of</strong> selection sort and quicksort using<br />
a balance scale. The report provides a graph without<br />
axes (Figure 2), which shows the difference between the two<br />
algorithms, and a good conclusion is drawn, that the “lower<br />
cost [<strong>of</strong> quicksort. . . ] will become more noticeable as the<br />
number <strong>of</strong> items compared increases.” This is an articulation<br />
<strong>of</strong> the difference between n 2 and n log n time; it does<br />
have a slight imprecision in that the word“compared”should<br />
be “sorted,” but it is reasonable evidence that the student<br />
un<strong>der</strong>stands the relationship, which is the goal <strong>of</strong> the standard.<br />
The HCI work in this exemplar mentions a visit to a<br />
university Usability Laboratory, and this would have been<br />
an excellent opportunity to see HCI evaluation happening<br />
in an authentic context and no doubt contributed to the<br />
12<br />
Figure 1: Notes and photo <strong>of</strong> experiment using balance<br />
scale activity from an Excellence exemplar<br />
Figure 2: Graph from an Excellence exemplar<br />
student’s appreciation <strong>of</strong> the topic.<br />
At the other end <strong>of</strong> the scale, the Achieved exemplars show<br />
work from students who have engaged with the topic, but<br />
haven’t made in-depth observations. For example, the first<br />
Achieved exemplar discusses bubble sort, and gives a trace<br />
<strong>of</strong> the algorithm being applied to 8 names (Figure 3). At the<br />
end <strong>of</strong> this example the writer gives the number <strong>of</strong> comparisons<br />
and swaps for that example, but makes no comment<br />
comparing this with different values <strong>of</strong> n, or with different<br />
algorithms. The second Achieved exemplar uses a different<br />
approach, showing a screenshot from an on-line visualisation<br />
<strong>of</strong> bubble sort (Figure 4), and using the statistics from the<br />
visualisation to determine the cost <strong>of</strong> the algorithm. In fact,<br />
the exemplar goes on to compare bubble sort and selection<br />
sort for different values on n, but there is no explanation <strong>of</strong><br />
how these were measured, and more significantly, the other<br />
sections on programming languages and usability are weak,<br />
and hence the overall grade is Achieved.<br />
Exemplars are also available for work meeting the Merit criteria.<br />
These show more advanced un<strong>der</strong>standing and discussion<br />
than the Achieved exemplars, but don’t have the<br />
completeness <strong>of</strong> the Excellence ones.<br />
Doing well in the report will be easier for students who are<br />
good at writing, and because there are also elements <strong>of</strong> math
Figure 3: A trace <strong>of</strong> bubble sort from an Achieved<br />
exemplar<br />
Figure 4: Using a visualisation to demonstrate an<br />
algorithm from an Achieved exemplar<br />
and scientific experimentation, students with strengths in<br />
science and math will also find it easier to do well. We consi<strong>der</strong><br />
this to be a positive, since writing, math and general<br />
science skills are valuable for computer scientists, but this<br />
emphasis represents a significant change from the computing<br />
classes before 2011 which were generally focussed more<br />
on using computers and tended not to attract students with<br />
good science and math skills.<br />
It’s also important to note that these students are working<br />
on topics that have traditionally been taught to university<br />
students, and one has to avoid the trap <strong>of</strong> expecting them<br />
to perform at the same level as more experienced students.<br />
Inevitably the average level <strong>of</strong> experience in writing, science<br />
and math will be lower, and it is important to calibrate<br />
expectations appropriately.<br />
3.3 Markers’ assessment report<br />
After the grading <strong>of</strong> the reports was complete (January 2012),<br />
the material was returned to the students and a report from<br />
the markers was released 6 .<br />
The markers’ report began by commenting on the quality <strong>of</strong><br />
the presentation. For example, the report notes that “small<br />
screen shots, text too small to read, and graphs with unlabeled<br />
axes did not contribute to a demonstration <strong>of</strong> un<strong>der</strong>standing.”<br />
Learning to conform to guidelines is good practice<br />
for future CS practitioners, and being in the habit <strong>of</strong> presenting<br />
data clearly is valuable! This is a benefit <strong>of</strong> requiring<br />
a report from students rather than examinations, and helps<br />
6 http://www.nzqa.govt.nz/nqfdocs/ncea-resource/reports/<br />
2011/level1/technology.pdf<br />
13<br />
to emphasise the importance <strong>of</strong> communication skills. It<br />
also reflects the value <strong>of</strong> students being aware <strong>of</strong> good experimental<br />
method (particularly for reporting on algorithm<br />
performance). Those who had a weaker science background<br />
might not be familiar with the process <strong>of</strong> collecting, reporting<br />
and discussing results in a precise manner.<br />
Another issue raised by markers was the personalisation <strong>of</strong><br />
the work — the markers said that “candidates who clearly<br />
demonstrated un<strong>der</strong>standing <strong>of</strong> basic concepts from computer<br />
science wrote in their own voice, providing evidence<br />
from their own work and experience to support any factual<br />
or referenced material.” The markers also pointed out that<br />
reports that were simply paraphrasing other sources didn’t<br />
make it clear that the student had un<strong>der</strong>stood the topic.<br />
The public availability <strong>of</strong> exemplars will make it even more<br />
important for the student’s “voice” to be heard i.e. to make<br />
sure the report is from personal experience on examples that<br />
are likely to be unique to the student.<br />
Regarding the human-computer interaction work, the markers<br />
observed the common problem <strong>of</strong> confusing “functionality<br />
<strong>of</strong> devices with usability.” Students very quickly focused<br />
on the physical form <strong>of</strong> an interface (buttons and menus),<br />
or the functionality <strong>of</strong> the s<strong>of</strong>tware, without identifying the<br />
human factors that could make it difficult for a user to find<br />
their way around the interface.<br />
The markers’ report concludes with a summary <strong>of</strong> what was<br />
required at each level, which is essentially the criteria <strong>of</strong><br />
the standard, although it includes typical problems found<br />
in “Not Achieved” work, which (apart from not meeting the<br />
criteria) were a lack <strong>of</strong> detail in descriptions, paraphrasing<br />
without un<strong>der</strong>standing, and not addressing all the requirements<br />
when using a template.<br />
The marking process is challenging because the standards<br />
are very open-ended, and work can be submitted based on<br />
a wide range <strong>of</strong> contexts. The standards are also new and<br />
there is no past history to draw on. We note that markers<br />
need to be well briefed, and provided with detailed guidelines.<br />
In particular they need to be familiar with relevant<br />
material from popular sources on the web (starting with but<br />
not limited to Wikipedia) so as to recognise material that<br />
has been used without un<strong>der</strong>standing. This is particularly<br />
important in some sub-areas (like programming languages).<br />
4. ANALYSIS OF STUDENT PAPERS<br />
To un<strong>der</strong>stand the student work better we manually analysed<br />
in detail 151 sample reports from the 654 AS91074<br />
submissions in 2011, noting the choice <strong>of</strong> examples that each<br />
student used and how they presented their un<strong>der</strong>standing <strong>of</strong><br />
the topic. The sample was selected by NZQA, who are responsible<br />
for grading the student work. All <strong>of</strong> the sampled<br />
papers were marked by one marker; the marking process<br />
has checks in place to ensure consistency between markers<br />
so this should not introduce a significant bias. The selection<br />
process could not be rigorously randomised, but because<br />
student submissions are essentially shuffled to prevent one<br />
marker getting all the reports for one school, the sample is<br />
reasonably broad. Table 2 shows the grades <strong>of</strong> the sampled<br />
reports compared with those <strong>of</strong> all submissions.
Table 2: Student success rates in AS91074 2011 (all<br />
submissions and sample analysed)<br />
Source Number N/A A M E<br />
All submissions 654 32.7% 37.2% 18.2% 11.9%<br />
Sample analysed 151 36.4% 35.1% 18.5% 9.9%<br />
Table 3: Number <strong>of</strong> pages in student submissions<br />
for AS91074 in 2011 (sample <strong>of</strong> 151 submissions)<br />
All Not Achv Achieved Merit Excellence<br />
Submitted pages<br />
Average 6.5 4.6 6.4 8.4 10.9<br />
Min 1 1 2 4 8<br />
Max 16 13 14 13 16<br />
Student-written content<br />
Average 5.4 3.4 5.2 7.3 9.8<br />
Min 0.5 0.5 1.5 3.5 6.5<br />
Max 16 12.5 11.5 12 16<br />
We discuss the student work here first in terms <strong>of</strong> the length<br />
<strong>of</strong> the document, then un<strong>der</strong> the three main topics <strong>of</strong> the<br />
standard: algorithms, programming languages and usability.<br />
We conclude by discussing general issues with the writing<br />
style, and the effect <strong>of</strong> teacher advice on the students’ work.<br />
As would have been the case for grading the student work,<br />
some <strong>of</strong> our analysis <strong>of</strong> the reports involved judgement calls<br />
on what a student’s intention was, including dealing with<br />
inaccurate descriptions and confusing text, so the figures<br />
reported below are approximate, but indicative <strong>of</strong> the kind<br />
<strong>of</strong> work that students submitted.<br />
4.1 Report page length<br />
The reports are limited to 14 pages (and markers will not<br />
mark any additional pages). An important question is whether<br />
this limit constrains the students. The number <strong>of</strong> pages submitted<br />
by students in the sample <strong>of</strong> 151 analysed documents<br />
is shown in Table 3. The first set <strong>of</strong> figures shows the number<br />
<strong>of</strong> pages the students submitted, and the second set shows<br />
our estimate <strong>of</strong> actual student-written content, which ignores<br />
the teacher-generated rubrics (which is the most common<br />
“padding” in the submissions), blank areas, tables <strong>of</strong><br />
contents, and reference lists (which were rare).<br />
Only 4 <strong>of</strong> the 151 sampled submissions used the full 14 pages<br />
(in fact, one <strong>of</strong> those 4 used more than 14 pages, but the<br />
extra pages would have been ignored by the markers). Most<br />
students did not seem to be unduly constrained by the limit,<br />
and 91% <strong>of</strong> the submissions used 11 pages or fewer.<br />
The average number <strong>of</strong> pages grows with the quality <strong>of</strong> the<br />
work, which isn’t surprising since more information is needed<br />
to cover all the criteria for Excellence. However, having a<br />
long document doesn’t necessarily predict a good grade; one<br />
13-page submission did not even reach the Achieved level<br />
because most <strong>of</strong> the space was taken up with a lot <strong>of</strong> detail<br />
<strong>of</strong> algorithm execution without discussing the bigger picture<br />
such as the performance <strong>of</strong> the algorithm. One student<br />
was awarded Excellence for just 6.5 pages <strong>of</strong> writing. In<br />
this case the student didn’t use any images, but had clear<br />
explanations and examples which showed un<strong>der</strong>standing <strong>of</strong><br />
the topic; for example, binary and linear search were compared<br />
for large values <strong>of</strong> n, providing a convincing contrast.<br />
14<br />
Two <strong>of</strong> the longest Excellence submissions used a lot <strong>of</strong> images,<br />
including several screen shots <strong>of</strong> multiple interfaces.<br />
The screen shots were relevant, although not essential, and<br />
the work could have been presented in slightly less space.<br />
Another 14-page Excellence submission was primarily text,<br />
and contained consi<strong>der</strong>ably more detail and covered a wi<strong>der</strong><br />
range <strong>of</strong> concepts than would be necessary to meet the Excellence<br />
criteria.<br />
Some <strong>of</strong> the longer submissions had extensive traces <strong>of</strong> algorithms<br />
that went over multiple pages. Often these were<br />
presented inefficiently — for example, one trace <strong>of</strong> a sorting<br />
program listed every comparison made, one per line, which<br />
was very difficult to read, and took four pages to describe<br />
sorting eleven numbers. Other students managed to convey<br />
similar traces in a fraction <strong>of</strong> the space by using better layout<br />
which more clearly showed the big picture <strong>of</strong> what was<br />
happening as well as making each step clear. Since layout<br />
isn’t one <strong>of</strong> the criteria, both approaches should be accepted,<br />
but teachers advising students on submission lengths would<br />
need to take into account how compact a student’s representations<br />
are.<br />
Based on our observations, the limit <strong>of</strong> 14 pages seems suitable<br />
as an absolute maximum, but students could be advised<br />
that about 10 pages <strong>of</strong> their own writing is usually sufficient<br />
to cover the Excellence criteria, and longer documents would<br />
be required only if they rely heavily on images in their descriptions.<br />
It is important that students focus on the criteria<br />
<strong>of</strong> the standard and cover all the requirements.<br />
Not all students addressed all three areas. Two students<br />
did not attempt the algorithms topic; seven made submissions<br />
that didn’t touch on programming languages, and nine<br />
didn’t tackle usability. It would appear to indicate that these<br />
few students simply ran out <strong>of</strong> time to complete all the criteria,<br />
but submitted their work anyway (none <strong>of</strong> them received<br />
a passing grade since there were criteria that were clearly not<br />
met.) Most <strong>of</strong> these gave the impression <strong>of</strong> a hasty attempt,<br />
but one student did a good job <strong>of</strong> two <strong>of</strong> the topics, and<br />
would have done well if they had submitted something on<br />
usability.<br />
4.2 Algorithms<br />
For the algorithms topic, the main criteria were to differentiate<br />
an algorithm from a program and informal instructions,<br />
and to discuss the “cost” <strong>of</strong> an algorithm. Almost any example<br />
could be used to discuss the first criterion, but measuring<br />
the cost required an algorithm that had interesting behavior<br />
based on its input size n. This criterion requires a good<br />
definition for the cost; generally it would be the running<br />
time <strong>of</strong> a program or the number <strong>of</strong> steps taken by the algorithm.<br />
In some cases memory requirements or disk accesses<br />
might be consi<strong>der</strong>ed as a cost, although this wasn’t used<br />
in any <strong>of</strong> the reports analysed here. Unfortunately some<br />
students misun<strong>der</strong>stood the cost to be the length <strong>of</strong> the program<br />
or algorithm, which painted them into a corner since<br />
this couldn’t be measured based on the size <strong>of</strong> the input, n.<br />
About 63% <strong>of</strong> the students used sorting algorithms as their<br />
example, and 19% used searching. Many students met the<br />
Achieved criteria or better for the algorithms part using<br />
these examples. About 9% <strong>of</strong> the students appeared to use
either their own program or another algorithm as an example,<br />
and these were generally not very convincing for meeting<br />
the criteria (most <strong>of</strong> them received “Not achieved” for<br />
the standard). An example used by one student that was<br />
(just) suitable for meeting the basic criteria was the Towers<br />
<strong>of</strong> Hanoi solution, although it wasn’t feasible to compare<br />
two different algorithms with different costs with this example,<br />
and hence couldn’t be used for the Excellence criterion.<br />
At least five students attempted to discuss the cost <strong>of</strong> an<br />
algorithm without using an example at all — none <strong>of</strong> their<br />
submissions met the criteria.<br />
The most popular sorting algorithms were bubble sort and<br />
quicksort (each was used in about 31% <strong>of</strong> the submissions).<br />
The high use <strong>of</strong> bubble sort is likely to reflect its popularity<br />
with teachers more than its pedagogical value [1]. Quicksort<br />
provides a strongly contrasting example to compare with<br />
quadratic time sorting algorithms. The other popular sorting<br />
algorithms were selection sort (used in about 22% <strong>of</strong><br />
submissions) and insertion sort (20%).<br />
To meet the Excellence criteria, students needed to choose<br />
two algorithms to compare, and this was attempted by about<br />
36% <strong>of</strong> the students. About 26% <strong>of</strong> them compared the<br />
performance <strong>of</strong> two different sorting algorithms, with about<br />
16% comparing an O(n log n) algorithm with an O(n 2 ) one,<br />
giving the opportunity to observe a strong contrast in performance.<br />
About 10% compared two O(n 2 ) sorting algorithms<br />
(mainly two <strong>of</strong> bubble sort, selection sort and insertion sort),<br />
which didn’t illustrate the contrasting performance differences<br />
that can occur between different algorithms for the<br />
same problem, although there were constant factor differences<br />
observed. About 11% <strong>of</strong> the students used two different<br />
searching algorithms for the comparison — all <strong>of</strong> these<br />
compared linear search with binary search, which provide a<br />
strong contrast in cost. One Excellence student also used<br />
hashing (based on the CS Unplugged searching activity),<br />
which led to a worthwhile discussion.<br />
The values <strong>of</strong> n that were chosen for comparing algorithms<br />
were <strong>of</strong>ten too small to make good observations. It wasn’t<br />
unusual for student to report performance for searching or<br />
sorting up to just 10 or 20 items, but it is at 100 or even 1000<br />
items that some algorithms start to show compelling performance<br />
differences. Very few students explicitly observed<br />
that the relationship between two algorithms was non-linear,<br />
although a few <strong>of</strong> the students who evaluated searching algorithms<br />
did recognise the significant difference between their<br />
measurements which reflected O(n) compared with O(log n)<br />
time — these were mainly students who scored Excellence<br />
overall. Note that students don’t have to implement binary<br />
search (which is notoriously hard to get right [12]), but can<br />
run a provided program that performs the search and counts<br />
comparisons, or use a visualisation (several are available online)<br />
that reports the time or number <strong>of</strong> comparisons made.<br />
Similarly, they don’t have to do a formal analysis <strong>of</strong> the cost,<br />
but could either plot data points and talk about trends, or<br />
reason about it based on repeatedly halving n.<br />
In general, students took a simplistic approach to analysing<br />
their algorithms, and they need to be made aware <strong>of</strong> the<br />
idea that algorithmic complexity can be non-linear, which<br />
is a crucial factor surrounding the scalability <strong>of</strong> algorithms<br />
15<br />
(for example, a quadratic sorting algorithm given 10 times<br />
as much input takes about 100 times as long to run). Some<br />
weren’t able to observe the difference because the algorithms<br />
they chose had the same complexity, and thus only showed<br />
a small constant difference; others didn’t notice significant<br />
differences because they only used small values <strong>of</strong> n. Only<br />
three <strong>of</strong> the submissions analysed used a graph to show the<br />
algorithm performance; others used tables, or simply presented<br />
results in an uncollated fashion. The lack <strong>of</strong> graphs<br />
would inevitably make it hard to see trends, although some<br />
students were able to describe them from tables <strong>of</strong> figures.<br />
Since the significant differences are likely to be unexpected,<br />
students need to be guided to look for them, for example,<br />
by encouraging them to experiment with very large values<br />
<strong>of</strong> n, and to chart the results through a spreadsheet, which<br />
makes it easier to deal with a larger number <strong>of</strong> observations.<br />
The choice <strong>of</strong> algorithms can clearly make a difference to<br />
the richness <strong>of</strong> the student report. The most effective choices<br />
were the pairs <strong>of</strong> algorithms with significantly different asymptotic<br />
complexities (binary search vs linear search, and quicksort<br />
vs insertion/selection/bubble sort). Both the binary/linear<br />
search and quicksort/quadratic sort pairs have costs that differ<br />
by a factor <strong>of</strong> n/ log n, which provides a obvious contrast<br />
for large values <strong>of</strong> n.<br />
Students don’t need to implement the algorithms to meet the<br />
criteria, and in fact the techniques needed (e.g. arrays/lists)<br />
are beyond the requirements <strong>of</strong> the programming standards<br />
at this level. For students who have more programming experience,<br />
the simpler algorithms (linear search and quadratic<br />
sorts) might serve as good programming exercises but the<br />
difficulty is that a student must get the programming exercise<br />
completely correct in or<strong>der</strong> to be able to use them<br />
for valid experiments (for example, one student gave results<br />
that showed Bubble sort to be 73 times faster than Insertion<br />
sort, which almost certainly reflects an issue with the<br />
implementation given that Bubble sort is usually slower than<br />
Insertion Sort). While quicksort is har<strong>der</strong> to implement, a<br />
list-based version <strong>of</strong> it (where the list is duplicated at each<br />
partition) is relatively easy to implement, and would still<br />
demonstrate the performance improvement required. Binary<br />
search is also easy to un<strong>der</strong>stand but difficult to get<br />
exactly right [12]. Interestingly, some students explained in<br />
their reports about why binary search is so hard to implement;<br />
it is good to see them sensitised to issues surrounding<br />
program correctness. In general we would recommend that<br />
students experiment with programs they are provided with<br />
(or visualisations), but interested students could be encouraged<br />
to implement their own version as an extra exercise to<br />
deepen their un<strong>der</strong>standing, but not use those versions to<br />
measure timing.<br />
The use <strong>of</strong> bubble sort tended to be a distraction, as it is<br />
very closely related to other quadratic sorting algorithms,<br />
involves more steps for students to trace, and has a tradition<br />
in algorithm pedagogy that paradoxically has made one <strong>of</strong><br />
the worst sorting algorithms one <strong>of</strong> the most well known [1].<br />
As a contrasting algorithm to quicksort, selection sort is<br />
useful because its performance is easily analysed using high<br />
school math, and insertion sort is useful because it has a<br />
contrasting best/average/worst case.
One complication with the algorithms topics is that a closely<br />
related standard (AS91075) on designing a computer program<br />
used the term “algorithmic structure” to describe the<br />
design <strong>of</strong> the program. This appears to have been confused<br />
with the use <strong>of</strong> “algorithm” in the AS91074 standard; some<br />
online postings from teachers indicate that they regarded<br />
them as describing the same thing, and at least one student<br />
stated that an algorithm is a program plan, which is essentially<br />
correct, but confusing in this context. Thus some<br />
submissions on algorithm “cost” are based on a student’s<br />
own design <strong>of</strong> a program, which typically doesn’t have any<br />
interesting behaviour to analyse, and in fact may be O(1)<br />
since the program does a very routine task. For example,<br />
several programs simply read in a value, did a calculation<br />
and printed the result. The use <strong>of</strong> the word “algorithm” has<br />
since been removed from the AS91075 standard to avoid this<br />
confusion.<br />
4.3 Programming languages<br />
The criteria for programming languages centre around the<br />
role <strong>of</strong> programming languages, high and low level languages,<br />
and how high level languages are translated to low level ones<br />
(i.e. compiling and interpreting).<br />
Most <strong>of</strong> the students described high and low level languages<br />
and compilers abstractly, but at most a third attempted<br />
some sort <strong>of</strong> concrete demonstration <strong>of</strong> a compiler being<br />
used. A demonstration is much more convincing <strong>of</strong> student<br />
un<strong>der</strong>standing because the student has personally had the<br />
experience <strong>of</strong> converting a program to a low level language<br />
and running it. Even better than demonstrating the process<br />
on a given program, students could have used their own<br />
program (for example, one done for AS91076) as the example,<br />
showing how it is compiled and/or interpreted. Because<br />
the standard discusses both interpreted and compiled programs,<br />
it’s unlikely that students would be familiar with<br />
two different languages and thus only one <strong>of</strong> the examples<br />
could reasonably be personalised, although some languages<br />
have both compiled and interpreted versions (e.g. Basic)<br />
and these could be contrasted. Another approach would be<br />
for students to learn the bare minimum <strong>of</strong> a language e.g.<br />
write a “Hello world” level program but substituting their<br />
own name. None <strong>of</strong> the submissions seemed to have taken<br />
this approach, and it would be interesting to evaluate, since<br />
it forces students through the process <strong>of</strong> getting the program<br />
to run. Using the student’s own program was risky for the<br />
algorithms topic, but for the programming languages topic it<br />
provides good personalisation and shows an awareness <strong>of</strong> the<br />
process, and it isn’t sensitive to the quality <strong>of</strong> the program<br />
(as long as it runs or compiles). Despite this, we observed<br />
few examples in our sample where students had obviously<br />
used their own program.<br />
In the student work analysed, the main languages mentioned<br />
as examples were Visual Basic/Basic (in about 18%<br />
<strong>of</strong> reports), Scratch (13%), Pascal (5%), Flash/Actionscript<br />
(4%), C (4%) and Alice (3%). Other languages used included<br />
C++, Java, JavaScript, Picaxe, and Python. About<br />
a half the reports didn’t use an example in a specific language,<br />
and these generally didn’t reflect a lot <strong>of</strong> un<strong>der</strong>standing<br />
since the concepts were explained in very abstract terms.<br />
Projects that compared a compiled and interpreted language<br />
16<br />
mainly used Visual Basic, C, Pascal and Java as examples<br />
<strong>of</strong> compiled languages, and Scratch, VBA in Excel, Alice,<br />
JavaScript and Python as examples <strong>of</strong> interpreted languages.<br />
The situation isn’t always simple, as mo<strong>der</strong>n languages can<br />
blur the distinction <strong>of</strong> compiling or interpreting (e.g. Java<br />
compiles to byte code, which is then interpreted). One student<br />
mentioned how Scratch can run as a Java applet itself,<br />
so it is interpreted twice and compiled! If students can fully<br />
un<strong>der</strong>stand these complexities then they will have a good<br />
grasp on the issues intended, but it will be important for<br />
teachers to provide guidance on the languages to explore.<br />
Another example used in several reports was Visual Basic<br />
(compiled) compared with Basic (VBA) macros in Excel (interpreted).<br />
The contrast is effective because it is essentially<br />
the same language and focuses the student on the different<br />
way it is run rather than the difference in language; the examples<br />
we observed using this contrast all appeared to meet<br />
the criteria for Excellence in the programming languages requirements.<br />
We would note that JavaScript is a useful language<br />
to explore because it is definitely interpreted, readily<br />
available, and is not confused by the drag-and-drop aspect<br />
<strong>of</strong> Scratch, whereas it is problematic to use HTML as an example<br />
<strong>of</strong> an interpreted language (as one project did) since<br />
it is not at all clear that HTML is usefully viewed as a programming<br />
language, even if it is a formal markup language.<br />
Some students presented their work on programming language<br />
as posters. This approach has potential since the<br />
student is writing for a particular audience (probably other<br />
students), and thus might be expected to focus the explanations<br />
and examples more. It also seemed to discourage<br />
paraphrasing, perhaps because they were thinking <strong>of</strong> it as<br />
communicating to peers rather than trying to write something<br />
that a teacher would recognise as correct.<br />
4.4 Usability<br />
The criteria in the standard emphasise factors <strong>of</strong> a user interface<br />
that contribute to its usability. Unfortunately many<br />
reports seemed to confuse usability with functionality, or focussed<br />
on the physical layout <strong>of</strong> the interface. For example,<br />
a student might comment that a cellphone can take photos<br />
(the functionality), but miss the point that the typical task<br />
that a user has (such as taking a photo and transferring it<br />
to an online photo album) might involve pressing mysterious<br />
key sequences, dealing with incompatible devices, and using<br />
a time-consuming procedure that is difficult to remember<br />
(usability). The better work focussed on typical tasks done<br />
by the user, and any problems encountered completing those<br />
tasks.<br />
Using a personally owned device and talking about the student’s<br />
own experience gives a clearer demonstration <strong>of</strong> un<strong>der</strong>standing<br />
than writing about interfaces abstractly. About<br />
61% <strong>of</strong> the reports provided a suitably personalised report.<br />
Those that didn’t discuss a personal device mainly gave general<br />
usability principles that appeared to be paraphrased<br />
from standard sources. In these cases, where illustrations<br />
were given they tended to be standard examples <strong>of</strong> usability<br />
issues, and once again it is hard to hear the “student voice”<br />
when this approach is used.<br />
The students who reported on experience with their own dig-
ital system used a variety <strong>of</strong> interfaces as examples: about<br />
29% <strong>of</strong> the reports evaluated some sort <strong>of</strong> mobile phone (simple<br />
phones, smart phones and iPhone related devices) or<br />
some feature <strong>of</strong> one (such as sending a text message or taking<br />
a photo). The main other commonly chosen devices were<br />
an iPod/Touch device, used in about 7% <strong>of</strong> the reports, and<br />
a video game controller (5%). Other interfaces appeared less<br />
frequently, including a calculator, CD player, DVD player,<br />
DVD recor<strong>der</strong>, email s<strong>of</strong>tware, games, web browsers, web<br />
sites, an mp3 player and a tape player. Even a toaster was<br />
used to provide a reasonable evaluation.<br />
Ten students reported on an interface that they had written<br />
themselves. These generally weren’t good as examples:<br />
few <strong>of</strong> them appeared to meet the criteria, and the remain<strong>der</strong><br />
were focussed on describing their interface without discussing<br />
it from the user’s point <strong>of</strong> view. It would appear that<br />
these students had tried to combine this standard with work<br />
done for the design and programming standards. For user<br />
interfaces this makes it difficult for the student to provide<br />
an objective and critical evaluation.<br />
An important step in usability assessment is to identify the<br />
tasks that the device is to be used for, which helps to take<br />
the focus away from features and layouts, and put it onto the<br />
sequence <strong>of</strong> operations that a user would perform to achieve<br />
their goals. It is not unusual for a device to be frustrating to<br />
use because in practical situations the functionality <strong>of</strong> the<br />
device can be difficult or slow to access, and this is revealed<br />
if a task is evaluated instead <strong>of</strong> just a feature <strong>of</strong> the device.<br />
Only about 36% <strong>of</strong> the submitted reports described the task<br />
that a user might do.<br />
Another useful approach that can pick up usability issues<br />
is observing another person using the interface. Only 9 students<br />
did this, and all these projects clearly met the criteria;<br />
a third-person observation would apparently make it easier<br />
to report on usability, although the sample is too small to<br />
draw firm conclusions. Some <strong>of</strong> the students who observed<br />
another person took notes <strong>of</strong> every step the person took.<br />
This approach was particularly convincing when the student<br />
described their observations, and it set the student up well<br />
to meet the Merit criterion for usability, and Excellence if<br />
they compared another device.<br />
In general it seemed that the students had little teaching<br />
on HCI and usability evaluation. Although not required at<br />
this level, introducing usability heuristics (e.g. from useit.<br />
com) provides a tool for students to analyse systems. Another<br />
approach is the cognitive walkthrough, which is described<br />
for high school students at www.cs4fn.org/usability/<br />
cogwalkthrough.php. This encourages students to observe<br />
someone else using the system, which makes the evaluator<br />
more aware <strong>of</strong> the steps being made.<br />
4.5 General writing issues<br />
A key phrase in the title and wording <strong>of</strong> the standard is for<br />
students to “demonstrate un<strong>der</strong>standing” in or<strong>der</strong> to meet<br />
the criteria. It is tempting to respond to criteria such as<br />
“explaining the need for programs to translate between high<br />
and low level languages” with abstract textual explanations<br />
that are paraphrased from books or online sources (and in<br />
fact the instructions for the report explicitly say that para-<br />
17<br />
phrasing is acceptable). Although the general instructions<br />
allow paraphrasing, a more compelling way to show un<strong>der</strong>standing<br />
would be for a student to provide a worked example<br />
using a context unique to themselves (for example, analysing<br />
the interface <strong>of</strong> their own mobile phone, or running experiments<br />
on the speed <strong>of</strong> an algorithm on their own computer).<br />
In many cases we found that “showing an un<strong>der</strong>standing”<br />
was interpreted as simply explaining the concept. In this<br />
case the process could end up grading Wikipedia’s knowledge,<br />
and any slips in the paraphrasing process only served<br />
to show that the student didn’t un<strong>der</strong>stand the concept!<br />
A common problem in the student reports is giving poor<br />
explanations <strong>of</strong> what they had done. For example, in explaining<br />
sorting, one student wrote that B is bigger than<br />
A, and A is bigger than C, but we weren’t told what A, B<br />
and C were — quite possibly they were sorting weights or<br />
labeled envelopes, but this wasn’t mentioned at all. Another<br />
problem was assuming that the marker would know things<br />
that the teacher did (perhaps because students are used to<br />
internal assessment); for example, referring to the “method<br />
that was used in class” rather than describing it.<br />
Many reports showed a lack <strong>of</strong> scientifically convincing reporting,<br />
such as drawing conclusions from just one speed<br />
measurement from each <strong>of</strong> two algorithms, or having results<br />
spread around the document and not explicitly comparing<br />
them. It was also not unusual for the student to assume that<br />
the rea<strong>der</strong> would go to the trouble <strong>of</strong> finding patterns in the<br />
results for themselves; for example, one student had a table<br />
<strong>of</strong> execution times that showed a clear difference between<br />
selection sort, bubble sort and quicksort, but they never articulated<br />
that they had noticed the difference, and thus got<br />
credit for measuring the algorithms, but not for showing an<br />
un<strong>der</strong>standing <strong>of</strong> the different costs.<br />
The work on the programming languages topic overwhelmingly<br />
used paraphrased quotes. This is largely because it is<br />
a lot har<strong>der</strong> to personalise (there are only so many compilers<br />
and machine languages that students are likely to encounter,<br />
and the role <strong>of</strong> interpreters and compilers is fairly well defined).<br />
A similar situation arises with the requirement to<br />
discuss the relationship between algorithms and programs;<br />
there are only a few ways to describe this accurately. In<br />
these cases the criteria aren’t ideal for a report and it would<br />
be better if they prompted examples rather than factual information.<br />
4.6 Guidance for students<br />
The nature <strong>of</strong> the standards is such that teachers can direct<br />
student work in a variety <strong>of</strong> ways to enable them to demonstrate<br />
their un<strong>der</strong>standing. Many <strong>of</strong> the projects reproduced<br />
questions and rubrics from teachers, and the nature <strong>of</strong> these<br />
instructions had an influence on how easy it would be for<br />
students to produce work that met the criteria.<br />
Several sets <strong>of</strong> the sampled reports had identical questions<br />
heading each section, which either would have come from<br />
the same teacher, or a group <strong>of</strong> teachers who had shared<br />
teaching ideas.<br />
Because the quality <strong>of</strong> student work depended a lot on the<br />
tasks that were set for them, we have attempted to identify
groups <strong>of</strong> submissions that seem to have been set the same<br />
tasks. Because teacher peer support means that teaching<br />
material is shared, this does not necessarily mean that the<br />
similar papers have the same teachers or are even at the<br />
same school, although inevitably those who are in the same<br />
class probably were given the same assignment.<br />
For example, in one set <strong>of</strong> reports which had the same questions,<br />
several students compared Insertion sort with Selection<br />
sort, which will have a constant-factor difference. Many<br />
<strong>of</strong> the students correctly concluded that Insertion sort is<br />
twice as fast as Selection sort, but missed the opportunity<br />
to explore the possibility <strong>of</strong> a difference that was more than<br />
a constant factor.<br />
Some <strong>of</strong> the questions were presumably intended to stimulate<br />
discussion, but seemed to be treated by students as an<br />
exam, resulting in very short answers. For example, a series<br />
<strong>of</strong> questions about choices <strong>of</strong> programming languages in<br />
one group resulted in very brief answers <strong>of</strong> just a few words.<br />
Although the questions on programming languages may not<br />
have had the intended effect, the questions on usability for<br />
the same group resulted in generally good responses from<br />
the students. For this the rubric prompted the students to<br />
choose someone who isn’t a digital native (e.g. parent or<br />
grandparent) and observe them using a device for about an<br />
hour.<br />
One group that did particularly well had prompting questions<br />
that didn’t just mirror the criteria, but started <strong>of</strong>f with<br />
a creative situation for the students to get them thinking<br />
beyond just checking <strong>of</strong>f the criteria. For algorithms, the<br />
students were prompted to get the names <strong>of</strong> favourite songs<br />
from five friends, and then explain searching algorithms using<br />
that list. This provided a familiar context (searching<br />
for songs on an mp3 player), and a meaningful task (selecting<br />
a favourite song). It also ensured unique lists for each<br />
student, even though they were doing the same task. The<br />
usability section for this group began by asking students to<br />
describe the context the device is used in, and the role <strong>of</strong> the<br />
interface. Once again, this prompt doesn’t directly address<br />
the criteria, but it starts the students thinking <strong>of</strong> the big<br />
picture, and resulted in good quality answers.<br />
Now that the standard has been in use for a year, teachers<br />
will be able to share best practices based on how well<br />
students were able to respond to the various prompts.<br />
5. CONCLUSION<br />
The analysis above has provided some specific ideas on how<br />
to teach and assess the three topics in the new standard. The<br />
following general observations are based on the analysis <strong>of</strong><br />
student work.<br />
Teacher pr<strong>of</strong>essional development and support is essential:<br />
Because the new standard was introduced very<br />
quickly, and because computer science is a new subject in<br />
New Zealand schools, teachers could not be expected to be<br />
well prepared to teach the new material, either in terms <strong>of</strong><br />
subject knowledge or pedagogical knowledge. Unfortunately<br />
this was reflected in the student work, particularly in the<br />
guidance <strong>of</strong> topics they were given for their investigations<br />
(e.g. comparing two algorithms that didn’t have contrast-<br />
18<br />
ing performance, or investigating only small values <strong>of</strong> n).<br />
Teachers need to be guided on what the main messages are<br />
that students should take away from this experience, and<br />
be given recommendations <strong>of</strong> examples that illustrate these<br />
in a compelling way — in this case, contrasting algorithms,<br />
contrasting programming languages, and interfaces that are<br />
small enough to evaluate but have interesting usability issues.<br />
Students do better when they personalize their work:<br />
When students report on their own experiences, whether it<br />
is running an algorithm on their own computer or evaluating<br />
a device they have chosen, their work carries a lot more<br />
authenticity than abstract discussions <strong>of</strong> a concept. It also<br />
moves their level <strong>of</strong> un<strong>der</strong>standing up Bloom’s taxonomy:<br />
descriptions <strong>of</strong> a concept <strong>of</strong>ten come across as a paraphrase<br />
that at best reflects remembering facts, while discussions<br />
based on a unique example reflect the student applying their<br />
knowledge.<br />
Assessment using reports: The AS91074 computer science<br />
standard is assessed by a student report submission.<br />
For most <strong>of</strong> the criteria this provides an opportunity for students<br />
to report on a rich personal experience, such as experiments<br />
with algorithms and evaluation <strong>of</strong> interface usability<br />
with real users. Through project work they get to experience<br />
the issues first-hand (such as a slow algorithm or a frustrated<br />
user), and get experience communicating in writing about<br />
computer science. However, some <strong>of</strong> the criteria (e.g. “describing...<br />
the function <strong>of</strong> a compiler”) invite responses that<br />
are just paraphrases <strong>of</strong> standard definitions, and would benefit<br />
from clarification or re-wording (e.g. “demonstrate the<br />
function <strong>of</strong> a compiler using an example”) to encourage students<br />
to explore the concept in a context rather than simply<br />
describe it.<br />
Using provided systems rather than having students<br />
build their own: The best student work was done when<br />
they assessed the performance <strong>of</strong> a provided implementation<br />
<strong>of</strong> an algorithm; students who implemented their own programs<br />
tended to be distracted by the implementation, were<br />
limited in the algorithms they could use, had potentially unreliable<br />
performance results, and were biased in their usability<br />
assessments. Because the standard is about algorithm<br />
performance and not programming, those who focussed on<br />
their own program were at risk <strong>of</strong> missing the big picture,<br />
or else might implement algorithms that didn’t show a good<br />
contrast. Likewise, students who discussed usability based<br />
on interfaces they had implemented couldn’t see their weaknesses.<br />
More advanced students might be expected to implement<br />
their own systems and evaluate them, but at the<br />
beginning high school level the messages about the key topics<br />
are likely to come across more clearly if students start<br />
with a working system.<br />
The introduction <strong>of</strong> computer science to NZ schools has addressed<br />
a need that was articulated by industry and teachers,<br />
and has happened remarkably quickly. The work done<br />
by students shows some pleasing levels <strong>of</strong> un<strong>der</strong>standing <strong>of</strong><br />
computer science for some; others have at least had experiences<br />
that indicate that they have had a taste <strong>of</strong> what kind<br />
<strong>of</strong> topics might come up in computer science courses. With<br />
very little lead-in time hundreds <strong>of</strong> students have studied
computer science, and with the lessons learned in the first<br />
year, the quantity and quality <strong>of</strong> students is only likely to<br />
increase in the future.<br />
6. ACKNOWLEDGEMENTS<br />
We are grateful to Scott Telfer (NZQA) for assisting with<br />
access to key data.<br />
7. REFERENCES<br />
[1] O. Astrachan. Bubble sort: An archaeological<br />
algorithmic analysis. ACM SIGCSE Bulletin,<br />
35(1):1–5, 2003.<br />
[2] O. Astrachan, J. Cuny, C. Stephenson, and C. Wilson.<br />
The CS10K project: mobilizing the community to<br />
transform high school computing. Proceedings <strong>of</strong> the<br />
42nd ACM Technical Symposium on Computer<br />
Science Education, SIGCSE 2011, pages 85–86, 2011.<br />
[3] T. Bell, P. Andreae, and L. Lambert. Computer<br />
Science in New Zealand High Schools. In T. Clear and<br />
J. Hamer, editors, ACE ’10: Proceedings <strong>of</strong> the 12th<br />
conference on Australasian Computing Education,<br />
volume 32 <strong>of</strong> Australian Computer Science<br />
Communications, pages 15–22, Brisbane, Australia,<br />
Jan. 2010. Australian Computer Society, Inc.<br />
[4] T. Bell, P. Andreae, and A. Robins. Computer Science<br />
in NZ High Schools: The First Year <strong>of</strong> the New<br />
Standards. In L. A. S. King, D. R. Musicant,<br />
T. Camp, and P. Tymann, editors, Proceedings <strong>of</strong> the<br />
43rd ACM technical symposium on Computer Science<br />
Education, Raleigh, NC, USA, pages 343–348, New<br />
York, 2012. ACM.<br />
[5] T. Brinda, H. Puhlmann, and C. Schulte. Bridging<br />
ICT and CS: Educational standards for computer<br />
science in lower secondary education. ACM SIGCSE<br />
Bulletin, 41(3):288–292, 2009.<br />
[6] A. Bruckman, M. Biggers, B. Ericson, T. McKlin,<br />
J. Dimond, B. DiSalvo, M. Hewner, L. Ni, and<br />
S. Yardi. “Georgia computes!”: Improving the<br />
Computing Education Pipeline. ACM SIGCSE<br />
Bulletin, 41(1):86–90, Mar. 2009.<br />
[7] T. Carrell, V. Gough-Jones, and K. Fahy. The future<br />
<strong>of</strong> Computer Science and Digital Technologies in New<br />
Zealand secondary schools: Issues <strong>of</strong> 21st teaching and<br />
learning, senior courses and suitable assessments.<br />
Technical report, 2008.<br />
[8] S. Furber, editor. Shut down or restart? The way<br />
forward for computing in UK schools. The Royal<br />
Society, London, 2012.<br />
[9] J. Gal-Ezer and C. Stephenson. The Current State <strong>of</strong><br />
Computer Science in U.S. High Schools : A Report<br />
from Two National Surveys. Journal for Computing<br />
Teachers, Spring, 2009.<br />
[10] J. Goode and J. Margolis. Exploring Computer<br />
Science. ACM Transactions on Computing Education,<br />
11(2):1–16, July 2011.<br />
[11] G. Grimsey and M. Phillipps. Evaluation <strong>of</strong><br />
technology achievement standards for use in New<br />
Zealand secondary school computing education.<br />
Technical report, New Zealand Computer Society<br />
(NZCS), Wellington, 2008.<br />
[12] R. E. Pattis. Textbook errors in binary searching.<br />
ACM SIGCSE Bulletin, 20(1):190–194, Feb. 1988.<br />
19<br />
[13] A. Robins. Learning edge momentum: A new account<br />
<strong>of</strong> outcomes. Computer Science Education, 20:37 – 71,<br />
2010.
Is Self-Efficacy in Programming Decreasing with the Level<br />
<strong>of</strong> Programming Skills?<br />
Michail N. Giannakos*<br />
Norwegian <strong>University</strong> <strong>of</strong> Science and<br />
Technology (NTNU)<br />
Dep. <strong>of</strong> Computer & Information Science<br />
Trondheim, NO-7491, Norway<br />
michail.giannakos@idi.ntnu.no<br />
ABSTRACT<br />
In this study, variables from the Unified Theory <strong>of</strong> Acceptance<br />
and Use <strong>of</strong> Technology and Social Cognitive Theory were chosen<br />
as important factors in students’ behavior and attitude towards<br />
Computer Science Education (CSE). This hybrid framework aims<br />
to measure the level <strong>of</strong> the selected key variables on CSE and<br />
identify potential differences among our different groups. The<br />
three different groups are consisted <strong>of</strong> students from: (1) programming<br />
courses <strong>of</strong> Greek Lyceums, (2) computer science<br />
courses at German Gymnasiums and (3) at the Department <strong>of</strong><br />
Informatics (freshmen) at the Ionian <strong>University</strong>, Greece. We<br />
asked the three different groups <strong>of</strong> students to complete a questionnaire<br />
that was <strong>der</strong>ived from this combination <strong>of</strong> the prior<br />
theories. The results revealed several differences in the measured<br />
variables, e.g. the freshmen who should have the highest programming<br />
competencies compared to the other two groups expressed<br />
the lowest degree <strong>of</strong> self-efficacy. The overall outcomes<br />
are expected to contribute to the un<strong>der</strong>standing <strong>of</strong> students’ likelihood<br />
to pursue computing related careers and promote the acceptance<br />
<strong>of</strong> CSE.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computer and Information Science Education]: Computer<br />
Science Education, Curriculum.<br />
General Terms<br />
Measurement, Experimentation, Human Factors.<br />
Keywords<br />
ICT courses, Programming courses, Informatics, Secondary education,<br />
Students’ beliefs, cross-cultural.<br />
1. INTRODUCTION<br />
As the success <strong>of</strong> Computer Science Education (CSE) might<br />
depend on students’ perceptions, attitude and beliefs, we aim to<br />
identify the differences in regard <strong>of</strong> these factors between students<br />
<strong>of</strong> a) secondary schools in <strong>Germany</strong>, b) secondary schools in<br />
Greece and c) freshmen <strong>of</strong> CS at a university in Greece. Based on<br />
existing theories as well as on prior work [11, 17], we have chosen<br />
variables related to students’ attitude and applied them to<br />
students from two different countries that attended three different<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that<br />
copies bear this notice and the full citation on the first page. To copy<br />
otherwise, or republish, to post on servers or to redistribute to lists,<br />
requires prior specific permission and/or a fee.<br />
Conference’10, Month 1–2, 2010, City, State, Country.<br />
Copyright 2010 ACM 1-58113-000-0/00/0010…$10.00.<br />
Peter Hubwieser<br />
Technische Universität München<br />
Fakultät für Informatik<br />
Boltzmannstr. 3. D-85478 Garching<br />
+49 89 289 17350<br />
Peter.Hubwieser@tum.de<br />
20<br />
Alexan<strong>der</strong> Ruf<br />
Technische Universität München<br />
Fakultät für Informatik<br />
Boltzmannstr. 3. D-85478 Garching<br />
+49 89 289 17350<br />
rufa@in.tum.de<br />
programming courses in secondary school and university, respectively.<br />
The goal <strong>of</strong> this investigation was to measure students’<br />
beliefs and to identify potential differences among different educational<br />
contexts. As students’ beliefs and attitude are highly<br />
correlated with their performance, students’ perceptions have an<br />
impact on what they will learn as well as how they plan to continue<br />
their course <strong>of</strong> study [15].<br />
The study presented in this paper was conducted on a large sample<br />
in Greece and <strong>Germany</strong> and seeks to provide insights for the<br />
following research aspects:<br />
a) To explore students’ perceptions regarding CS,<br />
b) To investigate potential differences in perceptions regarding<br />
CS amongst German and Greek upper secondary education<br />
students,<br />
c) To investigate potential differences in perceptions regarding<br />
CS amongst the ending <strong>of</strong> secondary education and the beginning<br />
<strong>of</strong> higher computing education students.<br />
The study attempts to fill the gaps in the CSE literature regarding<br />
students' perceptions in different countries and in different educational<br />
levels. Based on the empirical results we could further our<br />
un<strong>der</strong>standing <strong>of</strong> students CS perception in the transition from<br />
upper secondary education to higher computing education. The<br />
study is significant in that it focuses on the upper secondary education<br />
level, is situated in real conditions context collecting empirical<br />
data, while at the same time addressing not students' career<br />
choices with regard to CS, but their perceptions on several aspects<br />
regarding CS. The outcomes <strong>of</strong> the study could further our un<strong>der</strong>standing<br />
<strong>of</strong> students participation in the field <strong>of</strong> computing and<br />
could provide useful insight into the actions that should be devised<br />
within the educational system with a view to encouraging<br />
this participation.<br />
2. THEORETICAL BACKGROUND AND<br />
RELATED WORK<br />
Students’ perceptions and intentions are consi<strong>der</strong>ed as important<br />
determinants <strong>of</strong> the learning success. Reluctance towards adoption<br />
<strong>of</strong> informatics courses implies that research is needed to un<strong>der</strong>stand,<br />
more comprehensively, how students can be engen<strong>der</strong>ed.<br />
Although past research [3, 5] has empirically explained several<br />
issues regarding students perceptions and beliefs regards CSE, it<br />
is mostly focused on higher education and more specifically on<br />
CS departments.<br />
Several models and theories have been applied to address issues<br />
<strong>of</strong> students’ attitude, perceptions and to identify the cause and the<br />
* This work was carried out during the tenure <strong>of</strong> an ERCIM "Alain Bensoussan"<br />
Fellowship programme. The research leading to these results has<br />
received funding from the European Union Seventh Framework Programme<br />
(FP7/2007-2013) un<strong>der</strong> grant agreement no 246016.
effect <strong>of</strong> different factors on the adoption <strong>of</strong> science education [6].<br />
For instance, the Unified Theory <strong>of</strong> Acceptance and Use <strong>of</strong> Technology<br />
[18] and Social Cognitive Theory [2] are some <strong>of</strong> the most<br />
widely applied theories in the context <strong>of</strong> students’ behavior [6]. In<br />
addition, Confidence, Performance Expectancy, Social Influence,<br />
Satisfaction, Self-Efficacy and Perceive Behavioral Control are<br />
some <strong>of</strong> the most commonly used factors (i.e., [5, 6]) affecting<br />
students’ intention to attend a respective course.<br />
3. THE EDUCATIONAL CONTEXT<br />
The empirical study was conducted in three Greek Lyceums (3 rd<br />
grade), in classes <strong>of</strong> the 11 th and 12 th grade at two Bavarian (German)<br />
Gymnasiums and in the Department <strong>of</strong> Informatics at Ionian<br />
<strong>University</strong> in Corfu, Greece (1 st year). As there might be a substantial<br />
influence <strong>of</strong> the specific educational context on our results,<br />
we start with a short description <strong>of</strong> the school systems and<br />
the differences <strong>of</strong> the CS education policies in the two regarded<br />
countries. A very detailed description was presented recently by a<br />
Working Group named: Computer Science/Informatics in Secondary<br />
Education on the ITiCSE 2011 [13].<br />
3.1 The Greek Lyceum<br />
The 1 st grade <strong>of</strong> Greek Lyceum (Lykio), represents an orientation<br />
year with a general education program. The 2 nd and 3 rd grades<br />
<strong>of</strong>fer three curricular directions: Theoretical, Scientific and Technological.<br />
Students who follow the technological direction are<br />
taking a specific course named Applications Development in a<br />
Programming Environment (ADPE) that involves the development<br />
<strong>of</strong> algorithms and programming. This course has been taught<br />
for ten years. It focuses on the algorithmic approach and on the<br />
development <strong>of</strong> problem-solving skills in a programming environment.<br />
This subject is assigned to CS teachers.<br />
The overall aim <strong>of</strong> 3 rd Lyceum programming courses is to develop<br />
analytical and synthetic thinking, acquire methodological skills<br />
and be able to solve simple problems within a programming environment.<br />
This programming course has not been designed to<br />
educate programmers, and for this reason it is not designed to<br />
teach sophisticated programming techniques; it focuses on approaches<br />
and techniques <strong>of</strong> problem solving with emphasis on<br />
structured thinking. Many basic algorithmic and programming<br />
concepts, such as conditions, expressions and logical reasoning,<br />
are fundamentals <strong>of</strong> general knowledge and skills to be acquired<br />
in general education.<br />
The curriculum states that this subject must be taught (at least<br />
partially) in a computer lab. The Ministry <strong>of</strong> Education has certified<br />
specific Educational S<strong>of</strong>tware to support the lab work, especially<br />
for the Lyceum programming course. The Educational<br />
S<strong>of</strong>tware has been designed to support teaching, to complement<br />
the subject's needs and IT use and to help students to consolidate<br />
the material. The certified s<strong>of</strong>tware includes an activity space, a<br />
flow chart developer and a programming environment in accordance<br />
with the textbook.<br />
3.2 The Bavarian Gymnasium<br />
The organization <strong>of</strong> the schools in <strong>Germany</strong> is different for each<br />
<strong>of</strong> the 16 states. In the State <strong>of</strong> Bavaria, in which the survey was<br />
conducted, secondary school starts with the 5 th grade. There are<br />
three different types <strong>of</strong> secondary schools, which differ in their<br />
orientation and their level <strong>of</strong> difficulty. The type <strong>of</strong> secondary<br />
school that directly qualifies for the enrollment at universities is<br />
called Gymnasium, where students spend a total <strong>of</strong> 8 years. Currently<br />
about one third <strong>of</strong> all students in grade 8 attends the Gymnasium<br />
[4].<br />
21<br />
CS is a compulsory subject for all students <strong>of</strong> 6 th and 7 th grade at<br />
Gymnasium. During this time, the students learn to apply objectoriented<br />
modeling to standard s<strong>of</strong>tware applications (<strong>of</strong>fice,<br />
graphics, text processing, file management, hypertext structures,<br />
e-mail) as well as to describe an simulate simple algorithms. To<br />
this purpose, the students use commercial standard s<strong>of</strong>tware systems<br />
as well as custom-made s<strong>of</strong>tware tools and programming<br />
systems to solve simple problems applying the algorithmic control<br />
structures.<br />
Starting with 8 th grade, there are <strong>of</strong>fered four different educational<br />
directions at Gymnasium. In the scientific-technological direction<br />
CS continues to be a compulsory subject in the 9 th and 10 th grade.<br />
While programming skills are not very relevant for the 9 th grade –<br />
it deals primarily with functional modeling and relational database<br />
systems – the students in 10 th grade learn OOM and OOP (incl.<br />
inheritance). Nevertheless, the training programming skills is not<br />
the main focus, which lies on modeling competencies instead. The<br />
teachers are free to decide about the programming language and<br />
development environment the classes use, but they mostly agree<br />
to chose Java and BlueJ.<br />
In the last two grades (11 and 12) <strong>of</strong> Gymnasium, the so-called<br />
qualification stage, the students learn in courses that they can<br />
choose themselves, disregarding some subjects that are compulsory<br />
(e.g. Mathematics and German Language) and some additional<br />
regulations. To choose a CS course, the student must have selected<br />
the scientific-technological branch before. CS in the 11 th grade<br />
is based strongly on the learning outcomes <strong>of</strong> the 10 th grade – the<br />
learning content consists mainly <strong>of</strong> dynamic data structures and<br />
s<strong>of</strong>tware engineering – whereas the topics <strong>of</strong> the 12 th grade are<br />
elements <strong>of</strong> theoretical and technical CS. Unfortunately, in 2009<br />
only about 7% <strong>of</strong> all students <strong>of</strong> the qualification stage had chosen<br />
a CS course (according to the Bavarian School Administration).<br />
All German students who have completed our survey come from<br />
this group, most from 11 th grade (19 <strong>of</strong> 29).<br />
3.3 The Ionian <strong>University</strong><br />
The curriculum <strong>of</strong> the Department <strong>of</strong> Informatics <strong>of</strong> Ionian <strong>University</strong><br />
focuses on the broad area <strong>of</strong> CS/Informatics with two areas<br />
<strong>of</strong> specialization namely: (a) Humanistic Informatics and (b)<br />
Information Systems. The curriculum comprises a set <strong>of</strong> core<br />
courses and a set <strong>of</strong> elective courses through which the student<br />
specializes in one <strong>of</strong> the two aforementioned specializations. The<br />
curriculum was established early at the inception <strong>of</strong> the Department<br />
in 2004 and based on accepted international standards<br />
(ACM IS 2002). Regarding the first semester <strong>of</strong> the department<br />
which is in our interests, it has a general knowledge program (the<br />
orientations start at 5 th semester) and the following courses are<br />
being taught: 1) Introduction to Computer science, 2) Introduction<br />
to Programming, 3) Mathematical Analysis, 4) Linear Algebra<br />
and 5) Introduction to Information Society.<br />
4. THE SURVEY<br />
The questionnaire handed out to the students was divided into two<br />
parts. The first included questions on the demographics <strong>of</strong> the<br />
sample (age and gen<strong>der</strong>) and the second part included measures <strong>of</strong><br />
the various factors identified in the literature from previous researches.<br />
Table 4 in the Appendix lists the questionnaire factors,<br />
their operational definition, their items and the source from the<br />
literature review. In all cases, 7-point Likert scales were used.<br />
4.1 Sampling<br />
The survey was open in the middle <strong>of</strong> the school year 2011-2012<br />
at the Greek Lyceums and the German Gymnasiums and at the<br />
beginning <strong>of</strong> 2011-2012 year <strong>of</strong> study in the Department <strong>of</strong> In-
formatics. The final sample <strong>of</strong> respondents was comprised <strong>of</strong> 115<br />
Students. From the total <strong>of</strong> students, 55 (47.8%) attended the 3 rd<br />
<strong>of</strong> Greek Lyceum (16-17 years), 29 (25.2%) the 11 th or 12 th <strong>of</strong> a<br />
German Gymnasium (16-18 years) and 31 (27.0%) the first year<br />
<strong>of</strong> study at the Department <strong>of</strong> Informatics (17-18 years). 88 <strong>of</strong> the<br />
students were males (76.5%) and 27 (23.5%) females.<br />
4.2 Data Analysis and Results<br />
As proposed by Fornell and Larcker [9], there are three procedures<br />
to assess the convergent validity <strong>of</strong> any measure in a study:<br />
1) Composite reliability <strong>of</strong> each construct, 2) Item reliability <strong>of</strong><br />
the measure and 3) The average variance extracted (AVE).<br />
Therefore, we started with an analysis <strong>of</strong> composite reliability and<br />
dimensionality to check the validity <strong>of</strong> the scale used in the questionnaire.<br />
Concerning the reliability <strong>of</strong> the scales, Cronbach (CR)<br />
α indicator was applied [7] and inter-item correlations statistics<br />
for the items <strong>of</strong> the variable were calculated. As stated by to Fornell<br />
& Larcker [9], CR α value greater than 0.7 indicates a high<br />
reliability. Table 1 demonstrates the result <strong>of</strong> the test that revealed<br />
acceptable indices <strong>of</strong> internal consistency in all the factors.<br />
Table 1. Summary <strong>of</strong> Measurement Scales<br />
Factors Items Mean S.D. CR Loads AVE<br />
PE PE1 4.52 1.95 0.93 0.74 0.67<br />
PE2 4.16 1.75 0.85<br />
PE3 4.40 1.72 0.84<br />
PE4 4.51 1.69 0.83<br />
STF STF1 4.97 1.51 0.92 0.72 0.62<br />
STF2 4.98 1.58 0.73<br />
STF3 5.47 1.61 0.86<br />
STF4 5.23 1.68 0.82<br />
SI SN1 4.04 2.00 0.82 0.76 0.63<br />
SN2 4.12 2.04 0.82<br />
SEF SEF1 3.89 1.79 0.71 0.82 0.56<br />
SEF2 4.19 1.73 0.70<br />
SEF3 2.62 1.69 0.72<br />
BI BI1 4.90 2.09 0.96 0.82 0.73<br />
BI2 4.87 2.03 0.87<br />
BI3 4.36 2.07 0.88<br />
CPS CPS1 4.88 1.46 0.88 0.76 0.54<br />
CPS2 4.89 1.44 0.75<br />
CPS3 4.79 1.57 0.83<br />
CPS4 4.36 1.62 0.58<br />
C(CL)C CCC1 5.31 1.69 0.91 0.61 0.53<br />
CCC2 5.13 1.60 0.85<br />
CCC3 4.84 1.76 0.62<br />
CLC1 5.50 1.59 0.71<br />
CLC2 5.11 1.58 0.77<br />
CLC3 5.03 1.55 0.78<br />
CDS CDS1 4.74 1.68 0.90 0.80 0.59<br />
CDS2 4.44 2.03 0.83<br />
CDS3 4.82 1.84 0.82<br />
CDS4 4.48 1.97 0.75<br />
CDS5 4.07 1.92 0.63<br />
The reliability <strong>of</strong> an item was assessed by measuring its factor<br />
loading onto the un<strong>der</strong>lying construct. Hair et al. [12] recommended<br />
a factor loading <strong>of</strong> 0.5 to be good indicator <strong>of</strong> validity at<br />
the item level. The factor analysis identified eight distinct factors<br />
(Table 1): 1) Performance Expectancy (PE), 2) Satisfaction (STF),<br />
3) Social Influence (SI), 4) Self-Efficacy (SEF), 5) Behavioral<br />
Intention (BI), 6) Confidence with Problem Solving (CPS), 7)<br />
Confidence for using Data Commands (Conditional-Loop)<br />
(C(CL)C) and 8) Confidence for Data Structures (CDS).<br />
The third step for assessing the convergent validity is the average<br />
variance extracted (AVE); AVE measures the overall amount <strong>of</strong><br />
22<br />
variance that is attributed to the construct in relation to the amount<br />
<strong>of</strong> variance attributable to measurement error. Convergent validity<br />
is found to be adequate when the average variance extracted is<br />
equal or exceeds 0.50 [16].<br />
To examine the research questions regarding the differences in<br />
students’ perceptions among German Gymnasium, and CS freshmen,<br />
we used an Analysis <strong>of</strong> Variances (ANOVA), including the<br />
eight factors as dependent variables and the students’ group as<br />
independent variable. As we can see from the outcome data in<br />
Table 2, students’ group has a significant impact on students’ PE,<br />
STF, SI, BI and CDS. On the other hand students’ group does not<br />
exhibit significant difference on students’ SEF, CPS and C(CL)C.<br />
Table 2 displays our results regarding the significance <strong>of</strong> the<br />
differences, while Figure 3 shows the average results for each<br />
factor over the groups. SEF, CPS and C(CL)C have no significance<br />
difference among the three groups and from Figure 2 we<br />
can notice that these factors are on the same levels at each group.<br />
On the other hand PE, STF, SI, BI and CDS do have significant<br />
differences among the groups and in some cases these difference<br />
are quite remarkable (i.e., BI, PE).<br />
Table 2. The differences among the students’ groups<br />
Factor Mean (S.D.) F Result<br />
Lyceum Gymnas. CS<br />
(GR) (GE) Freshmen<br />
PE 4.10 (1.58) 3.75 (1.70) 5.61 (0.79) 14.97** S.D.<br />
STF<br />
SI<br />
5.03 (1.39)<br />
3.54 (1.90)<br />
4.65 (1.67)<br />
3.93 (1.79)<br />
5.90 (0.90)<br />
5.19 (1.34)<br />
6.86**<br />
9.15**<br />
S.D.<br />
S.D.<br />
SEF 3.58 (1.35) 3.77 (1.61) 3.34 (0.94) 0.78 I.D.<br />
BI 3.94 (1.97) 4.48 (1.59) 6.24 (0.90) 17.11** S.D.<br />
CPS 4.51 (1.46) 4.84 (1.44) 5.06 (0.79) 1.81 I.D.<br />
C(CL)C 5.14 (1.37) 4.74 (1.52) 5.55 (1.10) 2.73 I.D.<br />
CDS 4.10 (1.44) 4.66 (1.82) 5.08 (1.51) 4.03* S.D.<br />
**p
(GE) and freshmen CS and Lyceum (GR). Table 3 summarizes<br />
the significant results from the Games-Howell post hoc test.<br />
Table 3. Games-Howell post hoc test<br />
Freshmen CS<br />
Gymnasium (GE) PE*, STF*, SI*, BI*<br />
Lyceum (GR) PE*, STF*, SI*, BI*, CDS*<br />
* The mean difference is significant at the 0.05 level.<br />
5. CONCLUSION AND DISCUSSION<br />
Looking at Figure 2, we can easily notice that the scores <strong>of</strong> the<br />
Greek and German secondary education students are generally on<br />
the same levels. Additionally, our Games-Howell test showed that<br />
there is no significant difference in the perceptions <strong>of</strong> Greek and<br />
German secondary education students.<br />
On the other hand, compared to the school students, the Greek<br />
higher education students had significantly higher scores in 4<br />
(compared to German Gymnasium) respectively 5 (compared to<br />
Lyceum) variables. Nevertheless, there are 3 factors, where is no<br />
significant difference (SEF, CPS and C(CL)C). Notably, it seems<br />
that these factors ranged on the same levels although CS freshmen<br />
have more exposure on CS and Programming. This can be possibly<br />
based to the fact that in secondary education educators are<br />
using learner friendly environment (e.g. BlueJ, Alice or Scratch);<br />
and these environments increasing students’ self-efficacy and<br />
confidence regarding a CS and Programming [1, 8]. However, for<br />
the case <strong>of</strong> data structures (CDS) we identified a significant difference<br />
among the students <strong>of</strong> Lyceum (GR) and the CS freshmen,<br />
as the latter look more confident for data structures.<br />
The most astonishing result is the dramatic (relative) drop in Self-<br />
Efficacy (SEF), which might indicate that those “future CS pr<strong>of</strong>essionals”<br />
are not as self-confident regarding their programming<br />
abilities, as one would expect at the first glance, as this is the only<br />
factor indicating a lower level or the CS freshmen, although this<br />
difference is not significant, too. It might be due to the fact that<br />
CS freshmen had to deal with substantially more demanding<br />
problems that they found as hard to solve as the school students<br />
found their more easy ones. In contrast, PE, SI and STF are also<br />
indicating significance differences among the CS freshmen and<br />
the secondary education students. This may be possible explained<br />
to the wide enrolment and familiarity <strong>of</strong> freshmen with CS and<br />
programming and shows that they are quite content with their<br />
choice <strong>of</strong> career.<br />
The most significant difference among the two courses is indicated<br />
in students’ BI. This seems plausible, as the freshmen had<br />
chosen their career at this point <strong>of</strong> time and regarded themselves<br />
as specialists already.<br />
As with any empirical study, there are some limitations. First, in<br />
this study the respondents are Greek and German students, who<br />
had attended the respective educational systems; this may limit<br />
the extend <strong>of</strong> the generalization <strong>of</strong> the findings. However, this<br />
study is one <strong>of</strong> the few so far which combines empirical data from<br />
students’ beliefs in a cross-cultural framework; as such these<br />
findings are expected to shed light on that direction. Secondly, the<br />
data are based on self-reported method, other methods such as<br />
depth interviews and observations could provide a complimentary<br />
picture <strong>of</strong> the findings through data triangulation. Despite these<br />
limitations, the findings <strong>of</strong> this study generate valuable insights,<br />
which can be used as part <strong>of</strong> hypotheses for representative followup<br />
studies in students’ beliefs for CS education and in cross-<br />
country and cultural contexts <strong>of</strong> CS education research.<br />
23<br />
6. REFERENCES<br />
[1] An<strong>der</strong>son, M., et al., (2011). Affecting attitudes in first-year<br />
computer science using syntaxfree robotics programming.<br />
ACM Inroads 2, 3, 51-57. DOI=10.1145/2003616.2003635<br />
[2] Bandura A. (1986). The explanatory and predictive scope <strong>of</strong><br />
self-effıcacy theory. J Clin Soc Psychol, 4, 359–73.<br />
[3] Barker, LJ, McDowell, C, Kalahar. K. 2009. Exploring factors<br />
that influence computer science introductory course students<br />
to persist in the major. SIGCSE Bull. 41, 1, 153-157.<br />
[4] Bayerisches Landesamt für Statistik und Datenverarbeitung<br />
(2010). Verteilung <strong>der</strong> Schüler in <strong>der</strong> Jahrgangsstufe 8 nach<br />
Schularten und Regierungsbezirken - Schuljahr 2010/11.<br />
https://www.statistik.bayern.de/medien/statistik/bildungsozia<br />
les/verteilung_<strong>der</strong>_sch__ler_2010.2011.pdf<br />
[5] Biggers, M, Brauer, A and Yilmaz. T. (2008). Student perceptions<br />
<strong>of</strong> computer science: a retention study comparing<br />
graduating seniors with cs leavers. In SIGCSE 402-406.<br />
[6] Chen, K., Razi, M. and Rienzo, T. (2011), Intrinsic Factors<br />
for Continued ERP Learning: A Precursor to Interdisciplinary<br />
ERP Curriculum Design. Decision Sciences Journal <strong>of</strong><br />
Innovative Education, 9, 149–176.<br />
[7] Cronbach, L.J. (1951). Coefficient alpha and the internal<br />
structure <strong>of</strong> tests. Psychometrika, 16(3), 297-334.<br />
[8] Dillon, E., An<strong>der</strong>son, M., and Brown. M. 2012. Comparing<br />
feature assistance between programming environments and<br />
their "effect" on novice programmers. J. Comput. Sci. Coll.<br />
27, 5, 69-77.<br />
[9] Fornell, C., Larcker, D.F. (1981): Evaluating structural equation<br />
models with unobservable variables and measurement<br />
error. Journal <strong>of</strong> Marketing Research, 48, 39--50.<br />
[10] Games, P.A., Keselman, H.J., & Rogan, J.C. (1981). Simultaneous<br />
pairwise multiple comparison procedures for means<br />
when sample sizes are unequal. Psychological Bulletin, 90,<br />
594-598.<br />
[11] Giannakos, M. N et al. (2011). Programming in secondary<br />
education: benefits and perspectives. In Proc. <strong>of</strong> the ITiCSE<br />
'11, ACM, 349. DOI= 10.1145/1999747.1999863<br />
[12] Hair, J.F., et al., (2006): Multivariate data analysis, 6th edn.<br />
Upper saddle River, NJ: Prentice-Hall International<br />
[13] Hubwieser, P, et al., (2011). Computer science/informatics in<br />
secondary education. In Proc. <strong>of</strong> the ITiCSE-WGR '11,<br />
ACM, 19-38. DOI=10.1145/2078856.2078859<br />
[14] Lin, C. S., Wu, S. & Tsai, R. J. (2005). Integrated perceived<br />
playfulness into expectation– confirmation model for web<br />
portal context. Information & Management, 42(5), 683-693.<br />
[15] Metcalfe, J., & Finn, B. (2008). Evidence that judgments <strong>of</strong><br />
learning are causally related to study choice. Psychonomic<br />
Bulletin & Review, 15,174 –179.<br />
[16] Segars, A.H. (1997). Assessing the unidimensionality <strong>of</strong><br />
measurement: A paradigm and illustration within the context<br />
<strong>of</strong> information systems research. Omega, 25(1), 107-121.<br />
[17] Shih, H. (2008). Using a cognitive-motivation-control view<br />
to assess the adoption intention for Web-based learning.<br />
Computer & Education, 50, (1), 327-337.<br />
[18] Venkatesh, V. Morris, M.G. Davis, G.B. Davis, F.D. (2003)<br />
User acceptance <strong>of</strong> information technology: toward a unified<br />
view, MIS Quarterly 27(3), 425–478.
APPENDIX<br />
Table 4. The Factors, their Definitions and their items<br />
Factors Operational Definition Items Source<br />
Performance The degree to which an Using programming improves my performance in a task. (PE1) [18]<br />
Expectancy individual believes that Programming enhances my effectiveness in tasks progressing.<br />
(PE)<br />
attending the respective<br />
course is useful for him/her.<br />
(PE2)<br />
Programming would make it easier to complete a task. (PE3)<br />
Programming increases productivity in completing tasks. (PE4)<br />
Satisfaction The degree to which a I am satisfied with the programming experience. (STF1)<br />
[14]<br />
(STF)<br />
person positively feels with I am pleased with the programming experience. (STF2)<br />
the respective course. My decision to use programming was a wise one. (STF3)<br />
My feeling to use programming was good. (STF4)<br />
Social Influ- The degree to which an People who are important to me think that Ι should learn pro- [14]<br />
ence (SI) individual perceives that gramming. (SI1)<br />
most people who are im- People who influence my behavior encourage me to learn proportant<br />
to him think he<br />
should or should not attend<br />
the respective course.<br />
gramming. (SI2)<br />
Self-Efficacy The degree <strong>of</strong> conviction I could complete a programming task …<br />
[17]<br />
(SEF)<br />
that one can successfully if there was no one around to tell me what to do. (SEF1)<br />
execute the operation re- if I had only instructions for reference. (SEF2)<br />
quired to produce the outcomes.<br />
if I had never used it before. (SEF3)<br />
Behavioral The degree <strong>of</strong> students’ I intend to continue learning programming in the future. (BI1) [18]<br />
Intention (BI) willingness to attend the I will continue learning programming in the future. (BI2)<br />
respective course<br />
I will regularly learn programming in the future. (BI3)<br />
Confidence I feel confident in …<br />
Problem Solv- The degree <strong>of</strong> students’ un<strong>der</strong>standing the problems presented to me. (CPS1)<br />
[11]<br />
ingConfidence (CPS)<br />
confidence to successfully<br />
cope with various problems<br />
identifying the components <strong>of</strong> a problem. (CPS2)<br />
analyzing a problem to other simpler ones. (CPS3)<br />
posing a problem, formulating it accurately and completely.<br />
(CPS4)<br />
Confidence The degree <strong>of</strong> students’ formulating the forms <strong>of</strong> conditional statement if. (CCC1) [11]<br />
for using Data<br />
Commands<br />
confidence to successfully<br />
use data commands<br />
discerning the differences <strong>of</strong> the forms <strong>of</strong> conditional statement<br />
if. (CCC2)<br />
(Conditional-<br />
selecting the best form <strong>of</strong> conditional statement depending on the<br />
Loop)<br />
problem. (CCC3)<br />
formulating the loop statement.(CLC1)<br />
selecting the best loop statement. (CLC2)<br />
using the appropriate loop statement. (CLC3)<br />
Confidence The degree <strong>of</strong> students’ deciding whether it is necessary to use an array. (CDS1)<br />
[11]<br />
for Data Struc- confidence to successfully selecting the formula <strong>of</strong> array (one‐dimensional, two‐<br />
tures (CDS) use data structures<br />
dimensional, etc.). (CDS2)<br />
entering, processing and printing the items <strong>of</strong> an array. (CDS3)<br />
doing general exercises and exercises <strong>of</strong> searching and sorting<br />
using the structure <strong>of</strong> the array. (CDS4)<br />
defining the structures <strong>of</strong> the stack and queue with the correspondent<br />
operations. (CDS5)<br />
24
Performance<br />
Expectancy<br />
(PE)<br />
Satisfaction<br />
(STF)<br />
Social Influence<br />
(SI)<br />
Self-Efficacy<br />
(SEF)<br />
Behavioral<br />
Intention<br />
(BI)<br />
Problem<br />
Solving<br />
Confidence<br />
(CPS)<br />
Confidence<br />
for using<br />
Data Commands<br />
(Conditional-<br />
Loop)<br />
Confidence<br />
for Data<br />
Structures<br />
(CDS)<br />
Table 5. Complete Results <strong>of</strong> the Games-Howell post hoc test<br />
Independent Variables Mean Std.<br />
(I) (J)<br />
Differ.<br />
(I-J)<br />
Error<br />
Sig.<br />
3 rd Lyceum (GR) Gymnasium (GE) 0.35 0.38 0.63<br />
Freshmen CS -1.51* 0.26 0.00<br />
Gymnasium (GE) 3 rd Lyceum (GR) -0.35 0.38 0.63<br />
Freshmen CS -1.86* 0.35 0.00<br />
Freshmen CS 3 rd Lyceum (GR) 1.51* 0.26 0.00<br />
Gymnasium (GE) 1.86* 0.35 0.00<br />
3 rd Lyceum (GR) Gymnasium (GE) 0.38 0.36 0.55<br />
Freshmen CS -0.87* 0.25 0.00<br />
Gymnasium (GE) 3 rd Lyceum (GR) -0.38 0.36 0.55<br />
Freshmen CS -1.25* 0.35 0.00<br />
Freshmen CS 3 rd Lyceum (GR) 0.87* 0.25 0.00<br />
Gymnasium (GE) 1.25* 0.35 0.00<br />
3 rd Lyceum (GR) Gymnasium (GE) -0.39 0.42 0.62<br />
Freshmen CS -1.66* 0.35 0.00<br />
Gymnasium (GE) 3 rd Lyceum (GR) 0.39 0.42 0.62<br />
Freshmen CS -1.26* 0.41 0.01<br />
Freshmen CS 3 rd Lyceum (GR) 1.66* 0.35 0.00<br />
Gymnasium (GE) 1.26* 0.41 0.01<br />
3 rd Lyceum (GR) Gymnasium (GE) -0.19 0.35 0.86<br />
Freshmen CS 0.24 0.25 0.61<br />
Gymnasium (GE) 3 rd Lyceum (GR) 0.19 0.35 0.86<br />
Freshmen CS 0.43 0.34 0.44<br />
Freshmen CS 3 rd Lyceum (GR) -0.24 0.25 0.61<br />
Gymnasium (GE) -0.43 0.34 0.44<br />
3 rd Lyceum (GR) Gymnasium (GE) 0.54 0.46 0.47<br />
Freshmen CS -2.29* 0.32 0.00<br />
Gymnasium (GE) 3 rd Lyceum (GR) 0.54 0.46 0.47<br />
Freshmen CS -1.75* 0.40 0.00<br />
Freshmen CS 3 rd Lyceum (GR) 2.29* 0.32 0.00<br />
Gymnasium (GE) 1.75* 0.40 0.00<br />
3 rd Lyceum (GR) Gymnasium (GE) -0.32 0.33 0.60<br />
Freshmen CS -0.54 0.24 0.07<br />
Gymnasium (GE) 3 rd Lyceum (GR) 0.32 0.33 0.60<br />
Freshmen CS -0.22 0.30 0.75<br />
Freshmen CS 3 rd Lyceum (GR) 0.54 0.24 0.07<br />
Gymnasium (GE) 0.22 0.30 0.75<br />
3 rd Lyceum (GR) Gymnasium (GE) 0.40 0.34 0.47<br />
Freshmen CS -0.41 0.27 0.29<br />
Gymnasium (GE) 3 rd Lyceum (GR) -0.40 0.34 0.47<br />
Freshmen CS -0.81 0.35 0.06<br />
Freshmen CS 3 rd Lyceum (GR) 0.41 0.27 0.29<br />
Gymnasium (GE) 0.81 0.35 0.06<br />
3 rd Lyceum (GR) Gymnasium (GE) -0.56 0.39 0.33<br />
Freshmen CS -0.98* 0.33 0.01<br />
Gymnasium (GE) 3 rd Lyceum (GR) 0.56 0.39 0.33<br />
Freshmen CS -0.42 0.43 0.61<br />
Freshmen CS 3 rd Lyceum (GR) 0.98* 0.33 0.01<br />
Gymnasium (GE) 0.42 0.43 0.61<br />
*. The mean difference is significant at the 0.05 level.<br />
25
InfoSphere: An Extracurricular Learning Environment for<br />
Computer Science<br />
Nadine Bergner<br />
Lehr- und Forschungsgebiet<br />
Informatik 9, RWTH Aachen<br />
Ahornstr. 55<br />
52074 Aachen<br />
+49 241 8021933<br />
bergner@cs.rwth-aachen.de<br />
ABSTRACT<br />
This paper describes one <strong>of</strong> our measures to raise students’<br />
interest for Computer Science (CS) and to provide them with a<br />
realistic idea <strong>of</strong> the field. We outline the un<strong>der</strong>lying concepts and<br />
theories <strong>of</strong> InfoSphere – the extracurricular learning environment<br />
for CS at RWTH Aachen <strong>University</strong>. After explaining the<br />
theoretical, organizational, and infrastructural foundations, we<br />
provide an overview over the didactical concept <strong>of</strong> our 13<br />
different CS workshops for school students <strong>of</strong> different ages.<br />
Furthermore, we introduce the benefits <strong>of</strong> InfoSphere for our three<br />
different target groups: school students, university students in<br />
teacher training, and active CS teachers. Finally, we present first<br />
results from our ongoing evaluation <strong>of</strong> the school students’<br />
perception <strong>of</strong> the field <strong>of</strong> CS before and after visiting one <strong>of</strong> the<br />
InfoSphere workshops.<br />
Keywords<br />
Extracurricular learning environment; didactical concept;<br />
computer science workshops; school students; students in teacher<br />
training; CS teachers.<br />
1. MOTIVATION<br />
There is a near wide agreement <strong>of</strong> economy and society that we<br />
have a growing shortage <strong>of</strong> STEM (science, technology,<br />
engineering and mathematics) graduates in general and<br />
specifically a growing need for computer scientists. Despite the<br />
fact <strong>of</strong> splendid job opportunities, too few students major in<br />
Computer Science (CS). Main reasons for this mismatch are the<br />
low interest and little previous knowledge in STEM topics among<br />
school students [20]. For CS the situation is especially critical<br />
because it is no compulsory school subject in most German states,<br />
e.g. North Rhine-Westphalia. Therefore, mainly the social<br />
environment is responsible for the children’s interest in CS [18].<br />
There are a lot <strong>of</strong> prejudices and false images about computer<br />
scientists, e.g. that they are loners and do not like to work in<br />
teams [18]. As Engeser and Limbert showed it is specifically<br />
important necessary to get young kids into contact with CS topics,<br />
to prevent these misconceptions [8]. This is the only way to reach<br />
the students before they decide against a CS course, which is only<br />
an elective in school, and in the following against a CS study<br />
program at university because <strong>of</strong> the negative social stereotypes.<br />
In or<strong>der</strong> to raise interest in CS and STEM in general and to<br />
communicate a realistic idea about CS, the Learning Technology<br />
Research Group <strong>of</strong> RWTH Aachen <strong>University</strong> established<br />
Jan Holz<br />
Lehr- und Forschungsgebiet<br />
Informatik 9, RWTH Aachen<br />
Ahornstr. 55<br />
52074 Aachen<br />
+49 241 8021935<br />
holz@cs.rwth-aachen.de<br />
26<br />
Ulrik Schroe<strong>der</strong><br />
Lehr- und Forschungsgebiet<br />
Informatik 9, RWTH Aachen<br />
Ahornstr. 55<br />
52074 Aachen<br />
+49 241 8021930<br />
schroe<strong>der</strong>@cs.rwth-aachen.de<br />
InfoSphere 1 as an extracurricular learning environment (in<br />
German Schülerlabor) for CS, which opened in 2010. The main<br />
advantage <strong>of</strong> an extracurricular learning environment is its high<br />
flexibility in time and space which allows for active and selfregulated<br />
and explorative learning methodologies. Furthermore, it<br />
<strong>of</strong>fers extraordinary opportunities in educational technology and<br />
specifically prepared learning materials which cannot be found in<br />
normal school classes. Because <strong>of</strong> these possibilities various<br />
topics <strong>of</strong> CS are presented in an attractive way and thus help to<br />
represent an interesting and multifaceted image <strong>of</strong> CS in general<br />
and thus also counteract social stereotypes. At the moment we<br />
<strong>of</strong>fer 13 workshops about a variety <strong>of</strong> CS topics for school<br />
students from the age <strong>of</strong> eight up.<br />
Additionally, InfoSphere provides opportunities for students in<br />
teacher training for CS as well. They can gain first experiences in<br />
teaching in a prepared and safe environment with pupils and<br />
receive feedback from tutors. We let our CS teacher-students<br />
design learning materials and have them supervise InfoSphere<br />
workshops. The idea is to link the theoretical CS and teaching<br />
knowledge acquired in university courses with practical<br />
experiences in an early stage <strong>of</strong> their education for formative<br />
evaluation.<br />
Within InfoSphere we also address CS teachers <strong>of</strong> all school types<br />
as our third target group. Their situation is <strong>of</strong>ten difficult because<br />
CS is a novel school subject with too few learning materials<br />
provided by established school books. Also, many current CS<br />
teachers have not studied CS as a topic at university. Instead,<br />
they only took part in a further training. Furthermore, CS<br />
constantly changes and evolves. All this makes it is a complex<br />
task to teach CS in school. With our workshop <strong>of</strong>ferings we try to<br />
support and inspire teachers, by providing them the possibility to<br />
visit us with whole school classes. Moreover we make additional<br />
school materials available free <strong>of</strong> charge and support external<br />
teacher training programs.<br />
In summary, InfoSphere is targeted at school students, university<br />
students in teacher training, and active CS teachers. First <strong>of</strong> all,<br />
we provide an opportunity for school students to find out about<br />
various topics <strong>of</strong> CS and actively gain an own picture about what<br />
CS is and if it might be an interesting field for their studies and<br />
later pr<strong>of</strong>ession. On the other hand, we utilize the flexible and<br />
innovative learning environment to improve CS education for<br />
university students in teacher training as well as further education<br />
1 http://schuelerlabor.informatik.rwth-aachen.de
for active CS teachers. Furthermore, InfoSphere represents a<br />
potent research environment, where we can explore and analyze<br />
innovative learning methods, educational designs and media with<br />
school students. To set up research experiments would be a lot<br />
har<strong>der</strong> to achieve in regular school classes.<br />
2. THEORETICAL BACKGROUND<br />
There is ongoing research about extracurricular learning since<br />
several years. In 1994 Griffin described how learning in informal<br />
science settings would be most effective [11]. She looked into the<br />
students’ view <strong>of</strong> learning in such informal settings and identified<br />
orientation and preparation as the most important factors.<br />
Otherwise the unknown learning environment attracts most <strong>of</strong> the<br />
attention instead <strong>of</strong> the workshop itself. Thus, we connect our<br />
workshops with topics <strong>of</strong> regular school lessons. More<br />
specifically regarding extracurricular learning environments,<br />
Euler scrutinized how school students can act as researchers and<br />
what the resulting learning effects can be [9]. The main result is<br />
that it is important for the school students to learn on their own,<br />
which means they have to actively work with the contents instead<br />
<strong>of</strong> just listening to a teacher. Based on these findings we design<br />
our workshops with large proportions <strong>of</strong> action-oriented learning.<br />
In addition, Gu<strong>der</strong>ian studied the learning effects <strong>of</strong> pupils in the<br />
age <strong>of</strong> 10 to 14 after repeated visits <strong>of</strong> a physics lab [12]. In<br />
accordance to his findings we <strong>of</strong>fer additional workshops for<br />
individual participants in or<strong>der</strong> to keep up and raise students’<br />
interest.<br />
As a consequence <strong>of</strong> current findings about extracurricular<br />
learning environments and with regard to research results<br />
concerning learning concepts for STEM topics, we design our<br />
workshops to facilitate explorative, action-oriented learning. So<br />
far, there is little research about this didactical learning concept in<br />
the field <strong>of</strong> CS, but results from other STEM fields can also be<br />
applied to InfoSphere. For example, Bell has focused on<br />
experimental learning in physics [4]. He worked out the following<br />
activity model (translated by the authors):<br />
Table 1. Activity model following Bell<br />
1. defining<br />
2. exploring<br />
3. reflecting<br />
orienting<br />
formulating the problem<br />
assuming<br />
planning<br />
searching for information<br />
expressing knowledge / modeling<br />
experimenting<br />
analyzing / finding results<br />
presenting<br />
discussing / reflecting<br />
applying<br />
For InfoSphere workshops we distinguish three phases <strong>of</strong> defining<br />
the problem, exploring by planning/modeling/experimenting and<br />
reflecting by presenting/discussing learning and working results.<br />
The first stage focusses on the students’ self-directed process <strong>of</strong><br />
finding and defining the problem. An authentic, all-day situation<br />
27<br />
is presented to the students and questions lead them to discover a<br />
problem to be solved. The second stage is the main working phase<br />
during the InfoSphere workshops. The participants work in<br />
groups to solve tasks on different complexity levels. Thereby, we<br />
follow a scheme leaning towards a s<strong>of</strong>tware developing process <strong>of</strong><br />
planning, modeling, experimenting and realization. In the third<br />
stage we encourage reflection and additionally promote students’<br />
communication and presenting skills by letting them present their<br />
findings to the other participants.<br />
Aepkers et al. [1] studied exploratory, inquiry-based learning in<br />
general. Many <strong>of</strong> their findings describe effects on the learning<br />
process which are independent <strong>of</strong> the content and thus are applied<br />
to InfoSphere workshops as well.<br />
One specific goal <strong>of</strong> InfoSphere is to have participants develop a<br />
more realistic concept <strong>of</strong> the field <strong>of</strong> CS. Studies about novices’<br />
expectations and prior knowledge <strong>of</strong> CS e.g. [21] and [8] show<br />
that the general perception <strong>of</strong> CS does not match well with reality.<br />
Therefore, we consi<strong>der</strong> it important to foster a more realistic<br />
perception <strong>of</strong> CS, and to avoid stereotypes, and raise an interest<br />
for elective CS courses. In detail, <strong>Knobelsdorf</strong> evaluated the<br />
motives <strong>of</strong> successful CS university students and explored reasons<br />
for dropping out. Her analyses show that the most mentioned<br />
reason for problems in CS studies is “too challenging<br />
requirements” followed by “too low mathematical prior<br />
knowledge”, “lack <strong>of</strong> time” and “poor motivation” [16]. To avoid<br />
students from enrolling into a CS university program with wrong<br />
expectations we also inform about the contents and requirement <strong>of</strong><br />
CS study programs. This cannot be achieved only in InfoSphere<br />
workshops. We rather educate (future) CS teachers as multipliers,<br />
who observe the misconceptions concerning CS during<br />
workshops and while preparing InfoSphere learning media. For<br />
the other mentioned problems we also utilize InfoSphere for<br />
freshman before they start their study program in prep courses.<br />
Also in the first two terms we <strong>of</strong>fer InfoSphere workshops for<br />
students who experience difficulties in their normal courses.<br />
These workshops provide them with an alternative access to the<br />
current topics and give them a second chance catch up and pass<br />
the exam. These uses are not further described in the following <strong>of</strong><br />
this paper; see [2] and [14] for more details.<br />
3. INFOSPHERE CONCEPT<br />
In or<strong>der</strong> to enable building a broa<strong>der</strong> perception <strong>of</strong> CS we select<br />
exemplary topics from many different fields to demonstrate the<br />
wide range <strong>of</strong> CS. The collection <strong>of</strong> workshop topics is based on<br />
fundamental ideas <strong>of</strong> CS as described by Schwill [22]. In<br />
addition, we took German CS school curricula as well as<br />
educational standards [10] into account and link our workshops to<br />
regular school lessons [6].<br />
Furthermore, InfoSphere is targeted at a variety <strong>of</strong> target groups:<br />
school students, university students in CS teacher training, CS<br />
teachers, as well as other pr<strong>of</strong>essional or non-pr<strong>of</strong>essional people<br />
with an interest in CS. The following subsection describes the<br />
advantages for all target groups in more detail.<br />
3.1 Target Groups<br />
First <strong>of</strong> all, for primary and secondary school students InfoSphere<br />
is the place to get in contact with CS. They can gain an insight<br />
into the world <strong>of</strong> informatics without feeling school pressure. All<br />
the workshops are held by students in teacher training <strong>of</strong> RWTH<br />
Aachen <strong>University</strong>. So the participants’ performance does not<br />
influence their school grades. Besides the possibility to visit
InfoSphere with their whole class as part <strong>of</strong> their school activity,<br />
school students can take part in additional workshops on their<br />
own. For this purpose we frequently <strong>of</strong>fer special workshops<br />
which allow an individual registration. Thus, we can first reach all<br />
students <strong>of</strong> a class and try to raise an interest in CS topics and<br />
then continue to support especially interested kids.<br />
InfoSphere is integrated into our educational concept for CS<br />
teacher training, which makes up our second target group.<br />
Teacher students register for three consecutive courses in CS<br />
didactics over three semesters. After learning the basics <strong>of</strong> CS<br />
didactics during the first course, they create an InfoSphere<br />
workshop in groups <strong>of</strong> two to three students during the second<br />
course. By designing and preparing the learning materials and<br />
then mo<strong>der</strong>ating InfoSphere workshops with school students they<br />
gain practical experience about the didactical reduction <strong>of</strong> CS<br />
topics, didactical design, utilizing different innovative learning<br />
methodologies, and acting in front <strong>of</strong> a group <strong>of</strong> school students.<br />
To merge the specialized knowledge <strong>of</strong> CS at university on the<br />
one hand and the school students’ prior knowledge and skills on<br />
the other hand is a challenging task when designing a workshop.<br />
Especially the different challenges for different age groups are<br />
addressed during the second semester course. The third course in<br />
CS didactics is used to reflect the previous experiences un<strong>der</strong><br />
consi<strong>der</strong>ation <strong>of</strong> various pedagogical approaches which can be<br />
used to teach CS in school and in the InfoSphere.<br />
In addition to students from school and university, pr<strong>of</strong>essional<br />
CS teachers are our third target group. The main value <strong>of</strong><br />
InfoSphere for this group consists <strong>of</strong> the extracurricular learning<br />
units to various and novel CS topics. They can visit InfoSphere<br />
with their class to give their students an enthusing insight into CS.<br />
During the workshop the teachers experience their students from a<br />
neutral and observing perspective, which is almost impossible<br />
during regular lessons. Thereby they get to know much about the<br />
students’ learning methods and social skills. Apart from this, the<br />
teachers can download many additional learning materials for use<br />
in their own school lessons from our website 2 for free.<br />
3.2 Basic Conditions<br />
In the following section we present the basic conditions <strong>of</strong> our<br />
extracurricular learning environment. These specifications<br />
comprise our public relations, the room concept, the available<br />
technologies and the applied learning materials.<br />
Public Relations<br />
We have designed a website 2 , flyers and several posters. We<br />
present InfoSphere at several conferences across Europe and we<br />
regularly exhibit on events such as the science night <strong>of</strong> RWTH<br />
Aachen <strong>University</strong> or the local forum for STEM education. In<br />
addition, we present posters and spread out flyers at many<br />
different activities for school students, e.g. the information day at<br />
the RWTH Aachen <strong>University</strong>, Girl’s day, or go4IT!-workshops.<br />
Our website helps us to get in contact with CS teachers, interested<br />
school students and their parents. On the website you can find<br />
information about all provided workshops. Teachers, parents or<br />
ol<strong>der</strong> students can find out what the workshops are about, what<br />
are the target groups, the duration <strong>of</strong> the workshops and the<br />
expected prior knowledge. Accordingly, they can choose and<br />
register to the best fitting workshop for their classes, their children<br />
2 http://schuelerlabor.informatik.rwth-aachen.de or<br />
http://www.schuelerlabor-informatik.de<br />
28<br />
or themselves. Two different monthly newsletters inform school<br />
students and adults about current events. In the download area<br />
especially teachers find a lot <strong>of</strong> material to enhance their regular<br />
school lessons. For example we provide materials for self-study,<br />
which allows teachers to promote their students’ talent without<br />
neglecting the remaining class.<br />
The equipment <strong>of</strong> an extracurricular learning environment also<br />
plays an important role for its success; starting with the layout <strong>of</strong><br />
the rooms, the technological equipment up to the specifically<br />
designed hands-on materials for the workshops.<br />
Room Concept<br />
The building consists <strong>of</strong> several separated rooms which are placed<br />
on two different floors. The biggest one holds the “experimental<br />
area” and is located in the center <strong>of</strong> the building over both floors.<br />
The “teaching room” is located on level 2 and can be used for<br />
microteaching units. The further rooms are a contest area, an<br />
<strong>of</strong>fice and storerooms for all materials (see Figure 1).<br />
Figure 1. Room concept<br />
The students spend most <strong>of</strong> the time in the huge experimental<br />
area, which also contains five big tables for six to eight students.<br />
The specific designed furnishings support working in groups and<br />
utilizing technology, which are both typical for CS and especially<br />
s<strong>of</strong>tware development. This room is ideal for all active parts <strong>of</strong> the<br />
workshops, because the students are able to talk with each other<br />
about problems and encouraged to experiment with possible<br />
solutions and realize them in groups.<br />
During microteaching phases or students’ presentations we use<br />
the teaching room. The tables and chairs are flexible, so we can<br />
arrange them in rows <strong>of</strong> chairs, a circle <strong>of</strong> chairs or small group<br />
tables for four students. One big advantage <strong>of</strong> changing the<br />
location is that the students stop to work on their tasks and thus<br />
listen to the tutors or their classmates. During a workshop the
participants change between the rooms depending on the<br />
didactical concept <strong>of</strong> a phase.<br />
The contest area can be used by small groups <strong>of</strong> students to<br />
discuss or test their work or to prepare presentations. Finally, the<br />
storerooms are very important for us because we use a lot <strong>of</strong> selfdesigned<br />
hands-on materials for the workshops (see section 4.2<br />
Didactic Concept) which has to be stored.<br />
Additionally to the available seating in the different rooms there is<br />
a lot open space, which can be used to experiment with the handson-materials.<br />
The overall idea is to encourage the students to<br />
move around the building to find classmates to discuss their<br />
questions.<br />
Technologies<br />
The collection <strong>of</strong> technologies available in InfoSphere also makes<br />
a big difference to regular school classes. In brief, we can work<br />
with laptops, interactive whiteboards, smartphones, tablets,<br />
microcontrollers, multi-touch-tables, digital cameras, and several<br />
more specialized devices such as barcode-scanners or traffic light<br />
models.<br />
We have 32 laptops with Wi-Fi connection. Although it is<br />
theoretically possible to give each student one laptop we only do<br />
so during the evaluation phase at the end <strong>of</strong> the workshops (see<br />
section 5). During the workshops the participants share one laptop<br />
by two or three students because we want them to work in groups.<br />
This supports creative thinking and also counteracts the prejudice<br />
that computer scientists always work alone. We prefer mobile<br />
computers like laptops, tablets, or smartphones instead <strong>of</strong> desktop<br />
computers so that participants are able to move around and<br />
cooperate in flexible teams and working with varying technology<br />
or learning materials.<br />
Besides several flipcharts there are two interactive whiteboards,<br />
one in the experimental area and one in the micro-teaching room.<br />
They are height-adjustable and therefore also usable for smaller<br />
kids. These whiteboards are used for presentations or classroom<br />
discussions. A special advantage <strong>of</strong> InfoSphere over regular<br />
schools are 12 smartphones and 27 tablets (<strong>of</strong> different types),<br />
which run Android and can be programmed by the students. So<br />
far they are utilized in two different workshops; another one is<br />
currently in development. Furthermore, we are preparing two<br />
workshops based on microcontrollers. With this technological<br />
equipment we demonstrate that CS <strong>of</strong>fers opportunities beyond<br />
the user-side <strong>of</strong> systems and allows students to create and design<br />
technology according to their own imagination. In addition, there<br />
are three multi-touch-tables available. The big advantage <strong>of</strong> these<br />
is the possibility to use them with up to eight students for<br />
collaborative learning. Currently, we have only used them for<br />
research prototypes with university students (see [13]) but not yet<br />
during InfoSphere workshops with school students due to<br />
technical problems with the tables. For more information about<br />
our current workshop <strong>of</strong>fering (as <strong>of</strong> June 2012) see section 4.<br />
Materials<br />
Because <strong>of</strong> our goal to present many different facets <strong>of</strong> CS to<br />
primary and secondary school students <strong>of</strong> all ages it is<br />
fundamental to design a plethora <strong>of</strong> learning materials. Many <strong>of</strong><br />
these are hands-on materials to address different senses. Because<br />
there are no ready-to-use learning materials to all the different CS<br />
topics most <strong>of</strong> them are especially designed for the corresponding<br />
workshops. These materials consist <strong>of</strong> videos, simulations, handson<br />
models, physical circuits or self-developed s<strong>of</strong>tware programs.<br />
29<br />
Furthermore, we use the learning management system Moodle 3<br />
for some workshops. This helps us to provide the digital materials<br />
in a well-organized form. Within the system it is possible to<br />
present texts, pictures, videos or simulations to the participants.<br />
Sometimes the students’ have to answer question or solve little<br />
tasks to proceed to the next section. As a result <strong>of</strong> this the learning<br />
takes place in a self-regulated manner and there is no need for a<br />
tutor to check the answers. Therefore the teacher has time to<br />
observe her or his school students and the student instructors have<br />
time to help participants with individual problems.<br />
3.3 Workshop Organization<br />
All workshops take place in the InfoSphere environment with its<br />
special infrastructure and well prepared learning materials. It is<br />
also a positive effect that students come to new surroundings with<br />
a different atmosphere. The workshops are mentored by university<br />
students; who help the school students without assessing them.<br />
They are also close in their age. In this way InfoSphere also<br />
targets school students who are not yet interested in CS and<br />
allows them to experiment with different topics without being<br />
un<strong>der</strong> the pressure to perform.<br />
The second significant difference from regular CS lessons in<br />
school is the flexibility in time. There is no structure <strong>of</strong> 45 or 90<br />
minute slots. The workshops last from three hours up to three<br />
days. This is the reason why it is possible to use all kinds <strong>of</strong><br />
different learning methods (see 4.2).<br />
4. WORKSHOP CONCEPT<br />
4.1 Overview over the workshops<br />
The current InfoSphere workshops with the corresponding target<br />
groups are displayed in Table 2. There is one workshop for<br />
primary school students, six workshops for middle school students<br />
and six for high school students. Accordingly, the topics differ<br />
from binary numbers to artificial intelligence.<br />
Table 2. Overview over the existing workshops<br />
Nr. Title Age <strong>of</strong> the<br />
target group<br />
(in years)<br />
1 Magic <strong>of</strong> CS – A First Insight into CS<br />
[3, 7]<br />
8-9<br />
2 Easily Programming with Scratch 10-12<br />
3 Tour into the PC 10-12<br />
4 What’s about the Zebra Crossing? –<br />
EAN- & QR-Codes<br />
5 Treasure Hunt with Cryptographic<br />
Methods [5]<br />
10-12<br />
11-13<br />
6 Searching for the Shortest Path 13-16<br />
7 Object-oriented Programming with<br />
Alice<br />
15-17<br />
8 Artificial Intelligence 16-18<br />
9 Lego-Turing Machine 17-18<br />
10 Calculating with DNA 17-19<br />
11 GUI-Programming for Android 17-19<br />
3 http://subprogra.informatik.rwth-aachen.de/moodle2/
12 Smartphone-Remote Control for Lego-<br />
NXTs [13]<br />
18-19<br />
13 Insight into Computer Graphics 18-19<br />
These workshops have been designed with learning materials and<br />
detailed course plans by students in teacher training as seminar<br />
papers or thesis projects. Furthermore, some students work on the<br />
integration <strong>of</strong> the workshops into school lessons according to the<br />
German CS curriculum [19] and to educational standard<br />
recommendations <strong>of</strong> the German society for CS (GI) [10].<br />
4.2 Didactic Concept<br />
The fundamental concept <strong>of</strong> the InfoSphere workshops is<br />
explorative learning. The learning environment supports students<br />
to learn self-directed, which allows for differentiation between<br />
individual learners. We explicitly aim at students who are not yet<br />
interested in STEM topics and not only at highly motivated<br />
students who come to the lab in their spare time. Therefore,<br />
InfoSphere workshops contain open tasks which can be solved on<br />
different complexity levels. Some workshops also <strong>of</strong>fer a lot <strong>of</strong><br />
additional tasks, so very good students are challenged further.<br />
Phases in lecture style are reduced to a minimum (microteaching).<br />
Most tasks in InfoSphere are solved in teams with<br />
hands-on materials.<br />
The workshops can either be integrated into a sequence <strong>of</strong> school<br />
lessons or visited as separate extracurricular learning event. In<br />
future, we are going to design more learning materials to better<br />
incorporate workshops in regular sequences <strong>of</strong> school lessons.<br />
This should make it easier to integrate a study trip into the school<br />
curriculum [11]. On the other hand, separate workshops are<br />
necessary for all the students, who are not taking up CS in school<br />
yet.<br />
Our aim is to correct the social stereotype <strong>of</strong> computer scientists.<br />
Thus, InfoSphere workshops demonstrate that CS is not only<br />
attractive for “computer freaks”. We specifically address students<br />
who did not choose to study CS in school (yet), only because <strong>of</strong> a<br />
wrong CS perception. On the other hand, we also inform<br />
InfoSphere visitors, who seem already interested in the field, what<br />
it means to study CS at a university. As mentioned in section 2<br />
many CS students admit that they had a different idea about<br />
studying CS before enrolling into a program. Many pr<strong>of</strong>essional<br />
computer scientists develop innovative s<strong>of</strong>tware which is a<br />
creative process. That is why we explicitly value learning<br />
materials which encourage participants to creative problem<br />
solving. For example the workshop “Artificial Intelligence” ends<br />
up with the open task to implement a chat bot. For most<br />
InfoSphere tasks and projects there is not only one correct<br />
solution. At the end <strong>of</strong> the workshops students are asked to<br />
present their solutions and describe the problems they<br />
encountered during the solution process. discussion phases<br />
encourage reflecting what computational thinking can mean while<br />
increasing the students’ communication skills by the way. The<br />
workshop “Easily Programming with Scratch” is designed with<br />
the jigsaw teaching technique where is it necessary to help each<br />
other to achieve the goal within the whole group. Good<br />
communication skills will enable the students to learn and work in<br />
collaborative environments in their further life.<br />
Besides the social skills mentioned above many university<br />
students admit that they have not expected how much<br />
mathematics is expected in CS studies. To encounter wrong<br />
expectations, there are also workshops which explicitly make<br />
30<br />
mathematics as the foundation <strong>of</strong> many areas in CS visible. The<br />
“Insight into Computer Graphics” workshop combines matrix<br />
calculations with the corresponding visual effects on graphical<br />
objects. The students get to know how important mathematic<br />
foundations are to present images on a screen. They learn how<br />
matrix computations allow the enlargement, the translation and<br />
the rotation <strong>of</strong> any object in a three-dimensional coordinate<br />
system and furthermore which mathematical calculations are<br />
needed to represent a three-dimensional object on a twodimensional<br />
screen.<br />
In addition to choosing the right contents it is also important to<br />
utilize the appropriate methodology and learning media.<br />
Depending on the topic and the age <strong>of</strong> the target group the course<br />
format, technologies, and materials are selected from the variety<br />
<strong>of</strong> our educational pool. One <strong>of</strong> our important messages is that<br />
technologies can not only be used as they are but also be designed<br />
and adapted. For example in our two workshops “GUI-<br />
Programming for Android” and “Smartphone-Remote Control for<br />
Lego-NXTs” [15] the participants learn to program their own<br />
Android-Apps, so they see smartphones not only as useful devices<br />
but experience them also as platform which they can manipulate<br />
and design according to their own imagination. Simultaneously,<br />
we show them the significance <strong>of</strong> programming as a tool.<br />
Closing up this section there are some didactical decisions which<br />
are independent <strong>of</strong> the subject <strong>of</strong> the learning environment but<br />
nevertheless very important for its success. To <strong>of</strong>fer a learning<br />
environment without school pressure the workshops are held by<br />
university students. The unknown learning environment and the<br />
new and young tutors enable a learning situation which is very<br />
different from regular school lessons. That is one reason declared<br />
by several school students who were not interested in CS courses<br />
at school, but were quite motivated and interested when<br />
participating in an InfoSphere workshop. To further support these<br />
students we also <strong>of</strong>fer workshops with individual registration.<br />
These workshops have some other unique effects: the students<br />
experience to learn in groups with unknown students. So they<br />
learn to communicate with each other and to solve the problems in<br />
teams. In addition the participants work in groups <strong>of</strong> students <strong>of</strong><br />
different ages and prior knowledge, e.g. the workshop “Treasure<br />
hunt with cryptographic methods” [5] is open to kids at the age <strong>of</strong><br />
11 to 14. So they can learn from each other and combine their<br />
different skills to reach the best possible result. Thereby we try to<br />
support various competencies which are useful in school,<br />
university and vocational training.<br />
5. EVALUATION<br />
In general it is very interesting to conduct research for all three<br />
target groups to evaluate what are the main advantages <strong>of</strong> an<br />
extracurricular learning environment for CS for them. Currently<br />
we focus on the school students’ view. We explore how a visit <strong>of</strong><br />
the InfoSphere can change the students’ conception <strong>of</strong> CS. As<br />
described above our aim is to counteract social stereotypes about<br />
computer scientists and thus try to raise the number <strong>of</strong> school and<br />
university students in CS and to reduce the dropout rate.<br />
To examine the changes <strong>of</strong> the students’ conception <strong>of</strong> CS we<br />
conduct a quantitative evaluation before and after the workshops.<br />
For that purpose we developed two online questionnaires with<br />
partly equal questions. The students fill out the first one in school<br />
some days before their trip and the second one at the end <strong>of</strong> the<br />
workshop. In or<strong>der</strong> to correlate the two answers we integrate a<br />
code into the questionnaires to pseudonymise the data sets.
Altogether 830 participants visited the InfoSphere since its first<br />
workshops in January 2011. Figure 2 displays the distribution <strong>of</strong><br />
the 41 workshops to the different topics.<br />
Figure 2. Overview over the number <strong>of</strong> workshops<br />
After development and pretests the actual evaluation phase began<br />
in January 2012. Since then 325 school students registered for an<br />
InfoSphere workshop. Naturally some students take part in more<br />
than one course, so the number <strong>of</strong> different students who visit the<br />
InfoSphere is a bit lower. For our evaluation our sample consists<br />
<strong>of</strong> N = 215 school students (301 valid pretests, 265 valid<br />
posttests, 215 valid matches). In addition to this online-evaluation<br />
we also ask primary school students for their opinion but they do<br />
not have to fill out the extensive questionnaires. Therefore they<br />
are not part <strong>of</strong> the presented evaluation.<br />
More precisely in the first questionnaire we have evaluated the<br />
personal background <strong>of</strong> the participants, their conception <strong>of</strong> CS,<br />
their own interests, their performance at school, what they think<br />
about computer scientists and their expectations for the workshop.<br />
Afterwards we ask them the same questions about their<br />
conception <strong>of</strong> CS and what they think about becoming a computer<br />
scientist. In addition we want to hear their opinion about the<br />
workshop itself.<br />
Figure 3 displays the distribution <strong>of</strong> the participants by age and<br />
gen<strong>der</strong>. In contrast to the proportion <strong>of</strong> women in CS courses at<br />
the RWTH Aachen <strong>University</strong> <strong>of</strong> 10.7 % (215 out <strong>of</strong> 2016<br />
students) [17] the proportion <strong>of</strong> female students in our workshops<br />
is relatively high with 33.5 % females (101 out <strong>of</strong> 301<br />
participants). Students between 12 and 17 years form our main<br />
target group. The group in the age <strong>of</strong> 12 to 14 has the chance to<br />
select CS as their second elective at school. The ol<strong>der</strong> ones have<br />
to choose their courses in the final two years in school (senior<br />
classes – Oberstufe), select their advanced courses<br />
(Leistungskurse), and make up their mind what to study at<br />
university or during apprenticeship. For both groups it is very<br />
important to be informed what CS really mean, to make a wellconsi<strong>der</strong>ed<br />
decision.<br />
We try to reach all types <strong>of</strong> schools, but as we have mostly contact<br />
to grammar school teachers 80.3 % <strong>of</strong> our participants attend this<br />
type <strong>of</strong> school. In addition we try to reach students who have not<br />
been registered in a CS course so far. Up to now 17.8 % <strong>of</strong> the<br />
participants belong to this group. For this group it is interesting to<br />
analyze which workshops they are especially interested in and if<br />
these workshops help them to decide, whether CS could be a right<br />
topic to choose as an elective.<br />
31<br />
Figure 3. Distribution <strong>of</strong> the participants by age and gen<strong>der</strong><br />
We try to provide good access to CS topics for all different kind<br />
<strong>of</strong> school students. To evaluate what are their motives we ask the<br />
participants for their personal interests in computers and other<br />
technologies, discovering backgrounds, learning in groups and<br />
presenting their results. Figure 4 shows that most <strong>of</strong> the<br />
participants like working with a computer, develop programs on<br />
their own and get to know novel technologies best.<br />
Figure 4. School students' interests concerning CS (mean on a<br />
scale from 1 to 6)<br />
One specific goal <strong>of</strong> InfoSphere is to have participants develop a<br />
more realistic concept <strong>of</strong> the field <strong>of</strong> CS and at least consi<strong>der</strong> if it<br />
would be the right choice to become an computer scientist.<br />
Therefore we evaluate how much the students think they know<br />
about being a computer scientist. Figure 5 gives an overview over<br />
the results before and after the workshop. Both bar charts show a<br />
peak at number 6, which means that most <strong>of</strong> the students think<br />
they know quite much about being a computer scientist. The<br />
bigger changes are at number 3 & 4 and number 9 & 10. This<br />
shows that after the workshop more school students think that
they know what being a computer scientist means (67.5% chose a<br />
score <strong>of</strong> 6 to 10) than before the workshop (58.5% chose 6 to 10).<br />
Figure 5. How much students know about being a computer<br />
scientist (before and after the workshop)<br />
To get to know what the students think about being a computer<br />
scientist we also ask some more detailed questions like “Do you<br />
think CS is interesting?”, “Do you think computer scientists have<br />
to be creative?” or “Do you think CS is something for boys<br />
only?”. As a first analysis <strong>of</strong> the data Figure 6 gives an overview<br />
over the school students’ idea <strong>of</strong> CS divided by gen<strong>der</strong>.<br />
Figure 6. School students' image <strong>of</strong> CS (mean on a scale from 1<br />
to 10)<br />
It becomes obvious that boys and girls differ in some points. In<br />
particular, the differences in “CS is interesting and exciting” and<br />
“I could imagine working with CS in future” are our motivation to<br />
further support especially girls to get in contact with CS. In<br />
contrast to that more girls said that they can imagine what a<br />
computer scientist does. For further studies it is interesting what<br />
the girls exactly think about being a computer scientist.<br />
In conclusion, the evaluation shows that InfoSphere workshops<br />
have an effect on the school students’ thinking about CS. To<br />
research the exact effects we have to analyze the data in more<br />
32<br />
detail and develop follow up evaluations. Nevertheless we can<br />
only evaluate the short-term effects <strong>of</strong> a single workshop.<br />
Multiple measures are necessary to reach long-lasting effects.<br />
That is the reason why we <strong>of</strong>fer additional workshops with<br />
individual registration. So the interested students can take part in<br />
another workshop even if it is not possible to have second<br />
excursion with the whole class.<br />
6. SUMMARY & OUTLOOK<br />
In conclusion, InfoSphere is an effective way to raise interest in<br />
CS and to support interested school students. To communicate a<br />
realistic conception <strong>of</strong> CS to school students <strong>of</strong> all ages, we<br />
provide 13 workshops about different CS topics for kids from the<br />
age <strong>of</strong> eight years up.<br />
The flexibility in time and place in combination with innovative<br />
technologies enables numerous extravagant learning possibilities.<br />
This is a characteristic <strong>of</strong> our extracurricular learning environment<br />
which allows us to use action-oriented and self-directed learning.<br />
To further improve the effects <strong>of</strong> our workshops we are going to<br />
design learning materials to prepare and follow up the courses in<br />
school lessons. In addition we are going to extend our <strong>of</strong>ferings<br />
for single registration and evaluate the effect <strong>of</strong> repeated visits.<br />
Apart from the advantages for school students InfoSphere <strong>of</strong>fers<br />
additional possibilities for students in teacher training and CS<br />
teachers. Students in teacher training design the learning materials<br />
and are responsible for the courses. Thereby they gain useful<br />
experience for being a teacher. CS teachers can visit InfoSphere<br />
with their classes to enable the school students to get an insight<br />
into the vast field <strong>of</strong> CS. They have new possibilities to observe<br />
their studnets and can learn about new CS topics and innovative<br />
educational methodologies.<br />
Furthermore, InfoSphere is a magnificent environment for<br />
research in teaching methodology. Up to now we started with<br />
evaluating in which ways InfoSphere workshops affect the school<br />
students’ conception <strong>of</strong> CS but there are much more interesting<br />
research questions, e.g. what is the effect <strong>of</strong> the practical training<br />
for students in teacher training or how can extracurricular<br />
workshops be integrated in regular school lessons. For the near<br />
future our focus lies on further und detailed evaluation <strong>of</strong> the<br />
school students. We are going to have a closer look at the<br />
differences between boys and girls and how the results depend on<br />
the content and structure <strong>of</strong> the visited workshop.<br />
In addition to the 13 established workshops there are currently 8<br />
more workshops in development (as <strong>of</strong> June 2012) (see Table 3).<br />
Table 3. Overview over the planned workshops<br />
Title Target age (in<br />
years)<br />
App-programming with the App Inventor 11-15<br />
Algorithms 13-16<br />
Modeling a Crossroad with Traffic Lights 15-16<br />
Web Technologies 15-16<br />
Greenfoot 17-18<br />
Media Computation 17-18<br />
Game Theory 18-19<br />
House Automation 17-18
7. ACKNOWLEDGMENTS<br />
The initial equipment <strong>of</strong> the InfoSphere was financed by the zdi -<br />
Zukunft durch Innovation.NRW (future by innovation in North<br />
Rhine-Westphalia). Furthermore, InfoSphere cooperates with and<br />
is partly supported by the research project IGaDtools4MINT 4 .<br />
REFERENCES<br />
[1] Aepkers, M., Liebig, S., Bönsch, M., and Kaiser, A. 2002.<br />
Entdeckendes, forschendes und genetisches Lernen.<br />
Unterrichtskonzepte und -techniken 4. Schnei<strong>der</strong>-Verl.<br />
Hohengehren, Baltmannsweiler.<br />
[2] Apel, R., Berg, T. Bergner, N., Chatti, M.A., Holz, J.,<br />
Leicht-Scholten, C., and Schroe<strong>der</strong>, U. 2012. Ein<br />
vierstufiges För<strong>der</strong>konzept für die Studieneingangsphase in<br />
<strong>der</strong> Informatik. Accepted for publication in Proccedings <strong>of</strong><br />
5. Fachtagung zur Hochschuldidaktik <strong>der</strong> Informatik HDI<br />
2012. (in preparation)<br />
[3] Batur, F. and Bergner, N. 2012. Grundschulkin<strong>der</strong><br />
begeistern mit <strong>der</strong> Zauberschule Informatik. In Ideen und<br />
Modelle. 5. Münsteraner Workshop zur Schulinformatik.<br />
Books on Demand GmbH, Nor<strong>der</strong>stedt, 87–94.<br />
[4] Bell, T. 2006. Schülervorstellungen und Lernen von Physik.<br />
Forschendes Lernen. Piko-Brief 6. IPN Kiel, Kiel.<br />
[5] Bergner, N., Holz, J., and Schroe<strong>der</strong>, U. 2012.<br />
Cryptography for Middle School Students in an<br />
Extracurricular Learning Place. In CSEDU 2012 -<br />
Proceedings <strong>of</strong> the 4th International Conference on<br />
Computer Supported Education, Volume 2, Porto,<br />
Portugal, 16-18 April, 2012. SciTePress, 265–270.<br />
[6] Bergner, N., Holz, J., and Schroe<strong>der</strong>, U. 2012. Über<br />
fundamentale Ideen hinaus: Informatik im InfoSphere -<br />
Schülerlabor Informatik. In Ideen und Modelle. 5.<br />
Münsteraner Workshop zur Schulinformatik. Books on<br />
Demand GmbH, Nor<strong>der</strong>stedt, 77–86.<br />
[7] Bergner, N., Leonhardt, T., and Schroe<strong>der</strong>, U. 2011.<br />
Zauberschule Informatik ‐ Einblick in die Welt <strong>der</strong><br />
Informatik für Kin<strong>der</strong> im Grundschulalter. In Weigend, M.;<br />
Thomas, M.; Otte, F. (Hrsg.): Informatik mit Kopf, Herz<br />
und Hand -- Praxisbeiträge zur 14. GI-Fachtagung<br />
"Informatik und Schule" (INFOS2011) Münster: ZfL-<br />
Verlag, in print<br />
[8] Engeser, S., Limbert, N., and Kehr, H. 2010. Studienwahl<br />
Informatik. Abschlussbericht zur Untersuchung, München.<br />
[9] Euler, M. 2005. Schülerinnen und Schüler als Forscher:<br />
Informelles Lernen im Schülerlabor. Naturwissenschaften<br />
im Unterricht. Physik 16, 90, 4–12.<br />
[10] Gesellschaft für Informatik (GI) e. V., Ed. Grundsätze und<br />
Standards für die Informatik in <strong>der</strong> Schule.<br />
Bildungsstandards Informatik für die Sekundarstufe I.<br />
[11] Griffin, J. 1994. Learning to learn in informal science<br />
settings. Research in Science Education 24, 24, 121–128.<br />
[12] Gu<strong>der</strong>ian, P. 2007. Wirksamkeitsanalyse außerschulischer<br />
Lernorte. Der Einfluss mehrmaliger Besuche eines<br />
Schülerlabors auf die Entwicklung des Interesses an<br />
Physik. Dissertation, Humboldt-Universität zu Berlin.<br />
[13] Holz, J., Bergner, N., Schäfer, A., and Schroe<strong>der</strong>, U. 2012.<br />
Serious Games on Multi Touch Tables for Computer<br />
Science Students. In CSEDU 2012 - Proceedings <strong>of</strong> the 4th<br />
International Conference on Computer Supported<br />
4 www.igadtools4mint.de<br />
33<br />
Education, Volume 2, Porto, Portugal, 16-18 April, 2012.<br />
SciTePress, 519–524.<br />
[14] Holz, J., Bergner, N., and Schroe<strong>der</strong>, U. 2012.<br />
Anwendungsorientierte Gestaltung eines Informatik-<br />
Vorkurses als Studienmotivator. Accepted for publication<br />
in Proccedings <strong>of</strong> 5. Fachtagung zur Hochschuldidaktik <strong>der</strong><br />
Informatik HDI 2012. (in preparation)<br />
[15] Holz, J., Leonhardt, T., and Schroe<strong>der</strong>, U. 2011. Using<br />
smartphones to motivate secondary school students for<br />
informatics. In Proceedings <strong>of</strong> the 11th Koli Calling<br />
International Conference on Computing Education<br />
Research. Koli Calling ’11. ACM, New York, NY, USA,<br />
89‐94.<br />
[16] <strong>Knobelsdorf</strong>, M. 2011. Biographische Lern- und<br />
Bildungsprozesse im Handlungskontext <strong>der</strong><br />
Computernutzung. Dissertation, Freie Universität Berlin.<br />
[17] Kupferschläger, L. 2012. Zahlenspiegel <strong>der</strong> RWTH Aachen<br />
2011. RWTH Aachen, Aachen.<br />
[18] Lobbenmeier, D. 2003. Einstellungen von Schülerinnen<br />
zum Informatikunterricht und zur Informatik. Eine<br />
empirische Studie. Schriftliche Hausarbeit.<br />
[19] Ministerium für Schule und Weiterbildung NRW.<br />
Richtlinien und Lehrpläne für die Sekundarstufe II -<br />
Gymnasium/Gesamtschule in Nordrhein-Westfalen<br />
Informatik.<br />
[20] Reher, J. 2010. Dossier über die Informatik in <strong>der</strong><br />
allgemeinbildenden Schule für die Bundeslän<strong>der</strong><br />
Nordrhein-Westfalen und Nie<strong>der</strong>sachsen. Dossier,<br />
Universität Potsdam.<br />
[21] Schulte, C. and Magenheim, J. 2005. Novices' expectations<br />
and prior knowledge <strong>of</strong> s<strong>of</strong>tware development. Results <strong>of</strong> a<br />
Study with High School Students. In International Journal<br />
<strong>of</strong> Environmental & Science Education, 143–153.<br />
[22] Schwill, A. 2001. Fundamentale Ideen <strong>der</strong> Informatik.<br />
http://ddi.cs.uni-potsdam.de/Forschung/Schriften/ZDM.pdf.<br />
Accessed 21 February 2012.
The school experiment InTech – How to influence interest,<br />
self-concept <strong>of</strong> ability in Informatics and vocational<br />
orientation<br />
ABSTRACT<br />
Claudia Hildebrandt<br />
<strong>University</strong> <strong>of</strong> Oldenburg<br />
Computer Science Education<br />
26111 Oldenburg, <strong>Germany</strong><br />
claudia.hildebrandt@uni-oldenburg.de<br />
Engineers and pr<strong>of</strong>essionals in Informatics are very important<br />
nowadays but lacking in many European countries. The<br />
school experiment InTech (Informatics with technical aspects)<br />
wants to overcome this shortage. Therefore Informatics<br />
was established as a regular subject in the grades 7 to 9<br />
in 13 secondary schools in northern <strong>Germany</strong>. For 3 years<br />
the teachers <strong>of</strong> theses classes met and developed studentand<br />
context-oriented lessons in Informatics with technical<br />
aspects to encourage technical thinking and practice and<br />
influence students’ early career choices towards this direction.<br />
This paper gives an overview <strong>of</strong> related research, the<br />
organisation <strong>of</strong> the experiment and the research objectives<br />
<strong>of</strong> our long-term study. We also describe first results which<br />
show that this experiment influences interest, vocational orientation<br />
and self-concept <strong>of</strong> ability in Informatics for most<br />
students, especially for girls.<br />
Keywords<br />
Computer Science Education, interest <strong>of</strong> students in computer<br />
science, self-concept <strong>of</strong> ability in Informatics, vocational<br />
orientation, empirical study<br />
1. INTRODUCTION<br />
“Engineers are needed in this country.” This headline<br />
marks an important turning point in discussions on educational<br />
policy not only in <strong>Germany</strong>. And the demographic<br />
change is going to increase the need for qualified trainees for<br />
a lot <strong>of</strong> jobs in the fields <strong>of</strong> science, technology, engineering<br />
and mathematics (STEM subjects) (see [21], 80). But in<br />
most secondary schools in <strong>Germany</strong> technical thinking and<br />
practice are not rated highly. Most students <strong>of</strong> grammar<br />
schools (Gymnasium), which prepare for university, are not<br />
introduced to technical subjects and procedures. As a consequence<br />
too few students consi<strong>der</strong> the technical field as a<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WiPSCE 2012 Hamburg, <strong>Germany</strong><br />
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$10.00.<br />
34<br />
Ira Diethelm<br />
<strong>University</strong> <strong>of</strong> Oldenburg<br />
Computer Science Education<br />
26111 Oldenburg, <strong>Germany</strong><br />
ira.diethelm@uni-oldenburg.de<br />
career option. In or<strong>der</strong> to secure the country’s future economic<br />
success it is important to increase the number <strong>of</strong> students<br />
who want to take up a pr<strong>of</strong>ession in computer science<br />
or technology. The subject <strong>of</strong> Informatics is a good introduction<br />
to a technical education in general. Therefore the<br />
ministry <strong>of</strong> education encouraged some secondary schools in<br />
northern <strong>Germany</strong> in joining a pilot study teaching Informatics<br />
as a regular subject in grades 7 to 9, 2 to 3 lessons<br />
a week. Some sponsors from the regional metal industry<br />
support the project by providing additional teaching equipment.<br />
The question is now in what ways teaching Informatics at<br />
school can influence interest, the self-concept <strong>of</strong> ability in<br />
Informatics and vocational orientation <strong>of</strong> students towards<br />
a pr<strong>of</strong>ession in computer science or technology.<br />
We report on this pilot project with 13 schools in northern<br />
<strong>Germany</strong> and describe implementation strategies and its<br />
evaluation.<br />
We begin by discussing the research results <strong>of</strong> related studies.<br />
In section 3 we go on to describe the first and second<br />
part <strong>of</strong> the school experiment InTech and the organisation<br />
<strong>of</strong> the ongoing second InTech part. There we also present<br />
the objectives <strong>of</strong> our project. Our concept is based on establishing<br />
teams <strong>of</strong> Informatics teachers, so-called teacher sets.<br />
Section 4 deals with research questions and hypotheses. Section<br />
5 describes our research methods. Our first results will<br />
be presented and explained in section 6. We close with a<br />
final summary discussing important findings.<br />
2. RELATED RESEARCH<br />
2.1 Influence on vocational orientation<br />
Vocational orientation can be defined as a process <strong>of</strong> approximation<br />
and adjustment <strong>of</strong> young people to the world<br />
<strong>of</strong> employment. On the one hand young people must get<br />
to know their interests, wishes and competences in or<strong>der</strong><br />
to orientate themselves. On the other hand there are the<br />
requirements and needs <strong>of</strong> the world <strong>of</strong> employment and society<br />
towards which young people need to be oriented (see<br />
[1]).<br />
According to Taskinen, Asseburg and Walter (see [21],<br />
81f) the following characteristics <strong>of</strong> students can theoretically<br />
influence the pr<strong>of</strong>essional orientation <strong>of</strong> young people<br />
with regard to the STEM-pr<strong>of</strong>essions:<br />
• scientific and mathematical literacy
• self-concept in science<br />
• instrumental motivation<br />
• interest in studying science<br />
They analyzed within the framework <strong>of</strong> PISA 2006 to what<br />
extent the mentioned characteristics are associated with pr<strong>of</strong>essional<br />
orientation. Analyzing questionnaires filled out by<br />
about 7000 grade 9 students they came to primarily the following<br />
conclusions:<br />
36.6% <strong>of</strong> the young people answering the question on their<br />
pr<strong>of</strong>essional expectation, expect to pursue a STEM job at<br />
age 30 (see [21], 90). In comparison with young people not<br />
wanting to enter a STEM pr<strong>of</strong>ession as adults these students<br />
can be described as having a higher self-concept in<br />
science, stronger instrumental motivation and greater interest<br />
in studying science. The expected higher degree <strong>of</strong> scientific<br />
and mathematical literacy, however, can only be found<br />
in the partial sample <strong>of</strong> graduate pr<strong>of</strong>essional expectation.<br />
The girls’ self-concept in science is lower than the boys’.<br />
Furthermore those girls with a graduate STEM pr<strong>of</strong>essional<br />
expectation are characterized by a lower mathematical and<br />
scientific literacy as well as a lower interest in science than<br />
the boys. These results correspond with the findings <strong>of</strong> numerous<br />
other studies. In addition these girls also have a<br />
lower scientific competence than the boys. The instrumental<br />
motivation <strong>of</strong> girls with non-graduate STEM pr<strong>of</strong>essional<br />
expectation is significantly higher than that <strong>of</strong> the boys (see<br />
[21], 95).<br />
There are correlations between the investigated students’<br />
characteristics (scientific and mathematical literacy, self-concept<br />
in science, instrumental motivation and interest in studying<br />
science).<br />
In several analyses <strong>of</strong> STEM-pr<strong>of</strong>essional groups the correlations<br />
between the investigated students’ characteristics<br />
differ according to the examined pr<strong>of</strong>essional group. In some<br />
pr<strong>of</strong>essional groups a correlation between students’ characteristics<br />
and their pr<strong>of</strong>essional expectations could not be<br />
observed. The results <strong>of</strong> the analyses show that the pr<strong>of</strong>essional<br />
expectations <strong>of</strong> the students can only partially be<br />
explained by the theoretically deduced influencing factors <strong>of</strong><br />
the students’ choice <strong>of</strong> career.<br />
Consequently, pr<strong>of</strong>essional choice must be consi<strong>der</strong>ed as<br />
an interplay <strong>of</strong> numerous factors additionally influenced by<br />
gen<strong>der</strong>-specific mechanisms. The results are explained in<br />
detail by Taskinen, Asseburg and Walter in ([21], 90-103).<br />
According to them, instrumental motivation and interest<br />
can, among other things, affect young people’s vocational<br />
orientation. In or<strong>der</strong> to positively influence these students’<br />
characteristics teaching must be aware <strong>of</strong> the importance <strong>of</strong><br />
certain conditions <strong>of</strong> the learning environment.<br />
2.2 Influence on students’ motivation and interest<br />
in learning<br />
Prenzel et al. [17] explore the question to what extent the<br />
conditions <strong>of</strong> the learning environment, as perceived by the<br />
trainees, can assist or interfere with the development <strong>of</strong> selfdetermined<br />
motivation and interest. This question, along<br />
with some others, is researched in a longitudinal study which<br />
accompanied a small group <strong>of</strong> trainees (<strong>of</strong>fice administrators)<br />
in their training. Influencing factors which according<br />
to the state <strong>of</strong> research have a positive influence on the variants<br />
<strong>of</strong> self-determined motivation are assigned to six sets <strong>of</strong><br />
theoretical conditions ([17], 111):<br />
35<br />
• Contextual relevance <strong>of</strong> the learning material (e.g. references<br />
to applicability; relevance in real life; interdisciplinary<br />
approach)<br />
• Quality <strong>of</strong> instruction (e.g. precise structure; intelligibility)<br />
• Communicated interest <strong>of</strong> the lecturer (e.g. commitment;<br />
enthusiasm)<br />
• Social integration (e.g. cooperative style <strong>of</strong> working;<br />
friendly learning environment)<br />
• Support <strong>of</strong> competence (e.g. direct, pertinent and<br />
helpful feedback; individual reference standard)<br />
• Support <strong>of</strong> autonomy (e.g. <strong>of</strong>fer <strong>of</strong> alternatives; scope<br />
to experiment; support in exploring, planning, acting,<br />
learning independently)<br />
These factors are also taken into account by Parchmann et<br />
al. ([16], 17) in the ChiK project (see also below, section<br />
2.4).<br />
The trainees are questioned at intervals <strong>of</strong> about two months<br />
on that particular phase <strong>of</strong> their training and on the<br />
actual last day <strong>of</strong> that particular phase. The study gives an<br />
account <strong>of</strong> the results <strong>of</strong> the first ten months (five measuring<br />
points). There are 18 complete data sets (see [17], 111f).<br />
The theoretically justified assumption that there are systematic<br />
connections between the conditions relevant for motivation<br />
and the variants <strong>of</strong> learning motivation are confirmed<br />
by the findings (see [17], 124).<br />
Take, for example, the variation <strong>of</strong> the motivation “interested”<br />
1 . It indicates that the six supportive conditions are<br />
positively correlated with this variation at the 0,01 level <strong>of</strong><br />
significance. The highest correlation exists with the contextual<br />
relevance and the support <strong>of</strong> autonomy (the value <strong>of</strong><br />
the coefficient <strong>of</strong> correlation r = 0.62) (see [17], 118).<br />
2.3 Influence on students’ plans to take CIT<br />
Downes und Looker ([5], 179) explore the following factors<br />
that influence the participation rate in computer and<br />
information technology (CIT) subjects in senior secondary<br />
schools:<br />
• gen<strong>der</strong><br />
• parental education<br />
• access to and use <strong>of</strong> computers at home and at school<br />
• self-perceived ability in nine different IT home-related<br />
tasks<br />
• self-perceived ability in CIT subjects in school<br />
• attitudes towards CIT subjects and other school subjects<br />
(Mathematics and English) and schooling in general<br />
from which measures <strong>of</strong> “value” were constructed<br />
1 Apart from the factual incentive, the personal and general<br />
importance <strong>of</strong> the topic makes the student work hard to try<br />
and un<strong>der</strong>stand. Interested learning also means that the<br />
student wants to find out more about the topic and wants<br />
to study it (see [17], 109f).
Data in this study are drawn from 11 NSW (state <strong>of</strong> Australia)<br />
schools between 2005 and 2007. It includes 722 surveys.<br />
The male students present higher levels <strong>of</strong> home-based<br />
and school-related abilities than the females. Only for male<br />
students there is a direct connection between home-related<br />
confidence in one’s ability and plans to take CIT subjects<br />
(see [5], 193).<br />
Downes and Looker also find out that there are three key<br />
factors which influence plans to take CIT subjects in senior<br />
secondary years: gen<strong>der</strong>, amount <strong>of</strong> use at school and the<br />
value students place on CIT subjects (see [5], 194). The<br />
latter two factors are inter-related either directly or indirectly<br />
to home use and home- and school-related confidence<br />
in one’s ability. Consequently the authors conclude that<br />
“any school-based interventions that focus on increasing use<br />
<strong>of</strong> IT at school, and increasing the ’value’ <strong>of</strong> CIT subjects,<br />
need to also address increasing home use and self-perceived<br />
skill levels in tasks associated with both home and school<br />
use” ([5], 194f).<br />
2.4 Results from ChiK<br />
The project “Chemie im Kontext (short: ChiK)” is introduced<br />
representatively for the context-based projects biology,<br />
chemistry and physics. Contexts are chosen as starting<br />
points and structuring elements <strong>of</strong> teaching units. These<br />
contexts are either based on the environment <strong>of</strong> the learner<br />
or are relevant through social references or later vocational<br />
perspectives. The chemistry lessons are planned as practical<br />
training and a means to advance the learners’ development<br />
<strong>of</strong> competence (see [11] and [13]). Moreover, the<br />
symbiotic implementation strategy is employed (see [10]).<br />
So-called ”learning communities”, consisting <strong>of</strong> 8 - 12 teachers<br />
from different schools in cooperation with scientists and<br />
representatives <strong>of</strong> the education administration develop, test<br />
and improve teaching units on the basis <strong>of</strong> a context-based<br />
approach to teaching and learning chemistry (see [7], 54-<br />
64). The learning communities are named “sets”. Sets 2 are<br />
follow-up projects <strong>of</strong> sets 1 with low organisational changes.<br />
Two fundamental questions <strong>of</strong> this research are ([7], 65):<br />
• Do the students perceive changes in the teaching approach?<br />
Do the students <strong>of</strong> the sets 1- and sets 2teachers<br />
differ from each other in their perception <strong>of</strong> a<br />
change in teaching approaches? In or<strong>der</strong> to see the<br />
teaching practice from the students’ point <strong>of</strong> view,<br />
they were asked to describe their chemistry lessons<br />
using four characteristic features <strong>of</strong> teaching: systematic<br />
teaching <strong>of</strong> the subject, practical knowledge, crosscurricular<br />
competences and self-controlled learning.<br />
• Do motivation and interest develop equally in students<br />
<strong>of</strong> sets 1 and sets 2?<br />
The data collection for the sets 1 took place at the beginning<br />
<strong>of</strong> the school year 2002/2003 (t0) and at the end <strong>of</strong> the<br />
school years 2002/2003 (t1) and 2003/2004 (t2). The survey<br />
<strong>of</strong> the sets 2 was carried out at the beginning <strong>of</strong> the school<br />
year 2003/2004 and at the end <strong>of</strong> the school years 2003/2004<br />
and 2004/2005. The students had to indicate their level <strong>of</strong><br />
agreement or disagreement for a series <strong>of</strong> statements. The<br />
four-level answer scale ranged from “1” meaning “strongly<br />
disagree” to “4” meaning “strongly agree” (see [7], 66-69).<br />
The following results were obtained:<br />
The students did perceive changes in their chemistry teaching<br />
approach (see table 1). The increased perception <strong>of</strong> prac-<br />
36<br />
tical knowledge and self-controlled learning and at the same<br />
time the reduction <strong>of</strong> the dominance <strong>of</strong> systematic learning<br />
leads to the conclusion that students do notice the essential<br />
criteria <strong>of</strong> the ChiK concept.<br />
The motivation <strong>of</strong> the ChiK students in the chemistry<br />
lessons shows a declining development in the course <strong>of</strong> the<br />
second project year (see table 2). In another comparative<br />
study it could be shown that the motivation <strong>of</strong> the ChiKstudents<br />
declined significantly less than in classes which received<br />
traditional chemistry teaching (see [7], 75). A positive<br />
result is that in the two years un<strong>der</strong> observation there<br />
was no decline <strong>of</strong> interest (see table 2), whereas other scientific<br />
studies reviewing subjects like physics or biology show<br />
a continuous decline in interest from grade 5 to grade 9 (see<br />
[22], 367). The development concerning sustainable interest<br />
is not significant. Fussangel et al. describe the results in<br />
([7], 72-76).<br />
2.5 Summary <strong>of</strong> important results<br />
The studies discussed show that vocational orientation<br />
must be seen as an interplay <strong>of</strong> a lot <strong>of</strong> factors also influenced<br />
by gen<strong>der</strong>. Among other things interests and the selfconcept<br />
play a crucial role. Todt, who analyzes the importance<br />
<strong>of</strong> school for the development <strong>of</strong> childrens’ interests,<br />
also maintains that interests tend to be an important condition<br />
for the choice <strong>of</strong> pr<strong>of</strong>ession or <strong>of</strong> academic study (see<br />
[22], 373). Structures <strong>of</strong> interest are also closely connected<br />
to children’s general estimations and orientations. Specific<br />
and general interests are integrated into the individual selfconcept,<br />
especially towards the end <strong>of</strong> adolescence, i.e. the<br />
end <strong>of</strong> school education. There are a lot <strong>of</strong> indications, however,<br />
that interests can develop during the whole span <strong>of</strong><br />
life, provided they do not contradict the self-image and the<br />
activities involved can be mastered and lead to some satisfaction<br />
(see [22], 375). Certain conditions <strong>of</strong> teaching practices<br />
noticed by the students can influence the students’ characteristics<br />
”self-determined motivation” and ”interest”. These<br />
characteristics are interrelated with each other. On the basis<br />
<strong>of</strong> our research project InTech the teaching characteristics<br />
student- and context-orientation as well as the student<br />
characteristics interest, self-concept <strong>of</strong> ability in Informatics<br />
and vocational orientation will be explored. In the following<br />
chapters we give an overview <strong>of</strong> the history and the organisation<br />
<strong>of</strong> our experiment, the evaluating research and finally<br />
the first results.<br />
3. THE INTECH EXPERIMENT<br />
3.1 Description <strong>of</strong> part I<br />
For three school years, from 2005 to 2008, six secondary<br />
schools in Lower Saxony, <strong>Germany</strong>, tested ways <strong>of</strong> teaching<br />
Informatics in grades 7 to 9. Lessons in participating<br />
classes were held according to “Time Schedule I”, which allocated<br />
three hours per week in grade 7 and four hours in<br />
grades 8 and 9. Different organisational concepts and teaching<br />
units were developed, e.g. some schools cooperated with<br />
the subjects <strong>of</strong> arts and economy and others cooperated with<br />
astronomy or physics. All participating teachers met at regular<br />
intervals to exchange experiences, at least 4 times a<br />
year. Due to the fact that all schools decided to use robots<br />
in their classes and most schools in <strong>Germany</strong> cannot afford<br />
to buy 8 or more sets <strong>of</strong> robots at the same time, the project<br />
was sponsored by the “Stiftung Nie<strong>der</strong>sachsenMetall”.
mean at t0 mean at t2 mean at t0 mean at t2<br />
(sets 1, N = 549) (sets 1)<br />
(sets 2, N = 237) (sets 2)<br />
systematic<br />
subject<br />
teaching <strong>of</strong> the 3.28 3.16 3.09 2.90<br />
practical knowledge 2.37 2.45 2.33 2.50<br />
cross-curricular competences 2.42 2.20 2.33 2.50<br />
self-controlled learning 2.29 2.63 2.10 2.48<br />
Table 1: Description <strong>of</strong> teaching approach from a student-oriented point <strong>of</strong> view (see [7], 73)<br />
mean at t0 mean at t2 mean at t0 mean at t2<br />
(sets 1, N = 59) (sets 1)<br />
(sets 2, N = 238) (sets 2)<br />
motivation 3.08 2.81 3.00 2.74<br />
sustainable interest 1.98 2.04 2.09 2.08<br />
Table 2: Perception <strong>of</strong> motivation and interest from a student-oriented point <strong>of</strong> view (see [7], 74)<br />
Although there was no scientific evaluation during this<br />
first phase some results can yet be reported by Modrow and<br />
Reineke (see [14]):<br />
• At all participating schools the <strong>of</strong>fer was so much in<br />
demand that selection procedures had to be introduced<br />
in the following years.<br />
• Boys and girls were attracted to the <strong>of</strong>fer in equal numbers.<br />
• Motivation exceeded that <strong>of</strong> “normal” compulsory lessons.<br />
• Teachers were positively surprised at the students’ performances.<br />
• At some schools, groups could be motivated to participate<br />
in national and even international competitions<br />
for robotics. Two students from Intech even became<br />
world champions.<br />
• All schools developed teaching units involving robots<br />
(mostly using LEGO-Mindstorms). Boys and girls were<br />
equally enthusiastic about constructing and programming<br />
these.<br />
• All lessons were action/activity-oriented.<br />
• Teamwork was encouraged by the lessons’ project-oriented<br />
layout.<br />
• The participating teachers found it easy to introduce<br />
interdisciplinary questions and problems in Informatics<br />
classes.<br />
3.2 InTech - part II<br />
In or<strong>der</strong> to use and expand the experiences attained and<br />
to enable others to benefit from them as well, technicaloriented<br />
Informatics classes will be introduced or continued<br />
to be developed at 23 additional schools for three more<br />
years, beginning in grade 7. 13 <strong>of</strong> these schools are coordinated<br />
by us from the Carl von Ossietzky Universität Oldenburg<br />
and financed by “Arbeitgeberverband Nordmetall”<br />
and the “Stiftung des Verbands <strong>der</strong> Metall- und Elektroindustrie”.<br />
The other 10 schools are once more supported by<br />
the “Stiftung Nie<strong>der</strong>sachsenMetall” and coordinated by colleagues<br />
from the Georg-August-Universität Göttingen.<br />
37<br />
Schools wanting to participate must possess facilities that<br />
enable them to teach Informatics from the 7th grade onwards.<br />
A further requirement is the presence <strong>of</strong> two appropriately<br />
qualified teachers.<br />
The central concern <strong>of</strong> InTech in relation to the students<br />
is the following: Beginning with the 7th grade, the contextand<br />
student-oriented lessons in Informatics with technical<br />
aspect should encourage technical thinking and practice as<br />
well as introduce students to the functional principles and<br />
utilization <strong>of</strong> information and information systems. By doing<br />
so we want to support students in choosing a career in<br />
the Informatics business.<br />
In or<strong>der</strong> to ensure sustainability by the distribution <strong>of</strong> the<br />
results, we pursue the following arrangements regarding the<br />
teachers:<br />
• to establish a network <strong>of</strong> Informatics teachers,<br />
• to develop, test and upgrade student- and contextoriented<br />
teaching material,<br />
• to make the teaching material available via the Internet<br />
in or<strong>der</strong> to be used and developed further and<br />
• to distribute the teaching concepts by further pr<strong>of</strong>essional<br />
training.<br />
Finally, one goal <strong>of</strong> this second phase is also to conduct<br />
new research and to develop and expand further education<br />
opportunities for teachers.<br />
3.3 The concept <strong>of</strong> InTech<br />
The described goals will be realized by using a “symbiotic<br />
strategy <strong>of</strong> implementation” as explained by Frey et al. (see<br />
[6], 242-244) on the basis <strong>of</strong> the guideline“Technical thinking<br />
and action in meaningful contexts”, developed by our team.<br />
It contains the following strategy: Informatics teachers and<br />
researchers must establish a “working team” (so-called set <strong>of</strong><br />
teachers) which gets together on a regular basis, participates<br />
in different training courses and discusses, develops and realizes<br />
technical-based teaching units co-operatively. The set<br />
<strong>of</strong> teachers consists <strong>of</strong> about 24 teachers from 13 schools who<br />
participate in the project InTech (see [3], 97).<br />
The involvement with technical-related contents and working<br />
methods from different fields leads to the development<br />
<strong>of</strong> project-oriented and activity-based lessons. Options for
interdisciplinary lessons may be worked out. After testing<br />
the developed teaching material it will be improved cooperatively,<br />
if necessary. Afterwards, the teachers upload<br />
their material on the InTech-Portal.<br />
It is important to consi<strong>der</strong> that the success <strong>of</strong> the implementation<br />
process depends on numerous factors, such<br />
as the teachers’ attitudes, beliefs and skills as well as the<br />
co-operation and communication culture at the schools and<br />
the quality <strong>of</strong> the learning communities (see [10], 208-209).<br />
These aspects and the special conditions <strong>of</strong> Informatics teachers<br />
are taken into consi<strong>der</strong>ation in or<strong>der</strong> to develop and distribute<br />
innovative Informatics teaching concepts. These conditions<br />
and relevant questions are presented in [3].<br />
3.4 Teaching approach<br />
In or<strong>der</strong> to develop context-based and student-oriented<br />
lessons, Diethelm et al. <strong>of</strong>fer a framework (see ([4], 103f))<br />
and an exemplary unit (see [2]). The regular meetings <strong>of</strong> the<br />
teachers and researchers as well as the continuing work and<br />
the exchange <strong>of</strong> experiences and teaching units should support<br />
the teachers to change their teaching towards contextand<br />
student-orientation. But the way <strong>of</strong> teaching at school<br />
is determined by each teacher and could be very different.<br />
4. RESEARCH QUESTIONS<br />
The central intention <strong>of</strong> the InTech-project is to inspire<br />
and increase the interest in Informatics and engineering. By<br />
doing so we want to encourage students to choose a career<br />
in this field. In or<strong>der</strong> to judge the success <strong>of</strong> InTech, our<br />
research has to answer the following questions:<br />
1. In which ways did the teaching <strong>of</strong> Informatics change<br />
during the first year <strong>of</strong> the project from the students’<br />
point <strong>of</strong> view?<br />
We think that the co-operation in learning communities<br />
(the sets <strong>of</strong> teachers), different training courses<br />
in the framework <strong>of</strong> InTech, and the development <strong>of</strong><br />
teaching material improve the teaching competence <strong>of</strong><br />
the participating teachers and we expect the students<br />
to observe these changes. It will be analyzed if the students<br />
notice this student-oriented and context-oriented<br />
approach.<br />
2. In which ways do interest in Informatics, self-concept<br />
<strong>of</strong> ability and vocational orientation towards Informatics<br />
and technology change in the course <strong>of</strong> this first<br />
year?<br />
Keeping the results <strong>of</strong> ChiK (see section 2.4) and [22]<br />
in mind, we expect that the interest in Informatics<br />
will either be consistent or decrease a little. If the<br />
boys, as may be expected, overestimate their competence<br />
in Informatics, their self-perceived ability should<br />
decrease during the first project year. Since girls generally<br />
have a less strongly developed self-concept in<br />
science than the boys (see [21]), their self-assessment<br />
should increase between the two measuring points. If<br />
the teaching <strong>of</strong> Informatics is consi<strong>der</strong>ed to be more<br />
student- and context-oriented the average value on the<br />
scale <strong>of</strong> vocational orientation towards Informatics and<br />
technology should rise.<br />
The following research questions concern the correlation <strong>of</strong><br />
the different characteristics.<br />
38<br />
3. To what extent do the teaching characteristics studentand<br />
context-orientation influence vocational orientation,<br />
interest and the self-concept <strong>of</strong> ability in Informatics<br />
and technology during the first project year?<br />
Based on ChiK (see section 2.4) and the study on the<br />
<strong>of</strong>fice manager trainees (see section 2.2) we will analyze<br />
the correlations between the teaching characteristics<br />
student- and context-orientation and the student<br />
characteristics interest in Informatics, self-concept <strong>of</strong><br />
ability in Informatics and technology and vocational<br />
orientation. We expect that the teaching <strong>of</strong> Informatics<br />
will have a positive influence on vocational orientation,<br />
interest and self-concept.<br />
4. To what extent do vocational orientation, interest and<br />
self-concept <strong>of</strong> ability <strong>of</strong> the first and second surveys<br />
correlate?<br />
The studies discussed in 2 show that interests play a<br />
crucial role in the choice <strong>of</strong> pr<strong>of</strong>essional or academic careers<br />
and are closely connected to general estimations<br />
and orientations (see also [22]). We therefore examine<br />
the correlation <strong>of</strong> the three student characteristics.<br />
As described in ([21], 101) vocational orientation can be seen<br />
as an interplay <strong>of</strong> many factors, additionally influenced by<br />
gen<strong>der</strong>. It must be kept in mind that our results cannot<br />
do more than describe certain tendencies, since the samples<br />
are too small and only those students participated who were<br />
members <strong>of</strong> the InTech project. It can therefore be expected<br />
that they already have a basic interest in Informatics, having<br />
chosen the subject voluntarily.<br />
5. RESEARCH METHODS<br />
5.1 Timing and planning <strong>of</strong> the survey<br />
To find answers to the questions above we decided to use<br />
online-questionnaires at the beginning <strong>of</strong> the school year<br />
2009/2010 and again at the end <strong>of</strong> each school year for the<br />
next three years. Due to the fact that most <strong>of</strong> the teachers<br />
don’t teach a learning group for the whole three years and<br />
there also are fluctuations in the learning groups, especially<br />
in the sets, we decided to additionally question the students<br />
at the beginning <strong>of</strong> the last project year (see table 3) in or<strong>der</strong><br />
to notice time-related changes during the last year <strong>of</strong><br />
the project. We developed questionnaires for teachers and<br />
students based on the concept <strong>of</strong> the ChiK-project 2 .<br />
To adapt the concept to our needs, we added some items<br />
accommodating special requirements <strong>of</strong> our project.<br />
5.2 Participants and rate <strong>of</strong> response<br />
Initially, 24 teachers (four female teachers and 20 male<br />
teachers) and about 300 students from 13 different schools<br />
participated in the project. About 70% <strong>of</strong> the participants<br />
in the students’ survey are male, most <strong>of</strong> them were 12 to<br />
16 years old and were grammar school students (Gymnasiasten).<br />
Only a few <strong>of</strong> the participants came from general<br />
secondary schools (Realschule, Hauptschule, Gesamtschule).<br />
2 Note that the questionnaire <strong>of</strong> the ChiK-concept is also<br />
including parts <strong>of</strong> another project – the SINUS-project.<br />
Therefore, further information can be found in Ostermeier<br />
[15] and in unpublished scale handbooks about teachers and<br />
students <strong>of</strong> C. Gräsel, K. Fussangel and J. Schellenbach-Zell.
survey at the<br />
time t0<br />
time <strong>of</strong> survey beginning <strong>of</strong><br />
the school year<br />
2009/2010<br />
survey at the<br />
time t1<br />
end <strong>of</strong> the school<br />
year 2009/2010<br />
survey at the<br />
time t2<br />
end <strong>of</strong> the school<br />
year 2010/2011<br />
survey at the<br />
time t3<br />
beginning <strong>of</strong><br />
the school year<br />
2011/2012<br />
survey at the<br />
time t4<br />
end <strong>of</strong> the school<br />
year 2011/2012<br />
number <strong>of</strong> teachers 24 19 19 - in process<br />
number <strong>of</strong> students 215 221 160 235 in process<br />
Table 3: Overview <strong>of</strong> the times and the numbers <strong>of</strong> participants <strong>of</strong> the surveys<br />
Table 3 provides an overview <strong>of</strong> the times as well as the various<br />
numbers <strong>of</strong> participants in the surveys.<br />
59 students, who are involved in the InTech-project, participated<br />
in the first two surveys. This was identifiable by<br />
means <strong>of</strong> an individual code. In or<strong>der</strong> to observe the development<br />
during the time <strong>of</strong> the project, the study <strong>of</strong> the<br />
research questions (see section 6) is based on the data sets<br />
<strong>of</strong> these 59 students. Of these 59 participants 45 were boys<br />
and 12 were girls (two students gave contradicting information<br />
about gen<strong>der</strong>). Most <strong>of</strong> them were 13 and 14 years old<br />
at the time <strong>of</strong> survey t0 and 14 and 15 years at the time <strong>of</strong><br />
survey t1, and attended the 8th grade, about 85% <strong>of</strong> them<br />
at a grammar school (Gymnasium).<br />
5.3 Definitions <strong>of</strong> constructs with regards to<br />
the students<br />
In or<strong>der</strong> to gain information about the students’ perception<br />
<strong>of</strong> the teaching approach we use the following constructs:<br />
Student-orientation: The construct means that the choice<br />
<strong>of</strong> goals, the content settings and choice <strong>of</strong> methods<br />
have to be adjusted to the students’ needs. It includes<br />
the criterion “support <strong>of</strong> autonomy” from [17] (see section<br />
2.2).<br />
Context-orientation: This construct represents the idea<br />
that the given contexts in the lessons are oriented at<br />
real-world contexts which are <strong>of</strong> interest to the students.<br />
It corresponds with the criterion “contextual<br />
relevance” from [17] (see section 2.2).<br />
The following student-oriented constructs were also used in<br />
the questionnaires:<br />
Interest in Informatics: This construct includes aspects<br />
like enjoying Informatics-related topics and having an<br />
interest in gaining more knowledge in Informatics (for<br />
natural sciences – in the PISA studies 2006 – see ([20],<br />
128); for specific analyses see [18]).<br />
Self-concept <strong>of</strong> ability in Informatics: The construct includes<br />
self-awareness <strong>of</strong> competence.<br />
Vocational orientation: This construct reflects the motivation<br />
to learn a specific job in the field <strong>of</strong> Informatics.<br />
An overview <strong>of</strong> the constructs, their reliability and exemplary<br />
items can be found in table 4.<br />
5.4 Instruments<br />
The examination <strong>of</strong> the constructs (see above) takes place<br />
via items. Most <strong>of</strong> the items are answered on a scale system,<br />
ranging from “disagreement” to “agreement”, where “1”<br />
39<br />
means “I strongly disagree” and “6” means “I strongly agree”<br />
(Likert scale).<br />
A certain number <strong>of</strong> items form a construct. Using Cronbach’s<br />
Alpha-coefficient we can see how far a group <strong>of</strong> test<br />
items can be used as measurement <strong>of</strong> individual variables<br />
(here: constructs). The Cronbach’s Alpha value should lie<br />
between 0.7 and 1. The Cronbach’s Alpha value <strong>of</strong> the construct<br />
“student-orientation” lies slightly un<strong>der</strong> this value at<br />
the time t0 and at the time t1 the rounded-<strong>of</strong>f value is 0.7<br />
exactly. Therefore we include this construct into the analysis.<br />
In table 4 the constructs, exemplary items and the<br />
Cronbach’s Alpha value can be found.<br />
Answers on a Likert scale are typically ordinally-scaled.<br />
Because the answers on our Likert scale are equidistantly<br />
displayed, test participants should recognize the various possible<br />
answers as being equidistant. Consequently, for the<br />
analysis we use the Likert scale as an interval scale.<br />
6. FIRST RESULTS<br />
The first research question deals with the changes in<br />
the Informatics lessons based on the perception <strong>of</strong> the students.<br />
The constructs “student”- and “context-orientation”<br />
will be analyzed separately, according to their gen<strong>der</strong>, at<br />
the times <strong>of</strong> survey t0 and t1. The scale mean values and<br />
standard deviations <strong>of</strong> these constructs are shown in table<br />
5. The scale mean values <strong>of</strong> the construct “student-orientation”<br />
show a substantial increase between the beginning and<br />
the end <strong>of</strong> the first project year independently <strong>of</strong> gen<strong>der</strong>.<br />
The scale mean values <strong>of</strong> the construct“context-orientation”,<br />
however, have only risen slightly.<br />
The scale mean difference <strong>of</strong> the construct “student-orientation”<br />
is statistically significant only with regard to the<br />
boys. The girls’ scale mean difference <strong>of</strong> this construct is<br />
not statistically significant, the reason possibly being that<br />
too little data could be collected.<br />
It can be seen from table 5 that the scale mean values <strong>of</strong><br />
both constructs describing the Informatics lessons increase.<br />
Therefore it can be assumed that the students’ impression is<br />
that their teachers changed to a more student- and contextoriented<br />
approach towards the end <strong>of</strong> the first project year.<br />
This student perception agrees with the InTech-teachers’<br />
own assessment <strong>of</strong> their teaching. The results <strong>of</strong> the teachers’<br />
surveys show that the 18 teachers who participated in<br />
the first two questionnaires assess their teaching as being<br />
more student- and context-oriented and as using more variety<br />
<strong>of</strong> teaching and learning methods towards the end <strong>of</strong><br />
the first year than at the beginning <strong>of</strong> the project (studentorientation:<br />
scale mean value at the time <strong>of</strong> survey t0: 3.40,<br />
at the time <strong>of</strong> survey t1: 3.58; context-orientation: scale<br />
mean value at the time <strong>of</strong> survey t0: 3.46, at the time <strong>of</strong><br />
survey t1: 3.69; variety <strong>of</strong> teaching and learning methods:<br />
scale mean value at the time <strong>of</strong> survey t0: 3.90, at the time
characteristic number<br />
<strong>of</strong><br />
items<br />
exemplary items reliability<br />
(Cronbach’s<br />
Alpha)<br />
student-orientation 5 In the Informatics lessons we get the scope to set<br />
ourselves tasks and goals.<br />
context-orientation 5 In the Informatics lessons we are confronted with<br />
topics which we also like dealing with in our spare<br />
time.<br />
interest in Informatics 9 Informatics lessons are interesting to me because<br />
<strong>of</strong> the topics.<br />
self-concept <strong>of</strong> ability in Infor- 11 I believe that I am able to solve new problems in<br />
matics<br />
Informatics.<br />
vocational orientation 4 I attend Informatics lessons because it establishes<br />
a basis for my pr<strong>of</strong>essional choice.<br />
characteristics mean m<br />
(regarding the<br />
data<br />
boys)<br />
<strong>of</strong> the<br />
student-orientation t0: 3.60<br />
t1: 4.03<br />
context-orientation t0: 3.23<br />
t1: 3.34<br />
Table 4: Registration <strong>of</strong> characteristics <strong>of</strong> students<br />
difference mean m<br />
m1-m0 (regarding the<br />
data<br />
girls)<br />
<strong>of</strong> the<br />
+0.43 t0: 3.40<br />
t1: 3.98<br />
+0.11 t0: 2.73<br />
t1: 2.83<br />
t0<br />
difference StdDev<br />
m1-m0 (regarding the<br />
data<br />
boys)<br />
<strong>of</strong> the<br />
+0.58 t0: 0.92<br />
t1: 0.88<br />
+0.10 t0: 1.30<br />
t1: 1.25<br />
reliability<br />
(Cronbach’s<br />
Alpha)<br />
t1<br />
0.641 0.699<br />
0.830 0.825<br />
0.860 0.862<br />
0.911 0.926<br />
0.844 0.844<br />
StdDev<br />
(regarding the<br />
data <strong>of</strong> the<br />
girls)<br />
t0: 1.09<br />
t1: 1.15<br />
t0: 1.05<br />
t1: 1.14<br />
Table 5: Means and standard deviations regarding students’ perception <strong>of</strong> Informatics lessons<br />
<strong>of</strong> survey t1: 4.07; the mean value differences are not statistically<br />
significant). These tendencies could be attributed<br />
to the constructive work in the InTech-project, because the<br />
teachers pr<strong>of</strong>ited from the teaching materials and the experience<br />
<strong>of</strong> colleagues, and it can be taken to indicate a success<br />
<strong>of</strong> the implementation strategies <strong>of</strong> InTech.<br />
Looking at the construct “context-orientation” it stands<br />
out that the scale mean value <strong>of</strong> the boys is much higher than<br />
that <strong>of</strong> the girls. This mean value difference is, however,<br />
not statistically significant. The reason for this tendency<br />
could be the girls’ lower interest in Informatics (see table<br />
6). Therefore the girls become less aware <strong>of</strong> the contextorientation<br />
than the boys.<br />
For the second research question we examine the constructs<br />
”interest in Informatics”, ”self-concept <strong>of</strong> ability in<br />
Informatics” and ”vocational orientation”. The differences<br />
between the scale mean values at time t1 and time t0 and<br />
the standard deviations <strong>of</strong> the constructs are calculated (see<br />
table 6). Only the mean difference <strong>of</strong> the construct “interest<br />
in Informatics”regarding the result <strong>of</strong> the girls is statistically<br />
significant (based on the significance level <strong>of</strong> 5 %).<br />
As expected the interest in Informatics <strong>of</strong> the boys decreases<br />
a little (see section 4). The self-concept <strong>of</strong> ability,<br />
however, increases. This may mean that the teachers encouraged<br />
the students to embrace technical thinking and<br />
practice as well as introduced them to the functional principles<br />
and utilisation <strong>of</strong> information and information systems.<br />
The boys’ enthusiasm for careers in the Informatics business<br />
could not be increased. But the scale mean value <strong>of</strong> “vocational<br />
orientation towards Informatics” is already high with<br />
a value <strong>of</strong> 4. Moreover the table 6 shows that the scale mean<br />
values <strong>of</strong> the constructs regarding the girls increase. In con-<br />
40<br />
trast to the development in other natural-science subjects<br />
we find that there is a statistically significant increase <strong>of</strong> the<br />
scale mean value <strong>of</strong> the construct “interest in Informatics”<br />
during the first project year. Consequently the Informatics<br />
lessons represent an important instrument to make informatical<br />
and technical aspects accessible to the girls. They were<br />
able to find an interest in Informatics, increase their selfconcept<br />
<strong>of</strong> ability and their vocational orientation regarding<br />
Informatics.<br />
Looking at table 6 we also recognize that girls show much<br />
less interest in technical pr<strong>of</strong>essions and in Informatics. Their<br />
“self-concept <strong>of</strong> ability in Informatics” turns out to be lower<br />
than that <strong>of</strong> the boys. Comparing the three student characteristics<br />
the scale mean differences between the boys and<br />
girls are statistically significant at the time <strong>of</strong> survey t0 (level<br />
<strong>of</strong> significance 5 %). We notice comparable results for the<br />
scientific concepts <strong>of</strong> competence <strong>of</strong> the girls. There are indications<br />
that the scientific concept <strong>of</strong> competence is less<br />
developed in girls than in boys, see [12] and [19]. Possible<br />
reasons regarding the natural sciences are discussed in ([20],<br />
126f) and may be applied to Informatics.<br />
At the time <strong>of</strong> survey t1 the observed scale mean differences<br />
are not statistically significant any more (level <strong>of</strong><br />
significance 5 %). This may indicate that the differences<br />
between the boys and girls decrease in the course <strong>of</strong> the Informatics<br />
lessons, as far as the students’ characteristics are<br />
concerned which would be a good sign.<br />
The third research question is concerned with the correlation<br />
between the constructs “student-orientation” and<br />
“context-orientation”on the one hand and the constructs“interest<br />
in Informatics”, “self-concept <strong>of</strong> ability in Informatics”<br />
and “vocational orientation” on the other hand (see table 7
characteristics mean m<br />
(regarding the<br />
data<br />
boys)<br />
<strong>of</strong> the<br />
interest in Informatics t0: 3.79<br />
t1: 3.65<br />
self-concept <strong>of</strong> ability in In- t0: 3.82<br />
formatics<br />
t1: 4.10<br />
vocational orientation t0: 4.07<br />
t1: 4.00<br />
difference mean m<br />
m1-m0 (regarding the<br />
data<br />
girls)<br />
<strong>of</strong> the<br />
-0.14 t0: 2.92<br />
t1: 3.36<br />
+0.28 t0: 3.11<br />
t1: 3.35<br />
-0.07 t0: 2.95<br />
t1: 3.44<br />
difference StdDev<br />
m1-m0 (regarding the<br />
data<br />
boys)<br />
<strong>of</strong> the<br />
+0.44 t0: 0.80<br />
t1: 1.08<br />
+0.24 t0: 1.08<br />
t1: 0.97<br />
+0.49 t0: 1.17<br />
t1: 1.24<br />
Table 6: Scale mean values and standard deviations regarding student characteristics<br />
and 8). The correlation coefficients are Spearman’s Rho coefficients<br />
and provide a measure <strong>of</strong> the relationship between<br />
two sets <strong>of</strong> data.<br />
The correlation tables 7 and 8 show that apart from one<br />
correlation between “context-orientation” and “interest” for<br />
the boys there are no statistically significant correlations<br />
to be found at the time <strong>of</strong> the first survey, the explanation<br />
being that hardly any <strong>of</strong> the students had any teaching<br />
in Informatics at the time <strong>of</strong> the survey and therefore<br />
could not give an assessment. This is supported by free text<br />
comments <strong>of</strong> some students. At the time <strong>of</strong> survey t1, all<br />
students characteristics correlate with the teaching characteristic<br />
“student-orientation”. Gen<strong>der</strong> differences are to be<br />
noticed only in the level <strong>of</strong> the correlation coefficients and<br />
the level <strong>of</strong> significance. The correlations between the surveyed<br />
characteristics are stronger regarding the data <strong>of</strong> the<br />
girls. It must be taken into account, however, that there are<br />
only 12 data sets <strong>of</strong> girls as opposed to 45 <strong>of</strong> boys. 12 data<br />
sets do not allow any general statements.<br />
studentorientationcontextorientation<br />
interest in<br />
Informatics<br />
t0: 0.306<br />
t1: 0.673*<br />
t0: 0.439**<br />
t1: 0.298**<br />
selfconcept<br />
<strong>of</strong> ability<br />
t0: 0.297<br />
t1: 0.543*<br />
t0: 0.229<br />
t1: 0.290<br />
vocational<br />
orientation<br />
t0: - 0.145<br />
t1: 0.347**<br />
t0: 0.162<br />
t1: 0.128<br />
Table 7: Correlations regarding student characteristics<br />
(*p < 0.01, **p < 0.05, other correlations are<br />
non-significant). Only the data <strong>of</strong> the boys are taken<br />
into account.<br />
studentorientationcontextorientation<br />
interest in<br />
Informatics<br />
t0: 0.500<br />
t1: 0.773*<br />
t0: -0.008<br />
t1: 0.534<br />
self-concept<br />
<strong>of</strong> ability<br />
t0: 0.172<br />
t1: 0.874*<br />
t0: -0.239<br />
t1: 0.391<br />
vocational<br />
orientation<br />
t0: 0.305<br />
t1: 0.577**<br />
t0: -0.136<br />
t1: 0.330<br />
Table 8: Correlations regarding student characteristics<br />
(*p < 0.01, **p < 0.05, other correlations are<br />
non-significant). Only the data <strong>of</strong> the girls are taken<br />
into account.<br />
Contrary to our expectations, only the boys’ data present<br />
a low correlation between “context-orientation” and “interest<br />
in Informatics”. Consequently these results show that<br />
“context-orientation” has only little or no influence on the<br />
41<br />
StdDev<br />
(regarding the<br />
data <strong>of</strong> the<br />
girls)<br />
t0: 0.84<br />
t1: 1.20<br />
t0: 1.07<br />
t1: 1.23<br />
t0: 1.20<br />
t1: 1.20<br />
students’ characteristics. This outcome contradicts the results<br />
<strong>of</strong> [17] (see section 2.2). Further investigation must<br />
show if these findings will be confirmed.<br />
Our next research question referred to the correlations<br />
between the ascertained student characteristics. According<br />
to analyses in natural sciences some correlations were to be<br />
expected (see section 2.1). The correlations <strong>of</strong> the students’<br />
characteristics are shown in table 9 and 10, differentiated by<br />
gen<strong>der</strong>. The correlations are calculated by means <strong>of</strong> the correlation<br />
measure Spearmans Rho, which allows to calculate<br />
the relationship <strong>of</strong> ordinal scaled data.<br />
interest in<br />
Informatics<br />
self-concept<br />
<strong>of</strong> ability<br />
vocational<br />
orientation<br />
interest in<br />
Informatics<br />
1.00<br />
t0: 0.204<br />
t1: 0.534*<br />
t0: 0.537*<br />
t1: 0.599*<br />
self-concept<br />
<strong>of</strong> ability<br />
1.00<br />
t0: 0.365**<br />
t1: 0.230<br />
vocational<br />
orientation<br />
1.00<br />
Table 9: Correlations regarding student characteristics<br />
(*p < 0.01, **p < 0.05, other correlations are<br />
non-significant). Only the data <strong>of</strong> the boys are taken<br />
into account.<br />
interest in<br />
Informatics<br />
self-concept<br />
<strong>of</strong> ability<br />
vocational<br />
orientation<br />
interest in<br />
Informatics<br />
1.00<br />
t0: 0.763*<br />
t1: 0.792*<br />
t0: -0.023<br />
t1: 0.332<br />
self-concept<br />
<strong>of</strong> ability<br />
1.00<br />
t0: 0.156<br />
t1: 0.767*<br />
vocational<br />
orientation<br />
1.00<br />
Table 10: Correlations regarding student characteristics<br />
(*p < 0.01, other correlations are nonsignificant).<br />
Only the data <strong>of</strong> the girls are taken<br />
into account.<br />
Table 9 and 10 present that for the boys there are mo<strong>der</strong>ate<br />
correlations between “interest in Informatics” and “vocational<br />
orientation” (regardless <strong>of</strong> whether or not Informatics<br />
lessons take place) as well as a mo<strong>der</strong>ate correlation between<br />
“interest in Informatics” and “self-concept <strong>of</strong> ability in Informatics”<br />
(after one year <strong>of</strong> Informatics lessons) and a low<br />
correlation between “self-concept <strong>of</strong> ability in Informatics”<br />
and “vocational orientation” (only at the beginning <strong>of</strong> the<br />
project). For the girls there are strong correlations between
“self-concept <strong>of</strong> ability in Informatics” and “vocational orientation”<br />
(after one year <strong>of</strong> Informatics lessons) and between<br />
“self-concept <strong>of</strong> ability in Informatics” and “interest in Informatics”<br />
(regardless <strong>of</strong> whether or not Informatics lessons<br />
take place).<br />
Contrary to our assumptions no correlations can be determined<br />
mathematically for the girls between “interest in<br />
Informatics” and “vocational orientation” . We’ll have to<br />
wait and see whether this phenomenon can be confirmed in<br />
the surveys at the beginning and the end <strong>of</strong> the last project<br />
year. If the answer is positive we’ll have to search for the<br />
reasons.<br />
A strong correlation for the girls and a mo<strong>der</strong>ate correlation<br />
for the boys between “interest in Informatics” and<br />
“self-concept <strong>of</strong> ability in Informatics” exists at the level <strong>of</strong><br />
significance <strong>of</strong> p < 1%. The results demonstrate that the<br />
“interest in Informatics” plays a crucial role. In addition the<br />
“self-concept <strong>of</strong> ability in Informatics” is important for the<br />
girls and correlates strongly with the “vocational orientation”.<br />
Figure 1: Statistically significant correlations between<br />
the examined characteristics after the first<br />
project year<br />
The statistically significant correlations between the examined<br />
characteristics after the first project year are presented<br />
in figure 1. The continuous lines in this figure show<br />
the relationships between the characteristics regarding all<br />
students, the dashed lines only regarding the boys and the<br />
dotted lines display the results regarding only the girls.<br />
The correlation results can not give answers to the question<br />
<strong>of</strong> whether variable A influences variable B or vice versa.<br />
Therefore further research is required, for example observation<br />
<strong>of</strong> the teaching-learning situations from the perspectives<br />
<strong>of</strong> teachers and learners. For the verification <strong>of</strong> the positive<br />
outcomes <strong>of</strong> the InTech-project we have to compare classes<br />
that have Informatics lessons with classes that do not.<br />
7. CONCLUSIONS<br />
We can summarize the following results after the first<br />
project year:<br />
• The Informatics lessons are perceived to be more student-oriented.<br />
Consequently the competence <strong>of</strong> the<br />
teachers and thereby the quality <strong>of</strong> the Informatics<br />
lessons seem to increase.<br />
• “Student-orientation” correlates positively and significantly<br />
with the “interest in Informatics”, the “selfconcept<br />
<strong>of</strong> ability in Informatics” and the “vocational<br />
orientation regarding Informatics”after the first project<br />
year. These and further correlations are presented in<br />
figure 1.<br />
42<br />
• Regarding the girls, the “interest in Informatics”, the<br />
“vocational orientation regarding Informatics” and the<br />
“self-concept <strong>of</strong> ability in Informatics” increase.<br />
It is surprising and <strong>of</strong> course encouraging that the girls’<br />
“interest in Informatics” significantly increased contrary to<br />
the trend in science subjects like chemistry. Consequently<br />
they should get the opportunity to learn Informatics in the<br />
middle grades to enable them to discover possible interests<br />
in technical fields.<br />
According to the findings <strong>of</strong> Downes and Looker (2011,<br />
see [5], 191) home use and home-based ability belief regarding<br />
Informatics are indirectly related to the variable to take<br />
up CIT subjects (the second factor for the boys directly).<br />
Therefore the acquisition <strong>of</strong> knowledge about the functional<br />
principles and utilisation <strong>of</strong> information systems should not<br />
depend on coincidences such as social or regional origin, but<br />
should instead be part <strong>of</strong> general education. The subject<br />
<strong>of</strong> Informatics is well-suited as an introduction to a general<br />
technical education, as the required equipment and facilities<br />
are already in place at schools. With the help <strong>of</strong> the published<br />
experiences <strong>of</strong> the InTech-schools explaining how to<br />
introduce the subject Informatics and <strong>of</strong>fering the existing<br />
teaching material, more schools can start <strong>of</strong>fering the subject<br />
Informatics in the middle grades. And thus the number <strong>of</strong><br />
highly qualified graduates in technical pr<strong>of</strong>essions and study<br />
courses could increase. So far we can call the implementation<br />
<strong>of</strong> InTech a success.<br />
8. ACKNOWLEDGMENTS<br />
The realization <strong>of</strong> the project InTech was only possible<br />
with the support <strong>of</strong> many different people. At this point we<br />
would like to thank Eckart Modrow (Georg-August-Universität<br />
Göttingen), Vera Reineke, Astrid Tengen and Gudrun<br />
Köppen-Castrop (Nie<strong>der</strong>sächsisches Kultusministerium), Peter<br />
Golinski (Arbeitgeberverband NORDMETALL), Natascha<br />
Clasen and Sabine Stöhr (VME-Stiftung Osnabrück-<br />
Emsland), Andreas Breiter (ifib - Institut für Informationsmanagement<br />
Bremen GmbH), Ilka Parchmann (IPN - Leibnitz-Institut<br />
für die Pädagogik <strong>der</strong> Naturwissenschaften und<br />
Mathematik in Kiel), and last but not least all participating<br />
teachers and their schools.<br />
9. REFERENCES<br />
[1] Bundesinstitut für Berufsbildung (Hrsg.).<br />
Berufsorientierung. Version: 2012. demand:<br />
30.04.2012. http://www.good-practice.de/3337.php.<br />
[2] I. Diethelm. Wie forschend-entdeckendes Lernen<br />
gelingen kann - Forschendes und entdeckendes Lernen<br />
im Kontext zum Thema Datenschutz, Internet und<br />
Urheberrecht. LOG IN, (168):28–34, 2011.<br />
[3] I. Diethelm, C. Hildebrandt, and L. Krekeler.<br />
Implementation <strong>of</strong> computer science in context - a<br />
research perspective regarding teacher-training. 9th<br />
Koli Calling International Conference on Computing<br />
Education Research, Finland, pages 96–99, November<br />
2009.<br />
[4] I. Diethelm, J. Koubek, and H. Witten. IniK -<br />
Informatik im Kontext - Entwicklungen, Merkmale<br />
und Perspektiven. LOG IN, (169/170):97–105, 2011.<br />
[5] T. Downes and D. Looker. Factors that influence<br />
students’ plans to take computing and information
technology subjects in senior secondary schools.<br />
Comuter Science Education, 21(2):175–199, June 2011.<br />
[6] A. Frey, C. Gräsel, T. Puhl, and I. Parchmann.<br />
Implementation einer kontextorientierten<br />
Unterrichtskonzeption für den Chemieunterricht.<br />
Unterrichtswissenschaft, 32(3):238–256, 2004.<br />
[7] F. Fussangel, J. Schellenbach-Zell, and C. Gräsel. Die<br />
Verbreitung von Chemie im Kontext: Entwicklung <strong>der</strong><br />
symbiotischen Implementationsstrategie. In Chemie<br />
im Kontext. Von <strong>der</strong> Innovation zur nachhaltigen<br />
Verbreitung eines Unterrichtskonzepts, pages 49–82.<br />
Waxmann, 2008.<br />
[8] C. Gräsel, K. Fussangel, and J. Schellenbach-Zell.<br />
Skalenhandbuch Lehrer/-innen. (unpublished).<br />
[9] C. Gräsel, K. Fussangel, and J. Schellenbach-Zell.<br />
Skalenhandbuch Schüler/-innen. (unpublished).<br />
[10] C. Gräsel and I. Parchmann.<br />
Implementationsforschung - o<strong>der</strong>: <strong>der</strong> steinige Weg,<br />
Unterricht zu verän<strong>der</strong>n. Unterrichtswissenschaft,<br />
32(3):196–214, 2004.<br />
[11] C. Hössle, I. Parchmann, and M. Komorek.<br />
Naturwissenschaften im Kontext. demand: 30.04.2012.<br />
http://www.mathematik.uni-dortmund.de/ieem/<br />
BzMU/BzMU2009/Beitraege/Hauptvortraege/<br />
HOESSLE PARCHMANN KOMOREK 2009<br />
Naturwissenschaften.pdf.<br />
[12] P. Häussler and L. H<strong>of</strong>fmann. A curricular frame for<br />
physics education: Development, comparison with<br />
students’ interests, and impact on students’<br />
achievement and self-concept. Science Education,<br />
84:689 – 705, 2000.<br />
[13] I. Parchmann, C. Gräsel, A. Baer, P. Nentwig, R.<br />
Demuth, B. Ralle, and the ChiK Project Group.<br />
”Chemie im Kontext”: A symbiotic implementation <strong>of</strong><br />
a context-based teaching and learning approach.<br />
International Jounal <strong>of</strong> Science Education, 28(9):1041<br />
– 1062, 2006.<br />
[14] E. Modrow and V. Reineke. Der Modellversuch<br />
Intech08 - Informatikunterricht mit technischen<br />
Aspekten. information brochure, Stiftung<br />
Nie<strong>der</strong>sachsenMetall, 2008. http://www.stiftungnie<strong>der</strong>sachsenmetall.de/docs/SNM<br />
intech web.pdf.<br />
[15] C. Ostermeier. Kooperative Qualitätsentwicklung in<br />
Schulnetzwerken. Eine empirische Studie am Beispiel<br />
des BLK-Modellversuchsprogramms ”Steigerung <strong>der</strong><br />
Effizienz des mathematisch-naturwissenschaftlichen<br />
Unterrichts” (SINUS). Waxmann, Münster, 2004.<br />
[16] I. Parchmann, B. Ralle, and D.-S. D. Fuccia.<br />
Entwicklung und Struktur <strong>der</strong> Unterrichtskonzeption.<br />
Chemie im Kontex. In Chemie im Kontext. Von <strong>der</strong><br />
Innovation zur nachhaltigen Verbreitung eines<br />
Unterrichtskonzepts, pages 9–47. Waxmann, 2008.<br />
[17] M. Prenzel, A. Kristen, R. Dengler, P. Ettle, and<br />
T. Beer. Selbstbestimmt motiviertes und interessiertes<br />
Lernen in <strong>der</strong> kaufmännischen Erstausbildung. In<br />
Lehr-Lernprozesse in <strong>der</strong> kaufmännischen<br />
Erstausbildung. Wissenserwerb,<br />
Motivierungsgeschehen und Handlungskompetenzen,<br />
pages 108–127. Steiner, 1996.<br />
[18] M. Prenzel, K. Schütte, and O. Walter. Interesse an<br />
den Naturwissenschaften. In PISA 06. Die Ergebnisse<br />
<strong>der</strong> dritten internationalen Vergleichsstudie, pages<br />
43<br />
107–124. Waxmann, 2007.<br />
[19] S. Reis and S. Park. Gen<strong>der</strong> differences in<br />
high-achieving students in math and science. Journal<br />
for the Education <strong>of</strong> the Gifted, 25:52 – 73, 2001.<br />
[20] K. Schütte, A. C. Frenzel, R. Asseburg, and<br />
R. Pekrun. Schülermerkmale, naturwissenschaftliche<br />
Kompetenz und Berufserwartung. In PISA 06. Die<br />
Ergebnisse <strong>der</strong> dritten internationalen<br />
Vergleichsstudie, pages 125–146. Waxmann, 2007.<br />
[21] P. Taskinen, R. Asseburg, and O. Walter. Wer möchte<br />
später einen naturwissenschaftsbezogenen o<strong>der</strong><br />
technischen Beruf ergreifen? - Kompetenzen,<br />
Selbstkonzept und Motivationen als Prädiktoren <strong>der</strong><br />
Berufserwartungen in PISA 2006. In: Prenzel, M,<br />
Baumert (Hrsg.): Vertiefende Analysen zu PISA<br />
2006. Zeitschrift für Erziehungswissenschaft,<br />
Son<strong>der</strong>heft, pages 79–105, October 2008.<br />
[22] E. Todt. Die Bedeutung <strong>der</strong> Schule für die<br />
Entwicklung <strong>der</strong> Interessen von Kin<strong>der</strong>n und<br />
Jugendlichen. Unterrichtswissenschaft, (4):362–376,<br />
1985.
Uncovering Structure behind Function – the experiment as<br />
teaching method in computer science education<br />
ABSTRACT<br />
There are lots <strong>of</strong> reports about activities that aim at fostering and<br />
maintaining interest in computing. These activities rely on different<br />
ideas that should help to involve novices and young learners.<br />
In this article examples are given and explained from a disciplinespecific<br />
perspective based on the notion <strong>of</strong> a dual nature <strong>of</strong> digital<br />
artifacts, which has to be reconstructed and integrated during the<br />
process <strong>of</strong> computer science education. The claim is that learners<br />
either focus on the internal computational properties: the structure,<br />
or on the more external reasons to use them: the function –<br />
but rarely are able to develop an integrated perspective on both<br />
sides. Therefore learning experiences should be designed to<br />
bridge these perspectives and enable to perceive the dual nature <strong>of</strong><br />
digital artifacts. This idea <strong>of</strong> bridges between structure and function<br />
is used to design and analyze learning activities with regard<br />
to their potential in fostering and maintaining interest in computer<br />
science, especially to those not already interested. This article<br />
presents and discusses three examples for such experiments.<br />
Based on this, guidelines for experiments as teaching method, and<br />
questions for further research are <strong>der</strong>ived.<br />
Categories and Subject Descriptors<br />
K3.2 [Computers & Education]: Computer and Information<br />
Science Education – computer science education, information<br />
systems education.<br />
General Terms<br />
Experimentation, Human Factors.<br />
Keywords<br />
Secondary CS Education, K-12, Didactics, CS Ed Research,<br />
Pedagogy, Gen<strong>der</strong>, Teaching Methods, Dual Nature, Duality<br />
Reconstruction, Experiment<br />
1. INTRODUCTION<br />
Teaching computer science (CS) <strong>of</strong>ten faces problems like low<br />
enrollments, low retention, and low ratio <strong>of</strong> female computer<br />
science students. As these are urgent concerns, a lot <strong>of</strong> – <strong>of</strong>ten<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that<br />
copies bear this notice and the full citation on the first page. To copy<br />
otherwise, or republish, to post on servers or to redistribute to lists,<br />
requires prior specific permission and/or a fee.<br />
Conference’12, Month 1–2, 2010, City, State, Country.<br />
Copyright 2010 ACM 1-58113-000-0/00/0010…$10.00.<br />
Carsten Schulte<br />
Freie Universität Berlin<br />
Computing Education Research<br />
Königin-Luise Str. 24<br />
14195 Berlin <strong>Germany</strong><br />
schulte@inf.fu-berlin.de<br />
44<br />
locally successful – initiatives and activities take place, but still, in<br />
general the problems remain unsolved.<br />
Here only a fraction <strong>of</strong> such activities can be listed (and <strong>of</strong> course<br />
sometimes there are overlaps between the different variants):<br />
CS for fun / magical computer science, CS unplugged [6, 7,<br />
9]: These activities are introducing motivational examples to<br />
foster interest and motivation without the use <strong>of</strong> computers,<br />
in or<strong>der</strong> to refocus attention from the ‘technical or usage<br />
skills’ to the ‘thinking skills’ involved and useful in everyday<br />
life.<br />
Summer Schools, working groups, and weekend activities:<br />
Their aim is to enable a gentle introduction without a steep<br />
learning curve, without grades, and to help getting in contact<br />
with peers (e.g. courses designed specifically for female novices),<br />
in or<strong>der</strong> to lower the entry barriers to computer science.<br />
Competitions in different organizational forms, local, national<br />
or international, or related to specific tools like pedagogical<br />
IDE’s, robotics or others (see [5]): Competitions foster<br />
high engagement in the activities, and provide feedback from<br />
the engagement itself, but also from comparing results to<br />
others.<br />
Digital communities, like e.g. Scratch [22], build around a<br />
tool/topic: In general, they are aiming at similar aspects like<br />
the above mentioned summer schools, but provide more ongoing<br />
support.<br />
Innovation in the computer science classroom: This is <strong>of</strong>ten<br />
a mix <strong>of</strong> the above mentioned activities, but also specific<br />
ideas and interventions to foster interest. Like e.g. media<br />
computing [13], or computer science in context [29].<br />
Overall, there are lots <strong>of</strong> activities reported, that aim at fostering<br />
and maintaining interest in the computing disciplines (see e.g.. [1,<br />
10, 35, 36] for a comparison and evaluation <strong>of</strong> such types <strong>of</strong><br />
activities). These activities rely on different ideas that should help<br />
to engage novices and young learners. While these ideas <strong>of</strong>ten<br />
denote theories from other domains—like psychology with regard<br />
to motivation or attribution—in this article, a simple idea is presented<br />
which can be used to analyze and describe such activities<br />
and predict effects on learners with different attitudes, and – as the<br />
given examples should demonstrate – is powerful enough to drive<br />
the development <strong>of</strong> new activities.<br />
2. THE SITUATION IN CS ED<br />
The idea <strong>of</strong> introducing experiments draws on prevalent topics<br />
debated in CS Ed (section 2.1 ). Although a myriad <strong>of</strong> activities
aim to react on these topics, experiments focus on a new explanation<br />
<strong>of</strong> these issues, which is outlined in section 2.2 .<br />
2.1 Misconceptions <strong>of</strong> Computing<br />
The situation in CS education is <strong>of</strong>ten characterized by limited<br />
interest in computing, due to e.g. misconceptions, attitudes, gen<strong>der</strong>-bias,<br />
and similar issues [14, 20, 24, 28, 33, 35].<br />
For CS unplugged activities are some evaluative reports available:<br />
Feaster et. al. ([10], p.252) conclude that “the program had no<br />
statistically significant impact on student attitudes toward computer<br />
science or perceived content un<strong>der</strong>standing.” Taub et.al.<br />
[35] yield similar results (p.24) and found the following explanation<br />
(similar to [36]): “only some <strong>of</strong> the objectives were addressed<br />
in the activities, […] the activities do not engage with the students’<br />
prior knowledge and […] most <strong>of</strong> the activities are not<br />
explicitly linked to central concepts in CS” (p.1) and thus modifications<br />
to the activities are needed (A proposal for such modifications<br />
can be found in [2]).<br />
Basically, programs to increase interest (see section 1. ) aim at<br />
demonstrating a ‘true’ image <strong>of</strong> the discipline, showing that it is<br />
accessible for outsi<strong>der</strong>s or minorities, motivating by highlighting<br />
(different) career choices, and so on. In essence, these activities<br />
are trying to minimize misconceptions <strong>of</strong> computing. The – <strong>of</strong>ten<br />
implicit – approach is to exchange limited conceptions by more<br />
suitable ones.<br />
What, if students who are skeptical and don’t see the usefulness<br />
for their daily lives are right? Probably the activities described<br />
above are successful in triggering situational interest, but not in<br />
maintaining and developing individual interest [25]? What if their<br />
so-called misconceptions indeed are suitable for their current<br />
personal situations, in which they are mostly users <strong>of</strong> digital artifacts,<br />
and not programmers or computer scientists?<br />
Consequently, the approach presented here does not aim for (immediately)<br />
changing the perception <strong>of</strong> the discipline, but to<br />
change the perception <strong>of</strong> digital artifacts, and <strong>of</strong> suitable interaction<br />
patterns with digital artifacts. And this then is the starting<br />
point for being able to perceive the discipline within a different<br />
framework.<br />
2.2 Analysis <strong>of</strong> Insi<strong>der</strong> and Outsi<strong>der</strong><br />
The claim that the perception <strong>of</strong> digital artifacts is <strong>of</strong> more immediate<br />
concern is based on the research in computing biographies;<br />
its results will be briefly presented in this section.<br />
Crutzen [8] deconstructed the opposition <strong>of</strong> use and design in<br />
computer science. According to this analysis, “the symbolic<br />
meaning <strong>of</strong> use and design is constructed as an opposition in<br />
which design is active and virtuous and use is passive and not<br />
creative” – and in addition users are constructed as ‘outsi<strong>der</strong>s’,<br />
whereas designers are ‘Insi<strong>der</strong>s’.<br />
In Schulte and <strong>Knobelsdorf</strong> [32] the effects <strong>of</strong> prior experiences<br />
on attitudes and enrollment in computer science were analyzed.<br />
The dichotomy <strong>of</strong> ‘use’ (utilizing pre-given applications) and<br />
‘design’ (creating new programs and applications), based on [8],<br />
was used. Due to this dichotomy, Outsi<strong>der</strong>s perceive themselves<br />
as capable only <strong>of</strong> interacting with a computer in terms <strong>of</strong> using,<br />
whereas Insi<strong>der</strong>s perceive themselves as more competent due to<br />
the skill to design new digital artifacts.<br />
In some contrast to the original assumption <strong>of</strong> such a gap between<br />
‘use’ and ‘design’, the empirical results <strong>of</strong> the study revealed<br />
another biographical process <strong>of</strong> unaffiliated students. This lead to<br />
a third concept: ‘pr<strong>of</strong>essional use’. It refers to the ability to administrate<br />
digital devices, and a focus on interaction with the comput-<br />
45<br />
er in terms <strong>of</strong> installing, configuring, and solving (usage) problems<br />
(see [32]).<br />
This is connected with a misconception <strong>of</strong> the computing disciplines<br />
by Outsi<strong>der</strong>s, who perceive a computing pr<strong>of</strong>essional as a<br />
kind <strong>of</strong> ‘digital caretaker’: A person who takes care <strong>of</strong> and maintains<br />
digital infrastructures and helps when something malfunctions.<br />
Typical activities resemble cleaning, repairing, exchanging<br />
and updating some aspects <strong>of</strong> the whole, but not activities like<br />
designing, creating, or problem solving on a larger scale.<br />
(Aside: There is a popular distinction between digital immigrants<br />
and digital natives, where everybody born after 1980 is regarded<br />
as being a native. In our study, only digital natives participated.<br />
So the difference between use, design, pr<strong>of</strong>essional use, and <strong>of</strong><br />
Insi<strong>der</strong> and Outsi<strong>der</strong> is within the group <strong>of</strong> so-called Insi<strong>der</strong>s.)<br />
The overall conclusion <strong>of</strong> this is that Outsi<strong>der</strong>s are not discouraged<br />
by those affordances educators may see (difficulties in learning<br />
programming, abstraction, complexity), but by problems<br />
related to the everyday use <strong>of</strong> digital technologies: coping with<br />
errors, configuring, installing, and so forth. In other words, the socalled<br />
group <strong>of</strong> Outsi<strong>der</strong>s is not discouraged by perceiving an<br />
insurmountable gap between them as users and the goal to become<br />
a designer. Instead, they perceive a gap between use and being<br />
able to become a digital caretaker.<br />
Moreover, it seems as if this group perceives the digital world and<br />
its myriad artifacts more or less as a given, natural environment -<br />
like nature itself. This perception completely neglects the role <strong>of</strong><br />
human effort, creativity and engagement in developing and shaping<br />
the digital world. (Aside: In personal communication with<br />
other researchers’ they approved this diagnosis, but it still is – in<br />
my opinion - so dramatic that more effort should be done in researching<br />
this finding. See also section 4)<br />
Based on the above description, what can be done to foster interest<br />
for this group <strong>of</strong> outsi<strong>der</strong>s? And, moreover, how can this be<br />
done without deterring novices who are in the group <strong>of</strong> insi<strong>der</strong>s,<br />
and are already interested in topics like programming?<br />
In or<strong>der</strong> to find the answer, we need to explore the above outlined<br />
model a little further. First <strong>of</strong> all, both groups (Insi<strong>der</strong>s and Outsi<strong>der</strong>s)<br />
start as inexperienced users, e.g. as gamers or web surfers.<br />
But somehow several (the Insi<strong>der</strong>s) gain confidence, interest and<br />
are motivated to explore digital artifacts further, and eventually<br />
discover the possibilities <strong>of</strong> designing by adapting, configuring or<br />
developing web pages and by programming. In contrast the Outsi<strong>der</strong>s<br />
experience that they barely cope with the affordances <strong>of</strong><br />
everyday interaction with digital devices. The reason for this<br />
difficulty might be due to attributing an irrational behavior to<br />
digital devices, and/or a lack <strong>of</strong> specific ‘digital skills’ – as if only<br />
people with special abilities are able to persuade digital artifacts to<br />
do as they are supposed to.<br />
Interesting is the anthropomorphism in attributing ‘irrational<br />
behavior’ to digital devices. It shows a lack <strong>of</strong> knowledge <strong>of</strong> the<br />
internal principles behind the perceived functions; and this<br />
knowledge gap is filled by construing anthropomorphisms in the<br />
conception <strong>of</strong> digital devices; in the hope <strong>of</strong> being able to un<strong>der</strong>stand<br />
one’s own usage problems.<br />
And here we find two handles to foster interest: First, there is still<br />
some interest in un<strong>der</strong>standing ‘what’s going on’ internally. This<br />
should be used and deepened. Second, there is some immediate<br />
need to un<strong>der</strong>stand the structure <strong>of</strong> digital devices.
3. DUAL NATURE CONCEPT<br />
The concept (and term) ‘structure’ is borrowed from the philosophy<br />
<strong>of</strong> technology (see [31] for details). Due to the dual nature <strong>of</strong><br />
technical artifacts, there is the need to un<strong>der</strong>stand the function <strong>of</strong> a<br />
technical artifact as well as its internal design, the structure (structure),<br />
in or<strong>der</strong> to fully un<strong>der</strong>stand it.<br />
Function captures the perspective <strong>of</strong> the use, and the purpose <strong>of</strong><br />
the artifact: What should it be used for? What can be done with it?<br />
structure describes the internal mechanics: How it is made, how it<br />
works and which concepts are used in its (internal) design.<br />
Note, that ‘structure’ – in difference to the use <strong>of</strong> this term in<br />
computing – includes (data) structures as well as algorithms and<br />
processes.<br />
The need to un<strong>der</strong>stand the internal ‘mechanics’ or ‘principles’<br />
un<strong>der</strong>lying the perceived function (functionality) is not restricted<br />
to use. It is also acknowledged as a problem for novices learning<br />
programming, in [3], [23], and [34] similar dual conceptions are<br />
proposed. From now on and the rest <strong>of</strong> the article we use the terms<br />
‘structure’ and ‘function’ to refer to this philosophical meaning.<br />
The concept proposed here is to uncover structure behind function<br />
(by experimenting); and thus bridging the perceived gap between<br />
function and structure for Outsi<strong>der</strong>s. But at the same time, structure<br />
is still in focus, so that such teaching units should be <strong>of</strong> interest<br />
and motivating for Insi<strong>der</strong>s, too.<br />
The key idea is the perception <strong>of</strong> the dual nature <strong>of</strong> digital artifacts,<br />
which can be described by structure and function. Function<br />
is implemented by structure, and structure is designed with a<br />
purpose in mind; hence both are closely connected. Ideally, there<br />
is a ‘harmony’ between structure and function.<br />
Based on structure-knowledge, it should be easier to predict (and<br />
expect) certain functionalities in digital artifacts, to predict and<br />
expect certain ruptures, and to un<strong>der</strong>stand and memorize tactics<br />
and strategies in the use <strong>of</strong> digital artifacts (for such examples, see<br />
e.g. [3]).<br />
Vice-versa, based on function-knowledge, it should be easier to<br />
evaluate such an artifact. If a person truly un<strong>der</strong>stands the purpose<br />
<strong>of</strong> an artifact, she is better able to assess whether the chosen and<br />
implemented structure (the technical ‘solution’) fits the user expectations.<br />
Without such an integrated un<strong>der</strong>standing one is only<br />
able to state being unable to carry out the desired purpose with the<br />
tool – but not whether such usage problems are due to features <strong>of</strong><br />
the structure.<br />
Of course, this duality is an analytical concept; in ‘real life’ there<br />
exist only ‘complete’ digital artifacts – but the duality highlights<br />
typical pattern <strong>of</strong> perceiving and un<strong>der</strong>standing such artifacts.<br />
A last issue is the notion that immediate ‘jumping into’ discussing<br />
structure by a teacher might slow down the learning process,<br />
because learners are missing important aspects (namely the function)<br />
needed to really un<strong>der</strong>stand and memorize the teaching<br />
content. Even worse, if they don’t grasp the function related to<br />
structure, they aren’t able to see the usefulness <strong>of</strong> the learning<br />
content at all.<br />
3.1 Discussion <strong>of</strong> possible misconceptions<br />
In this section some important aspects <strong>of</strong> the concept are rephrased,<br />
in or<strong>der</strong> to discuss some misconception the author noticed<br />
when presenting the concept.<br />
These misconceptions might be due to the education <strong>of</strong> the typical<br />
audience <strong>of</strong> computing engineers. As Kroes [19] discuss, engineering<br />
education can be interpreted as learning to bridge struc-<br />
46<br />
ture and function – so to speak automatically and unconsciously.<br />
Therefore engineers are trained to overlook the gap and therefore<br />
are likely to regard the concept as not really important, as the<br />
main problems <strong>of</strong> the engineer are to refine and produce a suitable<br />
technical solution (the structure) for a given description <strong>of</strong> the<br />
function.<br />
1) The first misconception is therefore the notion that duality<br />
refers to the gap between the outside and the inside: The<br />
function is perceived by the (external) user, whereas the<br />
structure is in focus <strong>of</strong> the engineer, who focusses on the inside.<br />
In fact, the duality can be observed for or within every<br />
part <strong>of</strong> a digital system. Think e.g. about abstraction and the<br />
layered model <strong>of</strong> network protocols. Each layer is used for<br />
the next one, and itself hides the internal structure. Another<br />
example is abstract data types: Again, we see abstraction<br />
from structure, so that the engineer can focus on using the<br />
data type. (Note, however, that e.g. the java library gives<br />
hints about the internal structure when naming classes like<br />
ArrayList or LinkedList – so that the user <strong>of</strong> the class has<br />
some information whether it is better to use ArrayList <strong>of</strong><br />
LinkedList, also the functionality is the same)<br />
2) The second misconception, related to the first one is that<br />
people have problems declaring an aspect as structure, and<br />
another as function. It seems easy when looking at inside/outside<br />
the whole system (see misconception one). But<br />
is an ArrayList a function or a structure? The answer is – <strong>of</strong><br />
course – that it is both. Every part <strong>of</strong> a digital artifact as well<br />
as the whole can be perceived from either the viewpoint <strong>of</strong><br />
function, or from the viewpoint <strong>of</strong> structure. Because the artifact(s)<br />
embody a dual nature: The need to be un<strong>der</strong>stood in<br />
terms <strong>of</strong> function, <strong>of</strong> structure and in terms <strong>of</strong> their dual nature,<br />
there specific link between both sides. (The crucial observation<br />
from philosophy <strong>of</strong> technology is that a) it is quite<br />
hard to see both sides <strong>of</strong> the coin at the same time and b) that<br />
the two sides are based on quite different ontological and<br />
epistemological grounds)<br />
3) Misconceptions regarding structure: structure seen as in data<br />
structure, and missing e.g. algorithms or any other dynamic<br />
aspect <strong>of</strong> the implementation. Of course, structure is not a<br />
term from computing (as it is used here), but from technical<br />
philosophy and as such encompasses the mentioned aspects.<br />
A similar term used in computing might be mechanism.<br />
4) Misconception regarding function: as function as it is used in<br />
computing: the functionality, or as the use <strong>of</strong> the artifact, its<br />
design goals, requirements. To refer to the example used<br />
above, the dichotomy was mechanisms and goals. Similarly<br />
intention could be used. However, such an impression is too<br />
narrow, as function also includes the impact which was not<br />
intended (or as an engineer might say: mal-function).<br />
5) Another misconception might be triggered by the use <strong>of</strong> the<br />
terms insi<strong>der</strong> and outsi<strong>der</strong>, which associate the superiority or<br />
preference for insi<strong>der</strong>s. While it is a legitimate goal to foster<br />
interest for computing, the main goal <strong>of</strong> this approach is fostering<br />
self-determination - which includes <strong>of</strong> course the legitimate<br />
decision not to study computing. But such decisions, as<br />
well as a so-called Outsi<strong>der</strong>s attitude towards digital artifacts<br />
should be based on a somewhat truthful perception <strong>of</strong> the<br />
discipline and the nature <strong>of</strong> digital artifacts.<br />
4. Changing computing education<br />
In the end, this approach strives for a change in computing education<br />
on the school level. As can be seen in natural science or math,
the educational programs and rationales are strongly linked to the<br />
academic disciplines, but nevertheless are having a strong selfview<br />
<strong>of</strong> their unique role and importance for secondary education.<br />
(However, rea<strong>der</strong>s interested only in the actual method might skip<br />
this section. It gives an additional explanation and motivation for<br />
experiments as a method and why they are suggested in this specific<br />
form)<br />
So far, the role <strong>of</strong> computing in school is still somewhat unclear.<br />
Based on the discussion above – the duality <strong>of</strong> structure and function,<br />
and associated perceptions <strong>of</strong> a dichotomy between user and<br />
designer, and <strong>of</strong> Insi<strong>der</strong>-Outsi<strong>der</strong> – we define this role as supporting<br />
learners to develop along this continuum from Users~Outsi<strong>der</strong>s~function<br />
to Designers~Insi<strong>der</strong>s~duality. Learners<br />
should explore different steps between use and design. (And,<br />
experiments are likely to be one <strong>of</strong> the teaching methods supporting<br />
this development.)<br />
Interestingly, we can find several accounts within our discipline<br />
which – at least partially – refer to similar goals. Some <strong>of</strong> them<br />
will be briefly outlined within our framework <strong>of</strong> duality.<br />
First <strong>of</strong> all, the abstraction and theorizing <strong>of</strong> the work in computing<br />
biographies suggest a pattern <strong>of</strong> four different roles, which<br />
can be used to describe different biographical stages and development<br />
processes. The following is based on the thesis <strong>of</strong><br />
<strong>Knobelsdorf</strong> [17] p. 135ff. The empirically and theoretically<br />
<strong>der</strong>ived four biographical steps (or roles), are:<br />
1) checking out or trial; 2) use or apply; 3) configure or modify;<br />
and 4) create or produce.<br />
1. Trial: Focus is on getting to know the digital artifact (DA),<br />
trying out typical applications. Often motivated by curiosity.<br />
DA is perceived in a playful fashion, as a toy to tinker with.<br />
2. Use: Focus is on function; that is interacting with the DA in a<br />
way that is useful for a certain task or within a given context;<br />
and is therefore motivated by this need (which is external or<br />
unrelated to the artifact itself). The DA is consequently perceived<br />
as a medium, a communication device or a tool.<br />
3. Configure: Focus is on adapting the DA to the user's need, by<br />
changing parts <strong>of</strong> the hardware or s<strong>of</strong>tware. Motivation is<br />
based either on the desire to adapt the system to the individual<br />
need, or by coping with errors. Corresponding to these different<br />
motivations, the DA can be perceived as mysterious<br />
and incomprehensible, or as a kind <strong>of</strong> magical artifact with<br />
potentially unlimited possibilities.<br />
4. Create: Focus is on extending the DA by individually or selfbuilt<br />
items. Motivated by a need induced from use-context,<br />
or by the desire or joy <strong>of</strong> being creative. DA is perceived as a<br />
tool for creativity (a computer scientist might say, as a universal<br />
machine).<br />
Note, together with a change <strong>of</strong> the interaction pattern, the perception<br />
<strong>of</strong> the artifact changes as well; but this correlation is not<br />
causality. However, it seems possible that inducing a change in<br />
interaction might cause a change in perception, too.<br />
The second related area is design / end user programming. For<br />
example Fischer and Giaccardi [12] propose Meta-Design. Metadesign<br />
recommends a different role <strong>of</strong> design and designers,<br />
because at design-time not all user issues and needs occurring at<br />
use-time could be foreseen and fulfilled in a suitable manner.<br />
Therefor users need to be designers, too – at least in part. They<br />
also describe a continuum between end user usage and pr<strong>of</strong>essional<br />
design: 1) Passive consumer, 2) well-informed consumer,<br />
3) end-user, 4) power user, 5) domain designer, 6) meta-designer<br />
47<br />
[12]. Note this continuum (again) develops as an increase <strong>of</strong><br />
awareness, un<strong>der</strong>standing, and changing structure.<br />
They also argue: “Cultures are substantially defined by their<br />
media and their tools for thinking, working, learning, and collaborating.<br />
[…] The importance <strong>of</strong> meta-design rests on the fundamental<br />
belief that humans (not all <strong>of</strong> them, not at all times, not in all<br />
contexts) want to be and act as designers in personally meaningful<br />
activities. Meta-design encourages users to be actively engaged in<br />
generating creative extensions to the artifacts given to them and<br />
has the potential to break down the strict counterproductive barriers<br />
between consumers and designers” ([12], section 6.2). Reframing<br />
these arguments in terms <strong>of</strong> computer science education<br />
suggests that it is not only about avoiding usage problems or<br />
having proper tools at hand, but also about being actively engaged,<br />
creative, being able to shape one’s own (immediate) environment,<br />
and being able to develop oneself. The difference here is<br />
that the authors may believe that meta-designers as pr<strong>of</strong>essionals<br />
need an excellent education; while maybe they believe that users<br />
can act as designers without being (digitally) educated. But based<br />
on e.g. the above described biographical results, such a selfeducation<br />
process seems to be possible only for some people.<br />
A third example puts these ideas discussed even further. It was<br />
already postulated in the 1970es by Kay and Goldberg [16]. The<br />
Dynabook should allow people to change the DA they are using,<br />
and this would produce: “a metamedium, whose content would be<br />
a wide range <strong>of</strong> already-existing and not-yet-invented media.”<br />
This requires that the “burden <strong>of</strong> system design and specification<br />
is transferred to the user. This approach will only work if we do a<br />
very careful and comprehensive job <strong>of</strong> providing a general medium<br />
<strong>of</strong> communication which will allow ordinary users to casually<br />
and easily describe their desires for a specific tool.” In terms <strong>of</strong><br />
our framework such a description would mean to create a suitable<br />
idea <strong>of</strong> structure from un<strong>der</strong>standing the need for new or changed<br />
function. Note, this is a design process, and no direct translation<br />
from function to structure would be possible. So while from the<br />
perspective <strong>of</strong> the discipline (or the authors <strong>of</strong> [16] at that time)<br />
the idea was that ‘ordinary users’ will be capable <strong>of</strong> doing so - I<br />
think in light <strong>of</strong> current developments such users would need to be<br />
educated to be able to engage in such interaction with DA. And,<br />
they would need to be educated in CS.<br />
The authors also contrast two types <strong>of</strong> media or tools: Those with<br />
fixed purposes, and those with the ability to be adapted to new<br />
needs and ideas. While the first type, like cars and TV’s aim at<br />
anticipating users’ needs, the other type (like paper and pencil)<br />
“<strong>of</strong>fer[s] many dimensions <strong>of</strong> possibility and high resolution;<br />
these can be used in an unanticipated way by many, though tools<br />
need to be made or obtained to stir some <strong>of</strong> the medium’s possibilities<br />
while constraining others” [16].<br />
While the aforementioned ideas (users as designers) focused on<br />
personal development and self-fulfillment, here now the perspective<br />
is widening towards development <strong>of</strong> the society and culture as<br />
such. The vision is – so to speak– to unleash the full power <strong>of</strong> the<br />
universal machine. From an educational point <strong>of</strong> view, it’s not<br />
only about personal development, but also on participation: being<br />
able to participate in the discourse about advancement <strong>of</strong> society,<br />
and being able to take part in these developments.<br />
Of course there are different degrees <strong>of</strong> such participation – again<br />
from user to – maybe – meta-designer. Nevertheless, an educated<br />
person should be able to un<strong>der</strong>stand what is going on.<br />
While [12] and [16] the role <strong>of</strong> tool development and pr<strong>of</strong>essional<br />
design was consi<strong>der</strong>ed, now the necessity <strong>of</strong> people, and their<br />
education should be consi<strong>der</strong>ed. In summary, the grand vision
ehind the small suggestion to employ experiments in CS education<br />
is meant as a step towards this direction. The educational goal<br />
is to emancipate and educate all, so that they can use and modify<br />
or even create digital artifacts according to their own creativity<br />
and needs. And on a general level, to support well-being and<br />
development <strong>of</strong> society as such, by being a responsible and participating<br />
member <strong>of</strong> (the digital) society.<br />
4.1 Goals<br />
In summary, the duality <strong>of</strong> function and structure is an analytical<br />
perspective that should frame the perception <strong>of</strong> topics to learn for<br />
teachers and learners. It highlights some aspects, and is therefore<br />
connected to certain goals:<br />
Link computer science to ICT-experiences from everyday<br />
life.<br />
Foster interest and motivation.<br />
Foster integration in long term memory by linking new content<br />
(mainly from structure) to already known topics from<br />
everyday life (mainly from function).<br />
Develop / teach competencies to analyze different kinds <strong>of</strong><br />
digital artifacts (and thus foster self-efficacy and maybe<br />
change attribution).<br />
Demonstrate the relevance <strong>of</strong> CS for future development; in<br />
personal life as well as with regard to the wellbeing <strong>of</strong> society.<br />
This has three facets:<br />
Getting help for usage problems (reattribution, get problem<br />
solving ideas based on structure knowledge) and<br />
hence acknowledge the (immediate) value <strong>of</strong> the learning<br />
content,<br />
Un<strong>der</strong>standing the internal mechanics <strong>of</strong> digital artifacts,<br />
as a) a prerequisite or aspect <strong>of</strong> learning programming<br />
as well as b) an aspect <strong>of</strong> un<strong>der</strong>standing the<br />
dual nature <strong>of</strong> digital artifacts.<br />
Un<strong>der</strong>standing the need (and role) <strong>of</strong> design (structure);<br />
and therefore be able to alter or widen conceptions <strong>of</strong><br />
computing disciplines and the role <strong>of</strong> computing pr<strong>of</strong>essionals.<br />
Teaching and learning methods are needed, that provide bridges<br />
between structure and function. Experiments - un<strong>der</strong>stood similarly<br />
to science education - are promising candidates to build such<br />
bridges in the CS classroom. In science, experiments are rigorous<br />
methods developed to test (and sometimes develop) hypotheses<br />
about phenomena in nature. They are different from the everyday<br />
meaning in which experimenting is <strong>of</strong>ten associated with tinkering<br />
or trial and error. In the next section, examples <strong>of</strong> experiments as<br />
bridges between structure and function are presented and briefly<br />
discussed.<br />
5. DUALITY EXPERIMENTS<br />
In this section some examples are given, which show a variety <strong>of</strong><br />
experiments as teaching method in CS education. In doing so, the<br />
rationale behind and the desired effects are discussed.<br />
Note, however, that the examples are more or less based on literature.<br />
All <strong>of</strong> them were enacted, at least in basic form - but not yet<br />
empirically evaluated within the above outlined approach; therefore<br />
the intended effects are described on account <strong>of</strong> the described<br />
theory, and are not empirical results.<br />
5.1 The cell phone network<br />
Using cell phones / smart-phones is an everyday activity <strong>of</strong> school<br />
children. E.g. in <strong>Germany</strong> 96% <strong>of</strong> all teenagers (age 12-19) own a<br />
mobile phone [26].<br />
48<br />
The teaching unit focusses on location based data produced by<br />
mobile phones. In a first step, students are uncovering structure<br />
and function <strong>of</strong> the cellular network.<br />
A small experiment can be used as a motivating introduction[30]:<br />
Two cell phones and a metal box are needed. In the first experiment<br />
one phone is placed inside the box, which is closed. What<br />
happens if one tries to make a phone call to that phone? It<br />
wouldn’t ring because it is not available – no connection from the<br />
other phone is possible. In the second version <strong>of</strong> this experiment,<br />
the second phone is – after dialing the number <strong>of</strong> the first – also<br />
put inside the box. What happens now?<br />
This starting point raises the question how the cell phone network<br />
works – what is its structure? A part <strong>of</strong> it is the need to localize<br />
the mobile stations (the cell phones). And therefore, in the example<br />
above, both phones in the metal box cannot connect to the<br />
base station, and the called phone wouldn’t ring.<br />
In a second step – which also may be called an experiment, but<br />
more correctly is a role play – learners encounter the basic structure<br />
<strong>of</strong> a cellular network (mobile stations, base stations, and the<br />
central switching station with home location register) by enacting<br />
(or experimenting with) use cases like connecting to another<br />
phone. By enacting several scenarios learners will discover that<br />
besides the content data (what they are talking over the phone)<br />
also traffic data is transmitted and collected. Of these traffic data<br />
the location data is in focus <strong>of</strong> the following step.<br />
In [4] an example <strong>of</strong> location data collected by the phone service<br />
provi<strong>der</strong> is shown, including the raw data as a spreadsheet, an<br />
interactive visualization and an interpretation <strong>of</strong> the data.<br />
The learners will visualize the location data themselves, by using<br />
a predefined framework. This framework allows inserting a<br />
graphical representation (e.g. a dot) at the geographical coordinates<br />
in one <strong>of</strong> the typical online maps (like OpenStreetMap).<br />
Experiments with the data are made by changing the visualization,<br />
e.g. adding a counter for each position and/or filtering for time<br />
slots. For example, showing the position <strong>of</strong> the cell phone only<br />
during night time indicates the residence <strong>of</strong> the cell phone owner.<br />
These activities (producing different visualizations <strong>of</strong> the same<br />
data) demonstrate the information that can be revealed, and what<br />
possible (future) use <strong>of</strong> that information may be possible. This<br />
includes a demonstration and discussion <strong>of</strong> location based services<br />
like Google traffic, which predicts traffic conditions, based on<br />
accumulated location data <strong>of</strong> cell phone users [27].<br />
Students can learn – among other aspects – how computing is<br />
changing and devolving, and eventually reflect on their conception<br />
<strong>of</strong> computing disciplines.<br />
In summary, the above outlined experiment (or, if you will experiments<br />
(box, role play, and interactive visualization)) is aiming at<br />
visualizing structure, namely effects <strong>of</strong> the receiving signal<br />
strength, important parts <strong>of</strong> the cell phone network, and location<br />
data. The main role <strong>of</strong> the experiment is to build the above mentioned<br />
bridge from use experiences to perceiving the structure <strong>of</strong><br />
the cell phone network. It should do so by making the learners<br />
curious and start to ask, what will happen and how does it work.<br />
In other words, it should raise awareness for aspects <strong>of</strong> the otherwise<br />
hidden structure <strong>of</strong> the cell phone network. The playful<br />
approach should be motivating by making learners’ curious and<br />
asking themselves, what will happen.<br />
The immediate benefit for users is small, however. It may only a<br />
better un<strong>der</strong>standing <strong>of</strong> connection errors. In addition, the experiments<br />
might raise the awareness <strong>of</strong> differences with regard to the<br />
precision <strong>of</strong> one’s own location data, and that it can be influenced
y choice <strong>of</strong> handset, choice <strong>of</strong> network provi<strong>der</strong>, and <strong>of</strong> course<br />
by the individual pattern (also named habits, see [31]) in using the<br />
mobile phone.<br />
5.2 Algorithms in text-processing<br />
Typically text processing is associated with ICT, and with teaching<br />
usage. Sometimes, e.g. the object oriented analysis <strong>of</strong> text<br />
processing is part <strong>of</strong> teaching Computer science, but predominantly<br />
only function is analyzed (see e.g. [15]). But text processing<br />
can also be used to demonstrate structure so that such knowledge<br />
helps un<strong>der</strong>standing the sometimes ‘irrational’ or ‘unpredictable’<br />
behavior <strong>of</strong> text layout. A typical question <strong>of</strong> users might be:<br />
“Why aren’t my pictures where I wanted them to be?”<br />
In or<strong>der</strong> to answer such a question, a course can start with the<br />
historical example <strong>of</strong> Donald Knuth’s endeavor to capture the art<br />
<strong>of</strong> computer programming, and his task to develop TeX in or<strong>der</strong> to<br />
be able to present it in a suitable layout.<br />
One issue involved in developing such a text processing system is<br />
line breaking. Paragraphs can be automatically aligned, too. We<br />
are consi<strong>der</strong>ing justification (as in this article) as the desired style<br />
for paragraph alignment.<br />
In its simplest form, full justification is ren<strong>der</strong>ed by filling a line<br />
word by word until a word hasn't enough space, and then the<br />
spaces between words are widened until the line is in full justification<br />
(first-fit algorithm). A more advanced algorithm allows<br />
white space between words not only to grow, but also to shrink<br />
(best-fit algorithm).<br />
Here different line breaks are possible, and thus different possibilities<br />
have to be taken into account and evaluated in or<strong>der</strong> to decide<br />
whether shrinking or growing spaces leads to a better full justification.<br />
The most advanced algorithm takes into account not only<br />
the current line, but the whole paragraph. E.g. by omitting a local<br />
maximum the global maximum for the paragraph can be optimized.<br />
This is the total-fit algorithm developed together with TeX<br />
by Donald Knuth and Michael Plass [18].<br />
Thinking about these possibilities alone can help students to be<br />
able to perceive the structure <strong>of</strong> text processing. To do this, an<br />
experiment can be conducted in which students try to figure out<br />
whether the local text processor (like e.g. Open Office or Micros<strong>of</strong>t<br />
Word) uses total-fit.<br />
If so, within a paragraph line breaks can change, if the last line <strong>of</strong><br />
the paragraph is altered. This should be observable in WYSIWYG<br />
(“what you see is what you get”) text processors. Usually – at least<br />
in our experiments with students – they don’t see such changes, so<br />
the conclusion is that mo<strong>der</strong>n WYSIWIG processors do not use<br />
the best algorithm available. However, maybe the data used in the<br />
experiment didn’t reveal any changes, but with other text input it<br />
would. On the other hands students might argue that changing line<br />
breaks in already finished lines while typing would be confusing.<br />
A web search can be added to check if producers <strong>of</strong> word processors<br />
provide any information about the algorithms used.<br />
However, the main learning goal <strong>of</strong> the experiment is not to figure<br />
out which algorithm is used, but to un<strong>der</strong>stand that algorithms are<br />
used, and that there are differences, and a lot <strong>of</strong> things happening<br />
while the function experience suggests that simply characters<br />
from the keyboard are ‘typed onto the screen’.<br />
Does this help to better use text processing and e.g. prevent pictures<br />
from being placed by random within a document? Short<br />
answer: No (A longer answer will follow below). Is CS Education<br />
(for Outsi<strong>der</strong>s) therefore useless? Again: No. Let’s compare this<br />
to science: What is the ‘use’ <strong>of</strong> knowledge about the solar sys-<br />
49<br />
tems, and planets circling? A sunset is still a sunset – yet many<br />
people would agree that knowledge about the world is worthwhile<br />
in its own. Similarly, bridging function to structure enables learners<br />
to perceive structure, to get a better perception and un<strong>der</strong>standing<br />
<strong>of</strong> the digital world.<br />
In summary, there are three intended effects <strong>of</strong> examples like this:<br />
First, some learners should become intrigued to explore this world<br />
further and e.g. learn what an algorithm is in detail.<br />
Second, this un<strong>der</strong>standing should be transferred to other aspects<br />
<strong>of</strong> word processing, namely the example mentioned at the beginning:<br />
Why can it be that pictures are seemingly moved randomly<br />
around the text? A closer look here reveals that (usually) the user<br />
has the option to change which algorithm is used to place the<br />
picture on the page. So in a specific manner, this example might<br />
also reduce difficulties in using.<br />
Third, self-efficacy should be increased. As this experiment introduces<br />
algorithms, but in the context <strong>of</strong> everyday use <strong>of</strong> ICT, it<br />
builds a bridge to structure: Outsi<strong>der</strong>s should experience and<br />
notice that they have the ability to un<strong>der</strong>stand and gather such<br />
structure-knowledge.<br />
5.3 Search engine results<br />
In this example, the notion <strong>of</strong> ‘personalized web search’ is experimentally<br />
analyzed, using a method based on [11]. The question is<br />
how the result-lists <strong>of</strong> Google are compiled, especially whether<br />
everybody gets the same results.<br />
Freuz et al. [11] designed an experiment to analyze this. They<br />
opened three Google accounts (Kant, Nietzsche, and Foucault),<br />
and performed individually different searches, and so allowed<br />
Google to collect a search history <strong>of</strong> the accounts.<br />
Afterwards they compared results <strong>of</strong> the three accounts to some<br />
selected common searches. It became clear that within the first<br />
page <strong>of</strong> presented results, the three accounts got different answers<br />
to the same search term.<br />
The interesting aspect here is the methodology used: Focusing on<br />
the user experience (function), a scientific experiment is devised:<br />
Using a hypothesis, a methodological setup, and the gathering <strong>of</strong><br />
data which has to be analyzed and interpreted. The remarkable<br />
thing about this experiment is, that mainly operating within the<br />
usual user paradigm (within the framework <strong>of</strong> function), the experiment<br />
reveals novel insights in the function but at the same<br />
time also insights in the internal mechanics (structure).<br />
Students can learn that by engaging in thorough observation it is<br />
possible to infer structural aspects relying on normal usage only.<br />
6. DISCUSSION<br />
In science teaching, experiments are a well-known teaching method.<br />
E.g. science labs <strong>of</strong>fer courses for interested schools, where<br />
learners can experience interesting experiments. Although quite<br />
successful, such activities are sometimes labeled as “hands-on,<br />
minds <strong>of</strong>f”-approaches.<br />
While these activities - that’s the essence <strong>of</strong> the saying - are very<br />
good in triggering situational interest, as they provide most <strong>of</strong>ten a<br />
quite spectacular experience, they are less good in developing and<br />
maintaining individual interest.<br />
In research on interest, situational and individual interest is differentiated.<br />
Situational interest is conceptualized as being due to<br />
external causes, by catch facets like e.g. new approach to teaching<br />
or an interesting puzzle to solve [25]. Thus, the teacher can catch<br />
the attention, and situational interest is triggered. Individual inter-
est is developed by repeated experience <strong>of</strong> situational interest.<br />
According to Mitchell [25], meaningfulness and involvement are<br />
the two most important aspects to foster individual interest. Meaningfulness<br />
is the perception <strong>of</strong> a learner <strong>of</strong> the content as being<br />
valuable for her current life. Involvement is due to active engagement<br />
in the learning process.<br />
Based on this theoretical account, the following aspects are important<br />
affordances for developing and analyzing experiments as<br />
interest developing approach to teaching:<br />
If learners feel or believe they can influence the process (e.g. the<br />
learning process, or the process <strong>of</strong> interacting with a digital artifact)<br />
they are more likely to be involved.<br />
Teachers and learners <strong>of</strong>ten have different perceptions <strong>of</strong> what is<br />
<strong>of</strong> most importance in the learning process. How can Outsi<strong>der</strong>s be<br />
enabled or supported to see value (=meaningfulness) in learning<br />
computer science?<br />
Self-efficacy (or similar: self-confidence, self-view) should be<br />
preserved in or<strong>der</strong> to prevent learners from avoiding situations<br />
which are threatening to their self-esteem. And, vice-versa, learners<br />
should have opportunities to succeed, because “as achievement<br />
enhances self-image and confidence in an upward spiral<br />
in which increased levels <strong>of</strong> achievement enhance motivation<br />
which in turn leads to further increases in achievement” [21].<br />
The thesis is that building bridges between function and structure<br />
supports the perception <strong>of</strong> the value in learning structure (and<br />
hence CS, too).<br />
There are two issues connected to experiments: One is that such<br />
activities are interesting due to external factors (hands on), which<br />
are not transferred to the discipline. The other is that embedding<br />
the disciplines’ content in engaging activities can be seen only as<br />
one first step, and in addition to involvement also meaningfulness<br />
needs to be supported. And here we are talking about meaningfulness<br />
for Outsi<strong>der</strong>s, who are most likely are conceptualizing CS as<br />
related to computers and a kind <strong>of</strong> mysterious pr<strong>of</strong>essional pattern<br />
in interaction with such devices.<br />
Thus the main point is to shift attention from function to structure,<br />
while also preserving the link to the original function and context<br />
<strong>of</strong> an Outsi<strong>der</strong> as a user. When and how will computing become<br />
meaningful and valuable for such a person?<br />
In other words, the above described activities might be labeled as<br />
experiments, but only if they are building a bridge between function<br />
and structure. Most likely such a bridge can be built by starting<br />
from everyday use experiences with digital artifacts.<br />
Here we also see a difference to e.g. CS unplugged. Both approaches<br />
have some commonalities: They are aiming at triggering<br />
situational interest (<strong>of</strong> Outsi<strong>der</strong>s) by engaging them in unusual<br />
and intriguing activities, and both shift attention away from using<br />
the PC. The difference is that CS unplugged simply removes the<br />
distracting content and plugs out the PC, duality experiments start<br />
with use experiences and some kind <strong>of</strong> digital artifact from an<br />
unusual perspective that should raise curiosity in the usually<br />
hidden structural aspect. In terms <strong>of</strong> the briefly outlined theory <strong>of</strong><br />
interest, CS unplugged provides a good catch facet for situational<br />
interest, while a duality experiment aims at providing some hold<br />
facet in or<strong>der</strong> to develop individual interest by raising the impression<br />
<strong>of</strong> personal meaningfulness and value.<br />
From the above discussed experiments some conclusions can be<br />
drawn, in or<strong>der</strong> to highlight specific features and intended effects<br />
<strong>of</strong> experiments in computer science education.<br />
50<br />
Firstly, they do not aim to (immediately) change the perception <strong>of</strong><br />
the discipline, but to change the perception <strong>of</strong> digital artifacts, and<br />
the perception <strong>of</strong> suitable interaction patterns with digital artifacts.<br />
They introduce learners to the ‘internal mechanics’ (structure) <strong>of</strong><br />
digital artifacts so that they can reattribute the causes for usage<br />
problems.<br />
They should raise awareness for possible variants <strong>of</strong> the internal<br />
mechanics, so that learners can experience and explore a ‘design<br />
space’, in which one can un<strong>der</strong>stand that (within constraints)<br />
different structure-variants can lead to (sometimes subtle but<br />
important) differences in function.(see e.g. example in 5.2 ).<br />
Experiments should take into account different usage approaches,<br />
so to include different levels <strong>of</strong> use-competence and habits for<br />
interaction with a digital artifact (see e.g. 5.1 ).<br />
In or<strong>der</strong> to support a general change, and not only a local change<br />
in interacting with the currently analyzed digital artifact, transfer<br />
should be included; e.g. from the actual experiment to other digital<br />
artifacts, or other aspects <strong>of</strong> the same digital artifact (like e.g.<br />
from ‘text alignment’ to ‘picture alignment’ in section 5.2 ).<br />
7. CONCLUSIONS<br />
Overall, the examples described here are based on the idea <strong>of</strong><br />
duality reconstruction [31]. Nevertheless empirical research is<br />
needed to discern what in detail do students learn from such experiments.<br />
Also empirical data could help to un<strong>der</strong>stand which<br />
experiments really foster bridging structure and function and<br />
support learners to see the dual nature <strong>of</strong> digital artifacts. And <strong>of</strong><br />
course, it will be interesting to conduct empirical research on the<br />
question if learners change with regard to self-efficacy, motivation<br />
or interest as predicted.<br />
Structure and function are somewhat connected to Insi<strong>der</strong> (-<br />
>structure) and Outsi<strong>der</strong> (->function). Providing bridges should<br />
support the construction <strong>of</strong> learning-units that are appealing to<br />
both groups and their different interests (and perspectives on the<br />
topics).<br />
And, most important, the experiment should provide opportunities<br />
to perceive structure and function as essential. Especially using<br />
bridges between structure and function should increase the sense<br />
<strong>of</strong> value as it mutually supports the perspective that is not in focus<br />
<strong>of</strong> the learner.<br />
Questions yet to solve arise in the following aspects:<br />
The experiments outlined above allow learners to control parts <strong>of</strong><br />
the experiments. The actual interaction with an artifact as part <strong>of</strong><br />
the experiment usually can vary and is not as strictly defined as in<br />
‘real’ scientific experiments. More work should be done to discern<br />
important aspects <strong>of</strong> the teaching method, including e.g. link to<br />
learning theory and interaction <strong>of</strong> outcome and learner attributes.<br />
The above described experiments are fulfilling more or less the<br />
just mentioned affordances. In or<strong>der</strong> to further develop the teaching<br />
method, our discipline can learn from experiments in natural<br />
science - both positively and negatively. For example, we know<br />
from science education that it is possible to engage in contraproductive<br />
trial-and error, so called ‘Hands-on, minds-<strong>of</strong>f’ experiments.<br />
It is also an open question whether the intended transfer <strong>of</strong> a<br />
changed perception in the interaction with DA triggers a change in<br />
perception <strong>of</strong> the computing disciplines, too. Remember e.g. the<br />
digital caretaker, discussed in section 2.2 . Such a misconception<br />
<strong>of</strong> the discipline should be changed – but will it?<br />
These questions are answerable only by (empirical) research.
8. REFERENCES<br />
[1] Bell, T. et al. 2011. Overcoming obstacles to CS education<br />
by using non-programming outreach programmes. Proceedings<br />
<strong>of</strong> the 5th international conference on Informatics<br />
in Schools: situation, Evolution and Perspectives (Berlin,<br />
Heidelberg, 2011), 71–81.<br />
[2] Bell, T. and Newton, H. 2013. Unplugging Computer<br />
Science. Improving Computer Science Education. D.<br />
Kadijevich et al., eds. Routledge. 66–82.<br />
[3] Bhavnani, S.K. et al. 2008. Strategy-Based Instruction:<br />
Lessons Learned in Teaching the Effective and Efficient<br />
Use <strong>of</strong> Computer Applications. ACM Trans. Comput.-<br />
Hum. Interact. 15, 1 (May. 2008), 2:1–2:43.<br />
[4] Biermann, K. 2011. Data Protection: Betrayed by our own<br />
data. Die Zeit. http://www.zeit.de/digital/datenschutz/2011-<br />
03/data-protection-malte-spitz. Accessed: 2012-01-13<br />
[5] Competition Workshop 2006 - Accepted Papers:<br />
http://www.bwinf.de/competition-workshop/papers.html.<br />
Accessed: 2012-01-13.<br />
[6] Computer Science for Fun - cs4fn: HOME:<br />
http://www.cs4fn.org/. Accessed: 2012-01-13.<br />
[7] Computer Science Unplugged |: http://csunplugged.org/.<br />
Accessed: 2012-01-13.<br />
[8] Crutzen, C.K.M. 2000. Interaction, a world <strong>of</strong> differences.<br />
A vision on informatics from the perspective <strong>of</strong> gen<strong>der</strong><br />
studies. Open Universiteit Ne<strong>der</strong>land.<br />
[9] Curzon, P. et al. 2009. Enthusing & inspiring with reusable<br />
kinaesthetic activities. Proceedings <strong>of</strong> the 14th annual<br />
ACM SIGCSE conference on Innovation and technology in<br />
computer science education (New York, NY, USA, 2009),<br />
94–98.<br />
[10] Feaster, Y. et al. 2011. Teaching CS unplugged in the high<br />
school (with limited success). Proceedings <strong>of</strong> the 16th annual<br />
joint conference on Innovation and technology in<br />
computer science education (New York, NY, USA, 2011),<br />
248–252.<br />
[11] Feuz, Martin et al. 2011. Personal Web searching in the<br />
age <strong>of</strong> semantic capitalism: Diagnosing the mechanisms <strong>of</strong><br />
personalisation. First Monday. 16, 2 (2011).<br />
[12] Fischer, G. and Giaccardi, E. 2004. Meta-design: A<br />
Framework for the Future <strong>of</strong> End-User Development<br />
(Online-Version). End User Development. H. Lieberman et<br />
al., eds. Kluwer Academic Publishers, Dordrecht, The<br />
Netherlands. 1–26.<br />
[13] Guzdial, M. 2003. A media computation course for nonmajors.<br />
Proceedings <strong>of</strong> the 8th annual conference on Innovation<br />
and technology in computer science education (New<br />
York, NY, USA, 2003), 104–108.<br />
[14] Heersink, D. and Moskal, B.M. 2010. Measuring high<br />
school students’ attitudes toward computing. Proceedings<br />
<strong>of</strong> the 41st ACM technical symposium on Computer science<br />
education (New York, NY, USA, 2010), 446–450.<br />
[15] Hubwieser, P. 2006. Functions, Objects and States: Teaching<br />
Informatics in Secondary Schools. Informatics Education<br />
– The Bridge between Using and Un<strong>der</strong>standing Computers.<br />
R.T. Mittermeir, ed. Springer Berlin Heidelberg.<br />
104–116.<br />
[16] Kay, A. and Goldberg, A. 1977. Personal Dynamic Media.<br />
Computer. 10, 3 (1977), 31–41.<br />
51<br />
[17] <strong>Knobelsdorf</strong>, M. 2011. Biographische Lern- und Bildungsprozesse<br />
im Handlungskontext <strong>der</strong> Computernutzung. Freie<br />
Universität Berlin.<br />
[18] Knuth, D.E. and Plass, M.F. 1981. Breaking paragraphs<br />
into lines. S<strong>of</strong>tware: Practice and Experience. 11, 11 (Nov.<br />
1981), 1119–1184.<br />
[19] Kroes, P. 2002. Design methodology and the nature <strong>of</strong><br />
technical artefacts. Design Studies. 23, 3 (May. 2002),<br />
287–302.<br />
[20] Leiviskä, K. and Siponen, M. 2010. Attitudes <strong>of</strong> sixth form<br />
female students toward the IT field. SIGCAS Comput. Soc.<br />
40, 1 (Mar. 2010), 34–49.<br />
[21] Littlejohn, A. 2008. The Tip <strong>of</strong> the Iceberg: Factors Affecting<br />
Learner Motivation. RELC Journal. 39, 2 (Aug. 2008),<br />
214–225.<br />
[22] Maloney, J. et al. 2010. The Scratch Programming Language<br />
and Environment. Trans. Comput. Educ. 10, 4<br />
(2010), 1–15.<br />
[23] Mayer, R.E. 1981. The Psychology <strong>of</strong> How Novices Learn<br />
Computer Programming. ACM Comput. Surv. 13, 1 (1981),<br />
121–141.<br />
[24] McLachlan, C. et al. 2010. Student perceptions <strong>of</strong> ICT: a<br />
gen<strong>der</strong>ed analysis. Proceedings <strong>of</strong> the Twelfth Australasian<br />
Conference on Computing Education - Volume 103 (Darlinghurst,<br />
Australia, Australia, 2010), 127–136.<br />
[25] Mitchell, M. 1993. Situational interest: Its multifaceted<br />
structure in the secondary school mathematics classroom.<br />
Journal <strong>of</strong> Educational Psychology. 85, 3 (1993), 424.<br />
[26] MPFS 2011. JIM-Studie 2011: Jugend, Information, (Multi-)Media<br />
; Basisuntersuchung zum Medienumgang 12-<br />
bis 19jähriger. Medienpädagogischer Forschungsverbund<br />
Südwest.<br />
[27] Official Google Blog: The bright side <strong>of</strong> sitting in traffic:<br />
Crowdsourcing road congestion data:<br />
http://googleblog.blogspot.com/2009/08/bright-side-<strong>of</strong>sitting-in-traffic.html.<br />
Accessed: 2012-01-05.<br />
[28] Papastergiou, M. 2008. Are Computer Science and Information<br />
Technology still masculine fields? High school students’<br />
perceptions and career choices. Computers & Education.<br />
51, 2 (Sep. 2008), 594–608.<br />
[29] Pasternak, A. and Vahrenhold, J. 2010. Braided teaching in<br />
secondary CS education: contexts, continuity, and the role<br />
<strong>of</strong> programming. Proceedings <strong>of</strong> the 41st ACM technical<br />
symposium on Computer science education (New York,<br />
NY, USA, 2010), 204–208.<br />
[30] Redaktion Schulprojekt Mobilfunk 2008. Mobilfunk und<br />
Technik. Fächerübergreifende Sachinformationen für projektorientiertes<br />
Lernen. Klassen 5–10 sowie gymnasiale<br />
Oberstufe. Informationszentrum Mobilfunk e. V.<br />
[31] Schulte, C. 2008. Duality Reconstruction - Teaching Digital<br />
Artifacts from a Socio-technical Perspective. Proceedings<br />
<strong>of</strong> the 3rd international conference on Informatics in<br />
Secondary Schools - Evolution and Perspectives: Informatics<br />
Education - Supporting Computational Thinking (Torun,<br />
Poland, 2008), 110–121.<br />
[32] Schulte, C. and <strong>Knobelsdorf</strong>, M. 2007. Attitudes towards<br />
computer science-computing experiences as a starting<br />
point and barrier to computer science. Proceedings <strong>of</strong> the<br />
third international workshop on Computing education research<br />
(Atlanta, Georgia, USA, 2007), 27–38.
[33] Schulte, C. and Magenheim, J. 2005. Novices’ expectations<br />
and prior knowledge <strong>of</strong> s<strong>of</strong>tware development: results<br />
<strong>of</strong> a study with high school students. ICER 2005: Proceedings<br />
<strong>of</strong> the 1st International Computing Education Research<br />
Workshop; Seattle Washington USA October 1 - 2<br />
2005 ([New York, NY], 2005), 143–153.<br />
[34] Soloway, E. 1986. Learning to program = learning to construct<br />
mechanisms and explanations. Commun. ACM. 29, 9<br />
(1986), 850–858.<br />
[35] Taub, R. et al. 2012. CS Unplugged and Middle-School<br />
Students’ Views, Attitudes, and Intentions Regarding CS.<br />
Trans. Comput. Educ. 12, 2 (Apr. 2012), 8:1–8:29.<br />
[36] Thies, R. and Vahrenhold, J. 2012. Reflections on outreach<br />
programs in CS classes: learning objectives for “unplugged”<br />
activities. Proceedings <strong>of</strong> the 43rd ACM technical<br />
symposium on Computer Science Education (New<br />
York, NY, USA, 2012), 487–492.<br />
52
Agile Projects in High School Computing Education –<br />
Emphasizing a Learners’ Perspective<br />
Ralf Romeike<br />
<strong>University</strong> <strong>of</strong> Potsdam<br />
August-Bebel-Str. 89<br />
14482 Potsdam, <strong>Germany</strong><br />
romeike@cs.uni-potsdam.de<br />
ABSTRACT<br />
S<strong>of</strong>tware projects are seen as a methodology for secondary computing<br />
education which is highly appropriate and meets the demands<br />
and goals <strong>of</strong> Computer Science (CS). Yet the majority <strong>of</strong><br />
models and examples for project-based lessons rely on a traditional<br />
s<strong>of</strong>tware development approach: the waterfall model. In<br />
this paper such models are analyzed for their strength, problems,<br />
and deficiencies. Based on the results <strong>of</strong> the analysis a new approach<br />
to projects in secondary computing education is presented<br />
which uses the concept <strong>of</strong> didactic transposition to adapt agile<br />
s<strong>of</strong>tware development methods for project organization, management,<br />
and implementation in class. The resulting model applies<br />
valuable practices <strong>of</strong> eXtreme Programming and Scrum and provides<br />
a set <strong>of</strong> tools that allow high school s<strong>of</strong>tware projects to<br />
benefit from mo<strong>der</strong>n s<strong>of</strong>tware development methods. By emphasizing<br />
dynamic processes and a clear course <strong>of</strong> action an attractive<br />
perspective on CS is promoted.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers and Education]: Computer Science Education<br />
General Terms<br />
Human Factors, Theory.<br />
Keywords<br />
Secondary computing education, agile methods, project-based<br />
learning.<br />
1. INTRODUCTION<br />
In secondary computing education, s<strong>of</strong>tware projects are promoted<br />
to provide an appropriate and student-oriented approach to<br />
Computer Science (CS) [19, 23, 32, 39]. Yet, most projects in this<br />
context are mainly focused on sequential project layouts that resemble<br />
traditional s<strong>of</strong>tware development (SD) methodologies<br />
such as the waterfall model. In recent years it became apparent in<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that<br />
copies bear this notice and the full citation on the first page. To copy<br />
otherwise, or republish, to post on servers or to redistribute to lists, requires<br />
prior specific permission and/or a fee.<br />
Conference’10, Month 1–2, 2010, City, State, Country.<br />
Copyright 2010 ACM 1-58113-000-0/00/0010…$10.00.<br />
53<br />
Timo Göttel<br />
<strong>University</strong> <strong>of</strong> Hamburg<br />
Vogt-Kölln-Str. 30<br />
22527 Hamburg, <strong>Germany</strong><br />
tgoettel@acm.org<br />
pr<strong>of</strong>essional SE that such methodologies <strong>of</strong>ten fail to produce<br />
high quality products, bring forward delays in delivery, and insufficiently<br />
consi<strong>der</strong> customers’ needs (e.g. [26]). Analogue issues<br />
can be found in school projects: unfinished projects, missing time<br />
and motivation for testing, neglected documentation, and teachers’<br />
difficulties in managing s<strong>of</strong>tware projects are just some <strong>of</strong> the<br />
problems reported (cp. [19, 24, 32]). In mo<strong>der</strong>n SD, agile methods<br />
are promoted to provide a dynamic project management that relies<br />
on interaction and short design iterations. Agile methods build<br />
upon values and provide practices that are also highly expedient<br />
in high school contexts. Therefore, we present a new approach to<br />
projects in secondary computing education, which implements the<br />
theory <strong>of</strong> didactic transposition to adapt agile methods for project<br />
organization, management and implementation in classroom.<br />
Valuable agile practices <strong>of</strong> eXtreme Programming (XP) [2] and<br />
Scrum [39] will provide a set <strong>of</strong> tools allowing s<strong>of</strong>tware projects<br />
in high schools to reference mo<strong>der</strong>n SD by highlighting dynamic<br />
processes that help to focus on good results, a clear course <strong>of</strong><br />
action, and an attractive perspective on pr<strong>of</strong>essional CS by addressing<br />
common problems at the same time. In section 2, research<br />
on projects in computing education will be discussed and<br />
problems with prevalent models will be analyzed. The findings<br />
suggest a missing consi<strong>der</strong>ation <strong>of</strong> the learners’ perspective in<br />
project models. In section 3, agile methods in pr<strong>of</strong>essional and<br />
educational settings are discussed for their potential <strong>of</strong> supporting<br />
a learners’ perspective. By describing the agile model for school<br />
projects in computing education common agile practices are characterized<br />
and adapted. Finally, the model is discussed in the context<br />
<strong>of</strong> existing models, its potential, and issues in computing<br />
education.<br />
2. PROJECTS IN EDUCATION<br />
2.1 Project Based Learning<br />
Project-based learning (PBL) is an approach to teaching and<br />
learning in the classroom aiming for engaging students in explorative<br />
and problem-solving activities in authentic contexts. The<br />
concept is described to originate from teaching in Zurich and<br />
Paris in the 19th century. Since then, the idea has been picked up<br />
by various teachers and researchers and enriched with psychological,<br />
pedagogical and sociopolitical aspects, e.g. by Dewey und<br />
Kilpatrick [11]. Projects are un<strong>der</strong>stood as learning processes that<br />
draw on interests and demands <strong>of</strong> the students by striving for a<br />
complex result, <strong>of</strong>ten a product. This includes planning, problemsolving,<br />
analysis <strong>of</strong> different solutions and the evaluation <strong>of</strong> the<br />
process and its product. PBL is known for increasing students’<br />
motivation, for strengthening self confidence, and for fostering<br />
satisfaction in process and outcome (cp. [5]). Therefore, it is pro-
moted to foster high-level thinking skills including problemsolving<br />
and analysis skills. Thus, it helps to gain a deep un<strong>der</strong>standing<br />
<strong>of</strong> topics and processes (cp. [1, 25]). PBL is known to<br />
encourage peer interaction (cp. [25]).<br />
In <strong>Germany</strong>, Frey [16] elaborated PBL by describing the process<br />
along the following steps: A project starts with a project idea,<br />
which should be based on the interests <strong>of</strong> the students. Subsequently<br />
the idea is specified in or<strong>der</strong> to find agreement on what<br />
the class will be trying to achieve. After planning the necessary<br />
activities the project plan is carried out. Milestones and metacommunication<br />
serve as tools for supporting the process. This<br />
model is commonly accepted within <strong>Germany</strong> and was fundamental<br />
for an adapted model in German computing education (cp. 2.3)<br />
Summarizing, projects are characterized by being problemfocused<br />
and interdisciplinary, by allowing students a choice <strong>of</strong><br />
topic and foster personal responsibility. Projects are generally<br />
carried out over a longer time span and are <strong>of</strong>ten graded “differently”,<br />
e.g. by focusing more on processes and applying less pressure.<br />
2.2 PBL in Secondary Computing Education<br />
Because <strong>of</strong> the above mentioned benefits, PBL is widely used in<br />
higher computing education (cp. [6, 13]) and secondary computing<br />
education, especially in the context <strong>of</strong> s<strong>of</strong>tware projects [39,<br />
41]. Consequently, PBL is recommended as an appropriate approach<br />
to computing education in the majority <strong>of</strong> German curricula.<br />
However, there is limited research focusing on methodologies<br />
for SD projects in secondary education. Meerbaum-Salant and<br />
Hazzan [33] constitute “as far as we know, no general methodology<br />
has been developed for s<strong>of</strong>tware development projects in the<br />
high school“. Consequently, they propose an approach which<br />
focuses on a mentoring model for teachers (cp. 2.4).<br />
In <strong>Germany</strong>, a model for high school s<strong>of</strong>tware projects was proposed<br />
by Frey [15] and elaborated by Schubert and Schwill [39].<br />
The majority <strong>of</strong> published school SD projects can be attributed to<br />
this model. Therefore the model will be analyzed in the following.<br />
2.3 A Pr<strong>of</strong>essional Perspective<br />
A majority <strong>of</strong> publications concerning the use <strong>of</strong> projects in secondary<br />
computing education 1 stems from the 1980s and 1990s,<br />
generally analyzing and adopting the pr<strong>of</strong>essional approach to SD<br />
and adopting the waterfall model for the classroom. The idea was<br />
that a pr<strong>of</strong>essional model for SD would provide an appropriate<br />
framework for school projects: it <strong>of</strong>fers students a structured<br />
learning process and gives insights into pr<strong>of</strong>essional SD processes<br />
at the same time. In comparison to other subjects using PBL only<br />
as a teaching method without a connection to the subject matter<br />
itself, projects are scientifically anchored in CS [39]. Consequently,<br />
there is a large body <strong>of</strong> practice reports on SD projects.<br />
However, by analyzing publications <strong>of</strong> the major German conferences<br />
in computing education <strong>of</strong> the last 10 years we could not<br />
find any publication concerning methodologies for school projects<br />
in general secondary education. However, research shows that<br />
teachers consi<strong>der</strong> it as important that students get familiar with<br />
SD processes, e.g. by running “through the workflows <strong>of</strong> the waterfall<br />
model” [30].<br />
1 Research on this topic is rare. The statement reflects the situation<br />
in <strong>Germany</strong>. However, we are not aware <strong>of</strong> comparable<br />
publications from other countries.<br />
54<br />
Figure 1 illustrates the project steps <strong>of</strong> Schubert and Schwill’s<br />
model [39] with the corresponding output <strong>of</strong> each phase. The<br />
proposed model describes student activities along the s<strong>of</strong>tware life<br />
cycle. However, it does not provide methods or practices <strong>of</strong> how<br />
students can reach the expected outcome <strong>of</strong> each phase. In the<br />
problem analysis phase all important environmental conditions<br />
need to be gathered clearly and completely. Furthermore, this<br />
phase includes planning activities for time, team and equipment.<br />
The resulting requirements definition serves as a contract between<br />
teacher and students. In the subsequent system design process a<br />
model <strong>of</strong> the system is specified by dividing the “overall system”<br />
into modules. Until here, the activities shall be performed by the<br />
full team, which is allowed to split up into subgroups for minor<br />
tasks. Then, smaller groups handle a module each un<strong>der</strong> their own<br />
responsibility. The remaining phases follow the s<strong>of</strong>tware life cycle.<br />
Fig. 1. Project model by Schubert and Schwill [39] .<br />
Concerning the team structure, Schubert and Schwill recognize<br />
that the organization <strong>of</strong> the team cannot follow hierarchical structures<br />
as typically used in pr<strong>of</strong>essional SD teams. Instead, they<br />
suggest equal status and responsibility among team members and<br />
a “force” for communication and common goals. This goal is<br />
addressed by assigning eight different roles for student positions<br />
within the project (computer responsible, project supervisor, interface<br />
responsible, tester, documentation responsible, butler, session<br />
chair, secretary).<br />
Criticism <strong>of</strong> the methodology points out problems with a perceived<br />
bureaucratic overhead: “It is always the same: Students<br />
refuse to first plan on paper. Because only small programs are<br />
written, these are not documented. The taught principles <strong>of</strong> s<strong>of</strong>tware<br />
development are hardly noticed” [19]. Often, testing is omitted<br />
in the project realization due to a lack <strong>of</strong> time and a lack <strong>of</strong><br />
perceived importance, especially if the s<strong>of</strong>tware does not have a<br />
practical use after the project (cp. [24]). Other problems may result<br />
from the structure <strong>of</strong> the project: The time span that needs to<br />
be scheduled is up to half a year (Schubert and Schwill [39] suggest<br />
to perform one project per semester). This is difficult to plan<br />
for, especially if students lack project experience and supportive<br />
practices. Additionally, a sequential project layout collides with<br />
“project-unfriendly” circumstances <strong>of</strong> formal lessons such as lim-
ited time, heterogeneous student abilities and lessons spread over<br />
several weeks. Humbert [23] even summarizes that the pedagogical<br />
dimensions <strong>of</strong> PBL are not sufficiently consi<strong>der</strong>ed in such a<br />
project model. Additional problems <strong>of</strong> conducting school SD<br />
projects are outlined in the following by pursuing a teachers’<br />
perspective.<br />
2.4 A Teachers’ Perspective<br />
Meerbaum-Salant and Hazzan [32] analyzed difficulties encountered<br />
by teachers in mentoring SD projects in Israeli high schools.<br />
Even though the curricular background and objectives are somewhat<br />
different 2 than for projects in general educational settings as<br />
described in this paper, the results are similar to the problems<br />
reported from German teachers. Additional problems were identified<br />
in the contexts <strong>of</strong> scheduling the project, CS expertise <strong>of</strong> the<br />
teachers, consi<strong>der</strong>ing students’ individual performances, and<br />
evaluation <strong>of</strong> the project. Teachers describe mentoring <strong>of</strong> SD<br />
projects as a more complex task compared to traditional teaching.<br />
Meerbaum-Salant and Hazzan [33] express the need for a general<br />
methodology for SD projects in high school and address the previously<br />
identified problems in a mentoring methodology<br />
(ACMM). It is intended to support teachers, who are expected to<br />
be confident in a variety <strong>of</strong> knowledge types [32]. Therefore it<br />
describes a set <strong>of</strong> practices (Pedagogical Class Management Aspect,<br />
Social Aspect, Project Management Aspect) that shall be<br />
consi<strong>der</strong>ed by teachers while mentoring a project. The ACMM<br />
takes into account the principles <strong>of</strong> agile s<strong>of</strong>tware (such as communication,<br />
simplicity, feedback, respect), which basically are<br />
reflected in the teacher-student interaction.<br />
2.5 A Learners’ Perspective<br />
For learning settings, where a team <strong>of</strong> students is working cooperatively<br />
on projects, we see potential for taking the idea <strong>of</strong> applying<br />
agile methods further than it is described in the ACMM: project<br />
management can be done by the student team. This can be<br />
supported by straightforward and easy to use methods adopted<br />
from mo<strong>der</strong>n SD. Additionally, students may benefit from experiencing<br />
a SD process which also includes management aspects in<br />
addition to activities like analysis, designing, coding, and testing,<br />
as described in the other models. In the following we demonstrate<br />
how problems identified by Meerbaum-Salant and Hazzan [32]<br />
may be addressed in such a project by consi<strong>der</strong>ing agile methods<br />
as presented in section 3.<br />
Schedule: Teachers may need to catch up with teaching <strong>of</strong> material<br />
during the project. The sequential project approach does not<br />
allow for such a teacher’s intervention without disturbing the<br />
process. However, in an iterative project design, issues and success<br />
can be discussed in class regularly.<br />
Required CS knowledge: Some teachers admit a lack <strong>of</strong> project<br />
development knowledge. Students will need help with CS knowledge<br />
while solving problems. It meets the ideas <strong>of</strong> PBL if student<br />
2 German curricula emphasize computer science concepts in the<br />
context <strong>of</strong> general education which only partly includes algorithmic<br />
thinking and programming. In comparison, the un<strong>der</strong>lying<br />
curriculum <strong>of</strong> the study emphasizes foundations <strong>of</strong> algorithmic<br />
thinking and programming [17]. Additionally, we un<strong>der</strong>stand<br />
projects as teamwork where several students or the<br />
whole class are working on the same goal, Meerbaum-Salant<br />
and Hazzan [31] describe projects where students work individually<br />
and “each student has his or her own project subject”.<br />
55<br />
teams would be empowered to manage projects themselves. This<br />
can be supported by easy to follow practices and strategies which<br />
make teacher involvement almost unnecessary. Additionally,<br />
clearly defined practices may support teachers’ confidence. Heterogeneous<br />
student teams stimulate mutual assistance before requiring<br />
the help <strong>of</strong> a teacher.<br />
Students’ individual work: Teachers see a need for personal supervision<br />
in or<strong>der</strong> to achieve a timely completion <strong>of</strong> the project<br />
and meeting <strong>of</strong> the requirements. Agile methods allow for transferring<br />
this responsibility to the project team, hence relieving the<br />
teacher.<br />
Evaluation <strong>of</strong> project outcomes: Agile practices naturally lead to a<br />
variety <strong>of</strong> documents which can be consi<strong>der</strong>ed for project evaluation<br />
(e.g. estimates in planning poker or burn-down charts). Furthermore<br />
mutual assessment within teams may be performed.<br />
All suggested practices require a change <strong>of</strong> the project perspective<br />
from the teacher to the learner. In section 3 the mentioned agile<br />
practices are discussed in more detail and transferred to classroom<br />
settings by the use <strong>of</strong> didactic transposition.<br />
2.6 Didactic Transposition for Project Methodologies<br />
Didactic transposition describes the process <strong>of</strong> adapting pr<strong>of</strong>essional<br />
knowledge <strong>of</strong> a domain for teaching scenarios based on a<br />
didactic intent [9]. Hazzan et. al. [21] applied didactic transposition<br />
on agile SD methods with the intent to create a teaching<br />
framework and a mentoring methodology for s<strong>of</strong>tware projects.<br />
The model for school SD projects discussed in 2.1 can be attributed<br />
to didactic transposition as well. Here, the pr<strong>of</strong>essional process<br />
was adapted un<strong>der</strong> consi<strong>der</strong>ation <strong>of</strong> the un<strong>der</strong>lying principles<br />
<strong>of</strong> PBL. Schubert and Schwill [39] emphasize the advantage <strong>of</strong> a<br />
method which is learning activity and learning content at the same<br />
time. However, we see potential for shifting the focus from pr<strong>of</strong>essional<br />
process knowledge to mo<strong>der</strong>n pr<strong>of</strong>essional methods<br />
which may be adapted in a way that they address previously outlined<br />
problems in school SD projects.<br />
In the following, we discuss agile methods in pr<strong>of</strong>essional SD and<br />
in education. We applied didactical transposition for developing<br />
an agile approach to projects in computing education which emphasizes<br />
a learners’ perspective.<br />
3. AGILE PROJECTS IN COMPUTING<br />
EDUCATION<br />
3.1 Agile Methods in Pr<strong>of</strong>essional and Educational<br />
Settings<br />
Agile methods are popular amongst researchers and practitioners<br />
for enabling s<strong>of</strong>tware developers to create systems that are more<br />
likely to be accurate in meeting customers requests, finishing in<br />
time, building robust systems, and creating usable/readable code<br />
(cp. [22]). Therefore, in industry agile methods are currently replacing<br />
waterfall or other linear methodologies that are known for<br />
shortcomings in the above mentioned goals <strong>of</strong> SD. Agile methods<br />
are focused on social interactions and dynamic creative processes.<br />
Hence, developers in agile teams <strong>of</strong>ten report on a strong satisfaction<br />
in their work experience and strong confidence in their outcomes<br />
(cp. [27, 31]).<br />
The agile manifesto <strong>of</strong> 2001 [3] clearly presents values contrasting<br />
traditional linear methodologies and un<strong>der</strong>lying un<strong>der</strong>standings:
1. Individuals and interactions over processes and tools<br />
2. Working s<strong>of</strong>tware over comprehensive documentation<br />
3. Customer collaboration over contract negotiation<br />
4. Responding to change over following a plan<br />
Agile methods are implemented in various frameworks. XP [2]<br />
and Scrum [40] are the most prominent implementations applied<br />
in industry and academia. Those methodologies define practices<br />
to assure compliance with the agile values as described in the<br />
agile manifesto. Agile methods are mostly un<strong>der</strong>stood to support a<br />
development process comprehensively from start to the final<br />
stage. However, the individual practices can be classified according<br />
to their main targets. Some practices are designed to structure<br />
team processes and customer collaboration while other practices<br />
focus on the quality <strong>of</strong> code and outcome.<br />
Recently, several authors report on using agile methods in CS<br />
education at university level. Braught et al. [7] promote the use <strong>of</strong><br />
agile methods because it helps female students to engage in programming<br />
tasks through interaction with peers. Nagappan et al.<br />
[34] highlight social experiences, learning processes, and quality<br />
<strong>of</strong> code when using agile methods in CS1 courses. However, literature<br />
shows that there may be possible barriers hin<strong>der</strong>ing an<br />
implementation <strong>of</strong> agile methods in university scenarios. Rico and<br />
Sayani [37] present a study where they found that students had<br />
already established their own approaches and habits for SD that<br />
were opposing practices <strong>of</strong> agile methods. According to Rico and<br />
Sayani it was almost impossible to convince the students to adhere<br />
to the introduced practices <strong>of</strong> agile methods. In this connection,<br />
they recommend to introduce agile methods as early as possible.<br />
Consequently, a benefit is assumed in the use <strong>of</strong> agile methods<br />
in school contexts. Literature on agile methods at university<br />
levels is still discussing the possibility <strong>of</strong> conducting a fullyfledged<br />
project management according to an implementation as<br />
XP or Scrum. Some authors promote an almost complete implementation<br />
<strong>of</strong> XP (e.g. [28]), some recommend to use and adopt a<br />
subset <strong>of</strong> practices (e.g. [8]), while others exclusively use one<br />
practice (e.g. pair programming) to support computing education<br />
(e.g. [7]). Schnei<strong>der</strong> and Johnson [38] reviewed agile methods in<br />
computing education and highlight the importance <strong>of</strong> applying<br />
suitable practices according to their goals instead <strong>of</strong> fulfilling<br />
complete implementation <strong>of</strong> agile methods. Accordingly, Hazzan<br />
and Dubinsky [20] present ten reasons to consi<strong>der</strong> agile methods<br />
in computing education.<br />
Yet, literature on agile methods in secondary computing education<br />
is rare. Weigend [42] introduced elements <strong>of</strong> XP (user stories,<br />
spikes, test driven development, refactoring, and big visible<br />
charts) to provide a project-based iterative infrastructure that supports<br />
writing <strong>of</strong> high quality code. However, a sound methodology<br />
for connecting the presented elements in projects was not<br />
provided.<br />
The work in hand is based on encouraging classroom experiences<br />
presented by Göttel [18]. In various educational CS projects, agile<br />
methods, such as pair programming, standup meetings, informative<br />
workspaces, and user stories supported students in their project<br />
work and additionally helped students to discover social aspects<br />
<strong>of</strong> CS. This success provided our basis for developing a<br />
comprehensive agile model for school projects in computing education.<br />
56<br />
3.2 An Agile Model for Projects in<br />
Computing Education (AMoPCE)<br />
As discussed above, PBL represents a common teaching and<br />
learning method in computing education. However, even if common<br />
models suggest a structure and requirements for school s<strong>of</strong>tware<br />
projects, methods are described insufficiently for the individual<br />
phases. Agile methods and mo<strong>der</strong>n SD principles provide a<br />
set <strong>of</strong> clearly described strategies that seems well suited for an<br />
implementation in school contexts. As described in the agile<br />
manifesto, they emphasize communication, visualization, teamwork<br />
and common goals. In the following, we want to introduce a<br />
model for school s<strong>of</strong>tware projects that builds on the character <strong>of</strong><br />
agile methods in or<strong>der</strong> to address the problems outlined in section<br />
2. It follows the agile manifesto by focusing on<br />
1. Students and their interactions<br />
2. Rapid success and working s<strong>of</strong>tware<br />
3. Collaboration in or<strong>der</strong> to strive for a common goal over fulfilling<br />
a contract<br />
4. Responding to change and learning progress over following a<br />
plan<br />
The individual strategies and tools are illustrated in fig. 7 and will<br />
be described below in an agile model for projects in computing<br />
education (AMoPCE).<br />
In this description we focus on processes and methods that are<br />
central for the agile SD process. Additional pedagogical aspects<br />
such as triggering students’ motivation or finding agreement in<br />
choosing a project topic are not covered.<br />
The process contains various techniques adapted from pr<strong>of</strong>essional<br />
SD practices. They provide clear lines <strong>of</strong> action that can be<br />
followed by the students (e.g. generating user stories, planning<br />
poker, defining tasks). However, before applying them in a project<br />
we suggest introducing and practicing each method. On the<br />
other hand, an explorative learning approach is possible: The<br />
methods describe the processes in such detail that appropriate<br />
material can be created which allows students to learn and perform<br />
the processes independently.<br />
The methods will be first discussed from a pr<strong>of</strong>essional perspective<br />
3 and subsequently transferred in a way that they can be applied<br />
in classroom (italic text). Examples will illustrate the process.<br />
3.2.1 Preparation<br />
Creating s<strong>of</strong>tware requires competencies in programming and<br />
using tools. It is the responsibility <strong>of</strong> the teacher to make sure that<br />
the students have acquired the competencies needed for the s<strong>of</strong>tware<br />
project ahead or that they will be able to acquire them during<br />
the project (e.g. with provided teaching material). Another<br />
aspect consi<strong>der</strong>s establishing an appropriate infrastructure (see<br />
[33] for further elaboration).<br />
3.2.2 Ideas in<br />
The initial steps <strong>of</strong> a SD process usually are devoted to requirements<br />
analysis listing possible features, approaches, and needs <strong>of</strong><br />
3 In or<strong>der</strong> to maintain a consistent presentation, methods and practices<br />
described in this paper are based on [35], which may be<br />
referred to for elaboration and additional information on mo<strong>der</strong>n<br />
SD practices.
the target audience. This phase is based on interviews, observations,<br />
and brainstorming sessions with the customers.<br />
Building upon ideas and interests <strong>of</strong> students is a central characteristic<br />
<strong>of</strong> PBL. However, experience shows that students may not<br />
easily come up with ideas that can be implemented in such a project,<br />
especially at the first time. Therefore, an initial presentation<br />
<strong>of</strong> possible projects and the use <strong>of</strong> creativity techniques (e.g.<br />
brainstorming) are suggested. Resulting ideas shall be written<br />
down on individual Post-Its or cards for each activity that the<br />
s<strong>of</strong>tware needs to provide (cp. fig. 2).<br />
Fig 2. Post-its for recording ideas.<br />
3.2.3 User Stories<br />
User stories briefly describe features <strong>of</strong> a product that should be<br />
available to the actual user. Each user story addresses a specific<br />
activity <strong>of</strong> a user and is <strong>der</strong>ived from the ideas <strong>of</strong> the requirements<br />
analysis. They are written from the perspective <strong>of</strong> a customer.<br />
User stories should easily fit on index cards and also be un<strong>der</strong>stood<br />
by non-developers. They should provide additional space<br />
for an estimation <strong>of</strong> the work effort.<br />
In combination, user stories specify the entire intended product. A<br />
final state and amount <strong>of</strong> user stories has to be accomplished in<br />
agreement with the customer. Thereafter, stories are prioritized<br />
together with the customer by sorting user stories according to the<br />
importance <strong>of</strong> each story. Priorities are presented using incrementing<br />
numbers by powers <strong>of</strong> ten from 10 (most important) to<br />
50 (least important).<br />
Furthermore, the customer is asked to pick those stories that<br />
should be available in the initially delivered outcome or rather<br />
first major release. Consequently, discussion and reprioritizing<br />
stories may be necessary consi<strong>der</strong>ing the basic features wanted for<br />
the first release.<br />
User stories are created using various brainstorming techniques<br />
and take account <strong>of</strong> domain specific needs, knowledge, and approaches<br />
<strong>of</strong> the actual users specified in the requirements analysis.<br />
A user story<br />
- covers one activity that needs to be addressed<br />
- represents the perspective <strong>of</strong> the customer<br />
- is short, i.e. contains no more than three sentences<br />
- does not use technical terms<br />
- does not specify technology or tools<br />
57<br />
Fig. 3. Cards for user stories holding a title, a description, an<br />
estimate for workload, and a priority. Estimates will be added<br />
after completing the planning poker.<br />
In pr<strong>of</strong>essional SD projects one <strong>of</strong> the most important (and <strong>of</strong>ten<br />
unsuccessful) tasks is to find out what the customer wants. User<br />
stories provide a helpful way for achieving this goal. Since the<br />
students are going to implement their own ideas in their s<strong>of</strong>tware<br />
projects, this goal does not apply. However, the team needs to<br />
find an agreement on the requirements for the s<strong>of</strong>tware. These<br />
will be represented from a user’s perspective: User stories briefly<br />
describe how a user interacts with the s<strong>of</strong>tware. They can be developed<br />
from the previously recorded ideas. These need to be<br />
analyzed to find out, which interaction is really going to happen.<br />
This will be achieved by role playing and observation. Role playing<br />
is suggested as an attractive method for computing education<br />
to un<strong>der</strong>stand processes (cp. [4, 12, 14]). However, some <strong>of</strong> these<br />
examples use role plays in awkward contexts. In contrast, by role<br />
playing user interaction with the desired s<strong>of</strong>tware system, students<br />
use a method which is anchored in mo<strong>der</strong>n SD processes<br />
and helps to identify relevant processes. The rules <strong>of</strong> the role play<br />
are simple: One student pretends to be the s<strong>of</strong>tware and reacts<br />
accordingly. A sheet <strong>of</strong> paper may be used to illustrate the display.<br />
Another student takes the role <strong>of</strong> the user and instructs the<br />
s<strong>of</strong>tware about what he or she wants to do, according to the previously<br />
obtained ideas. The remaining students observe the situation<br />
carefully to un<strong>der</strong>stand details and constraints <strong>of</strong> the desired<br />
product. The role play should be repeated several times with<br />
changing actors until no more new requirements arise. With this<br />
experience it should be easy to formulate the requirements from a<br />
user’s perspective and write down the corresponding user stories<br />
(cp. fig. 3). Finally, the user stories receive a value for their priority.<br />
Priorities can be determined as a team, since generally<br />
agreement is quickly found.<br />
Communication is an essential element <strong>of</strong> agile SD. Even if a set<br />
<strong>of</strong> user stories will now describe the final goal, questions and<br />
changes will appear in the following process. Since there is no<br />
customer who can answer questions and make decisions in or<strong>der</strong><br />
to clarify yet open questions, a group member needs to take over<br />
this special role: the product owner. This position may be passed<br />
around with each iteration.<br />
3.2.4 Planning Poker<br />
Planning poker is a hands-on method helping participants to estimate<br />
time needed for the work packages and guarantees a fair and<br />
comprehensible approach amongst all team members. Each participant<br />
holds a deck <strong>of</strong> cards to estimate the workload <strong>of</strong> a user<br />
story. There should be cards representing estimates in comprehensible<br />
units (e.g. developer-days) and special cards allowing players<br />
to indicate a lack <strong>of</strong> information, a need for a break, and already<br />
finished functionalities as shown in fig. 4.
Fig. 4. Deck <strong>of</strong> cards used for the planning poker.<br />
Each play round is devoted to one user story. A user story is<br />
placed in the middle <strong>of</strong> the table by the dealer and all participants<br />
place a card specifying their estimate face down on the table. All<br />
played cards are turned at the same time. The dealer collects the<br />
played cards and sums up the estimates trying to set up an average<br />
estimate. The dealer should address outliers by asking for reasons<br />
explaining fundamental differences in the estimates. Furthermore,<br />
the dealer should reflect on average estimates referencing the<br />
differences in the played cards. After each round the acquired<br />
estimate is written on the card <strong>of</strong> the user story. Additionally, the<br />
individual estimates are written on the back <strong>of</strong> the user story card<br />
to keep track <strong>of</strong> the decision process.<br />
For a student team it is one <strong>of</strong> the most difficult tasks to estimate<br />
the workload and time demands <strong>of</strong> a given project due to a lack <strong>of</strong><br />
project experience. Additionally, very likely not all planning relevant<br />
aspects are known at this point, learning processes will happen<br />
and changes may be necessary. Planning poker describes a<br />
playful way for challenging all students <strong>of</strong> the team to engage in<br />
the planning process by analyzing user stories and tasks, relating,<br />
estimating, explaining and defending their calculations, thus<br />
practicing their communication skills and ability to give and receive<br />
criticism.<br />
For school s<strong>of</strong>tware projects the same card values can be used as<br />
in pr<strong>of</strong>essional SD. However, since these projects comprise a<br />
shorter working time, instead <strong>of</strong> days, 15 minute-periods seem to<br />
be appropriate. Each student estimates the time he or she believes<br />
he or she would need to implement the user story in focus. User<br />
stories should be presented with decreasing priority. Discussion<br />
<strong>of</strong> very divergent estimates will help resolving unspecified requirements<br />
and assumptions. After the planning poker is finished,<br />
the total workload for all user stories is divided by the number <strong>of</strong><br />
programmers or programming teams (if pair programming is<br />
used) and compared with the time available. Unlike in pr<strong>of</strong>essional<br />
SD, the priority is not to fulfill all requirements <strong>of</strong> the s<strong>of</strong>tware<br />
but it is indispensable to keep the time available. If there is<br />
a major difference, e.g. more than 20 per cent, amount or complexity<br />
<strong>of</strong> the user stories need to be modified.<br />
Again, for planning poker one team member needs to be the<br />
dealer. We suggest for this special role the position <strong>of</strong> a “teammaster”,<br />
who leads the planning poker as well as team meetings.<br />
In the subsequent process this position also may be passed around<br />
with each iteration.<br />
3.2.5 Tasks<br />
After the initial planning poker, user stories are broken down into<br />
tasks. Usually each user story can be seen as a collection <strong>of</strong> tasks.<br />
A task is a rough description <strong>of</strong> a work package that should be<br />
done by a single developer. A task should have a unique selfexplanatory<br />
name and should indicate its priority. The workload<br />
58<br />
<strong>of</strong> each task should also be estimated by planning poker. Afterwards,<br />
task estimates are summed up to double check them with<br />
the initial estimates on the corresponding user stories. Thus, these<br />
estimates should not differ tremendously.<br />
Fig. 5. Tasks for User Story “Control Lives”.<br />
While the user stories describe project goals from the perspective<br />
<strong>of</strong> the user, the students now need to change their perspective and<br />
look at them from a developer’s viewpoint. By dividing user stories<br />
into tasks, various design decisions need to be made. Experienced<br />
programmers will now benefit from their competencies and<br />
known best practices, less confident students at this point can<br />
benefit and learn from team members, processes and team discussion.<br />
3.2.6 Iterative Development, Prototypes, and Milestones<br />
Agile processes are designed to provide short iterations that constantly<br />
come up with working prototypes that can be used and<br />
discussed with users or customers. This allows for rapid feedback<br />
loops that help to uncover misun<strong>der</strong>standings, to detect issues in<br />
using the interface, and to adapt to new requests. An iteration is<br />
supposed to be short and has to be balanced according to implementing<br />
new features, fixing bugs, responding to change, and<br />
consi<strong>der</strong>ing group dynamics or individual demands. In pr<strong>of</strong>essional<br />
contexts, iterations vary between one week (5 working<br />
days) to one month (approximately 20 working days).<br />
Iterations are planned in a team meeting by consi<strong>der</strong>ing user stories’<br />
priorities, estimates, and the intended duration <strong>of</strong> an iteration.<br />
After deciding on the user stories to work on for the next<br />
iteration, they are pinned on a project board including associated<br />
tasks.<br />
In school s<strong>of</strong>tware projects, the planning <strong>of</strong> long development<br />
processes is reported to be difficult. Also, teachers report issues<br />
maintaining student motivation while they are not getting a grasp<br />
<strong>of</strong> the product until the whole project is assembled. The learning<br />
theory <strong>of</strong> constructionism emphasizes that learning happens especially<br />
felicitously in a context where learners are engaged with<br />
creating and investigating a personal relevant product (cp. [35]).<br />
Iterative development allows for creating a series <strong>of</strong> prototypes<br />
that can be analyzed, examined and played with in a constructionist<br />
sense. Also, such a design <strong>of</strong> the development processes allows<br />
for a higher flexibility in team organization and diversification<br />
due to more frequently changing tasks. Hence, each iteration is a<br />
mini-project containing each phase <strong>of</strong> the SD processes (requirements,<br />
design, code, test), but is easier to handle. This gives students<br />
the opportunity to perform the whole process several times<br />
within one project, to learn from and reflect on previous experiences<br />
and to take over several tasks in the team (in comparison to<br />
other project models, where tasks are more strictly divided<br />
amongst students). Besides the iterations, which should not be
longer than one to two weeks (which equals 2-4 lessons), milestones<br />
are used to structure the process and point out major<br />
achievements within the project. We suggest identifying 2-3 milestones<br />
for each project, representing versions <strong>of</strong> the final product<br />
with increasing value. However, only the achievement for the next<br />
milestone in the development is determined at a time. Milestones<br />
can be used for presenting the project progress for the rest <strong>of</strong> the<br />
class or teachers. Also, milestones should be positioned at times<br />
when the project pauses and teacher input is planned. Goals for<br />
the next milestone and the project progress are visualized at the<br />
project board.<br />
3.2.7 Project Board<br />
Project boards visualize goals and status <strong>of</strong> a current iteration and<br />
support target-oriented discussions. They present user stories and<br />
tasks in different status areas. Project boards are updated and<br />
discussed throughout the entire process. Thereby, it helps team<br />
members to keep track <strong>of</strong> the progress <strong>of</strong> the design process: the<br />
different areas <strong>of</strong> the board are used to present goals and accomplishments<br />
to the whole team. There are three main status areas:<br />
to-do user stories with associated tasks for the current iteration,<br />
tasks that are in progress, and completed tasks. Furthermore, there<br />
is an area to store user stories that need to be reconsi<strong>der</strong>ed in a<br />
future iteration. To provide a clear view, another area is reserved<br />
for finished user stories, allowing to take <strong>of</strong>f corresponding task<br />
cards. Figure 6 presents an accordant project board.<br />
Additionally, a burn-down chart is available on the project board.<br />
The chart visualizes the working time left in an iteration and work<br />
that needs to be done according to the task estimates. The chart is<br />
constantly keeping track <strong>of</strong> the progress by plotting the remaining<br />
sum <strong>of</strong> tasks at the end <strong>of</strong> a working unit.<br />
Likewise, in a school s<strong>of</strong>tware project all user stories with corresponding<br />
tasks are collected and presented at the team’s project<br />
board. The project board is the central organizational and informative<br />
workspace for the entire project and should be available<br />
at all times, e.g. by placement at the classroom wall. It is also the<br />
meeting point for the regular standup meetings.<br />
Fig. 6. Project Board including a burn-down chart.<br />
3.2.8 Standup Meetings<br />
Standup meetings provide a recurring fast and short update <strong>of</strong> the<br />
efforts <strong>of</strong> the team: Each team member has to report on accomplished<br />
tasks, possible issues in accomplishing certain goals, and<br />
59<br />
a plan for the work day. Meetings are done while standing to<br />
guarantee a fast and goal oriented session kicking <strong>of</strong>f a workday<br />
and should not exceed 5 to 15 minutes.<br />
In school projects, standup meetings can provide an elegant way<br />
for starting <strong>of</strong>f a lesson or working day within a project by encouraging<br />
team communication, sustaining motivation and identifying<br />
problems. The team gathers around the project board and<br />
recalls the project status, success and problems <strong>of</strong> the last working<br />
session and the goals for the day. After each team member has<br />
given a short statement the burn down chart is updated. If nonminor<br />
problems are identified, a longer meeting may be scheduled.<br />
3.2.9 Pair Programming<br />
Pair programming ensures an elaborated coding style: A pair <strong>of</strong><br />
programmers uses a single programming environment for coding.<br />
The person using the keyboard and mouse is adopting the role <strong>of</strong><br />
the “driver”. The driver is actually coding and asked to present his<br />
or her ideas to the second programmer (the “navigator”) verbally.<br />
Meanwhile, the navigator questions the coding outcomes, discusses<br />
possible misinterpretations, and seeks for alternative solutions<br />
that are more straightforward by keeping in focus the overall<br />
goals. The roles <strong>of</strong> driver and navigator are changed repeatedly<br />
during a workday. Programming in pairs helps to detect possible<br />
slips in the design and architecture <strong>of</strong> the code at early stages.<br />
Furthermore, it helps programmers to build upon social interaction<br />
uncovering misinterpretations <strong>of</strong> relations and intentions <strong>of</strong><br />
code parts.<br />
In many schools, two students share one computer due to limited<br />
hardware availability and hence <strong>of</strong>ten program in pairs. The agile<br />
method <strong>of</strong> pair programming supports this practice and adds a<br />
framework that encourages attention from both students, mutual<br />
learning and a notion <strong>of</strong> programming as a social activity.<br />
Fig. 7 illustrates the organization <strong>of</strong> a school s<strong>of</strong>tware project<br />
based on AMoPCE as described above. Other agile methods and<br />
ideas may be included in such a school project as well, e.g. test<br />
driven development, refactoring or “keep it simple”.<br />
3.3 Focusing on Programming Style and Outcome<br />
Several agile practices may be applied in the process to bring<br />
forward a high quality outcome. However, since these practices<br />
are optional for project organization and partly depend on programming<br />
environments, in the following they are only summarized<br />
but not adapted for school s<strong>of</strong>tware projects.<br />
3.3.1 Test Driven Development<br />
Test driven development replaces documentation and provides<br />
criteria to evaluate the code solutions: Programmers define intended<br />
functionalities by writing automatic tests covering all<br />
states and the correctness <strong>of</strong> the accordant results. First reports <strong>of</strong><br />
using this method in secondary education have been published<br />
(e.g. [10]).<br />
3.3.2 Refactoring<br />
Refactoring introduces the idea that every part <strong>of</strong> the code should<br />
be reconsi<strong>der</strong>ed and changed if a more accurate solution can be<br />
found: Developers should rewrite parts <strong>of</strong> code without adding<br />
functionality when there is a more straightforward solution available<br />
that passes all automatic tests. Emphasizing this aspect may<br />
raise students’ awareness <strong>of</strong> efficiency and for evaluating different<br />
solutions.
3.3.3 Keep It Simple<br />
This claim refers to code minimalism: each function should be<br />
solved by minimal and straightforward code snippets to ensure an<br />
elegant and readable code that is easy to maintain.<br />
4. DISCUSSION<br />
The proposed agile model for SD projects AMoPCE addresses a<br />
majority <strong>of</strong> the previously identified problems. Agile practices fill<br />
the learners’ gap between requirements and outcome by providing<br />
clearly defined strategies for handling difficult planning tasks.<br />
Based on the perception <strong>of</strong> PBL as a team activity, which is in<br />
line with mo<strong>der</strong>n SD, a team size <strong>of</strong> 4-6 students is recommended.<br />
Dividing the class into several teams, in comparison to having the<br />
full class working on one project as proposed e.g. by Schubert and<br />
Schwill [39], allows for addressing a broa<strong>der</strong> range <strong>of</strong> topics,<br />
hence it is easier to meet the interest <strong>of</strong> more students. Also, it<br />
should be much easier to find agreement within smaller teams.<br />
Adopting an iterative project design matches the formal circumstances<br />
<strong>of</strong> school projects. In most cases, projects will be worked<br />
at along the regular school timetable, sometimes in a so-called<br />
project week on several days. In both settings, pedagogical aspects<br />
<strong>of</strong> working in group settings, such as giving curricular input,<br />
intermediate project presentations or discussing <strong>of</strong> common problem<br />
solving strategies, can be taken into account easier, since<br />
meetings in plenum are part <strong>of</strong> the model. Correspondingly, projects<br />
are divided into mini-projects, which are easier to overview,<br />
plan and un<strong>der</strong>stand; bureaucratic overhead is reduced. Classroom-management<br />
aspects are addressed with pr<strong>of</strong>essional practices<br />
such as standup meetings. This includes recalling <strong>of</strong> the<br />
project status at the beginning <strong>of</strong> each lesson and quickly planning<br />
the individual and group activities for the day.<br />
Fig. 7. Agile Model for Projects in Computing Education (AMoPCE).<br />
60<br />
Within projects, ideas, students’ motivation, and skills change<br />
over time. Due to the limited time available to work on the projects<br />
per week, this is especially very likely for school s<strong>of</strong>tware<br />
projects. Agile methods welcome changes and provide mechanisms<br />
to adapt to them with <strong>of</strong>ten changing tasks and a straightforward<br />
implementation <strong>of</strong> PBL. Students’ confidence is addressed<br />
by increasing familiarity stemming from the iterative<br />
character <strong>of</strong> the process.<br />
In comparison to the ACMM, which provides solutions for teachers’<br />
difficulties that are grounded in teaching situations, the proposed<br />
agile model provides solutions for students’ difficulties in<br />
learning situations. This in return is expected to relieve the<br />
teacher. Giving students clearly defined practices to manage their<br />
development processes allows teachers to focus on supporting<br />
elaboration and implementation <strong>of</strong> students’ ideas, thus changing<br />
the teacher’s role from instructor to coach. This better meets the<br />
demands <strong>of</strong> PBL. Additionally, it allows teachers to highlight<br />
creativity and social aspects that are rarely seen in connection<br />
with computer science (cp. [8, 29]). Applying an adapted SD<br />
methodology in school that is also implemented by well known<br />
large scale companies may help teachers and students to build an<br />
adequate and appealing un<strong>der</strong>standing <strong>of</strong> computer science relying<br />
on creativity, dynamic change, feedback, and s<strong>of</strong>t skills.<br />
These attributes may support an attractive notion <strong>of</strong> computer<br />
science, as found in our first experimental settings with agile<br />
methods in school projects [18] .<br />
In summary, AMoPCE is suited for supporting the objectives <strong>of</strong><br />
PBL, for maintaining a pr<strong>of</strong>essional orientation and for easing the<br />
mentoring <strong>of</strong> s<strong>of</strong>tware projects. However, in this model curricular<br />
aspects such as content and size <strong>of</strong> a project are not explicitly<br />
consi<strong>der</strong>ed. Nevertheless, it provides the flexibility to fit into a<br />
variety <strong>of</strong> possible scenarios. Evaluation and assessment aspects,
e.g. assessing individual achievement in comparison to group<br />
achievement are not determined by the model. However, again<br />
practices <strong>of</strong> SD appear to be adaptable and can be consi<strong>der</strong>ed:<br />
Frequently, agile development teams use self assessments and<br />
subsequent peer reviews to verify individual workload and commitment.<br />
As outlined in section 3, there are further practices <strong>of</strong> agile methods<br />
that focus on the quality <strong>of</strong> the outcome: test driven development,<br />
collective code ownership, refactoring, and keep it simple.<br />
We acknowledge these practices to be also useful in educational<br />
settings but we un<strong>der</strong>stand them to be highly dependant on features<br />
and methodologies <strong>of</strong> the used programming environments<br />
and tools (e.g. code repositories, automatic testing environments).<br />
The use <strong>of</strong> agile practices in school SD projects has the potential<br />
for replacing the so far predominantly used sequential model.<br />
From discussions with teachers we know <strong>of</strong> the high interest, but<br />
a lack <strong>of</strong> knowledge and resources concerning the use <strong>of</strong> agile<br />
methods. This includes the demand for a revised project model.<br />
With AMoPCE, as outlined in this article, we present a model<br />
which explains ideas and realization possibilities <strong>of</strong> agile practices<br />
and which can be used as blueprint. In a next step <strong>of</strong> our<br />
research the model will be applied in classroom settings and it<br />
will be investigated, to which extend the expected benefits and<br />
problem solutions will be approved in practice.<br />
5. REFERENCES<br />
[1] Barak, M., Waks, S. and Doppelt, Y. (2000). Majoring in<br />
technology studies at high school and fostering learning.<br />
Learning Environments Research 3(2): 135-158.<br />
[2] Beck, K. and Andres, C. (2004). Extreme Programming Explained:<br />
Embrace Change (2nd Edition).<br />
[3] Beck, K., Beedle, M., et al. (2001). Manifesto for agile s<strong>of</strong>tware<br />
development. The Agile Alliance: 2002-04.<br />
[4] Bergin, J. (2000). The Object Game. An Exercise for Studying<br />
Objects. Online at:<br />
http://csis.pace.edu/~bergin/patterns/objectgame.html (visited<br />
01.07.2012).<br />
[5] Blumenfeld, P. C., Soloway, E., et al. (1991). Motivating<br />
Project-Based Learning: Sustaining the Doing, Supporting<br />
the Learning. Educational Psychologist 26: 369-398.<br />
[6] Blumenfeld, P. C., Soloway, E., et al. (1991). Motivating<br />
project-based learning: Sustaining the doing, supporting the<br />
learning. Educational Psychologist 26(3-4): 369-398.<br />
[7] Braught, G., Wahls, T. and Eby, L. M. (2011). The Case for<br />
Pair Programming in the Computer Science Classroom.<br />
Transactions on Computing Education 11: 2:1--2:21.<br />
[8] Caspersen, M. E. and Kölling, M. (2009). STREAM: A First<br />
Programming Process. Transactions on Computing Education<br />
9: 4:1--4:29.<br />
[9] Chevallard, Y. (1988). On didactic transposition theory:<br />
Some introductory notes. In Proceedings <strong>of</strong> the International<br />
Symposium on Research and Development in Mathematics,<br />
Bratislava, Czechoslavakia.<br />
[10] Christopher, G. J. (2004). Test-driven development goes to<br />
school. J. Comput. Small Coll. 20(1): 220-231.<br />
[11] Dewey, J. and Kilpatrick, W. H. (1935). Der Projekt-Plan:<br />
Grundlegung und Praxis, Hermann Böhlaus Nachfolger.<br />
[12] Diethelm, I. (2007). Strictly models and objects first - Unterrichtskonzept<br />
und -methodik für objektorientierte Modellierung<br />
im Informatikunterricht. Kassel, Universität Kassel.<br />
61<br />
[13] Fincher, S. and Petre, M. (1998). Project-based learning practices<br />
in computer science education. In Proceedings <strong>of</strong> the<br />
IEEE Frontiers in Education Conference (Tempe, Arizona).<br />
[14] Fothe, M. (2007). Algorithmen in spielerischer Form. In<br />
Proceedings <strong>of</strong> the Praxisband <strong>der</strong> GI-Fachtagung Informatik<br />
und Schule.<br />
[15] Frey, K. (1983). Die sieben Komponenten <strong>der</strong> Projektmethode<br />
- mit Beispielen aus dem Schulfach Informatik. LOG IN<br />
3(2): 16-20.<br />
[16] Frey, K. and Schäfer, U. (1982). Die Projektmethode. Weinheim<br />
[a.o.] Beltz.<br />
[17] Gal-Ezer, J., Beeri, C., et al. (1995). A high school program<br />
in computer science. Computer 28(10): 73-80.<br />
[18] Göttel, T. (2012). The image <strong>of</strong> Computer Science and social<br />
aspects <strong>of</strong> mo<strong>der</strong>n s<strong>of</strong>tware development processes in school<br />
contexts. Hamburg, <strong>University</strong> <strong>of</strong> Hamburg.<br />
[19] Hartmann, W., Näf, M. and Reichert, R. (2006). Informatikunterricht<br />
planen und durchführen, Springer.<br />
[20] Hazzan, O. and Dubinsky, Y. (2007). Why S<strong>of</strong>tware Engineering<br />
Programs Should Teach Agile S<strong>of</strong>tware Development.<br />
SIGSOFT S<strong>of</strong>tware Engineering Notes 32: 1-3.<br />
[21] Hazzan, O., Dubinsky, Y. and Meerbaum-Salant, O. (2010).<br />
Didactic transposition in computer science education. ACM<br />
Inroads 1(4): 33-37.<br />
[22] Highsmith, J. and Cockburn, A. (2001). Agile S<strong>of</strong>tware Development:<br />
The Business <strong>of</strong> Innovation. Computer 34: 120-<br />
122.<br />
[23] Humbert, L. (2005). <strong>Didaktik</strong> <strong>der</strong> Informatik, Teubner.<br />
[24] Koerber, B. (1992). Die Angst des Lehrers vorm Projektunterricht.<br />
LOG IN 12(5/6): 3.<br />
[25] Krajcik, J. S., Czerniak, C. and Berger, C. (1999). Teaching<br />
children science: A project-based approach, McGraw-Hill<br />
Boston, MA.<br />
[26] Larman, C. and Basili, V. R. (2003). Iterative and Incremental<br />
Development: A Brief History. Computer 36: 47-56.<br />
[27] Layman, L., Williams, L. and Cunningham, L. (2004). Exploring<br />
Extreme Programming in Context: An Industrial Case<br />
Study. Proceedings <strong>of</strong> the Agile Development Conference.<br />
Washington, DC, USA, IEEE Computer Society: 32-41.<br />
[28] L<strong>of</strong>tus, C. and Ratcliffe, M. (2005). Extreme Programming<br />
Promotes Extreme Learning? Proceedings <strong>of</strong> the 10th annual<br />
SIGCSE conference on Innovation and technology in computer<br />
science education. New York, NY, USA, ACM: 311-<br />
315.<br />
[29] Maass, S. and Wiesner, H. (2006). Programmieren, Mathe<br />
und ein bisschen Hardware... Wen lockt dies Bild <strong>der</strong> Informatik?<br />
Informatik-Spektrum 29(2): 125-132.<br />
[30] Magenheim, J., Nelles, W., et al. Competencies for informatics<br />
systems and modeling: Results <strong>of</strong> qualitative content<br />
analysis <strong>of</strong> expert interviews. In Proceedings <strong>of</strong>. IEEE.<br />
[31] Mann, C. and Maurer, F. (2005). A Case Study on the Impact<br />
<strong>of</strong> Scrum on Overtime and Customer Satisfaction. Proceedings<br />
<strong>of</strong> the Agile Development Conference. Washington, DC,<br />
USA, IEEE Computer Society: 70-79.<br />
[32] Meerbaum-Salant, O. and Hazzan, O. (2009). Challenges in<br />
Mentoring S<strong>of</strong>tware Development Projects in the High<br />
School: Analysis According to Shulman. Journal <strong>of</strong> Computers<br />
in Mathematics and Science Teaching 28(1): 21.<br />
[33] Meerbaum-Salant, O. and Hazzan, O. (2010). An Agile Constructionist<br />
Mentoring Methodology for S<strong>of</strong>tware Projects in<br />
the High School. Transactions on Computing Education 9:<br />
21:1--21:29.
[34] Nagappan, N., Williams, L., et al. (2003). Improving the CS1<br />
Experience with Pair Programming. ACM SIGCSE Bulletin<br />
35: 359-362.<br />
[35] Papert, S. and Harel, I. (1991). Situating Constructionism.<br />
Constructionism. S. Papert and I. Harel. Norwood, N.J.,<br />
Ablex Publishing.<br />
[36] Pilone, D. and Miles, R. (2008). Head first s<strong>of</strong>tware development,<br />
O'Reilly Media, Inc.<br />
[37] Rico, D. F. and Sayani, H. H. (2009). Use <strong>of</strong> Agile Methods<br />
in S<strong>of</strong>tware Engineering Education. Proceedings <strong>of</strong> the 2009<br />
Agile Conference. Washington, DC, USA, IEEE Computer<br />
Society: 174-179.<br />
[38] Schnei<strong>der</strong>, J.-G. and Johnston, L. (2005). eXtreme Programming:<br />
Helpful or Harmful in Educating Un<strong>der</strong>graduates?<br />
Journal <strong>of</strong> Systems and S<strong>of</strong>tware 74: 121-132.<br />
62<br />
[39] Schubert, S. and Schwill, A. (2011). <strong>Didaktik</strong> <strong>der</strong> Informatik,<br />
Springer.<br />
[40] Schwaber, K. and Beedle, M. (2001). Agile S<strong>of</strong>tware Development<br />
with Scrum.<br />
[41] Tucker, A., Deek, F., et al. (2003). A model curriculum for<br />
K-12 computer science: Final report <strong>of</strong> the ACM K-12 task<br />
force curriculum committee. New York, NY: The Association<br />
for Computing Machinery.<br />
[42] Weigend, M. (2005). Extreme Programming im Klassenraum.<br />
In Proceedings <strong>of</strong> the INFOS 2005 - 11. GI-<br />
Fachtagung Informatik und Schule (Dresden, <strong>Germany</strong>,<br />
2005). GI-Edition - Lecture Notes in Informatics (LNI).
Conceptual Change and Epistemological Belief<br />
Framework for Web Site Credibility Instruction<br />
Abstract—This paper describes instructional needs regarding<br />
Web-based instruction and reviews a line <strong>of</strong> research<br />
examining students’ determinations <strong>of</strong> Web site credibility. It<br />
proposes a framework for instructional interventions that<br />
draws upon two research traditions - scientific conceptual<br />
change research and epistemological beliefs about Web-based<br />
information. Recommendations for instruction and research<br />
based upon this model are delineated.<br />
Keywords-misconceptions, student engagement with<br />
technology, beliefs, conceptualizations <strong>of</strong> computing, teaching<br />
approaches, teaching methods, teaching with educational<br />
technology<br />
I. INTRODUCTION<br />
Is Web site credibility an important consi<strong>der</strong>ation for<br />
students doing research and using Web-based materials or are<br />
most students already knowledgeable in this area? This paper<br />
argues that consi<strong>der</strong>ations <strong>of</strong> Web site credibility are even<br />
more important than in the past and strongly tied to today’s<br />
critical thinking skills, both for computing education and for<br />
learning in general. Further, it is erroneous to assume that<br />
students are knowledgeable in this area. To illustrate, I<br />
describe research that investigates students’ and pre-service<br />
teachers (i.e., teachers-in-training) determination <strong>of</strong> Web site<br />
credibility. This research un<strong>der</strong>scores the need for<br />
instructional interventions at elementary, secondary and postsecondary<br />
(university levels), and provides the impetus for<br />
developing effective instructional interventions. Thus, the<br />
main focus <strong>of</strong> this paper is in the development <strong>of</strong> a theoretical<br />
framework that can guide instructional interventions. In<br />
developing this framework, we borrow upon research<br />
frameworks provided in the conceptual change research<br />
related to science learning and epistemological belief research<br />
related to the Internet.<br />
II. CREDIBILITY RESEARCH<br />
Do students make determinations <strong>of</strong> Web site credibility<br />
that differ from determinations <strong>of</strong> experts (pr<strong>of</strong>essionals in<br />
various fields and doctoral students)? What sorts <strong>of</strong><br />
determinations do they make? What kinds <strong>of</strong> instructional<br />
interventions are effective for students?<br />
Marie Iding, <strong>University</strong> <strong>of</strong> Hawaii<br />
College <strong>of</strong> Education<br />
1776 <strong>University</strong> Avenue<br />
Honolulu, HI, 96822, U.S.A.<br />
miding@hawaii.edu<br />
63<br />
In an initial study, Klemm, Iding and Speitel [1]<br />
compared scientists’ and pre-service elementary and<br />
secondary teachers’ determination <strong>of</strong> information credibility<br />
<strong>of</strong> various science resources, including information from the<br />
Web. They found that secondary pre-service teachers (who<br />
had more science training and required coursework in the<br />
area) were more like scientists in their determinations.<br />
However, pre-service teachers were more likely to rate<br />
popular information sources such as newsmagazines and<br />
popular news television shows as credible information<br />
sources. In contrast, scientists rated these information sources<br />
as low in credibility. There was some agreement on museum,<br />
aquarium and nature center resources, which were rated<br />
uniformly highly by all 3 groups.<br />
In subsequent research investigating tenth gra<strong>der</strong>s<br />
credibility determinations regarding biology information on<br />
Web sites, an instructional intervention was employed [2].<br />
Over a four-day period <strong>of</strong> time, a teacher provided instruction<br />
in examining specific areas relevant to making credibility<br />
determinations – Web site authors and institutional<br />
affiliations, information veracity, and organization <strong>of</strong> Web<br />
sites. Students’ lists <strong>of</strong> criteria affecting their own credibility<br />
determinations were more extensive after the intervention, and<br />
students reported that they would spend more time evaluating<br />
Web resources for credibility in the future.<br />
In other research Iding, Crosby, Auernheimer and Klemm<br />
[3] investigated university students’ credibility determinations<br />
regarding content related to topics that they had covered in<br />
class (i.e., computer science or education topics). Findings<br />
indicated that students exhibited awareness <strong>of</strong> and suspicion<br />
regarding vested interests <strong>of</strong> Web site authors with<br />
commercial interests, although they did not as critically regard<br />
vested interests <strong>of</strong> Web site authors whose purposes were<br />
viewed as solely educational in nature.<br />
In work with Norwegian un<strong>der</strong>graduate and graduatelevel<br />
university students enrolled in film studies classes, Iding,<br />
Nordbotten and Singh [4] noted that the more educated<br />
students were, the less confident they were about their Web
site credibility determinations. This finding makes sense<br />
intuitively, as it un<strong>der</strong>scores the difficulties inherent in<br />
information credibility determinations.<br />
In other research, credibility determinations were affected<br />
by literacy levels and socioeconomic status (SES) [5]. For<br />
example, with respect to health care posted on Web sites, HIV<br />
positive patients with lower educational levels and literacy<br />
skills were less critical <strong>of</strong> fraudulent claims regarding<br />
purported “cures” for AIDS than were their more educated<br />
counterparts from higher SES levels.<br />
Furthermore, Fogg, Soohoo, Danielson, Marable,<br />
Stanford and Taber [6] found that members <strong>of</strong> the general<br />
public could describe adequate bases upon which to make<br />
credibility determinations. However, when faced with actual<br />
Web sites, they abandoned these criteria in favor <strong>of</strong><br />
determinations based on Web site appearance (or “design<br />
look”).<br />
Anecdotally, the author has found that while teaching in a<br />
Pacific Island US territory in June, 2012, university level<br />
students appeared to be less aware <strong>of</strong> Web site credibility<br />
issues in citing sources than were their US counterparts. It<br />
should be noted that these students are from a developing part<br />
<strong>of</strong> the Pacific and the majority would fit US definitions <strong>of</strong><br />
living below the poverty level. Thus, many do not have<br />
computers at home or regular easily obtainable access to the<br />
Internet. Additionally, the university-level courses in which<br />
they were enrolled were only for several weeks, not allowing<br />
extensive work with resources or time to make effective<br />
credibility determinations.<br />
These examples un<strong>der</strong>score the ongoing need for critical<br />
Web credibility evaluation instruction in educational contexts,<br />
at all levels, or wherever and whenever computers are<br />
introduced into instruction. While it is clearly possible to<br />
teach Web and information evaluation without reference to a<br />
theoretical basis un<strong>der</strong>lying instruction, I propose that it may<br />
be valuable to develop instruction that is consistent with<br />
frameworks known to be associated with the growth <strong>of</strong> critical<br />
learning skills. Because determinations about information and<br />
credibility implicitly involve epistemological consi<strong>der</strong>ations, I<br />
draw briefly on research that investigates epistemological<br />
beliefs regarding information on the Internet. Secondly, since<br />
science and scientific learning is an area where students must<br />
counter their initial conceptions or preconceptions in or<strong>der</strong> to<br />
learn effectively, and because science-related Web sites are<br />
replete with misinformation, I discuss conceptual change<br />
research and how it might be drawn on analogically to<br />
facilitate the development <strong>of</strong> aspects <strong>of</strong> an instructional model<br />
for effective instruction in credibility evaluation.<br />
III. EPISTEMOLOGICAL BELIEFS AND THE INTERNET<br />
People implicitly make epistemological determinations or<br />
exercise epistemological beliefs in selecting information from<br />
the Web; reading, watching or listening to it; and finally in<br />
64<br />
citing it or using that information to make decisions or to<br />
influence their thinking in some way. Braten, Stromso, and<br />
Samuelstuen [7] conducted a relevant study in this area and<br />
adopted the definition <strong>of</strong> personal epistemology originally<br />
proposed by H<strong>of</strong>er as, “how knowing occurs, what counts as<br />
knowledge and where it resides, and how knowledge is<br />
constructed and evaluated” (p. 1).<br />
Such a definition provides a useful basis for instruction,<br />
especially in different content areas. Braten et al. [7] carried<br />
out research investigating Norwegian university students’<br />
epistemological beliefs around information on the Internet,<br />
using dimensions <strong>of</strong> epistemological beliefs proposed by<br />
H<strong>of</strong>er and Pintrich [8]. Results indicated that two dimensions<br />
<strong>of</strong> epistemological beliefs were confirmed via factor analyses.<br />
As Braten et al. explained, “The first dimension, general<br />
Internet epistemology, ranged from the integrated view that<br />
the Internet is an essential source <strong>of</strong> true, specific facts to<br />
doubt about the Internet as a good source <strong>of</strong> true factual<br />
knowledge. The other dimension, justification for knowing,<br />
ranged from the view that Internet-based knowledge claims<br />
can be accepted without critical evaluation to the view that<br />
knowledge claims encountered on the Internet should be<br />
checked against other sources, reason, and prior knowledge”<br />
(p. 141).<br />
Clearly, effective instruction for credibility<br />
determinations should ideally include an introduction to<br />
epistemological consi<strong>der</strong>ations that can be general or content<br />
specific. Consi<strong>der</strong>ations such as those raised by H<strong>of</strong>er in the<br />
definition above can guide the discussion, and inspire<br />
questions such as: What is knowledge? Who creates it? Who<br />
determines whether it is accepted in a field? What are the<br />
characteristics <strong>of</strong> credible knowledge? How does one<br />
personally determine whether knowledge expressed on the<br />
Web is credible?<br />
IV. CONCEPTUAL CHANGE RESEARCH<br />
A classic means <strong>of</strong> countering science misconceptions<br />
was summarized by Chinn and Brewer [9]. Their instructional<br />
model consisted <strong>of</strong> the following points:<br />
1) “Consi<strong>der</strong> a physical scenario whose outcome is not<br />
known” (p. 38).<br />
2) “Predict the outcome” (p. 38).<br />
3) “Construct competing theoretical explanations to<br />
support the predictions” (p. 38).<br />
4) “Observe the outcome (anomalous data)” (p. 38).<br />
5) “Modify competing theoretical explanations, if<br />
necessary” (p. 38)<br />
6) “Evaluate competing explanations” (p. 38)<br />
7) “Reiterate the preceding steps with different data.” (p.<br />
38)<br />
Chinn and Brewer [9] described various factors that can<br />
influence the likelihood <strong>of</strong> conceptual change, including how<br />
anomalous data is handled – whether it is accepted, ignored,
einterpreted, or excluded, for example. They explained that<br />
individuals’ background knowledge, competing theories,<br />
characteristics <strong>of</strong> the anomalous data, and strategies used also<br />
affect the likelihood <strong>of</strong> conceptual change.<br />
Chinn and Brewer [9] argued that only by consi<strong>der</strong>ing<br />
one’s preconceptions in light <strong>of</strong> anomalous data deeply, (<strong>of</strong>ten<br />
prompted by explaining one’s thoughts to others) can true<br />
conceptual change occur. They also convincingly explain<br />
that, “In or<strong>der</strong> to learn epistemological commitments<br />
appropriate to evaluating evidence and theories, students may<br />
need to participate in a community that regularly debates<br />
alternative theories discusses responses to anomalous data,<br />
and evaluates evidence and theories. During this process <strong>of</strong><br />
enculturation, students are like apprentices learning the craft<br />
<strong>of</strong> scientific reasoning and teachers can use strategies <strong>of</strong><br />
modeling (by thinking out loud in front <strong>of</strong> students), coaching,<br />
providing scaffolding and gradually withdrawing it, and<br />
reflecting upon the cognitive strategies used (A. Collins,<br />
Brown & Newman, 1989)” (p. 33).<br />
Pintrich, Marx and Boyle [10] also addressed the<br />
conceptual change research, emphasizing that countering<br />
preconceptions at the logical, cognitive level may leave out<br />
important non-rational, motivational and affective elements <strong>of</strong><br />
the determinations. These elements include motivational<br />
aspects <strong>of</strong> conceptual change, including goals, interests,<br />
values, and self-efficacy beliefs. They called for further<br />
research in this area. Clearly, affective and motivational<br />
aspects also play major roles in people’s credibility<br />
determinations, as the work <strong>of</strong> Fogg et al. [6], mentioned<br />
earlier, illustrates. To what specific extent and how remain<br />
potentially fruitful questions for further research in Web site<br />
credibility determinations as well.<br />
Thus, if we borrow from conceptual change research in<br />
science learning, we need to consi<strong>der</strong> students’ goals in<br />
credibility determinations and possible motives (or nonmotives).<br />
Further, we need to consi<strong>der</strong> whether there is an<br />
analogous event that can trigger conceptual change in much the<br />
same way conceptual change is trigger by anomalous events or<br />
information that does not fit with preconceptions. Within<br />
context <strong>of</strong> the conceptual change research, perhaps an<br />
analogous triggering event could be confrontation <strong>of</strong> Webbased<br />
information on a Web site that has a very credible<br />
appearance. Exercises with Web sites from different content<br />
areas and with different levels <strong>of</strong> information quality should<br />
facilitate learning.<br />
V. GENERAL INSTRUCTIONAL RECOMMENDATIONS<br />
Based upon this summary <strong>of</strong> research on Web credibility,<br />
and syntheses <strong>of</strong> suggestions that can be drawn from the<br />
research in epistemological beliefs related to the Internet, and<br />
conceptual change in science, the following suggestions are<br />
proposed for inclusion in effective Web information<br />
credibility instruction:<br />
65<br />
• Consi<strong>der</strong>ation <strong>of</strong> content specific epistemological<br />
approaches and beliefs. Discussion <strong>of</strong> the questions<br />
such as the following can be pr<strong>of</strong>itable: What is<br />
knowledge? Who evaluates it and determines whether<br />
it is accepted?<br />
• Discussion <strong>of</strong> specific aspects <strong>of</strong> Web site credibility,<br />
including sources, possible vested interests <strong>of</strong> Web site<br />
authors, corroboration <strong>of</strong> information with other<br />
sources<br />
• Teachers should model and think aloud as they<br />
evaluate the credibility <strong>of</strong> actual Web sites <strong>of</strong> varying<br />
degrees <strong>of</strong> credibility [10,11]<br />
• Students should work collaboratively to evaluate actual<br />
sites <strong>of</strong> varying credibility, and be encouraged to<br />
develop own strategies.<br />
• Students should be encouraged to articulate own<br />
explanations for their determinations to others.<br />
• Self-regulation should be central. Thus, if we consi<strong>der</strong><br />
Web site credibility determinations to involve to some<br />
extent a process <strong>of</strong> conceptual change, students should<br />
be encouraged to address the following questions when<br />
consi<strong>der</strong>ing a Web site as an information source:<br />
Firstly, what is my initial conception or credibility<br />
determination regarding this Web site? Secondly, why?<br />
What characteristics caused me to make this<br />
determination – Web site appearance, author’s title,<br />
educational or company affiliation? Lastly, when<br />
comparing my initial conception to criteria I have<br />
learned, do I still consi<strong>der</strong> this information/this Web<br />
site to be credible? Why or why not?<br />
• Teachers should provide opportunities for continued<br />
discussions <strong>of</strong> successes or failures and suggestions for<br />
improving credibility determinations. Simple<br />
questions such as the following can elicit this<br />
information: What’s working well? What isn’t? What<br />
suggestions do you have for improvement?<br />
VI. CONCLUSION<br />
This paper has summarized a line <strong>of</strong> research<br />
investigating students’ Web site credibility determinations. It<br />
has argued for the need for critical Web credibility instruction<br />
at all levels and has summarized research in epistemological<br />
beliefs about information from the Internet, and conceptual<br />
change in science learning. By drawing on these perspectives,<br />
instructional recommendations have been synthesized. These<br />
recommendations provide a foundation for further<br />
development in instruction and research that addresses the<br />
need for critical information credibility determinations when<br />
using Web-based information.<br />
REFERENCES<br />
[1] E.B. Klemm and M. Iding, Do scientists amd teachers agree on the<br />
credibility <strong>of</strong> media information sources, International Journal <strong>of</strong><br />
Instructional Media, vol. 28, 2001, pp. 83-91.
[2] M. Iding, R. Landsman, and T. Nguyen. Critical evaluation <strong>of</strong> scientific<br />
websites by high school students. In D. Watson & J. An<strong>der</strong>son (Eds.),<br />
Networking the Learner: Computers in Education: Seventh IFIP World<br />
Conference on Computers in Education Conference Proceedings.<br />
Dordrecht, Netherlands: Kluwer Academic Publishers, pp. 373-382,<br />
2002.<br />
[3] M. Iding, M.E. Crosby, B. Auernheimer, and E. B. Klemm, E.B. Web<br />
site credibility: Why do people believe what they believe? Instructional<br />
Science, 37, 2009, pp. 43-63<br />
[4] M. Iding, J. Nordbotten, and J.M. Singh, Credibility: Norwegian<br />
students evaluate media studies Web sites. In D. Kuma & J. Turner<br />
(Eds.), Education for the 21 st century: Impact <strong>of</strong> ICT and digital<br />
resources: WCC 2006 Santiago, Chile, (pp. 327-331). New York:<br />
Springer, 2006.<br />
[5] E. G. Benotsch, S. Kalichman, and L.S. Weinhardt. HIV–AIDS patients’<br />
evaluation <strong>of</strong> health information on the Internet: The digital divide and<br />
vulnerability to fraudulent claims. Journal <strong>of</strong> Consulting and Clinical<br />
Psychology, 72(6), 2004, pp. 1004-1011.<br />
[6] Fogg, B.J., Soohoo, C., Danielson, D., Marable, L., Stanford, J. &<br />
Tauber, E. R. 2002 How do people evaluate a Web site’s credibility:<br />
Results from a large study.<br />
http://www.consumerwebwatch.org/dynamic/web-credibility-reportevaluate.cfm<br />
66<br />
[7] I. Braten, H. I. Stromso, and M. Samuelstuen. The relationship between<br />
Internet-specific epistemological beliefs and learning within Internet<br />
technologies. J. Educational Computing Research, vol. 33(2), 2005, pp.<br />
141-171.<br />
[8] B. K. H<strong>of</strong>er, and P. R. Pintrich. The development <strong>of</strong> epistemological<br />
theories: Beliefs about knowledge and knowing and their relationship to<br />
learning. Review <strong>of</strong> Educational Research, vol. 67, 1997, pp. 88-140.<br />
[9] C.A. Chinn, and W. F. Brewer. The role <strong>of</strong> anomalous data in<br />
knowledge acquisition: A theoretical framework and implications for<br />
science instruction. Review <strong>of</strong> Educational Research, vol. 63, 1993, pp.<br />
1-49<br />
[10] P.R. Pintrich, R.W. Marx, and R. A Boyle. Beyond cold conceptual<br />
change: The role <strong>of</strong> motivational beliefs and clssroom contextual factors<br />
in the process <strong>of</strong> conceptual change. Review <strong>of</strong> Educational Research,<br />
vol. 63, 1993, pp. 167-199.<br />
[11] D. Schunk. Social cognitive theory and sefl-regulated learning. In B.<br />
Zimmerman and D. Schunk (Eds.), Self-regulated learning and academic<br />
achievement: Theory, research and practice, New York: Springer, 1989,<br />
pp. 83-110.
How Teachers in Different Educational Systems Value<br />
Central Concepts <strong>of</strong> Computer Science<br />
Peter Hubwieser<br />
Technische Universität München<br />
TUM School <strong>of</strong> Education<br />
Fakultät für Informatik<br />
Boltzmannstr. 3. D-85478 Garching<br />
+49 89 289 17350<br />
Peter.Hubwieser@tum.de<br />
ABSTRACT<br />
The 16 German states exhibit substantial differences regarding the<br />
organization as well as the substantial focus <strong>of</strong> computer science<br />
education at their schools. This empirical study investigates how<br />
teachers from two German states with different educational systems<br />
assess the value <strong>of</strong> central concepts <strong>of</strong> computer science. We<br />
asked 120 teachers in each country to complete our questionnaire,<br />
received 38 responses and applied a specific split-plot design to<br />
evaluate the results. The findings show that the assessments by the<br />
two groups differ regarding the content concepts model, system,<br />
computer, and information. Additionally, we detected differences<br />
in the rating <strong>of</strong> some individual process concepts (analyzing,<br />
classifying, finding relationships, generalizing, comparing, and<br />
or<strong>der</strong>ing) in relation to the content concept model. These results<br />
are consistent with the differences in the focus <strong>of</strong> the curricula as<br />
well as with the content <strong>of</strong> the teacher education programs in the<br />
two states.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computer and Information Science Education ]: Computer<br />
science education.<br />
General Terms<br />
Human Factors<br />
Keywords<br />
computer science education, teacher education, subject domain<br />
knowledge, central concepts.<br />
1. INTRODUCTION<br />
The educational systems <strong>of</strong> the 16 German states exhibit substantial<br />
structural differences. This applies to the implementation <strong>of</strong><br />
computer science education in schools as well as to the teacher<br />
education programs. This empirical study investigates how teachers<br />
from two German states differ in their evaluations <strong>of</strong> central<br />
concepts <strong>of</strong> computer science that are given particular emphasis in<br />
the curricula respectively in the teacher training programs <strong>of</strong> these<br />
states.<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that<br />
copies bear this notice and the full citation on the first page. To copy<br />
otherwise, or republish, to post on servers or to redistribute to lists,<br />
requires prior specific permission and/or a fee.<br />
Conference’10, Month 1–2, 2010, City, State, Country.<br />
Copyright 2010 ACM 1-58113-000-0/00/0010 …$15.00.<br />
Andreas Zendler<br />
<strong>University</strong> <strong>of</strong> Education Ludwigsburg<br />
Institute <strong>of</strong> Mathematics and Computer<br />
Science<br />
Reuteallee 46 D-71634 Ludwigsburg<br />
+49 7141 140 683<br />
zendler@ph-ludwigsburg.de<br />
67<br />
The provision <strong>of</strong> effective computer science education (CSE) in<br />
schools requires computer science (CS) teachers who are able to<br />
draw on three types <strong>of</strong> pr<strong>of</strong>essional knowledge [41], [48], [7, 26],<br />
[1]:<br />
(1) disciplinary knowledge <strong>of</strong> the subject taught (content<br />
knowledge),<br />
(2) general pedagogical knowledge (pedagogical knowledge),<br />
and<br />
(3) knowledge <strong>of</strong> how to teach specific subject matter (pedagogical<br />
content knowledge).<br />
A crucial tasks for teacher training programs is to define what the<br />
subject <strong>of</strong> computer science is and what it is not: “One <strong>of</strong> the<br />
challenges we face when discussing computer science education is<br />
that the field <strong>of</strong> computer science seems to progress so quickly<br />
that is difficult even for computer scientists to clearly define its<br />
contents and proscribe its boundaries” [12].<br />
This study builds on different approaches to teacher training<br />
programs (See sections “Theoretical Background” and “Related<br />
Work”). Specifically, we investigated whether and how computer<br />
science teachers differ from the two different German states Baden-Württemberg<br />
(BW) and Bavaria (BY) in their evaluations <strong>of</strong><br />
the relationships between central content concepts and central<br />
process concepts <strong>of</strong> the discipline. We also examine whether<br />
Bavarian computer science teachers evaluate the concepts given in<br />
their training programs (data, model, process, structure, information)<br />
[43] as being <strong>of</strong> particular importance. By addressing<br />
these research questions, we hope to provide insights that can<br />
inform the pre- and in-service training <strong>of</strong> computer science teachers.<br />
2. THEORETICAL BACKGROUND<br />
In the last decade the research-based approach to teacher education<br />
has been gaining ground [24], [35], [12], [11] . It is characterized<br />
by three main desi<strong>der</strong>ata:<br />
(1) teachers need in-depth knowledge <strong>of</strong> recent research advances<br />
in the subject they teach,<br />
(2) teacher training should itself be an object <strong>of</strong> research,<br />
(3) teachers need to internalize a research-based approach to<br />
their work.<br />
In implementing curricular reforms, moreover, it is important to<br />
consi<strong>der</strong> a bottom-up approach [10], [14] that explicitly draws on<br />
teachers’ knowledge and perspectives: “To un<strong>der</strong>stand the role <strong>of</strong><br />
teachers with respect to educational reform, it has been suggested<br />
that their beliefs and views (…), or their practical knowledge (…)<br />
be analyzed” (see [10], p. 138). Fincher and Tenenberg [13] as<br />
well as Ni [34] have discussed the application <strong>of</strong> the bottom-up<br />
approach to training programs for computer science teachers.
Deciding on the subject matter to be taught is a key challenge for<br />
those involved in the design <strong>of</strong> computer science education [44].<br />
The spectrum currently ranges from training the use <strong>of</strong> computer<br />
<strong>of</strong>fice s<strong>of</strong>tware to programming courses or to the theoretical solution<br />
<strong>of</strong> algorithmic problems. Teachers’ uncertainty is evident in<br />
the fact that course content tends to reflect current trends and<br />
developments and to draw on short-lived product knowledge [16].<br />
Given the rapid pace <strong>of</strong> development in the field <strong>of</strong> information<br />
technology, however, this knowledge soon becomes obsolete.<br />
Instead, computer science education should equip students with<br />
knowledge and skills that will remain relevant in the longer term,<br />
that they can use in their everyday lives, and that are to some<br />
extent representative <strong>of</strong> the subject [33].<br />
The contents to be covered in computer science education have<br />
previously been discussed primarily in the context <strong>of</strong> fundamental<br />
ideas [39], first introduced by [5]. According to [39], a fundamental<br />
idea is a scheme <strong>of</strong> thinking, action, description, or explanation<br />
that satisfies four criteria: It must be relevant in multiple domains<br />
<strong>of</strong> a discipline (horizontal criterion). It must be teachable on every<br />
intellectual level (vertical criterion). It must remain relevant in the<br />
longer term (time criterion). And it must be related to everyday<br />
language and/or thinking (sense criterion). Arguments based on<br />
these criteria have been applied also to justify the learning content<br />
covered in the new school subject <strong>of</strong> computer science in Bavaria.<br />
Additionally, however, it has been postulated that the modeling <strong>of</strong><br />
real or planned systems using special representation techniques<br />
from s<strong>of</strong>tware engineering (e.g., object and class diagrams or data<br />
flow charts) is <strong>of</strong> particular importance in the context <strong>of</strong> general<br />
education [19]. Computer science instruction is expected to explain<br />
the basic functioning <strong>of</strong> computer systems, and the abstract<br />
representation <strong>of</strong> these systems in the form <strong>of</strong> such models is<br />
expected to further this aim. Moreover, computer science instruction<br />
is expected to develop students’ ability to structure complex<br />
systems, which is clearly also the goal <strong>of</strong> such modeling activities<br />
[4].<br />
3. RELATED WORK<br />
Several scientists have proposed catalogues <strong>of</strong> the basic concepts<br />
or fundamental ideas <strong>of</strong> computer science [36], [39]; [2]; [8], [27],<br />
[17], [47].<br />
However, these catalogues have a number <strong>of</strong> shortcomings in<br />
terms <strong>of</strong> validity:<br />
(1) they are based on the subjective judgments <strong>of</strong> a single author<br />
or small group <strong>of</strong> authors,<br />
(2) they lack empirical verification,<br />
(3) they relate only to subdomains <strong>of</strong> computer science,<br />
(4) their validity has been established only for the national context<br />
in which they were developed.<br />
We have addressed points (1) to (3) in previous research. Specifically,<br />
the results <strong>of</strong> our three studies [49], [50], [51] were based<br />
on the judgments <strong>of</strong> a larger sample <strong>of</strong> experts, were empirically<br />
<strong>der</strong>ived, and related to various subdomains <strong>of</strong> computer science.<br />
In the current discussion on curriculum development in computer<br />
science education, the combination <strong>of</strong> two scientifically informed<br />
oriented approaches is consi<strong>der</strong>ed crucial: first, the structure <strong>of</strong> the<br />
discipline approach introduced by [5]; second, the process as<br />
content approach, based on the work <strong>of</strong> Parker and Rubin [37],<br />
which has more recently enjoyed a renaissance thanks to Costa<br />
and Liebmann [6].<br />
In 2003, Meyer and Land introduced the idea <strong>of</strong> threshold concepts,<br />
core concepts in a discipline that are transformative, irre-<br />
68<br />
versible and integrative [28]. Recently Sorva presented a comprehensive<br />
overview <strong>of</strong> research about this issue [42], while San<strong>der</strong>s<br />
et al. investigated the relations between threshold concepts and<br />
threshold skills in computing [38].<br />
In the study by Zendler, Spannagel, and Klaudt [51], which drew<br />
on these two approaches and on a constructivist theory <strong>of</strong> learning<br />
[3], [29]. For the study 24 German computer science pr<strong>of</strong>essors<br />
rated the relevance <strong>of</strong> 15 content concepts (e.g., algorithm, problem,<br />
and model) with respect to 16 process concepts (e.g., analyzing,<br />
categorizing, and classifying) on a 6-point scale from (“no<br />
importance”) to 5 (“great importance”). The main finding <strong>of</strong> the<br />
study was that there are specific groups <strong>of</strong> content concepts that<br />
should be taught in combination with specific groups <strong>of</strong> process<br />
concepts. In total, 15 blocks <strong>of</strong> content and process concepts were<br />
identified as being particularly relevant (e.g., the blocks <strong>of</strong> the<br />
content concepts algorithm, data, information, problem, model,<br />
and structure in combination with the process concepts categorizing,<br />
classifying, finding cause-and-effect relationships, and generalizing).<br />
In 2009 we have investigated the view <strong>of</strong> active teachers on the<br />
recently installed compulsory subject <strong>of</strong> informatics in the state <strong>of</strong><br />
Bavaria [32], [22]. Regarding the valuation <strong>of</strong> CS concepts we<br />
found that the teachers could be assigned to one <strong>of</strong> three different<br />
clusters as far as their valuation <strong>of</strong> the CS concepts is concerned<br />
[32]:<br />
(1) “<strong>of</strong>fice users”, concentrating on the application <strong>of</strong> s<strong>of</strong>tware<br />
systems,<br />
(2) “fans <strong>of</strong> the curriculum“ and “anti-programmers”, preferring<br />
object-oriented modeling (OOM) instead <strong>of</strong> programming<br />
and algorithmic concepts,<br />
(3) “traditional computer scientists” that focus on traditional<br />
algorithmic views instead <strong>of</strong> OOM.<br />
4. COMPUTER SCIENCE EDUCATION<br />
4.1 General Education in BY and BW<br />
The educational systems <strong>of</strong> BY and BW are quite similar, at least<br />
as far as the structure <strong>of</strong> the general education is concerned (see<br />
Fig. 1). In both states the 4-yeared primary education takes place<br />
in the Grundschule. Following this, the students split according to<br />
their performance level in 3 types <strong>of</strong> secondary schools:<br />
(1) Gymnasium (8 years),<br />
(2) Realschule (6 years),<br />
(3) Hauptschule (5-6 years).<br />
Figure 1. School System in BW and BY<br />
As our data were gathered among teachers at Gymnasium, we will<br />
restrict the following explanations to this school type, which is
attended by about a 33-40 % <strong>of</strong> all school students in grade 5<br />
currently.<br />
4.2 CSE in BY and BW<br />
By the occasion <strong>of</strong> the extensive structural reform project <strong>of</strong> the<br />
Bavarian Gymnasium in 2004 (from G9 to G8) a new (at least<br />
partly) compulsory subject <strong>of</strong> CS was incorporated. Fig. 2 displays<br />
the organization <strong>of</strong> the new subject and the years it has<br />
started in the different grades. In grade 6 and 7, CS is incorporated<br />
formally in the subject combination Nature and Technology<br />
(NuT) that comprises the Biology (in grade 5 and 6), Physics (in<br />
grade 7) and CS (in grade 6 and 7). Nevertheless, all three subjects<br />
are taught separated in a prescribed number <strong>of</strong> lessons per<br />
week by teachers that need a university degree in the respective<br />
subject.<br />
Figure 2. The subject <strong>of</strong> CS in Bavarian Gymnasiums<br />
In the Science & Technology direction <strong>of</strong> study, the students have<br />
to attend CS as a compulsory subject in grade 9 and 10. In average<br />
about 50% <strong>of</strong> the students choose this direction. In grade 11 and<br />
12, an elective CS course might be chosen instead <strong>of</strong> a second<br />
natural science or a second foreign language. In the final examination,<br />
CS can be chosen for written as well as for oral examination.<br />
The learning topics <strong>of</strong> the subject <strong>of</strong> CS are prescribed in specific<br />
curricula, which we have described in several publications, e.g.<br />
[18], [19], [20].<br />
In contrary to the situation in Bavaria, there is no compulsory<br />
subject <strong>of</strong> CS in BW. Despite there are working groups beside the<br />
regular schedule up to grade 10. From this point <strong>of</strong> time, the<br />
students can choose an elective course that is designed according<br />
to the proposed educational standards <strong>of</strong> the German Gesellschaft<br />
für Informatik [15].<br />
5. TEACHER EDUCATION<br />
Regarding the curricula <strong>of</strong> regular teacher education programs,<br />
there aren’t big differences between the two states, at least as far<br />
as the teachers at Gymnasiums are concerned. The education takes<br />
place at the Universities in contrary to the teachers for the other<br />
school types in BW, who are educated at specific pedagogical<br />
colleges (Pädagogische Hochschulen). The content <strong>of</strong> the teacher<br />
education in CS was standardized in 2008 by the German Kultusministerkonferenz<br />
(KMK), the national board <strong>of</strong> the 16 secretary<br />
<strong>of</strong> states that are in charge for the schools, which worked out<br />
Standards for Teacher Education [40]. Thus we can assume that<br />
the subject knowledge content <strong>of</strong> the regular teacher education<br />
courses is not much different in the two states.<br />
Nevertheless, in Bavaria, many <strong>of</strong> the teachers that are active<br />
teaching currently were educated in specific courses that were<br />
installed in or<strong>der</strong> to support the introduction <strong>of</strong> the new compulsory<br />
subject. At this time, the dilemma associated with all new<br />
subjects has to be addressed: On the one hand, if teachers are<br />
69<br />
trained before a new subject is introduced, there is a risk that they<br />
will not find employment after graduation. On the other hand, a<br />
new subject can be successfully introduced only if sufficient<br />
numbers <strong>of</strong> appropriately trained teachers are available. When<br />
computer science was first implemented as a compulsory subject<br />
in academic-track schools (called Gymnasia) in the southern<br />
German state <strong>of</strong> Bavaria, this dilemma was resolved by providing<br />
specific in-service training programs for teachers who had already<br />
been appointed to teach two other subjects before [43].<br />
These programs emphasized the importance <strong>of</strong> modeling techniques<br />
in computer science. In particular, the modeling <strong>of</strong> processes<br />
[21] and object-oriented modeling formed the core <strong>of</strong> the<br />
SIGNAL and FLIEG in-service-courses attended by the majority<br />
<strong>of</strong> the currently active computer science teachers in Bavaria, who<br />
have received further training to date. Specifically, the courses<br />
contained the following modules M1 to M8:<br />
- M1: Data modeling and database systems<br />
- M2: Modeling processes<br />
- M3: Object-oriented modeling and programming<br />
- M4: Algorithms and data structures<br />
- M5: S<strong>of</strong>tware technology<br />
- M6: Technological computer science, data security<br />
- M7: Theoretical computer science<br />
- M8: The teaching <strong>of</strong> computer science.<br />
As the content <strong>of</strong> these specific courses is well-tailored to the<br />
curriculum <strong>of</strong> the Bavarian subject <strong>of</strong> CS, there might exist substantial<br />
differences to the regular teacher education in BW.<br />
6. METHODOLOGY<br />
6.1 Study design<br />
The hypothesis was tested in a 2•15×16 split-plot design (3-factor<br />
design with repeated measures <strong>of</strong> factors B and C, see Fig. 3; [46];<br />
[25]. Factor A comprised the p = 2 groups surveyed, with factor<br />
level a1 representing group G1 <strong>of</strong> n1 Baden-Württemberg (BW)<br />
teachers <strong>of</strong> computer science and factor level a2 representing<br />
group G2 <strong>of</strong> n2 Bavarian (BY) teachers <strong>of</strong> computer science teachers.<br />
Factor B represented the q = 15 content concepts (CC) b 1, ...,<br />
b15: problem, information, model, algorithm, data, structure,<br />
system, computation, process, s<strong>of</strong>tware, program, test, communication,<br />
language, and computer. Factor C consisted <strong>of</strong> the r = 16<br />
process concepts (PC) c 1, ..., c 16: analyzing, classifying, problem<br />
solving and problem posing, categorizing, investigating, finding<br />
relationships, generalizing, creating and inventing, comparing,<br />
finding cause-and-effect relationships, questioning, transferring,<br />
communicating, presenting, collaborating, and or<strong>der</strong>ing.<br />
…<br />
…<br />
…<br />
…<br />
…<br />
…<br />
…<br />
Figure 3. Layout <strong>of</strong> the 2•15×16 split-plot design
While the factors A, B and C were regarded as independent variables,<br />
the dependent variable was the respondents’ evaluation <strong>of</strong><br />
the importance <strong>of</strong> a specific process concept for a specific content<br />
concept. Ratings were given on a 6-point scale from 0 (“no importance”)<br />
to 5 (“great importance”).<br />
6.2 Power analysis<br />
A power calculation <strong>of</strong> type II, N being a function <strong>of</strong> power (1–β),<br />
Δ, and α, was used to determine the necessary sample size for the<br />
2•15×16 split-plot design (see [31]): With a power (1–β) <strong>of</strong> 0.99,<br />
only large effects (Δ = 0.80) on the dependent variable being<br />
consi<strong>der</strong>ed significant, and a significance level <strong>of</strong> α = 0.05, a total<br />
sample <strong>of</strong> approximately N*= 30 ( n1* = 15 Baden-Württemberg<br />
teachers <strong>of</strong> computer science teachers, n 2* = 15 Bavarian teachers<br />
<strong>of</strong> computer science teachers), would be required, based on the<br />
power computations <strong>of</strong> Mueller and Barton [30] or Mueller et al.<br />
[31] for ε-corrected F tests.<br />
6.3 Operational hypothesis<br />
Given the study design and the above specification <strong>of</strong> the independent<br />
and dependent variables, the operational hypothesis <strong>of</strong> the<br />
study can be formulated as follows: “CS teachers from BW differ<br />
from CS teachers from BY in their evaluations <strong>of</strong> the relations<br />
between central content concepts <strong>of</strong> computer science (CC, see<br />
section 2.1) and central process concepts <strong>of</strong> computer science<br />
(PC, see section 2.1), as operationalized by their rating on a sixpoint<br />
scale <strong>of</strong> the importance <strong>of</strong> a specific process concept for a<br />
specific content concept."<br />
6.4 Sampling<br />
A total <strong>of</strong> 120 CS teachers in BW and 120 CS teachers in BY<br />
were contacted and invited to complete a questionnaire pertaining<br />
to computer science concepts. The questionnaire began with a<br />
short introduction, in which the 15 central content concepts and<br />
the 16 central process concepts were listed in tabular form in<br />
alphabetical or<strong>der</strong>. Following this, the q = 15 content concepts and<br />
the r = 16 process concepts were presented in alphabetical or<strong>der</strong><br />
in a matrix, with the content concepts in the rows and the process<br />
concepts in the columns. Participants were asked to rate the following<br />
statement for each <strong>of</strong> the 15×16 = 240 cells <strong>of</strong> the matrix:<br />
Each cell represents a combination <strong>of</strong> a concept and a process and<br />
requires an integer from 0 (no importance) to 5 (great importance)<br />
indicating the relevance <strong>of</strong> the combination. Participants filled in<br />
each cell on a 6-point scale from 0 (“no importance”) to 5 (“great<br />
importance”).<br />
To maximize the return rate, we mailed both samples the questionnaires<br />
in sealed, personalized envelopes, enclosing a preaddressed<br />
return envelope franked with stamps showing flower<br />
designs (see [9] for recommendations on increasing return rates).<br />
The return rate for the BW teachers was 14.2% (n1 = 17 valid<br />
questionnaires), which can be consi<strong>der</strong>ed reasonable for a postal<br />
survey (see [45]). The return rate for the BY teachers was 17.5%<br />
(n2 = 21 valid questionnaires).<br />
6.5 Data Analysis<br />
The procedure used to analyze the experimental data was as follows:<br />
First, we performed a descriptive analysis (7.1-7.2), focusing<br />
on the content concepts. Second, we conducted a three-factor<br />
analysis <strong>of</strong> variance with repeated measures (7.3) in accordance<br />
with the SPF-2•15×16 split-plot design (see Winer [46], chapter<br />
7). Third, we conducted a posteriori comparisons <strong>of</strong> means to test<br />
for significant effects <strong>of</strong> the A × B interaction (7.4) and the A × B<br />
× C interaction (7.5). The process concepts were included in<br />
analyses (2) to (4).<br />
70<br />
Data analyses were conducted using SPSS 17.0; the power analysis<br />
was computed with PASS 8.0.9.<br />
7. RESULTS<br />
Fig. 6 in the appendix displays the original data <strong>of</strong> the mean ratings<br />
obtained from the BW teachers (a 1) and the BY teachers (a 2)<br />
for each <strong>of</strong> the 15 × 16 combinations <strong>of</strong> content concepts × process<br />
concepts (repeated measures factors B × C).<br />
7.1 Means<br />
The four content concepts with the highest averages (see Appendix)<br />
are the same for the two groups <strong>of</strong> teachers: problem, information,<br />
model and algorithm. The concept with the lowest average<br />
is also the same, namely computer. Major differences in the<br />
assessment <strong>of</strong> content concepts between the two groups <strong>of</strong> teachers<br />
can be found for the content concepts information (2.79 vs.<br />
3.11), model (2.62 vs. 3.30), system (1.90 vs. 2.44) and computer<br />
(1.62 vs. 2.04).<br />
7.2 Process-related coverage<br />
To determine differences in the assessment <strong>of</strong> content and process<br />
concepts by the two groups <strong>of</strong> teachers, the process-related coverage<br />
can be used: A content concept has high process-related coverage<br />
if it is rated as highly important (> 2.50) for many <strong>of</strong> the<br />
process concepts; it has lower process-related coverage if it is<br />
rated as less important (≤ 2.50) for many <strong>of</strong> the process concepts.<br />
It is striking that the content concepts information, model and<br />
system are rated highly in relation to more process concepts by the<br />
BY teachers compared to the BW teachers. On the other hand,<br />
teachers from Bavaria rate more process concepts highly in combination<br />
with the content concepts information (14 vs. 11), model<br />
(15 vs. 12) and system (6 vs. 1) compared to the BW teachers,<br />
while the latter rate more process concepts highly in combination<br />
with computation (8 vs. 3) and program (8 vs. 5).<br />
7.3 Analysis <strong>of</strong> Variance<br />
To examine whether the BW teachers differed from the BY teachers<br />
in their evaluations <strong>of</strong> the relationships between the content<br />
concepts and the process concepts, we formulated three statistical<br />
hypotheses, which were tested at the significance level <strong>of</strong> α =<br />
0.05. The three null hypotheses were chosen as follows:<br />
i) The means <strong>of</strong> the content concepts µ1 un<strong>der</strong> factor level a1<br />
(BW teachers) and µ2 un<strong>der</strong> factor level a2 (BY teachers) are<br />
equal, such that:<br />
H 0: µ 1 = µ 2.<br />
ii) The means <strong>of</strong> the content concepts µ1�1, µ1�2, ..., µ2�15 un<strong>der</strong><br />
the 2 � 15 levels <strong>of</strong> the factor combinations A × B are equal,<br />
such that:<br />
H0: µ1�1 = µ1�2 = ... = µ2�15.<br />
iii) The means <strong>of</strong> the content concepts µ1�1×1, µ1�1×2, ..., µ2�15×16<br />
un<strong>der</strong> the 2 � 15 ×16 levels <strong>of</strong> the factor combinations A × B<br />
× C are equal, such that:<br />
H0: µ1�1×1 = µ1�1×2 = ... = µ2�15×16.<br />
For an analysis <strong>of</strong> variance <strong>of</strong> a split-plot design, the data must<br />
satisfy the condition <strong>of</strong> sphericity. This assumption was tested<br />
using Mauchly’s W test for sphericity, with the test statistic W<br />
being compared to a chi-square distribution to assess the adequacy<br />
<strong>of</strong> the sphericity assumption.
The assumption <strong>of</strong> sphericity was not met for either the content<br />
concepts (W = 0.003, χ 2 104 = 178.35, p < 0.001) or the process<br />
concepts (W < 0.001, χ 2 119 = 248.60, p < 0.001) at the α level <strong>of</strong><br />
0.05. In the further analyses, we therefore applied the ε correction<br />
<strong>of</strong> degrees <strong>of</strong> freedom proposed by [23], as presented in table 1.<br />
Table 1. Results <strong>of</strong> the ANOVA with Huynh–Feldt ε correction<br />
<strong>of</strong> degrees <strong>of</strong> freedom<br />
The main effect A (BW vs. BY teachers) was not significant at the<br />
α level <strong>of</strong> 0.05 (F1, 36 = 0.24, p < 0.63). The corresponding H0 was<br />
therefore not rejected: The teachers from BW respectively BY did<br />
not differ in their global evaluations <strong>of</strong> the content concepts.<br />
The interaction effect A × B (group × content concept) was significant<br />
at the α level <strong>of</strong> 0.05 (F10, 360 = 2.26, p < 0.02). The corresponding<br />
H0 was therefore rejected: The teachers from BW respectively<br />
BY differed significantly in their evaluations <strong>of</strong> individual<br />
content concepts.<br />
The interaction effect A × B × C (group × content concept ×<br />
process concept) was not significant at the α level <strong>of</strong> 0.05 (F78, 2809<br />
= 1.09, p < 0.29). The corresponding H0 was therefore not rejected:<br />
The teachers from BW respectively BY did not differ in their<br />
evaluations <strong>of</strong> the relationships between individual content concepts<br />
and individual process concepts.<br />
7.4 Individual Comparisons for the A × B Interactions<br />
The global test <strong>of</strong> the A × B interaction revealed a significant<br />
overall effect <strong>of</strong> group × content concept. Therefore we evaluated,<br />
which concepts were rated differently by comparing the mean<br />
values, applying t-tests in or<strong>der</strong> to test simple AB effects for<br />
p•q×r split-plot designs (see [46], pp. 535–536), concerning the ε<br />
correction <strong>of</strong> the degrees <strong>of</strong> freedom (see section 7.3). Fig. 4<br />
visualizes the means <strong>of</strong> the A × B interaction. As the figure shows,<br />
the content concept model was rated significantly different by the<br />
two groups <strong>of</strong> teachers (a1, a2) at the α level <strong>of</strong> 0.05 (t396 = 2.23, p<br />
< 0.027). The differences for system, computer, and information<br />
are remarkable, but not significant.<br />
7.5 Individual Comparisons for the A × B × C<br />
Interaction<br />
The global test <strong>of</strong> the A × B × C interaction did not reveal a significant<br />
overall effect <strong>of</strong> group × content concept × process concept.<br />
Taking into account the significant difference regarding the<br />
content concept model (see section 7.4), it makes sense to compare<br />
only this concept with respect to the process concepts a<br />
posteriori. Fig. 5 displays the comparisons <strong>of</strong> the means on the<br />
concept model regarding the different process concepts.<br />
71<br />
These were calculated by applying 16 t-tests to analyze simple<br />
AC effects for SPF-p•q×r experimental designs ([46] pp. 535-<br />
536); An ε correction <strong>of</strong> degrees <strong>of</strong> freedom was taken into account<br />
again (see above). The t-tests were calculated at an adjusted<br />
α level <strong>of</strong> 0.05/16 = 0.0031. It turned out that the ratings from BW<br />
teachers (a 1) respectively from BY teachers (a 2) differed significantly<br />
regarding the content concept model related to the following<br />
process concepts: classifying, finding relationships, generalizing,<br />
comparing, questioning, and or<strong>der</strong>ing.<br />
Content concepts<br />
b1 = problem<br />
b2 = information<br />
b3 = model<br />
b4 = algorithm<br />
b5 = data<br />
b6 = structure<br />
b7 = system<br />
b8 = computation<br />
b9 = process<br />
b10 = s<strong>of</strong>tware<br />
b11 = program<br />
b12 = test<br />
b13 = communication<br />
b14 = language<br />
b15 = computer<br />
0.0<br />
1.75<br />
Rating<br />
2.0 2.25 2.5 2.75 3.0 3.25 3.5<br />
Groups<br />
a1= BW teachers <strong>of</strong> computer science<br />
a2= BY teachers <strong>of</strong> computer science<br />
a1<br />
a2<br />
__<br />
simple AB effects<br />
0.85-<br />
0.63-0.84<br />
0.53-0.62<br />
0.31-0.52<br />
0.16-0.30<br />
0.00-0.15<br />
Figure 4. Comparisons for the factor level combinations A × B<br />
Rating<br />
4.0 4.5<br />
4.25<br />
3.75<br />
2.0 2.25 2.5 2.75 3.0 3.25 3.5<br />
1.75<br />
c1 = analyzing<br />
c2 = classifying<br />
c3 = problem solving and posing<br />
c4 = categorizing<br />
c6 = finding relationships<br />
c5 = investigating<br />
c7 = generalizing<br />
c8 = creating and inventing<br />
c9 = comparing<br />
c10 = finding cause-and-effect r.<br />
c11 = questioning<br />
c12 = transferring<br />
c13 = communicating<br />
c14 = presenting<br />
c15 = collaborating<br />
c16 = or<strong>der</strong>ing<br />
p < .0006<br />
p < .003<br />
p < .006<br />
1.05-<br />
0.91-1.04<br />
0.67-0.90<br />
0.45-0.66<br />
0.23-0.44<br />
0.00-0.22<br />
p < .01<br />
p < .05<br />
p < .10<br />
Figure 5. Comparisons for the content concept model<br />
8. DISCUSSION<br />
The results <strong>of</strong> the performed evaluations support the research<br />
hypothesis that computer science teachers from Baden-<br />
Württemberg differ from computer science teachers from Bavaria<br />
in the assessment <strong>of</strong> key content concepts <strong>of</strong> computer science<br />
related to central process concepts <strong>of</strong> computer science. Already<br />
from the descriptive evaluation, it has become clear that there are<br />
differences in the assessment <strong>of</strong> content concepts by the two<br />
groups <strong>of</strong> computer science teachers. There were differences for<br />
the concepts <strong>of</strong> content model, system, computer, and information.<br />
The analysis <strong>of</strong> variance and the individual comparisons showed<br />
that the two groups rated the individual process concepts classifying,<br />
finding relationships, generalizing, comparing, questioning,<br />
and or<strong>der</strong>ing differently with respect to the content concept model.<br />
As the regular teacher education programs in the German states<br />
are standardized by the KMK (see section 5), it is not likely that<br />
those differences would be caused by the courses <strong>of</strong> lessons in CS<br />
that the teachers had attended at their universities. Therefore, the
eason has to be sought outside the regular programs. In Bavaria<br />
this might have been a specific in-service program like SIGNAL<br />
or FLIEG (see section 5), attended by many Bavarian teachers<br />
instead <strong>of</strong> regular programs, a specific didactical training programs<br />
for the new Bavarian subject <strong>of</strong> CS or the daily teaching<br />
environment at schools. It would be plausible to assume one <strong>of</strong><br />
those three reasons because all those courses or environments<br />
focus on modeling and thus might cause a quite special view on<br />
the concept model as well as on the salient process concepts, e.g.<br />
classifying or finding relationship.<br />
On the other hand, there are certainly some threats to the validity<br />
<strong>of</strong> our results. Firstly, the question <strong>of</strong> the interview could be more<br />
precise, e.g. asking to assess the value <strong>of</strong> the concepts for CS<br />
courses in school or as background <strong>of</strong> the teachers. A stimulating<br />
text passage before the question might help also. Secondly, the<br />
teachers had not much time to think about the 240 combinations<br />
<strong>of</strong> content and process concepts.<br />
9. CONCLUSION AND FUTURE WORK<br />
Suggested by this comparison <strong>of</strong> two different states, we aim to<br />
conduct a larger study, interviewing teachers in <strong>Germany</strong> about<br />
their valuation <strong>of</strong> content and process concepts in the future.<br />
Our results show that there are significant differences in the valuation<br />
<strong>of</strong> certain concepts between the teachers <strong>of</strong> those two states.<br />
It would be very interesting to investigate the perceptions <strong>of</strong> those<br />
concepts by the teachers, trying to detect if the differences concern<br />
only the degree <strong>of</strong> valuation or even different subject<br />
knowledge about these aspects.<br />
Further, it should be consi<strong>der</strong>ed to incorporate our results in the<br />
teacher education programs. The first step could be to analyze the<br />
coverage <strong>of</strong> the salient concepts by those programs. In the case<br />
that there are differences, those should be removed. Additionally,<br />
there might be some fuzziness in the German standards for teacher<br />
education o the KMK regarding those concepts.<br />
10. REFERENCES<br />
[1] Abell, S. K. 2007. Research on science teacher knowledge.<br />
In Handbook <strong>of</strong> research on science, S. K. Abell and N. G.<br />
Le<strong>der</strong>man, Eds. Lawrence Erlbaum, London, 1105–1149.<br />
[2] Baumann, R. 1998. Fundamentale Ideen <strong>der</strong> Informatik –<br />
gibt es das? In Informatische Bildung in Deutschland. Perspektiven<br />
für das 21. Jahrhun<strong>der</strong>t ; Festschrift für Pr<strong>of</strong>essor<br />
Dr. Wolfgang Arlt aus Anlass seines Eintritts in den Ruhestand,<br />
B. Koerber and W. Arlt, Eds. Log-In-Verl., Berlin.<br />
[3] Ben-Ari, M. 1998. Constructivism in computer science<br />
education. ACM SIGCSE Bulletin 30, 1, 257–261.<br />
[4] Breier, N. and Hubwieser, P. 2002. An Information-<br />
Oriented Approach to Informatical Education. Informatics<br />
in Education 1, 31–42.<br />
[5] Bruner, J. S. 1960. The process <strong>of</strong> education. [a searching<br />
discussion <strong>of</strong> school education opening new paths to learning<br />
and teaching]. A Vintage book V-234. Vintage Books,<br />
New York NY.<br />
[6] Costa, A. L. and Liebmann, R. M. 1997. The Processcentered<br />
school. Sustaning a renaissance community. Corwin<br />
Press, Thousand Oaks, Calif.<br />
[7] Davis, E. A. 2004. Knowledge integration in science teaching:<br />
Analyzing teachers’ knowledge development. Research<br />
in Science Education, 34 (2004), 21–53.<br />
[8] Denning, P. J. 2003. Great principles <strong>of</strong> computing. Communications<br />
<strong>of</strong> the ACM, 46(11) (2003), 15–20.<br />
72<br />
[9] Dillman, D. A. 2000. Mail and Internet surveys: the tailored<br />
design method. Wiley, New York.<br />
[10] Driel van, J. H., Beijaard, D., and Verloop, N. 2001. Pr<strong>of</strong>essional<br />
development and reform in science education: the<br />
role <strong>of</strong> teachers' practical knowledge. Journal <strong>of</strong> Research<br />
in Science Teaching, 38(2) (2001), 137–158.<br />
[11] EC – European Commission. 2009. Common european<br />
principles for teacher competences and qualifications. Accessed<br />
10 August 2009.<br />
[12] Ericson, B. 2008. Ensuring exemplary teaching in an essential<br />
discipline. Addressing the crisis in computer science<br />
teacher certification. ACM, New York.<br />
[13] Fincher, S. and Tenenberg, J. 2007. Warren's questions. In<br />
Proccedings <strong>of</strong> Third International Computing Education<br />
Research Workshop, ICER, Ed. ACM, New York, 51–60.<br />
[14] Fishman, B. and Davis, E. 2006. Teacher learning research<br />
and the learning sci-ences. In The Cambridge handbook <strong>of</strong><br />
the learning sciences, R. K. Sawyer, Ed. Cambridge <strong>University</strong><br />
Press., New York, 535–550.<br />
[15] Gesellschaft für Informatik e. V., Ed. 2008. Grundsätze und<br />
Standards für die Informatik in <strong>der</strong> Schule. Bildungsstandards<br />
Informatik für die Sekundarstufe I. Empfehlungen <strong>der</strong><br />
Gesellschaft für Informatik e. V., erarbeitet vom Arbeitskreis<br />
»Bildungsstandards«, Bonn.<br />
[16] Hartmann, W. and Nievergelt, J. 2002. Informatik und<br />
Bildung zwischen Wandel und Beständigkeit. Informatik<br />
Spektrum 25, 6, 465–476.<br />
[17] Hartmann, W., Näf, M., and Reichert, R. 2006. Informatikunterricht<br />
planen und durchführen. eXamen.press. Springer,<br />
Berlin.<br />
[18] Hubwieser, P. 2003. Object Models <strong>of</strong> IT-Systems: Supporting<br />
Cognitive Structures in Novice Courses <strong>of</strong> Informatics.<br />
In Informatics and the digital society: Social, ethical<br />
and cognitive issues ; IFIP TC3/WG 3.1 & 3.2 Open<br />
Conference on Social, Ethical, and Cognitive Issues <strong>of</strong> Informatics<br />
and ICT, July 22 - 26, 2002, <strong>Dortmund</strong>, <strong>Germany</strong>.<br />
IFIP / International Fe<strong>der</strong>ation for Information Processing.<br />
Kluwer Academic Publishers, Boston, Mass.,<br />
129‐140.<br />
[19] Hubwieser, P. 2006. Functions, Objects and States: Teaching<br />
Informatics in Secondary Schools: Invited talk. In Informatics<br />
Education - The Bridge between Using and Un<strong>der</strong>standing<br />
Computers. Lecture notes in computer science.<br />
Springer.<br />
[20] Hubwieser, P. 2007. A smooth way towards object oriented<br />
programming in secondary schools. In Informatics, Mathematics<br />
and ICT: A golden triangle: Proceedings <strong>of</strong> the<br />
Working Joint IFIP Conference: WG3.1 Secondary Education,<br />
WG3.5 Primary Education; College <strong>of</strong> Computer and<br />
Information Science, Northeastern <strong>University</strong> Boston, Massachusetts,<br />
USA 27th ‐ 29th June 2007, D. Benzie and M.<br />
Iding, Eds., Boston,MA.<br />
[21] Hubwieser, P. and Aiglstorfer, G. 2004. Fundamente <strong>der</strong><br />
Informatik: Ablaufmodellierung, Algorithmen und Datenstrukturen.<br />
Oldenbourg, München.<br />
[22] Hubwieser, P. and Mühling, A. 2009. Ausgewählte Ergebnisse<br />
einer Befragung von Lehrkräften zum Unterrichtsfach<br />
Informatik in Bayern. durchgeführt im November/Dezember<br />
2009 als Online-Umfrage unter Lehrkräften, die an bayerischen<br />
Gymnasien Informatik unterrichten.<br />
http://www.ddi.edu.tum.de/fileadmin/tueds10/www/materia<br />
l/Forschung/Lehrerbefragung_2009/Lehrerbefragung-2009-<br />
Abschlussbericht.pdf. Accessed 29 February 2012.
[23] Huynh, H. and Feldt, L. S. 1976. Estimation <strong>of</strong> the Box<br />
correction for degrees <strong>of</strong> freedom from sample data in randomised<br />
block and split-plot designs. Journal <strong>of</strong> Educational<br />
Statistics, 1 (1976), 69–82.<br />
[24] Kansanen, P. 2006. Constructing a research-based program<br />
in teacher education. In Competence oriented teacher training,<br />
F. Oser, Achtenhagen F. and U. Renold, Eds. Sense<br />
Publishers, Rotterdam, 11–22.<br />
[25] Kirk, E. 1994. Experimental design. Wadsworth, Belmont.<br />
[26] Linn, M. C., Eylon, B.-S., and Davis, E. A. 2004. The<br />
knowledge integration perspective on learning. In Internet<br />
environments for science education, M. C. Linn, E. A. Davis<br />
and P. Bell, Eds. Erlbaum, Mahwah, NJ, 29–46.<br />
[27] Loidl, S., Mühlbacher, J. R., and Schauer, H. 2005. Preparatoy<br />
knowledge: propaedeutic in informatics. In From computer<br />
literacy to informatics fundamentals, Proceedings <strong>of</strong><br />
international conference on informat-ics in secondary<br />
schools—evolution and perspectives, R. T. Mittermeir, Ed.<br />
Springer, New York, 104–115.<br />
[28] Meyer, J. H. F. and Land, R. 2003. Threshold concepts and<br />
troublesome knowledge 1: Linkages to ways <strong>of</strong> thinking<br />
and practising within the disciplines. In Improving Student<br />
Learning – Ten Years On., C. Rust, Ed. OCSLD, Oxford.<br />
[29] Moreno, L., González, C., Castilla, I., González, E. J., and<br />
Sigut, I. 2007. Use <strong>of</strong> constructivism and collaborative<br />
teaching in an ILP processors course. IEEE Transactions on<br />
Education 50, 2, 101–111.<br />
[30] Mueller, K. E. and Barton, C. N. 1989. Approximate power<br />
for repeated-measures ANOVA lacking sphericity. Journal<br />
<strong>of</strong> the American Statistical Association, 84(406) (1989),<br />
549–555.<br />
[31] Mueller, K. E., LaVange, L. E., Ramey, S. L., and Ramey,<br />
C. T. 1992. Power calcu-lations for general linear multivariate<br />
models including repeated measures appli-cations.<br />
Journal <strong>of</strong> the American Statistical Association, 87(420)<br />
(1992), 1209–1226.<br />
[32] Mühling, A., Hubwieser, P., and Brinda, T. 2010. Exploring<br />
Teachers‘ Attitudes Towards Object Oriented Modelling<br />
and Programming in Secondary Schools. In ICER ‘10: Proceedings<br />
<strong>of</strong> the Sixth international workshop on Computing<br />
education research. ACM, New York, NY, USA, 59–68.<br />
[33] Müller, P. and Hubwieser, P. 2000. Informatics as a Mandatory<br />
Subject at Secondary Schools in Bavaria. In Open<br />
Classrooms in the Digital Age, EDEN, Ed., 129‐140.<br />
[34] Ni, L. 2009. What makes CS teachers change? Factors<br />
influencing CS teachers' adoption <strong>of</strong> curriculum innovations.<br />
SIGCSE Bulletin, 41(1) (2009), 544–548.<br />
[35] Niemi, H. and Jakku-Sihvonen, R. 2006. Research-based<br />
teacher education. In Research-based teacher education in<br />
Finland, R. Jakku-Sihvonen and H. Niemi, Eds. Finnish<br />
Educational Research Association, Turku, 31–50.<br />
[36] Nievergelt, J. 1990. Computer Science for Teachers: A<br />
quest for classics and how to present them. In Computer Assisted<br />
Learning, D. H. Norrie and Six H. W., Eds. Lecture<br />
notes in computer science 438. Springer, Berlin, 2–15.<br />
[37] Parker, J. C. and Rubin, L. J. 1966. Process as content.<br />
Curriculum design and the application <strong>of</strong> knowledge. Rand<br />
McNally, Chicago.<br />
[38] San<strong>der</strong>s, K., Boustedt, J., Eckerdal, A., McCartney, R.,<br />
Moström, J. E., Thomas, L., and Zan<strong>der</strong>, C. 2012. Threshold<br />
concepts and threshold skills in computing. In Proceed-<br />
73<br />
ings <strong>of</strong> the ninth annual international conference on International<br />
computing education research. ICER ’12. ACM,<br />
New York, NY, USA, 23‐30.<br />
[39] Schwill, A. 1994. Fundamental ideas <strong>of</strong> computer science.<br />
Bulletin European Association for Theoretical Computer<br />
Science 53, 274–274.<br />
[40] Sekretariat <strong>der</strong> ständigen Konferenz <strong>der</strong> Kultusminister <strong>der</strong><br />
Län<strong>der</strong> in <strong>der</strong> Bundesrepublik Deutschland, Ed. 2008. Län<strong>der</strong>gemeinsame<br />
inhaltliche Anfor<strong>der</strong>ungen für die Fachwissenschaften<br />
und Fachdidaktiken in <strong>der</strong> Lehrerbildung. Beschluss<br />
<strong>der</strong> Kultusministerkonferenz vom 16.10.2008 i. d. F.<br />
vom 16.09.2010, Berlin.<br />
[41] Shulman, L. S. 1986. Those who un<strong>der</strong>stand: Knowledge<br />
growth in teaching. Educational Researcher, 15(2) (1986),<br />
4–14.<br />
[42] Sorva, J. 2010. Reflections on threshold concepts in computer<br />
programming and beyond. In Proceedings <strong>of</strong> the 10th<br />
Koli Calling International Conference on Computing Education<br />
Research. Koli Calling ’10. ACM, New York, NY,<br />
USA, 21‐30.<br />
[43] Spohrer, M. 2009. Konzeption und Analyse neuer Maßnahmen<br />
in <strong>der</strong> Fort- und Weiterbildung von Informatiklehrkräften.<br />
Dissertation, Technische Universität.<br />
[44] Stephenson, C., Gal-Ezer, J., Haberman, B., and Verno, A.<br />
2005. The New Educational Imperative: Improving High<br />
School Computer Science Education. Using worldwide research<br />
and pr<strong>of</strong>essional experience to improve U. S.<br />
Schools. Final Report <strong>of</strong> the CSTA Curriculum Improvement<br />
Task Force. ACM, CSTA, New York.<br />
[45] Vaux, A. and Briggs, C. S. 2005. Conducting mail and<br />
Internet surveys. In The psychology research handbook, F.<br />
T. L. Leong and J. T. Austin, Eds. Sage, Thousand Oaks,<br />
CA, 186–209.<br />
[46] Winer, B. J., Brown, D. R., and Michels, K. M. 1991. Statistical<br />
principles in ex-perimental design. McGraw-Hill,<br />
Boston.<br />
[47] Wursthorn, B. 2005. Fundamental concepts <strong>of</strong> computer<br />
science in a Logo-environment. In Digital tools for lifelong<br />
learning. Proceedings <strong>of</strong> the tenth European Logo Conference<br />
(pp. 219-227). Warsaw, Poland Centre for Informatics<br />
and Technology Education inDigital tools for lifelong<br />
learning., G. Gregorczyk, A. Walat, W. Kranas and M.<br />
Borowiecki, Eds., Warsaw, 219–227.<br />
[48] Zeidler, D. L. 2002. Dancing with maggots and saints:<br />
visions for subject matter knowledge, pedagogical<br />
knowledge, and pedagogical content knowledge in science<br />
teacher education reform. Journal <strong>of</strong> Science Teacher Education,<br />
13(1) (2002), 27–42.<br />
[49] Zendler, A. and Spannagel, C. 2008. Empirical foundation<br />
<strong>of</strong> central concepts for computer science education. ACM<br />
Journal on Educational Resources in Computing, 8(2)<br />
(2008), Article No. 6.<br />
[50] Zendler, A., Spannagel, C., and Klaudt, D. 2008. Process as<br />
content in computer science education: empirical determination<br />
<strong>of</strong> central processes. Computer Science Education,<br />
18(4) (2008), 231–245.<br />
[51] Zendler, A., Spannagel, C., and Klaudt, D. 2011. Marrying<br />
content and process in computer science education. IEEE<br />
Transactions on Education 54, 3, 387–397.
APPENDIX<br />
Groups<br />
a1 = BW teachers <strong>of</strong> computer science<br />
a2 = BY teachers <strong>of</strong> computer science<br />
a1<br />
a2<br />
Process concepts<br />
c1 = analyzing<br />
= finding relationships<br />
Content concepts c6<br />
b1 = problem<br />
b2 = information<br />
b3 = model<br />
b4 = algorithm<br />
b5 = data<br />
b6 = structure<br />
b7 = system<br />
b8 = computation<br />
b9 = process<br />
b10 = s<strong>of</strong>tware<br />
b11 = program<br />
b12 = test<br />
b13 = communication<br />
b14 = language<br />
b15 = computer<br />
grand means<br />
b1 = problem<br />
b2 = information<br />
b3 = model<br />
b4 = algorithm<br />
b5 = data<br />
b6 = structure<br />
b7 = system<br />
b8 = computation<br />
b9 = process<br />
b10 = s<strong>of</strong>tware<br />
b11 = program<br />
b12 = test<br />
b13 = communication<br />
b14 = language<br />
b15 = computer<br />
grand means<br />
c2 = classiyfing<br />
c3 = problem solving and posing<br />
c4 = categorizing<br />
c5 = investigating<br />
c7 = generalizing<br />
c8 = creating and inventing<br />
4.71 3.53 4.00 3.47 3.29 3.24 3.29 3.12 2.71 3.24 3.00 3.29 2.59 2.71 2.76 2.41<br />
3.53 3.35 2.41 3.47 3.00 2.24 2.18 1.94 2.82 1.88 2.65 2.65 3.41 3.53 2.76 2.88<br />
3.41 2.88 2.88 2.88 2.53 2.53 2.65 2.76 2.41 2.65 1.88 3.11 2.53 2.94 2.18 1.71<br />
4.41 2.82 3.88 2.53 2.24 3.35 3.41 3.82 3.00 3.53 2.29 3.24 2.24 2.35 1.82 1.76<br />
3.82 3.76 2.00 3.47 2.29 2.82 1.59 1.35 3.12 2.00 1.82 1.82 2.76 3.06 1.94 3.35<br />
4.00 3.35 3.24 3.06 2.71 2.71 3.29 2.47 2.53 2.65 1.82 2.59 2.24 2.41 2.18 2.53<br />
2.94 2.41 1.88 2.41 2.00 1.94 1.94 1.35 2.12 1.59 1.47 2.06 1.65 1.35 1.65 1.65<br />
2.94 2.41 2.88 2.29 2.65 2.71 2.94 2.12 2.53 3.00 1.94 2.59 1.53 2.00 1.71 1.88<br />
3.59 2.71 2.29 2.59 2.53 2.06 2.82 2.24 1.76 2.59 2.35 2.41 2.35 2.41 2.06 1.71<br />
2.82 2.24 2.65 2.29 2.35 1.82 1.76 3.12 2.00 2.18 1.47 1.88 2.00 2.35 2.47 1.41<br />
3.53 2.71 3.24 2.41 2.88 2.18 2.82 3.12 2.35 2.65 1.94 2.65 2.18 2.41 2.47 1.53<br />
2.71 2.35 2.65 2.41 2.29 2.47 1.65 2.53 2.65 2.76 2.06 2.00 1.53 2.35 1.41 1.35<br />
2.29 2.35 2.18 2.00 2.71 2.24 2.12 2.59 2.06 2.18 2.88 1.88 4.47 3.18 3.88 1.82<br />
2.65 2.18 1.94 2.41 2.00 2.06 2.00 1.71 2.12 1.59 2.41 2.24 3.59 2.71 2.71 1.18<br />
2.18 1.29 1.53 1.53 1.65 1.53 1.82 1.41 2.00 1.94 1.41 1.47 2.06 1.76 1.18 1.18<br />
c9 = comparing<br />
3.30 2.69 2.64 2.62 2.47 2.39 2.42 2.38 2.41 2.43 2.09 2.39 2.47 2.50 2.21 1.89 2.46<br />
4.57 2.67 4.00 2.86 2.90 3.67 3.43 2.52 3.19 3.29 3.14 2.95 2.95 2.05 3.10 2.14<br />
3.86 3.95 2.29 3.57 3.00 3.29 2.76 2.33 3.24 2.57 3.19 2.57 3.57 3.71 2.57 3.33<br />
4.29 3.86 3.62 3.43 2.95 4.00 4.00 3.14 3.43 3.48 2.71 3.14 3.05 2.76 2.29 2.71<br />
4.05 2.52 4.43 2.00 2.48 3.14 2.90 3.43 2.62 3.14 2.38 3.00 1.57 1.48 1.76 2.24<br />
4.10 3.62 1.95 3.24 2.76 2.67 3.05 1.57 2.81 2.05 1.81 1.81 2.57 2.62 1.76 3.38<br />
4.38 3.57 2.76 3.33 2.43 3.43 3.57 2.48 3.57 3.71 2.43 2.90 2.62 2.24 2.52 2.57<br />
3.43 3.05 2.48 2.62 2.29 2.76 2.52 1.90 2.38 2.71 1.95 2.33 2.29 1.86 2.48 2.10<br />
2.43 1.90 3.00 1.81 2.29 2.29 2.52 2.52 2.33 2.48 2.10 2.14 1.81 2.10 1.67 1.81<br />
3.76 2.52 2.43 2.43 2.05 2.62 2.43 1.67 2.19 3.38 2.00 2.29 2.38 1.76 2.29 2.33<br />
3.05 2.29 2.90 2.19 2.43 2.10 2.00 2.14 2.81 2.71 2.38 2.48 2.43 1.81 1.86 2.05<br />
3.43 2.19 2.90 2.43 2.29 1.86 2.29 2.95 2.05 2.95 1.95 2.19 2.43 2.14 2.57 1.81<br />
3.05 2.29 2.90 2.19 2.43 2.10 2.00 2.14 2.81 2.71 2.38 2.48 2.43 1.81 1.86 2.05<br />
2.67 1.76 2.29 1.67 2.19 2.76 2.29 1.95 2.14 2.81 3.00 2.00 3.71 2.95 3.38 2.10<br />
2.71 2.24 2.76 2.00 1.67 2.24 2.57 2.00 2.57 1.71 2.38 1.76 2.95 1.90 2.33 1.71<br />
2.67 2.33 2.00 2.24 2.00 2.29 2.14 1.43 2.14 2.14 1.48 1.57 2.14 1.71 2.29 2.10<br />
3.50 2.73 2.82 2.54 2.37 2.74 2.72 2.26 2.66 2.73 2.31 2.34 2.60 2.20 2.34 2.26<br />
Figure 6. Means for the 2•15×16 split-plot design (n1 = 17; n2 = 21)<br />
74<br />
c10 = finding cause-and-effect r.<br />
c11 = questioning<br />
c12 = transferring<br />
c13 = communicating<br />
c14 = presenting<br />
c15 = collaborating<br />
c16 = or<strong>der</strong>ing<br />
grand means<br />
3.21<br />
2.79<br />
2.62<br />
2.92<br />
2.56<br />
2.74<br />
1.90<br />
2.38<br />
2.40<br />
2.18<br />
2.57<br />
2.20<br />
2.55<br />
2.22.<br />
1.62<br />
3.09<br />
3.11<br />
3.30<br />
2.70<br />
2.61<br />
3.03<br />
2.45<br />
2.20<br />
2.41<br />
2.17<br />
2.40<br />
2.35<br />
2.48<br />
2.22<br />
2.04<br />
2.57
Ways <strong>of</strong> Planning Lessons on the Topic <strong>of</strong> Networks and<br />
the Internet<br />
ABSTRACT<br />
Ana-<strong>Maria</strong> Mesaro¸s<br />
<strong>University</strong> <strong>of</strong> Oldenburg<br />
Computer Science Education<br />
26111 Oldenburg, <strong>Germany</strong><br />
ana.maria.mesaros@uni-oldenburg.de<br />
Changing conditions and different educational concepts are<br />
wide-spread problems <strong>of</strong> Computer Science education. Due<br />
to this fact suitable teacher training can only be developed<br />
if the preconditions <strong>of</strong> CS teachers are being consi<strong>der</strong>ed. In<br />
or<strong>der</strong> to find out about teachers’ subjective theories on planning<br />
lessons we asked them how they would plan lessons on<br />
the specific topic <strong>of</strong> networks and the Internet. In this paper<br />
we show the research framework for examining this question<br />
and describe the organization <strong>of</strong> our study and first results.<br />
The different ways <strong>of</strong> planning lessons for this specific topic<br />
we found and describe here differ enormously.<br />
Keywords<br />
subjective theories, educational reconstruction, planning lessons<br />
1. INTRODUCTION<br />
Without a basis <strong>of</strong> generally accepted standards and with<br />
varying qualifications Computer Science (CS) teachers make<br />
their own decisions on what topics to teach and how to teach<br />
them. This leads to a situation in which the planning <strong>of</strong><br />
lessons on a specific topic differs greatly from teacher to<br />
teacher. The reasons for this are different subjective theories<br />
about CS and different ideas about teaching approaches.<br />
For those who want to provide effective teacher training<br />
like us those subjective theories are very important to know.<br />
To gain this basic knowledge, we conducted a survey on the<br />
question on how teachers would plan a lesson on the specific<br />
topic <strong>of</strong> networks and the Internet.<br />
Parts <strong>of</strong> this research have already been introduced in [9]<br />
where we reported on a pre-test. We therefore confine ourselves<br />
here to giving only a short introduction to the research<br />
framework. The focus <strong>of</strong> this paper is on the different possibilities<br />
to plan lessons on the topic <strong>of</strong> networks and the<br />
Internet. Therefore, our next step is to describe this framework<br />
shortly, the design <strong>of</strong> the interview and the sample. We<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WiPSCE 2012 Hamburg, <strong>Germany</strong><br />
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$10.00.<br />
75<br />
Ira Diethelm<br />
<strong>University</strong> <strong>of</strong> Oldenburg<br />
Computer Science Education<br />
26111 Oldenburg, <strong>Germany</strong><br />
ira.diethelm@uni-oldenburg.de<br />
then present the results achieved so far: How the planning<br />
<strong>of</strong> lessons defers in the reasons for teaching this topic, in<br />
the contents <strong>of</strong> the topic, in the structuring the content, in<br />
the learning objectives pursued with it and the used teaching<br />
methods and teaching material. We also report on the<br />
different ways <strong>of</strong> integrating students’ perspectives in the<br />
planning <strong>of</strong> lessons. Finally we draw our conclusions and<br />
give a preview <strong>of</strong> our future work.<br />
2. EDUCATIONAL RECONSTRUCTION<br />
FOR TEACHER EDUCATION<br />
The research questions presented above are embedded in<br />
the framework <strong>of</strong> Educational Reconstruction for Teacher<br />
Education which shows the significance <strong>of</strong> teachers’ subjective<br />
theories on the planning <strong>of</strong> their lessons.<br />
This model is based on the model <strong>of</strong> Educational Reconstruction.<br />
The main idea is that to plan lessons means to<br />
structure a content consi<strong>der</strong>ing the students’ perspectives<br />
[6]. Therefore the scientific content has to be broken down<br />
and rearranged, so that the students’ perspectives are taken<br />
into account. This study is based on the adaptation <strong>of</strong> the<br />
model <strong>of</strong> Educational Reconstruction for Teacher Education.<br />
This adaptation, see fig. 2, contains some changes in<br />
Figure 1: Educational Reconstruction for Teacher<br />
Education [6]<br />
comparison with the original one. The content for teacher<br />
training is defined by the scientific view <strong>of</strong> a topic and by<br />
the domain-specific educational concepts. As in the original<br />
model <strong>of</strong> Educational Reconstruction the content has to
include learners’ perspectives. While talking about Teacher<br />
Education, the teacher himself becomes the learner. This<br />
implies that teachers’ perspectives have to supplement the<br />
content given by the educational concepts. Teachers’ perspectives<br />
<strong>of</strong> planning lessons in CS are based on their subjective<br />
theories. Based on Kelly [5] and Groeben [4] we define<br />
subjective theories as individual cognitive structures <strong>of</strong><br />
self- and worldviews that have the function to explain and<br />
to predict.<br />
This model must be un<strong>der</strong>stood as an iterative one. The<br />
educational concepts have to be structured anew once the<br />
subjective theories <strong>of</strong> teachers are known. Similarly the subjective<br />
theories can be un<strong>der</strong>stood better, if the educational<br />
concepts are used as an aid. When developing guidelines for<br />
teacher training one has to make sure that these educational<br />
concepts and subjective theories are taken into account.<br />
Therefore, there are several steps to achieve the goal <strong>of</strong><br />
creating guidelines for Teacher Education in CS. In the following<br />
chapter we show by the design <strong>of</strong> our investigation<br />
how we will achieve it.<br />
3. EXPLORING TEACHERS’ SUBJECTIVE<br />
THEORIES<br />
The intention <strong>of</strong> this study is to explore subjective theories<br />
<strong>of</strong> CS teachers on planning CS lessons. The method we used<br />
was a semi-structured guided interview. For our aim it is<br />
important to detect a wide range <strong>of</strong> subjective theories <strong>of</strong><br />
CS teachers, which calls for a qualitative research approach.<br />
In or<strong>der</strong> to increase the quality <strong>of</strong> the answers and for a<br />
better comparison we decided to do the investigation on a<br />
specific topic. For several reasons we decided on the topic<br />
<strong>of</strong> networks and the Internet. This topic does not only have<br />
high relevance in students’ everyday life [8], but the idea<br />
<strong>of</strong> communication is one <strong>of</strong> Dennings’ principles [1] and it is<br />
also mentioned in the Education Standards for CS published<br />
by the German Association <strong>of</strong> CS [3]. Additionally, students’<br />
perspectives have been investigated lately on this topic [2]<br />
and we can use these results for our investigation.<br />
In the following chapter we will describe the structure <strong>of</strong><br />
the interview and the sample.<br />
3.1 The Structure <strong>of</strong> the Interview<br />
For the collection <strong>of</strong> data we chose a semi-structured guided<br />
interview. This means that there are set questions, but<br />
they may be asked in a different or<strong>der</strong> or put in a different<br />
way. During the interview a communicative confirmation<br />
took place. The interviewer repeated the given answer in<br />
his own words and asked the teacher if he had un<strong>der</strong>stood<br />
him correctly.<br />
As an introduction we asked the teacher how he or she<br />
had planned the past school year. This question does<br />
not only provide an overview <strong>of</strong> the topics the teacher has<br />
taught in the past year, but it directs the focus on the (pr<strong>of</strong>essional)<br />
expertise <strong>of</strong> the teacher. In addition, this question<br />
shows how the teacher plans other topics in CS and thereby<br />
helps to generalize the results <strong>of</strong> this study.<br />
In the following, questions were asked about the planning<br />
<strong>of</strong> lessons on the Internet. Therefore we asked<br />
if this topic has been taught, and how relevant this topic<br />
is. Afterwards we asked in which grade they would teach<br />
this topic and which contents it would contain. Questions<br />
on the learning objectives, the used methods and materials<br />
76<br />
followed. We will come to the answers <strong>of</strong> this section later.<br />
The next set <strong>of</strong> questions concerned students’ perspectives.<br />
Teachers were asked for their assessment <strong>of</strong> their<br />
students’ un<strong>der</strong>standing <strong>of</strong> the functioning <strong>of</strong> the Internet.<br />
They were also asked if they consi<strong>der</strong>ed it important to<br />
know their students’ perspectives and which part this aspect<br />
played in planning the lessons. Since during a previous pretest<br />
most teachers had not been able to imagine students’<br />
perspectives, the interviewer read a student’s statement on<br />
how the internet works [2] to them.<br />
The following questions focus on the other subjects <strong>of</strong><br />
teachers because a teacher in <strong>Germany</strong> normally teaches<br />
two subjects and not just one like in many other countries.<br />
With these answers we un<strong>der</strong>stand better in which ways CS<br />
differs from other subjects in the teachers’ perception. This<br />
might be important for the concepts for teacher training.<br />
The interview ends with questions about the teachers’<br />
biography. After describing the methods for the data collection<br />
we will go on to give some information about the<br />
sample.<br />
3.2 The Sample<br />
For the aim <strong>of</strong> discovering different subjective theories<br />
<strong>of</strong> CS teachers on planning lessons, not the frequency is<br />
important but the variety <strong>of</strong> theories. Our focus was on<br />
questioning teachers with possibly different subjective theories.<br />
Therefore, we selected teachers with varying experience,<br />
those who had been teaching for thirty years and<br />
others who had just started a few weeks before.<br />
We also interviewed teachers with different qualifications.<br />
Some <strong>of</strong> them had an <strong>of</strong>ficial certificate for teaching CS,<br />
others just taught the subject because they were interested<br />
in it or because they have been asked to. Not all teachers<br />
had studied CS at university.<br />
We also took the different school types in <strong>Germany</strong> into<br />
account and chose teachers from all types <strong>of</strong> secondary schools<br />
in which CS is being taught. All these characteristics led<br />
to a heterogeneous group <strong>of</strong> fifteen teachers whose different<br />
subjective theories we set out to examine.<br />
3.3 Evaluation <strong>of</strong> the Interviews<br />
The interviews took place in school and were recorded.<br />
All interviews were transcribed. The transcription only included<br />
the spoken word without showing differences in intonation.<br />
We consi<strong>der</strong>ed this way <strong>of</strong> transcription sufficient<br />
for our objective. The transcribed interviews were analyzed<br />
in several steps. First they were coded with the method<br />
<strong>of</strong> qualitative content analysis [7]. In a first step the categories<br />
<strong>of</strong> codings were built in a deductive way from the<br />
interview structure. The second coding was an inductive<br />
one. Therefore the given categories from the interview were<br />
supplemented by new categories, taken from the data.<br />
In the following, the results <strong>of</strong> this two first coding steps<br />
are shown.<br />
4. DIFFERENT WAYS OF PLANNING<br />
LESSONS<br />
In the following we describe the different approaches to<br />
teaching the topic networks and the Internet.<br />
4.1 Reasons for the Topic<br />
Most <strong>of</strong> the interviewed teachers did not handle this topic<br />
in class or only included some aspects <strong>of</strong> it in lessons on some
other topic. The reasons given varied consi<strong>der</strong>ably. Some <strong>of</strong><br />
the teachers who did teach this topic consi<strong>der</strong>ed it to be<br />
important because it is very interesting for the students.<br />
Other teachers gave various reasons for not teaching this<br />
topic. Some <strong>of</strong> them thought that the students were not<br />
ready for this difficult topic yet or that they were not interested.<br />
Others consi<strong>der</strong>ed it more important to do some<br />
programming or teach the topics needed for the university<br />
entrance qualification. Some teachers just said that they did<br />
not plan to teach this topic in their school and others explained<br />
that their knowledge <strong>of</strong> this topic was not sufficient.<br />
But most <strong>of</strong> them thought that the topic <strong>of</strong> networks and<br />
the Internet was relevant for students, because they use it<br />
daily and it is a part <strong>of</strong> general knowledge to know how the<br />
Internet works in their opinion.<br />
An explanation for that might be that teachers find this<br />
topic important but their own knowledge is not sufficient.<br />
When we asked them how they felt about their knowledge<br />
<strong>of</strong> networks and the Internet most <strong>of</strong> them expressed some<br />
uncertainty. They did have ideas about how to teach the<br />
topic even though they had not done so, yet.<br />
The answers on how many lessons should be dedicated<br />
to this topic showed that it might be a topic with lots <strong>of</strong><br />
contents. Nearly all <strong>of</strong> the teachers consi<strong>der</strong>ed a lot <strong>of</strong> time<br />
necessary for teaching this topic. The answers varied from<br />
nine periods <strong>of</strong> 90 minutes to half a school year.<br />
The grade in which they would teach this topic differs as<br />
well. Some teachers would teach networks and the Internet<br />
in the sixth or seventh grade, but would focus on using<br />
the Internet and learning about the hazards. Others would<br />
teach it from the ninth grade onwards and consi<strong>der</strong> different<br />
contents as important.<br />
4.2 Contents <strong>of</strong> the Topic<br />
The answers on which topics teachers think should be<br />
taught in lessons about network and the Internet can be<br />
classified in four groups. The first group consists <strong>of</strong> all the<br />
topics concerning the function or the history <strong>of</strong> the Internet.<br />
Included in this category is the question on how the Internet<br />
works. Almost all teachers would include the function <strong>of</strong> the<br />
Internet.<br />
Another topical group is programming/creating something<br />
in the Internet. Teachers would teach how to create a personal<br />
homepage. They would teach HTML or PHP, as an<br />
method how to create an individual part <strong>of</strong> the Internet.<br />
Another theme would be different network technologies,<br />
like emails. This group also contains cryptology. Here teachers<br />
would describe how an email gets from the sen<strong>der</strong> to the<br />
recipient and what kind <strong>of</strong> encryption can be used.<br />
Yet another topic is the utilization <strong>of</strong> the Internet. Themes<br />
like social networks and privacy are part <strong>of</strong> that. This content<br />
seems to be important to CS teachers, because nearly<br />
all <strong>of</strong> them mentioned it.<br />
4.3 Structuring the Content<br />
The different possibilities <strong>of</strong> introducing the subject depend<br />
on the topics the teacher has decided to teach: like the<br />
function <strong>of</strong> the Internet, the creation <strong>of</strong> a personal homepage,<br />
social networks, data security or privacy. Some <strong>of</strong> the<br />
teachers would want their introduction to be problem-based.<br />
Therefore they would present a real problem to be solved by<br />
the students during the lessons. Another possibility would<br />
be to start with a problem <strong>of</strong> current interest.<br />
77<br />
Other teachers have no plans for structuring lessons, they<br />
are very flexible and prepared to structure the topics spontaneously.<br />
They are guided by the students’ interests and<br />
decide together on the content <strong>of</strong> the lessons. Some <strong>of</strong> them<br />
would introduce the topic <strong>of</strong> networks and the Internet by<br />
asking students how they imagine the Internet works, or by<br />
asking them what contents they think are important to learn<br />
about this topic. Other teachers are convinced that students<br />
do not have a perspective on this topic, so they would begin<br />
the lessons without taking students’ perspectives into account.<br />
If they should later realize that the students do have<br />
some perspectives on this topic, they would consi<strong>der</strong> them.<br />
4.4 Learning Objectives<br />
We also asked about the learning objectives teachers pursued<br />
with their lessons. We found out that the objectives<br />
named can be classified in four groups. The first group is<br />
about content knowledge. Students should know how the<br />
Internet works, they should know something about the history<br />
<strong>of</strong> the Internet and they should know some cryptology<br />
algorithms.<br />
Another group is about comprehension. Students should<br />
un<strong>der</strong>stand how the Internet works so that they can use it<br />
better, and they should also un<strong>der</strong>stand what was wrong<br />
about their perspectives. This group <strong>of</strong> learning objectives<br />
can be characterized very well by a statement one teacher<br />
made: ”‘Students should un<strong>der</strong>stand what they use”’.<br />
Another group is about using the Internet/social networks<br />
correctly. The last group <strong>of</strong> learning objectives is about s<strong>of</strong>t<br />
skills, like teamwork or presentation skills.<br />
4.5 Teaching Methods<br />
Teachers use different methods to achieve their learning<br />
objectives. Some <strong>of</strong> them prepare the content and just tell<br />
the students how everything works. Other teachers inquire<br />
about students’ interests and plan their lessons based on<br />
that information. In this case topics and methods are based<br />
on students’ interests.<br />
Most <strong>of</strong> the teachers use the method <strong>of</strong> projects. Students,<br />
mostly in groups, have to do their own research on a topic,<br />
un<strong>der</strong>stand it, prepare it for presentation and <strong>of</strong> course do<br />
the presentation. This is the opposite <strong>of</strong> teachers’ input,<br />
because it is a very student-based method.<br />
Other teachers explained that visualization is very important<br />
for this topic. By visualization teachers mean different<br />
things. It might be a kind <strong>of</strong> game, or a simulation where<br />
the students act as part <strong>of</strong> a network and can see how the<br />
data gets from one computer to another. Visualization can<br />
also mean watching a movie in which the pathway <strong>of</strong> the<br />
data through the network is shown. Furthermore visualization<br />
can also mean to have a look at the real network, such<br />
as the school network or the network at home.<br />
4.6 Teaching Materials<br />
On the question about the materials used in the lessons<br />
the interviewed teachers did not only give examples <strong>of</strong> what<br />
kind <strong>of</strong> materials they would use, but also mentioned the<br />
sources <strong>of</strong> the material. Some teachers use their own knowledge<br />
and creativity as a source and produce their material<br />
themselves. Others fall back on prepared materials and just<br />
use them. Some <strong>of</strong> the interviewed teachers had no idea<br />
where they would get the materials for the topic <strong>of</strong> networks<br />
and the Internet.
Some teachers would use applets or s<strong>of</strong>tware to do a simulation<br />
<strong>of</strong> a network or to show the pathway <strong>of</strong> the data<br />
through the Internet. Others would use films as a kind <strong>of</strong><br />
visualization. Some would use newspaper articles to describe<br />
a problem that could then be handled in class. Worksheets<br />
are also popular materials.<br />
4.7 Integrating Students’ Perspectives<br />
Most teachers agreed that students’ perspectives are important<br />
in planning the lessons. Only a few said that students’<br />
perspectives are not relevant. They are convinced<br />
that students are not interested in finding out how the Internet<br />
works. One teacher compared the students with fish<br />
which do not need to know what water is.<br />
Those teachers who agree that students’ perspectives have<br />
to be taken into account, claim to observe students’ perspectives<br />
during classes and to take them into consi<strong>der</strong>ation.<br />
About half <strong>of</strong> the teachers said that the lessons they plan<br />
would change the students’ conceptions <strong>of</strong> the Internet.<br />
We also asked teachers to describe their students’ conceptions<br />
<strong>of</strong> the Internet. Teachers think that students imagine<br />
the Internet...<br />
• to be like a Super-Computer.<br />
• to be like a cloud that surrounds the whole world, and<br />
you can connect everywhere with this cloud.<br />
• to be like a cupboard filled with information<br />
• to consist <strong>of</strong> unstructured connections between the computer<br />
to have something like rays to a radio tower or<br />
a satellite<br />
• to function like a phone call<br />
5. CONCLUSIONS AND FUTURE WORK<br />
In the previous chapter we discussed the different teaching<br />
methods and learning objectives which we discovered in<br />
our survey. These answers only refer to that part <strong>of</strong> the interviews<br />
dealing with the question <strong>of</strong> the planning <strong>of</strong> lessons<br />
on the topic <strong>of</strong> networks and the Internet.<br />
An interesting result is that most teachers think that networks<br />
and the Internet is an important topic to teach, but<br />
for various reasons they do not do it. One widespread reason<br />
seems to be the problem <strong>of</strong> not having enough content<br />
knowledge about this topic. This problem can fortunately<br />
be solved easily by developing teacher training courses.<br />
An obvious result is that teachers have a clear idea on how<br />
to plan lessons, although they feel a lack <strong>of</strong> knowledge about<br />
this topic. Their pedagogical knowhow seems to help them<br />
in overcoming this lack <strong>of</strong> knowledge. This might be important<br />
in developing teacher training. It shows that teachers<br />
rather need training on content knowledge than on pedagogical<br />
knowhow.<br />
An astonishing result for us was the fact that in CS it<br />
seems to be usual that students were allowed to choose the<br />
topics they wanted to learn. This can probably be explained<br />
by the lack <strong>of</strong> generally accepted standards and by the teachers’<br />
insufficient confidence in their own knowledge. But some<br />
teachers also mentioned that they thought their students<br />
might know more about CS than they did themselves and<br />
that might explain why students were allowed to make decisions<br />
on the content <strong>of</strong> their lessons.<br />
78<br />
Contrary to our expectation teachers claim to include students’<br />
perspectives in planning lessons. It is not unlikely<br />
that they may just have given the expected answer. Another<br />
explanation is a presentation about students’ perspectives<br />
that some <strong>of</strong> the teachers <strong>of</strong> our sample attended.<br />
The different ways <strong>of</strong> planning lessons we discussed have<br />
an internal structure, a kind <strong>of</strong> a pattern which expands over<br />
the other topics taught by the teachers. These patterns are<br />
formed by the subjective theories <strong>of</strong> teachers on planning CS<br />
lessons. Therefore our next step will be to discover the types<br />
<strong>of</strong> planning CS lessons, based on this subjective theories.<br />
These types <strong>of</strong> planning lessons determine not only the topic<br />
networks and the Internet but any kind <strong>of</strong> lesson planning.<br />
By using these types, our final objective is the development<br />
<strong>of</strong> guidelines for teacher training. We do not only want<br />
to show in which different ways teachers plan their lessons<br />
but also how to include this knowledge into teacher trainingfor<br />
the benefit <strong>of</strong> all teachers. Therefore, the differing ways<br />
<strong>of</strong> planning presented here are a big step towards this goal.<br />
6. REFERENCES<br />
[1] P. Denning. Great principles <strong>of</strong> computing.<br />
Communications <strong>of</strong> the ACM, 46(11):15–20, November<br />
2003.<br />
[2] I. Diethelm and S. Zumbrägel. An investigation <strong>of</strong><br />
secondary school students’ conceptions on how the<br />
internet works. In Proceedings <strong>of</strong> the 12th Koli Calling<br />
International Conference on Computing Education<br />
Research, Koli, Finland, 2012. (accepted).<br />
[3] Gesellschaft für Informatik (GI) e. V. Grundsätze und<br />
Standards für die Informatik in <strong>der</strong> Schule. Number<br />
28,150/151 in LOG IN. Berlin, 2008.<br />
[4] N. Groeben, D. Wahl, J. Schlee, and B. Scheele.<br />
Forschungsprogramm Subjektive Theorien. Eine<br />
Einführung in die Psychologie des reflexiven Subjekts.<br />
A. Francke Verlag, Tübingen, 1988.<br />
[5] G. A. Kelly. The psychology <strong>of</strong> personal constructs.<br />
Routledge, London, 1991.<br />
[6] M. Komorek and U. Kattmann. The model <strong>of</strong><br />
educational reconstruction. In S. Mikelskis-Seifert,<br />
U. Ringelband, and M. B. (Eds.), editors, Four Decades<br />
<strong>of</strong> Research in Science Education - from Curriculum<br />
Development to Quality Improvement, chapter 7, pages<br />
171–188. Waxmann, 2008.<br />
[7] P. Mayring. Qualitative content analysis. Forum<br />
Qualitative Sozialforschung / Forum: Qualitative Social<br />
Research, 1(2):10, 2000.<br />
[8] Medienpädagogischer Forschungsverbund Südwest. JIM<br />
2010 Jugend, Information, (Multi-)Media.<br />
Forschungsberichte. Baden-Baden, 2010.<br />
[9] A.-M. Mesaros and I. Diethelm. Exploring computer<br />
science teachers’ subjective theories on designing their<br />
lessons. In Proceedings <strong>of</strong> the 5th International<br />
Conference on Informatics in Schools: Situation,<br />
Evolution and Perspectives ISSEP, Selected Papers,<br />
Bratislava, Slovenia, 2011.
Promoting Computational Thinking with Programming<br />
Cynthia C. Selby<br />
<strong>University</strong> <strong>of</strong> Southampton<br />
Highfield<br />
Southampton UK<br />
44 (0) 2380 593475<br />
ABSTRACT<br />
The term computational thinking has received some discussion in<br />
the field <strong>of</strong> computer science education research. The term is<br />
defined as the concept <strong>of</strong> thinking about problems in a way that<br />
can be implemented in a computing device. Of course, after<br />
having thought about a problem using computational thinking<br />
skills, the next step should be to use programming skills to<br />
implement the solution. This work in progress is exploring ways<br />
in which programming can be employed as a tool to teach<br />
computational thinking and problem solving. Data is collected<br />
from teachers, academics, and pr<strong>of</strong>essionals from various<br />
industries. They are purposively selected because <strong>of</strong> their<br />
knowledge <strong>of</strong> or interest in the topics <strong>of</strong> problem solving,<br />
computational thinking, and the teaching <strong>of</strong> programming. This<br />
data is analyzed within the paradigm <strong>of</strong> the grounded theory<br />
approach. The results <strong>of</strong> an initial analysis imply an or<strong>der</strong>ing <strong>of</strong><br />
complexity associated with computational thinking skills, imply<br />
connections between computational thinking skills and<br />
programming activities, and imply a relationship between<br />
computational thinking skills and other taxonomies <strong>of</strong> learning.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers and Education]: Computers and Education,<br />
Curriculum<br />
General Terms<br />
Design, Experimentation, Theory<br />
Keywords<br />
Computational thinking, pedagogy <strong>of</strong> programming, problem<br />
solving<br />
1. INTRODUCTION<br />
Problem solving skills are used in developing or implementing<br />
strategies to solve problems in many domains. These skills are<br />
<strong>of</strong>ten expressed as heuristics [12], appropriate and plausible<br />
approaches to a problem. Effective problem solving has been<br />
promoted by the use <strong>of</strong> strategies including means-ends analysis,<br />
schema acquisition, algorithmic approaches, and targeted<br />
frameworks [19, 12, 9].<br />
C.Selby@soton.ac.uk<br />
79<br />
However, in the domain <strong>of</strong> computer science, some research [11]<br />
has found that learners do not naturally solve problems in ways<br />
that can be translated to computing devices by the use <strong>of</strong><br />
programming. This disparity is highlighted in the Lister study<br />
[6], where it is suggested that ineffective problem solving skills,<br />
including the ability to work through lines <strong>of</strong> logic, may be the<br />
cause <strong>of</strong> ineffective programming skills. Additional studies [15,<br />
13] suggest that problems in learning to program are exacerbated<br />
by a lack <strong>of</strong> strategic tools.<br />
The strategic tools, identified as useful for those attempting to<br />
solve problems with the aid <strong>of</strong> computational devices in various<br />
domains, include, but are not limited to, decomposition,<br />
abstraction, simulation, and generalization [10]. The name given<br />
to these specialized mental skills, resulting in solutions to<br />
problems directly translatable to a computing device, is<br />
computational thinking [21]. Actually implementing these<br />
solutions requires a different set <strong>of</strong> skills.<br />
Programming skills are the specific technical skills needed to<br />
produce specific solutions using a set <strong>of</strong> defined digital tools,<br />
<strong>of</strong>ten associated with a programming language [9]. Research<br />
<strong>of</strong>ten reports that learners struggle with programming skills such<br />
as tracing [3, 6] and un<strong>der</strong>standing a model <strong>of</strong> the machine [2, 9].<br />
Having recognized this issue, other researchers [7, 20] highlight<br />
the need for a defined hierarchy <strong>of</strong> programming skills. One<br />
study [16] attempted to provide such a hierarchy for objectoriented<br />
programming. The researchers found that teachers<br />
interpreted the hierarchy as a capability hierarchy.<br />
Given that a hierarchy <strong>of</strong> generic programming skills could be<br />
developed and interpreted as capability, the levels could be<br />
aligned with existing hierarchies, such as the cognitive domain <strong>of</strong><br />
Bloom’s Taxonomy. These same programming skills could be<br />
mapped to the higher-level computational thinking skills that they<br />
evidence, thereby defining a hierarchy <strong>of</strong> computational thinking<br />
skills. This setting provides the context for an ongoing<br />
investigation into the relationship between the teaching <strong>of</strong><br />
programming and its effect on the acquisition <strong>of</strong> computational<br />
thinking skills by learners.<br />
2. STUDY METHOD<br />
This study is based on a grounded theory approach employing<br />
qualitative data collection methods and qualitative data analysis<br />
techniques. The first activity is the administration <strong>of</strong> an Internet<br />
based questionnaire. The second activity is the collection <strong>of</strong> data<br />
from an Internet based community <strong>of</strong> practice forum. The third<br />
activity is the administration <strong>of</strong> a face-to-face, audio recorded,<br />
semi-structured interview schedule for respondents previously<br />
identified by an analysis <strong>of</strong> the questionnaire results and<br />
community <strong>of</strong> practice discussions. All data is iteratively<br />
augmented and analyzed guided by Strauss and Corbin’s [18]<br />
grounded theory procedures and techniques until theoretical
saturation. It is anticipated that a product <strong>of</strong> the theory generation<br />
may be a model <strong>of</strong> relationships between problem solving skills,<br />
computational thinking skills, and programming skills.<br />
2.1 Participants and Sampling<br />
The participants in this research all have some interest in the<br />
teaching <strong>of</strong> programming, computational thinking, problem<br />
solving, or any combination <strong>of</strong> the three. Not all participants are<br />
teachers. Participants may be employed in industries where<br />
computational thinking skills and programming skills are useful<br />
or required. Other participants may be members <strong>of</strong> pr<strong>of</strong>essional<br />
communities <strong>of</strong> practice, representing industry, academia, or<br />
education. They are still perceived to have an interest in and<br />
appropriate knowledge <strong>of</strong> the research context.<br />
An individual participant may not engage with every data<br />
collection instrument. Participants are matched to instruments. In<br />
the case <strong>of</strong> the first instrument, an online questionnaire, the<br />
targeted sample consists <strong>of</strong> members <strong>of</strong> organizations whose<br />
ideologies promote the teaching <strong>of</strong> programming or<br />
computational thinking skills. In the case <strong>of</strong> the second, the<br />
online community <strong>of</strong> practice, conversation threads are chosen<br />
purposively for their applicability to the context <strong>of</strong> this research,<br />
without regard to the identity <strong>of</strong> the poster. From the<br />
questionnaire responses and the community <strong>of</strong> practice<br />
conversations, a further purposive selection is made to identify<br />
participants for the interviews. This purposive sampling is<br />
supported by Strauss and Corbin who affirm that theoretical<br />
sampling is a foundation stone <strong>of</strong> grounded theory which, “…<br />
enables the researcher to choose those avenues <strong>of</strong> sampling that<br />
can bring about the greatest theoretical return” ([18], p. 202).<br />
2.2 Data Collection<br />
The questionnaire and interview schedule have been designed<br />
specifically to elicit responses applicable to the topics <strong>of</strong> problem<br />
solving, computational thinking, and the teaching <strong>of</strong><br />
programming. To ensure the same level <strong>of</strong> appropriateness <strong>of</strong><br />
response, a set <strong>of</strong> keyword criteria has been developed on which<br />
the community <strong>of</strong> practice messages are searched.<br />
2.2.1 Online Questionnaires<br />
The questionnaire makes use <strong>of</strong> some closed questions but the<br />
majority <strong>of</strong> questions are open-ended to allow participants to<br />
respond as they wish. The or<strong>der</strong>ing <strong>of</strong> the questions is from<br />
general to specific, divided into major sections. Results are<br />
submitted one screen or page at a time. In this way, even the<br />
results <strong>of</strong> abandoned questionnaires have the potential to be used.<br />
Personal information is requested early in the response process to<br />
identify participants. This provides a mechanism for contacting<br />
the participant, should he or she be selected for an interview. The<br />
design <strong>of</strong> the resulting questionnaire aims to be as open as<br />
possible to facilitate depth <strong>of</strong> response, while controlling for<br />
researcher and question bias.<br />
2.2.2 Community <strong>of</strong> Practice<br />
The community <strong>of</strong> practice, whose discussions and opinions are<br />
<strong>of</strong> interest in this study, is computer-mediated. Simply by<br />
contributing, the members signify some interest in the topics that<br />
overlap with this study. Although computer-mediated, some<br />
individuals share collaborative practices in the classroom. There<br />
are also face-to-face meetings, <strong>of</strong> varying scale, held throughout<br />
the year.<br />
In or<strong>der</strong> to identify the most appropriate threads for inclusion in<br />
the dataset, discussions are keyword searched. The keywords<br />
80<br />
have been chosen to correspond to the terminology used in the<br />
initial research literature and early questionnaire responses.<br />
These terms include computational thinking, abstraction,<br />
decomposition, algorithm, and problem solving. Discussions,<br />
composed <strong>of</strong> individual and related messages, are consi<strong>der</strong>ed as a<br />
whole [17]. Regardless <strong>of</strong> the age <strong>of</strong> a discussion, once it has<br />
been identified as pertinent, every individual message in that<br />
discussion is read and coded, in line with the questionnaire and<br />
interview data.<br />
2.2.3 Interviews<br />
The design <strong>of</strong> the interview schedule used in this research is based<br />
on a semi-structured approach, as defined by Cohen, Manion, and<br />
Morrison [1]. In particular, the question wording and sequences<br />
are specified in advance <strong>of</strong> the interview. The interviewer is<br />
granted the flexibility to provide additional questions in or<strong>der</strong> to<br />
elicit greater depth in the responses. The interviewer is also<br />
granted the flexibility to record non-verbal indicators, such as<br />
body language or gestures. This semi-structured approach should<br />
provide sufficient control to ensure comparability <strong>of</strong> results,<br />
sufficient flexibility to ensure depth <strong>of</strong> responses, and sufficient<br />
consistency to support the simultaneous collection and analysis <strong>of</strong><br />
data indicated by the grounded theory paradigm.<br />
3. FINDINGS<br />
The current, non-saturated dataset is being analyzed in line with<br />
grounded theory, first as conceptual free nodes, then as<br />
categories. These categories and concepts may change, as more<br />
data is added and processed. Three <strong>of</strong> the categories presented<br />
here, problem solving skills, computational thinking skills, and<br />
programming skills have been introduced above.<br />
3.1 Problem Solving Skills<br />
In the context <strong>of</strong> this study, problem solving skills are not specific<br />
to programming, but are a wi<strong>der</strong> skill set applicable in many<br />
domains. Participants have highlighted problem un<strong>der</strong>standing<br />
and persistence as important concepts in this category. Analysis<br />
<strong>of</strong> the data indicates that a common key first step in both learning<br />
to solve problems and learning to program is being able to<br />
un<strong>der</strong>stand the problem and its constraints. This observation<br />
agrees with Pólya’s problem solving approach [12]. The theme <strong>of</strong><br />
persistence is <strong>of</strong>ten linked with the idea <strong>of</strong> “not giving up”.<br />
Puzzles and games are named as appropriate activities to promote<br />
persistence. They are identified as providing sustained and<br />
lengthy problem solving with discrimination <strong>of</strong> useful data, back<br />
tracking, and constant evaluation. Many participants recognize<br />
that the opportunity to problem solve is relevant in many different<br />
contexts.<br />
3.2 Computational Thinking Skills<br />
Participants in this study identified explicit examples <strong>of</strong><br />
computational thinking skills, as defined by the National Research<br />
Council [10], and related them specifically to problem solving.<br />
Recognizable computational thinking skills such as<br />
decomposition, modeling, and algorithm design are found in the<br />
responses, along with other skills such as planning, justifying, and<br />
evaluating.<br />
Decomposition, the skill to break problems down, is viewed as<br />
being taken for granted. However, for some students, this is<br />
reported as being very difficult and requiring explicit teaching.<br />
Modeling is identified in the sense <strong>of</strong> high-level systems that are<br />
decomposed into smaller parts, with each individual part<br />
modeling behavior <strong>of</strong> a subsystem. The act <strong>of</strong> planning an
algorithm or a product is viewed as a high-level computational<br />
thinking skill. Algorithm design is expressly tied to problem<br />
solving by the participants. It is defining the steps, using some<br />
accepted convention, to solve a problem. This is viewed<br />
differently to program design, which is seen as the translation <strong>of</strong><br />
an algorithm into automation un<strong>der</strong>standable by a computing<br />
device. Unexpectedly, participants also included learning to ask<br />
questions about alternatives, identifying trade-<strong>of</strong>fs, justifying<br />
decisions, identifying limitations, refining solutions, and<br />
evaluating results. These are frequently complemented by the<br />
term analytical thinking, which is perceived to involve comparing<br />
alternatives, precisely describing, explaining how, and criticizing<br />
weaknesses.<br />
In general, participants in this study agree with the National<br />
Research Council [10] and Wing [21] concerning the broad<br />
definition <strong>of</strong> computational thinking and none limited the use <strong>of</strong><br />
the term or the skill set identified by the use <strong>of</strong> the term only to<br />
the domain <strong>of</strong> computer science.<br />
3.3 Teaching Programming Skills<br />
The concepts in this category represent high-level concerns for<br />
the participants. Included here are the concepts <strong>of</strong> logical<br />
thinking, programming as a tool, and collaboration as a pedagogic<br />
strategy.<br />
The term logical thinking occurs prolifically in the dataset and<br />
appears to be associated closely with programming constructs<br />
such as sequence, selection, and iteration. This association is<br />
anticipated and parallels that <strong>of</strong> Saeli [9] who reports that the<br />
most identified big idea <strong>of</strong> programming is control structures.<br />
The idea <strong>of</strong> programming as a vehicle for teaching computational<br />
thinking crosses boundaries between respondents, encompassing<br />
academics, teachers, and industry pr<strong>of</strong>essionals. The components<br />
<strong>of</strong> computational thinking, such as decomposition and<br />
generalization, are also reported by Saeli [9] to be a big idea <strong>of</strong><br />
programming. Collaboration is identified, by participants, as an<br />
effective strategy for teaching computational thinking. This is<br />
usually described as paired or group work, most commonly<br />
involving discussions at the analysis or design phases <strong>of</strong> s<strong>of</strong>tware<br />
development. Notably, there are currently no responses<br />
indicating provision for group implementation or paired<br />
programming.<br />
While it is not surprising that participants associate the teaching<br />
<strong>of</strong> programming constructs, decomposition, and generalization<br />
with computational thinking, it is surprising that an established<br />
pedagogic technique, collaboration, has not been extended to<br />
opportunities for paired programming.<br />
4. CONCLUSION<br />
4.1 Preliminary Model<br />
Although the dataset has not yet been shown to be saturated, as<br />
proscribed by grounded theory, it can form the basis for<br />
preliminary theory generation. As indicated in the introduction,<br />
participants’ responses are used directly to <strong>der</strong>ive a model <strong>of</strong> the<br />
relationships between computational thinking skills, programming<br />
skills, and the cognitive domain <strong>of</strong> Bloom’s Taxonomy. Figure 1,<br />
a preliminary model, has been <strong>der</strong>ived to illustrate some <strong>of</strong> these<br />
relationships.<br />
81<br />
Figure 1: Preliminary Model<br />
The computational thinking skills, reflecting the terminology [10,<br />
21] introduced previously, are represented by an increasing level<br />
<strong>of</strong> complexity. This hierarchy can be discerned from the<br />
participants’ responses and the reported or<strong>der</strong> <strong>of</strong> introduction in<br />
the classroom. For example, breaking problems down,<br />
decomposition, is one technique introduced early in the teaching<br />
<strong>of</strong> both problem solving and programming. The programming<br />
activities column represents those activities that participants view<br />
as promoting computational thinking. For example, the<br />
collaborative work, reported by participants and described in 3.3,<br />
usually takes place during the analysis or design phase where a<br />
problem is broken down into subcomponents. Interestingly, the<br />
terminology used in Bloom’s Taxonomy, the last column, is<br />
represented directly in the participants’ responses. The terms<br />
analyze and un<strong>der</strong>stand are also used to describe activities found<br />
in the initial stages phases <strong>of</strong> problem solving or programming<br />
task. Although the dataset on which the model is based will<br />
change and grow, possible relationships can already be discerned<br />
4.2 Implications<br />
This study assumes, in line with Isbell and colleagues [5], that<br />
computational thinking skills are a requirement <strong>of</strong> 21 st century<br />
society and that these skills must be taught. This research<br />
contributes to the body <strong>of</strong> knowledge that may be used to inform<br />
the issue <strong>of</strong> effective teaching strategies for both programming<br />
and computational thinking. By more explicitly defining the<br />
relationship between computational thinking and programming,<br />
educators may be motivated to move the focus <strong>of</strong> activities from<br />
the production <strong>of</strong> an artifact to the acquisition <strong>of</strong> computational<br />
thinking skills. In the context <strong>of</strong> the current educational<br />
requirements to include more computer science at all key stages,
the results <strong>of</strong> this study could influence the design <strong>of</strong> curriculums<br />
aiming to incorporate the development <strong>of</strong> computational thinking<br />
skills. In addition, this research responds directly to Guzdial’s<br />
call [4] for more research into how to teach computing in a way<br />
that enforces computational thinking.<br />
4.3 Future Work<br />
Although the current study has not yet reached its conclusion,<br />
areas for further study have already been exposed by analysis <strong>of</strong><br />
the data. These include:<br />
How do learners move from the specifics <strong>of</strong> programming,<br />
such as language constructs or blocks, to more abstract<br />
concepts, such as sorting an array or finding an average,<br />
which aid higher-level problem solving?<br />
How could the explicit teaching <strong>of</strong> general problem solving<br />
skills and high-level problem solving strategies influence the<br />
development <strong>of</strong> computational thinking skills?<br />
More work into the relationships between problem solving,<br />
computational thinking, and programming could lead to improved<br />
classroom lessons, improved curriculums, and an improved<br />
un<strong>der</strong>standing <strong>of</strong> the skills required in 21 st century society.<br />
5. REFERENCES<br />
[1] Cohen, L., Manion, L. & Morrision, K. 2007. Research<br />
Methods in Education, Abingdon, England, Routledge.<br />
[2] Du Boulay, B. 1989. Some difficulties <strong>of</strong> learning to<br />
program. In: SOLOWAY, E. & SPOHRER, J. C. (eds.)<br />
Studying the novice programmer. Hillsdale, NJ: Lawrence<br />
Erlbaum.<br />
[3] Fitzgerald, S., Simon, B. & Thomas, L. 2005. Strategies that<br />
students use to trace code: an analysis based in grounded<br />
theory. Proceedings <strong>of</strong> the first international workshop on<br />
Computing education research. Seattle, WA, USA: ACM.<br />
[4] Guzdial, M. 2008. Education: Paving the way for<br />
computational thinking. Commun. ACM, 51, 25-27.<br />
[5] Isbell, C. L., Stein, L. A., Cutler, R., Forbes, J., Fraser, L.,<br />
Impagliazzo, J., Proulx, V., Russ, S., Thomas, R. & Xu, Y.<br />
2010. (Re)defining computing curricula by (re)defining<br />
computing. SIGCSE Bull., 41, 195-207.<br />
[6] Lister, R., Adams, E. S., Fitzgerald, S., Fone, W., Hamer, J.,<br />
Lindholm, M., McCartney, R., Moström, J. E., San<strong>der</strong>s, K.,<br />
Seppälä, O., Simon, B. & Thomas, L. 2004. A multi-national<br />
study <strong>of</strong> reading and tracing skills in novice programmers.<br />
Working group reports from ITiCSE on Innovation and<br />
technology in computer science education. Leeds, United<br />
Kingdom: ACM.<br />
[7] Lopez, M., Whalley, J., Robbins, P. & Lister, R. 2008.<br />
Relationships between reading, tracing and writing skills in<br />
introductory programming. Proceeding <strong>of</strong> the Fourth<br />
international Workshop on Computing Education Research.<br />
Sydney, Australia: ACM.<br />
[8] Ma, L., Ferguson, J., Roper, M. & Wood, M. 2011.<br />
Investigating and improving the models <strong>of</strong> programming<br />
82<br />
concepts held by novice programmers. Computer Science<br />
Education, 21, 57 - 80.<br />
[9] McCracken, M., Almstrum, V., Diaz, D., Guzdial, M.,<br />
Hagan, D., Kolikant, Y., Laxer, C., Thomas, L., Utting, I. &<br />
Wilusz, T. 2001. A multi-national, multi-institutional study<br />
<strong>of</strong> assessment <strong>of</strong> programming skills <strong>of</strong> first-year CS<br />
students. Working group reports from ITiCSE on Innovation<br />
and technology in computer science education. Canterbury,<br />
UK: ACM.<br />
[10] National Research Council. 2010. Report <strong>of</strong> a Workshop on<br />
the Scope and Nature <strong>of</strong> Computational Thinking. Available:<br />
http://www.nap.edu/catalog.php?record_id=12840 [Accessed<br />
10-05-2011].<br />
[11] Pane, J. F., Ratanamahatana, C. A. & MYERS, B. A. 2001.<br />
Studying the language and structure in non-programmers'<br />
solutions to programming problems. International Journal <strong>of</strong><br />
Human-Computer Studies, 54, 237-264.<br />
[12] Pólya, G. 1985. How To Solve It, London, Penguin.<br />
[13] Robins, A., Rountree, J. & Rountree, N. 2003. Learning and<br />
Teaching Programming: A Review and Discussion.<br />
Computer Science Education, 13, 137 - 172.<br />
[14] Saeli, M. 2012. Teaching Programming for Secondary<br />
School: a Pedagogical Content Knowledge Based Approach.<br />
PhD, Eindhoven <strong>University</strong> <strong>of</strong> Technology.<br />
[15] Saknini, V. & Hazzan, O. 2008. Reducing Abstraction in<br />
High School Computer Science Education: The Case <strong>of</strong><br />
Definition, Implementation, and Use <strong>of</strong> Abstract Data Types.<br />
J. Educ. Resour. Comput., 8, 1-13.<br />
[16] Schulte, C. & Bennedsen, J. 2006. What do teachers teach in<br />
introductory programming? Proceedings <strong>of</strong> the second<br />
international workshop on Computing education research.<br />
Canterbury, United Kingdom: ACM.<br />
[17] Sixsmith, J. & Murray, C. 2001. Ethical issues in the<br />
documentary analysis <strong>of</strong> e-mail posts and archives.<br />
Qualitative Health Research, 11, 423-432.<br />
[18] Strauss, A. & Corban, J. 1998. Basics <strong>of</strong> qualitative<br />
research: Techniques and procedures for developing<br />
grounded theory, London, Sage Publications Ltd.<br />
[19] Sweller, J. 1988. Cognitive Load During Problem Solving:<br />
Effects on learning. Cognitive Science, 12, 257-285.<br />
[20] Venables, A., Tan, G. & Lister, R. 2009. A closer look at<br />
tracing, explaining and code writing skills in the novice<br />
programmer. Proceedings <strong>of</strong> the fifth international workshop<br />
on Computing education research workshop. Berkeley, CA,<br />
USA: ACM.<br />
[21] Wing, J. 2011. Research Notebook: Computational Thinking<br />
- What and Why? The Link. Pittsburgh, PA: Carneige<br />
Mellon.
Preparing Teachers for Teaching Informatics: Theoretical<br />
Consi<strong>der</strong>ations and Practical Implications<br />
Vassilios Dagdilelis<br />
Department <strong>of</strong> Educational and Social Policy<br />
<strong>University</strong> <strong>of</strong> Macedonia<br />
Thessaloniki, Greece<br />
+30 2310 891 336<br />
dagdil@uom.gr<br />
ABSTRACT<br />
The integration <strong>of</strong> ICT in Primary and Secondary Education is<br />
consi<strong>der</strong>ed <strong>of</strong> great importance for the enhancement <strong>of</strong> both<br />
teaching and learning. However, a successful integration <strong>of</strong> ICT in<br />
the teaching practice requires a well-organized training <strong>of</strong><br />
teachers. In this paper, we present qualitative results from a<br />
national-scale program aiming at training Secondary Education<br />
teachers in the usage <strong>of</strong> ICT. Specifically, we focus on the training<br />
<strong>of</strong> Informatics and Computer Science teachers that were for the<br />
first time included in this training program taking place the last 12<br />
years in Greece. These trainee-teachers will act as trainers <strong>of</strong><br />
teachers after the successful completion <strong>of</strong> their own training.<br />
Several important aspects <strong>of</strong> organizing the training <strong>of</strong> trainers are<br />
examined, while some findings concerning the major problems <strong>of</strong><br />
preparing teachers for teaching Informatics are analysed.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers and Education]: Computer and Information<br />
Science Education – Computer science education, literacy.<br />
General Terms<br />
Teachers <strong>of</strong> Computer Science.<br />
Keywords<br />
Secondary Computing Education, Teacher’s Training.<br />
1. INTRODUCTION<br />
Over these last 12 years in Greece, a very large program has<br />
gradually been implemented whose aim is to integrate ICT into<br />
education generally, and into teaching practice specifically.<br />
Recently, Informatics and Computer Science teachers were also<br />
included. Training <strong>of</strong> teachers is an important part <strong>of</strong> this project<br />
– especially Informatics’ teachers. Preparation <strong>of</strong> the trainers is<br />
also an important part <strong>of</strong> the project. It is obvious that the success<br />
<strong>of</strong> the whole program lies heavily in the successful training <strong>of</strong> the<br />
relatively small number <strong>of</strong> teachers that are going to train a great<br />
number <strong>of</strong> teachers in the next face <strong>of</strong> the program. In this paper,<br />
several important aspects <strong>of</strong> this preparation <strong>of</strong> trainers are<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that<br />
copies bear this notice and the full citation on the first page. To copy<br />
otherwise, or republish, to post on servers or to redistribute to lists,<br />
requires prior specific permission and/or a fee.<br />
Conference’10, Month 1–2, 2010, City, State, Country.<br />
Copyright 2010 ACM 1-58113-000-0/00/0010 …$15.00.<br />
83<br />
Stelios Xinogalos<br />
Department <strong>of</strong> Technology Management<br />
<strong>University</strong> <strong>of</strong> Macedonia<br />
Naoussa, Greece<br />
+30 23320 52469<br />
stelios@uom.gr<br />
examined and some findings, concerning the major problems <strong>of</strong><br />
preparing teachers for teaching Informatics are.<br />
2. ORGANIZING THE TRAINING OF<br />
TEACHERS<br />
The preparation <strong>of</strong> trainers takes place in training schools named<br />
with the Greek acronym “PAKE” which stands for<br />
PAnepistimiaka Kendra Ekpedefsis, meaning "<strong>University</strong> Centers<br />
for Educating (Trainers)". These centers periodically operate as<br />
schools preparing trainers. In other words, they are a very special<br />
kind <strong>of</strong> educational establishment where trainers <strong>of</strong> teachers are<br />
instructed by highly specialized educators.<br />
Integrating effectively technology into teaching is a very<br />
demanding task. Mishra και Koehler ([8], [9]) proposed the<br />
TPACK framework. Accordingly to this framework, effective<br />
technology integration for teaching specific content or subject<br />
matter requires a deep un<strong>der</strong>standing and a negotiation between<br />
three components: Technology, Pedagogy, and Content. Their<br />
combination is necessary, as a form <strong>of</strong> a specific expertise <strong>of</strong> the<br />
teacher. This framework, eventually, should be consi<strong>der</strong>ed in a<br />
broa<strong>der</strong> sense: in many cases the social dimension <strong>of</strong> a specific<br />
subject matter is also important – and Informatics/Computer<br />
Science belongs to this category.<br />
Preparing trainers is an even more demanding goal, from two<br />
points <strong>of</strong> view:<br />
(a) in the international literature there is a relatively sufficient<br />
number <strong>of</strong> research findings on the preparation <strong>of</strong> teachers in<br />
various subjects. There are however, far less data regarding<br />
the trainers <strong>of</strong> teachers and trainers <strong>of</strong> secondary pr<strong>of</strong>essors.<br />
(b) the trainers, both during their preparation, and when actually<br />
training other teachers, have to operate almost simultaneously<br />
at different levels: they have to prepare teachers who, in turn,<br />
will teach students. Consequently, they must learn how to get<br />
prepared, in or<strong>der</strong> to teach teachers who will teach students.<br />
The following diagram, with each arrow meaning “educates”,<br />
may better explain the role <strong>of</strong> PAKE:<br />
Figure 1. Educating trainers in “PAKE”.
Thus, training <strong>of</strong> trainers in PAKE requires negotiation on three<br />
distinct levels:<br />
� Level 1: teachers teach programming to students. Several<br />
research findings point out the difficulties faced by students<br />
when dealing with programming concepts and students’<br />
perceptions when attempting to solve problems [2], [7]. The<br />
concept <strong>of</strong> teaching scenario was created for the teaching <strong>of</strong><br />
programming. A teaching scenario is a detailed description <strong>of</strong><br />
a teaching module which can last more than one hour. Apart<br />
from the description <strong>of</strong> the learning objectives and the lesson<br />
itself, each teaching scenario includes information concerning<br />
classroom organization, the learning theories which the course<br />
is founded on, and when necessary, elements on epistemology.<br />
The teaching scenario attempts to provide a broad and indepth<br />
analysis <strong>of</strong> the course the teacher has prepared. So, in<br />
this context the teacher must know how to create teaching<br />
scenarios, which means that he/she must have knowledge and<br />
skills related to the Didactics <strong>of</strong> the specific subject (i.e. <strong>of</strong><br />
Informatics): teaching methods, issues associated with<br />
learning and the problems faced by students, assessment<br />
techniques and a wide range <strong>of</strong> other knowledge and skills.<br />
� Level 2: trainers instructing teachers. Besides the general<br />
principles <strong>of</strong> adult education, there is no substantial research<br />
on findings specifically on the training <strong>of</strong> Informatics and<br />
Computer Science teachers. The training scenario - similar to<br />
the teaching scenario- is the central idea <strong>of</strong> these courses. A<br />
training scenario describes in detail a unit (a piece <strong>of</strong><br />
knowledge) which is taught to teachers. As is the case with<br />
the teaching scenario, where a very important element is<br />
students’ misconceptions, a significant part <strong>of</strong> the training<br />
scenarios makes reference to the basic problems that may<br />
arise in the training <strong>of</strong> teachers. A characteristic problem is<br />
Greek teachers’ awareness (or lack <strong>of</strong>) the necessity <strong>of</strong> such<br />
training. A typical response is: what can 140 training hours <strong>of</strong><br />
seminars <strong>of</strong>fer to teachers with 10-15 years <strong>of</strong> teaching<br />
experience? Teachers also tend to pose specific objections:<br />
why, for instance, they have to be trained in the teaching <strong>of</strong><br />
Logo and Logo-like environments (such as Scratch and the<br />
Greek environment Xelonokosmos, the StarLogo and Turtle<br />
Art) – environments, specifically constructed for teaching<br />
purposes (and thus not really helpful)? The program tries to<br />
include all the necessary elements in or<strong>der</strong> to provide<br />
complete training to Informatics/ Computer Science teachers<br />
(CS/ICT teachers) and to provide satisfactory answers to this<br />
kind <strong>of</strong> questions as well [3].<br />
� Level 3: training trainers (educators at PAKE teach trainers).<br />
Few research data exist in this area – if any. In this case there<br />
is no particular scenario to follow, as is the case with levels 1<br />
& 2 above, as the content and purpose <strong>of</strong> training trainers is<br />
too broad to fit into a single scenario model. A key element in<br />
the training <strong>of</strong> trainers was the distinction between levels 1<br />
and 2. For instance, the trainers themselves many times<br />
questioned the need to actually create training seminars. The<br />
training <strong>of</strong> trainers does not only include preparing them to<br />
deal with new technologies that are known but have not yet<br />
been introduced in education - such as tablets and<br />
smartphones, but also to prepare them for technologies that<br />
are completely unknown. How does one prepare trainers in<br />
technologies that are unknown? Among other things, certain<br />
recent changes in the Computer Science course at school, has<br />
resulted in added obstacles to training. In Greek schools, the<br />
interdisciplinary approach to the various subjects has recently<br />
been introduced along with a ‘projects’ approach (usually<br />
84<br />
students working in groups on a specific project topic over the<br />
semester). Even in IT, there is an evident orientation to digital<br />
literacy or computers/information literacy. These changes are<br />
creating a new framework, a new "school ecology" in which<br />
the IT course is a part <strong>of</strong>. As is perhaps to be expected, these<br />
changes were not immediately accepted by all teachers thus<br />
compounding the task <strong>of</strong> training [4], [5], [6].<br />
3. DATA<br />
Teachers <strong>of</strong> CS/ICT followed an intensive program in or<strong>der</strong> to<br />
become trainers <strong>of</strong> other teachers <strong>of</strong> CS/ICT. This program lasted<br />
380 hours and almost 60 selected teachers followed the program.<br />
The recording <strong>of</strong> data related to all levels <strong>of</strong> teachers’ preparation,<br />
is part <strong>of</strong> a larger system for recording and evaluating the progress<br />
<strong>of</strong> the whole project, a project related to the pedagogical and<br />
didactical use <strong>of</strong> ICT. These data are <strong>of</strong> course a useful guide for<br />
the further development <strong>of</strong> the whole project.<br />
Along with this material, we had access to the material trainers<br />
display themselves at various sites - with open access for all. This<br />
material either falls within the framework <strong>of</strong> their training or not -<br />
for instance, can be personal blogs. Finally, other data, such as<br />
<strong>of</strong>ficial documents and papers written and presented by the<br />
trainers, were gathered from various sources.<br />
In this paper we refer briefly to some <strong>of</strong> our findings that emerged<br />
from the qualitative analysis <strong>of</strong> the data. The data analysis has not<br />
yet completed, but the first important results are already<br />
emerging. The time range, the volume (quantity) and the diversity<br />
<strong>of</strong> the data, even if they do not guarantee the reliability <strong>of</strong> our<br />
results, certainly are an important factor <strong>of</strong> this reliability.<br />
4. SOME FINDINGS<br />
Among the subjects that are taught in the context <strong>of</strong> trainers’<br />
training for Informatics/Computer Science, programming<br />
occupies a dominant position. Taking into account the fact that<br />
the training focuses on secondary education, it is obvious that the<br />
subjects <strong>of</strong> Information Technology that are taught, do not cover<br />
the whole field <strong>of</strong> science. Programming stands out among these<br />
subjects, since it is consi<strong>der</strong>ed to be more closely connected with<br />
mental processes, such as problem solving, in comparison to other<br />
subjects that occur in sciences. Thus, the value <strong>of</strong> programming<br />
and algorithm design is <strong>of</strong>ten consi<strong>der</strong>ed to be more general, and<br />
as such, surpasses the boundaries <strong>of</strong> the school course and has a<br />
more general educational value. Most <strong>of</strong> the findings mentioned<br />
below refer to programming activities, since in this field the<br />
didactical phenomena we present become clearer.<br />
The transformation <strong>of</strong> teachers from instructors to trainers <strong>of</strong><br />
teachers requires getting familiar with a variety <strong>of</strong> new theoretical<br />
concepts, as well as acquiring knowledge <strong>of</strong> various theories<br />
related to the learning and teaching <strong>of</strong> IT issues. This field is<br />
generally known as Didactics <strong>of</strong> Informatics. It is obvious that<br />
these theoretical concepts have no value per se, but as tools with<br />
which the teacher can better un<strong>der</strong>stand didactical phenomena and<br />
act accordingly. However, these concepts are <strong>of</strong>ten related to<br />
views on education and learning, which do not get easily accepted<br />
by teachers. Typical examples <strong>of</strong> such concepts are the concepts<br />
<strong>of</strong> constructivism by J. Piaget and the theories <strong>of</strong> L. Vygotsky.<br />
Mo<strong>der</strong>n theories on the organization <strong>of</strong> a course (such as the<br />
project method or the so-called student-centered teaching) are<br />
based on the socio-constructivism theories. Consequently, these<br />
theories have a great practical usefulness.
However, even when such learning theories get accepted by<br />
teachers, they are sometimes superficially interpreted and<br />
applied. For example, adopting a Logo or Logo-like environment<br />
for teaching programming, does not necessarily mean that a<br />
constructivism teaching approach has also been "automatically"<br />
adopted. The teaching scenario plays an important role too in<br />
achieving good learning results. Asking students to print 10 or<br />
more times a “Hello world” message in Scratch, in or<strong>der</strong> to<br />
comprehend the “repeat N times” programming construct, cannot<br />
be consi<strong>der</strong>ed as a good example <strong>of</strong> constructivism.<br />
Comprehending the potential benefits <strong>of</strong> various learning theories<br />
and using them as tools for devising successful didactical<br />
interventions cannot be easily taught to trainers and mainly<br />
applied by them. The theories that seem to be more popular<br />
among trainers are collaborative learning, constructivism,<br />
exploratory/discovery learning, social development theory and<br />
cultural context. The most referenced (by trainers) proponents <strong>of</strong><br />
such theories are Piaget, Vygotsky and Bruner. Unfortunately,<br />
this "popularity" does not seem to be applied into teaching<br />
actions: in most cases, trainers are using typical teaching methods<br />
in their practice, claiming however that these methods are<br />
constructivist ones. In fact, the example <strong>of</strong> learning theories to<br />
which we refer is <strong>of</strong> great importance, because it is part <strong>of</strong> the socalled<br />
"theoretical topics" for which, too <strong>of</strong>ten, teachers and<br />
consequently their trainers, express strong objections.<br />
More precisely, an important obstacle <strong>of</strong> the training process lies<br />
in the unwillingness <strong>of</strong> some trainers to participate actively in<br />
training seminars or courses on topics they believe that have a<br />
theoretical nature and/or they already have experience in. It is<br />
not that trainers don't believe Vygotsky's Theory or Piaget's<br />
points <strong>of</strong> view: they believe that it is useless to learn any learning<br />
theory - such theories have nothing to do with "real" teaching and<br />
real classes. A typical example is the teaching <strong>of</strong> programming<br />
itself, a topic that most Informatics teachers have more or less<br />
experienced. The personal experience <strong>of</strong> trainers is definitely<br />
important, but the experience that has been gathered from decades<br />
<strong>of</strong> rigorous research on the teaching and learning <strong>of</strong> programming<br />
is equally, or even more, important. Several teachers have little or<br />
no knowledge <strong>of</strong> the results <strong>of</strong> such research and unfortunately<br />
some <strong>of</strong> them do not get easily persuaded that these results can<br />
truly help them in teaching programming more successfully. It is<br />
not a matter <strong>of</strong> theory, it is a matter <strong>of</strong> practice. The educator has<br />
to work heavily on making such a seminar interesting for trainers.<br />
The educator has to present research results along with specific<br />
examples for utilizing them in or<strong>der</strong> to devise examples and<br />
assignments with the aim <strong>of</strong> avoiding common misconceptions<br />
and difficulties that are the source <strong>of</strong> students' more severe errors.<br />
So, subjects such as "learning theories" and didactic concepts, for<br />
example "misconceptions <strong>of</strong> students", are not consi<strong>der</strong>ed as<br />
really useful. The integration <strong>of</strong> programming into the school<br />
curriculum requires a specific didactic transposition: scientific<br />
knowledge and pr<strong>of</strong>essional practice is transferred to the school<br />
and included in the educational context [1]. Thus, the concepts are<br />
simplified to make them more easily un<strong>der</strong>stood; the course is<br />
organized into 45-50-minute sessions, which is the duration <strong>of</strong> a<br />
teaching period in Greek schools; the subject matter has been<br />
organized into introductory units, exercises, questions, problems.<br />
However, during this transposition <strong>of</strong> concepts into the school<br />
environment, another transformation takes place: namely, the<br />
school education <strong>of</strong> an entire scientific field favours certain types<br />
<strong>of</strong> problems, while marginalizing other aspects <strong>of</strong> the key<br />
concepts, such as algorithms, data structures, and programming<br />
85<br />
generally. For example, data structures as a means <strong>of</strong> encoding<br />
entities <strong>of</strong> the external world, such as images, text, etc., are rarely<br />
referred to in the school environment. Trainee-trainers in many<br />
cases did not fully acknowledge the importance <strong>of</strong> teaching or the<br />
activities on such issues, consi<strong>der</strong>ing them <strong>of</strong> no use, since they<br />
were not directly related to current concepts <strong>of</strong> programming – at<br />
least the “school version” <strong>of</strong> programming.<br />
Concepts, such as the didactic contract and the didactic<br />
transposition, research data related to concepts such as variable,<br />
repetition and selection structures, and data about teaching<br />
recursion are included in the curriculum <strong>of</strong> PAKE. Empirical<br />
research on the different programming environments is included<br />
as well. This approach allowed a clear distinction between the<br />
teaching scenarios and the training scenarios, as the theoretical<br />
scheme <strong>of</strong> reciprocal interactions and "observations" <strong>of</strong> a teaching<br />
system (see Figure 1) that clearly defines the framework within<br />
which these teaching interactions take place. The theoretical<br />
approach also (was supposed that) helped the un<strong>der</strong>standing <strong>of</strong><br />
didactic phenomena as phenomena, i.e. as observable events that<br />
are not random, but have causes producing them and therefore can<br />
be studied. As we stated earlier, theory <strong>of</strong> Didactics <strong>of</strong><br />
Informatics and especially <strong>of</strong> programming was accepted by the<br />
trainers - but it is almost sure that they are not all convinced for<br />
the value and usefulness <strong>of</strong> such "theoretical" subjects.<br />
It is also worth noting that, in a general way, when trainers are<br />
invited to invent new problems and situations for teaching new<br />
concepts, they <strong>of</strong>ten produce very questionable examples and<br />
scenarios. For instance, as mentioned earlier, when they were<br />
asked to invent situations for the introduction <strong>of</strong> the repetitive<br />
structure in Logo-like environments, usually they adopted a<br />
typical approach: draw a "regular" geometrical figure - such as a<br />
regular polygon - with the help <strong>of</strong> the "turtle" and show step-bystep<br />
how this figure could be constructed directly with a repetitive<br />
statement. However, sometimes the proposed figures are very<br />
complex and practically impossible to be constructed by young<br />
pupils. In other cases, the proposed initial activities are totally<br />
inappropriate for an introduction to repetitive statements, since<br />
they deal with complex arithmetic operations.<br />
Teachers in general are enthusiastic with educational<br />
programming environments that have adopted more or less the<br />
notion <strong>of</strong> game-based learning, such as Kodu, GameMaker and<br />
especially Scratch. These environments are consi<strong>der</strong>ed to<br />
motivate students and engage them in the cognitive demanding<br />
process <strong>of</strong> programming. Environments that <strong>of</strong>fer some kind <strong>of</strong><br />
structure editor for developing programs and avoiding syntax<br />
errors are consi<strong>der</strong>ed by teachers even more effective for<br />
introducing young students to programming. It is possible that<br />
trainers are very positive for this kind <strong>of</strong> environments, because<br />
they believe that programming is boring for young pupils. So, the<br />
adoption <strong>of</strong> "cool" environments and programming goals like the<br />
construction <strong>of</strong> games could be a motivation for the pupils.<br />
In spite <strong>of</strong> that, a significant number <strong>of</strong> teachers do not decide to<br />
use such environments in cases where programming is part <strong>of</strong> the<br />
written exams that take place at the end <strong>of</strong> the school year, in the<br />
3 rd grade <strong>of</strong> Gymnasium for example. Their main argument is that<br />
there is no way to examine in paper students' knowledge <strong>of</strong><br />
programming, since they were not obliged to learn the syntax <strong>of</strong><br />
the language. Of course, this argument is not valid since there are<br />
several ways <strong>of</strong> examining students’ knowledge, such as<br />
providing them pictures <strong>of</strong> statements, giving them segments <strong>of</strong><br />
code to find bugs, or fill in and so on. After all, the curriculum
states clearly that the aim is for students to be able to solve<br />
problems in an algorithmic way and not to learn a specific<br />
programming language. Things are worst in the 3 rd Grade <strong>of</strong><br />
Greek Lyceum where students are examined in Pan-Hellenic level<br />
in a subset <strong>of</strong> a Pascal-like pseudolanguage translated in Greek in<br />
or<strong>der</strong> to enter <strong>University</strong>. Although specialized tools have been<br />
developed, which <strong>of</strong>fer s<strong>of</strong>tware visualisation features that have<br />
proven valuable for students, such as step by step execution and<br />
visualisation <strong>of</strong> variables' state both for programs and structured<br />
flowcharts, several teachers decide not to use them in class. It is<br />
surprising that there are teachers teaching the specific course that<br />
have never used such tools, or at least have not encouraged their<br />
students to use them at home for their assignments, in or<strong>der</strong> to<br />
execute them step by step for testing their assumptions and facing<br />
their difficulties with flow <strong>of</strong> control. Their main arguments for<br />
not using such tools are the limited number <strong>of</strong> teaching hours and<br />
the large proportion <strong>of</strong> students per computer. It is obvious that<br />
such facts <strong>of</strong> the educational system act as great obstacles for<br />
persuading teachers towards the incorporation <strong>of</strong> valuable tools<br />
in the educational process. The problem, <strong>of</strong> course, remains<br />
identical when teachers follow courses in PAKE in or<strong>der</strong> to<br />
become trainers.<br />
A number <strong>of</strong> trainees, fortunately small, are moreover negative in<br />
getting trained in subjects that are not currently taught at Greek<br />
schools. Such a subject is object-oriented programming (OOP)<br />
that can be taught only in an elective basis in Secondary<br />
education. Although trainees recognize the fact that OOP is the<br />
dominant programming technique nowadays, and they accept the<br />
fact that several educational environments exist that can be used<br />
for a successful teaching <strong>of</strong> OOP to students, they still have<br />
objections. However, most <strong>of</strong> the trainees believe that sooner or<br />
later OOP should be included in the curriculum and are positive<br />
towards utilizing programming microworlds, such as Jeroo or<br />
objectKarel, for younger students and educational programming<br />
environments, such as BlueJ and Greenfoot for technologicaloriented<br />
Lyceum students. Especially trainees with no experience<br />
on OOP seem to appreciate even more the importance <strong>of</strong><br />
programming microworlds for teaching basic OOP concepts.<br />
5. FINAL REMARKS AND CONCLUSIONS<br />
The exploitation <strong>of</strong> ICT in Secondary Education is consi<strong>der</strong>ed<br />
necessary, as it is expected to enhance both teaching and learning.<br />
In Greece, a large-scale program is being carried out over the last<br />
12 years with the aim <strong>of</strong> integrating ICT into Secondary<br />
Education teaching practice. However, teachers <strong>of</strong> Informatics<br />
were not included in this program until recently. The training <strong>of</strong><br />
the very first Informatics' teachers, that are going to train<br />
massively other teachers in the usage <strong>of</strong> ICT in Secondary<br />
Computing Education, has just finished. As expected, this training<br />
<strong>of</strong> teachers carried out in or<strong>der</strong> to prepare them for training other<br />
teachers in the usage <strong>of</strong> ICT was not easy.<br />
Teachers <strong>of</strong> Informatics tend to give more emphasis on practice,<br />
rather than theory. Furthermore, they consi<strong>der</strong> themselves experts<br />
in technological issues and do not get easily, at least as easily as<br />
teachers in other disciplines, enthusiastic with ICT. Their<br />
transformation from instructors to trainers <strong>of</strong> teachers requires<br />
getting familiar with a variety <strong>of</strong> new theoretical concepts, as well<br />
as acquiring knowledge <strong>of</strong> various theories related to Didactics <strong>of</strong><br />
Informatics. As mentioned, several factors make this process<br />
difficult: unwillingness to participate actively in seminars with a<br />
theoretical nature, or relevant to issues they consi<strong>der</strong> they have<br />
experience in; learning theories and didactic concepts are not<br />
86<br />
consi<strong>der</strong>ed useful; superficial adoption <strong>of</strong> learning theories and<br />
misapplication <strong>of</strong> acquired knowledge; objections in getting<br />
trained in subjects that are not currently taught at schools;<br />
teachers' adherence in school books and their gathered experience<br />
do not let them innovate in their classrooms; adherence to<br />
intricacies <strong>of</strong> the educational system prevents them from utilizing<br />
ICT tools that they consi<strong>der</strong> useful. Of course, the aforementioned<br />
limitations are true for some <strong>of</strong> the trainees only.<br />
It is obvious that in or<strong>der</strong> to face these difficulties special<br />
attention must be paid in the material that is used during the<br />
preparation <strong>of</strong> trainers and the way it is presented. Theoretical<br />
issues have to be presented in conjunction with good examples <strong>of</strong><br />
their application in teaching specific IT issues in the classroom<br />
and not in general. An abundance <strong>of</strong> examples and practical ideas,<br />
as well as material that trainees can easily work on/modify and<br />
use in classroom has to be <strong>of</strong>fered. The experience gathered from<br />
the first cycle <strong>of</strong> preparing trainers for teachers <strong>of</strong> Informatics<br />
provides great help towards this direction. Our plans for further<br />
research include the rigorous investigation <strong>of</strong> the vast amount <strong>of</strong><br />
data that has been collected, in or<strong>der</strong> to shed more light on and<br />
enhance the training <strong>of</strong> Secondary Computing Education teachers.<br />
6. REFERENCES<br />
[1] Chevallard Y. 1994. Les processus de transposition<br />
didactique et leur théorisation, Contribution à l’ouvrage<br />
dirigé par G. Arsac, Y. Chevallard, J.-L. Martinand, Andrée<br />
Tiberghien (éds), La transposition didactique à l’épreuve, La<br />
Pensée sauvage, Grenoble, 135-180.<br />
[2] Dagdilelis V., Satratzemi M., Evangelidis G., 2004.<br />
Introducing Secondary Education Students to Algorithms<br />
and Programming. Education and Information Technologies,<br />
vol.9, no.2, 159-173.<br />
[3] Demetriadis et. Al. 2003. Cultures in negotiation: teachers’<br />
acceptance/resistance attitudes consi<strong>der</strong>ing the infusion <strong>of</strong><br />
technology into schools, Computers & Education, 41(3), 19-<br />
37.<br />
[4] Drent, M., and Meelisen, M. 2008. Which factors obstruct or<br />
stimulate teacher educators to use ICT innovatively?<br />
Computers & Education, 51, 187-199.<br />
[5] Jung, I. 2005. ICT-Pedagogy Integration in Teacher<br />
Training: Application Cases Worldwide. Educational<br />
Technology & Society, 8 (2), 94-101.<br />
[6] Khe Foon Hew, and Brush T. 2007. Integrating technology<br />
into K-12 teaching and learning: current knowledge gaps and<br />
recommendations for future research, Educational<br />
Technology Research and Development, 55 (3), 223-252.<br />
[7] Schwill A. 1997 Computer science education based on<br />
fundamental ideas, In: Passey, D.; Samways, B. (eds.):<br />
Information Technology. Supporting change through teacher<br />
education. Chapman Hall, 285–291.<br />
[8] Mishra, P. & Koehler, M. J. 2006. Technological<br />
Pedagogical Content Knowledge: A new framework for<br />
teacher knowledge. Teachers College Record, 108 (6), 1017-<br />
1054.<br />
[9] Koehler, M. J. & Mishra, P. 2008. Introducing TPCK. In J.<br />
A. Colbert, K. E. Boyd, K. A. Clark, S. Guan, J. B. Harris,<br />
M. A. Kelly & A. D. Thompson (Eds.), Handbook <strong>of</strong><br />
Technological Pedagogical Content Knowledge for<br />
Educators (pp. 1–29). NewYork: Routledge.
Grand challenges for the UK: Upskilling teachers to teach<br />
Computer Science within the Secondary Curriculum<br />
ABSTRACT<br />
Sue Sentance<br />
Anglia Ruskin <strong>University</strong><br />
Chelmsford, Essex, UK<br />
sue.sentance@anglia.ac.uk<br />
Mark Dorling<br />
Langley Grammar School<br />
Langley, Slough, UK<br />
markdorling@lgs.slough.sch.uk<br />
Recent changes in UK education policy with respect to ICT<br />
and Computer Science (CS) have meant that more teachers<br />
need the skills and knowledge to teach CS in schools.<br />
This paper reports on work in progress in the UK researching<br />
models <strong>of</strong> continuing pr<strong>of</strong>essional development (CPD)<br />
for such teachers. We work with many teachers who either<br />
do not have an appropriate academic background to teach<br />
Computer Science, or who do and have not utilised it in<br />
the classroom due to the curriculum in place for the last<br />
fifteen years. In this paper we outline how educational policy<br />
changes are affecting teachers in the area <strong>of</strong> ICT and<br />
Computer Science; we describe a range <strong>of</strong> models <strong>of</strong> CPD<br />
and discuss the role that local and national initiatives can<br />
play in developing a hybrid model <strong>of</strong> transformational CPD,<br />
briefly reporting on our initial findings to date.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers & Education ]: Computers and Information<br />
Science Education- Computer Science Education<br />
Keywords<br />
Computer Science Education, Continuing Pr<strong>of</strong>essional Development,<br />
CPD, K-12 Curriculum, High School<br />
1. INTRODUCTION<br />
Recent curriculum changes in the UK have meant that<br />
there is a huge demand for continuing pr<strong>of</strong>essional development<br />
(CPD) 1 in Computer Science (CS) and for more focus<br />
1 In this paper the term continuing pr<strong>of</strong>essional development<br />
will be used as equivalent to the term in-service training<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WIPSCE 2012 Hamburg, <strong>Germany</strong><br />
Copyright 2012 ACM X-XXXXX-XX-X/XX/XX ...$15.00.<br />
87<br />
Adam McNicol<br />
Long Road Sixth Form College<br />
Cambridge, UK<br />
amcnicol@longroad.ac.uk<br />
Tom Crick<br />
Cardiff Metropolitan <strong>University</strong><br />
Cardiff, Wales, UK<br />
tcrick@cardiffmet.ac.uk<br />
on CS in pre-service training (initial teacher training) for<br />
ICT teachers. Finding ways to support the re-skilling <strong>of</strong><br />
many teachers in the UK to be able to teach CS in school is<br />
<strong>of</strong> interest to universities, industry, government departments<br />
and teacher associations. This paper reports on the work being<br />
done to investigate models <strong>of</strong> CPD that will meet the<br />
needs <strong>of</strong> teachers who wish, or are required to, teach GCSE 2<br />
CS in their schools, and do not have sufficient up-to-date<br />
subject knowledge. Much research has been done on a variety<br />
<strong>of</strong> models for building and sustaining an effective and<br />
confident teaching workforce. In this paper, we consi<strong>der</strong> the<br />
context <strong>of</strong> CS Education CPD in the UK, and draw on the<br />
framework <strong>of</strong> CPD outlined by Aileen Kennedy [9], proposing<br />
a range <strong>of</strong> strategies that would lead to transformative<br />
CPD in the area <strong>of</strong> CS in schools.<br />
2. BACKGROUND TO CS EDUCATION IN<br />
SCHOOL IN THE UK<br />
There have been significant changes to compulsory education<br />
in the UK over the last 25 years, which resulted over<br />
time in CS essentially disappearing as a curriculum subject<br />
for un<strong>der</strong>-16 year olds and being replaced by Information &<br />
Communication Technology (ICT) [4]. However this situation<br />
in the UK is currently being reversed. Computing At<br />
School (CAS) is an organisation formed in 2009, to promote<br />
CS education in the UK and support CS teachers. Its efforts<br />
have been augmented by the effect <strong>of</strong> a lecture by Eric<br />
Schmidt, Executive Chairman <strong>of</strong> Google, criticising the lack<br />
<strong>of</strong> computer science education in UK schools [13], and also a<br />
report by The Royal Society describing the teaching <strong>of</strong> computer<br />
science in many schools as “highly unsatisfactory” [14,<br />
p.1]. The Royal Society report’s recommendations included:<br />
• increasing the number <strong>of</strong> teachers trained to teach Computer<br />
Science<br />
• improving in-service training for teachers<br />
• providing more technical resources for schools.<br />
The UK government subsequently announced that the National<br />
Curriculum for ICT in England was to be disapplied<br />
from September 2012, removing a prescriptive programme <strong>of</strong><br />
2 GCSE is a qualification in the UK that is taken at age 16.<br />
GCSE Computing and CS have just been re-introduced in<br />
the UK after approximately15 years
study and facilitating the teaching <strong>of</strong> more Computer Science<br />
in school. Whilst this is a hugely positive shift in education<br />
policy for the UK, and it is apparent that there are<br />
many schools willing to <strong>of</strong>fer CS to students, there are many<br />
teachers who feel they do not have the skills and an urgent<br />
need for more CPD for schools in this area.<br />
3. MODELS OF CPD<br />
There is a consi<strong>der</strong>able corpus <strong>of</strong> research on CPD spanning<br />
several decades, including hundreds <strong>of</strong> individual studies<br />
<strong>of</strong> different types <strong>of</strong> CPD and the evaluation <strong>of</strong> CPD.<br />
Teachers can participate in many hours <strong>of</strong> training or other<br />
prescribed CPD in school or college and take up external<br />
courses in or<strong>der</strong> to improve their own skills in teaching and<br />
learning but sometimes actual change is difficult to achieve,<br />
particularly as “change is a gradual and difficult process for<br />
teachers”[7]. As Bell & Gilbert report[2], sometimes even<br />
the most well-intentioned efforts to change do not succeed:<br />
“. . . many teachers, even after attending an inservice<br />
course, for example, feel unable to use<br />
the new teaching activities, curriculum materials<br />
or content knowledge to improve the learning<br />
<strong>of</strong> their students . . . Many teachers are aware <strong>of</strong><br />
this pattern and feel frustrated in their attempts<br />
to change” [2, p.9]).<br />
A particularly effective form <strong>of</strong> CPD is collaborative, which<br />
can be defined as “teachers working together on a sustained<br />
basis and/or teachers working with LEA or HEI or other<br />
pr<strong>of</strong>essional colleagues” [3] . In all but one <strong>of</strong> 266 studies<br />
<strong>of</strong> collaborative CPD reviewed by Cordingley, Bell, Rundell<br />
& Evans [3] there was a definite teacher improvement as a<br />
result.<br />
Kennedy [9] consi<strong>der</strong>s a wide range <strong>of</strong> models <strong>of</strong> CPD and<br />
proposes a framework through which they can be analysed.<br />
She defines nine different categories <strong>of</strong> CPD and places them<br />
on a spectrum in terms <strong>of</strong> their “potential capacity for transformative<br />
practice and pr<strong>of</strong>essional autonomy” [9, p.236].<br />
Kennedy’s nine categories <strong>of</strong> CPD are as follows:<br />
• Transmissional<br />
– training<br />
– award-bearing<br />
– deficit<br />
– cascade<br />
• Transitional<br />
– standards-based<br />
– coaching/mentoring<br />
– community <strong>of</strong> practice<br />
• Transformative<br />
– action research<br />
– transformative<br />
There is a notion that the transmissional type <strong>of</strong> CPD<br />
is less successful in facilitating teacher change than those<br />
that <strong>of</strong>fer a degree <strong>of</strong> teacher autonomy. Fraser, Kennedy,<br />
Reid & McKinney posit that “Formal planned opportunities,<br />
which are essentially transmissive, are unlikely to result<br />
in transformative pr<strong>of</strong>essional learning for teachers, because<br />
they attend primarily to occupational aspects <strong>of</strong> pr<strong>of</strong>essional<br />
learning” [6, p165]. In contrast, they consi<strong>der</strong> that transformational<br />
learning is more likely to take place where the<br />
opportunities for learning attend to the personal and social<br />
aspect <strong>of</strong> pr<strong>of</strong>essional learning [6]. This will be an important<br />
factor in establishing an effective model <strong>of</strong> CPD in the UK.<br />
88<br />
Another category <strong>of</strong> CPD is the community <strong>of</strong> practice,<br />
where there is a joint enterprise, mutuality and a shared<br />
repertoire <strong>of</strong> communal resources [15]. In practice, teachers<br />
working together towards a common goal, for example,<br />
implementing a new strategy, who share their experiences,<br />
talk the same language, and are willing to learn from one<br />
another, can be said to be a community <strong>of</strong> practice.<br />
Being mentored or coached is another way that a teacher<br />
can develop pr<strong>of</strong>essionally. With a peer coaching model [8],<br />
teachers <strong>of</strong> equal status work together; in contrast, mentoring<br />
assumes that the mentor has a higher level <strong>of</strong> expertise<br />
than the mentee [5]. Mentoring may be less likely to be<br />
transformative than coaching, whereas the coaching experience<br />
is designed such that the coachee is able to solve their<br />
own problems and thus become empowered to be able to<br />
effect change.<br />
As Lipowski, Jorde, Prenzel & Seidel report [10], there is<br />
also a need for institutional support within CPD. In a recent<br />
study, experts from a range <strong>of</strong> countries report an urgent<br />
need to “reform existing insititutional conditions, including<br />
existing cooperation or coordination structures between institutions<br />
involved in the TPD system” [10, p694]. The impact<br />
<strong>of</strong> effective CPD can be directly linked to school improvement.<br />
Opfer and Ped<strong>der</strong> report [12] that teachers in the<br />
highest performing schools reported participating in pr<strong>of</strong>essional<br />
learning activities with higher levels <strong>of</strong> effectiveness:<br />
they were <strong>of</strong> longer duration, were more active, and teachers<br />
shared what they had learned with colleagues more <strong>of</strong>ten.<br />
This demonstrates that achieving good-quality CPD can affect<br />
the performance <strong>of</strong> the school. An Ofsted [11] 3 report<br />
on pr<strong>of</strong>essional development (PD) supports this by stating<br />
that the weakest link in the chain is the way the schools<br />
evaluate the effectiveness <strong>of</strong> their PD activities.<br />
In terms <strong>of</strong> the particular case <strong>of</strong> CPD for CS teachers,<br />
particularly relevant is the balance between subject matter<br />
knowledge (SMK) and pedagogical content knowledge<br />
(PCK) as described in a recent review by Armoni [1]. Armoni<br />
emphasises the importance <strong>of</strong> learning how to teach<br />
CS as well as a teacher’s own un<strong>der</strong>standing <strong>of</strong> the subject<br />
and we hope to incorporate this work in our study, as well<br />
as the focus on a constructivist approach to the preparation<br />
<strong>of</strong> teachers in this area. However, our study will consi<strong>der</strong><br />
preparation <strong>of</strong> in-service teachers as well as pre-service<br />
teachers in the rapidly changing curriculum in the UK.<br />
4. RESEARCH QUESTIONS<br />
This paper reports on the beginning <strong>of</strong> a study to investigate<br />
the nature <strong>of</strong> the CPD required to facilitate more Computer<br />
Science in school in the UK (particularly England and<br />
Wales). The research questions addressed at the beginning<br />
<strong>of</strong> this study are as follows:<br />
• To what extent teachers without CS-related degrees<br />
can be trained and given the confidence to be able to<br />
teach CS in the curriculum up to age 16 (GCSE level)?<br />
• What are the effective models <strong>of</strong> teachers’ CPD and to<br />
what extent they can be applied to CS?<br />
• How does CPD for teachers in CS impact on improved<br />
provision <strong>of</strong> courses in England and Wales, and correspondingly<br />
on pupil learning in this area?<br />
There are many emerging initiatives in the UK around the<br />
3 Office for Standards in Education, Children’s Services and<br />
Skills
first two questions. We maintain that what is needed is to<br />
be able to address the third question, the impact on pupils.<br />
In aiming for an upskilling <strong>of</strong> teachers, the transformation<br />
will only be achieved if pupil learning is enhanced and there<br />
are increased opportunities for pupils.<br />
5. TRANSFORMATIVE CPD FOR<br />
SECONDARY TEACHERS IN THE UK<br />
A successful model <strong>of</strong> CPD can be bottom-up, top-down,<br />
or a combination <strong>of</strong> both. In the UK at present there are several<br />
initiatives arising from individual institutions becoming<br />
aware <strong>of</strong> local need for pr<strong>of</strong>essional development. In addition<br />
subject associations can have a role to play in establishing<br />
a national framework for upskilling the teaching workforce.<br />
5.1 Local initiatives: training<br />
An initial study has been carried out using two models<br />
<strong>of</strong> training for ICT teachers wishing to include more CS in<br />
their teaching. These courses are taking place at a university<br />
in the east <strong>of</strong> England. There are two approaches to<br />
university based training that have been trialled at Anglia<br />
Ruskin <strong>University</strong> by the first and second authors: (i) twilight<br />
sessions over a period <strong>of</strong> 10 weeks, and (ii) intensive<br />
courses over a number <strong>of</strong> days.<br />
The twilight model has been so far aimed at existing teachers<br />
who may find it difficult to attend an intensive course as<br />
it requires so much time away from school, and the intensive<br />
course at trainee teachers who have just finished their<br />
initial teacher education. In both approaches the delivery<br />
has been by current teachers/teacher educators who have<br />
experience in teaching CS in a school environment. It has<br />
been assumed that current practitioners are best placed to<br />
explain how to deliver this content to students and to avoid<br />
overcomplicating the process with unnecessary detail.<br />
5.1.1 Twilight Model<br />
The twilight model involves a 2.5-hour session each week<br />
over the course <strong>of</strong> a 10-week period. Each session is split<br />
into 1 hour <strong>of</strong> CS theory and 1.5 hours <strong>of</strong> programming.<br />
There are no expectations <strong>of</strong> existing knowledge or skills<br />
in either component but there is flexibility to react to the<br />
requirements <strong>of</strong> a particular cohort.<br />
Each theory session focuses on a different part <strong>of</strong> the CS<br />
specification each week. This is delivered through a mix<br />
<strong>of</strong> lectures, practical activities and discussion. Each theory<br />
session stands alone and therefore it is possible for teachers<br />
to miss some weeks and still gain benefit from attending further<br />
sessions without having to make significant investments<br />
in time covering missed material. Programming concepts<br />
are delivered through a mix <strong>of</strong> demonstration and paired<br />
working. The focus is on getting to the practical programming<br />
quickly so that the instructors can support the attendees.<br />
Both theory and programming are backed by websites,<br />
which contain the material taught during the sessions and<br />
supplementary material including video tutorials.<br />
5.1.2 Intensive Courses<br />
Intensive courses are provided over a period <strong>of</strong> five days<br />
where attendees will focus completely on developing the<br />
knowledge and skills required to deliver the CS specification.<br />
The content is similar to the twilight model. The intensive<br />
model has focused solely on developing programming skills<br />
89<br />
in attendees as this is an area that many trainees have no<br />
experience in. The intensive nature <strong>of</strong> the course gives attendees<br />
the opportunity to focus and practice without other<br />
competing demands on their time.<br />
5.1.3 Initial Findngs<br />
At the end <strong>of</strong> the intensive programme, 90% said that they<br />
had the skills to teach introductory programming. Teachers<br />
commented that: “The training was brilliant. I feel that I<br />
have learnt heaps and this has definitely sparked an interest<br />
in computing for me. I’m keen to learn more!”. Four months<br />
later, the attendees were asked if they had used what they<br />
had learned on the course. There was a low response but<br />
four out <strong>of</strong> six reported that they were applying the training<br />
and one teacher reported: “ It has given me the impetus to<br />
drive forward with introduction <strong>of</strong> GCSE Computing in my<br />
school.”<br />
After the twilight model, the respondents gave similar enthusiastic<br />
feedback on their experience <strong>of</strong> the course, for<br />
example: “My expectations have been exceeded. I’ve learned<br />
more about Python than I thought or hoped and the computer<br />
science lessons have been very thorough.”. Teachers<br />
were asked to rate their confidence levels before and after<br />
the course and the average confidence level rose from 2.9<br />
to 7.7 from the 10-week program. Some teachers did find<br />
the course very challenging: “ [I would like] . . . more time<br />
programming or do this first as by the later time I was really<br />
tired and found it more difficult to focus.” . Attending<br />
training after a long teaching day is demanding on teachers.<br />
Whilst there has been very positive feedback from teachers<br />
themselves, the trainers‘ observations are that some teachers<br />
are finding it difficult to fully engage with the course,<br />
particularly in the twilight model. Many teachers appear to<br />
find it difficult to set aside time to practise programming<br />
between sessions and unfortunately this significantly hampers<br />
their ability to become competent programmers. Like<br />
learning any new skill if it is not practised the knowledge<br />
quickly fades.<br />
In terms <strong>of</strong> the intensive model, the focus is also on getting<br />
to practical activity as soon as possible and the concepts are<br />
again conveyed through a mix <strong>of</strong> demonstration and paired<br />
working. Since there is no delay between sessions there is little<br />
opportunity for knowledge to fade so there is not the same<br />
imperative to practise between sessions. Once the course is<br />
over the same dangers relating to practice still apply.<br />
Other models <strong>of</strong> CPD will now be consi<strong>der</strong>ed which are<br />
being initiated by CAS.<br />
5.2 Local initiatives: community <strong>of</strong> practice<br />
Local hub meetings are held after school for groups <strong>of</strong><br />
teachers in the areas across the UK to discuss CS teaching<br />
issues. Guest speakers are invited to share their own areas<br />
<strong>of</strong> expertise. Typically, hub meetings take place two or<br />
three times per year with about 20 to 30 attendees, although<br />
this varies. Hubs provide a community <strong>of</strong> practice for participating<br />
teachers where they can discuss issues relating to<br />
teaching Computer Science in school and find out about new<br />
developments and resources.<br />
5.3 National initiatives: the Network<br />
<strong>of</strong> Excellence<br />
The Network <strong>of</strong> Computer Science Teaching Excellence is<br />
an initiative that has been set up by CAS and BCS Academy
<strong>of</strong> Computing, the learned society which is dedicated to advancing<br />
Computing (CS & IT) as an academic discipline. It<br />
is designed to utilise and formalise the hub system set up<br />
within CAS, with schools and universities across the country<br />
registering to support one another. The ambitious aim<br />
<strong>of</strong> the network is to establish CS teaching in at least 1000<br />
schools by 2015. It is planned that initially one university<br />
will support twenty-five secondary schools. Using the university<br />
as a central point <strong>of</strong> reference it is hoped that they<br />
will be able to better identify and adapt their support for the<br />
needs <strong>of</strong> the local schools. To make this model sustainable,<br />
schools will then support at least one other school.<br />
5.4 National initiatives: the Master Teacher<br />
To support the universities with developing and delivering<br />
teacher training materials that meet the needs <strong>of</strong> local<br />
schools, CAS is recruiting Master Teachers to form a local<br />
provi<strong>der</strong> team. Master Teachers will work with universities<br />
in the Network <strong>of</strong> Excellence. In the medium term it is<br />
hoped that this will create CS departments who are more<br />
aware <strong>of</strong> the needs <strong>of</strong> local schools and how to meet them<br />
as well as a national network <strong>of</strong> advanced skills teachers<br />
in CS. These ‘CAS Master teachers’ will be responsible for<br />
the delivery <strong>of</strong> CPD to schools in their region working in<br />
association with HE and industry. The structure and content<br />
<strong>of</strong> the courses can be determined by the local provi<strong>der</strong><br />
team but will be influenced by the CAS Curriculum, and<br />
will point to suitable resources on the CAS Community site.<br />
Each resource would be mapped to the points <strong>of</strong> study in<br />
the CAS Curriculum and in the long term ensure curriculum<br />
coverage.<br />
6. DISCUSSION/NEXT STEPS<br />
Using Kennedy (2005) classification we have a series <strong>of</strong><br />
different models that are planned to be used in the UK:<br />
• National<br />
– Cascade – the Network <strong>of</strong> Excellence<br />
– Coaching/mentoring – the Master Teacher model<br />
• Local<br />
– Training/deficit courses – subject-knowledge and<br />
pedagogy provided by universities<br />
– Community <strong>of</strong> Practice – local CAS hubs supporting<br />
a network <strong>of</strong> teachers<br />
Together, as a hybrid collection, we posit that these form<br />
a transformational model <strong>of</strong> CPD.<br />
This range <strong>of</strong> models <strong>of</strong> CPD has been planned to tackle<br />
a national emerging situation in teacher education in England<br />
and Wales. An online questionnaire has been designed<br />
to collate the perceived needs <strong>of</strong> ICT teachers in terms <strong>of</strong><br />
their preparation to teach CS in secondary school. This<br />
will enable us to answer the first research question above.<br />
The next focus in this research study will be on the third<br />
research question and on developing a research instrument<br />
to measure the impact <strong>of</strong> the pr<strong>of</strong>essional development on<br />
both pupils, teachers and schools. The initiatives outlined<br />
in this paper and the impact <strong>of</strong> the various models will be<br />
evaluated.<br />
7. CONCLUSION<br />
It cannot be assumed that providing training or facilitating<br />
pr<strong>of</strong>essional development activities for teachers will necessarily<br />
bring about the transformation <strong>of</strong> Computer Science<br />
90<br />
education in the UK that we require. There is a need in the<br />
UK for teachers to have confidence at an academic level to<br />
teach Computer Science; however pr<strong>of</strong>essional development<br />
relating to pedagogy must not be ignored. The question as<br />
to who is responsible for upskilling the teachers is increasingly<br />
important. Local initiatives may be most valued by<br />
teachers as they create networks and interpersonal relationships<br />
to support teachers. A local approach, however, can<br />
be ad-hoc and areas <strong>of</strong> the country will be neglected.<br />
8. REFERENCES<br />
[1] M. Armoni. Looking at secondary teacher preparation<br />
through the lens <strong>of</strong> computer science.<br />
Trans.Comput.Educ., 11(4):23:1–23:38, nov 2011.<br />
[2] B. Bell and J. Gilbert. Teacher development: a model<br />
from science education. Falmer Press, London, 1996.<br />
[3] P. Cordingley, M. Bell, B. Rundell, and D. Evans.<br />
The impact <strong>of</strong> collaborative CPD on classroom<br />
teaching and learning: how does collaborative<br />
continuing pr<strong>of</strong>essional development (CPD) for<br />
teachers <strong>of</strong> the 5-16 age range affect teaching and<br />
learning? Technical report, Social Research Unit,<br />
Institute <strong>of</strong> Education, 2003.<br />
[4] T. Crick and S. Sentance. Computing at school:<br />
Stimulating Computing Education in the UK. In<br />
Proceedings <strong>of</strong> the 11th Koli Calling International<br />
Conference on Computing Education Research, Koli<br />
Calling ’11, pages 122–123, NY, USA, 2011. ACM.<br />
[5] Department for Education and Skills. National<br />
framework for mentoring and coaching. Technical<br />
report, CUREE, 2005.<br />
[6] C. Fraser, A. Kennedy, L. Reid, and L. McKinney.<br />
Teachers‘ continuing pr<strong>of</strong>essional development:<br />
contested concepts, un<strong>der</strong>standing and models.<br />
Journal <strong>of</strong> In-Service Education, 33(22):153–169, 2007.<br />
[7] T. R. Guskey. Pr<strong>of</strong>essional development and teacher<br />
change. Teachers and Teaching, 8(3):381–391, 08/01;<br />
2012/06 2002.<br />
[8] B. Joyce and B. Showers. The evolution <strong>of</strong> peer<br />
coaching. Educational Lea<strong>der</strong>ship, 53(6):12–16, 1996.<br />
[9] A. Kennedy. Models <strong>of</strong> continuing pr<strong>of</strong>essional<br />
development: a framework for analysis. Journal <strong>of</strong><br />
In-Service Education, 31(2):235–250, 2005.<br />
[10] K. Lipowski, D. Jorde, M. Prenzel, and T. Seidel.<br />
Expert views on the implementation <strong>of</strong> teacher<br />
pr<strong>of</strong>essional development in European countries.<br />
Pr<strong>of</strong>essional Development in Education,<br />
37(5):685–700, 11/01; 2012/06 2011.<br />
[11] Ofsted. The logical chain: continuing pr<strong>of</strong>essional<br />
development in effective schools. Technical report,<br />
Ofsted, 2006.<br />
[12] V. D. Opfer. The lost promise <strong>of</strong> teacher pr<strong>of</strong>essional<br />
development in England. European Journal <strong>of</strong> Teacher<br />
Education, 34(1):3–24, 2011.<br />
[13] E. Schmidt. Eric Schmidt‘s MacTaggart lecture - full<br />
text, Friday 26th August 2011 2011.<br />
[14] The Royal Society. Shut Down or Restart? The way<br />
forward for Computing in UK Schools. Technical<br />
Report January 2012, DES 2448, 2012.<br />
[15] E. Wenger. Communities <strong>of</strong> Practice and Learning<br />
Systems. Organization, 7(2):225–246, 2000.
(Some) Grand Challenges <strong>of</strong> Computer Science Education<br />
in the Digital Age: A Socio-Cultural Perspective<br />
ABSTRACT<br />
The goal <strong>of</strong> this paper is to articulate (some <strong>of</strong>) the grand<br />
challenges that computer science education (CSE) at the school<br />
level faces in the digital age. Based on the socio-cultural<br />
theoretical idea that learning means entering a culture, I suggest<br />
viewing schooling as an encounter between intertwined cultures.<br />
Computer science (CS) students can be viewed as members <strong>of</strong><br />
many intertwined cultures: (a) they are newcomers to the<br />
pr<strong>of</strong>essional computing culture, but (b) most are also old timers<br />
in a “user” culture, living in a world surrounded by informationcommunication<br />
technologies (ICT), and also have informal<br />
learning experience (and values) within ICT, mostly from out-<strong>of</strong>school<br />
experience; and finally, (c) they are members <strong>of</strong> the<br />
school culture which itself is currently in a process <strong>of</strong><br />
transformation due to the digital age). Using this framework, I<br />
discuss two interrelated grand challenges <strong>of</strong> CSE in K-12 school<br />
levels: (1) the need to adjust the CS curriculum to better overlap<br />
with life-long learning skills; and (2) the need to better learn the<br />
characteristics <strong>of</strong> the “digital” generation and attune education to<br />
address these needs.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers and Education]: Computer and Information<br />
Science Education – computer science education.<br />
General Terms<br />
Human Factors<br />
Keywords<br />
Digital age, sociocultural theories, cultural encounter, challenges<br />
in CSE, K-12<br />
1. INTRODUCTION<br />
Due to the accelerating processes <strong>of</strong> globalization and<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, or<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
Conference’04, Month 1–2, 2004, City, State, Country.<br />
Copyright 2004 ACM 1-58113-000-0/00/0004…$5.00.<br />
Yifat Ben-David Kolikant<br />
School <strong>of</strong> Education<br />
Hebrew <strong>University</strong> <strong>of</strong> Jerusalem, Jerusalem<br />
Israel 91905<br />
Tel: 972-2-5882056<br />
Yifat.Kolikant@mail.huji.ac.il<br />
91<br />
digitalization, educational goals in many countries have changed.<br />
More emphasis is put on innovation and knowledge creation in<br />
or<strong>der</strong> to maintain or increase a country's competitiveness in a<br />
global economy. To this end, the new primary objective <strong>of</strong><br />
schooling is to provide opportunities for students to develop lifelong<br />
learning skills. Moreover, much work has been invested in<br />
transforming the school agenda to move beyond information and<br />
teaching practices and from being teacher-centered to being more<br />
student-centered. This creates a great challenge to education, and<br />
as I argue in this paper, CSE is no exception.<br />
This paper <strong>of</strong>fers a socio-cultural perspective to discuss (some<br />
<strong>of</strong>) the grand challenges that the digital age poses to CSE. The<br />
socio-cultural perspective is described in Section 2. Section 3<br />
presents a socio-culturally inspired viewpoint on CSE as a<br />
cultural encounter. Section 4 presents the challenges to education<br />
in the digital age. Section 5 discusses the challenges and<br />
opportunities CSE deals with due to the digital age and finally in<br />
Section 6, conclusions and implications are provided.<br />
2. THE SOCIO-CULTURAL<br />
THEORETICAL FRAMEWORK<br />
According to Lave and Wegner [10], learning is a process <strong>of</strong><br />
enculturation into a culture or a community <strong>of</strong> practice by<br />
newcomers' participation in peripheral yet legitimate, genuine<br />
activities <strong>of</strong> the target culture. The newcomers gradually, through<br />
continuous negotiation on the meaning <strong>of</strong> actions achieve a<br />
holistic viewpoint <strong>of</strong> pr<strong>of</strong>essional practice and increase their<br />
participation, ultimately becoming full-fledged participants.<br />
Instructional models that rely on this theoretical framework<br />
emulate "real world" participation in a culture in school. In these<br />
emulations, the teacher usually represents an old-timer in the<br />
culture whereas students are viewed as newcomers to the culture<br />
(e.g.,[4]). Figure 1 is a schematic representation <strong>of</strong> the social<br />
fabric as pictured by this approach. This power relationship<br />
faithfully reproduces the power relationship between newcomers<br />
and old-timers described by Lave and Wenger [10] in craft<br />
apprenticeship.<br />
Nevertheless, I claim that in certain domains, the metaphorical<br />
view <strong>of</strong> the educational milieu as a "busy tailoring shop" [4]<br />
where the expert tailor is the teacher and the students are merely<br />
apprentices (or newcomers) does not capture the complicated<br />
social fabric <strong>of</strong> the milieu and is therefore limited in its ability to
explain and anticipate learning difficulties. Today's students<br />
come<br />
Figure 1. A schematic representation <strong>of</strong> a cognitiveapprenticeship<br />
viewpoint on schooling<br />
to school with a wealth <strong>of</strong> informal knowledge that includes<br />
learning practices which have proved useful in out-<strong>of</strong>-school<br />
learning experiences, such as their interaction with the Internet<br />
as consumers and producers <strong>of</strong> knowledge as well as their<br />
gaming experiences (for further discussion <strong>of</strong> the impact <strong>of</strong><br />
students' computer experience on their cognitive skills, see Gee,<br />
[6]. As our expectations <strong>of</strong> schooling sociology shift so that we<br />
no longer assume that the teacher is the “knowledge authority”<br />
and instead expect students to assume responsibility for creating<br />
knowledge, our metaphors for educational milieus should reflect<br />
this change.<br />
Figure 2. A schematic representation <strong>of</strong> a cultural encounter<br />
viewpoint on schooling<br />
I suggest using the metaphor <strong>of</strong> a cultural encounter. This<br />
metaphor faithfully maintains students' “newcomerness” to the<br />
culture represented by the instructional setting but, at the same<br />
time, regards students as longstanding members <strong>of</strong> other,<br />
intertwined cultures. Moreover, this metaphor also enables us to<br />
regard school goals and practices as being driven from multiple,<br />
not necessarily overlapping cultures, thereby adding to the<br />
picture by emphasizing the multicultural nature <strong>of</strong> the sociology<br />
<strong>of</strong> schooling today. Figure 2 is a schematic representation <strong>of</strong> this<br />
viewpoint.<br />
92<br />
3. COMPUTER SCIENCE EDUCATION AS<br />
A CULTURAL ENCOUNTER<br />
In any domain, the teacher and the student are active in two<br />
cultures simultaneously, or metaphorically, each <strong>of</strong> them wears<br />
two hats simultaneously. The first hat is the hat <strong>of</strong> the target<br />
community <strong>of</strong> practice. The teacher, like the entire instructional<br />
setting, represents the community <strong>of</strong> the studied practice, and the<br />
student is the newcomer to that community <strong>of</strong> practice. Indeed,<br />
the CS curriculum is oriented toward the academic community<br />
whose un<strong>der</strong>standing <strong>of</strong> the computer world involves the<br />
abstraction, solution, and pro<strong>of</strong> <strong>of</strong> algorithmic problems.<br />
The second hat is the hat <strong>of</strong> participants in a school milieu.<br />
School is a cultural environment in itself. Brousseau [3] defines<br />
school activity as creating and playing didactical milieus,<br />
designed by the teachers according to the knowledge they want to<br />
devolve, with consi<strong>der</strong>ation <strong>of</strong> the students they teach. Learning<br />
is achieved when the student develops strategies to make sense<br />
<strong>of</strong> the milieu.<br />
CS students wear a third hat, that <strong>of</strong> a computer user or local<br />
developer. Their viewpoint regarding the CS world—what<br />
constitutes a good problem, an accountable approach to a<br />
solution, and a satisfactory solution <strong>of</strong> the problem—were shaped<br />
without much interaction with the other two "pr<strong>of</strong>essional" CS<br />
cultures: the CS academia and industry [1]. A user is defined by<br />
Turkle [18] as someone who is “involved in the machine in a<br />
hands-on way, but is not interested in the technology except as it<br />
enables an application” [18, p.32]. Many students, thus, are<br />
veterans in the world <strong>of</strong> computers; this shapes their<br />
un<strong>der</strong>standing <strong>of</strong> and interests in this world. Furthermore, as oldtimers<br />
in this world, they might delegitimize the CS curriculum<br />
as relevant to their CS life, which ultimately leads to a culture<br />
clash.<br />
Often the bor<strong>der</strong>line between users and programmers is not<br />
clear-cut [7, 12]. In fact, it is all a matter <strong>of</strong> the context. Students<br />
might mistakenly mis-contextualize educational computer-related<br />
situations and approach them “wearing” the user hat rather than<br />
the programmer’s hat. For example, in previous work, it was<br />
evident that both the user culture and school culture nurtured<br />
students (mis)conceptions <strong>of</strong> correctness as relative and partial.<br />
“Relatively correct” meant that that the program “worked,” i.e.<br />
produced the desired output with additional irrelevant output.<br />
Relative correctness meant that there was a “grain <strong>of</strong><br />
correctness” in the program, that is, something in the code was<br />
written correctly. In fact, students referred to program correctness<br />
as the sum <strong>of</strong> the correctness <strong>of</strong> the code constituents, as teachers<br />
would grade their programs. Moreover, some attributed the<br />
teacher’s judgment that the program was incorrect to<br />
‘‘pettiness’’ [1, 2].<br />
4. EDUCATIONAL CHALLENGES IN THE<br />
DIGITAL AGE<br />
4.1 The change in school culture<br />
Traditional schooling is characterized by its focus on information<br />
and teacher-centered practices. Resnick [16] and other scholars<br />
contend that these schooling foci are unsuitable for current times<br />
and should therefore change, moving beyond information and
ecoming more student-centered. Indeed, the primary educational<br />
goals in many countries have changed. More emphasis is put on<br />
innovation and knowledge creation. In other words, the school<br />
culture is un<strong>der</strong>going process <strong>of</strong> transformation.<br />
The partnership for 21 st century skills (P21), a broad coalition <strong>of</strong><br />
education, business, government and foundations, has advanced a<br />
powerful learning framework that captures the three essential<br />
learning goals most needed in our times. P21 indicates major sets<br />
<strong>of</strong> skills needed by tomorrow’s citizens: (a) learning and<br />
innovation, with emphasis on team-work skills, (b) information,<br />
media, and ICT skills, and (c) life and career skills since<br />
students need to develop the “ability to navigate the complex life<br />
and work environments in the globally competitive information<br />
age” (www.p21.org). P21 suggests reducing the number <strong>of</strong> core<br />
domains taught and then use these core domains as a basis to<br />
help students develop those skills, and moreover, weaving 21st<br />
century interdisciplinary themes into core subjects, such as<br />
financial, economic, business and entrepreneurial literacy and<br />
global awareness.<br />
4.2 The change in students’ capital<br />
Another body <strong>of</strong> work is devoted to un<strong>der</strong>standing the nature <strong>of</strong><br />
this generation who grew up surrounded by ICT, and adjust<br />
education accordingly. Some claim that digital natives are<br />
different and better learners than the so-called digital<br />
immigrants, those who were already adults when these<br />
technologies were introduced (e.g., [14, 15]). In contrast, others<br />
are concerned that this generation poses many challenges to<br />
educators because it has many attributes that hin<strong>der</strong> learning in<br />
traditional settings. According to Twenge [19], the generation<br />
born after 1970, which she terms Generation Me, is more selffocused,<br />
to the point <strong>of</strong> narcissism, and less motivated to learn<br />
something unless its immediate benefits are clear. Furthermore,<br />
while more assertive, students in this generation are less selfreliant.<br />
5. THE GRAND CHALLENGES TO CSE<br />
5.1 The rapid technological changes: a<br />
catalyst <strong>of</strong> a cultural clash<br />
The rapid technological changes in the world might intensify the<br />
culture clash between students who hold user-oriented<br />
viewpoints and educators aiming at facilitating students’<br />
un<strong>der</strong>standing <strong>of</strong> the pr<strong>of</strong>essional CS culture. Educators might<br />
find themselves racing to catch up with new innovative<br />
programming environments, chosen not purely due to their<br />
pedagogical potential but rather as a means to be judged as less<br />
“backward” by their students. Moreover, since ICT and learningby-doing,<br />
hands-on ICT incrementally touches on all school<br />
domains, students would be less likely to elect CS only because<br />
they get hands-on computer time. Educators thus face the need to<br />
balance between the need to maintain the students' sense <strong>of</strong><br />
relevance by adjusting the CS educational environment to the<br />
outside world while at the same time maintaining pedagogical<br />
and curricular merit. This challenge is not new, only the<br />
increment in the frequency <strong>of</strong> changes together with students’<br />
expectations intensifies it.<br />
93<br />
5.2 The ongoing transformation <strong>of</strong> schooling<br />
to a 21 st century education<br />
Some would assume that CS need not consi<strong>der</strong> a change since it<br />
is a high-tech related domain and hence is already aligned with<br />
the new goals <strong>of</strong> school, i.e. preparing students for life-long<br />
learning with ICT. They might argue that, programming, as well<br />
as the rest <strong>of</strong> CS, are important computer-related skills and<br />
knowledge. Moreover, CSE has always been embedded with<br />
hands-on student-centered activities. Most CS curricula go<br />
beyond information, aiming at nurturing problem-solving<br />
capabilities by means <strong>of</strong> algorithmic thinking and programming,<br />
an important asset for life-long learning.<br />
However, other 21st century skills, such as collaboration,<br />
communication, and career skills are not necessarily emphasized<br />
in these CS curricula. Principals, parents, and students might<br />
doubt the benefit <strong>of</strong> studying CS since it loses its unique<br />
importance as nurturing computing-oriented problem-solving<br />
skills. Some might argue that CS should be omitted from school<br />
programs in favor <strong>of</strong> other domains which emphasize more 21 st<br />
century skills. This, in itself, is a legitimate idea given the<br />
tendency to reduce the number <strong>of</strong> domains taught. Therefore,<br />
CSE goals should be extended to involve, in addition to<br />
acculturation into the pr<strong>of</strong>essional CS culture, the preparation <strong>of</strong><br />
students for life-long learning in the 21 st century.<br />
CSE can be re-designed to leverage 21 st century skills without<br />
compromising the knowledge, values, and practices <strong>of</strong> the<br />
pr<strong>of</strong>essional culture. Obviously, as any domain, more ICT tools<br />
can be used to support students’ learning, starting with<br />
simulations (<strong>of</strong> algorithms, scripts, the machine, and so forth).<br />
More importantly, CS prides itself on nurturing the required<br />
skills. It encourages creativity and critical thinking. It can also<br />
encourage communication and collaboration. For example,<br />
documentation can be taught as a way for communicating<br />
effectively with others at work. Additionally, CS can and should<br />
be used to facilitate cross-disciplinary instruction. For example,<br />
programming environments, such as Scratch [17] and Alice [5]<br />
enable students to express ideas and tell stories, and therefore<br />
can be used in humanities’ educational environments. In fact, CS<br />
has a unique advantage over other domains for advancing the<br />
goals <strong>of</strong> knowledge creation, creativity, and entrepreneurial skills<br />
because it enables students to create new technological tools. CS<br />
can also serve as a rich arena or teaching life and career skills.<br />
The young history <strong>of</strong> CS is replete with people and companies<br />
who have thought out <strong>of</strong> the box, made bold decisions,<br />
sometimes failed and sometimes succeeded and managed from<br />
there on. This type <strong>of</strong> career action resembles the future <strong>of</strong> these<br />
students: tomorrow’s adults in a rapidly changing world.<br />
Students can benefit from the exposure to the life paths <strong>of</strong> people<br />
and products in this field.<br />
5.3 Teaching "generation me"<br />
The fact that ICT has become more and more prevalent in<br />
students' lives in and out <strong>of</strong> school might intensify their useroriented<br />
computer-related capital (see, for example the report <strong>of</strong><br />
the OECD (2006) on the growth <strong>of</strong> student use <strong>of</strong> ICT,[13]). This<br />
computer-related capital influences students’ judgment as to<br />
what is important and where to invest their efforts to learn.<br />
Moreover, according to Twenge [19], today's students are less
motivated to learn something unless the immediate benefits are<br />
clear. Hence, CS teachers might experience even greater<br />
difficulties in introducing the pr<strong>of</strong>essional CS culture as<br />
legitimate, let alone the desired one. Simple introductory<br />
problems previously used to teach programming with gradual<br />
levels <strong>of</strong> difficulty, such as a program that prints "hello world"<br />
on the screen, might influence students' judgment for the worse<br />
about the worthiness <strong>of</strong> taking CS classes given that they can<br />
(or know someone who can) achieve more attractive outcomes in<br />
the use <strong>of</strong> the computer. Students' high level <strong>of</strong> frustration and<br />
their low self-reliance is also an obstacle to CS education.<br />
Debugging, for example, can be frustrating. Hence, educators<br />
face the tension between the need to provide students with<br />
attractive problems and less “hello-world”-like problems, so that<br />
the students will become engaged in their solution, while taking<br />
into consi<strong>der</strong>ation students’ “newcomerness” programming<br />
capabilities.<br />
This concern leads educators to <strong>of</strong>fer students problems the<br />
students might value as worth <strong>of</strong> their efforts, such as the<br />
development <strong>of</strong> computer games [11] and the processing <strong>of</strong><br />
media [8, 9]. Some teach introductory courses using visual<br />
programming environments that enable work processes more<br />
reminiscent <strong>of</strong> students’ experience with the user interface. For<br />
example, Alice 3 [5] enables students to program 3-D characters<br />
using a graphical drag-and- drop programming. Similarly,<br />
Scratch was designed to motivate all students to program [17].<br />
These environments enabled teachers to provide problems that<br />
students perceive as attractive while at the same time avoiding or<br />
postponing frustrating actions, such as debugging syntax.<br />
Both approaches rely on students' computer-related capital, as<br />
users, in an attempt to meet students where they are, provide<br />
them with an immediate sense <strong>of</strong> the relevance <strong>of</strong> school<br />
experience to their lives, and at the same time introduce them to<br />
the principles in programming in a way that the students might<br />
value and will be motivated to further their experience.<br />
Educators, however, should keep in mind that meeting students<br />
where they are should not aspire to leave them in their current<br />
cultural horizons, but rather to help the students expand their<br />
cultural perspective and cross the boundaries towards the<br />
pr<strong>of</strong>essional culture.<br />
6. CONCLUSIONS AND IMPLICATIONS<br />
The challenges articulated in this paper emerge from the<br />
recognized need to transform schooling so that it better prepares<br />
students for life within a digitalized and globalized world. The<br />
cultural-encounter metaphor is useful to detect challenges as<br />
interaction between two (or more) intertwined cultures. Part <strong>of</strong><br />
the resolution <strong>of</strong> these challenges includes the awareness <strong>of</strong><br />
educators <strong>of</strong> the cultural capital <strong>of</strong> their students, accumulated<br />
through the students’ in and out-<strong>of</strong> school experience with ICT.<br />
The other part is to bridge between students’ current<br />
perspectives and the desired target practices and values. To this<br />
end, in previous studies [1], it has been suggested that<br />
educational situations should be designed as fertile zones <strong>of</strong><br />
cultural encounter (FZCE), a metaphorical zone that enables<br />
using the students’ cultural capital as a bridge to the perspective<br />
represented by the teachers. FZCE occurs when the educational<br />
milieu clearly reflects the distinction between the intertwined<br />
94<br />
cultures, while making the cultural tools represented by the<br />
instructional setting accessible and relevant to the students.<br />
Creating FZCEs for today’s students is a great challenge. First,<br />
the target is complicated: it is about enculturation within the CS<br />
pr<strong>of</strong>essional culture while, at the same time, nurturing students’<br />
life-long learning capacities. Moreover, as part <strong>of</strong> the general<br />
transformation in school, instructional milieus should be<br />
designed with more student-centeredness, allowing for and<br />
encouraging students to manage their learning actions. However,<br />
CS has great potential to serve as an arena to teach 21 st century<br />
skills in ways students perceive as relevant and therefore are<br />
more motivated to invest mental effort.<br />
7. ACKNOWLEDGEMENT<br />
I thank M. Ben-Ari and S. Pollack for their insightful comments.<br />
8. REFERENCES<br />
[1] Ben-David Kolikant, Y., & Ben Ari, M. Fertile zones <strong>of</strong><br />
cultural encounter in computer science education, Journal <strong>of</strong><br />
the Learning Science, 2008, 1,17,1-32.<br />
[2] Ben-David Kolikant, Y., & Mussai, M. 'So my program does<br />
not run': definition, origins, and practical expressions <strong>of</strong><br />
students' (mis)conceptions <strong>of</strong> correctness, CSE, 18, 2, 131-<br />
151.<br />
[3] Brousseau, G. Theory <strong>of</strong> didactical situations in<br />
mathematics, Dordrecht: Kluwer Academic, 1997.<br />
[4] Collins, A., Brown, J. S., & Holum, A. (1991). Cognitive<br />
apprenticeship: Making thinking visible. American<br />
Educator, 12(6), 38-47.<br />
[5] Dann, W., & Cooper, S. Alice 3: concrete to abstract.<br />
Communications <strong>of</strong> the ACM, 2009, 52(8), 27–29.<br />
[6] Gee, P. J. (2003).What video games have to teach us about<br />
learning and literacy. New York: Palgrave Macmillan.<br />
[7] Goodell, H., Maulsby, D., Kuhn, S., & Traynor, C. End-user<br />
programming / informal programming. A workshop<br />
conducted at the ACM CHI 1999, SIGCHI Bulletin, 31 (4),<br />
1999, 17-21.<br />
[8] Guzdial, M. (2003). A media computation course for nonmajors.<br />
In Proceeding <strong>of</strong> the ITiCSE (pp. 104–108).<br />
[9] Guzdial, M. & Tew, A. E. Imagineering inauthentic<br />
legitimate peripheral participation: an instructional design<br />
approach for motivating computing education. ICER 2006,<br />
ACM Press, New York, NY. (pp. 51–58).<br />
[10] Lave, J., & Wenger, E. (1991). Situated learning:<br />
Legitimate peripheral participation. Cambridge, UK:<br />
Cambridge <strong>University</strong> Press.<br />
[11] Leutenegger, S. and Edgington, J. 2007. A games first<br />
approach to teaching introductory programming. In SIGCSE<br />
'07. ACM, New York, NY, 115-118.<br />
[12] Nardi, B. A. (1993). A small matter <strong>of</strong> programming:<br />
Perspectives on end user computing. Cambridge: MIT Press.<br />
Organisation for Economic Co-operation and Development
[13] OECD. (2006). Are students ready for a technology-rich<br />
world? What PISA studies tell us.<br />
http://www.oecd.org/document/31/0,3343,en_32252351_32<br />
236173_35995743_1_1_1_1,00.html. Retrieved 09.08.09.<br />
[14] Papert, S. (1996). The connected family: Bridging the<br />
digital generation gap. Atlanta, GA: Longstreet Press.<br />
[15] Prensky, M. (2001). Digital natives, digital immigrants. On<br />
the Horizon, 9(5), 1–6.<br />
[16] Resnick, M. 2002. Rethinking learning in the digital age. In<br />
The global information technology report: Readiness for the<br />
networked world, ed. G. Kirkman. New York: Oxford<br />
<strong>University</strong> Press.<br />
95<br />
[17] Resnick, M., Maloney, J. , Monroy-Hernández, A., Rusk,<br />
N., Eastmond, E., Brennan, K., Millner, A., Rosenbaum, E.,<br />
Silver, J., Silverman, B. Kafai., Y. (2009). Scratch:<br />
programming for all. Commun. ACM 52, 11, 60-67.<br />
[18] Turkle, S. (1995). Life on the screen: Identity in the age <strong>of</strong><br />
the Internet. New York: Simon and Schuster.<br />
[19] Twenge, J. (2006). Generation Me: Why today’s young<br />
Americans are more confident, assertive, entitled – and<br />
more miserable than ever before. New York: Free Press.
ABSTRACT<br />
Challenge and Creativity: Using .NET Gadgeteer In<br />
Schools<br />
Sue Sentance<br />
Anglia Ruskin <strong>University</strong><br />
Chelmsford, UK<br />
sue.sentance@anglia.ac.uk<br />
This paper reports on a study carried out in secondary<br />
schools in the UK with students learning to use .NET Gadgeteer,<br />
a rapid prototyping platform for building small electronic<br />
devices [32]. A case study methodology has been<br />
used. Some <strong>of</strong> the students involved in this four-monthlong<br />
project had some prior background in computer programming<br />
whereas for others this was completely new. The<br />
teaching materials provided a two-phase model <strong>of</strong> learning:<br />
an instruction phase followed by a creative phase, the latter<br />
utilising a bricolage approach to learning programming [30].<br />
The aim <strong>of</strong> the pilot was to generate an interest in building<br />
devices and stimulate creativity. The research found that<br />
the tangible nature <strong>of</strong> the .NET Gadgeteer modules helped<br />
to engage the students in becoming creative, and that students<br />
valued challenges with which they were not usually<br />
presented within the curriculum.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers & Education]: Computers and Information<br />
Science Education - Computer Science Education<br />
Keywords<br />
Secondary education, .NET Gadgeteer, high schools, computer<br />
programming, creativity, bricolage<br />
1. INTRODUCTION<br />
Micros<strong>of</strong>t .NET Gadgeteer 1 is a platform that enables<br />
rapid prototyping <strong>of</strong> small electronic gadgets and embedded<br />
hardware devices. It combines the advantages <strong>of</strong> objectoriented<br />
programming, sol<strong>der</strong>less assembly <strong>of</strong> electronics using<br />
a kit <strong>of</strong> hardware modules, and the quick fabrication <strong>of</strong><br />
physical enclosures using computer-aided design. The fact<br />
that .NET Gadgeteer covers a variety <strong>of</strong> sophisticated computer<br />
science and engineering skills, but requires minimal<br />
1 http://netmf.com/gadgeteer<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WIPSCE 2012 Hamburg, <strong>Germany</strong><br />
Copyright 2012 ACM X-XXXXX-XX-X/XX/XX ...$15.00.<br />
96<br />
Scarlet Schwi<strong>der</strong>ski-Grosche<br />
Micros<strong>of</strong>t Research Cambridge<br />
Cambridge, UK<br />
scarlets@micros<strong>of</strong>t.com<br />
prior knowledge, makes it especially suitable for education.<br />
.NET Gadgeteer has great potential in schools as it can be<br />
used to teach students simple electronics and computer programming<br />
as well as computer-aided design. Moreover, it is<br />
very motivating for young people to be able to build their<br />
own gadgets.<br />
A research project was initiated to investigate whether<br />
.NET Gadgeteer has the potential to be used in schools.<br />
This was designed to elicit students’ and teachers’ perceptions<br />
<strong>of</strong> the platform, and suggest directions for further research<br />
and development. A case study methodology was<br />
used, as defined by Stake [26], whereby different sources were<br />
drawn together to give an accurate description <strong>of</strong> how the<br />
pilot use <strong>of</strong> this platform was perceived by both teachers,<br />
students and researchers.<br />
This paper reports on the findings <strong>of</strong> this first study. In<br />
the paper, we will give an outline <strong>of</strong> the current situation<br />
with regard to Computer Science (CS) education in schools<br />
in the UK which provides some motivation for the current<br />
project. A brief description <strong>of</strong> .NET Gadgeteer will help to<br />
set the scene and the next section <strong>of</strong> the paper will focus on<br />
the methodology and findings <strong>of</strong> the research. The findings<br />
will be discussed and then further work suggested.<br />
2. COMPUTER SCIENCE EDUCATION IN<br />
SCHOOLS<br />
2.1 Background<br />
A recent report by the Royal Society, an influential academic<br />
body in the UK, begins with the statement “The current<br />
delivery <strong>of</strong> Computing education in many UK schools<br />
is highly unsatisfactory” [29, p.1]. Prior to this report, there<br />
had been an increasing awareness <strong>of</strong> the need for more Computer<br />
Science in schools in England and Wales. The Computing<br />
At School group 2 (CAS) has been very active since<br />
2008 in advocating the need for more Computer Science in<br />
the curriculum and supporting teachers on the ground [6].<br />
Since the publication <strong>of</strong> the Royal Society report, the government<br />
has started to implement change in the curriculum;<br />
more qualifications are being launched to feed the need<br />
<strong>of</strong> schools and pupils to study this subject within the curriculum.<br />
The Royal Society states that “Computer Science<br />
is sufficiently important and foundational that it should be<br />
recognised as a high status subject in schools, like mathematics,<br />
physics or history” [29, p.34]. These developments<br />
in the UK mirror those that have been happening elsewhere<br />
2 http://www.computingatschool.org.uk
in the world, for example, as described by Wilson et al. in<br />
the USA’s Running on Empty report produced by the CSTA<br />
[33], and as implemented in Israel [10].<br />
It is apparent that there are many schools keen to <strong>of</strong>fer<br />
Computing-related content to students aged 11-14 and<br />
also school qualifications in Computing/Computer Science<br />
to students aged 14-18. Within this climate, providing engaging<br />
resources and vehicles for learning Computer Science<br />
is very timely.<br />
As the emphasis grows on secondary Computing education,<br />
more research in this area is needed. It is important<br />
to look at the pedagogical approaches to bridge the gap between<br />
computer literacy and Computer Science as well as<br />
what should be taught within the curriculum [3]. Constructivist<br />
learning theories applied to Computer Science<br />
emphasise the active, subjective and constructive character<br />
<strong>of</strong> knowledge, placing students at the centre <strong>of</strong> the learning<br />
process [13]. Specifically, constructivist learning is based on<br />
students’ active participation in problem-solving and critical<br />
thinking regarding a learning activity which they found<br />
relevant and engaging [11]. As a learning theory, it has pr<strong>of</strong>oundly<br />
influenced the teaching <strong>of</strong> programming [2]. Experiential<br />
learning stems from constructivism and is a term<br />
which can be used to describe the design <strong>of</strong> activities which<br />
engage learners in a very direct way. It describes the process<br />
<strong>of</strong> engaging learners in an authentic experience in which<br />
they can make discoveries and experiment with knowledge<br />
at first hand. Through reflection, students construct new<br />
knowledge and ways <strong>of</strong> thinking about themselves, leading<br />
to deeper learning.<br />
Research to date in secondary computing education has<br />
investigated a range <strong>of</strong> approaches to structuring the curriculum<br />
[27, 10, 1], whilst other research has focused on<br />
tools and environments that may motivate and engage young<br />
people in the classroom such as Scratch[17], Alice[5], Greenfoot[15],<br />
and Kodu[16]. It is important, however, to address<br />
the approach to learning that un<strong>der</strong>lies these tools or environments.<br />
Scratch, for example, builds on work in Logo and<br />
on the constructionist ideas <strong>of</strong> Papert [17]. Papert used the<br />
term constructionism which is a combination <strong>of</strong> ‘constructivism’<br />
and ‘construction’ [21]. This is particularly relevant<br />
for .NET Gadgeteer which involves physically constructing<br />
devices in an exploratory way.<br />
2.2 Engaging students<br />
There have been many discussions in the literature about<br />
how to engage students with Computer Science and suggestions<br />
for reasons why students do not have a positive<br />
attitude to the subject, for example [25, 9, 34, 4]. Downes<br />
and Looker suggest that the more IT students use at school,<br />
the more they are likely to take up computing-related subjects<br />
when they are given a choice [8]. Brinda, Puhlmann<br />
and Schulte discuss how to introduce Computer Science by<br />
working from what students already know from their ICT<br />
education and making Computer Science relevant to their<br />
own experience [3]. Pollock and Harvey integrated a range <strong>of</strong><br />
pedagogical approaches in or<strong>der</strong> to engage students more effectively<br />
and demonstrated the effectiveness <strong>of</strong> collaborative<br />
work and reflection on learning [22]. Collaborative working<br />
is especially important in our .NET Gadgeteer trials. Cutts,<br />
Esper and Simon discuss how to <strong>of</strong>fer all students Computing<br />
education in relation to “what a computer can do and<br />
how one can interact with it” [7] by giving students an un-<br />
97<br />
<strong>der</strong>standing that computers are deterministic, precise and<br />
comprehensible. This un<strong>der</strong>standing can be gained without<br />
necessarily learning to program. With .NET Gadgeteer, we<br />
hope that we can engage all students in terms <strong>of</strong> a better<br />
un<strong>der</strong>standing <strong>of</strong> how the devices and technology all around<br />
us works.<br />
Schulte and <strong>Knobelsdorf</strong> look at attitudes towards Computer<br />
Science using a biographical approach, and note the<br />
differences between those that regard themselves as insi<strong>der</strong>s<br />
and outsi<strong>der</strong>s [25]. They recommend that teachers “should<br />
intertwine introduction to CS (e.g. learning programming<br />
- a design activity) with learning pr<strong>of</strong>essional use . . . a major<br />
problem is to teach another world-image <strong>of</strong> CS” [p.37].<br />
The interest <strong>of</strong> students in real devices and current technology<br />
makes .NET Gadgeteer quite engaging in this regard,<br />
as users can relate it to the pr<strong>of</strong>essional world <strong>of</strong> developing<br />
devices.<br />
2.3 Using tangible environments<br />
Besides .NET Gadgeteer, other tangible devices are available<br />
which enable students to write programs using actual<br />
hardware components. These include Lego Mindstorms 3 ,<br />
the Scratch Pico Board 4 , Open <strong>University</strong>‘s SenseBoard [23]<br />
and Arduino 5 . All these hardware kits have different features<br />
but <strong>of</strong>fer the same experience <strong>of</strong> hardware in addition<br />
to s<strong>of</strong>tware, thus broadening the exposure to the way that<br />
technology works. The Raspberry Pi 6 is now available which<br />
also has the appeal <strong>of</strong> being tactile and exposed; this has generated<br />
a lot <strong>of</strong> enthusiasm, although it is a computer rather<br />
than a means to build devices such as the other devices described<br />
here.<br />
Marshall [18] and Horn et al. [12] both describe how tangible<br />
environments can have a very positive effect on collaborative<br />
and active learning, as they enable students to<br />
share work together in a very visible way. They also utilise<br />
concrete physical manipulation which can facilitate more effective<br />
or natural learning. Working with physical devices<br />
can encourage an exploratory or bricolage approach, as discussed<br />
next.<br />
2.4 The bricolage approach<br />
The concept <strong>of</strong> bricolage was introduced by Levi-Strauss<br />
in The Savage Mind [14]. It refers to a science <strong>of</strong> concrete<br />
development as an alternative to abstract planning, and was<br />
applied to the area <strong>of</strong> computer programming by Turkle and<br />
Papert [30]. It represents a mode <strong>of</strong> learning based on ‘tryit-and-see-what-happens’<br />
[2], <strong>of</strong> which Ben-Ari is rather critical<br />
claiming that “A student who exclusively uses such techniques<br />
is ultimately not qualified to work on the s<strong>of</strong>tware <strong>of</strong><br />
embedded and operating system, which requires the ability to<br />
create and test abstract hypotheses” [2]. However, Stiller [28]<br />
successfully used a bricolage approach when teaching programming<br />
by encouraging students to build on previous programs,<br />
based on a pedagogy <strong>of</strong> incremental problem-solving.<br />
The intention <strong>of</strong> this project was to use a bricolage approach<br />
when introducing .NET Gadgeteer to students.<br />
3 http://mindstorms.lego.com<br />
4 http://wiki.scratch.mit.edu/wiki/PicoBoard<br />
5 http://www.arduino.cc/<br />
6 http://www.raspberrypi.org/
Figure 1: .NET Gadgeteer modules<br />
Figure 2: Designer view <strong>of</strong> a .NET Gadgeteer device<br />
3. .NET GADGETEER<br />
.NET Gadgeteer is a platform for creating your own electronic<br />
devices using a wide variety <strong>of</strong> hardware and a powerful<br />
programming environment [32]. .NET Gadgeteer hardware<br />
consists <strong>of</strong> an ever increasing range <strong>of</strong> mainboards and<br />
modules. The environment is open source and therefore,<br />
new modules can be developed by any enthusiast. Students<br />
with little or no Computing background can build robot-like<br />
devices made up <strong>of</strong> components that sense and react to their<br />
environment using switches, displays, motor controllers, and<br />
more. Components are plugged into a mainboard and subsequently<br />
programmed to make them work together (see Figure<br />
1).<br />
.NET Gadgeteer originated at Micros<strong>of</strong>t Research Cambridge,<br />
UK. It was designed as a tool for researchers to make<br />
it faster and easier to prototype new kinds <strong>of</strong> devices. For<br />
example, a digital camera can be built in about half an hour.<br />
One <strong>of</strong> the motivations for the Gadgeteer project has been<br />
to provide “both a low threshold for entry – allowing non<br />
expert developers and designers to quickly sketch and construct<br />
functional devices – together with a high ceiling that<br />
allows experienced users to create sophisticated and capable<br />
devices that can be used in practice” [31]. This is ideal<br />
for education where there is a need to stretch and challenge<br />
youngsters to achieve their highest potential. Since<br />
then, the platform has proven to be <strong>of</strong> interest to hobbyists<br />
and for secondary and tertiary education. Micros<strong>of</strong>t<br />
Research has launched .NET Gadgeteer as open source s<strong>of</strong>tware/hardware,<br />
and .NET Gadgeteer kits are now available<br />
from a variety <strong>of</strong> hardware vendors.<br />
A starter .NET Gadgeteer kit consists <strong>of</strong> a mainboard,<br />
and various modules including a camera, joystick, buttons,<br />
LEDs, potentiometer, Ethernet port, and touch-sensitive<br />
display. In this study, the FEZ Spi<strong>der</strong> Starter Kit from<br />
98<br />
Figure 3: Example Visual C# code<br />
Figure 4: Aspects <strong>of</strong> learning with .NET Gadgeteer<br />
GHI Electronics 7 was used. In addition, there are many<br />
other sensors and other modules that can be added separately.<br />
Devices can be constructed by connecting modules<br />
with cables and then programming it with respect to the<br />
events triggered when using the device (for example, a ButtonPressed<br />
or PictureCaptured event in case <strong>of</strong> a digital<br />
camera). The programming language used is Visual C#<br />
with support for Visual Basic .NET coming soon. Figure<br />
1 shows the .NET Gadgeteer hardware with a range <strong>of</strong><br />
modules connected together. Figure 2 shows the graphical<br />
Gadgeteer Designer that is used to generate code for<br />
each module being used. Figure 3 shows the programming<br />
environment in Visual C#.<br />
Programming a .NET Gadgeteer device involves learning<br />
to program in Visual C#, but there are other skills involved<br />
in designing and implementing a gadget. Connecting the<br />
input and output modules to the mainboard will give students<br />
more <strong>of</strong> an un<strong>der</strong>standing <strong>of</strong> hardware, and there is<br />
the option to teach students the un<strong>der</strong>lying electronics <strong>of</strong><br />
the board. A so-called exten<strong>der</strong> board can be used to attach<br />
third-party hardware. In addition, the design <strong>of</strong> the case<br />
and the interaction <strong>of</strong> the user with the device will involve<br />
an un<strong>der</strong>standing <strong>of</strong> user-interface design and physical form<br />
factor. Being able to program the screen will also teach<br />
students an un<strong>der</strong>standing <strong>of</strong> graphics objects. Overall, the<br />
experience <strong>of</strong> using .NET Gadgeteer will cover a variety <strong>of</strong><br />
aspects <strong>of</strong> the CS school curriculum, as shown in Figure 4.<br />
4. CASE STUDY<br />
The case study was designed as an observation <strong>of</strong> how<br />
students and teachers used .NET Gadgeteer and their experiences.<br />
4.1 Aims and objectives<br />
7 http://www.ghielectronics.com/
The aims <strong>of</strong> the case study were to investigate the potential<br />
<strong>of</strong> .NET Gadgeteer to consi<strong>der</strong> which age group, which<br />
type <strong>of</strong> lessons, and which type <strong>of</strong> activities would be suitable<br />
for learning with the use <strong>of</strong> this tool. Specifically, the<br />
following research questions were addressed:<br />
1. Does .NET Gadgeteer yield an engaging and motivating<br />
environment to work with in schools?<br />
2. Are the initial teaching materials sufficient for students<br />
in lower and upper secondary schools to build .NET<br />
Gadgeteer devices?<br />
3. Could .NET Gadgeteer be used to support student<br />
learning in Computer Science in school?<br />
4. Is .NET Gadgeteer most suitable in schools as an extracurricular<br />
activity or could it have a place in the main<br />
curriculum (in England and Wales)?<br />
4.2 Participants<br />
In this study, .NET Gadgeteer was used for the first time<br />
with secondary school students. Eight local schools in Cambridgeshire<br />
(UK) volunteered to be involved. The schools<br />
were a mixture <strong>of</strong> age 11-16 and age 11-18 schools, state<br />
schools (seven) and private schools (one), mixed schools<br />
(six), girls’ schools (one) and boys’ schools (one). The teachers<br />
were initially trained in the use <strong>of</strong> .NET Gadgeteer and<br />
provided with lesson plans and example projects. Munson<br />
et al. [20] point out that in or<strong>der</strong> to engage students, it is<br />
also necessary to support teachers. We devised this pilot<br />
project with an initial learning session for teachers at Micros<strong>of</strong>t<br />
Research Cambridge, for them to cascade to pupils.<br />
Teachers introduced .NET Gadgeteer to their schools in the<br />
form <strong>of</strong> after-school or lunchtime clubs. The ages <strong>of</strong> the students<br />
attending the club ranged from 11 to 15 year, with one<br />
school choosing to use .NET Gadgeteer with an ol<strong>der</strong> group<br />
<strong>of</strong> 17 year olds. The students worked in groups <strong>of</strong> three to<br />
four with one .NET Gadgeteer kit per group. There was<br />
no requirement, given the existing ICT curriculum, for the<br />
schools to have taught programming before and it was acknowledged<br />
that it would not be possible to teach significant<br />
amounts <strong>of</strong> C# programming in the course <strong>of</strong> the pilot.<br />
The materials developed consisted <strong>of</strong> eight lesson plans<br />
with approximately one hour’s teaching material in each.<br />
The materials included learning objectives, aims <strong>of</strong> the session,<br />
a starter session as well as the main session in several<br />
steps, and extension material. The lesson plan contained all<br />
the instructions for carrying out the tasks. The session plans<br />
had aims and outcomes, in terms <strong>of</strong> programming skills to be<br />
learned, but the approach taken was actually to get devices<br />
working by following through instructions.<br />
The students were loaned .NET Gadgeteer kits for four<br />
month. Teachers planned to run either one after-school club<br />
or one lunchtime club each school week. The plan was for<br />
them to run around ten hours <strong>of</strong> activity.<br />
4.3 Methodology<br />
The students’ progress with .NET Gadgeteer was monitored<br />
by observation visits and their experiences evaluated<br />
at the end <strong>of</strong> the project. A case study methodology was<br />
used [24], taking the form <strong>of</strong> a collective case study,<br />
whereby a variety <strong>of</strong> sources were used to give an overall<br />
picture <strong>of</strong> how .NET Gadgeteer could be used in schools.<br />
An online questionnaire was designed to give teachers the<br />
99<br />
Figure 5: Extract from teachers’ questionnaire<br />
opportunity to give feedback in their own time. This was<br />
available via a web link and had a mixture <strong>of</strong> open and closed<br />
questions. Teachers were also interviewed informally about<br />
their experiences <strong>of</strong> the project. An extract <strong>of</strong> the teachers’<br />
questionnaire is shown in Figure 5.<br />
To collect verbal responses from students about their experiences,<br />
short interviews were held with students who attended<br />
at an event associated with the pilot. This represented<br />
a sample <strong>of</strong> 16 out <strong>of</strong> 84 students who took part in<br />
the pilot. The students volunteered to be interviewed, and<br />
had the option <strong>of</strong> speaking to the researcher individually or<br />
in pairs. This made it less stressful for the participants, although<br />
the sampling was therefore not <strong>of</strong> a representative<br />
group <strong>of</strong> the participating students. The interviewer asked<br />
each <strong>of</strong> the students the following four questions:<br />
1. What sort <strong>of</strong> things have you worked on in your .NET<br />
Gadgeteer sessions?<br />
2. What has been your favourite part <strong>of</strong> working with<br />
.NET Gadgeteer and why?<br />
3. Would you continue to use .NET Gadgeteer if you had<br />
a kit at school or home?<br />
4. Has working with .NET Gadgeteer increased your un<strong>der</strong>standing<br />
<strong>of</strong> how computers and electronic devices<br />
work?<br />
The researcher filmed a demonstration <strong>of</strong> their gadget if<br />
they had made one. These interviews were transcribed and<br />
coded and key themes drawn out <strong>of</strong> the resultant data. The<br />
questions were designed to generate a range <strong>of</strong> data about<br />
engagement with the project.<br />
The short interviews were followed by an in-depth focus<br />
group at one <strong>of</strong> the pilot schools. Two girls and two boys<br />
were invited to participate in the focus group which was<br />
conducted by the first author with a teacher present at all<br />
times. A focus group has many advantages when seeking<br />
to find out attitudes as the one-to-many dialogue, perhaps<br />
including a difference <strong>of</strong> opinion, may allow other issues to<br />
be raised which may not arise in an interview where there
Figure 6: Pedagogical model<br />
is no opposing view. The focus group was planned such<br />
that a teacher was present and the students all knew each<br />
other. It included more in-depth questions about the students’<br />
engagement with technology and computer programming<br />
more generally, with the intention <strong>of</strong> providing an elaboration<br />
<strong>of</strong> the points made in the short interviews and in the<br />
teacher questionnaires. The questions were based around<br />
the following themes: experiences <strong>of</strong> the .NET Gadgeteer<br />
project, learning new technologies, and aptitudes and difficulties<br />
with learning to program. The focus group discussion<br />
was carried out by the first author, and recorded, transcribed<br />
and then coded using TAMS analyser 8 .<br />
4.4 Pedagogical approach<br />
The pedagogical approach taken with the teaching materials<br />
was based on a two-phase model whereby the first<br />
phase involved instruction and the second phase involved<br />
presenting students with a challenge. The students needed<br />
some instruction on how to use .NET Gadgeteer and the<br />
instructions led them through the implementation <strong>of</strong> three<br />
devices. Teachers utilised the materials differently, some setting<br />
student tasks based on the materials, and others letting<br />
the students work through the materials at their own pace.<br />
Each teacher had their own different style <strong>of</strong> teaching and<br />
this was noted in observations made by the first author at<br />
visits to each <strong>of</strong> the schools. Despite the teacher-differences,<br />
there was a common approach to teaching which led into the<br />
challenge phase <strong>of</strong> the project. Students were encouraged to<br />
create their own device for the end <strong>of</strong> the project. In or<strong>der</strong><br />
to investigate how students were able to learn with the ‘tryit-and-see’<br />
bricolage approach, students were encouraged to<br />
be as inventive as they could in coming up with their own<br />
ideas for devices. Figure 6 shows the rationale behind the<br />
simple two-phase pedagogical model adopted.<br />
5. RESULTS<br />
Table 1 shows the schools, numbers <strong>of</strong> students attending<br />
sessions, and the number <strong>of</strong> session plans they managed to<br />
complete, although some schools ran more sessions and some<br />
students used some <strong>of</strong> their own time to complete projects.<br />
After completing the session plans, students then worked<br />
on their own projects in Phase 2 <strong>of</strong> the pilot, applying what<br />
they had learned to their own ideas. Students worked in<br />
teams <strong>of</strong> between two and five students and developed a<br />
range <strong>of</strong> small devices. Students also enjoyed building the<br />
housing for the gadgets, which in some cases was quite sophisticated<br />
using moulded plastic, and for other gadgets,<br />
just as effective, using polystyrene or cardboard. Examples<br />
8 http://tamsys.sourceforge.net/<br />
100<br />
Figure 7: Examples <strong>of</strong> devices created by students<br />
with .NET Gadgeteer: “Robber Gadget” takes a picture<br />
<strong>of</strong> a burglar when a precious item is taken <strong>of</strong>f<br />
the sensor platform; “Rainbow Press” is a reaction<br />
game; “Alien Invasion” is a shooting game; “Gadgea-sketch”<br />
is a drawing device.<br />
<strong>of</strong> the gadgets developed by students are shown in Figure 7.<br />
Table 1: Participants in the project<br />
School Number Number Sessions Lowest Highest<br />
sessions students (out <strong>of</strong> 8) year year<br />
group group<br />
A 10 8 5 Y9 Y10<br />
B 4 6 4 Y7 Y13<br />
C 4 24 3 Y7 Y8<br />
D 5 6 6 Y11 Y13<br />
E >10 10 6 Y8 Y11<br />
F >10 3 7 Y9 Y10<br />
G >10 10 6 Y7 Y11<br />
H 7 18 4 Y9 Y9<br />
(Y7 = age 11/12 (equivalent to 6 in USA);<br />
Y13 = age 17/18 (equivalent to 12 in USA)<br />
Several schools ran more than ten sessions although engagement<br />
between the eight schools varied. The number <strong>of</strong><br />
students attending at each school varied from three to 24<br />
students. None <strong>of</strong> the schools completed all <strong>of</strong> the eight session<br />
plans provided, indicating that the material took longer<br />
to cover and that the difficulty level may have been un<strong>der</strong>estimated.<br />
We originally aimed the project at Y9 students<br />
(aged 13-14) but gave teachers the freedom to select appropriate<br />
students as school environments differ. This had the<br />
benefit <strong>of</strong> giving us some examples <strong>of</strong> projects from students<br />
from Y7 to Y13, demonstrating the wide potential <strong>of</strong> .NET<br />
Gadgeteer.<br />
Teachers were asked to rate how they thought that the students<br />
had mastered key programming concepts. It was not<br />
expected that novice programmers would be able to achieve<br />
a thorough un<strong>der</strong>standing <strong>of</strong> Visual C# programming in just<br />
10 hours <strong>of</strong> sessions, and we had no prior hypothesis about<br />
how much students in such a mixed age group at different
Figure 8: Acquisition <strong>of</strong> programming concepts (average<br />
across 8 schools)<br />
schools with different teaching styles would learn. However,<br />
it was encouraging to see that most schools felt that their<br />
students had an un<strong>der</strong>standing <strong>of</strong> assignment, data types,<br />
variables and selection as a result <strong>of</strong> the pilot. Figure 8<br />
shows the relative acquisition <strong>of</strong> programming concepts as<br />
rated by their teachers.<br />
Teachers were broadly positive about the achievements<br />
<strong>of</strong> their pupils and the motivation provided by .NET Gadgeteer.<br />
One teacher commented that “It was fun and really<br />
nice to have things to touch and build. Some students were<br />
very engaged in the whole thing” (School H, teacher).<br />
There were a few technical problems experienced by teachers<br />
as the platform had never been tested before on school<br />
networks. These are discussed in Section 5.4.<br />
5.1 Interviews with students<br />
Approximately 85 students used .NET Gadgeteer in school<br />
as part <strong>of</strong> this pilot and 50 attended the end-<strong>of</strong>-school pilot<br />
event. Of these, 16 were interviewed during the day (14<br />
boys, two girls). The comments from the student interviews<br />
and the focus group clearly identified four emerging themes.<br />
These were around the following features <strong>of</strong> the programme:<br />
• Challenge and difficulty<br />
• Creativity and freedom<br />
• The tactile nature <strong>of</strong> .NET Gadgeteer<br />
• The concept <strong>of</strong> “real programming”<br />
These will be discussed in turn.<br />
5.1.1 Challenge and difficulty<br />
Without being prompted, students commented on the fact<br />
that they had found working with .NET Gadgeteer quite<br />
challenging at times. For example, one student said “I enjoyed<br />
it but it was hard. And it was a challenge”(School C,<br />
male), and another found it “. . . a big learning curve really”<br />
(School G, male). One student tried to explain why some<br />
other students from his school found the programming difficult:<br />
“. . . but it took a great deal <strong>of</strong> effort, and not everyone’s<br />
ready to . . . you know, actually step up to the challenge”<br />
(School B, male). However, the level <strong>of</strong> challenge seemed<br />
to be popular with several students: “. . . on Gadgeteer it’s<br />
101<br />
so much better because it’s har<strong>der</strong>, but that’s the good <strong>of</strong> it<br />
you know” (School B, male). One young student with some<br />
programming experience commented that it was “. . . not too<br />
simple, so that you’re not really not learning that much, but<br />
it’s not too complex, that you‘re not getting any help with<br />
debugging” (School G, male).<br />
There were fewer comments to the extent that it was not<br />
too difficult, and these were from two students who had had<br />
extensive programming experience from working on projects<br />
at home and self-teaching. There was recognition from several<br />
students that they had learned a lot from their experience:<br />
“. . . we never learnt code before at all, so it most <strong>of</strong><br />
the things, well everything we learnt for Gadgeteer was stuff<br />
that was new” (School A, male).<br />
5.1.2 Creativity and freedom<br />
The pilot was designed with a second phase whereby students<br />
were to design a gadget or device <strong>of</strong> their choice, with<br />
freedom limited only by the modules that were available in<br />
the particular .NET Gadgeteer kit they were working with.<br />
Students referred to the fact that they liked the freedom<br />
to be creative and that they liked the tactile nature <strong>of</strong> the<br />
modules: “You’re in control you can take an idea anywhere<br />
and use the hardware that’s available to make it and without<br />
limited . . . so the key it’s versatile” (School A, male). Another<br />
student particularly liked the camera project, which<br />
could be extended in many ways: “What I enjoyed most<br />
about .NET Gadgeteer is the creativity you can have and the<br />
challenge it poses . . . especially with the camera, I really enjoyed<br />
that. Also trying to build your own sort <strong>of</strong> gadgets, and<br />
that was, you could really use your imagination” (School G,<br />
male). Another student commented on the freedom given to<br />
use .NET Gadgeteer as they wanted: “You’re allowed to be<br />
sort <strong>of</strong> creative and sort <strong>of</strong> like make anything so you weren’t<br />
really limited to what you can make” (School A, male). An<br />
ol<strong>der</strong> student, with experience <strong>of</strong> programming and the Arduino<br />
platform, was very impressed by the flexibility it gave<br />
him: “. . . you just plug it together and it works, you don’t<br />
have to figure out the circuits for yourself . . . , but it’s brilliant,<br />
really fun” (School D, male).<br />
5.1.3 The tactile nature <strong>of</strong> .NET Gadgeteer<br />
Six students made comments to the extent that the tangible<br />
nature <strong>of</strong> the modules made it exciting to build and<br />
program devices. For example, one <strong>of</strong> the boys commented<br />
that “. . . you don’t just like have it all as pictures on the<br />
screen, you actually have the stuff that you can put together<br />
. . . ” (School C, male). During the event, the students were<br />
proud to be able to demonstrate the devices that they had<br />
made to other students, and their achievement was more<br />
visible in its physical form. Another student commented<br />
that developing a physical gadget meant that they could<br />
build something that had a purpose: “You actually have<br />
something you could hold and manipulate, that’s, it’s sort<br />
<strong>of</strong> feeling a practical use for your programs, instead <strong>of</strong> just<br />
doing it for the sake <strong>of</strong> it” (School G, male). Having both<br />
aspects <strong>of</strong> development, with the hardware and the s<strong>of</strong>tware,<br />
also appealed to another student: “You need to program it,<br />
you need to put it together, and it just makes it a whole lot<br />
better” (School B, male).<br />
5.1.4 Doing “real programming”<br />
Students made reference to the fact that they knew that
Visual C# was a programming language used in industry by<br />
pr<strong>of</strong>essionals, and that they were using a tool that felt very<br />
‘adult’. For example, one student commented that “This<br />
was much better because it was more pr<strong>of</strong>essional and, you<br />
could do a lot more with it” (School G, male). The comment<br />
made was in comparison to Scratch, which is the environment<br />
with which many <strong>of</strong> the students were already familiar.<br />
Another student compared .NET Gadgeteer to his experience<br />
with Scratch also: “.NET Gadgeteer . . . sort <strong>of</strong> like<br />
got proper programming if you know what I mean, . . . with<br />
Scratch you’ve got words already set out for you and you can<br />
sort <strong>of</strong> like make building blocks . . . but with .NET Gadgeteer<br />
you’ve definitely got to have that more adult aspect <strong>of</strong> it and<br />
the programming . . . ” (School G, male).<br />
The pilot had encouraged students to consi<strong>der</strong> doing more<br />
programming in the future. Some students reflected that<br />
their views on Computer Science as a subject had changed<br />
as a result <strong>of</strong> the engagement with .NET Gadgeteer: “ . . . I<br />
definitely want to take Computing and I’m looking to take<br />
a job inside Computing as well when I’m ol<strong>der</strong>” (School G,<br />
male). A young student mused about the possibilities that<br />
might open for him if he pursued a career in Computer Science:<br />
“It would be a nice career to take on . . . and so I’d<br />
really like to be working at a place like Silicon Valley, it<br />
would be a really nice opportunity” (School G, male).<br />
Overall, the interviews elicited mostly positive comments<br />
about .NET Gadgeteer. The results <strong>of</strong> the focus group discussion<br />
will be reported next.<br />
5.2 Focus group findings<br />
This group <strong>of</strong> students included three from Y10 and one<br />
from Y9 (this translates to the beginning upper secondary<br />
school in Europe and the beginning <strong>of</strong> high school in the<br />
USA). The focus group was held in School A with two boys<br />
and two girls who had participated in the project. They were<br />
asked similar questions to the short interviews initially and<br />
similar themes emerged, for example, relating to freedom<br />
and creativity: “I liked designing for the competition, the<br />
fact that we got to design our own product”(School A, male).<br />
The students also reported that they liked the tactile nature<br />
<strong>of</strong> the kits, and that the first project was quite accessible:<br />
“ . . . like the ease <strong>of</strong> it, it was quite easy to integrate the bits,<br />
for instance making the camera was quite simple” (School<br />
A, male). Other comments were made that mirrored the<br />
content <strong>of</strong> the short interviews already reported on. They<br />
commented that .NET Gadgeteer made the concept <strong>of</strong> the<br />
technologies they use more accessible “because essentially we<br />
had many <strong>of</strong> the bits <strong>of</strong> an iPod Touch, a camera and a touch<br />
screen, and you can get the audio jack and stuff, to give you<br />
an idea that you could perhaps make something like that and<br />
that these technologies weren’t far away” (School A, male).<br />
Students also discussed some <strong>of</strong> the key aspects that they<br />
felt appealed about Computer Science to young people, what<br />
helped them to learn, what qualities were useful in learning<br />
programming and why other students might not succeed.<br />
The first quality to be suggested was to be open-minded:<br />
“Open-minded . . . because it’s like some people if they don’t<br />
want to learn something then they most probably won’t listen<br />
to it” (School A, female). The student went on to explain<br />
that the stereotype <strong>of</strong> being “weird and nerds” could put<br />
people <strong>of</strong>f doing Computer Science if they were not openminded.<br />
The students agreed between them a set <strong>of</strong> qualities that<br />
102<br />
might make students good at programming, or working with<br />
.NET Gadgeteer, as follows:<br />
• Open-mindedness<br />
• Common sense<br />
• Problem-solving skills<br />
• Patience<br />
• Imagination<br />
• Intuition<br />
Common sense was defined by the student as “just being<br />
able to look at something, and then work out, kind <strong>of</strong> maybe<br />
use your brain to go through what the computer is doing, the<br />
way the computer is looking at it. There’s a lot <strong>of</strong> people<br />
can’t do that. I think they find it a lot more difficult to<br />
program” (School A, male). From the student’s point <strong>of</strong><br />
view (this student had taught himself some programming<br />
already) this was common sense, whereas another might see<br />
this as logical thinking. This was an interesting example <strong>of</strong> a<br />
student with experience perhaps un<strong>der</strong>estimating what was<br />
difficult to a non-programmer [25]. The two girls were more<br />
able to empathise with the issues that other students might<br />
have in learning to program.<br />
The notion <strong>of</strong> patience being a key to success in programming<br />
was mentioned at other times in the focus group and<br />
in one <strong>of</strong> the short interviews: “Problem-solving skills and<br />
patience, because when you’re learning it, you have to have<br />
a lot <strong>of</strong> patience in you because it might take some time to<br />
learn everything, you can’t exactly rush through it otherwise<br />
you won’t learn it properly” (School A, male).<br />
Students also reported on the sources that they used to<br />
assist them when they had problems and needed help with<br />
their work in computing programming. They demonstrated<br />
that they knew where they could look for help on the Internet:<br />
“Looking at other people’s code” (School A, male), and<br />
also: “I read tutorials and find the answer on the Internet”<br />
(School A, male). The two girls were more reliant on teachers<br />
and friends to help them with their coding: “Mainly the<br />
teachers who help me try and learn and help me go through it<br />
so it sticks in my head a bit more” (School A, female) and:<br />
“Just having people explain it to you and if you go wrong<br />
having people explaining exactly where you went wrong and<br />
how you’re supposed to do it right, and things like that would<br />
be quite helpful. Having someone that can really explain it<br />
to you” (School A, female).<br />
This illustrates some <strong>of</strong> the gen<strong>der</strong> differences, as discussed<br />
briefly below.<br />
5.3 Girls and .NET Gadgeteer<br />
In this study, we had not particularly focused on comparing<br />
the responses from different gen<strong>der</strong> groups in our design,<br />
and we had a much smaller number <strong>of</strong> girls in the project<br />
than boys. However, the comments from girls (two in the<br />
short interviews and two more in the focus group) suggested<br />
some surprise that they had managed to be successful with<br />
.NET Gadgeteer. Girls reported that prior to using .NET<br />
Gadgteer “I really thought I wouldn’t be able to do anything”<br />
(School G, female). Another girl admitted that it hadn’t really<br />
appealed to her: “. . . just thought it was all a bit, sort <strong>of</strong><br />
hard, just like why would I need to know” (School E, female).
The focus group girls were more confident: “Well, maybe a<br />
little bit overwhelming at first but once you un<strong>der</strong>stand what<br />
it all means it’s not that difficult once you know what you’re<br />
doing” (School A, female).<br />
The girls reported more confidence after they had finished<br />
creating their gadgets with .NET Gadgeteer: “Yeah, it’s definitely<br />
opened up more about what I know about programming”<br />
(School E, female). Another <strong>of</strong> the girls said that she<br />
was more confident since building her (highly innovative)<br />
gadget: “I’ve really surprised myself, and I’d have to say,<br />
I’ve grown in confidence as well” (School G, female).<br />
One <strong>of</strong> the teachers (School C), in an interview, commented<br />
on what girls enjoyed about the programme: “The<br />
girls, I think they were just more interested in getting the<br />
camera working, and they all wanted to be able to take pictures<br />
<strong>of</strong> themselves, which kind <strong>of</strong> gives them that sense <strong>of</strong><br />
ownership”.<br />
With such a small number <strong>of</strong> girls, we can only note these<br />
gen<strong>der</strong> differences; intuitively, though, it seems as if working<br />
with this platform could appeal widely to both gen<strong>der</strong>s, as<br />
gadgets with a socially useful function could be developed<br />
over time that are more appealing to girls. The next section<br />
addresses some <strong>of</strong> the limitations and problems we found<br />
with .NET Gadgeteer on this first use in UK schools.<br />
5.4 Problems with .NET Gadgeteer<br />
In terms <strong>of</strong> their dissatisfaction with the project the main<br />
complaint from students was about lack <strong>of</strong> time to complete<br />
their projects: “ . . . we had after-school clubs every Wednesday,<br />
but towards the end we ended up doing it at lunchtimes<br />
as well” (School G, male). Another student reported spending<br />
much time on his .NET Gadgeteer trying to complete a<br />
project: “ . . . mainly just in lunchtimes, then Thursday and<br />
Wednesday I’ve been going after school, and every opportunity<br />
possible really . . . I’ve been trying to get things finished”<br />
(School E, male).<br />
Teachers also would have liked more time to complete<br />
projects, indicating that working with .NET Gadgeteer can<br />
encourage longer-lasting project work than we had originally<br />
envisaged: “I would have loved to take it further if we had<br />
had more time, and to see the possibilities and how far our<br />
students would be able to take it. What could our students<br />
create if they had another term? How advanced could our<br />
students take it? What new systems, concepts could they interpret<br />
and un<strong>der</strong>stand to produce bigger and better systems?<br />
Just more time for students to play and to see the real potential<br />
<strong>of</strong> the gadgets they create” (School A, teacher).<br />
An issue for teachers was some <strong>of</strong> the technical difficulties<br />
they had had at the outset when the pilot started. These<br />
schools were some <strong>of</strong> the first public users ever <strong>of</strong> .NET<br />
Gadgeteer, so some locked-down school networks posed a<br />
challenge initially. This links to the fact that the students<br />
felt that they did not have enough time to build what they<br />
wanted to. For our research project, we had set these time<br />
constraints. For future work with .NET Gadgeteer it is likely<br />
that a full academic year would be needed to really benefit<br />
and learn, as with any new skill. .NET Gadgeteer was<br />
very new when the project started, as the teachers received<br />
some <strong>of</strong> the first kits that were shipped over to the UK. The<br />
students were using the .NET Gadgeteer environment before<br />
the schools had had time to test it properly on their systems:<br />
“I still have not been able to load the required s<strong>of</strong>tware onto<br />
the school system, and have had to work using a teacher<br />
103<br />
laptop that I had to battle 8 weeks for to get admin rights<br />
to load the programs onto it” (School C, teacher). These<br />
teething problems have now been eradicated as indicated<br />
by this teacher: “ . . . there were some minor glitches with a<br />
mismatch <strong>of</strong> s<strong>of</strong>tware and hardware at the start” (School G,<br />
teacher).<br />
The teachers also wanted more materials, as identified<br />
here: “Some form <strong>of</strong> ‘student book’ to allow them to track<br />
and take notes might help” (School H, teacher). Another<br />
teacher wanted more model projects to inspire students: “I<br />
would have quite liked a very advanced gadget to have been<br />
created and code provided, not for students to see code, but so<br />
we could easily show some <strong>of</strong> the high potential that could be<br />
achieved using Gadgeteer systems” (School A, teacher). Finally,<br />
one teacher felt that their Y7 students were too young<br />
and that the ol<strong>der</strong> students got more out <strong>of</strong> the experience:<br />
“Great ideas and all the students were enthusiastic despite<br />
<strong>of</strong>ten making very little progress. Our mistake - we opened<br />
it to all age groups and while the younger students were very<br />
keen, they lacked the concentration and basic programming<br />
experience to get as much from it as the ol<strong>der</strong> students did”<br />
(School G, teacher).<br />
6. DISCUSSION<br />
The results <strong>of</strong> this study <strong>of</strong> .NET Gadgeteer in secondary<br />
schools demonstrate that students were engaged by .NET<br />
Gadgeteer, although they did find it difficult. This is backed<br />
up by their teachers’ comments. Challenge is seen as a positive<br />
feature as reported by the students in the short interviews;<br />
many students referred to this although it was not<br />
specifically elicited. There were some very able students<br />
within the pilot groups for whom the opportunity to use<br />
.NET Gadgeteer provided them with a stimulating resource<br />
that they very much enjoyed. However, these students had<br />
volunteered to be part <strong>of</strong> the pilot groups, and more investigation<br />
is needed to see how less motivated students experience<br />
the platform.<br />
Students automatically seemed to compare .NET Gadgeteer<br />
with the Scratch programming environment, although<br />
again this was not mentioned by the interviewer, presumably<br />
because that is the experience that students had had so far<br />
and as such the only experience with which they were able<br />
to compare .NET Gadgeteer. Their comments implied that,<br />
after using Scratch, they felt that .NET Gadgeteer allowed<br />
them to write their own functions and have more freedom to<br />
customise their own programs. What .NET Gadgeteer <strong>of</strong>fers<br />
after having learned Scratch is an introduction to a real<br />
programming language and this was popular with students.<br />
The introduction to programming provided by Scratch, if<br />
incorporated in a teaching programme at school, will help<br />
them prior to moving forward with .NET Gadgeteer.<br />
Although a challenge, students seem to have enjoyed working<br />
together in groups and creating something <strong>of</strong> their own.<br />
The production <strong>of</strong> a physical device to demonstrate was<br />
a motivating factor. Observations <strong>of</strong> students working revealed<br />
that they worked well in groups in the main, with<br />
teachers being flexible about sizes and make-up <strong>of</strong> groups.<br />
Within groups, students assigned themselves different roles,<br />
with some having more responsibility for the design <strong>of</strong> the<br />
case or user-interface, with others writing more <strong>of</strong> the code.<br />
This is one <strong>of</strong> the outcomes <strong>of</strong> the project that we had not<br />
expected. To develop their own gadget, students were necessarily<br />
involved in both problem-solving and some inde-
pendent research. They had to experiment with different<br />
modules to find out exactly how they worked. The wi<strong>der</strong><br />
skills they gain through engagement in these activities have<br />
a positive effect on developing their personal, learning and<br />
thinking skills.<br />
The pedagogical model that we used encouraged a bricolage<br />
approach. We wanted to provide some instruction, but<br />
then to inspire children to experiment with modules and devices<br />
on their own and come up with their own devices. For<br />
some <strong>of</strong> the children, though, the gap between our Phase 1<br />
and Phase 2 was too great and the teachers did not obviously<br />
have the background or sufficient training in the platform to<br />
support them. We need to build on the incremental problem<br />
solving as used successfully by Stiller [28] and provide<br />
smaller gaps in future materials, to maintain both teacher<br />
and student confidence.<br />
We have not specifically tested the students’ un<strong>der</strong>standing<br />
<strong>of</strong> programming principles in this project. Meerbaum-<br />
Salant carried out a study whereby she measured the un<strong>der</strong>standing<br />
<strong>of</strong> key programming concepts <strong>of</strong> students who<br />
had participated in a course using the Scratch programming<br />
environment [19]. She reports that “the bottom-up programming<br />
habit is clearly encouraged by the characteristics <strong>of</strong> the<br />
Scratch environment and is in line with Papert’s philosophy<br />
<strong>of</strong> constructionism . . . and with bricolage” [p.171] and<br />
concludes that “. . . we are disturbed by the habits <strong>of</strong> programming<br />
that we uncovered. These habits are not at all what one<br />
expects as the outcome <strong>of</strong> learning computer science” [p.172].<br />
It is not clear whether she is blaming the Scratch environment<br />
in being too “easy” or the constructionist and bricolage<br />
approach in her concerns about long-term effects <strong>of</strong> an exploratory<br />
approach to programming. We would hope to be<br />
able to show that an exploratory, and bricolage-centred, pedagogical<br />
approach can be both motivating and teach good<br />
programming habits, in the context <strong>of</strong> a real programming<br />
language that works behind a motivating hardware environment.<br />
As the platform is still new and materials are still un<strong>der</strong><br />
development, we have not yet been able to replicate this<br />
level <strong>of</strong> analysis <strong>of</strong> learning for .NET Gadgeteer. However,<br />
although a bricolage approach seems implicit in ‘gadgetbuilding’,<br />
it may not be as bottom-up as the Scratch programming<br />
that Meerbaum-Salant describes. Where students<br />
assemble their gadget before programming, they have put<br />
together all the modules they need for the project and then<br />
can break down the programming into events. This provides<br />
more <strong>of</strong> a top-down, but also very concrete experience<br />
<strong>of</strong> program development. More research is obviously<br />
needed to discover if this is effective in the long-term acquisition<br />
<strong>of</strong> key concepts. This will involve the development<br />
<strong>of</strong> a suitable research instrument to measure learning with<br />
.NET Gadgeteer.<br />
Figure 9 adds the development <strong>of</strong> the wi<strong>der</strong> personal,<br />
learning and thinking skills to that shown in Figure 4. This<br />
illustrates some <strong>of</strong> the potential wi<strong>der</strong> skills that can be<br />
gained by students working on .NET Gadgeteer in groups<br />
on their own projects.<br />
7. CONCLUSIONS AND FURTHER WORK<br />
In this paper, we have presented our experiences <strong>of</strong> using<br />
.NET Gadgeteer in schools with students <strong>of</strong> various ages<br />
and backgrounds over a four-month period. Our findings<br />
have revealed that students are very engaged by it, and are<br />
104<br />
Figure 9: Wi<strong>der</strong> skills development with .NET Gadgeteer<br />
inspired to build devices for which they can see a real purpose.<br />
Through analysis <strong>of</strong> data from students and teachers,<br />
we have uncovered some key features <strong>of</strong> .NET Gadgeteer<br />
that provide motivation and interest: challenge, freedom to<br />
explore, tangibility and exposure to the real world <strong>of</strong> Computing.<br />
We are partway towards answering our research questions:<br />
1. Is .NET Gadgeteer an engaging and motivating environment<br />
to work with in schools? We have found that<br />
students reported positively about using .NET Gadgeteer<br />
for a range <strong>of</strong> reasons, as described in this paper.<br />
2. Can students in lower and upper secondary schools use<br />
.NET Gadgeteer to build devices with the initial teaching<br />
materials developed? Students have been successful<br />
in building devices; however, the teachers indicated<br />
that the materials need to be developed further.<br />
3. Could .NET Gadgeteer be used to support student learning<br />
in Computer Science in school? With the changes<br />
in the curriculum in the UK, involving the introduction<br />
<strong>of</strong> more Computer Science in secondary schools, .NET<br />
Gadgeteer provides an environment that is both engaging<br />
and encourages experiential learning. Working<br />
with .NET Gadgeteer has the potential to develop students’<br />
ability to work collaboratively. Programming<br />
skills can be developed through a mixture <strong>of</strong> instruction<br />
and an exploratory approach to learning.<br />
4. Is .NET Gadgeteer most suitable in schools as an extracurricular<br />
activity or could it have a place in the curriculum<br />
(in England and Wales)? The research revealed<br />
that working through the materials and creating<br />
the devices was quite intensive and required more time<br />
than available in an after-school club. Further work is<br />
needed to write longer courses that would work with<br />
curricular in UK schools and within examined courses.<br />
The .NET Gadgeteer platform is still in its infancy as it<br />
was only launched in August 2011. At the time <strong>of</strong> writing,<br />
more modules are being developed by an increasing number<br />
<strong>of</strong> manufacturers.<br />
The pilot project was successful in its aims and objectives<br />
which were to establish that .NET Gadgeteer can be a useful
tool for teachers and students in the classroom. Its tangible<br />
nature engen<strong>der</strong>s curiosity and creativity, and motivates<br />
students to explore, bricolage-fashion, how to program their<br />
device. The physical nature <strong>of</strong> the platform encourages a<br />
learning-by-doing experiential approach to learning and in<br />
addition lends itself to collaborative working, whereby students<br />
with a range <strong>of</strong> skills and abilities can support and<br />
learn from each other.<br />
As an initial pilot project with a new platform, this research<br />
has highlighted particular areas <strong>of</strong> further work. In<br />
future studies using .NET Gadgeteer, we would like to investigate<br />
how students’ perception <strong>of</strong> the importance <strong>of</strong> challenge<br />
and creativity links to the development <strong>of</strong> secure and<br />
robust programming un<strong>der</strong>standing. It is thus proposed that<br />
further studies with .NET Gadgeteer focus on the facilitation<br />
<strong>of</strong> the acquisition <strong>of</strong> certain programming concepts. The<br />
two-phase pedagogical model utilised in the pilot project will<br />
be developed further to provide more staged support, whilst<br />
retaining an exploratory and experiential approach.<br />
8. ACKNOWLEDGMENTS<br />
With thanks to Steve Hodges and Steven Johnston, Micros<strong>of</strong>t<br />
Research, and to all the participating teachers and<br />
students for their assistance and enthusiasm during the pilot<br />
project. The first author would like to thank Micros<strong>of</strong>t<br />
Research for supporting her work.<br />
9. REFERENCES<br />
[1] O. Astrachan, J. Cuny, C. Stephenson, and C. Wilson.<br />
The CS10K Project: Mobilizing the Community to<br />
Transform High School Computing. In Proceedings <strong>of</strong><br />
the 42nd ACM Technical Symposium on Computer<br />
Science Education, SIGCSE ’11, pages 85–86, New<br />
York, NY, USA, 2011. ACM.<br />
[2] M. Ben-Ari. Constructivism in Computer Science<br />
Education. In Proceedings <strong>of</strong> the 29th SIGCSE<br />
Technical Symposium on Computer Science Education,<br />
pages 257–261. ACM, 1998.<br />
[3] T. Brinda, H. Puhlmann, and C. Schulte. Bridging<br />
ICT and CS - Educational Standards for Computer<br />
Science in Lower Secondary Education. Proceedings <strong>of</strong><br />
ITICSE’09, Paris, France, July 6-9 2009.<br />
[4] L. Carter. Why Students with an Apparent Aptitude<br />
for Computer Science Don’t Choose to Major in<br />
Computer Science. In Proceedings <strong>of</strong> the 37th SIGCSE<br />
Technical Symposium on Computer Science Education,<br />
pages 27–31. ACM, 2006.<br />
[5] S. Cooper, W. Dann, and R. Pausch. Alice: a 3-D<br />
Tool for Introductory Programming Concepts.<br />
J.Comput.Sci.Coll., 15(5):107–116, Apr. 2000.<br />
[6] T. Crick and S. Sentance. Computing at School:<br />
Stimulating Computing Education in the UK. In<br />
Proceedings <strong>of</strong> the 11th Koli Calling International<br />
Conference on Computing Education Research, Koli<br />
Calling ’11, pages 122–123, New York, USA, 2011.<br />
ACM.<br />
[7] Q. Cutts, S. Esper, and B. Simon. Computing as the<br />
4th R: a General Education Approach to Computing<br />
Education. In Proceedings <strong>of</strong> the 7th International<br />
Workshop on Computing Education Research, ICER<br />
’11, pages 133–138, New York, NY, USA, 2011. ACM.<br />
105<br />
[8] T. Downes and D. Looker. Factors that Influence<br />
Students’ Plans to Take Computing and Information<br />
Technology Subjects in Senior Secondary School.<br />
Computer Science Education, 21(2):175–199, 2011.<br />
[9] A. Fisher and J. Margolis. Unlocking the Clubhouse:<br />
the Carnegie Mellon Experience. SIGCSE Bulletin,<br />
34(2):79–83, June 2002.<br />
[10] O. Hazzan, J. Gal-Ezer, and L. Blum. A Model for<br />
High School Computer Science Education: the Four<br />
Key Elements that Make It! In Proceedings <strong>of</strong> the 39th<br />
SIGCSE Technical Symposium on Computer Science<br />
Education, pages 281–285. ACM, 2008.<br />
[11] F. T. H<strong>of</strong>stetter. Multimedia Literacy.<br />
Irwin/McGraw-Hill, Boston, 2nd edition, 1997.<br />
[12] M. S. Horn, R. J. Crouser, and M. U. Bers. Tangible<br />
Interaction and Learning: the Case for a Hybrid<br />
Approach. Personal Ubiquitous Computing,<br />
16(4):379–389, Apr. 2012.<br />
[13] M. Kordaki, M. Miatidis, and G. Kapsampelis. A<br />
Computer Environment for Beginners’ Learning <strong>of</strong><br />
Sorting Algorithms: Design and Pilot Evaluation.<br />
Computers and Education, 51:708–723, 2008.<br />
[14] C. Lévi-Strauss, J. Weightman, and D. Weightman.<br />
The Savage Mind. <strong>University</strong> <strong>of</strong> Chicago Press, 1966.<br />
[15] M. Kolling. The Greenfoot Programming<br />
Environment. Transactions on Computing Education,<br />
10(4):14:1–14:21, Nov. 2010.<br />
[16] M. MacLaurin. Kodu: End-User Programming and<br />
Design for Games. In Proceedings <strong>of</strong> the 4th<br />
International Conference on Foundations <strong>of</strong> Digital<br />
Games, FDG ’09, New York, USA, 2009. ACM.<br />
[17] J. Maloney, M. Resnick, N. Rusk, B. Silverman, and<br />
E. Eastmond. The Scratch Programming Language<br />
and Environment. Transactions on Computing<br />
Education, 10(4):16:1–16:15, Nov. 2010.<br />
[18] P. Marshall. Do Tangible Interfaces Enhance<br />
Learning? In Proceedings <strong>of</strong> the 1st International<br />
Conference on Tangible and Embedded Interaction,<br />
TEI ’07, pages 163–170, New York, NY, USA, 2007.<br />
ACM.<br />
[19] O. Meerbaum-Salant, M. Armoni, and M. Ben-Ari.<br />
Habits <strong>of</strong> Programming in Scratch. In Proceedings <strong>of</strong><br />
the 16th Annual Joint Conference on Innovation and<br />
Technology in Computer Science Education, ITiCSE<br />
’11, pages 168–172, New York, NY, USA, 2011. ACM.<br />
[20] A. Munson, B. Moskal, A. Harriger,<br />
T. Lauriski-Karriker, and D. Heersink. Computing at<br />
the High School Level: Changing What Teachers and<br />
Students Know and Believe. Computing Education,<br />
57(2):1836–1849, Sept. 2011.<br />
[21] S. Papert. Mindstorms: Children, Computers, And<br />
Powerful Ideas. Basic Books, 1993. 92053249.<br />
[22] L. Pollock and T. Harvey. Combining Multiple<br />
Pedagogies to Boost Learning and Enthusiasm. In<br />
Proceedings <strong>of</strong> the 16th Annual Joint Conference on<br />
Innovation and Technology in Computer Science<br />
Education, ITiCSE ’11, pages 258–262, New York,<br />
NY, USA, 2011. ACM.<br />
[23] M. Richards, M. Petre, and A. K. Bandara. Starting<br />
with Ubicomp: Using the Senseboard to Introduce<br />
Computing. In Proceedings <strong>of</strong> the 43rd ACM Technical<br />
Symposium on Computer Science Education, SIGCSE
’12, pages 583–588, New York, NY, USA, 2012. ACM.<br />
[24] R.Stake. The Art <strong>of</strong> Case Study Research. London,<br />
Sage, 1995.<br />
[25] C. Schulte and M. <strong>Knobelsdorf</strong>. Attitudes Towards<br />
Computer Science-Computing Experiences as a<br />
Starting Point and Barrier to Computer Science. In<br />
Proceedings <strong>of</strong> the 3rd International Workshop on<br />
Computing Education Research, ICER ’07, pages<br />
27–38, New York, NY, USA, 2007. ACM.<br />
[26] R. E. Stake. The Art <strong>of</strong> Case Study Research. Sage<br />
Publications, 1995. 95004979.<br />
[27] C. W. Starr, D. Bergman, and P. Zaubi. The<br />
Development and Implementation <strong>of</strong> a Context-Based<br />
Curricular Framework for Computer Science<br />
Education in High Schools. In Proceedings <strong>of</strong> the 14th<br />
Annual ACM SIGCSE Conference on Innovation and<br />
Technology in Computer Science Education, pages<br />
283–287. ACM, 2009.<br />
[28] E. Stiller. Teaching Programming Using Bricolage.<br />
J.Comput.Sci.Coll., 24(6):35–42, June 2009.<br />
[29] The Royal Society. Shut Down or Restart? The Way<br />
Forward for Computing in UK Schools. Technical<br />
Report January 2012 DES2448, The Royal Society,<br />
2012.<br />
[30] S. Turkle and S. Paper. Epistemological Pluralism and<br />
the Revaluation <strong>of</strong> the Concrete. Journal <strong>of</strong><br />
Mathematical Behaviour, 11(1):pp 3 – 33, 1992.<br />
[31] N. Villar, J. Scott, and S. Hodges. Prototyping with<br />
Micros<strong>of</strong>t .NET Gadgeteer. In Proceedings <strong>of</strong> the 5th<br />
International Conference on Tangible, Embedded, and<br />
Embodied Interaction, TEI ’11, pages 377–380, New<br />
York, NY, USA, 2011. ACM.<br />
[32] N. Villar, J. Scott, S. Hodges, K. Hammill, and<br />
C. Miller. .NET Gadgeteer: A Platform for Custom<br />
Devices. Proceedings <strong>of</strong> Pervasive 2012, Lecture Notes<br />
in Computer Science, 2012.<br />
[33] C. Wilson, L. A. Sudol, C. Stephenson, and<br />
M. Stehlik. Running on Empty: The Failure to Teach<br />
K–12 Computer Science in the Digital Age. Technical<br />
report, CSTA, 2010.<br />
[34] S. Yarosh and M. Guzdial. Narrating Data Structures:<br />
the Role <strong>of</strong> Context in CS2. J.Educ.Resour.Comput.,<br />
7(4):6:1–6:20, Jan. 2008.<br />
106
eledSQL – A New Web-Based Learning Environment for<br />
Teaching Databases and SQL at Secondary School Level<br />
ABSTRACT<br />
Andreas Grillenberger<br />
<strong>University</strong> <strong>of</strong> Erlangen-Nuremberg<br />
Didactics <strong>of</strong> Informatics<br />
Martensstr. 3<br />
91058 Erlangen<br />
mail@andreas-grillenberger.de<br />
Data modeling using databases and SQL is a fundamental<br />
part <strong>of</strong> the curriculum <strong>of</strong> secondary computing education<br />
in <strong>Germany</strong>. Pr<strong>of</strong>essional database tools like HeidiSQL,<br />
phpMyAdmin or Micros<strong>of</strong>t Access are <strong>of</strong>ten used in class as<br />
a “learning s<strong>of</strong>tware”, although these tools have been developed<br />
for managing complex databases, <strong>of</strong>ten for companies,<br />
and not for educational purposes. Such tools <strong>of</strong>fer a wide<br />
range <strong>of</strong> functions <strong>of</strong> which only a small part is required by<br />
secondary computing education. At the beginning <strong>of</strong> such<br />
instruction, students can hardly work independently with<br />
these programs, as they do not know the database language<br />
SQL by then. This <strong>of</strong>ten leads to theory-loaded introductory<br />
phases <strong>of</strong> such classes or alternatively to a usage <strong>of</strong> such<br />
tools like spreadsheet programs. To address this problem,<br />
a new web-based learning environment for databases and<br />
SQL (named eledSQL), also suitable for mobile devices and<br />
only with the functionality needed for secondary computing<br />
education, was developed. The basic idea was to initially<br />
allow students to make database queries using natural language<br />
and then gradually introduce them to the use <strong>of</strong> SQL.<br />
Starting with a problem analysis and a discussion <strong>of</strong> related<br />
work in the field <strong>of</strong> teaching databases and SQL, in this paper<br />
the conception <strong>of</strong> eledSQL, its implementation and first<br />
experiences with its practical use are described.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers and Education]: Computer and Information<br />
Science Education—Computer science education<br />
General Terms<br />
Design, Human Factors, Languages<br />
Keywords<br />
Databases, SQL, Secondary Education, Learning Environment,<br />
Web-based Learning, Mobile Learning, Tools<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WiPCSE 2012 Hamburg, <strong>Germany</strong><br />
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$15.00.<br />
107<br />
Torsten Brinda<br />
<strong>University</strong> <strong>of</strong> Erlangen-Nuremberg<br />
Didactics <strong>of</strong> Informatics<br />
Martensstr. 3<br />
91058 Erlangen<br />
brinda@cs.fau.de<br />
1. MOTIVATION<br />
In secondary computing education in <strong>Germany</strong> data modeling<br />
using databases and SQL is a fundamental part <strong>of</strong> the<br />
curriculum [1]. As “learning s<strong>of</strong>tware” pr<strong>of</strong>essional database<br />
tools such as HeidiSQL 1 , phpMyAdmin 2 or Micros<strong>of</strong>t Access<br />
are used in class – tools, which were not designed for educational<br />
purposes but for the pr<strong>of</strong>essional management <strong>of</strong><br />
databases. In supplementary material to the Bavarian computer<br />
science high school curriculum [16], for example, either<br />
the use <strong>of</strong> the XAMPP package 3 (with web server, MySQL<br />
database server, phpMyAdmin) or <strong>of</strong> Micros<strong>of</strong>t Access is recommended<br />
by the authors. These pr<strong>of</strong>essional tools have in<br />
common that on the one hand they <strong>of</strong>fer a wide range <strong>of</strong><br />
functions, <strong>of</strong> which only a small part is needed for secondary<br />
computing education, and on the other they require previous<br />
knowledge about databases and the SQL language for<br />
its competent use. The use <strong>of</strong> such programs in introductory<br />
lessons on databases is therefore hardly possible. As a<br />
consequence, <strong>of</strong>ten either theory-loaded introduction phases<br />
can be observed in class or the database s<strong>of</strong>tware is used<br />
like a spreadsheet program there. The second option entails<br />
the risk that the students – because <strong>of</strong> the similarities <strong>of</strong> the<br />
user interfaces <strong>of</strong> for example Micros<strong>of</strong>t Excel and the table<br />
view <strong>of</strong> Micros<strong>of</strong>t Access – which supports their orientation<br />
in the beginning – <strong>of</strong>ten have difficulties in recognizing the<br />
differences between the two system categories and do not or<br />
not fully un<strong>der</strong>stand the difference between a spreadsheet<br />
and a database. Other problems arise, when the learners<br />
find out how to use the QBE function (query by example)<br />
within Micros<strong>of</strong>t Access, where the un<strong>der</strong>lying SQL code is<br />
generated automatically [5].<br />
The use <strong>of</strong> XAMPP has the advantage that with the appropriate<br />
user rights configuration the learners can experience<br />
the advantages <strong>of</strong> multi-user access to a database.<br />
With Micros<strong>of</strong>t Access this is potentially also possible but<br />
consi<strong>der</strong>ably more complicated. Another advantage <strong>of</strong> the<br />
client-server architecture is that the learners can access the<br />
database from outside the school (e.g. to do homework),<br />
if the database server is accessible on the Internet at all.<br />
However, secondary school servers are <strong>of</strong>ten configured very<br />
restrictively and <strong>of</strong>ten the installation <strong>of</strong> a database server<br />
is not allowed there for security reasons. In this case, client<br />
and server are sometimes installed in the computer room(s)<br />
<strong>of</strong> the school on every single workstation, which is very la-<br />
1 www.heidisql.com (Jun 20th, 2012)<br />
2 www.phpmyadmin.net (Jun 20th, 2012)<br />
3 www.apachefriends.org/en/xampp.html (Jun 20th, 2012)
orious, contrary to the client-server idea and excludes the<br />
experience <strong>of</strong> multi-user access.<br />
Summing up, although these systems cover all concepts to<br />
be taught in class by corresponding functions, they appear,<br />
mainly because <strong>of</strong> the required prior knowledge and the variety<br />
<strong>of</strong> functions contained in them, not very recommendable<br />
for a use in introduction phases into databases and SQL in<br />
computer science classes in secondary schools. So the question<br />
is: how should a system better suitable for this purpose<br />
be designed?<br />
2. TEACHWARE FOR DATABASES / SQL<br />
As a next step published research on learning environments<br />
in the area <strong>of</strong> databases and SQL was analyzed. In<br />
the literature a number <strong>of</strong> such environments are described<br />
which, however, refer mostly to higher education. Most <strong>of</strong><br />
these environments can be categorized as systems, which visualize<br />
interactive SQL examples using multimedia or those<br />
that provide SQL problems and evaluate student solutions<br />
[2].<br />
The first category includes, e.g. the desktop application<br />
eSQL [10], which visualizes stepwise the selection <strong>of</strong> output<br />
data instead <strong>of</strong> the presentation <strong>of</strong> a result for given SQL<br />
queries, the QueryViz system [6] with a special focus on the<br />
visualization <strong>of</strong> nested queries, the web-based SAVI system<br />
(SQL Advanced Visualization [4]), which allows navigation<br />
<strong>of</strong> the animations in both directions, and ADbC (Animated<br />
Database Courseware) [13], which provides a large number<br />
<strong>of</strong> interactive components for the visualization <strong>of</strong> concepts<br />
<strong>of</strong> the fields <strong>of</strong> databases and database security. The ADbC<br />
collection is extensive, sees itself as a blended learning material<br />
and provides interactive visualizations <strong>of</strong> e.g. various<br />
notations in the area <strong>of</strong> entity relationship (ER) modeling or<br />
<strong>of</strong> prepared questions, e.g. the transfer <strong>of</strong> given ER diagrams<br />
into tables with a visualization <strong>of</strong> the consequences <strong>of</strong> incorrect<br />
answers. In addition, visualizations <strong>of</strong> the execution <strong>of</strong><br />
predefined SQL queries are provided. Other systems focus<br />
on the visualization <strong>of</strong> specific aspects such as on the process<br />
<strong>of</strong> ER modeling [9] or the normalization <strong>of</strong> a database [11].<br />
The second category includes systems such as SQLTutor<br />
[12] and SQLator [15]. The learners are confronted with<br />
problems <strong>of</strong> everyday life and must create respective SQL<br />
queries to solve these problems. The solutions are then analyzed<br />
by these systems using heuristics, correct solutions are<br />
presented and any errors explained as far as possible. Elements<br />
<strong>of</strong> both categories can be found in web-based systems<br />
such as SQLzoo.net and Teradata SQL Assistant, which provide<br />
a set <strong>of</strong> predefined and not modifiable tables and queries<br />
plus a detailed feedback upon execution [5].<br />
The most far-reaching integrated approach finally comes<br />
from Brusilovsky et al. [2, 3], who combined interactive<br />
examples, an SQL knowledge tester, adaptivity and SQL<br />
problems in a self-developed LMS using an un<strong>der</strong>lying SQL<br />
ontology.<br />
All these systems are pure learning environments. Some<br />
include hard-wired examples that are not changeable, others<br />
focus only on parts <strong>of</strong> the school-relevant content but all<br />
emphasize the explanation <strong>of</strong> concepts and neglect productive<br />
work. A system suitable for secondary schools should<br />
combine learning and working. This also requires an adjustment<br />
to the special needs <strong>of</strong> schools with teachers, classes<br />
and students.<br />
108<br />
3. REQUIREMENTS ANALYSIS<br />
Based on the results <strong>of</strong> the analysis <strong>of</strong> database tools<br />
and learning environments, we propose a new system better<br />
suited to the needs <strong>of</strong> introductory school classes. Such<br />
a system should be accessible with any web browser over<br />
the Internet and <strong>of</strong>fer s<strong>of</strong>tware-ergonomic user interfaces for<br />
both desktop pcs and mobile devices <strong>of</strong> all kinds in accordance<br />
with the currently changing media use in schools [14].<br />
The system’s user interface should be easily adaptable to the<br />
corporate design <strong>of</strong> an educational institution. Installation<br />
on a school server or an external web space (in the case <strong>of</strong><br />
school safety concerns) should be as simple as possible in or<strong>der</strong><br />
to keep the resulting workload for teachers low. A common<br />
weakness <strong>of</strong> previously published learning environments<br />
(see sect. 2) is that they provide many different useful visualizations<br />
and problems but that their integration into basic<br />
learning management system functionality adapted to the<br />
school needs is missing. The roles and user rights concept<br />
to be implemented allows differentiated functions and rights<br />
for teachers, school classes, learners and groups <strong>of</strong> learners<br />
as well as an administrator. This provides the possibility <strong>of</strong><br />
adapting such a system to meet the specific needs <strong>of</strong> different<br />
computer science courses <strong>of</strong> a school and its students and to<br />
enable their simultaneous work with the system. Concerning<br />
content, the students should be introduced to working with<br />
databases and SQL with the system. This results in the<br />
requirement that the system must be customizable to the<br />
learning progress <strong>of</strong> learners and groups <strong>of</strong> learners. This<br />
could be realized by three different modes to interact with<br />
databases. At the lowest level the formulation <strong>of</strong> queries<br />
is partly based on forms and partly on natural language.<br />
Prior SQL knowledge is not required there. On the next<br />
higher level queries can be entered as on the lowest level, in<br />
addition, however, the un<strong>der</strong>lying SQL syntax appears automatically<br />
(see fig. 1). On the third level the interaction with<br />
the database should be only possible with SQL. The system<br />
should provide only the functionality needed in class and try<br />
not to compete with pr<strong>of</strong>essional database tools. The implemented<br />
functionality should cover typical secondary schools’<br />
database curricula like the Bavarian one [16].<br />
Users <strong>of</strong> the role <strong>of</strong> teachers should be able to create student<br />
accounts or import accounts from a file and manage<br />
their user rights. Students should be assignable to a class<br />
or a course by them. Furthermore, groups <strong>of</strong> students for<br />
group work should be definable. Another point <strong>of</strong> criticism<br />
<strong>of</strong> existing learning environments was a lack <strong>of</strong> adaptability<br />
to the specific target group (see sect. 2). The teacher should<br />
be able to set the query mode (see above) for entire classes.<br />
Many learning environments provide predefined tables and<br />
queries, which can not be changed and adapted to a specific<br />
course. Therefore, a teacher should be able to create own<br />
table templates, import such templates from files in CSV<br />
format and share them with classes, groups or learners as a<br />
copy. By this, tables can easily be recovered when needed<br />
again. To focus the educational processes on the essential,<br />
SQL functions, such as select, update, insert, delete,<br />
show (or their natural language representation), should be<br />
activatable and also deactivatable for specific tables. At various<br />
locations in the system, the teacher should be able to<br />
adjust teaching texts and provide stepwise assistance for the<br />
SQL queries. A teacher should finally be able to switch to<br />
the perspective <strong>of</strong> any learner and use all his or her features.<br />
A user in the administrator should additionally be able to
create and manage teacher accounts.<br />
Users in the role <strong>of</strong> students should be able to perform all<br />
activated operations on the tables provided by the teacher.<br />
As the students work only on table copies all activated functions<br />
can be executed without any risk <strong>of</strong> destruction. This<br />
also supports exploratory learning. There is also the possibility<br />
that students, groups or classes can create, edit and<br />
delete tables together, if this feature has been enabled by<br />
the teacher. The aim is to make the learners experience the<br />
multi-user mode <strong>of</strong> a database. The tables provided by the<br />
teacher can only be deleted though, if this has explicitely<br />
been approved by him or her. Finally, the learners should<br />
be able to see only those materials provided to them, their<br />
group or class and not those provided to any others.<br />
4. DESIGN AND IMPLEMENTATION<br />
The development process <strong>of</strong> the specified system, the Erlangen<br />
learning environment for databases and SQL – eled-<br />
SQL, with requirements analysis, use cases, database design,<br />
class diagrams and implementation <strong>of</strong> the s<strong>of</strong>tware is<br />
described in detail in [7]. In the following, we sketch only<br />
some important aspects. Due to the requirements that the<br />
system should be web-based and that the installation efforts<br />
should be kept low for teachers, the decision for a serverbased<br />
system with browser-based access was made. To meet<br />
school conditions, it was taken into account that on the one<br />
hand, the components needed on an internal school server to<br />
install eledSQL are standard components free <strong>of</strong> charge and<br />
that the s<strong>of</strong>tware can also be run via an external web hoster,<br />
if the school safety conception is contrary to the installation<br />
on site. External web hosters usually <strong>of</strong>fer pre-structured<br />
packages where, for example, usable programming languages<br />
and databases are predefined. Because the language PHP<br />
and MySQL databases are consi<strong>der</strong>ably more widespread<br />
than their alternatives, both have been chosen for the development<br />
<strong>of</strong> eledSQL. PHP is very well suited for the development<br />
<strong>of</strong> web applications and also supports object-oriented<br />
s<strong>of</strong>tware development, so that the system was developed in<br />
an object-oriented way because <strong>of</strong> the resulting structuring<br />
benefits, where possible and appropriate. Conceptually, the<br />
s<strong>of</strong>tware was structured in three parts following the modelview-controller<br />
pattern. The view contains templates that<br />
correspond to the user interface (see fig. 1) and can be<br />
Figure 1: User interface <strong>of</strong> eledSQL system<br />
109<br />
easily adapted to the corporate design <strong>of</strong> a school without<br />
deep PHP knowledge. In or<strong>der</strong> to support different language<br />
versions <strong>of</strong> the s<strong>of</strong>tware, all texts <strong>of</strong> the user interface<br />
have been moved to a language file. The model contains<br />
functional classes that generate content based on user input<br />
and database access. The controller mediates between<br />
these levels (see fig. 2). In or<strong>der</strong> to use potentially other<br />
Template<br />
Controller<br />
Functions<br />
MySQL<br />
database<br />
Login<br />
Single column<br />
Database<br />
Learner<br />
Database<br />
access<br />
MySQL any<br />
PDO-compatible<br />
database<br />
Two columns<br />
Teacher<br />
Invocation <strong>of</strong> the responsible controller<br />
Role-dependant<br />
functions<br />
Invocation<br />
<strong>of</strong> the template<br />
Invocation<br />
<strong>of</strong> functions<br />
Initialisation <strong>of</strong> function classes<br />
Initialisation <strong>of</strong> the used database<br />
PDO<br />
database<br />
Legend<br />
Template<br />
index file<br />
main class<br />
PDO<br />
database<br />
Architectural level<br />
User<br />
Index file<br />
Main class<br />
Registry<br />
(Procedural) Index file<br />
(Object-oriented) Classes<br />
(Object-oriented) Database<br />
classes<br />
Figure 2: Architecture <strong>of</strong> the eledSQL system<br />
databases like PostgreSQL an additional abstraction layer<br />
from the database queries using PHP Data Objects (PDO)<br />
was integrated into the system. This makes the system even<br />
more flexible. To implement the security requirements first<br />
separate databases for system data, classes, groups and individuals<br />
were consi<strong>der</strong>ed. Since external web hosters typically<br />
limit the number <strong>of</strong> databases this option was dismissed.<br />
For this reason, the system uses a single database<br />
in which the table names reflect the intended addressees.<br />
System tables (marked with “sys”, such as user accounts)<br />
are processable only by means <strong>of</strong> the management function<br />
built-in into eledSQL or an external program. Template tables<br />
(marked with “tpl”) can be created, filled with data<br />
and then be made available to the learners as a copy by<br />
a teacher. Tables marked with an identifier <strong>of</strong> the respective<br />
class, group or learner are available only for the referred<br />
learners. Before executing an SQL statement it can be easily<br />
verified by checking the label <strong>of</strong> each involved table whether<br />
the permission to do so is given. Furthermore, it is easy to<br />
check whether a student is allowed to perform the respective<br />
functions on the tables involved. Also, SQL injections are<br />
therewith preventable.<br />
The web-based user interface <strong>of</strong> the system <strong>of</strong> eledSQL<br />
was tested with major browsers <strong>of</strong> the operating systems<br />
Windows, Linux and MacOS. Furthermore, its usability was<br />
checked with some probands. Identified problems were eliminated.
5. FIRST CLASSROOM EXPERIENCES<br />
An early version <strong>of</strong> eledSQL was used within a teaching<br />
internship <strong>of</strong> the first author at a secondary school in Erlangen<br />
(<strong>Germany</strong>), where a 9 th grade class (learners 14 to<br />
15 years old) was introduced into the field <strong>of</strong> databases. Before<br />
the teaching internship the class had learned to deal<br />
with spreadsheets. The experiences within this course were<br />
throughout positive – the learners could use the tool without<br />
a detailled introduction into the s<strong>of</strong>tware itself. Thereby<br />
they were able to discover the differences between the wellknown<br />
spreadsheets and the new databases by dealing with<br />
exercises prepared by the teacher. For example, they were to<br />
find the address <strong>of</strong> a person in a spreadsheet and a database<br />
with about 1.000 entries each or all persons with names including<br />
a defined string. While dealing with these exercises,<br />
they un<strong>der</strong>stood the different fields <strong>of</strong> application <strong>of</strong><br />
databases and spreadsheets. This is a huge adavantage over<br />
the conventional way <strong>of</strong> teaching – which is in many cases<br />
to tell the differences to the pupils, because they cannot<br />
discover them theirselves due to missing SQL knowledge.<br />
This sometimes results in motivational problems, especially<br />
when the pupils cannot realize why “the same application<br />
with another name” should have these pros and cons, when<br />
only using the tabular view at the beginning.<br />
6. SUMMARY AND OUTLOOK<br />
In this paper we motivated the need for a new learning<br />
environment on databases and SQL for the computing education<br />
at secondary level. We analyzed database s<strong>of</strong>tware<br />
and learning environments used in education and <strong>der</strong>ived<br />
from the analysis results and curricular needs requirements<br />
for the new learning environment, whose conception, design,<br />
implementation and first experiences <strong>of</strong> the educational use<br />
we described afterwards. The current version <strong>of</strong> eledSQL<br />
is available for free (Open Source License) at Sourceforge 4 .<br />
The development <strong>of</strong> this s<strong>of</strong>tware is part <strong>of</strong> a bigger student<br />
project that is continued as part <strong>of</strong> another bachelor thesis,<br />
where a detailed analysis <strong>of</strong> the classroom use in two parallel<br />
classes is made. A further development <strong>of</strong> the system<br />
is also planned. In a next release assessments and teaching<br />
material should be integrated.<br />
7. REFERENCES<br />
[1] T. Brinda, H. Puhlmann, and C. Schulte. Bridging<br />
ICT and CS: educational standards for computer<br />
science in lower secondary education. In Proceedings <strong>of</strong><br />
the 14th annual ACM SIGCSE conference on<br />
Innovation and technology in computer science<br />
education, ITiCSE ’09, pages 288–292, New York, NY,<br />
2009. ACM.<br />
[2] P. Brusilovsky, S. Sosnovsky, D. H. Lee, M. Yudelson,<br />
V. Zadorozhny, and X. Zhou. An open integrated<br />
exploratorium for database courses. In Proceedings <strong>of</strong><br />
the 13th annual conference on Innovation and<br />
technology in computer science education, ITiCSE ’08,<br />
pages 22–26, New York, NY, 2008. ACM.<br />
[3] P. Brusilovsky, S. Sosnovsky, M. V. Yudelson, D. H.<br />
Lee, V. Zadorozhny, and X. Zhou. Learning SQL<br />
programming with interactive tools: From integration<br />
to personalization. Transactions on Computing<br />
Education, 9(4):19:1–19:15, Jan. 2010.<br />
4 http://eledsql.sourceforge.net/<br />
110<br />
[4] M. Cembalo, A. De Santis, and U. Ferraro Petrillo.<br />
SAVI: a new system for advanced SQL visualization.<br />
In Proceedings <strong>of</strong> the 2011 conference on Information<br />
technology education, SIGITE ’11, pages 165–170, New<br />
York, NY, 2011. ACM.<br />
[5] J. Cigas and B. Kushan. Experiences with online SQL<br />
environments. Journal <strong>of</strong> Computing Sciences in<br />
Colleges, 25(5):251–257, May 2010.<br />
[6] J. Danaparamita and W. Gatterbauer. Queryviz:<br />
helping users un<strong>der</strong>stand SQL queries and their<br />
patterns. In Proceedings <strong>of</strong> the 14th International<br />
Conference on Extending Database Technology,<br />
EDBT/ICDT ’11, pages 558–561, New York, NY,<br />
2011. ACM.<br />
[7] A. Grillenberger. Entwurf und Implementierung einer<br />
Lernumgebung für den Informatikunterricht zur<br />
Unterrichtssequenz Datenbanksysteme (in German).<br />
Bachelor’s thesis, <strong>University</strong> <strong>of</strong> Erlangen-Nuremberg,<br />
Didactics <strong>of</strong> Informatics, Erlangen, <strong>Germany</strong>, 2012.<br />
[8] M. Guimaraes and M. Murray. Using animation<br />
courseware in the teaching <strong>of</strong> database security. In<br />
Proceedings <strong>of</strong> the 8th ACM SIGITE conference on<br />
Information technology education, SIGITE ’07, pages<br />
253–258, New York, NY, USA, 2007. ACM.<br />
[9] L. Hall and A. Gordon. A virtual learning<br />
environment for entity relationship modelling. In<br />
Proceedings <strong>of</strong> the 29th ACM SIGCSE technical<br />
symposium on Computer science education, SIGCSE<br />
’98, pages 345–349, New York, NY, 1998. ACM.<br />
[10] R. Kearns, S. Shead, and A. Fekete. A teaching system<br />
for SQL. In Proceedings <strong>of</strong> the 2nd Australasian<br />
conference on Computer science education, ACSE ’97,<br />
pages 224–231, New York, NY, 1996. ACM.<br />
[11] H.-J. Kung and H.-L. Tung. A web-based tool for<br />
teaching data modeling. Journal <strong>of</strong> Computing<br />
Sciences in Colleges, 26(2):231–237, Dec. 2010.<br />
[12] A. Mitrovic. Learning sql with a computerized tutor.<br />
In Proceedings <strong>of</strong> the twenty-ninth SIGCSE technical<br />
symposium on Computer science education, SIGCSE<br />
’98, pages 307–311, New York, NY, 1998. ACM.<br />
[13] M. Murray and M. Guimaraes. Animated database<br />
courseware: using animations to extend conceptual<br />
un<strong>der</strong>standing <strong>of</strong> database concepts. Journal <strong>of</strong><br />
Computing Sciences in Colleges, 24(2):144–150, Dec.<br />
2008.<br />
[14] C. Piech and E. Roberts. Informatics education using<br />
nothing but a browser. Proceedings <strong>of</strong> the IFIP<br />
Conference on ICT and Informatics in a Globalised<br />
World <strong>of</strong> Education, Mombasa, Kenya, August 2011,<br />
2011. http://cs.stanford.edu/˜eroberts/papers/<br />
NothingButABrowser.pdf.<br />
[15] S. Sadiq, M. Orlowska, W. Sadiq, and J. Lin.<br />
SQLator: an online SQL learning workbench. In<br />
Proceedings <strong>of</strong> the 9th annual ACM SIGCSE<br />
conference on Innovation and technology in computer<br />
science education, ITiCSE ’04, pages 223–227, New<br />
York, NY, 2004. ACM.<br />
[16] Staatsinstitut für Schulqualität und<br />
Bildungsforschung. Informatik am<br />
Naturwissenschaftlich-technologischen Gymnasium,<br />
Jahrgangsstufe 9 (Handreichung, in German), 2004.
ABSTRACT<br />
Bringing Contexts into the Classroom –<br />
A Design-Based Approach<br />
Detlef Rick, Marcel Morisse, Ingrid Schirmer<br />
Universität Hamburg, Department <strong>of</strong> Informatics<br />
Vogt-Kölln-Straße 30<br />
D-22527 Hamburg, <strong>Germany</strong><br />
{rick | morisse | schirmer}@informatik.uni-hamburg.de<br />
Project-based learning and application-oriented approaches<br />
drawing on real-world issues and examples, have always been<br />
widely used in both secondary and higher Computer Science<br />
(CS) education. Recent demands for teaching CS or Informatics<br />
1 in context and building courses on coherent application<br />
areas outside CS go even further [4, 23]. In <strong>Germany</strong> the<br />
open working group IniK (Informatik im Kontext) compiles<br />
educational material and concepts for teaching Informatics<br />
in context [42, 24, 10, 8]. However, there is only little theoretical<br />
foundation for IniK as an educational approach and<br />
its implementation at schools and universities [7].<br />
In this paper we suggest a process model for bringing<br />
‘real-world’ application contexts into the Informatics classroom.<br />
The proposed model is grounded on a design-based<br />
study and has evolved from theoretical and practical insights<br />
gained in five university-level project courses and five associated<br />
school projects. The model focuses on the development<br />
<strong>of</strong> context artifacts which originate from the real-world context,<br />
and teaching units building on the context artifacts.<br />
When teaching units are instantiated in the classroom, context<br />
artifacts are recontextualized in an educational context<br />
and are employed as learning equipment fostering the students’<br />
imagination, and thus anchoring the real-world application<br />
context within the classroom context.<br />
In the tradition <strong>of</strong> classical German <strong>Didaktik</strong> models [18,<br />
28], the process model aims at providing a tool and a terminology<br />
for the preparation and reflection <strong>of</strong> instruction. On<br />
the basis <strong>of</strong> the model we discuss both the potential value<br />
and the drawbacks <strong>of</strong> teaching Informatics in context.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers and Education]: Computer and Information<br />
Science Education<br />
1 We use the term Informatics in a broad sense with a view<br />
to describe the disciplines <strong>of</strong> Computer Engineering, Computer<br />
Science, S<strong>of</strong>tware Engineering, Information Systems<br />
and others.<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WiPSCE 2012 Hamburg, <strong>Germany</strong><br />
Copyright 2012 ACM X-XXXXX-XX-X/XX/XX ...$10.00.<br />
111<br />
General Terms<br />
Design, Human Factors<br />
Keywords<br />
Informatics in context, learning environment, process model,<br />
greenfoot, design-based research<br />
1. INTRODUCTION<br />
The Great Hall at Hogwarts School <strong>of</strong> Witchcraft and<br />
Wizardry is equipped with a ceiling that has been bewitched<br />
to look like the sky outside. So for J. K. Rowling’s famous<br />
first-year student Harry Potter 2 it was “hard to believe there<br />
was a ceiling there at all, and that the Great Hall didn’t<br />
simply open on to the heavens” (p. 129). Apparently the<br />
foun<strong>der</strong>s <strong>of</strong> Hogwarts for some reason deemed it necessary<br />
to let the school’s inside resemble the outside world in some<br />
degree.<br />
Recent demands for teaching Computer Science and Informatics<br />
in context [4, 7, 23, 24, 39] request educators to<br />
contextualize their teaching and to bring the outside world<br />
into the classroom. Although CS “would seem to want to<br />
connect to the real world” and although for some projectbased<br />
and application-oriented courses it is not at all uncommon<br />
to build on a consistent real-world context, Cooper and<br />
Cunningham [4] criticize that “computer science curriculum<br />
recommendations are silent about the opportunity to do so”<br />
and that “most widely used texts do not seem to have much<br />
content addressing real-world questions” (p. 5).<br />
Guzdial [15] qualifies this critique, remarking that “it is<br />
reasonable to fear that students who learn computing within<br />
a context might over-specialize that knowledge” and that<br />
“the additional knowledge taught about the context might be<br />
a distraction, especially for the lower-ability students” (p. 5).<br />
Guzdial comes to the conclusion that context can provide<br />
additional relevance if there is evidence that students can’t<br />
see meaning and usefulness <strong>of</strong> decontextualized content.<br />
Instead <strong>of</strong> trying to answer the question <strong>of</strong> whether contextoriented<br />
teaching approaches actually result in higher student<br />
motivation and commitment and thus in better learning<br />
and deeper un<strong>der</strong>standing [39], in the following we propose<br />
a process model for bringing ‘real-word’ contexts into the<br />
classroom environment. The model has evolved from five<br />
cycles <strong>of</strong> our university-level project course for Informatics<br />
students majoring in Computer Science (CS), S<strong>of</strong>tware<br />
2 J. K. Rowling. Harry Potter and the Philosopher’s Stone.<br />
Bloomsbury, London, 1997.
Engineering (SE), Information Systems (IS), or Computer<br />
Science Education (CSEd).<br />
Our students were assigned to develop school projects in<br />
the context <strong>of</strong> our department’s outreach programs conducted<br />
in cooperation with schools and local s<strong>of</strong>tware development<br />
enterprises. Thus our project course followed a<br />
contextualized approach in that our students were to develop<br />
project management and s<strong>of</strong>tware development skills<br />
in the context <strong>of</strong> preparing Informatics teaching units and<br />
designing computational artifacts for the classroom, i. e. in<br />
an educational context outside Informatics. In the project<br />
our students implemented Greenfoot [41] scenarios simulating<br />
some technical and non-technical issues <strong>of</strong> ‘real-world’<br />
contexts, namely the context <strong>of</strong> airport baggage handling<br />
and the context <strong>of</strong> inventory control in retail stores.<br />
The proposed process model is design-based in two respects.<br />
On the one hand the development <strong>of</strong> teaching units<br />
and the design <strong>of</strong> artifacts fitting the classroom context<br />
are consi<strong>der</strong>ed to be main activities in preparing contextoriented<br />
teaching. Thus design activities are an integral part<br />
<strong>of</strong> the proposed process. And on the other hand the model<br />
itself is the result <strong>of</strong> an iterative design process.<br />
Albeit grounded on our specific project experience the<br />
model is reasonably abstract to present a reference framework<br />
within which different conceptions <strong>of</strong> teaching Informatics<br />
in context can be discussed.<br />
2. DESIGN-BASED METHODOLOGY<br />
In 1969 Simon introduced the concept <strong>of</strong> design sciences<br />
as the sciences <strong>of</strong> the artificial [36]. According to Simon the<br />
artificial is investigated not only in terms <strong>of</strong> what is, but also<br />
in terms <strong>of</strong> what ought to be. Design studies therefore aim at<br />
both the development <strong>of</strong> artifacts – e. g. s<strong>of</strong>tware, processes,<br />
institutions, or interventions [20, 40] – and the generation <strong>of</strong><br />
significant theories and usable knowledge, e. g. design principles<br />
or patterns [1, 29]. Moreover design-based approaches<br />
are characterized by an iterative design process, where the<br />
artifacts are developed in a process <strong>of</strong> ‘progressive refinement.’<br />
Thus, for instance, IS and SE can be characterized<br />
as design sciences [17].<br />
In education research, design-based research methodology<br />
(DBR) is being increasingly utilized [1]. According to<br />
Collins et al. [3] design experiments “bring together two critical<br />
pieces in or<strong>der</strong> to guide us to better educational refinement:<br />
a design focus and assessment <strong>of</strong> critical design elements”<br />
(p. 21). While design experiments are conducted<br />
in specific educational contexts, they are aiming at general<br />
design principles.<br />
The artifact we developed in the course <strong>of</strong> our design experiments<br />
was the process model for bringing contexts into<br />
the classroom, described in section 4. The process model<br />
was instantiated in five university-level s<strong>of</strong>tware development<br />
project courses and five associated week-long school<br />
projects. In the project courses the students assumed diverse<br />
roles. They developed (educational) s<strong>of</strong>tware, prepared<br />
school projects, and finally taught classes in the ‘classroom<br />
context.’ Thus the assessment <strong>of</strong> the process model<br />
was primarily based on the evaluation <strong>of</strong> the project courses,<br />
grounded on:<br />
• assessment <strong>of</strong> the s<strong>of</strong>tware and other artifacts developed<br />
by the students with respect to requirements,<br />
• students’ project reports and course achievements,<br />
112<br />
• student feedback (oral, questionnaire),<br />
• feedback from the boys and girls who attended the<br />
associated school projects (oral, questionnaire), and<br />
• feedback from the teachers <strong>of</strong> our cooperation school<br />
(oral).<br />
In the following section we describe how the process model<br />
evolved from our un<strong>der</strong>standing <strong>of</strong> s<strong>of</strong>tware development as<br />
decontextualization and recontextualization, in combination<br />
with the seminal design <strong>of</strong> Greenfoot, and the idea <strong>of</strong> teaching<br />
Informatics ‘in context.’<br />
3. ORIGINATION OF THE MODEL<br />
Our process model evolved in a multi-institutional project<br />
setting including multiple arenas. In a project-based s<strong>of</strong>tware<br />
development course for Informatics students, first <strong>of</strong>fered<br />
in 2009/10, our students were assigned to develop an<br />
Informatics school project. In the following we describe the<br />
prerequisites that informed the design <strong>of</strong> our first project<br />
course and the corresponding process model, as well as the<br />
subsequent iterations <strong>of</strong> progressive refinement.<br />
3.1 Phase 0: Prerequisites<br />
Our objectives for the project course were tw<strong>of</strong>old. On<br />
the one hand our Informatics students were to develop, for<br />
instance, project management and social skills, s<strong>of</strong>tware development<br />
methods and so forth. In this project we had external<br />
clients – the cooperating schools – and a firm project<br />
deadline. On the other hand we defined requirements concerning<br />
the school project – the product.<br />
With the school project we wanted to rouse the boys’ and<br />
girls’ interest in Informatics and thus induce them to choose<br />
optional Informatics courses at school. According to <strong>Knobelsdorf</strong><br />
and Schulte [23] context-oriented approaches are<br />
auspicious with regard to increasing participation in Computing.<br />
Therefore we were positive that a context-oriented<br />
approach was motivating not only for our students but also<br />
for the boys and girls who participated in the school project.<br />
In or<strong>der</strong> to present a broad image <strong>of</strong> our discipline we aimed<br />
to expose the youngsters to a complex real-world context<br />
where information technology can make a difference. Moreover<br />
we wanted to provide insight into a few CS core concepts,<br />
like variables, control structures and object orientation,<br />
as well as into more application-oriented issues. So our<br />
students were expected to develop multiple views <strong>of</strong> Computing<br />
and to share these with the pupils, drawing examples<br />
from one coherent application context. In cooperation with<br />
a local s<strong>of</strong>tware development enterprise, we chose the context<br />
<strong>of</strong> airport baggage handling [31] and made it accessible<br />
to the students by arranging an on-site inspection at the<br />
Cologne Bonn Airport, and talks with domain experts from<br />
both the airport and the s<strong>of</strong>tware producer PSI Logistics<br />
GmbH. Thus our students were provided insight into the<br />
context they had to analyze and to un<strong>der</strong>stand in or<strong>der</strong> to<br />
present it to the class later on in the school project.<br />
As a technological framework we used the educational<br />
Java programming environment Greenfoot [41] which is also<br />
a framework for the development <strong>of</strong> microworlds, i. e. 2D<br />
games and simulations, or scenarios [16, 26]. Hence the s<strong>of</strong>tware<br />
our students were to develop, essentially was Greenfoot<br />
scenarios.
3.1.1 Greenfoot as a framework for designing<br />
learning environments<br />
The idea <strong>of</strong> using Greenfoot for the project course arose<br />
from the fact that Greenfoot is not just the programming environment<br />
itself but features an easy-to-use Java API for the<br />
development <strong>of</strong> 2D-graphical games and simulations. Since<br />
the Greenfoot framework is based on Java, it facilitates the<br />
development <strong>of</strong> quite complex scenarios: “While it is possible<br />
to create simple games quickly and easily in Greenfoot, it<br />
is equally possible to build highly sophisticated simulations<br />
<strong>of</strong> complex systems [...]” (Kölling [27], p. 2).<br />
Thus Greenfoot is, in principle, well suited to provide both<br />
an introductory development environment and meaningful<br />
contexts. Henriksen and Kölling suggest that Greenfoot<br />
would open new possibilities for course design because it allows<br />
for using multiple scenarios <strong>of</strong> varied complexity and<br />
from different application areas [16]. Moreover scenarios<br />
can, in principle, be adapted to different learning theories<br />
and to different learning objectives related to both object<br />
orientation and other informatical issues beyond programming.<br />
These new possibilities are promising. But obviously the<br />
freedom to develop and adapt scenarios, or maybe even just<br />
to select them to match specific course objectives, involves<br />
additional time and effort for course preparation (figure 1).<br />
Traditional microworlds Greenfoot<br />
Implement framework<br />
(with fixed scenario)<br />
Design exercise<br />
Solve exercise<br />
Framework<br />
implementor<br />
Teacher<br />
Student<br />
Implement framework<br />
Design scenario<br />
Design exercise<br />
Solve exercise<br />
Framework<br />
implementor<br />
Student<br />
Teacher<br />
Figure 1: Roles in creation and use <strong>of</strong> traditional<br />
microworlds and Greenfoot scenarios, according to<br />
Henriksen and Kölling [16]. Greenfoot scenarios –<br />
games and simulations – can be shared in the Greenfoot<br />
gallery or maybe along with some ‘nifty assignments’<br />
[11] in the Greenroom. (Adapted from [16])<br />
Our project course aimed for fathoming the possibilities<br />
provided by the Greenfoot framework, and the limitations.<br />
In consi<strong>der</strong>ation <strong>of</strong> Nygaard’s advice that teaching objectorientation<br />
must start with a “sufficiently complex example”<br />
[30], we focused on developing scenarios that were conspicuously<br />
more complex than conventional microworlds. The<br />
challenge was to design these complex scenarios in such a<br />
way that they would fit well into the framework and at the<br />
same time match the classroom requirements so as to arrange<br />
learning environments around them.<br />
While Greenfoot as an IDE is aimed at young students<br />
[43], for advanced students it’s well-designed API gives a<br />
good and concise example <strong>of</strong> an object-oriented framework.<br />
In developing against the Greenfoot framework (and yet using<br />
a pr<strong>of</strong>essional development environment) our Informatics<br />
students learn how to elegantly handle and extend a given<br />
framework.<br />
113<br />
3.1.2 S<strong>of</strong>tware development as decontextualization<br />
and recontextualization<br />
Our department has a strong tradition in questioning the<br />
“separability <strong>of</strong> production from the use <strong>of</strong> products in social<br />
contexts” ([14], p. 49), that is associated with e. g. Floyd<br />
and colleagues [12, 14], and that has been influenced by<br />
the so-called Scandinavian Approach [13]. Typically, when<br />
our students enroll for the project course they are already<br />
acquainted with the ideas <strong>of</strong> agility [2] and the perspective<br />
on s<strong>of</strong>tware development as an evolutionary process that<br />
intertwines s<strong>of</strong>tware production, embedment and use.<br />
Therefore another starting point for the development <strong>of</strong><br />
our process model is our un<strong>der</strong>standing <strong>of</strong> s<strong>of</strong>tware development<br />
as a design process and a social activity incorporating<br />
alternate decontextualization and recontextualization<br />
activities. Instead <strong>of</strong> bringing up Dijkstra’s ‘firewall’ between<br />
the realm <strong>of</strong> the ‘correctness problem’ and the realm<br />
<strong>of</strong> the ‘pleasantness problem’ [9], the dialectic <strong>of</strong> decontextualization<br />
and recontextualization emphasizes that s<strong>of</strong>tware<br />
is situated in human contexts.<br />
Decontextualization refers to the analysis and translation<br />
<strong>of</strong> contextualized meaning and ‘situated action’ [38] into<br />
models, algorithms, programs and other computational artifacts.<br />
On the other hand recontextualization is the transfer<br />
and embedment <strong>of</strong> formalized decontextualized activities,<br />
programs etc. into a human context and results in transforming<br />
it. Figure 2 illustrates this ‘socio-technical core’ on<br />
the basis <strong>of</strong> Rolf and colleagues [25, 33, 44].<br />
Along with the ‘roles in creation and use <strong>of</strong> Greenfoot<br />
scenarios’ shown in figure 1, we adopted this model as a<br />
central pattern for the design <strong>of</strong> our process model.<br />
Context<br />
Recontextualization<br />
Decontextualization<br />
Artifacts<br />
Figure 2: The ‘socio-technical core’ (cf. [25, 33, 44])<br />
<strong>of</strong> alternate decontextualization and recontextualization<br />
processes in s<strong>of</strong>tware development activities<br />
serves us as a starting point and as a pattern for our<br />
model design.<br />
3.2 Phase 1: First project course<br />
The outline <strong>of</strong> our first project course <strong>of</strong>fered in winter<br />
2009/10 can be consi<strong>der</strong>ed as our first prototype process<br />
model. Our course <strong>of</strong> action featured many aspects we later<br />
elaborated in our model. A central activity was developing<br />
computational artifacts like models, Greenfoot simulations<br />
and other programs and documents that would be employed<br />
as learning equipment in the classroom context <strong>of</strong> a school<br />
project. But we also gave attention on theoretical and transdisciplinary<br />
foundations like learning styles, gen<strong>der</strong> and diversity<br />
issues [32].<br />
The most important result that emerged from this first<br />
project course was a terminology related to our course setting.<br />
In contrast to other s<strong>of</strong>tware development projects in<br />
our project the context that was to be analyzed <strong>of</strong>ten was not
the same as the context the s<strong>of</strong>tware was being developed<br />
for. The artifacts were never to be recontextualized into the<br />
original context they were <strong>der</strong>ived from. Moreover conversations<br />
with our cooperation partners from different domains<br />
occasionally caused confusion about what ‘the project’, ‘the<br />
requirements’ or ‘the application context’ was.<br />
For example, in our first project we had a Requirements<br />
team who analyzed the context <strong>of</strong> airport baggage handling<br />
and who held interviews with domain experts. In the course<br />
<strong>of</strong> the project it became apparent that requirements came<br />
from multiple different contexts (see figure 3) and that the<br />
s<strong>of</strong>tware had to meet not only the requirements from the application<br />
context but also the above-mentioned requirements<br />
that were defined by the course instructors for the school<br />
project. We therefore accustomed ourselves and our students<br />
to distinguish, for instance, between the ‘real-world’<br />
application context and the classroom or school context.<br />
Evaluation <strong>of</strong> phase 1.<br />
In their feedback our students confirmed that the topic<br />
<strong>of</strong> the project course (Knowledge acquisition, Gen<strong>der</strong> and<br />
Diversity in Informatics and IT ) was interesting and important.<br />
They also liked trying things out on their own.<br />
” gutes, wichtiges Thema; gute Einbindung <strong>der</strong><br />
Theorie in die Praxis; Möglichkeit des Ausprobierens“<br />
[good, important topic; good integration<br />
<strong>of</strong> theory in practice; opportunity to try things<br />
out]<br />
” Interessantes Thema, viel eigenständige Arbeit“<br />
[interesting topic, a good deal <strong>of</strong> independent<br />
work]<br />
On the other hand students criticized that project organization,<br />
assignments and purpose were not clear enough.<br />
” nicht immer gut organisiert, klare Zielsetzung<br />
hat gefehlt“ [not always well organized, clear<br />
purpose was missing]<br />
Also the students’ suggestions for improvement focused on<br />
project organization and clarity <strong>of</strong> assignments:<br />
” klare Aufgaben / Ziele“ [clear mission / purpose]<br />
” aktive Organisationsgruppe, um die Interaktion<br />
zwischen den Gruppen zu verbessern“ [establish<br />
an active organization team in or<strong>der</strong> to improve<br />
interaction between teams]<br />
Therefore in the next iteration we concentrated on structuring<br />
the aspects and tasks we identified during phase one.<br />
The development <strong>of</strong> Greenfoot scenarios and other artifacts<br />
became the focus <strong>of</strong> all subsequent project courses while<br />
theoretical, technological, contextual and educational perspectives<br />
were conditioning requirements and design aims.<br />
Figure 3 represents the structure <strong>of</strong> this task, thus extending<br />
the model on the right side <strong>of</strong> figure 1.<br />
Not only the university-level project course but also the<br />
school projects during phase one were very successful. The<br />
Greenfoot Baggage Handling System (GBHS) developed by<br />
the students was used in three school projects with pupils<br />
at different ages (13–15; 18–20; 17–18). As Informatics-incontext<br />
projects these hands-on classes aimed at bringing<br />
a real-world context into the classroom, not only programming.<br />
Therefore our students included issues from IS in the<br />
114<br />
Application<br />
context,<br />
domain<br />
knowledge<br />
Greenfoot<br />
framework<br />
Design <strong>of</strong><br />
Greenfoot<br />
scenario<br />
Teaching- learning<br />
situation, classroom<br />
environment<br />
Different<br />
views on<br />
Computing<br />
Figure 3: The requirements for the design <strong>of</strong> Greenfoot<br />
scenarios come from multiple contexts.<br />
school project [31]. This was extremely motivating. Especially<br />
the young pupils (13–15) were very focused when<br />
they calculated key indicators in or<strong>der</strong> to optimize the layout<br />
<strong>of</strong> their conveyor systems (see section 5). The cooperating<br />
teacher came to the conclusion:<br />
” Die Kin<strong>der</strong> haben spielerisch gelernt, ohne<br />
es wirklich zu merken, wir konnten für etwas mehr<br />
Informatikinhalte in <strong>der</strong> Schule werben und einige<br />
Eltern haben mich sogar angesprochen, was<br />
wir denn mit den Kin<strong>der</strong>n gemacht hätten, so<br />
überschwänglich hätten sie noch nie von <strong>der</strong> Schule<br />
berichtet.“ [The children learned through play<br />
without really noticing it. We succeeded in promoting<br />
Informatics in schools: some parents even<br />
asked what we’ve done to the kids because they’ve<br />
never before been so rhapsodic about school!]<br />
3.3 Phase 2: Design aims for context artifacts<br />
Subsequent to our first iteration we developed a process<br />
model by combining the tasks identified in phase one (figure<br />
3) with a modified ‘socio-technical core’ where decontextualization<br />
and recontextualization concern two different<br />
contexts (figure 2). Furthermore in phase two we abstracted<br />
from the concrete endeavor <strong>of</strong> designing Greenfoot scenarios.<br />
In fact from the beginning <strong>of</strong> our first project course our<br />
students have also created other artifacts, e. g. models and<br />
diagrams. Thus in figure 4 the Greenfoot scenarios, along<br />
with the other artifacts, are depicted by the centric trapezoid<br />
shapes. The Greenfoot framework now is part <strong>of</strong> the<br />
technological platform the artifacts are usually based on.<br />
In or<strong>der</strong> to make clear that the artifacts are related to<br />
a consistent context and that not any and every artifact<br />
can be included in the Informatics-in-context classroom, we<br />
suggest the notion <strong>of</strong> context artifacts. With this term we<br />
refer to both artifacts that have been decontextualized from<br />
their original context where they have a situated meaning,<br />
and newly designed artifacts objectifying situated meaning,<br />
e. g. formal models, algorithms and s<strong>of</strong>tware.<br />
As a result from phase two we present the following design<br />
aims for context artifacts (figure 4):<br />
Technological framework. Developers need to know their<br />
tools. Context artifacts should be designed to elegantly<br />
build on and seamlessly fit into the technological framework.<br />
The technological framework also includes, for instance, the
Real-world<br />
context<br />
Views<br />
<strong>of</strong><br />
Computing<br />
Context<br />
Artifacts<br />
Technological framework;<br />
tools and media<br />
Classroom<br />
context<br />
Figure 4: When context artifacts (e. g. Greenfoot<br />
scenarios) are developed as a vehicle for bringing<br />
context into the classroom, they need to be designed<br />
in such a way that they integrate not only with the<br />
technological framework and the educational context<br />
but also with the ‘real-world’ context and the<br />
different perspectives on Computing.<br />
school network infrastructure and learning management system.<br />
Educational context. Special attention must be paid to<br />
the educational context the artifacts are designed for, i. e.<br />
subject-matter, learning objectives, curriculum and educational<br />
principles like learning theories, gen<strong>der</strong> and diversity<br />
aspects etc. Often context artifacts need to be specifically<br />
adapted to the classroom context.<br />
Real-world application context. Moreover, in an Informatics-in-context<br />
learning environment, context artifacts serve<br />
as vehicles bringing the real-world context into the classroom.<br />
They are not just learning objects but represent the<br />
‘real world’ and translate meaning from the real-world application<br />
context into the classroom context. Therefore they<br />
must be designed to bear a meaning when contextualized in<br />
or related with the real-world context.<br />
Informatical perspective. Finally Informatics-in-context<br />
classes are Informatics classes. Hence context artifacts must<br />
be selected or designed on the basis <strong>of</strong> their relevance from<br />
an informatical point <strong>of</strong> view. On the one hand curricula define<br />
what the body <strong>of</strong> knowledge <strong>of</strong> Informatics is and what<br />
is relevant for educational purposes. On the other hand, as<br />
we see it, Informatics in context aims at presenting a broad<br />
image <strong>of</strong> Informatics. Therefore the design <strong>of</strong> context artifacts<br />
should integrate multiple views on Computing and<br />
Informatics.<br />
Evaluation <strong>of</strong> phase 2.<br />
The phase-two students’ feedback proved that clarity and<br />
project organization had improved. At the same time we<br />
succeeded in maintaining the strengths <strong>of</strong> the project:<br />
” Guter Themenbereich, eher ungewöhnlich für<br />
FB Informatik“<br />
[Good topic, quite unusual for the department <strong>of</strong><br />
Informatics]<br />
” Die Veranstalter waren immer gut vorbereitet.<br />
Die Aufgabenstellungen waren größtenteils klar<br />
115<br />
definiert. Durch die Projektwoche war die Arbeit<br />
nicht so einseitig.“<br />
[The lecturers were always well prepared. The assignments<br />
were mostly clear. The school project<br />
<strong>of</strong>fered a diversion from the other activities.]<br />
Moreover many <strong>of</strong> the students appreciated the guidance by<br />
the lecturers. But there were still also many critical voices:<br />
” teilweise unübersichtliche Organisation, schwammig<br />
gestellte Aufgaben, was zu Unklarheiten führte“<br />
[sometimes unclear organization, vaguely formulated<br />
assignments resulting in confusion]<br />
Other objections were related to the composition <strong>of</strong> teams.<br />
Some students complained about “unequal” teams and demanded<br />
“fair group forming” for future projects. Therefore<br />
in phase three we gave attention to roles.<br />
3.4 Phase 3: Pre-educational stage<br />
and educational stage<br />
At the beginning <strong>of</strong> phase three we started to ask which<br />
<strong>of</strong> the proposed tasks teachers can effectively assume when<br />
preparing lessons. While in the first cycle <strong>of</strong> our project<br />
course there were many students studying IS (14 in 21),<br />
the students in the second course had a very diverse background,<br />
majoring in CS, SE, IS or CSEd, so that they focused<br />
on different aspects <strong>of</strong> the project. Students from CS,<br />
SE and IS concentrated mostly on different aspects <strong>of</strong> decontextualization<br />
(analysis <strong>of</strong> context and construction <strong>of</strong><br />
context artifacts), whereas CSEd students were interested<br />
in the recontextualization (preparation and implementation<br />
<strong>of</strong> teaching units).<br />
Thus the idea to discern a pre-educational and an educational<br />
stage seemed appropriate. The pre-educational stage<br />
is independent from a concrete educational context though<br />
it nevertheless assumes an educational purpose. Because <strong>of</strong><br />
its independence the first stage can be treated by external<br />
experts, or educators can be supported by them in doing so,<br />
e. g. in cooperation with universities. The educational stage<br />
needs to be conducted by educators.<br />
The distinction <strong>of</strong> pre-educational and educational stage<br />
brought about another distinction. As a result from this<br />
iteration we also distinguish between context artifacts and<br />
teaching units. Teaching units are a class <strong>of</strong> artifacts augmented<br />
with context artifacts. In contrast to context artifacts,<br />
teaching units are not decontextualized from the realworld<br />
application context but they provide plans for classroom<br />
activities and teaching-learning situations. They can<br />
describe a single lesson, a whole course or any meaningful<br />
chapter <strong>of</strong> it.<br />
Evaluation <strong>of</strong> phase 3.<br />
In the first iterations <strong>of</strong> the project we focused on the<br />
design <strong>of</strong> the university-level project course. Now we broadened<br />
our scope to include the adoption <strong>of</strong> Informatics-incontext<br />
in schools.<br />
The context artifacts the students developed were not only<br />
used in the school projects but also in the Informatics Pr<strong>of</strong>ilkurs<br />
(advanced placement course) <strong>of</strong> one <strong>of</strong> the cooperating<br />
schools. Pr<strong>of</strong>iles – implemented only recently and only in<br />
part at German secondary schools – are meant to promote<br />
competition among schools and to encourage the formation
<strong>of</strong> partnerships and cooperation models in or<strong>der</strong> to diversify<br />
educational opportunities. Since educational standards<br />
must be met by all schools pr<strong>of</strong>iles are closely related to<br />
context-oriented approaches.<br />
Therefore for the evaluation <strong>of</strong> phase three we paid more<br />
attention to the two pr<strong>of</strong>ile teachers’ feedback. They reported<br />
that the pr<strong>of</strong>ile is very attractive for pupils from both<br />
cooperating schools and even from other schools. Since 2010<br />
when the pr<strong>of</strong>ile was implemented, the number <strong>of</strong> pupils who<br />
change schools because <strong>of</strong> the <strong>of</strong>fered pr<strong>of</strong>ile, has increased.<br />
Both teachers explain the pr<strong>of</strong>ile’s success by their cooperation<br />
with external partners and the school projects:<br />
” Ohne Euch hätten wir das nie geschafft.“<br />
[Without you we would never have accomplished<br />
this.]<br />
” Unser Informatikpr<strong>of</strong>il ist erfolgreich, es kommt<br />
wie<strong>der</strong> eins zustande, auch Dank <strong>der</strong> tollen Angebote,<br />
die wir durch Euch machen können.“ [Our<br />
Pr<strong>of</strong>ilkurs is successful: we can <strong>of</strong>fer it again this<br />
year thanks to the great opportunities we can<br />
give due to your support.]<br />
” Für uns ist es wichtig, dass die Projekte klein<br />
genug sind für die Schule, aber dennoch mit praxisbezogenem<br />
Hintergrund. [Das Projekt] hat die<br />
Schüler unheimlich motiviert und es war in unser<br />
Semester eingebettet.“ [For us it is important<br />
that the school projects are small enough to be<br />
practical at school, but at the same time reflect<br />
a pr<strong>of</strong>essional background. The pupils were very<br />
motivated, and the project was well integrated in<br />
the Pr<strong>of</strong>ilkurs this school semester.]<br />
However, the success <strong>of</strong> school courses doesn’t end with<br />
the motivation <strong>of</strong> the pupils but with learning outcomes<br />
which have to meet educational standards.<br />
3.5 Phase 4: From input to outcome<br />
For the first three project courses our process <strong>of</strong> bringing<br />
contexts into the classroom essentially resembled a waterfall<br />
model. It was only after the decontextualization was<br />
completed that the recontextualization started: the school<br />
project took place not before end <strong>of</strong> term. Our students<br />
who developed context artifacts and teaching units had no<br />
second opportunity to improve their products.<br />
However, during the school projects we observed how the<br />
pupils got along with the Greenfoot scenarios and gained<br />
first-hand insights into how the context artifacts and the<br />
teaching units fitted into the classroom context. The teachers<br />
from the cooperation school who were temporarily replaced<br />
by our students also gave valuable feedback. Moreover<br />
the boys and girls developed their own ideas on how<br />
to improve the s<strong>of</strong>tware. Some <strong>of</strong> them even engaged in actively<br />
implementing new features and thus created context<br />
artifacts themselves. These results informed the requirements<br />
for our next project course.<br />
An additional aspect to be consi<strong>der</strong>ed in the model is the<br />
role <strong>of</strong> educational standards. In our school project educational<br />
standards have been playing a minor role. But by<br />
now a cooperation school had integrated some <strong>of</strong> our course<br />
material into their curriculum. Since current educational<br />
standards regulate content only to a minor degree and instead<br />
focus on learning outcome, it was perfectly possible for<br />
them to adapt our material to their courses. However in this<br />
116<br />
case courses need to be aligned to educational standards.<br />
Evaluation <strong>of</strong> phase 4.<br />
The evaluation <strong>of</strong> phases four and five is object <strong>of</strong> further<br />
research because it requires a comparison <strong>of</strong> learning outcomes<br />
<strong>of</strong> both traditional Informatics courses and Informaticsin-context<br />
courses, which is beyond our scope. But our<br />
school projects are already producing results. There are<br />
pupils who attended the first school projects in 2010 at the<br />
age <strong>of</strong> 14 and who now chose the Pr<strong>of</strong>ilkurs <strong>of</strong>fered by our<br />
cooperating school. And by now there are also pupils who<br />
successfully finished the Pr<strong>of</strong>ilkurs and enrolled for Informatics<br />
or engineering degree programs.<br />
3.6 Phase 5: Consolidation<br />
Within the current fifth phase we have consolidated the<br />
graphical representation <strong>of</strong> the suggested model in or<strong>der</strong> to<br />
communicate it to other researchers and practitioners. Furthermore<br />
we want to discuss both the relevance <strong>of</strong> the model<br />
and the choice <strong>of</strong> context. The current version is described<br />
in the following section and illustrated in figure 5.<br />
4. THE PROCESS MODEL<br />
In the following we will describe our process model for<br />
bringing contexts into the classroom along the lines <strong>of</strong> the<br />
distinction <strong>of</strong> the pre-educational and the educational stages<br />
(see 3.4). The model is shown in figure 5.<br />
4.1 Pre-Educational Stage<br />
At the pre-educational stage the real-world context is analyzed<br />
with an essentially informatical stance. This stage is<br />
characterized by the decontextualization <strong>of</strong> context artifacts<br />
from their original context.<br />
4.1.1 Decontextualization <strong>of</strong> Context Artifacts<br />
The simplest instance <strong>of</strong> a context artifact is when a physical<br />
artifact – e. g. a punch card – has lost its original meaning,<br />
yet is still associated with a context. We call these<br />
kind <strong>of</strong> context artifacts aboriginal or primary. Most artifacts<br />
cannot simply be removed from their original context<br />
though, and therefore must be mediated. Examples <strong>of</strong><br />
mediated or secondary context artifacts are photographs or<br />
videos. However in or<strong>der</strong> to un<strong>der</strong>stand the situated meaning<br />
<strong>of</strong> activities and artifacts in a complex system, they need<br />
to be fully decontextualized and reassembled. Instances for<br />
these tertiary newly designed context artifacts are models or<br />
simulations, and other objectifications.<br />
As described above (3.1.2) decontextualization, i. e. formalization<br />
and constructing artifacts, is usually one important<br />
aspect <strong>of</strong> s<strong>of</strong>tware development. The other is recontextualization,<br />
i. e. implementation <strong>of</strong> the artifacts into the<br />
context. In our model two contexts must be analyzed: the<br />
real-world context and the classroom context. Thus the preeducational<br />
stage is pre-educational in that educational purposes<br />
are assumed in the decontextualization <strong>of</strong> context artifacts.<br />
Yet for this activity an informatical stance might be<br />
predominant.<br />
4.1.2 Guiding Informatical Principles (GP)<br />
The decontextualization is guided by fundamental informatical<br />
principles because the artifacts are selected and developed<br />
from an informatical point <strong>of</strong> view. Guiding principles<br />
(GP) may be found in the Great Principles <strong>of</strong> Comput-
Real-world<br />
context<br />
GP<br />
Context<br />
Artifacts<br />
EP<br />
Classroom<br />
context<br />
Technological framework;<br />
tools and media<br />
Teaching<br />
units<br />
Learning<br />
outcomes<br />
Educational<br />
standards<br />
Figure 5: The process model for bringing real-world contexts into the classroom focuses on the design <strong>of</strong><br />
context artifacts and teaching units. In the process two stages can be identified: at the pre-educational<br />
stage context artifacts are decontextualized from their original context; at the educational stage they are<br />
recontextualized into the classroom context.<br />
ing [5] or Fundamental Ideas in Informatics [35] frameworks.<br />
The notion <strong>of</strong> Computational Thinking [45] might be helpful<br />
as well. At this stage the perspective should not be too<br />
narrow though, because the artifacts should be designed to<br />
translate many aspects <strong>of</strong> the real-world context into the<br />
classroom context and open up a broad image <strong>of</strong> Informatics.<br />
In figure 5 the guiding informatical principles are represented<br />
by triangles side by side with the educational principles.<br />
4.1.3 Technological Framework<br />
Often information and communication technology (ICT)<br />
are needed in or<strong>der</strong> to bring real-world contexts into the<br />
classroom. The technological framework provides the infrastructure,<br />
tools, media etc. needed for developing artifacts<br />
and recontextualizing them into the classroom context.<br />
Thus both the artifacts and the classroom context build on<br />
the technological platform. This belongs to both the preeducational<br />
stage and the educational stage.<br />
4.2 Educational Stage<br />
During the educational stage an educational stance is predominant.<br />
The guiding principles for the recontextualization<br />
<strong>of</strong> context artifacts into the classroom context therefore are<br />
first and foremost educational.<br />
4.2.1 Recontextualization <strong>of</strong> Context Artifacts<br />
Artifacts are meaningless unless they are contextualized.<br />
Thus recontextualizing formalized action or computational<br />
artifacts is as important for s<strong>of</strong>tware development as decontextualization.<br />
In our process, context artifacts are to be<br />
implemented into the classroom context. According to figure<br />
1 this requires designing exercises. That is by far not all.<br />
Learning objectives have to be defined; the artifacts must be<br />
adapted to the concrete classroom context and possibly augmented<br />
with additional material; the learning environment<br />
must be arranged around the artifacts; methods <strong>of</strong> instruc-<br />
117<br />
tion are to be selected and prepared etc.<br />
Recontextualized into the classroom context, context artifacts<br />
become part <strong>of</strong> the learning environment. Moreover<br />
they foster the students’ imagination and thus anchor the<br />
real-world context in the classroom context.<br />
4.2.2 Guiding Educational Principles (EP)<br />
The development <strong>of</strong> context artifacts is guided by informatical<br />
principles as well as educational principles (EP). In<br />
the present version <strong>of</strong> our model we distinguish between educational<br />
principles and educational standards. Educational<br />
standards regulate which learning outcomes students must<br />
meet. Principles guide how this is to be achieved. These<br />
include learning theories, teaching methodology, gen<strong>der</strong> and<br />
diversity aspects, etc.<br />
4.2.3 Outcome-oriented Standards and<br />
Learning Outcomes<br />
Current educational standards are outcome-oriented, i. e.<br />
they describe what the students must learn in terms <strong>of</strong> outcome<br />
like competencies. Since the outcome artifacts produced<br />
in the classroom should match the descriptions in the<br />
standards, the teaching units must be adjusted in such a<br />
way that they address the required competencies and skills.<br />
5. EXAMPLE: PREPARATION OF THE<br />
SCHOOL PROJECT<br />
In this section we will illustrate the application <strong>of</strong> the<br />
model described above. The context used in several <strong>of</strong> our<br />
project courses was baggage handling at airports. We started<br />
with an important airport component, the baggage handling<br />
system (BHS). Using the example <strong>of</strong> baggage handling, the<br />
different informatical principles, organizational as well as<br />
socio-technical concepts could be covered in teaching units.<br />
The initial difficulties with the BHS at Heathrow Airport’s<br />
Terminal 5 in 2008 gave a vivid example and were great<br />
motivation for both our students and the youngsters.
Figure 6: Two pupils presenting their conception <strong>of</strong><br />
airport baggage handling after they were introduced<br />
to the context by a YouTube video.<br />
Starting with the first stage <strong>of</strong> the proposed model, the<br />
pre-educational stage, the students began to analyze the<br />
context <strong>of</strong> baggage handling. First, they searched the Internet<br />
to find diverse artifacts describing the context. The students<br />
found pictures, videos and technical documents some<br />
<strong>of</strong> which could also be used as (secondary) context artifacts.<br />
Furthermore, they had the chance to visit an international<br />
airport and to inspect the baggage handling system on site.<br />
They took notes and additional pictures and had interviews<br />
with domain experts.<br />
Moreover, since we used the Greenfoot as a framework<br />
for the development <strong>of</strong> a “micro baggage handling world,”<br />
the students had to learn how to develop Greenfoot scenarios<br />
with pr<strong>of</strong>essional development tools 3 . For the school<br />
project the students implemented a Greenfoot simulation <strong>of</strong><br />
an airport baggage handling system. They modelled animated<br />
passengers who check in their luggage and go to their<br />
flights while their luggage is transported to the correct airplane.<br />
The context artifacts Greenfoot scenario, the interviews,<br />
photos, videos and further documents were the result<br />
<strong>of</strong> the first stage.<br />
At the educational stage the context artifacts and educational<br />
standards were used to develop teaching units. The<br />
developed teaching units consisted <strong>of</strong> a timetable, descriptions<br />
<strong>of</strong> basic definitions and concepts and different exercises.<br />
For example, the pupils were introduced to the context<br />
<strong>of</strong> airport baggage handling by videos and prosa text.<br />
After that they were to draw a schematic <strong>of</strong> the BHS (figure<br />
6).<br />
In another teaching unit the boys and girls started to investigate<br />
a Greenfoot simulation <strong>of</strong> a BHS, that intentionally<br />
exhibited some serious problems. After a few minutes the<br />
baggage cumulated on the conveyor (figure 7). The first exercise<br />
was to figure out the problem and to rearrange the<br />
layout <strong>of</strong> the conveyor system. This was done by drag-anddrop<br />
and interactively exploring the objects’ interfaces.<br />
3 As a didactical programming environment Greenfoot is not<br />
well suited for pr<strong>of</strong>essional teamwork or code refactoring.<br />
Therefore we use both Greenfoot and Eclipse for the development<br />
<strong>of</strong> the Greenfoot scenarios.<br />
118<br />
Figure 7: Greenfoot simulation <strong>of</strong> an airport baggage<br />
handling system, used in the school projects <strong>of</strong><br />
2010 [31]. In this scenario the BHS layout is flawed<br />
so that after a few minutes the conveyor system is<br />
jammed with luggage.<br />
In this section we have illustrated how we have used the<br />
model in or<strong>der</strong> to bring the context <strong>of</strong> baggage handling into<br />
the classroom. We have described the decontextualization <strong>of</strong><br />
context artifacts based on informatical principles and supported<br />
by the Java programming environment Greenfoot.<br />
In the second stage <strong>of</strong> the proposed model we have developed<br />
teaching units using the context artifacts. The designed<br />
teaching units have been brought into the classroom<br />
during different school projects. The feedback, we received<br />
from the pupils, were mainly positive because they enjoyed<br />
to get insight into complex real world context like baggage<br />
handling.<br />
6. DISCUSSION<br />
The proposed process model has been developed in a specific<br />
course setting with multiple subprojects. The different<br />
roles and the division <strong>of</strong> work established in our project<br />
are obviously reflected in our model. Thus our model may<br />
not hold universally. Nevertheless we argue that the division<br />
<strong>of</strong> work in our project results from the fact that the<br />
different contexts involved are essentially separated in the<br />
first place. After the discussion <strong>of</strong> the our methodology, we<br />
therefore consi<strong>der</strong> other possible constellations and a fundamental<br />
dilemma for Informatics teachers – especially at<br />
school.<br />
6.1 Model Validity<br />
Although the process model for bringing contexts into the<br />
classroom is so abstract that it can only serve as a reference<br />
framework, there is good reason to challenge it. In our<br />
projects it has been proven to be viable, but further inquiries<br />
may remain necessary. Nevertheless we claim ‘validity by design’<br />
for our model arguing along the following principles for<br />
design-based research in IS, that were introduced by Hevner<br />
et al. [17]:<br />
Design as an Artifact. The goal <strong>of</strong> design-based approaches<br />
is the creation <strong>of</strong> appropriate and reliable artifacts,
i. e. theories, models, frameworks, technology or tools. In<br />
our research project it is the process model for bringing contexts<br />
into the classroom.<br />
Problem relevance. A design-based research project must<br />
deliver a solution to a relevant and significant problem. In<br />
the complexity and amount <strong>of</strong> task during analysis <strong>of</strong> realworld<br />
contexts, the translation into the classroom context,<br />
as well as in the additional workload for teachers we see the<br />
major problems <strong>of</strong> CS in context. At the moment there is<br />
little research addressing these problems. Our model doesn’t<br />
provide a solution to them, but it can help identifying them<br />
(see 6.2 and 6.3).<br />
Design evaluation. The relevance, quality and viability<br />
<strong>of</strong> designed artifacts have to be shown via elaborate evaluation<br />
methods. We tested our process model during each<br />
iteration in at least two classroom contexts: at least one<br />
school project and the project course itself which can also<br />
be described by the model. Here the context <strong>of</strong> the school<br />
project was the ‘real-world’ context. Furthermore we found<br />
our model to be coherent with other proposed models (e. g.<br />
[6]) as well as practical (e. g. anchored instruction [19]) and<br />
theoretical (e. g. boundary objects [37]) conceptions. The<br />
evaluations <strong>of</strong> each iterations produced new aspects, which<br />
were addressed in the next iteration.<br />
Research Contribution. A concrete contribution to educational<br />
research has to be provided by the design-science<br />
research. Our contribution is a process model which serves<br />
as a reference framework with its own terminology based on<br />
approaches to s<strong>of</strong>tware development.<br />
Research rigor. A very important aspect <strong>of</strong> design science<br />
is “the application <strong>of</strong> rigorous methods in both the construction<br />
and evaluation <strong>of</strong> the designed artifact” (Hevner et<br />
al. [17], p. 87). In our design we used a methodology similar<br />
to s<strong>of</strong>tware development methods. We also used prospective<br />
and reflective approaches supporting our research process<br />
[29]. In the current phase we want to communicate our<br />
model to other researchers and practitioners and, given the<br />
model, to survey their own experience.<br />
Design as a Search Process. Design science must be conducted<br />
via an iterative process with construction and evaluation<br />
periods in or<strong>der</strong> to mature [36]. In our research<br />
project, four consecutive research phases (both construction<br />
and evaluation) have been accomplished. Currently a fifth<br />
iteration is ongoing.<br />
Communication <strong>of</strong> results. The results will be communicated<br />
to practitioners and researchers enabling to benefit<br />
from the constructed artifacts as well as to discuss and evaluate<br />
the results. On the one hand, we have introduced our<br />
proposed process model to students and teachers, on the<br />
other hand, we will bring the model to the research community<br />
via the present publication.<br />
6.2 Contexts<br />
As described above, the ‘real-world’ context normally is<br />
different from classroom context. Therefore new challenges<br />
for teachers and Computing education experts occur in the<br />
decontextualization and recontextualization processes <strong>of</strong> Informatics-in-context<br />
course development and preparation.<br />
But how can an appropriate real-world context be chosen?<br />
First, the real-world context should be in the same sociocultural<br />
frame as the classroom context. Otherwise too<br />
much time and effort must be spent on un<strong>der</strong>standing the<br />
context instead <strong>of</strong> learning Informatics, and the context might<br />
119<br />
turn out to be unattractive for the pupils. For example<br />
in low-income neighborhoods baggage handling systems or<br />
inventory control systems might be unattractive contexts.<br />
Also in rural areas information systems in animal husbandry<br />
might be more motivating than the baggage handling system.<br />
Second, contexts can arise from Informatics itself. For<br />
example Informatics contexts like artificial intelligence or<br />
compiler construction provide rich contexts <strong>of</strong>fering the possibility<br />
to include many Informatics aspects and principles<br />
(provided that the first principle is taken into consi<strong>der</strong>ation<br />
and the students un<strong>der</strong>stand the practical purposes).<br />
Finally, <strong>of</strong>ten technological frameworks like specific hardware<br />
or tools constitute contexts that are attractive at least<br />
for technophiles. For example robotics classes <strong>of</strong>ten don’t<br />
get beyond playful experimentation toward serious application<br />
contexts.<br />
6.3 CS Teachers’ Working Conditions<br />
As outlined by Diethelm et al. [7], the teachers’ role and<br />
demands on CS teachers have changed in the last years.<br />
In addition to teaching, new contexts have to be analyzed,<br />
teaching units must be designed and transformed for the<br />
actual classroom setting as well as the outcome has to be<br />
evaluated. All tasks are highly communicative and cooperative<br />
and require a set <strong>of</strong> various skills and competencies CS<br />
teachers do not always have [7].<br />
The pre-educational examination <strong>of</strong> contexts is a timeconsuming<br />
endeavor that requires a broad background in<br />
Informatics and the real-world context. Most teachers don’t<br />
have time and background to develop contexts for Informaticsin-context<br />
courses.<br />
These kind <strong>of</strong> problems are not new and have already<br />
been discussed in German <strong>Didaktik</strong> discourse in the 1950s<br />
by Roth [34] and Klafki [21]:<br />
“We believe that it would be demanding too<br />
much <strong>of</strong> teachers in terms <strong>of</strong> time and mental<br />
energy to expect them to ‘rationalize’ about the<br />
contents in a pre-pedagogical context [or stance]<br />
whenever they set out to prepare themselves for<br />
teaching. This would involve, for example, adopting<br />
the role <strong>of</strong> a scientist who sees the contents in<br />
question as a research exercise in a specific field.<br />
We are <strong>of</strong> the opinion that this applies not only to<br />
teachers at primary, junior secondary and vocational<br />
level, but also to those at senior secondary<br />
level!” (Klafki, [22], p. 17)<br />
Although the problems remain the same today’s schools<br />
look different from the Volksschulen <strong>of</strong> the 1950s. Teachers<br />
work in teams and can collaborate even if they are distributed<br />
all over the globe.<br />
Our supposed model tries to support teachers in their<br />
work. It names necessary steps and requirements for bringing<br />
real world contexts into the classroom as well as introduces<br />
a pre-educational stages allowing to “outsource” certain<br />
steps (esp. analysis <strong>of</strong> contexts) to other experts. Thus<br />
we hope to strengthen the link between theory and practice<br />
in teaching Informatics in context.<br />
7. CONCLUSION<br />
Learning always is situated in contexts. But usually the<br />
classroom context is not the ‘real-world’ context where ac-
tivities, artifacts, and concepts have a situated meaning implying<br />
real consequences. In the classroom things bear a<br />
special meaning according to the ‘hidden curriculum’. Hence<br />
teaching is a performative act in that it constitutes and reinforces<br />
its own prerequisites. Moreover teaching in context<br />
requires the classroom context to resemble aspects <strong>of</strong> a realworld<br />
context but is still relying on both the teacher’s and<br />
the students’ imagination and play-acting. Therefore the<br />
development, planning and preparation <strong>of</strong> context-oriented<br />
teaching units is an intricate and time-comsuming endeavor.<br />
The ceiling <strong>of</strong> the Great Hall at Hogwarts was supposedly<br />
intentionally designed to bring the outside world into the<br />
school. Unlike the teachers at Hogwarts, Informatics teachers<br />
at our schools cannot make use <strong>of</strong> magic in or<strong>der</strong> to bring<br />
the outside world into the classroom. They can, however,<br />
use computing technology and develop context artifacts so<br />
as to arrange the learning environment around them.<br />
Though our process model for bringing contexts into the<br />
classroom emerged from a special project constellation, it<br />
reveals that contextualized teaching requires not only more<br />
careful preparation but also some preparation steps that<br />
possibly are beyond the means <strong>of</strong> most teachers. Educators<br />
and most notably primary and secondary school teachers<br />
who want to contextualize their teaching would therefore be<br />
well advised to draw on the strengths <strong>of</strong> their institution<br />
and to take advantage <strong>of</strong> cooperations with other educational<br />
and research facilities as well as local enterprises.<br />
8. REFERENCES<br />
[1] T. An<strong>der</strong>son and J. Shattuck. Design-based research:<br />
A decade <strong>of</strong> progress in education research?<br />
Educational Researcher, 41(1):16–25, 2012.<br />
[2] K. Beck et al. Manifesto for agile s<strong>of</strong>tware<br />
development. URL http://agilemanifesto.org/<br />
(08/2012), 2001.<br />
[3] A. Collins, D. Joseph, and K. Bielaczyc. Design<br />
research: Theoretical and methodological issues. The<br />
Journal <strong>of</strong> the Learning Sciences, 13(1):15–42, 2004.<br />
[4] S. Cooper and S. Cunningham. Teaching computer<br />
science in context. ACM Inroads, 1(1):5–8, 2010.<br />
[5] P. J. Denning and C. Martell. Great principles <strong>of</strong><br />
computing; <strong>of</strong>ficial web site, 2005 ff. URL<br />
http://cs.gmu.edu/cne/pjd/GP/ (April 2012).<br />
[6] I. Diethelm and C. Dörge. From context to<br />
competencies. In N. Reynolds and<br />
M. Turcsányi-Szabó, editors, Key Competencies in the<br />
Knowledge Society, IFIP TC 3 International<br />
Conference, KCKS 2010, pages 67–77, Berlin, 2010.<br />
Springer.<br />
[7] I. Diethelm, C. Hildebrandt, and L. Krekeler.<br />
Implementation <strong>of</strong> Computer Science in Context – a<br />
research perspective regarding teacher-training. In<br />
Koli Calling 2009 : 9th International Conference on<br />
Computing Education Research, pages 93–99, 2010.<br />
[8] I. Diethelm, J. Koubek, and H. Witten. IniK –<br />
Informatik im Kontext : Entwicklungen, Merkmale<br />
und Perspektiven. LOG IN, (169/170):97–105, 2011.<br />
[9] E. W. Dijkstra. On the cruelty <strong>of</strong> really teaching<br />
computing science [a debate on teaching computer<br />
science]. Communications <strong>of</strong> the ACM,<br />
32(12):1397–1414, 1989.<br />
[10] D. Engbring and A. Pasternak. Einige anmerkungen<br />
120<br />
zum begriff inik. In I. Diethelm, C. Dörge,<br />
C. Hildebrandt, and C. Schulte, editors, <strong>Didaktik</strong> <strong>der</strong><br />
Informatik : Möglichkeiten empirischer<br />
Forschungsmethoden und Perspektiven <strong>der</strong><br />
Fachdidaktik [6. Workshop <strong>der</strong> GI-Fachgruppe DDI],<br />
number P-168 in Lecture Notes in Informatics (LNI) –<br />
Proceedings, pages 119–124, Bonn, 2010. Gesellschaft<br />
für Informatik.<br />
[11] S. Fincher, M. Kölling, I. Utting, N. Brown, and<br />
P. Stevens. Repositories <strong>of</strong> teaching material and<br />
communities <strong>of</strong> use: Nifty assignments and the<br />
Greenroom. In ICER ’10: Proceedings <strong>of</strong> the 6th<br />
International Computing Education Research<br />
Workshop, pages 107–114, 2010.<br />
[12] C. Floyd. Steps – a methodical approach to pd<br />
[paricipatory design]. Communications <strong>of</strong> the ACM,<br />
36(4):83, June 1993.<br />
[13] C. Floyd, W.-M. Mehl, F.-M. Reisin, G. Schmidt, and<br />
G. Wolf. Out <strong>of</strong> scandinavia: Alternative approaches<br />
to s<strong>of</strong>tware design and system development.<br />
Human-Computer Interaction, 4(4):253–350, 1989.<br />
[14] C. Floyd, F.-M. Reisin, and G. Schmidt. Steps to<br />
s<strong>of</strong>tware development with users. In Proceedings <strong>of</strong> the<br />
2nd European S<strong>of</strong>tware Engineering Conference, ESEC<br />
’89, pages 48–64, London, UK, 1989. Springer-Verlag.<br />
[15] M. Guzdial. Does contextualized computing education<br />
help? ACM Inroads, 1(4):4–6, 2010.<br />
[16] P. Henriksen and M. Kölling. Greenfoot: Combining<br />
object visualisation with interaction. In OOPSLA ’04:<br />
Proceedings <strong>of</strong> the 19th Annual ACM Conference on<br />
Object-Oriented Programming, Systems, Languages,<br />
and Applications, pages 73–82, 2004.<br />
[17] A. R. Hevner, S. T. March, J. Park, and S. Ram.<br />
Design science in information systems research. MIS<br />
Quarterly, 28(1):75–105, 2004.<br />
[18] W. Jank and H. Meyer. Didaktische Modelle.<br />
Cornelsen, Berlin, 10th edition, 2002.<br />
[19] M. Kariuki and M. Duran. Using anchored instruction<br />
to teach preservice teachers to integrate technology in<br />
the curriculum. Journal <strong>of</strong> Technology and Teacher<br />
Education, 12(3):431–445, 2004.<br />
[20] A. E. Kelly. Design research in education: Yes, but is<br />
it methodological? The Journal <strong>of</strong> the Learning<br />
Sciences, 13(1):115–128, 2004.<br />
[21] W. Klafki. Didaktische Analyse als Kern <strong>der</strong><br />
Unterrichtsvorbereitung. Die Deutsche Schule,<br />
50(10):450–471, 1958.<br />
[22] W. Klafki. Didactic analysis as the core <strong>of</strong> preparation<br />
<strong>of</strong> instruction. Journal <strong>of</strong> Curriculum Studies,<br />
27(1):13–30, 1995. Translated and abridged version <strong>of</strong><br />
[21].<br />
[23] M. <strong>Knobelsdorf</strong> and C. Schulte. Computer science in<br />
context – pathways to computer science. In Koli<br />
Calling 2007 – Proceedings <strong>of</strong> the Seventh Baltic Sea<br />
Conference on Computing Education Research, pages<br />
65–76, 2008.<br />
[24] J. Koubek, C. Schulte, P. Schulze, and H. Witten.<br />
Informatik im kontext (inik) : Ein integratives<br />
unterrichtskonzept für den informatikunterricht. In<br />
B. Koerber, editor, Zukunft braucht Herkunft : 25<br />
Jahre ” INFOS – Informatik und Schule“ [INFOS<br />
2009, 13. GI-Fachtagung Informatik und Schule],
number P-156 in Lecture Notes in Informatics (LNI) –<br />
Proceedings, pages 268–279, Bonn, 2009. Gesellschaft<br />
für Informatik.<br />
[25] D. Krause, A. Rolf, M. Christ, and E. Simon. Wissen,<br />
wie alles zusammenhängt. Informatik-Spektrum,<br />
29(4):263–273, 2006.<br />
[26] M. Kölling. The greenfoot programming environment.<br />
ACM Transactions on Computing Education,<br />
10(4):14:1–14:21, 2010.<br />
[27] M. Kölling. Introduction to Programming with<br />
Greenfoot : Object-Oriented Programming in Java with<br />
Games and Simulations. Pearson, Boston, MA, 2010.<br />
[28] R. Künzli. German didaktik: Models <strong>of</strong><br />
re-presentation, <strong>of</strong> intercourse, and <strong>of</strong> experience. In<br />
I. Westbury et al., editors, Teaching as a Reflective<br />
Practice, pages 41–54. Lawrence Erlbaum, Mahwah,<br />
NJ, 1999.<br />
[29] Y. Mor and N. Winters. Design approaches in<br />
technology-enhanced learning. Interactive Learning<br />
Environments, 15(1):61–75, 2007.<br />
[30] K. Nygaard. A process-oriented conceptual platform<br />
for informatics. URL<br />
http://hci.stanford.edu/courses/cs547/abstracts/01-<br />
02/011019-nygaard.html (July 2012),<br />
2001.<br />
[31] D. Rick, J. Ludwig, S. Meyer, C. Reh<strong>der</strong>, and<br />
I. Schirmer. Introduction to business informatics with<br />
greenfoot using the example <strong>of</strong> airport baggage<br />
handling. In Koli Calling ’10: Proceedings <strong>of</strong> the 10th<br />
Koli Calling International Conference on Computing<br />
Education Research, pages 68–69, 2010.<br />
[32] D. Rick and I. Schirmer. ” IT-Aneignung, Gen<strong>der</strong> und<br />
Diversity“: Ein transdisziplinäres Projektseminar für<br />
Informatik-Studierende. In M. Thomas and<br />
M. Weigend, editors, Informatik und Kultur : 4.<br />
Münsteraner Workshop zur Schulinformatik, pages<br />
64–77, Münster, <strong>Germany</strong>, 2010. Zentrum für<br />
Lehrerbildung. URL http://miami.unimuenster.de/servlets/DocumentServlet?id=5304<br />
(June<br />
2010).<br />
[33] A. Rolf. Mikropolis 2010 : Menschen, Computer,<br />
Internet in <strong>der</strong> globalen Gesellschaft.<br />
Metropolis-Verlag, Marburg, 2008.<br />
[34] H. Roth. Pädagogische Psychologie des Lehrens und<br />
Lernens. Schroedel, Hannover, 7th edition, 1963<br />
[1957].<br />
121<br />
[35] A. Schwill. Computer science education based on<br />
fundamental ideas. In Proceedings <strong>of</strong> the IFIP TC3<br />
WG3.1/3.5 joint working conference on Information<br />
technology : supporting change through teacher<br />
education: supporting change through teacher<br />
education, pages 285–291. Chapman & Hall, Ltd.,<br />
1997.<br />
[36] H. A. Simon. The Sciences <strong>of</strong> the Artificial.<br />
MIT Press, Cambridge, MA, first edition, 1969.<br />
[37] S. L. Star and J. R. Griesemer. Institutional ecology,<br />
‘translations’ and boundary objects: Amateurs and<br />
pr<strong>of</strong>essionals in berkeley’s museum <strong>of</strong> vertebrate<br />
zoology, 1907-39. Social Studies <strong>of</strong> Science,<br />
19(3):387–420, 1989.<br />
[38] L. A. Suchman. Plans and situated actions : the<br />
problem <strong>of</strong> human-machine communication.<br />
Cambridge <strong>University</strong> Press, 1987.<br />
[39] A. E. Tew, B. Dorn, W. D. Leahy, Jr., and<br />
M. Guzdial. Context as support for learning computer<br />
organization. ACM Journal on Educational Resources<br />
in Computing, 8(3):8:1–8:18, 2008.<br />
[40] The Design-Based Research Collective. Design-based<br />
research: An emerging paradigm for educational<br />
inquiry. Educational Researcher, 32(1):5–8, 2003.<br />
[41] The Greenfoot Team. Greenfoot. Free download,<br />
2006ff. URL http://www.greenfoot.org/ (April 2012).<br />
[42] The IniK Group. Informatik im Kontext – IniK für<br />
alle. IniK Website, 2007ff. URL<br />
http://www.informatik-im-kontext.de/ (July 2012).<br />
[43] I. Utting, S. Cooper, M. Kölling, J. Maloney, and<br />
M. Resnick. Alice, greenfoot, and scratch – a<br />
discussion. Transactions on Computing Education,<br />
10(4):17:1–17:11, November 2010.<br />
DOI: 10.1145/1868358.1868364.<br />
[44] J.-H. Wah<strong>of</strong>f, J. P. de Albuquerque, and A. Rolf. The<br />
mikropolis model: A framework for transdisciplinary<br />
research <strong>of</strong> information systems in society. In Y. K.<br />
Dwivedi, M. R. Wade, and S. L. Schneberger, editors,<br />
Information Systems Theory, volume 29 <strong>of</strong> Integrated<br />
Series in Information Systems, pages 367–386.<br />
Springer New York, 2012.<br />
[45] J. M. Wing. Computational thinking.<br />
Communications <strong>of</strong> the ACM, 49(3):33–35, 2006.
Email for You (only?) – Design and Implementation<br />
<strong>of</strong> a Context-based Learning Process on<br />
Internetworking and Cryptography<br />
Andreas Gramm<br />
1. Schulpraktisches Seminar<br />
Charlottenburg-Wilmersdorf<br />
Otto-Suhr-Allee 100<br />
10585 Berlin, <strong>Germany</strong><br />
+49 (0)30 90 29 13 280<br />
gramm@gymnasium-tiergarten.de<br />
ABSTRACT<br />
The didactical approach <strong>of</strong> teaching computer science in context<br />
aims at enabling learners to un<strong>der</strong>stand concepts <strong>of</strong> computer<br />
science better through the help <strong>of</strong> concrete illustration and<br />
meaning. This paper describes a learning arrangement in which<br />
students in lower secondary school education are motivated to<br />
engage with cryptographic algorithms ranging from Caesar to<br />
RSA by making them discover the challenges <strong>of</strong> a private and<br />
trustable communication over public networks. We also describe<br />
experiences we made by developing and testing the context-based<br />
learning process in several classes <strong>of</strong> different age. While<br />
designing and implementing context-based teaching material<br />
proved to be demanding, we were rewarded with a high level <strong>of</strong><br />
both interest and un<strong>der</strong>standing on the part <strong>of</strong> the students<br />
suggesting that context-based learning will prove a promising<br />
didactical tool for computer science teachers.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computer and Information Science Education]: Computer<br />
science education<br />
General Terms<br />
Algorithms, Reliability, Security, Human Factors.<br />
Keywords<br />
Context-based computer science education, computer science in<br />
context, IniK, communication protocol, email protocol,<br />
communication security, cryptology, RSA<br />
1. COMPUTER SCIENCE IN CONTEXT<br />
In the light <strong>of</strong> new insights into learning through neurobiological<br />
research and with regard to a constructivist un<strong>der</strong>standing <strong>of</strong><br />
learning, it has recently been argued that while educators wish to<br />
implement largely transferable, abstract knowledge, concrete<br />
contexts are necessary for learners to un<strong>der</strong>stand the nature,<br />
meaning and use <strong>of</strong> an abstract idea or concept. Cooper and<br />
Cunningham showed how this approach has been applied to the<br />
design <strong>of</strong> introductory courses at university level using 3D<br />
animations (Alice), media computation, computer graphics or<br />
robotics as contexts [6]. According to their analysis, suitable<br />
Malte Hornung<br />
Freie Universität Berlin<br />
Didactics <strong>of</strong> Informatics<br />
Königin-Luise-Straße 24-26<br />
14195 Berlin, <strong>Germany</strong><br />
+49 (0)30 83 87 51 87<br />
malte.hornung@fu-berlin.de<br />
122<br />
Helmut Witten<br />
Gesellschaft für Informatik (GI)<br />
Working group “CS education in<br />
Berlin and Brandenburg (IBBB)”<br />
Brandenburgische Straße 23<br />
10797 Berlin, <strong>Germany</strong><br />
+49 (0)30 88 68 25 69<br />
helmut@witten-berlin.de<br />
contexts are a source <strong>of</strong> illustrative examples, project ideas,<br />
motivation and meaning. Guzdial points out that the use <strong>of</strong><br />
contexts improves retention <strong>of</strong> students: Students are more likely<br />
to continue CS studies when they un<strong>der</strong>stand the usefulness and<br />
value <strong>of</strong> what they are learning [17]. But he also warns that<br />
learners might overgeneralise a concept when only seen in a<br />
single context. He therefore calls for concepts to be presented in<br />
various different contexts to allow for an adequate<br />
decontextualisation.<br />
While the international debate <strong>of</strong> learning CS in context primarily<br />
discusses introductory courses to computer science at universities,<br />
the discussion in <strong>Germany</strong> focuses on secondary school<br />
education. An open project group called “Informatik im Kontext –<br />
IniK” (translates to “computer science in context”) was established<br />
to develop and promote the approach <strong>of</strong> context-based<br />
computer science secondary school education. The project’s<br />
website [21] describes the concept <strong>of</strong> IniK and provides teaching<br />
material and didactic instructions for teachers for a number <strong>of</strong><br />
contexts and possible teaching units.<br />
IniK is partly modelled on projects promoting context-based<br />
learning <strong>of</strong> sciences (chemistry, physics and biology), which have<br />
received substantial funding from both the Fe<strong>der</strong>al Government <strong>of</strong><br />
<strong>Germany</strong> as well as the different fe<strong>der</strong>al states (known as<br />
Bundeslän<strong>der</strong>). Out <strong>of</strong> these three projects “Chemie im Kontext -<br />
ChiK” [19] was the first and has been the project most influential<br />
for IniK (cf. [10]). In contrast to the three science projects, IniK is<br />
more <strong>of</strong> the character <strong>of</strong> a grass root development with almost no<br />
funding. To our knowledge, Berlin is the only fe<strong>der</strong>al state to<br />
grant a small amount <strong>of</strong> resources to a group <strong>of</strong> teachers<br />
committed to the IniK approach. Apart from teachers interested in<br />
mo<strong>der</strong>n teaching and learning approaches it is mainly pr<strong>of</strong>essors<br />
and students from several universities who further develop the<br />
ideas and concepts <strong>of</strong> IniK.<br />
While the idea <strong>of</strong> a context-based approach to computer science<br />
has been discussed in <strong>Germany</strong> before (see e.g. [12] and [7]), the<br />
IniK approach was first fully described in 2009 by Koubek,<br />
Schulte, Schulze and Witten in [24]. They identify three<br />
prerequisites for a learning process to qualify as being contextbased<br />
in the field <strong>of</strong> secondary school computer science<br />
education:
1. Context-based education in computer science makes use<br />
<strong>of</strong> a context that is relevant to students. The context is a<br />
concrete situation, in which aspects <strong>of</strong> different<br />
dimensions are significant to un<strong>der</strong>stand the situation<br />
and to find adequate solutions.<br />
2. The competencies acquired through the learning process<br />
are compliant with standards for teaching computer<br />
science in secondary school education. The Society for<br />
Informatics in <strong>Germany</strong> Gesellschaft für Informatik (GI)<br />
has published principles and minimal standards for<br />
computer science in lower secondary school education<br />
[1] (for an English summary <strong>of</strong> these standards see [4]<br />
and [25]). Referring to the definition <strong>of</strong> Weinert, the<br />
competencies described in the standards are to be<br />
un<strong>der</strong>stood as “the cognitive abilities and skills<br />
possessed by or able to be learned by individuals that<br />
enable them to solve particular problems, as well as the<br />
motivational, volitional and social readiness and capacity<br />
to use the solutions successfully and responsibly in<br />
variable situations” ([20], p. 65).<br />
3. Context-based education in computer science shows a<br />
variety <strong>of</strong> teaching and learning methods to activate<br />
different types <strong>of</strong> learners and enhance cooperative<br />
student-student interaction in class.<br />
Since the founding <strong>of</strong> IniK, several further ideas on how to design<br />
and implement context-based learning <strong>of</strong> computer science have<br />
been developed. Diethelm, Borowski and Weber suggested a way<br />
to find contexts which are relevant to learners by indirectly asking<br />
them what interests they have in a certain informatics devices [8].<br />
Furthermore, Diethelm and Dörge described a way to develop<br />
competencies using a context [9]. In 2011, Diethelm, Koubek and<br />
Witten summarised the development, features and perspectives <strong>of</strong><br />
IniK and state criteria for the selection <strong>of</strong> suitable contexts and as<br />
to how units could best be documented [11].<br />
The goal <strong>of</strong> the IniK-working group was to design several<br />
context-based learning environments with a high permeability to<br />
everyday school practice and reflect on its implementation in<br />
secondary schools both to prove the feasibility <strong>of</strong> the concept to<br />
other CS teachers as well as to gain experience and further<br />
develop the concept. The method used for design and reflection<br />
has an alignment to research methods like Action Research [2]<br />
and Design Based Research [3]. Note that therefore this report’s<br />
conclusions are mostly based on in-service reflection, not on a<br />
systematic empirical approach. “Email for You (only?)” is the title<br />
<strong>of</strong> a context-based learning environment on internetworking and<br />
cryptography which is presented and discussed in this report.<br />
2. DESIGN AND IMPLEMENTATION OF<br />
THE LERANING PROCESS<br />
We have developed a context-based learning process on security<br />
issues <strong>of</strong> internetwork communication (for a more detailed<br />
documentation <strong>of</strong> the learning process in German see [22], [16],<br />
[14] and [13]). In doing so, we intended to show the feasibility <strong>of</strong><br />
the IniK concept for a context-based learning <strong>of</strong> computer science<br />
in everyday computer science classrooms as well as to gather<br />
experience that can be helpful for the further development <strong>of</strong> the<br />
concept and other context-based learning processes.<br />
The indented age group is from year 9 to year 12 (puplis aged 14<br />
to 18). From prior experience we can say that students <strong>of</strong>ten enjoy<br />
123<br />
the topic cryptography, because transmitting secrets is associated<br />
with adventure and decrypting or even cracking a cipher<br />
resembles a strategic game. However, just discussing cryptographic<br />
algorithms such as Caesar doesn’t help raising awareness<br />
for the fact, that we as users make use, or at least are well-advised<br />
to make use <strong>of</strong> cryptography whenever communicating over<br />
public networks such as the Internet. Most students communicate<br />
over the Internet on a daily basis, be it via instant messengers,<br />
social networks or email. The recent annual survey among<br />
German teenagers “Jugend, Information, (Multi-) Media (JIM)<br />
2011” [18] shows that they spend almost half <strong>of</strong> their online time<br />
communicating. The context <strong>of</strong> security issues in private communication<br />
over public networks is therefore part <strong>of</strong> students’ daily<br />
life, only that in most cases they are not aware <strong>of</strong> it. We believe<br />
that rising awareness <strong>of</strong> both advantages and challenges <strong>of</strong> using<br />
computer systems is a central aim <strong>of</strong> secondary school computer<br />
science education.<br />
The context can be explored with regard to different dimensions.<br />
On the one hand, assessing security risks requires a thorough<br />
un<strong>der</strong>standing <strong>of</strong> the un<strong>der</strong>lying network technology and the<br />
communication protocols employed. On the other hand, security<br />
always requires additional resources. In the case <strong>of</strong> private<br />
communication, it requires the exchange <strong>of</strong> public keys. Today,<br />
many users don’t use encryption even though the technology is<br />
available for free. This clearly indicates to human factors such as<br />
too little awareness <strong>of</strong> security issues and possible consequences.<br />
We structured the learning process by five questions:<br />
1. How is an email communicated from my computer to<br />
the computer <strong>of</strong> the addressee?<br />
2. What dangers challenge a private communication over a<br />
public network, such as the Internet?<br />
3. How can I achieve privacy?<br />
4. How can I assure the integrity <strong>of</strong> a message and the<br />
authenticity <strong>of</strong> a sen<strong>der</strong>?<br />
5. Why should I communicate privately?<br />
The answers to questions 2 to 4 respectively build on the<br />
knowledge students acquire to answer the questions asked before.<br />
This way, they shall experience this knowledge as being useful to<br />
solve tasks such as providing security in an unsafe environment.<br />
The last question could alternatively be asked at the beginning <strong>of</strong><br />
the learning process, but in our view it is more rewarding to<br />
discuss it when students have got clear ideas <strong>of</strong> the challenges <strong>of</strong><br />
communication over public networks, which they might not have<br />
at the beginning <strong>of</strong> the learning process. It also seems more<br />
attractive to us that students discover the possible dangers in a<br />
hands-on activity rather than in an academic discussion.<br />
Of course, showing students ways to disturb email communication<br />
raises ethical questions. In fact, the Bundesverfassungsgericht<br />
(<strong>Germany</strong>’s High Court) has discussed the question whether using<br />
a network analysing tool is in breach <strong>of</strong> German law. The court<br />
decided that the law in question only applies when such tools are<br />
used with malevolent intentions. We prefer that our students know<br />
about existing dangers and take protective measures rather than<br />
leaving this knowledge to possibly more malevolent experts who<br />
might use their lack <strong>of</strong> knowledge to infect their computer with<br />
malware. However, we took the discussion as an occasion to look<br />
for a tool which restricts the accessed network traffic to those
packets that are directly addressed to the client computer, which<br />
suits our purposes sufficiently.<br />
Table 1 shows the un<strong>der</strong>lying model <strong>of</strong> competencies on three<br />
different levels. It is up to teachers, the age <strong>of</strong> the learners and the<br />
time available to decide on how deeply the insights into security<br />
provided by cryptography should be developed. It is possible to<br />
just do parts <strong>of</strong> the activities in one year and continue with a more<br />
pr<strong>of</strong>ound examination <strong>of</strong> e.g. the RSA key generation algorithm in<br />
a following year. By pointing out these levels <strong>of</strong> competency, we<br />
enable teachers to adjust the material to their needs more flexibly.<br />
Table 1. Competency model for the un<strong>der</strong>standing <strong>of</strong> the level<br />
<strong>of</strong> security provided for communication by cryptography.<br />
Level 1 Students describe dangers <strong>of</strong> communicating via<br />
public networks (eavesdropping, manipulation <strong>of</strong><br />
messages, false sen<strong>der</strong>’s identity) and state<br />
prerequisites for a secure communication (privacy,<br />
integrity <strong>of</strong> the message, authenticity <strong>of</strong> the sen<strong>der</strong>).<br />
Students use tools to create a pair <strong>of</strong> keys, exchange<br />
public keys, encrypt and digitally sign emails and<br />
verify incoming emails.<br />
Students explain on a general level how a computer<br />
verifies the integrity <strong>of</strong> a message and the<br />
authenticity <strong>of</strong> its sen<strong>der</strong>.<br />
Level 2 Students assess the level <strong>of</strong> security provided by<br />
RSA using keys <strong>of</strong> different length based on<br />
computer experiments for a reconstruction <strong>of</strong> a<br />
private key and an internet research on the RSA<br />
Challenge.<br />
Level 3 Students create pairs <strong>of</strong> keys using the RSA key<br />
generation algorithm with small prime numbers and<br />
use them to encrypt and decrypt messages manually.<br />
Students reconstruct private keys with small prime<br />
numbers manually, explore algorithms to find large<br />
prime numbers and define the relation between the<br />
length <strong>of</strong> keys and the level <strong>of</strong> security provided<br />
using the problem to factor large semi-prime<br />
numbers as a mathematical argument. They assess<br />
the level <strong>of</strong> security provided by RSA using keys <strong>of</strong><br />
different length based on mathematical reasoning.<br />
In the German documentation <strong>of</strong> the learning process, links to<br />
relevant sections <strong>of</strong> the GI standards for teaching computer<br />
science are provided for each section <strong>of</strong> the learning process.<br />
Relevant competencies are mostly from the content standards<br />
informatics systems and informatics, man and society as well as<br />
the process standards reason and evaluate, for example:<br />
- Students un<strong>der</strong>stand the basic structure and functionality <strong>of</strong><br />
informatics systems.<br />
- Students react appropriately to risks arising from the use <strong>of</strong><br />
informatics systems.<br />
- Students ask questions and state hypotheses on matters <strong>of</strong><br />
informatics.<br />
- Students measure different criteria and assess their adequacy<br />
for their own actions.<br />
124<br />
Relevant links to the K–12 Computer Science Standards<br />
published by the Computer Science Teachers Association (CSTA)<br />
in 2011 [5] are not pointed out but can be easily found, for<br />
example:<br />
- Students explain the multiple levels <strong>of</strong> hardware and<br />
s<strong>of</strong>tware that support program execution (e.g., […]<br />
networks), […] describe how the Internet facilitates global<br />
communication. [Computers and Communications Devices]<br />
- Students exhibit legal and ethical behaviors when using<br />
information and technology and discuss the consequences <strong>of</strong><br />
misuse, analyze the positive and negative impacts <strong>of</strong><br />
computing on human culture. [Community, Global, and<br />
Ethical Impacts]<br />
- Students explain the principles <strong>of</strong> security by examining<br />
encryption, cryptography, and authentication techniques<br />
[Computing Practice and Programming].<br />
While the material is in German, we have translated selected<br />
extracts <strong>of</strong> the material into English, to enable rea<strong>der</strong>s to<br />
un<strong>der</strong>stand how the materials are intended to work. The full<br />
material as well as a didactic comment can be found at [22].<br />
In the following sections, we describe the activities and the<br />
material developed to stimulate and enable the activities in the<br />
context-based learning process.<br />
2.1 Discovering Email Protocols<br />
The first question asked is: “How is an email communicated from<br />
my computer to the computer <strong>of</strong> the addressee?” To un<strong>der</strong>stand<br />
internetwork communication, students need to find out about the<br />
structure <strong>of</strong> interconnected computernetworks such as the Internet<br />
as well as end-to-end communication protocols between client<br />
computer and email server such as the Simple Mail Transfer<br />
Protocol (SMTP) for sending email and the Post Office Protocol<br />
(POP) for retrieving email from a mail server.<br />
To un<strong>der</strong>stand the characteristics and function <strong>of</strong> communication<br />
protocols, students are first asked to communicate a single word<br />
through a door by pulling a cord that passes un<strong>der</strong>neath a door.<br />
Students will soon come up with ideas for coding letters as long<br />
or short, strong or weak pulls or pulls to one <strong>of</strong> two possible<br />
sides. Inevitably from time to time, students will make mistakes in<br />
their coding or decoding and thus will agree on signs to cancel the<br />
transmission <strong>of</strong> a letter and start again from the beginning. They<br />
will also decide for a sign to show the end <strong>of</strong> a letter, e.g. a long<br />
break. Some groups agree on a sign for the receiver to signal an<br />
acknowledgement. This way, without any explicit knowledge <strong>of</strong><br />
what a communication protocol is, they have defined such a<br />
protocol including the separation <strong>of</strong> operational and data signals.<br />
When comparing the protocols <strong>of</strong> different groups they will find<br />
different quality <strong>of</strong> service criteria such as speed and reliability.<br />
The next step will be to have an insight into real life email<br />
network traffic. To this end, we set up a new email server in the<br />
classrooms. We use the email server s<strong>of</strong>tware Hamster which<br />
needs no installation and runs from a USB flash drive, that the<br />
teacher provides during the computer science classes. Students<br />
will sign up for an individual account one after another setting a<br />
secret password, which should not be one they are already using<br />
for another account.
Figure 1. Network traffic for authentification via POP.<br />
After configuring their email account in the email client<br />
application Mozilla Thun<strong>der</strong>bird, they start sending each other<br />
emails and watch the generated email network traffic with a<br />
network traffic analyzer called Socket Sniff (cf. Figure 1).<br />
Figure 2. Pupils are to rearrange the messages in the<br />
correct or<strong>der</strong> using the network traffic analyzed.<br />
Students are asked to analyze their network traffic to reconstruct<br />
either the Simple Mail Transfer Protocol or the Post Office<br />
Protocol. Figure 2 shows a scrambled collection <strong>of</strong> generalized<br />
messages. Using the example <strong>of</strong> their concrete network<br />
connection, students will one by one identify these messages in<br />
their concrete network traffic and thus bring the general messages<br />
into the or<strong>der</strong> specified in the emailing protocol in question. The<br />
left side shows terms for some more general steps, which are to be<br />
assigned to a number <strong>of</strong> messages exchanged to identify typical<br />
phases <strong>of</strong> communication such as initiating a communication<br />
125<br />
channel, authorizing the communication partners, and closing the<br />
communication channel.<br />
In a following cooperative exchange phase students will explain<br />
the protocol they have reconstructed to a student who has<br />
reconstructed the other protocol and then discuss similarities and<br />
differences <strong>of</strong> the two communication protocols.<br />
2.2 Discovering possible Dangers<br />
The second question structuring the learning process is: “What<br />
dangers challenge a private communication over a public network,<br />
such as the Internet?” This question is not explicitly asked in the<br />
class room. Instead, while students are still working on<br />
reconstructing the emailing protocols they experience the teacher<br />
exploiting the dangers <strong>of</strong> an unencrypted, plain-text communication<br />
over a public network in the didactic environment <strong>of</strong> the<br />
class room’s local area network, the different dangers are then<br />
collected in a discussion based on the students’ experiences. This<br />
is actually the reason why we use a separate mail server for the<br />
computer science classes and not the student’s private email<br />
accounts. Therefore, students should be advised at the beginning<br />
to choose a password other than any passwords they already use.<br />
Figure 3. Unsolicited mail from the German Chancellor?<br />
First, the teacher sends an email with a false sen<strong>der</strong> address. We<br />
used an email which claims, that the German Chancellor Angela<br />
Merkel suggests buying stocks <strong>of</strong> the Russian gas company<br />
Gazprom (see Figure 3). While many email provi<strong>der</strong>s check the<br />
correctness <strong>of</strong> a sen<strong>der</strong>’s address, this is not required by the<br />
SMTP protocol or the Hamster email server, The email shows<br />
several features that identify it as spam: It shows frequent spelling<br />
mistakes, shows no specific knowledge about the addressee (Be<br />
honest, who is really friend with a head <strong>of</strong> state?), shows an<br />
inappropriate choice <strong>of</strong> colloquial language and the link leads to a<br />
Url that differs from the text shown for the link. The last aspect is<br />
also the reason why mo<strong>der</strong>n versions <strong>of</strong> email client applications<br />
identify the mail as a possible fraud mail such as phishing mails.<br />
These features can be collected in the discussion to raise the<br />
students’ awareness <strong>of</strong> spam and phishing mails.
While spam and phishing mails can be easily detected when users<br />
are aware <strong>of</strong> the features discussed above, users are more likely to<br />
react to emails that seem more personal. If we go back to Figure 1,<br />
we can see that anyone between the client computer and the mail<br />
server can read the email. With this knowledge, a malevolent<br />
communication participant could send a more individual email, in<br />
which he or she relates to something personal he or she knows<br />
from reading intercepted email. Not only can such a person<br />
retrieve personal information. The Post Office Protocol requires<br />
users to send their password in plain text. Intercepting password<br />
and username, a malevolent communication participant could<br />
access the mailbox and send mail from this mailbox as if it were<br />
his or her own. To demonstrate intercepting email on the<br />
communication path between client an mail server, the server is<br />
connected to the school network via a computer which is extended<br />
by a second network card and configured to bridge the two<br />
network cards. While this computer is technically a network<br />
bridge, it is perfectly fit to simulate an inter-network router. We<br />
have decided to provide students with a network traffic analyzer<br />
with very limited potential. It only reads the traffic assigned to a<br />
process running on the machine that the analyzer is running on.<br />
To intercept messages, teachers need to make use <strong>of</strong> a more<br />
powerful analyzer tool such as Wireshark.<br />
Thirdly, a person with access to the mail server’s memory could<br />
modify the content <strong>of</strong> emails before they are downloaded by the<br />
user. Here, the teacher manipulates emails by opening them from<br />
the computer’s file system using a simple text editor, changes<br />
some facts such as the date <strong>of</strong> a suggested appointment and saves<br />
the changes made to the file. Students follow the manipulation on<br />
a video projector attached to the computer on which the mail<br />
server is running.<br />
The discussion <strong>of</strong> the dangers to a private communication over a<br />
public network should lead to describing the requirements for<br />
secure communication over public networks shown in Table 2:<br />
Table 2. Dangers to and requirements for secure<br />
communication over public networks.<br />
danger requirement<br />
intercepting messages privacy<br />
manipulating messages integrity <strong>of</strong> received messages<br />
faking the sen<strong>der</strong>’s address authenticity <strong>of</strong> sen<strong>der</strong>s<br />
These three requirements structure the steps striving to find<br />
mechanisms that provide the desired security feature as described<br />
in the next section.<br />
2.3 Achieving Privacy through Encryption<br />
By retracing the genesis <strong>of</strong> cryptographic algorithms, students<br />
should not only learn <strong>of</strong> certain algorithms but develop an<br />
awareness <strong>of</strong> important criteria for a cryptographic algorithm and<br />
learn to question the level <strong>of</strong> security provided by a certain<br />
security feature. They should also learn that security always<br />
requires a certain amount <strong>of</strong> trust into people and technology and<br />
can never be fully guaranteed.<br />
The desire to hide information on the communication path to<br />
prevent it from being intercepted by enemies is not at all a new<br />
requirement. The development <strong>of</strong> cryptographic algorithms dates<br />
back to ancient times. The most prominent historic algorithm is<br />
126<br />
probably the Caesar algorithm, where Roman emperors<br />
substituted the characters <strong>of</strong> a message by other characters,<br />
retrieved by shifting an alphabet by a certain amount <strong>of</strong> letters.<br />
The key here is how far the two alphabets are shifted. This<br />
algorithm can easily be applied by children <strong>of</strong> all ages by shifting<br />
two stripes showing two alphabets each, as shown in Figure 4.<br />
Figure 4. Alphabets to apply the Caesar algorithm.<br />
In or<strong>der</strong> to remind students <strong>of</strong> the overall aim to achieve means<br />
for secure emailing, they are asked to exchange messages<br />
encrypted with the Caesar algorithm via email. They are then also<br />
challenged to crack a Caesar cipher <strong>of</strong> which the key is unknown.<br />
By providing a tool called Krypto 1.5 by Michael Kühn to de- and<br />
encrypt Caesar ciphers, we enable students to quickly test the 25<br />
possible keys for a Caesar encryption.<br />
This way it becomes clear that despite having been in use for<br />
centuries the Caesar algorithm is not safe at all. One could argue<br />
that if the letters in the target alphabet were to be distributed to<br />
arbitrary positions and not in the alphabetic or<strong>der</strong>, there would be<br />
26! = 403.291.461.126.605.635.584.000.000 possible keys. While<br />
it is true, that it seems impossible to test all possible keys, longer<br />
texts can still be cracked by comparing the frequency <strong>of</strong> signs to<br />
that <strong>of</strong> the typical frequency <strong>of</strong> letters in the assumed language <strong>of</strong><br />
the original message. This mechanism is well explained in the<br />
story “The Gold Bug” by Edgar Allen Poe.<br />
A more promising way seems to work with multiple alphabets, as<br />
the Vigenère algorithm does, probably the best known symmetric<br />
polyalphabetic algorithm. Here several keys are employed,<br />
described by the letters <strong>of</strong> a keyword. The letter at the current<br />
position <strong>of</strong> the keyword determines how far the alphabet for a<br />
Caesar encryption <strong>of</strong> this letter is shifted.<br />
Figure 5. Animation <strong>of</strong> Vigenère with Krypto 1.5.<br />
The tool Krypto 1.5 provides an excellent animation <strong>of</strong> the<br />
algorithm. A screenshot <strong>of</strong> this animation is shown in Figure 5.<br />
With this animation students can watch the program slowly<br />
encrypting a text.
A weak point <strong>of</strong> the Vigenère algorithm is that when the key is<br />
used repeatedly, typical frequent short sequences <strong>of</strong> letters such as<br />
“and”, “in”, and “but” are likely to translate to the same sequences<br />
in the cipher. The least common multiple <strong>of</strong> the positions <strong>of</strong> such<br />
parallel sequences hints to the length <strong>of</strong> the key. If the length <strong>of</strong><br />
the key is guessed, a frequency analysis can be conducted and the<br />
key can be reconstructed. To secure 100% security, the key used<br />
must be <strong>of</strong> the same length as the message. This lead to the<br />
creation <strong>of</strong> code books called “one time pads”.<br />
While the cipher generated from an arbitrarily generated one-time<br />
pad cannot be cracked, sen<strong>der</strong> and receiver still need to hold a<br />
copy <strong>of</strong> the same one-time-pad. The exchange <strong>of</strong> secret keys can<br />
only be avoided by means <strong>of</strong> asymmetric cryptography such as the<br />
RSA algorithm. While the un<strong>der</strong>lying mathematical concepts <strong>of</strong><br />
RSA are not trivial, the idea <strong>of</strong> using pairs <strong>of</strong> keys where each<br />
communication participant holds a secret and a public key can be<br />
un<strong>der</strong>stood even by young learners.<br />
The image <strong>of</strong> a padlock and a key can be used to a certain extend.<br />
We could imagine handing out a number <strong>of</strong> unlocked padlocks to<br />
communication partners, who can now lock a box by closing the<br />
lock. If we keep the key, we are the only person able to unlock the<br />
box. Of course, we cannot communicate padlocks digitally, so we<br />
need a function on numbers to do a comparable job. Here the<br />
modulo function comes into play.<br />
Figure 6. Animation <strong>of</strong> asymmetric cryptography.<br />
To introduce students to the idea <strong>of</strong> asymmetric cryptography we<br />
have developed an online form, where students can exchange<br />
messages and encrypt them with public and private keys. Figure 6<br />
shows the form in use. Very small numbers are used as keys so<br />
that students can better follow the flow <strong>of</strong> information. The form<br />
is accompanied by instructions on how to apply the keys. If they<br />
un<strong>der</strong>stand the instructions correctly, they will be able to<br />
successfully decrypt messages encrypted with their public key.<br />
Once the idea <strong>of</strong> asymmetric cryptography has been established,<br />
the algorithm for generating RSA keys is presented and students<br />
manually en- and decrypt their birthday using very small keys.<br />
Students are then asked to challenge the encryption with small<br />
numbers by trying to reconstruct a possible private key for a given<br />
public key. They will find out, that using larger keys makes it<br />
more difficult to reconstruct a private key. This raises the question<br />
127<br />
how large a key must be so that mo<strong>der</strong>n computers will not be<br />
able to crack an RSA cipher in a realistic time span. We use the<br />
tool CrypTool to challenge larger RSA ciphers and ask students to<br />
gather information on the RSA Factoring Challenge.<br />
2.4 Verifying a digital Signature<br />
Figure 6 shows, that from the texts a hash value is generated, that<br />
is unique for a given sequence <strong>of</strong> characters. When the sen<strong>der</strong><br />
sends the hash value as well, the receiver can detect changes made<br />
to the message, because the hash value calculated from the<br />
received text does not match the one sent by the sen<strong>der</strong>.<br />
Unfortunately, a person could manipulate both the message and<br />
the hash value. This means, that the hash value has to be<br />
protected. The sen<strong>der</strong> encrypts the hash value using his or her<br />
own private key. Now anyone can decrypt the signature using the<br />
sen<strong>der</strong>’s public key to verify the signature. But nobody except the<br />
sen<strong>der</strong> can produce a signature that can be correctly decrypted<br />
with the sen<strong>der</strong>’s public key. This way, both the message’s<br />
integrity and the sen<strong>der</strong>’s authenticity can be verified.<br />
Students are <strong>of</strong>ten confused that now the or<strong>der</strong> in which the keys<br />
are employed differs from the one used to encrypt messages to<br />
achieve privacy. Teachers are best advised to point out and<br />
discuss this difference and stress the fact, that this mechanism<br />
suits a very different purpose.<br />
To acquire also practical competencies in using digital signatures<br />
students should first encrypt and sign emails using the animation<br />
from the project website. Then, teachers should motivate them to<br />
configure the PGP-plugin Enigmail for Mozilla Thun<strong>der</strong>bird,<br />
generate a set <strong>of</strong> keys, exchange public keys with their fellow<br />
students and encrypt and sign, decrypt and verify emails and<br />
signatures with Enigmail.<br />
2.5 Motivation for secure Communication<br />
In a jigsaw puzzle students first study texts to become experts for<br />
one <strong>of</strong> the following four topics:<br />
1. Freedom <strong>of</strong> communication as a request in countries<br />
un<strong>der</strong> autocratic rule such as Iran or China.<br />
2. The Echelon project as an example <strong>of</strong> an intelligence<br />
system analyzing email traffic.<br />
3. De-Mail as an example <strong>of</strong> attempts for commercial<br />
email cryptography with trust centres.<br />
4. Pretty Good Privacy (PGP) as an example <strong>of</strong> royalty<br />
free cryptography with a web <strong>of</strong> trust.<br />
Students then exchange the expert knowledge they have acquired<br />
in mixed groups.<br />
3. LESSONS LEARNED<br />
One question frequently discussed is: Should teachers find subject<br />
matter that suits an interesting context or should they find a<br />
context that is suitable to teach a certain subject matter?<br />
Traditionally, curricula point out fields <strong>of</strong> subject matter and<br />
assign them to certain years <strong>of</strong> learning. Teachers therefore are<br />
used to decide on the subject matter first, and then select a<br />
suitable contextualisation or application <strong>of</strong> an abstract scientific<br />
concept. More recently, curricula also point out fields <strong>of</strong><br />
competences as well as fields <strong>of</strong> content. Teachers now have to<br />
decide on which competences they want their students to acquire,<br />
which content from the curriculum is suitable to develop these<br />
competences and which context can be used to show the concept’s
elevance. Within the IniK working group, we have started out<br />
designing context-based units both from a context, brainstorming<br />
on subject matter that could best be acquired, as well as from a<br />
certain subject matter, brainstorming on what context could best<br />
demonstrate the concept's relevance to students. In the case <strong>of</strong><br />
“Email for You (only?)” we first grouped and rearranged parts <strong>of</strong><br />
activities developed earlier and then modified and added material<br />
to bridge gaps where needed. In practise, both ways seem<br />
possible, practical and worth exploring.<br />
Figure 7. Fundamental and optional modules.<br />
In the student-oriented learning process students spend a lot <strong>of</strong><br />
time actively and individually engaging with the different topics.<br />
It is at times challenging to select just some <strong>of</strong> the many possible<br />
aspects and dimensions <strong>of</strong> a topic. Several very different units are<br />
possible for a single context. If a learning process turns out to<br />
require a lot <strong>of</strong> time, it makes sense to mark several aspects as<br />
optional and others as mandatory parts, to encourage teachers to<br />
first try out parts <strong>of</strong> the material. We have grouped the different<br />
learning activities into modules as shown in Figure 7.<br />
While some modules are necessary for an un<strong>der</strong>standing <strong>of</strong><br />
following modules, others only lead to a more thorough<br />
un<strong>der</strong>standing <strong>of</strong> aspects such as the concrete implementation <strong>of</strong><br />
algorithms or the level <strong>of</strong> security they provide.<br />
The proposed learning arrangement involves a lot <strong>of</strong> s<strong>of</strong>tware. All<br />
the s<strong>of</strong>tware proposed can be used free <strong>of</strong> royalties. While most <strong>of</strong><br />
them are available for Windows, Mac and Linux operating system,<br />
some are only available for the Windows platform. Here,<br />
alternative applications should be found for teachers working with<br />
other platforms. For a simple start we have gathered portable<br />
versions <strong>of</strong> the s<strong>of</strong>tware into a Zip-File, which can be extracted to<br />
a USB flash drive. Programs can then be started directly from the<br />
USB flash drive without any need for setup routines. This is<br />
extremely practical for teachers working in an environment, where<br />
s<strong>of</strong>tware installation on clients is laborious. A disadvantage <strong>of</strong><br />
providing the Zip-File is, that the applications are quickly<br />
outdated, especially in the case <strong>of</strong> Mozilla Thun<strong>der</strong>bird. This is<br />
also true for the directions given to students. Several times, these<br />
directions had to be adjusted after the dialogue for setting up an<br />
email account in Thun<strong>der</strong>bird was changed. (As a matter <strong>of</strong> fact, it<br />
128<br />
has been improved in the sense that users are now warned <strong>of</strong><br />
connecting to a mail server without SSL or TLS encryption. We<br />
have found a way to work around the warning so that students can<br />
still discover the dangers <strong>of</strong> plain text communication with an<br />
email server.)<br />
While we believe we have made good use <strong>of</strong> the context for<br />
purposes <strong>of</strong> providing illustration and meaning, we must admit<br />
that with using the material as it is little is done to decontextualise<br />
newly gained knowledge by applying it to similar problems in<br />
other fields <strong>of</strong> application, e. g. designing a protocol for the<br />
communication <strong>of</strong> different parts or layers <strong>of</strong> services within a<br />
single device or discussing security issues in other situations, e. g.<br />
large permanent storages such as data bases. It is our impression<br />
that most <strong>of</strong> the other learning processes proposed at the projects<br />
website [21] lack opportunities for decontextualisation. Since the<br />
curricula only give orientation and concrete courses in computer<br />
science vary a lot between different schools and teachers, it is<br />
difficult to suggest connections to what students have learned<br />
before. Thus, it is left to teachers to realize these connections – so<br />
far no guidance is given on how to achieve decontextualisation.<br />
Finally, designing, implementing, testing and readjusting “Email<br />
for You (only?)” has taken a lot <strong>of</strong> time. If the project is to be<br />
successful in establishing the concept in everyday CS classes, it<br />
will be necessary to acquire further resources to give teachers and<br />
university staff the time needed to develop such material. A<br />
promising approach could be to include university students and<br />
teacher trainees in the design and reflection <strong>of</strong> context-based<br />
learning processes.<br />
4. FUTURE WORK<br />
Further contexts should be explored to estimate their potential for<br />
teaching computer science in secondary school. The IniK project<br />
group has started a list <strong>of</strong> contexts [23] which is grouped into<br />
ideas for contexts with a potential but need for developing<br />
suitable teaching materials, ideas for contexts where it is unclear,<br />
which concrete aspects <strong>of</strong> the contexts are useful to the teaching<br />
<strong>of</strong> computer science in secondary school education, and ideas for<br />
contexts which at a first glance seem attractive but after a more<br />
thorough examination don’t seem useful to the teaching <strong>of</strong><br />
computer science in secondary school education. It would be<br />
helpful to have a more detailed un<strong>der</strong>standing as to what potential<br />
different contexts have. A suggested learning process will always<br />
discuss only a selection <strong>of</strong> possible aspects, so eventually we<br />
could have different units exploring the same contexts with<br />
different focus. More teaching materials for further topics should<br />
be developed. Today, there is e. g. no unit that motivates the<br />
design and use <strong>of</strong> databases – visualization using open data seems<br />
an interesting approach here.<br />
But next to the development <strong>of</strong> further material, experiences made<br />
in the design and implementation <strong>of</strong> context-based computer<br />
science courses should be collected, shared and discussed in or<strong>der</strong><br />
to develop guidelines and advice for the development <strong>of</strong> contextbased<br />
learning processes in secondary school computer science<br />
education. An example would be concepts for a sustainable<br />
decontextualisation, which are missing so far.<br />
5. CONCLUSION<br />
“Email four You (only?)” shows that a context-based learning <strong>of</strong><br />
computer science is possible and helps students un<strong>der</strong>stand the<br />
relevance <strong>of</strong> knowledge they acquire in computer science classes.
We have learnt that we can both look for contexts where a certain<br />
topic is <strong>of</strong> relevance as well as screen an interesting context as to<br />
which topics <strong>of</strong> computer science education are relevant for it.<br />
Either way, it is a long way <strong>of</strong> selecting suitable aspects out <strong>of</strong> a<br />
wide choice <strong>of</strong> possible aspects and developing material which<br />
makes sense at a certain point <strong>of</strong> the learning process. If in doubt,<br />
one should decide on whether aspects are necessary for the<br />
following sections <strong>of</strong> the learning process. If this is not the case,<br />
they can be marked as optional. Whenever s<strong>of</strong>tware is involved,<br />
the s<strong>of</strong>tware should be easy to use, if possible without installation.<br />
We have by now presented our proposed learning arrangement in<br />
several hands-on workshops for computer science teachers<br />
throughout <strong>Germany</strong> and also in the journal LOG IN [16], which<br />
is widely distributed amongst computer science teachers<br />
throughout <strong>Germany</strong>. Since then, we have received a lot <strong>of</strong><br />
positive feedback from colleagues stating that their students are<br />
highly motivated to un<strong>der</strong>stand both the un<strong>der</strong>lying technology <strong>of</strong><br />
email as well as the different cryptographic algorithms. While we<br />
are happy to see our own impressions affirmed, we are indeed<br />
aware that this is only a first but nevertheless promising hint as to<br />
the potential <strong>of</strong> context-based learning <strong>of</strong> computer science in<br />
secondary school education. Reaching a stage where we now have<br />
several learning processes readily prepared and testes in class, it<br />
now seems important to validate our theses on context-based<br />
learning <strong>of</strong> computer science in a more through, scientific<br />
evaluation.<br />
6. REFERENCES<br />
[1] Arbeitskreis »Bildungsstandards« <strong>der</strong> Gesellschaft für<br />
Informatik (ed.), 2008. Grundsätze und Standards für die<br />
Informatik in <strong>der</strong> Schule – Bildungsstandards Informatik<br />
für die Sekundarstufe I. Empfehlungen <strong>der</strong> Gesellschaft<br />
für Informatik e. V. vom 24. Januar 2008. In LOG IN ,<br />
supplementary to vol. 150/151 (2008). Retrieved<br />
October 10, 2012. http://informatikstandards.de<br />
[2] Altrichter, H., Posch, P., Somekh, B. 1993. Teachers<br />
Investigate their Work: Introduction to the Method <strong>of</strong><br />
Action Research. Routledge, London, New York 1993.<br />
[3] Barab, S. and Squire, K. 2004. Introduction: Design-<br />
Based Research: Putting a Stake in the Ground. The<br />
Journal <strong>of</strong> learning sciences. 13,1 (2004), 1-14.<br />
[4] Brinda, T., Puhlmann, H., and Schulte, C., 2009.<br />
Bridging ICT and CS – Educational Standards for<br />
Computer Science in Lower Secondary Education. In<br />
ITiCSE '09: Proceedings <strong>of</strong> the 14th annual ACM<br />
SIGCSE conference on Innovation and technology in<br />
computer science education, 288-292.<br />
[5] Computer Science Teachers Association (CSTA) 2011.<br />
K-12 Computer Science Standards. Revised 2011.<br />
Retrieved October 10, 2012. http://csta.acm.org/<br />
Curriculum/sub/CurrFiles/CSTA_K-12_CSS.pdf<br />
[6] Cooper, S. and Cunningham, S. 2010. Teaching<br />
computer science in context. ACM Inroads 1, 5‐8.<br />
[7] Coy, W. 2005. Informatik … im Großen und Ganzen. In<br />
LOG IN 136 (2005), pp.17-23.<br />
[8] Diethelm, I., Borowski, C., and Weber, T. 2010.<br />
Identifying relevant CS contexts using the miracle<br />
question. In Proceedings <strong>of</strong> the 10th Koli Calling<br />
International Conference on Computing Education<br />
129<br />
Research. Koli Calling 10. ACM, New York, NY, USA,<br />
74‐75.<br />
[9] Diethelm, I. and Dörge, C. 2010. From Context to<br />
Competencies. In Key Competencies in the Knowledge<br />
Society, N. Reynolds and M. Turcsányi-Szabó, Eds. IFIP<br />
Advances in Information and Communication<br />
Technology. Springer Boston, 67–77.<br />
[10] Diethelm, I., Hildebrandt, C., and Krekeler, L. 2009.<br />
Implementation <strong>of</strong> Computer Science in Context - a<br />
research perspective regarding teacher-training. In Koli<br />
Calling 2009. 9th International Conference on<br />
Computing Education Research, A. Pears and C. Schulte,<br />
Eds., 96–99.<br />
[11] Diethelm, I., Schulte, C. and Witten, H. 2011. Informatik<br />
im Kontext (IniK) – Entwicklungen, Merkmale und<br />
Perspektiven. In Praxisberichte zur 14. GI-Fachtagung<br />
Informatik und Schule (INFOS), Münster 2011.<br />
[12] Engbring, D. 2005. Informatik im Kontext. In LOG IN,<br />
vol. 136 (2005), pp.28-33.<br />
[13] Esslinger, B., Gramm, A., Hornung, M. and Witten, H.<br />
2011. Asymmetrische Kryptographie für die Sek I - RSA<br />
(fast) ohne Mathematik? In Informatik mit Kopf, Herz<br />
und Hand. Praxisbeiträge zur 14. GI-Fachtagung<br />
Informatik und Schule (INFOS 2011) (Münster,<br />
<strong>Germany</strong>, September 13 - 15, 2011) Zentrum für Lehrerbildung<br />
<strong>der</strong> Universität Münster 2011, pp. 225-235.<br />
[14] Esslinger, B., Gramm, A., Hornung, M. and Witten, H.<br />
2012. Kann man RSA vertrauen? In LOG IN vol.<br />
172/173 (2012), in print.<br />
[15] Gramm, A. 2012. Animation <strong>of</strong> Asymmetric<br />
Cryptography. Retrieved October 10, 2012<br />
http://it-lehren.de/asymcrypt<br />
[16] Gramm, A., Hornung, M. and Witten, H. 2011. E-Mail<br />
(nur?) für Dich - Eine Unterrichtsreihe des Projekts<br />
Informatik im Kontext. In LOG IN , supplementary to<br />
vol. 169/170 (2011).<br />
[17] Guizdal, M. 2010. Does contextualized computing<br />
education help? ACM Inroads 1, 4, 4‐6.<br />
[18] mpfs – Medienpädagogischer Forschungsverbund<br />
Südwest (ed.), 2011. JIM-Studie 2011 – Jugend,<br />
Information, (Multi-)Media. Basisuntersuchung zum<br />
Medienumgang 12- bis 19-Jähriger. Stuttgart:<br />
Medienpädagogischer Forschungsverbund Südwest,<br />
2011, p. 32. Retrieved October 10, 2012.<br />
http://www.mpfs.de/fileadmin/JIM-pdf11/JIM2011.pdf<br />
[19] Parchmann, I., Gräsel, C., Baer, A., Demuth, R. and<br />
Ralle, B. 2007. Chemie im Kontext - a symbiotic<br />
implementation <strong>of</strong> a context-based teaching and learning<br />
approach. In International Journal <strong>of</strong> Science Education<br />
28, 09 (2007) pp. 1041-1062. DOI=<br />
http://doi.acm.org/10.1080/09500690600702512<br />
[20] Klieme, E. et al. 2004. The Development <strong>of</strong> National<br />
Educational Standards. An Expertise. Bundesministerium<br />
für Bildung und Forschung, Berlin.<br />
[21] Koubek, J. 2012. Website <strong>of</strong> the “Informatik im Kontext”<br />
project. Retrieved October 10, 2012.<br />
http://informatik-im-kontext.de
[22] Koubek, J. 2012. Website <strong>of</strong> the “Informatik im Kontext”<br />
project – section “Entwürfe » Email (nur?) für Dich”.<br />
Retrieved October 10, 2012. http://informatik-imkontext.de/index.php/entwuerfe/email-nur-fuer-dich/<br />
[23] Koubek, J. 2012. Website <strong>of</strong> the “Informatik im Kontext“<br />
project – section “Kontexte und Ideen”. Retrieved<br />
October 10, 2012.<br />
http://informatik-im-kontext.de/index.php/kontextideen/<br />
[24] Koubek, J.; Schulte, C.; Schulze, P. and Witten, H. 2009.<br />
Informatik im Kontext (IniK) – Ein integratives<br />
130<br />
Unterrichtskonzept für den Informatikunterricht. In GI-<br />
Edition LNI – Lecture Notes in Informatics, vol. P156.<br />
Zukunft braucht Herkunft – 25 Jahre »INFOS –<br />
Informatik und Schule«. INFOS 2009 – 13. GI-<br />
Fachtagung Informatik und Schule 21.–24. September<br />
2009 in Berlin, pp. 268–279.<br />
[25] Saeli, M., Schulte, S. To be published 2013. Applying<br />
Standards to Computer Science Education. In Improving<br />
Computer Science Education (by Kadijevic, D., Angeli,<br />
C., Schulte, C.), 117-131.
Comparing CSTA K-12 Computer Science Standards<br />
with Austrian Curricula<br />
Daniel L. Egger<br />
Alpen-Adria-Universität Klagenfurt<br />
Universitätsstraße 65-67<br />
A- 9020 Klagenfurt<br />
+43 650 519 7333<br />
egger.it@edu.uni-klu.ac.at<br />
ABSTRACT<br />
In 2011, the CSTA has published a review <strong>of</strong> its K-12 Computer<br />
Science Standards. Aiming to assess the situation <strong>of</strong> Computer<br />
Science Education in Austria, we used these standards as an external<br />
level rod. For this purpose, we investigated to which degree<br />
a well-chosen subset <strong>of</strong> these standards is implemented in Austrian<br />
curricula for computer science education. We analyzed curricula<br />
<strong>of</strong> several types <strong>of</strong> secondary schools (AHS, HTL for<br />
Chemistry, HTL for Informatics, HLW). Our findings reveal that<br />
the CSTA standards are quite poorly implemented at this point <strong>of</strong><br />
time in Austria.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computer and Information Science Education]: Computer<br />
Science Education, Curriculum.<br />
General Terms<br />
Human Factors, Standardization<br />
Keywords<br />
Computer Science Education, Secondary School, Educational<br />
Standards, Curricula<br />
1. INTRODUCTION<br />
During the last years, more and more countries have come to the<br />
conclusion that it would be necessary to incorporate serious programs<br />
for Computer Science Education (CSE) in their school<br />
systems. As stated in 2011 by the US Computer Science Teacher<br />
Association (CSTA): “To function in society, every citizen in the<br />
21 st century must un<strong>der</strong>stand at least the principles <strong>of</strong> computer<br />
science. […] Elementary and secondary schools have a unique<br />
opportunity and responsibility to address this need.” [10]. Nevertheless,<br />
as the education systems <strong>of</strong> the countries differ substantially,<br />
the implementation and organization <strong>of</strong> CSE has to be also<br />
very diverse from country to country. In or<strong>der</strong> to compare the<br />
outcomes <strong>of</strong> such different education systems, suitable educational<br />
standards have to be defined and applied.<br />
This paper, which was produced mainly by the students <strong>of</strong> a<br />
teacher education course at our university, aims to demonstrate<br />
how such a comparison might look like. We investigated, to<br />
which degree a suitable subset <strong>of</strong> the K-12 standards that have<br />
been most recently published by the CSTA [10] is adopted by the<br />
Austrian education system. Our work was inspired by the CSTA<br />
study “Running On Empty” that compared the outcomes <strong>of</strong> CSE<br />
in the 50 US states. The comparison will show to which degree<br />
the current Austrian programs for CSE fulfill the US standards.<br />
Sabrina M. Elsenbaumer<br />
Alpen-Adria-Universität Klagenfurt<br />
Universitätsstraße 65-67<br />
A- 9020 Klagenfurt<br />
+43 650 830 3302<br />
selsenba@edu.uni-klu.ac.at<br />
131<br />
Peter Hubwieser<br />
Technische Universität München<br />
Fakultät für Informatik<br />
Boltzmannstr. 3, D-85478 Garching<br />
+49 89 289 17350<br />
Peter.Hubwieser@tum.de<br />
Although there are other educational standards for CSE, this<br />
paper will not deal with any standards in detail other than the<br />
CSTA standards. It is also not the purpose <strong>of</strong> this paper to discuss<br />
or argue for the appropriateness <strong>of</strong> the CSTA standards in general.<br />
We will simply take them for granted and for a reference <strong>of</strong> comparison<br />
for the means <strong>of</strong> this paper.<br />
As we do not intend to discuss the process <strong>of</strong> standardization for<br />
CSE in general, we refer to [5] and [3], where the current situation<br />
regarding standardization in CSE in connection with curriculum<br />
topics is discussed systematically.<br />
We will start with the explanation <strong>of</strong> the theoretical background<br />
and <strong>of</strong> the Austrian school system and continue with a short overview<br />
<strong>of</strong> the CSTA standards. A final comparison <strong>of</strong> the individual<br />
curricula will detect the differences in implementation, which will<br />
give reason for a general conclusion.<br />
2. THEORETICAL BACKGROUND<br />
Klieme et al. (see [4] p. 42) define educational standards as “a<br />
clear and concise statement <strong>of</strong> what matters in [a] school system”.<br />
However, educational standards shall not be confused with regulations<br />
for grading and examination. They do not substitute for<br />
whole curricula, but rather specify fundamental themes in key<br />
learning fields (see [4], pp. 85-86). It seems clear that standards<br />
can only be effective if they mirror parents’, teachers’ and students’<br />
expectations <strong>of</strong> what should be taught/learned and if they<br />
only serve a purpose other than intensification <strong>of</strong> the already<br />
predominant (see [7], pp. 203). Also, it seems as if educational<br />
standards could provide an instrument to foster system-wide<br />
educational equality and comparability.<br />
Yet, some negative aspects have to be consi<strong>der</strong>ed. First, it seems<br />
obvious that some teachers may perceive educational standards as<br />
an annoyance, irritation or burden limiting their flexibility and<br />
freedom in teaching. Furthermore, the teaching staff will have to<br />
agree upon how set standards can be met, which definitely can<br />
lead to major arguments among the staff (see [4],pp. 42-49).<br />
In addition, there is a strong disagreement about whether standards<br />
in education lead to standardization rather than to diversification<br />
<strong>of</strong> learning processes. Reigeluth [7] warns that we have to<br />
discriminate between two uses/types <strong>of</strong> standards: standards that<br />
foster uniformity and standards that support customization,<br />
whereby the latter is definitely to be given priority. Uniform<br />
standards will make the education system increasingly inflexible<br />
and in “the broadest sense, [such standards can be seen as] an<br />
expression <strong>of</strong> pr<strong>of</strong>essional values in the context <strong>of</strong> a test’s purpose”<br />
(see [6], p. 464) thus they could have a huge impact on<br />
business and economy (see [1], p. 266).
On the contrary, customizing standards allow for a greater degree<br />
<strong>of</strong> flexibility in teaching as they respect that different students<br />
learn at different rates. This means standards must specify minimal<br />
standards on the one hand and extended standards, in which<br />
students can pursue their personal interests further, on the other.<br />
In doing so, Reigeluth suggests establishing measurable, crossgradual<br />
standard levels without relation to timetables, permitting<br />
students to choose specifications freely within certain limits. Such<br />
standards work against the sorting-out <strong>of</strong> students and promote<br />
motivation building in reaching the respective standards [7].<br />
Though educational standards affect everyone dealing with (public)<br />
education ([7], p. 202), special attention should be given to<br />
their implications on school curricula. “The experiences […] with<br />
educational standards show that the implementation <strong>of</strong> this new<br />
form <strong>of</strong> output-driven management is certain to prompt changes<br />
in the work <strong>of</strong> schools, but the implications for the curricula are<br />
not as clear-cut; rather, the options here are diverse and openended”,<br />
state Klieme et al. [4] on p. 82.<br />
3. RELATED WORK<br />
The publication that was certainly the most influential for this<br />
paper is the report Running on Empty: The Failure to Teach K–12<br />
Computer Science in the Digital Age by Wilson et al. [12]. It<br />
describes how computer science education is represented in the<br />
curricula <strong>of</strong> the United States. Wilson et al. apply a methodology<br />
that is very similar to ours, providing evidence for the implementation<br />
<strong>of</strong> the 55 CSTA K-12 Standards in the version <strong>of</strong> 2006 [8]<br />
in the mentioned curricula: “The researchers were very liberal in<br />
the analysis; that is, they required minimal evidence to consi<strong>der</strong> a<br />
particular ACM/ CSTA standard as adopted. If the state standards<br />
made any reference to the general idea detailed in the<br />
ACM/CSTA standards, it was marked as adopted“ see [12], p. 34.<br />
The report covers all 50 states <strong>of</strong> the USA and all 3 levels <strong>of</strong> the<br />
CSTA standards from 2006.<br />
Unfortunately, it provides a disillusioning result: It finds “that<br />
roughly two-thirds <strong>of</strong> the entire country has few computer science<br />
standards for secondary school education, K–8 computer science<br />
standards are deeply confused, few states count computer science<br />
as a core academic subject for graduation, and computer science<br />
teacher certification is deeply flawed” [11].<br />
The individual results vary largely from one state to another,<br />
since there are states with hardly any standards implemented<br />
while some conform to (almost) all <strong>of</strong> them. A neatly arranged,<br />
interactive overview <strong>of</strong> the state-by-state results can be found at<br />
[2] . The overall results are displayed in Fig. 1.<br />
Figure 1. Overall results <strong>of</strong> the CSTA survey [2] .<br />
As our investigation refers to the revised version <strong>of</strong> the CSTA<br />
standards from 2011 ([10], see below), which is different from the<br />
2006 version in many respects (e.g. the overall number <strong>of</strong> stand-<br />
132<br />
ards, the definition <strong>of</strong> the levels and the strands) it is , unfortunately,<br />
not possible to compare our results with those <strong>of</strong> Wilson et<br />
al. [12].<br />
4. THE EDUCATIONAL CONTEXT<br />
Since the CSTA Standards were developed in the context <strong>of</strong> the<br />
US educational system, whereas we refer to the Austrian system,<br />
it seems appropriate to outline the major differences and similarities<br />
<strong>of</strong> these two systems (see Fig. 2). The relevance <strong>of</strong> national<br />
policies and educational systems to CSE was discussed closely by<br />
a recent ITiCSE working group [3].<br />
In the Austrian school system, children are required to attend<br />
school for 9 years (respectively grades). This section <strong>of</strong> compulsory<br />
schooling includes 4 years <strong>of</strong> primary school education, 4<br />
years <strong>of</strong> lower secondary school education and one additional<br />
year from the upper secondary education level. For the upper<br />
secondary education (up to the final ‘Reifeprüfungen’ that qualify<br />
for enrollment at universities) there are various types <strong>of</strong> schools<br />
with various specifications and also different numbers <strong>of</strong> years<br />
that can be chosen by the students after they have passed grade 8,<br />
e.g.<br />
- AHS (“Allgemeinbildende höhere Schule”, aiming to general<br />
education),<br />
- HAK (“Handelsakademie”), focused on trade and business,<br />
- HLW (“Höhere Lehranstalt für wirtschaftliche Berufe”),<br />
mainly educating for tourism and business,<br />
- HTL (“Höhere Technische Lehranstalt”), mainly technically<br />
and business-oriented),<br />
- BMS (“Berufsbildende mittlere Schulen”) that lead to a<br />
vocational qualification.<br />
Besides these types, there are specific schools for Vocational<br />
Education and Training, specializing in one specific pr<strong>of</strong>ession.<br />
Figure 2: School Systems <strong>of</strong> Austria and the USA<br />
In the US school system, the number <strong>of</strong> years (respectively<br />
grades) which students are required to attend school differs from<br />
one state to the next. Elementary schooling includes nursery<br />
schools and kin<strong>der</strong>gartens on the lowest level, which are followed<br />
by elementary schools. The primary education may end with
grade 4, 5, 6 or 8, depending on choice <strong>of</strong> the parents (and the<br />
system <strong>of</strong> the respective state). After primary schools, the children<br />
may attend a middle school, a 4-year high school a junior high<br />
school, followed by a senior high school or a combined juniorsenior<br />
high school up to the 12th grade. After graduation from<br />
high school, it is possible to attend various forms <strong>of</strong> postsecondary<br />
education such as colleges, universities and vocational institutions.<br />
5. CSTA STANDARDS<br />
In addition to a revision <strong>of</strong> the ACM K-12 curriculum from 2003<br />
[9] the CSTA published the first version <strong>of</strong> the K-12 CSE standards<br />
in 2006 [8]. To keep track with the current flow <strong>of</strong> technological<br />
development, the standards were revised again in 2011<br />
[10]. This included also a substantial raise <strong>of</strong> the number <strong>of</strong> standards<br />
(from 55 in 2006 to 174 in 2011).<br />
5.1 Levels<br />
The CSTA developed a three-level-model for K-12 computer<br />
science, which means that each <strong>of</strong> the three levels aims to represent<br />
different age- and knowledge-levels for students from Kin<strong>der</strong>gartens<br />
(K) up to grade 12. Level one is valid for K to grade 6,<br />
level two from grade 6 to 9 and finally level three for grades 9 to<br />
12.<br />
Fig. 3 visualizes the three levels. Level 1 is called “Computer<br />
Science and Me”, is designed for young elementary school students<br />
and aims to support those with a basic knowledge in technology.<br />
It demands simple concepts <strong>of</strong> computational thinking.<br />
Furthermore, the students in K-6 should develop a first un<strong>der</strong>standing<br />
<strong>of</strong> the importance <strong>of</strong> computer science in an engaging,<br />
creative and explorative atmosphere. Often these first ideas are<br />
embedded within suitable contexts from social science, language,<br />
arts or mathematics.<br />
Figure 3. Three-Level-Model for CSTA Standards [10].<br />
Level 2, titled “Computer Science and Community”, demands<br />
deeper knowledge in or<strong>der</strong> to un<strong>der</strong>stand computational thinking<br />
as a problem solving tool. Moreover, grade 6-9 students should<br />
apply computers as devices which facilitate communication and<br />
collaboration within the computer science classroom and also in<br />
other curricular areas.<br />
Finally, according to level 3, the students in grades 9 to 12 should<br />
be able to apply learned concepts and create real-world solutions.<br />
The students are advanced learners <strong>of</strong> computer science, who<br />
should already be able to contribute to the development <strong>of</strong> new<br />
technologies. Level 3 is divided into three different categories,<br />
each focusing on a diverse field <strong>of</strong> computer science. Level 3A,<br />
called “Computer Science in the Mo<strong>der</strong>n World”, represents basic<br />
abilities that should be acquired in grade 9 or 10. It addresses all<br />
133<br />
students, demanding consolidated knowledge and abilities. The<br />
students should be able to make competent decisions regarding<br />
computer systems and computational techniques and transfer<br />
these abilities to whatever job they might be working in later.<br />
Apart from making use <strong>of</strong> their knowledge, they should also be<br />
aware <strong>of</strong> and appreciate the benefits <strong>of</strong> computers but also un<strong>der</strong>stand<br />
the social and ethical impact <strong>of</strong> mo<strong>der</strong>n technologies in<br />
today’s society.<br />
Level 3B and 3C are intended for more advanced studies. In<br />
grades 10 or 11, according to Level 3B (“Computer Science concepts<br />
and practices”), students should learn algorithmic thinking<br />
and problem-solving strategies, they should deepen their<br />
knowledge <strong>of</strong> the principles <strong>of</strong> computer science and become<br />
experts in working collaboratively, using collaboration tools when<br />
solving problems. Level 3C, represents in-depth studies in one<br />
particular field. It is recommended for grade 11 or 12 students<br />
who wish to gain pr<strong>of</strong>essional computing certification or equivalent<br />
education, helping them to carve out their career (see [10],<br />
pp. 7-9).<br />
5.2 Strands<br />
To take the complexity <strong>of</strong> computer science into account, the<br />
CSTA standards distinguish five “complementary and essential<br />
strands” (see [10], p. 9), displayed by Fig. 4. In the following we<br />
give a short description <strong>of</strong> these five strands.<br />
Computational Thinking (CT) is one <strong>of</strong> the core elements <strong>of</strong><br />
computer science. The main idea <strong>of</strong> CT is the problem-solving<br />
methodology. Concepts like abstraction, recursion or analysis not<br />
only target for educational consumers but also for producers <strong>of</strong><br />
new technological devices.<br />
Collaboration (CO) starts at school, when students work cooperatively<br />
together, gathering information or using a variety <strong>of</strong> communication<br />
tools, to name just a few examples, and leads to an<br />
effective collaboration on important projects with colleagues in<br />
future careers.<br />
Figure 4. Strands according to CSTA Standards [10].<br />
Computing Practice & Programming (CP). As though programming<br />
is an essential part <strong>of</strong> computer science, it is not the only<br />
one. Other abilities, like developing homepages, using s<strong>of</strong>tware<br />
tools, handling databases, organizing files and fol<strong>der</strong>s, dealing<br />
with computational problems, etc. are good examples <strong>of</strong> what<br />
computing practice might also include.
Computers and Communications Devices (CC): the students<br />
should un<strong>der</strong>stand the elements <strong>of</strong> mo<strong>der</strong>n computers and communication<br />
devices and be able to use them competently.<br />
Community, global and ethical impacts (CG) deals with ethical<br />
norms, responsibility and general awareness. The students need to<br />
be informed about privacy and security topics, and develop a<br />
certain sense for ethically correct choices and appropriate social<br />
networking behavior. It is important to un<strong>der</strong>stand the positive but<br />
also negative impacts computers and technology might have on<br />
everyone’s life.<br />
6. METHODOLOGY<br />
6.1 Selection <strong>of</strong> Curricula<br />
As already mentioned in the introduction, the results that are<br />
presented here were produced by students during a seminar that is<br />
part <strong>of</strong> a teacher education program. Un<strong>der</strong> these circumstances it<br />
was not possible to investigate the adoption <strong>of</strong> all 174 standards.<br />
Also, it would have taken too much time to analyze the curricula<br />
<strong>of</strong> all Austrian school types (which are numerous, see section 4).<br />
Thus we had to select several curricula and an appropriate subset<br />
<strong>of</strong> standards for our research. As computer science education in<br />
Austria does not take place before secondary education, we decided<br />
to focus on the curricula <strong>of</strong> several important types <strong>of</strong> secondary<br />
schools, which are described in the following subsections.<br />
6.1.1 AHS<br />
The “Allgemeinbildende Höhere Schule” is the school type with<br />
the broadest form <strong>of</strong> higher education. Schools that belong to this<br />
category include the so-called “Gymnasium” and the “Bundesoberstufenrealgymnasium”<br />
(BORG). An AHS is a school with<br />
focus on a comprehensive and broad general education that lays<br />
the foundations for the studies at a university. The AHS is divided<br />
into an “Unterstufe”, comprising the grades 5 to 8, and an “Oberstufe”<br />
that consists <strong>of</strong> grades 9 to 12. After successfully passing<br />
the “Unterstufe”, students may subsequently attend the “Oberstufe”,<br />
but instead they might also choose to attend a different<br />
school like a BHS, which will be explained in the chapter below,<br />
or might switch to vocational education and training. Furthermore,<br />
the AHS distinguishes between three kinds <strong>of</strong> branches:<br />
The “Gymnasium”, the “Realgymnasium” and the<br />
“wirtschaftskundliche Realgymnasium”. After the sixth grade,<br />
students have to decide for one <strong>of</strong> these three branches. All <strong>of</strong><br />
them follow almost the same curriculum, except a few specializations<br />
for each branch. The main focus <strong>of</strong> the “Gymnasium” is on<br />
languages, whereas the “Realgymnasium” goes deeper into mathematics<br />
and natural sciences. The “wirtschaftkundliche Realgymnasium”,<br />
specializes in economics, chemistry, psychology and<br />
philosophy. However, these three types are still targeted on general<br />
education and cannot be compared to e.g. the BHS, which are<br />
much more specialized schools aiming for a certain career path.<br />
6.1.2 HTLs for Informatics and for Chemistry<br />
This type <strong>of</strong> school (“Höhere Technische Lehranstalt”) is mostly<br />
associated with a strong technical and scientific education. However,<br />
there are numerous fields <strong>of</strong> specialization like Computer<br />
Science, Business Administrations, Chemistry, IT, Networking,<br />
Mechanics, Design and many more. We analyzed the curricula <strong>of</strong><br />
two branches that might represent the opposite end <strong>of</strong> the CSE<br />
spectrum: First, the branch dedicated specifically to computer<br />
sciences, and second, the branch for Chemistry that has probably<br />
the minimal CSE implementation. These two extreme values<br />
should delimit the range <strong>of</strong> CSE in all the HTL curricula.<br />
134<br />
Because <strong>of</strong> the large variety <strong>of</strong> specializations, each branch <strong>of</strong><br />
HTL has its own specialized curriculum. However, there is also a<br />
basic/core curriculum for all <strong>of</strong> the branches which describes the<br />
respective competencies that the students should have by the end<br />
<strong>of</strong> their studies. The specialized branch-curricula then make<br />
changes and amendments to the basic/core curriculum in or<strong>der</strong> to<br />
put importance on the respective focus <strong>of</strong> the branch.<br />
6.1.3 HLW<br />
The “Höhere Lehranstalt für Wirtschaftliche Berufe” has its focus<br />
clearly on business administrations and tourism, although other<br />
specializations exist. This curriculum should serve as a counterexample<br />
to the very technical HTL curricula.<br />
Other schools like HAK (Handelsakademien) or Tourism schools<br />
have different specializations, though they cover pretty much the<br />
same computer science topics, as HLWs. For this, and timesaving<br />
reasons, only the curriculum for HLWs will be sampled<br />
out. Since 2009, the curriculum regarding computer science<br />
courses got revised and adapted. Actuality is very important and,<br />
therefore, the revised curriculum <strong>of</strong> 2009 will serve as the basis<br />
for comparison. The new curriculum proposes two mandatory<br />
computer science courses for all students attending a HLW. These<br />
two are titled Office- and Information-Management and Applied<br />
Informatics.<br />
6.2 Selection <strong>of</strong> Standards<br />
When looking at the subset <strong>of</strong> selected curricula, it becomes<br />
obvious that all <strong>of</strong> them are located at the higher secondary level,<br />
from the 9 th to the 12 th /13 th grade. Comparing this to the CSTA<br />
standards framework, it seems suitable to limit the regarded set <strong>of</strong><br />
standards to level 3A as, according to the CSTA, the standards <strong>of</strong><br />
this level should be implemented throughout the whole range <strong>of</strong><br />
upper secondary schools. Level 3B and 3C already represent indepth<br />
specializations that are not to be intended by schools <strong>of</strong> all<br />
types, but only at the ones that focus on computer sciences. In<br />
consequence, as the viewpoint <strong>of</strong> this paper lies on general education,<br />
we will restrict our investigations to the competencies subset<br />
<strong>of</strong> level 3A.<br />
The table 8 in the appendix lists the standards <strong>of</strong> CSTA Level 3A,<br />
which we have chosen as a reference for our research about the<br />
adoption by the curricula <strong>of</strong> the selected school types listed<br />
above.<br />
6.3 The Measure <strong>of</strong> Adoption<br />
To measure the degree <strong>of</strong> the adoption <strong>of</strong> the standards by the<br />
curricula, we had to define, un<strong>der</strong> which condition a standard can<br />
be consi<strong>der</strong>ed as implemented.<br />
It has to be mentioned that it is not possible to look for literal<br />
incorporations, since, first, Austrian curricula are not written or<br />
available in English and, secondly, this would not detect all forms<br />
<strong>of</strong> implementations. Thus it seems more suitable to look for<br />
equivalents in content, meaning that the described competencies<br />
have to be featured in the respective curricula in some respect.<br />
For this, it seems appropriate to not only rate if standards were<br />
fully implemented or not, but also whether the standard has been<br />
implemented partially. For this, we applied the scoring scale that<br />
is shown in table 1.
Table 1. Scoring Scale for the Implementation <strong>of</strong> a Standard<br />
Rating Description<br />
The specific competence is implemented in the<br />
1 respective curriculum more or less totally (i.e. the<br />
specific competence is described in the curriculum).<br />
The specific competence is implemented partially<br />
0.5 in the respective curriculum (i.e. something similar<br />
bot not equivalent can be found in the curriculum).<br />
The specific competence is not at all implemented<br />
in the respective curriculum (i.e. nothing similar to<br />
0<br />
the specific competence can be found in the curriculum).<br />
6.4 The Rating Process<br />
To ensure a fair amount <strong>of</strong> objectivity and reliability, all ratings<br />
were cross-examined and double checked by the two student<br />
authors <strong>of</strong> this paper, who generally worked in parallel on separate<br />
parts <strong>of</strong> the curricula. Thereby, the rating itself happened<br />
exclusively on the basis <strong>of</strong> the statements featured in the curricula.<br />
Due to the sometimes very universal statements in the curricula,<br />
it was occasionally difficult to decide which rating to apply.<br />
As an example we will draw on the standard 4.2 Develop criteria<br />
for purchasing or upgrading computer system hardware (cf. table<br />
8 in appendix). The HTL core curriculum states “assess a PC<br />
configuration and make purchasing decisions” 1 which is basically<br />
the same as specified by the respective standard. Consequently,<br />
the rating 1 was applied. The HLW curriculum, however, only<br />
states “hardware and s<strong>of</strong>tware requirements” 2 . We decided that,<br />
although one could infer the ability to make purchasing decisions<br />
from someone’s knowledge <strong>of</strong> the hardware requirements, this<br />
was too general and that therefore only the rating 0.5 should be<br />
applied. By and large, we judged strictly and usually opted for<br />
giving the worse rating in bor<strong>der</strong>line cases.<br />
7. RESULTS AND DISCUSSION<br />
First, we present the results <strong>of</strong> our investigation for the selected<br />
school types. This is followed by a comparison and summary <strong>of</strong><br />
the results <strong>of</strong> all types.<br />
7.1 Results for the Selected School Types<br />
7.1.1 AHS<br />
For the purpose <strong>of</strong> further elaborations, only the curricula <strong>of</strong> the<br />
“Oberstufe” <strong>of</strong> AHS will be investigated in or<strong>der</strong> to compare and<br />
contrast them with the standards <strong>of</strong> level 3A, as they are recommended<br />
approximately for the same age levels and grades. However,<br />
computer science courses are only mandatory in the ninth<br />
grade only. Students in grade 10 to 12 are <strong>of</strong>fered elective computer<br />
science courses, which is referred to as “Wahlpflichtfach”<br />
and follows a separate curriculum, which is included in the analysis<br />
as well.<br />
Table 2. Selected Standards in the AHS Curriculum<br />
Standard<br />
No.<br />
Strand Strand Strand Strand<br />
CT CO CP CC<br />
x=1 x=2 x=3 x=4 x=5<br />
x.1 0 0 0 0 0<br />
x.2 0 0.5 0 0 0<br />
x.3 0 0.5 0 0.5 0<br />
Strand<br />
CG<br />
1<br />
German: “eine PC-Konfiguration bewerten und Anschaffungsentscheidungen<br />
treffen“<br />
2<br />
German: „Hard- und S<strong>of</strong>twareanfor<strong>der</strong>ungen“<br />
135<br />
x.4 0.5 0 0 0 1<br />
x.5 0<br />
0 0 0<br />
x.6 0.5 0.5 0 0.5<br />
x.7 0 0 0 0<br />
x.8 0.5 0 0.5 0<br />
x.9 0.5 0 0 0<br />
x.10 0 1 0.5 0.5<br />
x.11 0 0.5<br />
0<br />
x.12 0<br />
Sum 2 / 11 1 / 4 2 / 12 1.5 / 10 2 / 11<br />
Strand (18%) (25%) (17%) (15%) (18%)<br />
Overall<br />
Sum<br />
8.5 / 48 (18%)<br />
The results are shown in table 2, where e.g. 2/11 represents ”score<br />
2 <strong>of</strong> possible 11”. The abbreviations <strong>of</strong> the strands and the numbers<br />
<strong>of</strong> the standards are listed in table 8 in the appendix. It turned<br />
out that only a small number (18%) <strong>of</strong> recommended CSTA level<br />
3A standards can be regarded as implemented in the Austrian<br />
AHS curriculum. As the AHS curriculum is kept quite short and<br />
generally formulated, it is difficult to compare it to the standards<br />
that are described very precisely. However, the curriculum frequently<br />
mentions basic skills and general concepts the students<br />
should gain, without going too much into depth. The main focus<br />
<strong>of</strong> AHS lies on developing abstract thinking, making use <strong>of</strong> information<br />
technology and creating a general picture <strong>of</strong> computation<br />
and its impacts on society (http://www.bmukk.gv.at/).<br />
7.1.2 HTLs for Informatics and for Chemistry<br />
Initially we will rate the implementation <strong>of</strong> the CSTA level 3A<br />
standards in the basic/core curriculum, before rating the specialized<br />
curricula for Chemistry and Informatics. The curricula give<br />
rather detailed information on the required competencies and<br />
skills and, therefore, were rather straightforward to compare with<br />
the standards. The results <strong>of</strong> this process are depicted in table 3,<br />
table 4 and table 5.<br />
Table 3. Selected Standards in the Core HTL Curriculum<br />
Stand- Strand Strand Strand Strand Strand<br />
ard No. CT CO CP CC CG<br />
x.1 0.5 0 0.5 0 0<br />
x.2 0 0 0 1 0.5<br />
x.3 0.5 0 0 1 0<br />
x.4 0 0 0 0 1<br />
x.5 0.5<br />
0 0 0<br />
x.6 0 0 0.5 0.5<br />
x.7 0.5 0.5 0 1<br />
x.8 0 0.5 1 0<br />
x.9 0 0 0.5 0.5<br />
x.10 0 0 0 0.5<br />
x.11 0 0<br />
0<br />
x.12 0<br />
Sum 2 / 11 0 / 4 1.5 / 12 4 / 10 4 / 11<br />
Strand (18%) (0%) (13%) (40%) (36%)<br />
Overall<br />
Sum<br />
11.5 / 48 (24%)<br />
Table 4. Selected Standards in the HTL for Chemistry<br />
Stand- Strand Strand Strand Strand Strand<br />
ard No. CT CO CP CC CG<br />
x.1 0.5 0 0.5 0 0<br />
x.2 0 0 0 1 0.5<br />
x.3 0.5 0 0 1 0
x.4 0 0 0 0 1<br />
x.5 0.5<br />
0 0 0<br />
x.6 0 0 0.5 0.5<br />
x.7 0.5 0.5 0 1<br />
x.8 0 0.5 1 0<br />
x.9 0 0 0.5 0.5<br />
x.10 0 0 0 0.5<br />
x.11 0 0<br />
0<br />
x.12 0<br />
Sum 2 / 11 0 / 4 1.5 / 12 4 / 10 4 / 11<br />
Strand (18%) (0%) (13%) (40%) (36%)<br />
Overall<br />
Sum<br />
11.5 / 48 (24%)<br />
Table 5. Selected Standards in the in the HTL for Informatics<br />
Stand- Strand Strand Strand Strand Strand<br />
ard No. CT CO CP CC CG<br />
x.1 1 0 1 0.5 0<br />
x.2 1 0.5 0 1 0.5<br />
x.3 1 0 1 1 0<br />
x.4 0.5 0 1 0.5 1<br />
x.5 1<br />
0.5 1 0<br />
x.6 1 1 0.5 0.5<br />
x.7 1 0.5 1 1<br />
x.8 0.5 0.5 1 0<br />
x.9 0.5 1 0.5 0.5<br />
x.10 1 0 0 1<br />
x.11 0 0.5<br />
0<br />
x.12 0<br />
Sum 8.5 / 11 0.5 / 4 7 / 12 7 / 10 4.5 / 11<br />
Strand (77%) (13%) (58%) (70%) (41%)<br />
Overall<br />
Sum<br />
27.5 / 48 (57%)<br />
As the tables 3, 4 and 5 show, there is no difference at all between<br />
the basic/core curriculum (25%) and the specialization in Chemistry.<br />
However, the Informatics curriculum specifies a lot more<br />
skills and competencies in terms <strong>of</strong> computer science education<br />
than the other two curricula do. Consequently, the degree <strong>of</strong><br />
implementation rises from 25% to 56%. Since Chemistry and<br />
Informatics certainly represent two extreme points on a continuum,<br />
one can interpolate these scores in or<strong>der</strong> to estimate the implementation<br />
percentage <strong>of</strong> an average HTL curriculum.<br />
7.1.3 HLW<br />
Finally we analyze the curriculum <strong>of</strong> the HLW. The results are<br />
displayed in table 6.<br />
Table 6. Selected Standards in the HLW Curriculum<br />
Stand- Strand Strand Strand Strand Strand<br />
ard No. CT CO CP CC CG<br />
x.1 0 0 0.5 0 0<br />
x.2 0 0.5 0 0.5 0.5<br />
x.3 0 0 0 0.5 0<br />
x.4 0.5 0 0 0 1<br />
x.5 0<br />
0 0 0.5<br />
x.6 0.5 0 0.5 0.5<br />
x.7 0 0 0.5 0.5<br />
x.8 0 0 1 0<br />
x.9 0 0 0 0<br />
x.10 0 0 0 0.5<br />
x.11 0 0<br />
0<br />
x.12 0<br />
136<br />
Sum<br />
Strand<br />
Overall<br />
Sum<br />
1 / 11<br />
(9%)<br />
0.5 / 4<br />
(13%)<br />
0.5 / 12<br />
(4%)<br />
8.5 / 48 (18%)<br />
3 / 10<br />
(30%)<br />
3.5 / 11<br />
(32%)<br />
As table 6 shows, the implementation <strong>of</strong> the CSTA standards in<br />
the HLW is quite poor. It only scores an overall sum <strong>of</strong> 16% over<br />
all five strands. Similar to the AHS, programming skills are ignored<br />
in the HLW curriculum. It mainly focuses on <strong>of</strong>fice- and<br />
information-management and user skills, which students may<br />
need for future pr<strong>of</strong>essions in the field <strong>of</strong> economics, tourism<br />
and/or commerce. Moreover, the HLW’s curriculum demands a<br />
lot <strong>of</strong> general concepts and ideas on computer science from students,<br />
which are, in some cases, quite difficult to interpret. According<br />
to the curriculum, students should, for example, be able to<br />
master a current operating system, which is very complex to<br />
define. Knowing or mastering an operating system might imply to<br />
un<strong>der</strong>stand the general concepts and how to use it, but might also<br />
go far deeper into other subject matters. Personal experience and<br />
the overall structure <strong>of</strong> the curriculum, led to a very general interpretation<br />
<strong>of</strong> this and similar statements. Therefore, it is assumed<br />
that certain declarations, such as the one mentioned above, do not<br />
purport to focus on depth but on giving overviews related to<br />
practice. However, this fact might have also diminished the results,<br />
as the CSTA standards for level 3A, are partly a lot more<br />
specialized and ask for in-depth knowledge as well.<br />
7.2 Comparison <strong>of</strong> the Curricula<br />
Following the isolated inspection <strong>of</strong> the selected curricula, it<br />
seems interesting to compare the scores <strong>of</strong> the curricula. Table 8<br />
clearly depicts the differences in implementation <strong>of</strong> the CSTA<br />
standards in the respective curricula.<br />
Table 7. CSTA Standards in Curricula Selection<br />
Strands AHS<br />
(1) CT<br />
(2) CO<br />
(3) CP<br />
(4) CC<br />
(5) CG<br />
Sum<br />
Average<br />
2 / 11<br />
(18%)<br />
1 / 4<br />
(25%)<br />
2 / 12<br />
(17%)<br />
1.5 / 10<br />
(15%)<br />
2 / 11<br />
(18%)<br />
8.5 / 48<br />
(18%)<br />
HTL<br />
HTL<br />
Chemistry Informatics<br />
2 / 11 8.5 / 11<br />
(18%) (77%)<br />
0 / 4<br />
0.5 / 4<br />
(0%)<br />
(13%)<br />
1.5 / 12 7 / 12<br />
(13%) (58%)<br />
4 / 10 7 / 10<br />
(40%) (70%)<br />
4 / 11 4.5 / 11<br />
(36%) (41%)<br />
11.5 / 48 27.5 / 48<br />
(24%) (57%)<br />
14 / 48<br />
(29%)<br />
HLW<br />
1 / 11<br />
(9%)<br />
0.5 / 4<br />
(13%)<br />
0.5 / 12<br />
(4%)<br />
3 / 10<br />
(30%)<br />
3.5 / 11<br />
(32%)<br />
8.5 / 48<br />
(18%)<br />
Note: Since the basic/core HTL curriculum is the same as<br />
the one for the specialization in Chemistry, only the latter<br />
is featured in the table above.<br />
As expected beforehand, the AHS, being the school with the most<br />
general background and aim in education, also features the lowest<br />
score (8.5/48). Likewise, the HLW, since specialized in business<br />
and tourism, does not exceed this score (8.5/48). The HTL curriculum<br />
for Chemistry, which has the same features as the basic/core<br />
HTL curriculum, tops them by a few points (11.5/48) which might<br />
be due to the more technical focus <strong>of</strong> the school. However, the<br />
HTL for Informatics is the only specimen that manages to break<br />
through the 50% barrier (27.5/48) and is thus the school type with
the highest rate <strong>of</strong> implementation <strong>of</strong> the CSTA Level 3A standards.<br />
Calculating an average score from these four curricula gives<br />
us a mean rating <strong>of</strong> 14/48 (29%).<br />
Figure 5. Overview <strong>of</strong> results.<br />
Looking at the relative adoption <strong>of</strong> the 5 strands, it is manifest<br />
that the HTL for Informatics has percentages over 50% regarding<br />
most <strong>of</strong> the strands except Collaboration (only 13%) and Community,<br />
global and ethical impacts (41%). Oppositely, at the AHS<br />
Collaboration has the highest scores and therefore seems to be<br />
encouraged there much more. On the other hand, at the HTL for<br />
Chemistry and the HLW, the strands Computers and Communications<br />
Devices (CC) and Community, global and ethical impacts<br />
(CG) produced the highest percentages, indicating the very different<br />
focuses <strong>of</strong> these schools compared to AHS. As a whole, the<br />
results reflect the schools’ top priorities (or pr<strong>of</strong>iles) that are<br />
generally associated with them (cf. section 4), i.e. the HTL has a<br />
very strong technical markedness, the AHS a stronger emphasis<br />
on collaborative and social aspects, and the HLW focusses on<br />
communication, international and global aspects.<br />
8. CONCLUSION AND FUTURE WORK<br />
The purpose <strong>of</strong> this paper was to demonstrate to which degree a<br />
subset <strong>of</strong> CSTA Level 3 standards (i.e. level 3A) is implemented<br />
in a selection <strong>of</strong> Austrian school curricula, which lead to a clear<br />
and unambiguous result: The incorporation <strong>of</strong> CSTA standards<br />
into Austrian school curricula is, to a very large degree, unsatisfying.<br />
Even the HTL for informatics, with its special focus on computer<br />
sciences, does not reach more than 57% (27.5/8) adoption<br />
rate. This seems even more disappointing when consi<strong>der</strong>ing that<br />
only the Level 3A standards were compared to the Austrian curricula<br />
and the more elaborate standards <strong>of</strong> Level 3B and 3C,<br />
which feature more in-depth competencies, were not taken into<br />
account at all.<br />
However, it has to be consi<strong>der</strong>ed that the Austrian curricula do<br />
not go into much detail when describing the required competencies<br />
and skills, but use rather abstract terms. This makes it quite<br />
difficult to compare the (English) CSTA standards to the (German)<br />
Austrian curricula since a lot <strong>of</strong> freedom, in what specifically<br />
to teach, is still present.<br />
Prospective future research in this area could involve a more<br />
exhaustive elaboration <strong>of</strong> curricula from other school types or<br />
countries. In doing so, it would be necessary to broaden the range<br />
<strong>of</strong> curricula to be examined in the course <strong>of</strong> the rating process.<br />
Furthermore, the rating scale/key should be redefined more precisely<br />
in or<strong>der</strong> to assure a more objective scoring. Also, the scoring<br />
should be done by more people to ensure the objectiveness<br />
and reliability <strong>of</strong> the whole process. Last, but not least, it seems<br />
137<br />
somewhat obvious that also the rest <strong>of</strong> the CSTA standards (1, 2,<br />
3B, 3C) could/should be compared to the respective curricula <strong>of</strong><br />
other levels <strong>of</strong> education. To our regret, it was not possible to<br />
elaborate on these aspects in more detail due to the limited time<br />
and resources un<strong>der</strong> which this paper has been produced.<br />
9. REFERENCES<br />
[1] Betts, J. 1998. The Impact <strong>of</strong> Educational Standards on the<br />
Level and Distribution <strong>of</strong> Earnings. The American Economic<br />
Review 88, 1, 266–275.<br />
[2] CSTA – Computer Science Teachers Association. 2010.<br />
Running On Empty. State-by-State Results.<br />
http://www.acm.org/runningonempty/roemap.html. Accessed<br />
22 June 2012.<br />
[3] Hubwieser, P., Armoni, M., Brinda, T., Dagiene, V., Diethelm,<br />
I., Giannakos, M. N., <strong>Knobelsdorf</strong>, M., Magenheim, J.,<br />
Mittermeir, R., and Schubert, S. 2011. Computer science/informatics<br />
in secondary education. In Proceedings <strong>of</strong><br />
the 16th annual conference reports on Innovation and technology<br />
in computer science education - working group reports.<br />
ITiCSE-WGR ’11. ACM, New York, NY, USA,<br />
19‐38.<br />
[4] Klieme, E., Avenarius, H., Blum, W., Döbrich, P., Gruber,<br />
H., Prenzel, M., Reiss, K., Riquarts, K., Rost, J., Tenorth, H.-<br />
E., and Vollmer, H. J. 2004. The Development <strong>of</strong> National<br />
Educational Standards. An Expertise. Bundesministerium für<br />
Bildung und Forschung, Berlin.<br />
[5] Micheuz, P. 2008. Harmonization <strong>of</strong> Informatics Education -<br />
Science Fiction or Prospective Reality? In Informatics Education<br />
- Supporting Computational Thinking, Third International<br />
Conference on Informatics in Secondary Schools -<br />
Evolution and Perspectives, ISSEP 2008, Torun, Poland, July<br />
1-4, 2008. Lecture notes in computer science. Springer,<br />
317–326.<br />
[6] Norcini, J. J. 2003. Setting standards on educational tests.<br />
Medical Education 37, 5, 464‐469.<br />
[7] Reigeluth, C. M. 1997. Educational Standards: To Standardize<br />
or to Customize Learning? Phi Delta Kappan 78, 3, 202–<br />
206.<br />
[8] Tucker, A., Deek, F., Jones, J., McCowan, D., Stephenson,<br />
C., and Verno, A. 2006. A . Final Report <strong>of</strong> the ACM K–12<br />
Task Force Curriculum Committee, New York.<br />
[9] Tucker, A., Ed. 2003. A . Final Report <strong>of</strong> the ACM K–12<br />
Task Force Curriculum Committee October 2003, New<br />
York.<br />
[10] Tucker, A., Seehorn, D., Carey, S., Moix, D., Fuschetto, B.,<br />
Lee, I., O’Grady-Cuniff, D., Stephenson, C., and Verno, A.<br />
2011. CSTA K-12 Computer Science Standards. Revised<br />
2011. CSTA Standards Task Force.<br />
[11] Wilson, C., Sudol, L. A., Stephenson, C., and Stehlik, M.<br />
2010. Running on Empty. Executive Summary.<br />
http://csta.acm.org/runningonempty/fullreport.pdf. Accessed<br />
21 June 2011.<br />
[12] Wilson, C., Sudol, L. A., Stephenson, C., and Stehlik, M.<br />
2010. Running on Empty.<br />
http://csta.acm.org/runningonempty/fullreport.pdf. Accessed<br />
21 June 2011
APPENDIX<br />
Table 8. CSTA Level 3A Standards [10]<br />
No. Level 3A Standards<br />
Strand 1: Computational Thinking (CT)<br />
1.1 Use predefined functions and parameters, classes and<br />
methods to divide a complex problem into simpler<br />
parts.<br />
1.2 Describe a s<strong>of</strong>tware development process used to solve<br />
s<strong>of</strong>tware problems (e.g., design, coding, testing, verification).<br />
1.3 Explain how sequence, selection, iteration, and recursion<br />
are building blocks <strong>of</strong> algorithms.<br />
1.4 Compare techniques for analyzing massive data collections.<br />
1.5 Describe the relationship between binary and hexadecimal<br />
representations.<br />
1.6 Analyze the representation and trade-<strong>of</strong>fs among<br />
various forms <strong>of</strong> digital information.<br />
1.7 Describe how various types <strong>of</strong> data are stored in a<br />
computer system.<br />
1.8 Use modeling and simulation to represent and un<strong>der</strong>stand<br />
natural phenomena.<br />
1.9 Discuss the value <strong>of</strong> abstraction to manage problem<br />
complexity.<br />
1.10 Describe the concept <strong>of</strong> parallel processing as a strategy<br />
to solve large problems.<br />
1.11 Describe how computation shares features with art<br />
and music by translating human intention into an<br />
artifact.<br />
Strand 2: Collaboration (CO)<br />
2.1 Work in a team to design and develop a s<strong>of</strong>tware<br />
artifact.<br />
2.2 Use collaborative tools to communicate with project<br />
team members (e.g., discussion threads, wikis, blogs,<br />
version control, etc.).<br />
2.3 Describe how computing enhances traditional forms<br />
and enables new forms <strong>of</strong> experience, expression,<br />
communication, and collaboration.<br />
2.4 Identify how collaboration influences the design and<br />
development <strong>of</strong> s<strong>of</strong>tware products.<br />
Strand 3: Computing Practice and Programming (CP)<br />
3.1 Crate and organize Web pages through the use <strong>of</strong> a<br />
variety <strong>of</strong> web programming design tools.<br />
3.2 Use mobile devices/emulators to design, develop, and<br />
implement mobile computing applications.<br />
3.3 Use various debugging and testing methods to ensure<br />
program correctness (e.g., test cases, unit testing,<br />
whit box, block box, integration testing)<br />
3.4 Apply analysis, design, and implementation techniques<br />
to solve problems (e.g., use one or more s<strong>of</strong>tware<br />
lifecycle models).<br />
3.5 Use Application Program Interfaces (APIs) and libraries<br />
to facilitate programming solutions.<br />
3.6 Select appropriate file formats for various types and<br />
uses <strong>of</strong> data.<br />
3.7 Describe a variety <strong>of</strong> programming languages available<br />
to solve problems and develop systems.<br />
3.8 Explain the program execution process.<br />
3.9 Explain the principles <strong>of</strong> security by examining encryption,<br />
cryptography, and authentication techniques.<br />
3.10 Explore a variety <strong>of</strong> careers to which computing is<br />
138<br />
No. Level 3A Standards<br />
central.<br />
3.11 Describe techniques for locating and collecting small<br />
and large-scale data sets.<br />
3.12 Describe how mathematical and statistical functions,<br />
sets, and logic are used in computation.<br />
Strand 4: Computers and Communications Devices (CC)<br />
4.1 Describe the unique features <strong>of</strong> computers embedded<br />
in mobile devices and vehicles (e.g., cell phones, automobiles,<br />
airplanes).<br />
4.2 Develop criteria for purchasing or upgrading computer<br />
system hardware.<br />
4.3 Describe the principal components <strong>of</strong> computer organization<br />
(e.g., input, output, processing, and storage).<br />
4.4 Compare various forms <strong>of</strong> input and output.<br />
4.5 Explain the multiple levels <strong>of</strong> hardware and s<strong>of</strong>tware<br />
that support program execution (e.g., compilers, interpreters,<br />
operating systems, networks).<br />
4.6 Apply strategies for identifying and solving routine<br />
hardware and s<strong>of</strong>tware problems that occur in everyday<br />
life.<br />
4.7 Compare and contrast client-server and peer-to-peer<br />
network strategies.<br />
4.8 Explain the basic components <strong>of</strong> computer networks<br />
(e.g., servers, file protection, routing, spoolers and<br />
queues, shared resources, and fault-tolerance).<br />
4.9 Describe how the Internet facilitates global communication.<br />
4.10 Describe the major applications <strong>of</strong> artificial intelligence<br />
and robotics.<br />
Strand 5: Community, Global, and Ethical Impacts (CG)<br />
5.1 Compare appropriate and inappropriate social networking<br />
behaviors.<br />
5.2 Discuss the impact <strong>of</strong> computing technology on business<br />
and commerce (e.g., automated tracking <strong>of</strong> goods,<br />
automated financial transactions, e-commerce, cloud<br />
computing).<br />
5.3 Describe the role that adaptive technology can play in<br />
the lives <strong>of</strong> people with special needs.<br />
5.4 Compare the positive and negative impacts <strong>of</strong> technology<br />
on culture (e.g., social networking, delivery <strong>of</strong><br />
news and other public media, and intercultural communication).<br />
5.5 Describe strategies for determining the reliability <strong>of</strong><br />
information found on the Internet.<br />
5.6 Differentiate between information access and information<br />
distribution rights.<br />
5.7 Describe how different kinds <strong>of</strong> s<strong>of</strong>tware licenses can<br />
be used to share and protect intellectual property.<br />
5.8 Discuss the social and economic implications associated<br />
with hacking and s<strong>of</strong>tware privacy.<br />
5.9 Describe different ways in which s<strong>of</strong>tware is created<br />
and shared and their benefits and drawbacks (commercial<br />
s<strong>of</strong>tware, public domain s<strong>of</strong>tware, open source<br />
development).<br />
5.10 Describe security and privacy issues that relate to<br />
computer networks.<br />
5.11 Explain the impact <strong>of</strong> the digital divide on access to<br />
critical information.
Information Theory on Czech Grammar Schools: First<br />
Findings<br />
ABSTRACT<br />
Daniel Lessner<br />
Department <strong>of</strong> S<strong>of</strong>tware and Computer Science Education,<br />
Faculty <strong>of</strong> Mathematics and Physics,<br />
Prague, Czech Republic<br />
lessner@ksvi.mff.cuni.cz<br />
Computing science is not a part <strong>of</strong> standard grammar school<br />
(secondary) education in the Czech Republic. As a part <strong>of</strong><br />
an effort to change this, we have developed an introductory<br />
computing science course.<br />
It consists <strong>of</strong> multiple month long modules. The first module<br />
deals with the notion <strong>of</strong> information. Students are expected<br />
to un<strong>der</strong>stand it sufficiently for comparing informational<br />
gain <strong>of</strong> statements or questions. Along with binary<br />
numeral system, they are expected to un<strong>der</strong>stand the relation<br />
between data length and the amount <strong>of</strong> information<br />
stored.<br />
In this paper we describe the module itself and present the<br />
results from the first run <strong>of</strong> experimental teaching. The aim<br />
was to find out the limits <strong>of</strong> such topic on grammar schools<br />
and suitability <strong>of</strong> chosen teaching methods.<br />
It turned out that no matter how abstract is the term<br />
information itself, students are capable <strong>of</strong> acquiring it deep<br />
enough so that it would be beneficial for them.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers and education]: Computer and Information<br />
Science Education—Curriculum, computer science<br />
education<br />
General Terms<br />
Human factors, Experimentation<br />
Keywords<br />
Secondary education, computing science, information theory<br />
1. INTRODUCTION<br />
In this paper we first briefly discuss the unfortunate situation<br />
<strong>of</strong> computing science (CS) education on Czech grammar<br />
schools 1 . Then we describe a basic CS course developed with<br />
1 General secondary education, students aged 15 – 19.<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WiPSCE ’12 Hamburg, <strong>Germany</strong><br />
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$10.00.<br />
139<br />
an intent to change this state. This course is tested with two<br />
seminar groups. The aim <strong>of</strong> this experimental teaching is to<br />
find out whether it is even possible to teach CS on a grammar<br />
school level in Czechia.<br />
We have decided to focus this paper on the first module,<br />
which helps students study information. We describe<br />
its specific objectives and content. The necessary aspects<br />
such as the course organization and students characteristics<br />
are described in a separate section. The last section finally<br />
shows the results and findings and their interpretation.<br />
Grammar schools in the Czech Republic are one <strong>of</strong> the<br />
secondary education branches. The curriculum is defined in<br />
the Framework Education Programme for Secondary General<br />
Education (FEP) [1]. From it the individual school programmes<br />
are <strong>der</strong>ived. According to FEP, grammar schools<br />
shall prepare students for university studies. This shall<br />
be accomplished through development <strong>of</strong> key competences.<br />
FEP defines six <strong>of</strong> them: to solve problems, to communicate,<br />
to learn, civic, entrepreneurial and personal and social<br />
competence. Classical subjects are means to develop these<br />
key competences 2 .<br />
We use the term computing science for the general area<br />
<strong>of</strong> efficient (i.e. automated) information processing. It has<br />
very little to do with digital literacy. It is also not programming.<br />
Computer is merely the tool, not the subject.<br />
A similar view can be seen in Computer Science Unplugged<br />
from New Zealand [2]. It does not employ computers by its<br />
definition, what certainly says enough about its relation to<br />
digital literacy or programming.<br />
The K-12 CS curriculum is not widely applied [5], but<br />
serves as a good prototype and reference. Given a much<br />
longer time frame, the programme can afford to be much<br />
more complex then the other examples here. The goals are<br />
similar to ours: to introduce basics <strong>of</strong> CS on an appropriate<br />
level. All topics included in our course except for a few<br />
details are covered in this curriculum.<br />
FEP casually mentions a few CS topics. It includes so<br />
called informatics 3 as a compulsory subject, which is focused<br />
on using digital technology. There is no systematic approach<br />
to efficient information processing.However, consi<strong>der</strong>ing the<br />
grammar school educational goals and the concept <strong>of</strong> key<br />
competences, CS seems to be very promising. Moreover, it<br />
2 Yet they are not unimportant – to communicate we need<br />
languages, to solve problem sciences, to be good citizen history<br />
etc.<br />
3 Informatics in Czech is an umbrella term, which can mean<br />
anything related to computers or digital technology, depending<br />
on the current context.
is a part <strong>of</strong> general knowledge. CS is shaping our world<br />
significantly and most <strong>of</strong> grammar school graduates are not<br />
even aware <strong>of</strong> its existence.<br />
We consi<strong>der</strong> the absence <strong>of</strong> CS in grammar schools a mistake.<br />
We believe we are missing here a fruitful opportunity<br />
to achieve secondary education goals easier. In an imaginary<br />
ideal case, CS is a regular school subject, such as physics or<br />
history. More thorough reasons are given in [3]. We intend<br />
to develop, carry out and evaluate a basic CS course to have<br />
a solid base for a possible future discussion about including<br />
CS into grammar schools.<br />
This paper presents results from one course module. The<br />
question is, whether it is possible to teach information and<br />
to what extent. We have planned the module (objectives,<br />
content, activities, assignments, final test) and carried it out<br />
with two seminar groups.<br />
2. INTRODUCTORY CS COURSE<br />
In an effort to move towards our above mentioned ideas,<br />
we have developed a program for a basic CS course. It covers<br />
what we think any grammar school student should know.<br />
General goals <strong>of</strong> the course are <strong>der</strong>ived from secondary education<br />
goals. We must enhance students problem solving<br />
skills along with other key competences and thus prepare<br />
students for studying university, for their career and for a<br />
life in our society [1].<br />
Further, as with other sciences, the aim <strong>of</strong> their study<br />
on grammar schools is to get familiar with their subjects,<br />
methods, connections to other areas, fundamental results<br />
and their implications and applications. Last but not least,<br />
students shall be aware <strong>of</strong> some fundamental limits to CS,<br />
such as the existence <strong>of</strong> practically or computationally unsolvable<br />
problems. Students shall be able to actually use the<br />
basics <strong>of</strong> CS in everyday life.<br />
Based on the usual situation on our grammar schools, we<br />
have planned the course to last one school year, 90 minutes<br />
a week. It is structured into approximately one month (four<br />
lessons) lasting modules. Each module deals explicitly with<br />
one fundamental idea, while other ideas are present in the<br />
background. The main topics chosen for individual modules<br />
are information, graph, problem, algorithm and efficiency.<br />
More detailed information can be found in [4].<br />
3. INFORMATION: MODULE PLAN<br />
Information is the first module out <strong>of</strong> 8 in the course plan.<br />
The word informatics used in Czech is based on the word information,<br />
so it seems logical to deal with it in the beginning<br />
<strong>of</strong> the course. But most importantly, this module contains<br />
many notions that will come later (such as algorithm, efficiency<br />
or recursion) in a simple and conceivable form.<br />
3.1 Objectives<br />
The basic objective to achieve is un<strong>der</strong>standing <strong>of</strong> the term<br />
information on an appropriate level and un<strong>der</strong>standing <strong>of</strong><br />
what is it good for.Along with the term information itself,<br />
there are others to be un<strong>der</strong>stood: communication. message,<br />
data, knowledge, encoding, language, reliability, bit.<br />
Further, students shall be able to calculate the amount <strong>of</strong><br />
information contained in a message.<br />
By a step higher is efficient querying in a given situation.<br />
This goes along with the concept <strong>of</strong> halving the set<br />
<strong>of</strong> possibilities. Students shall un<strong>der</strong>stand the concept <strong>of</strong><br />
140<br />
decision tree, use it properly, judge its efficiency and optimize<br />
it in simple cases (including non-uniform distribution<br />
<strong>of</strong> answers)Students shall be able to construct binary codes<br />
for objects and work correctly with the relation between the<br />
number <strong>of</strong> codes and their length.<br />
As the CS course opens with this module, we use it to prepare<br />
and develop concepts for upcoming topics, such as decomposition,<br />
algorithm (a reliable workflow <strong>of</strong> simple steps,<br />
e.g. a decision tree) or efficiency (related to defined criteria).<br />
3.2 Process<br />
Here we describe briefly the educational activities and related<br />
content included in the module. They result from the<br />
above discussed objectives and methods. There is some slack<br />
to vary the activities between each study group.<br />
Students are at first introduced to the course and to the<br />
content and objectives <strong>of</strong> the first module. Then the teacher<br />
helps them realize the importance <strong>of</strong> the term information<br />
and the fact that they actually can not explain what it is.<br />
They are asked to search their resources and come up with<br />
an explanation themselves. Discussing their various results<br />
and opinions, key features and links emerge.<br />
Guessing games<br />
To make things less abstract and vague, the teacher proposes<br />
an animal guessing game. He thinks <strong>of</strong> an animal,<br />
students ask him yes or no questions in an effort to find the<br />
animal as fast as possible. Students intuitively review the efficiency<br />
<strong>of</strong> each question after the game and formulate their<br />
first hypotheses on the optimal strategy. Frequent finding<br />
is for example that direct guessing is useless in the beginning,<br />
negated ”reassuring“ question is useless completely in<br />
noise-free environment, making the teacher answer yes may<br />
be emotionally satisfying, though it is meaningless for the<br />
game.<br />
However, animals turn out to be a rather too complex<br />
model. They are not a well bounded domain, success in<br />
the game depends on some knowledge <strong>of</strong> animals, and it is<br />
surprisingly tricky to answer some questions (e.g. elephants<br />
do live in Africa, as well as in India, and in captivity all over<br />
the world).<br />
That is why we switch to a much better arranged domain –<br />
natural numbers. They still model the situation well enough<br />
to give us important results. Numbers allow students to<br />
clear their minds from animals and focus on the problem<br />
itself. The task is to formulate a good strategy, i.e. find<br />
a set <strong>of</strong> rules to create efficient questions. Some heuristics<br />
emerge eventually in most <strong>of</strong> the pairs. Then we play one<br />
more game together so that each pair can see their results<br />
in practice.<br />
Decision trees and information definitions<br />
One <strong>of</strong> the ways to describe a strategy is a decision tree<br />
scheme. It shows the efficiency also visually. Looking at<br />
a tree scheme, one may work out the formula to the relate<br />
number <strong>of</strong> possibilities and and number <strong>of</strong> questions asked<br />
to isolate one. This finally allow us to quantify information.<br />
Consi<strong>der</strong>ing the scheme, it is practical to measure uncertainty<br />
by number <strong>of</strong> questions necessary to remove it, i.e.<br />
⌈log(|S|)⌉, where S denotes the set <strong>of</strong> possible outcomes.<br />
Base <strong>of</strong> the logarithm is two in case <strong>of</strong> binary questions.<br />
Amount <strong>of</strong> information is then the decrease <strong>of</strong> such uncertainty<br />
after getting the answer. More formally, the amount
<strong>of</strong> information in a message M, represented as a set <strong>of</strong> eliminated<br />
outcomes, is I(M) = log(|S|) − log(|S \ M|). This<br />
approach is a special case <strong>of</strong> the well known approach using<br />
entropy, using uniform probability distribution. This makes<br />
the basic idea <strong>of</strong> reduction <strong>of</strong> uncertainty more accessible as<br />
reduction <strong>of</strong> possible outcomes. This simplified definition is<br />
sufficiently useful to lead to the main ideas behind.<br />
The second definition <strong>of</strong> information we use says that information<br />
consists <strong>of</strong> data and their interpretation. This is<br />
a fruitful starting point to further observations too. It is in<br />
accordance with students intuitive concepts. It can be illustrated<br />
using the decision tree scheme too. Another closer<br />
look to the decision tree for guessing numbers reveals the<br />
tight connection to binary numbers. It also opens the opportunity<br />
to think <strong>of</strong> an informational value <strong>of</strong> an individual<br />
character, e.g. a bit.<br />
Students realize sooner or later that not every animal have<br />
the same chance to take part in the game. The teacher<br />
provides them with some more examples where non-uniform<br />
distribution is important, e.g. assessing stroke risk factor<br />
or other medical diagnosis tasks. This topic leands to a<br />
fundamental heuristic: what comes <strong>of</strong>ten shall be at hand.<br />
Additional activities<br />
The objectives <strong>of</strong> this module are covered with activities<br />
above. Below suggested activities are optional and utilized<br />
mostly based on students interests. They serve to review the<br />
matter and to investigate it from different points <strong>of</strong> view.<br />
They also imply some real world applications. Not all students<br />
in the group must work on the same task <strong>of</strong> these.<br />
The guessing game can be extended for real numbers. It<br />
<strong>of</strong> course deepens the un<strong>der</strong>standing <strong>of</strong> mathematics. For<br />
CS, it opens the question <strong>of</strong> a process finiteness. Another<br />
domain for guessing game may be subsets <strong>of</strong> a given finite<br />
set. Halving the possibilities leads to the simplest encoding<br />
by characteristic function. And again, students reach deeper<br />
un<strong>der</strong>standing <strong>of</strong> sets and subsets, which will be useful later<br />
in the course. An even more advanced strategy is required<br />
to play mastermind.<br />
To remind students that binary questions are just a convention,<br />
we let them solve problems with balance scale. One<br />
kind <strong>of</strong> problem is to find a heavier item in otherwise indistinguishable<br />
group. Another kind is about designing somehow<br />
optimal set <strong>of</strong> masses.<br />
Popular TV shows, such as Who Wants to Be a Millionaire?<br />
can be a good resource. Students can think <strong>of</strong> heuristics<br />
on which kind <strong>of</strong> help shall be used in which situation. It<br />
is easier to explain these heuristics using the new knowledge<br />
<strong>of</strong> information theory. A little bit similar and very openly<br />
defined project deals with designing cheat sheets, where efficiency<br />
is <strong>of</strong> course crucial.<br />
Efficiency <strong>of</strong> information encoding in natural languages<br />
can be investigated by comparing same texts in different<br />
languages. Another question <strong>of</strong> this field is the efficiency<br />
<strong>of</strong> Morse code for different languages, and to find a way to<br />
enhance it, if possible. Also DNA encoding amino acids is<br />
an interesting system to be investigated from the informational<br />
point <strong>of</strong> view. The last one to mention is the START<br />
method (Simple triage and rapid treatment) utilized in mass<br />
disasters.<br />
3.3 Feedback and evaluation<br />
Students work mostly in groups, what allows them to re-<br />
141<br />
ceive continuous feedback on their thoughts from their peers<br />
as well as from other groups. They also consult the teacher<br />
when needed during their work, what allows him to provide<br />
some more informal feedback. Another source <strong>of</strong> feedback<br />
are submitted home assignments, commented by the teacher.<br />
They are usually problematic or require some more intensive<br />
work, which can be done outside the classroom.<br />
Students write a test at the end <strong>of</strong> each module. These<br />
tests are rather difficult. The matter itself is quite abstract,<br />
the tests last the whole lesson (90 minutes) and contain some<br />
unusual tasks. Some aspects on the other hand make the<br />
tests easier. The required skills are stated very clearly in<br />
advance. Further, students may use their lecture notes and<br />
internet resources.<br />
4. EXPERIMENTAL LESSONS<br />
Here we describe the actual teaching to allow the rea<strong>der</strong> to<br />
better un<strong>der</strong>stand the below presented results. We worked<br />
with two groups, denoted A and B. Both groups are <strong>of</strong> 12<br />
students. Majority <strong>of</strong> students does not intend to continue<br />
their carrier in CS anyhow. Their motivation is only to<br />
pass the subject as effortlessly as possible. Almost a half <strong>of</strong><br />
them have signed up for the seminar as for ”the best <strong>of</strong> bad<br />
options“, because they just had to pick one. This fact <strong>of</strong><br />
course hin<strong>der</strong>s teaching CS. On the other hand, it increases<br />
the validity <strong>of</strong> our research. Our target is a regular grammar<br />
school student. The less interested students we work with,<br />
the more valuable results we obtain.<br />
All twelve students from group A are in their last year.<br />
Only one <strong>of</strong> them intends to pass the school-leaving exam<br />
in informatics. Their school specializes on sports. This determines<br />
their attitude to sciences well enough (and again,<br />
it increases the value <strong>of</strong> our results). They all have been<br />
through a one year lasting optional seminar on programming<br />
in Pascal. The reason they took the seminar was the<br />
usual, i.e. picking ”the best <strong>of</strong> bad options“ and the result<br />
corresponds this attitude.<br />
Group B is on an average grammar school with no specific<br />
specialization. Four students are in their last year and<br />
they all intend to pass the school-leaving exam in informatics.<br />
The others (eight students) are one year younger and<br />
their decision is still to be made. The students have various<br />
programming experience.<br />
The module on information took six weeks (90 minutes<br />
teaching time each) instead <strong>of</strong> the originally planned four.<br />
This is mostly due to building in the necessary mathematics<br />
reviews (e.g. numeral systems and logarithms).<br />
5. RESULTS<br />
Students accepted the module surprisingly well (consi<strong>der</strong>ing<br />
it being quite deep and abstract and out <strong>of</strong> the <strong>of</strong>ficial<br />
curriculum). Some <strong>of</strong> them realized the importance <strong>of</strong> the<br />
topic and were glad to un<strong>der</strong>stand it better, some were challenged<br />
by the proposed problems.<br />
5.1 Observation during lessons<br />
An unexpectedly fruitful discussion arose from ”Is it even?“<br />
question during numbers guessing. It is informatically optimal.<br />
Some students therefore insist to use it as the first<br />
question. Some realize that they can ask it also at the end<br />
in a simpler way, when they are choosing between the last<br />
two alternatives. This shows an important fact: difficulty
<strong>of</strong> answers management is important in practice. The question<br />
does not allow us to construct consequent questions<br />
straightforwardly with the same principle. Yet, after being<br />
reminded <strong>of</strong> binary system, students are able to realize the<br />
connection. They are doing almost the same as bisecting,<br />
only reversed. The value <strong>of</strong> these thoughts lay in the fact<br />
that they were driven mostly by students themselves.<br />
Disturbing (however expected) finding is low level <strong>of</strong> un<strong>der</strong>standing<br />
<strong>of</strong> certain parts <strong>of</strong> mathematics among students.<br />
Logarithms are taught intensively, yet no (!) student have<br />
seen a direction to calculate the depth <strong>of</strong> a tree. Our guidance<br />
had to be very explicit. A few months later, most <strong>of</strong><br />
the students actively knew that it is related to exponential<br />
growth and therefore also to logarithms.<br />
Some <strong>of</strong> the ideas encountered in the first module were<br />
needed again during the school year. More than a half <strong>of</strong><br />
students (in both groups) recalled them actively. This includes<br />
bisection, decision trees, binary numeral system and<br />
efficiency modeled using some kind <strong>of</strong> elementary step.<br />
5.2 Test results<br />
Further explanations and remarks are based on collected<br />
test sheets examination and our own notes from the lessons<br />
and consecutive discussions with students. We have not used<br />
any formal pre-test. It was more than clear from the initial<br />
discussions held for this purpose that students had difficulties<br />
to even un<strong>der</strong>stand the given questions. Student can<br />
get a certain amount <strong>of</strong> points for each task. Furhther given<br />
percentages indicate the average ratio <strong>of</strong> acquired points,<br />
illustrating the success on each task.<br />
Brief explanation.<br />
Students were to explain briefly the meaning <strong>of</strong> the term<br />
information. Any meaningful answer was rewarded with<br />
points. Definition or equally complete description got all the<br />
points, less precise expressions or examples got only partial<br />
score. The resulting success is lower than expected, given<br />
that it is a question <strong>of</strong> knowledge and outer resources were<br />
available. We see the reason in the lack <strong>of</strong> communication<br />
skills. Most <strong>of</strong> the students un<strong>der</strong>stood the term well, as we<br />
have seen during other tasks. To express it was however too<br />
difficult.<br />
Information and logarithm.<br />
This was a double task. First was to determine how many<br />
yes/no questions does one need to reliably identify a Czech<br />
deputy (they are 200). The second was to determine the<br />
minimal sufficient length for their hypothetical binary ID<br />
code. Very few students have used logarithm for any <strong>of</strong> the<br />
tasks. Majority <strong>of</strong> them chose to simulate the known process<br />
(sequential halving and doubling, respectively. Only two<br />
students have realized that the result is the same for both<br />
questions and the second calculation is not needed.<br />
Suboptimal decision tree.<br />
There was a small (7 leaves, 4 levels) decision tree with<br />
given rates per leaf. First task was to calculate the expected<br />
number <strong>of</strong> questions asked. The second was to lower<br />
that number by adjusting the tree.Those who made some<br />
calculations usually got both parts right. Less than a quarter<br />
<strong>of</strong> students (in both groups) passed the calculation, as<br />
it was still too mysterious for them. The bigger surprise is<br />
then that some <strong>of</strong> them un<strong>der</strong>stood the situation well enough<br />
142<br />
to optimize the tree (swapping frequent nodes towards the<br />
root).<br />
Inquiries comparison.<br />
Students were to sort seven given questions (such as “Is it<br />
a prime?”) according to the expected amount <strong>of</strong> information<br />
provided by the answers. All the questions were bound to<br />
the same situation, identifying an integer number between<br />
0 and 255. This task was very dependent on mathematical<br />
skills, therefore the results are rather poor.<br />
Statements comparison.<br />
This last task was not included in the same test with the<br />
others. We put it into another test six months later. There<br />
were four different smiley-like crime suspect faces and four<br />
statements regarding the <strong>of</strong>fen<strong>der</strong>, such as “The <strong>of</strong>fen<strong>der</strong> has<br />
no hair”. The task was to sort these statements according<br />
to the amount <strong>of</strong> information included. Unlike with the<br />
previous task, here it was obvious enough that those who had<br />
a mistake in their resulted did not un<strong>der</strong>stand the matter<br />
at all. Hence we gave all the points or none. The success<br />
rates were unexpectedly high, consi<strong>der</strong>ing the time passed.<br />
We can conclude that the educational results hold well.<br />
6. CONCLUSION<br />
Based on the results above, we may conclude that students<br />
are capable <strong>of</strong> un<strong>der</strong>standing the notion <strong>of</strong> information and<br />
its basic applications. The goals <strong>of</strong> the module were set<br />
surprisingly close to the real need. The limiting factors lay<br />
mostly in students mathematics skills. The teaching methods<br />
are found useful and efficient by the students, although<br />
they have recommended some tweaks, mostly in activities<br />
distribution in time.<br />
We described the concept <strong>of</strong> an introductory CS course<br />
for Czech grammar schools and gave more detailed specification<br />
about the module about information. Then we could<br />
proceed to the results obtained during the first testing year<br />
<strong>of</strong> the course.<br />
We intended to find out whether the module is feasible in<br />
the given constraints and conditions. The result shown here<br />
is that the proposed concept works, although it needs a few<br />
adjustments. The most distinctive feature <strong>of</strong> the proposed<br />
course is the shift in goals from CS itself to key competences<br />
and connections to other grammar school subjects.<br />
7. REFERENCES<br />
[1] Framework Education Programme for Secondary<br />
General Education (Grammar Schools). V´yzkumn´y<br />
ústav pedagogick´y v Praze, 2007.<br />
[2] M. R. Fellows, T. Bell, and I. Witten. Computer<br />
Science Unplugged... <strong>of</strong>fline activities and games for all<br />
ages: Original Activities Book. Computer Science<br />
Unplugged, 1996.<br />
[3] J. Hromkovič and B. Steffen. Why teaching informatics<br />
in schools is as important as teaching mathematics and<br />
natural sciences. In Proceedings <strong>of</strong> ISSEP, ISSEP’11,<br />
pages 21–30, Berlin, Heidelberg, 2011. Springer-Verlag.<br />
[4] D. Lessner. Computer science curriculum proposal for<br />
czech grammar schools. In Zborník príspevkov, ITAT<br />
2011, pages 99–104, Seňa, Slovakia, 2011. PONT s. r. o.<br />
[5] A. Verno. Supporting k-12 computer science education.<br />
J. Comput. Sci. Coll., 23(3):145–146, jan 2008.
Data modeling and database systems<br />
as part <strong>of</strong> general education in CSE<br />
Claudia Strödter<br />
<strong>University</strong> <strong>of</strong> Jena<br />
Didactics <strong>of</strong> mathematics and computer science<br />
Ernst-Abbe-Platz 2<br />
07743 Jena<br />
0049/3641946202<br />
claudia.stroedter@uni-jena.de<br />
ABSTRACT<br />
Everyday we unconsciously send queries and actions to databases.<br />
The used applications are part <strong>of</strong> school and working life and in<br />
specific areas <strong>of</strong> leisure. In or<strong>der</strong> to teach a responsible handling<br />
with such applications it will be necessary that “data modeling<br />
and database systems” will be a part <strong>of</strong> computer science<br />
education. By analyzing selected teaching material it is<br />
investigated if and how it is possible to teach this topic as a<br />
general education. The study shows that it´s possible to teach<br />
general contents and processes with selected material.<br />
Furthermore, a first proposal for a grading <strong>of</strong> the topic “data<br />
modeling and database systems” reveals.<br />
Categories and Subject Descriptors<br />
K3.2 [Computers & Education]: Computer and Information<br />
Science Education – computer science education, information<br />
systems education.<br />
General Terms<br />
Documentation, Experimentation, Human Factors, Theory,<br />
Standardization, Verification<br />
Keywords<br />
teaching in heterogeneous classes, education standards in lower<br />
secondary education, curricular aspects, didactical approaches,<br />
data modeling and database systems<br />
1. MOTIVATION<br />
Computing systems have influence on several parts <strong>of</strong> our<br />
everyday life. Activities in our working life and in specific areas<br />
<strong>of</strong> leisure (e.g. credit card payment, shopping, seeing medical<br />
pr<strong>of</strong>essionals) are supported by informatics applications.<br />
Unconsciously we initiate several procedures to read and process<br />
data and send queries to databases. Young people or students<br />
spend much time a day using the internet. In addition, they use<br />
smartphones, chip cards, and several vending machines. In sum,<br />
the handling <strong>of</strong> computing systems is an important element in<br />
young people’s life. Mostly, they are unaware <strong>of</strong> the associated<br />
informatics contents and processes. In or<strong>der</strong> to ensure a<br />
responsible use <strong>of</strong> computing systems and personal data this<br />
knowledge should be a part <strong>of</strong> school education, especially <strong>of</strong><br />
computer science education (CSE). Therefore, the CSTA K-12<br />
Standards for Computer Science (CS) focus on the following<br />
goals. Students should:<br />
143<br />
(1) un<strong>der</strong>stand the nature <strong>of</strong> CS and its place in the mo<strong>der</strong>n<br />
world,<br />
(2) un<strong>der</strong>stand that CS interweaves concepts and skills,<br />
(3) be able to use CS skills (especially computational thinking)<br />
in their problem-solving activities in other subjects. […] (see<br />
[4] page 7).<br />
Similar demands are used in the basic experiences by Fothe.<br />
“CSE is a general education that allows students the following<br />
basic experiences:<br />
(1) to discover and to un<strong>der</strong>stand computing systems<br />
(informatics systems) in several parts <strong>of</strong> life,<br />
(2) to recognize, that done or planned activities can be formulated<br />
as an algorithm and transferred into a computer<br />
program, that modeling allows to transfer parts <strong>of</strong> reality into<br />
a computer-compatible form and that computing systems<br />
(informatics systems) are designed by humans,<br />
(3) to acquire problem-solving skills in dealing with exercises,<br />
which are useful at school and real life (“modeling<br />
techniques” means as techniques <strong>of</strong> mental work).<br />
These experiences are connected to each other in several ways. 1 ”<br />
[6] In summary, to un<strong>der</strong>stand CSE as a general education it is<br />
necessary to un<strong>der</strong>stand general informatics contents and<br />
processes and to use this knowledge in everyday life.<br />
As mentioned earlier many activities in our everyday life access<br />
databases. In or<strong>der</strong> to ensure a general education the topic “data<br />
modeling and database systems” should be a part <strong>of</strong> CSE. In the<br />
CSTA K-12 Standards for CS is required that students should be<br />
able to select appropriate database formats for a particular<br />
computational problem and to form abstract ideas about specific<br />
components (e.g., input, output, processors, and databases) and<br />
their roles in the computational spectrum (see [4] page 11). There<br />
exist many ideas and concepts to teach “data modeling and<br />
database systems” for university students. Other concepts focus<br />
only on the use <strong>of</strong> specific database tools. Antonitsch uses another<br />
approach. He suggests to develop structure awareness by starting<br />
with analyzing and querying ready-to-use databases provided by<br />
the teacher [1].<br />
Nevertheless, it remains unclear which general informatics<br />
contents and processes are necessary to teach “data modeling and<br />
database systems”. In German school this topic is a part <strong>of</strong> CSE in<br />
lower secondary education. The objective <strong>of</strong> this article is to<br />
provide criteria for exercises (informatics contents and processes,<br />
requirement areas) in or<strong>der</strong> to teach “data modeling and database<br />
systems” in CSE in lower secondary education. In this study,<br />
existing teaching material is used to get an idea <strong>of</strong> teaching this<br />
topic as a general education.<br />
1 All translations (German – English) were done by the author.
2. DATA MODELING AND DATABASE<br />
SYSTEMS IN GERMAN SCHOOLS<br />
In consequence <strong>of</strong> the results <strong>of</strong> international comparative studies<br />
(e.g. International student assessment - PISA 2000) in <strong>Germany</strong><br />
the “Principles and Standards for School Informatics –<br />
Educational Standards <strong>of</strong> Informatics in Lower Secondary<br />
Education” (translation in accordance with: [8], abbreviated:<br />
educational standards <strong>of</strong> informatics SI) were developed. These<br />
are output-orientated minimum standards and specify the<br />
competencies which should be acquired when completing 7th and<br />
10th grade. They consist <strong>of</strong> process standards and content<br />
standards (see Figure 1) and were proposed to bridge the<br />
knowledge about information and communication technology<br />
(ICT) and CS (see [3]). Therefore, in this study these competencies<br />
are used to evaluate exercises.<br />
Figure 1. Process and content standards 2<br />
“Data modeling and database systems” influences several process<br />
and content standards. For this study it is reasonable to focus on a<br />
few main competencies. Most contents and processes associated<br />
with “data modeling and database systems” are summarized in the<br />
content standard Information and data and the process standard<br />
Model and implementation (see table 1).<br />
Table 1. Associated competencies <strong>of</strong> the “educational<br />
standards <strong>of</strong> informatics SI”<br />
Standards Competencies<br />
Content<br />
Standard 3<br />
Information<br />
and data<br />
Process<br />
Standard<br />
Model and<br />
implementation<br />
Learners <strong>of</strong> all age should …<br />
…un<strong>der</strong>stand the connection between data and information<br />
and different forms <strong>of</strong> representations <strong>of</strong> data,<br />
…un<strong>der</strong>stand operations on data, and interpret these<br />
with regard to the represented information,<br />
…be able to trigger suitable operations on data.<br />
Learners <strong>of</strong> all age should …<br />
…create informatics models based on given facts,<br />
…implement models with suitable applications,<br />
…reflect models and their implementation.<br />
Currently, in <strong>Germany</strong> no educational standards for upper<br />
secondary education <strong>of</strong> informatics exist. The “Einheitliche<br />
Prüfungsanfor<strong>der</strong>ungen Informatik” [9] (translation: standardized<br />
requirements for final examinations in CSE, abbreviated: EPA)<br />
were developed to ensure transparency, comparability, and<br />
standardization <strong>of</strong> final examinations. They define examination<br />
areas and assist in compiling exercises (especially requirement<br />
areas). “Data modeling and database systems” is included in all<br />
examination areas. The requirement areas are <strong>of</strong>ten used for<br />
grading <strong>of</strong> competencies and exercises (e.g. see [10]). In this<br />
2 [8], translated by [3].<br />
3 The translations <strong>of</strong> the content standards are taken over by [3].<br />
144<br />
study the levels reproduction performance, transfer performance<br />
and constructive performance are also used to evaluate the<br />
cognitive complexity <strong>of</strong> exercises. On the first level students are<br />
able to account known facts, to explain and present methods and<br />
principles <strong>of</strong> CS. Transfer performances allow the independent<br />
use <strong>of</strong> known facts, methods and principals in or<strong>der</strong> to solve new<br />
problems. Learners on the last level independently decide on<br />
methods in new and complex problem situations (see [12]).<br />
The “educational standards <strong>of</strong> informatics SI” and the EPA are<br />
structured in several ways. Nevertheless, it is possible to assign<br />
the competencies and contents <strong>of</strong> the EPA to the process<br />
standards and content standards <strong>of</strong> the “educational standards <strong>of</strong><br />
informatics SI” [7]. Therefore, in this study parts <strong>of</strong> both<br />
educational principals are used to evaluate exercises.<br />
3. RESEARCH QUESTION AND DESIGN<br />
Education is influenced by many details. There are, for example:<br />
the teaching methods, the scientific contents and processes, the<br />
asked questions, and the used exercises. This study focusses on<br />
the contents and processes in exercises.<br />
Material<br />
By comparing five recent German schoolbooks, it was shown that<br />
the topic “data modeling and database systems” is divided into<br />
exercises about storage <strong>of</strong> large data volumes, data modeling,<br />
implementation and queries. Developing schoolbooks is a longterm<br />
process, in which the authors’ entire expert-knowledge about<br />
education and science is used and discussed in several ways.<br />
Therefore, it is assumed that exercises in schoolbooks are typical<br />
exercises. This selection enables a nonreactive research, which<br />
excludes experimenter effects, test or other answer falsifications.<br />
In this study, 61 exercises taken out <strong>of</strong> one frequently used<br />
schoolbook (see [5]) for lower secondary education are analyzed.<br />
The book is permitted in all funeral states <strong>of</strong> <strong>Germany</strong> and allows<br />
to achieve the existing curricula. The chosen exercises include the<br />
whole topic “data modeling and database systems” and are similar<br />
to the exercises in the other four schoolbooks.<br />
Analysis method<br />
To analyze the teaching material the qualitative content analysis<br />
by Mayring [11], more precisely the structuring analysis is used.<br />
Table 2. Schematic representation <strong>of</strong> the analysis method<br />
Operation Explanation<br />
Selection <strong>of</strong> material 61 exercises taken from one school book<br />
Determination <strong>of</strong><br />
analyzing units<br />
Deductive development<br />
<strong>of</strong> categories<br />
Ensuring <strong>of</strong> the interrater<br />
reliability<br />
One exercise is one unit<br />
Developing <strong>of</strong> categories, anchors, and<br />
coding rules<br />
The categories were discussed with<br />
members <strong>of</strong> the department <strong>of</strong> didactics<br />
<strong>of</strong> mathematics and computer science.<br />
Rating All exercises were rated by two Raters.<br />
Interpretation Structuring analysis<br />
In or<strong>der</strong> to provide criteria for exercises in “data modeling and<br />
database systems”, the following questions are posed to the<br />
material:<br />
Question 1: Which contents and processes are necessary to<br />
achieve the competencies <strong>of</strong> the “educational standards <strong>of</strong><br />
informatics SI”?<br />
The “educational standards <strong>of</strong> informatics SI” allow an outputorientated<br />
teaching, simultaneously they were proposed to bridge
the knowledge about ICT and CS (see [3]). CSE as a general<br />
education in German school should comply with these standards.<br />
Question 2: Which contents and processes are used to teach “data<br />
modeling and database systems” according to the individual level<br />
<strong>of</strong> students (means as different existing knowledge)?<br />
The knowledge about handling computing systems and the<br />
experiences students have in their everyday life are mostly<br />
different. In or<strong>der</strong> to un<strong>der</strong>stand general informatics contents and<br />
processes and to use this knowledge in everyday life it is<br />
necessary to use exercises according to the individual level <strong>of</strong><br />
students. Doing so, it will be also possible to combine the already<br />
existing knowledge about handling computing systems with<br />
knowledge about general informatics contents and processes.<br />
Category systems<br />
The first question focusses on the “educational standards <strong>of</strong><br />
informatics SI”. The competencies relating to “data modeling and<br />
database systems” are used as categories to evaluate the exercises<br />
(see table 3, CS1-PS3). It is investigated which informatics<br />
contents and processes enable the achievement <strong>of</strong> the competencies.<br />
The presented competencies are connected to each other in<br />
several ways. It was possible to assign more than one category for<br />
each exercise.<br />
The individual level <strong>of</strong> students can be consi<strong>der</strong>ed in different<br />
ways (e.g. information representation, teaching arrangement, time<br />
to work on exercises, context <strong>of</strong> exercises, complexity <strong>of</strong><br />
exercises). By using the method applied in this study, it is<br />
possible to investigate the cognitive complexity 4 . In or<strong>der</strong> to<br />
evaluate the exercises three categories (see table 3, ABI-ABIII)<br />
are used. Only one category can be assigned for each exercise.<br />
Table 3. Category systems for question 1 and 2<br />
Code Category and Anchor<br />
Learners un<strong>der</strong>stand the connection between data and<br />
information and different forms <strong>of</strong> representations <strong>of</strong> data.<br />
CS1 “Choose a problem area where large data volumes are<br />
stored and connected to the internet. What information on<br />
objects, persons, or events are stored in the database?”<br />
Learners un<strong>der</strong>stand operations on data, and interpret<br />
these with regard to the represented information.<br />
CS2 “Send the following described queries to the given<br />
database mail-or<strong>der</strong> business.<br />
Give an overview <strong>of</strong> all clients <strong>of</strong> a specific city.”<br />
Learners are able to trigger suitable operations on data.<br />
“The following SQL statement is given<br />
CS3 SELECT Artikel.ANr, Artikel.Bezeichnung, Lieferant.LNr,<br />
Lieferant.Name FROM Artikel, Lieferant;<br />
Send a query to the database and interpret the results.”<br />
Learners create informatics models based on given facts.<br />
“Decide which columns are necessary to acquire data in<br />
PS1<br />
the table CD. Complete the current schema with the new<br />
columns.”<br />
Learners implement models with suitable applications.<br />
PS2 “Transfer your current data model into a relational model.<br />
Use the necessary rules and specify the developed tables.”<br />
Learners reflect models and their implementation.<br />
“Check your database mail-or<strong>der</strong> business in accordance<br />
PS3<br />
with the given rules. If necessary, optimize your<br />
database.”<br />
4 Cognitive complexity involves the requirement, the level <strong>of</strong><br />
difficultly etc. (see [2]).<br />
145<br />
Reproduction performance<br />
ABI “Which information can you receive from the tables in the<br />
database CD? Formulate five questions.”<br />
Transfer performance<br />
“In the last exercise you verbally formulated possible<br />
ABII queries to your database. Describe your queries as<br />
selection, projection or as a combination <strong>of</strong> both. Express<br />
your queries as SQL-statements.”<br />
Constructive performance<br />
ABIII “Does your database mail-or<strong>der</strong> business guarantee<br />
referential integrity? If necessary, update your database.”<br />
Rating<br />
All exercises were evaluated by one teacher and one student<br />
teacher <strong>of</strong> CSE. For all categories the inter-rater reliability was<br />
higher than 0.65 (Cohen-Kappa). To get an idea <strong>of</strong> the rating<br />
method an example for rating an exercise is given first:<br />
Exercise: “A part <strong>of</strong> the data model CD rental is given. Find other<br />
useful classes/entities for this data model. Complete the data<br />
model with appropriate attributes. Define the cardinality <strong>of</strong> the<br />
relationships.”<br />
Rating <strong>of</strong> Rater 1 (n1) and 2 (n2): PS1, PS2, CS1, ABII<br />
All equal rated exercises (N=45) were used to receive a common<br />
grading <strong>of</strong> the content and process competencies. The grading is<br />
presented in the next chapter.<br />
4. RESULTS<br />
The “educational standards <strong>of</strong> informatics S1” are developed as<br />
minimum standards. Nevertheless, there are informatics contents<br />
and processes above and below these standards. The answers <strong>of</strong><br />
question 1 and 2 emphasize contents and processes to achieve the<br />
standards in three requirement areas.<br />
In or<strong>der</strong> to get an overview <strong>of</strong> the used material some quantitative<br />
data (see table 4) are given first.<br />
Table 4. Quantitative data <strong>of</strong> the analysis<br />
Rater CS1 CS2 CS3 PS1 PS2 PS3 ABI ABII ABIII<br />
n1 30 30 26 8 6 21 13 32 16<br />
n2 31 29 24 8 6 23 11 31 19<br />
In regard to answering the first part (contents) <strong>of</strong> question 1 a<br />
summary <strong>of</strong> the content standards in exercises is shown in table 5.<br />
Combining the answers <strong>of</strong> question 1 and 2 results in a grading <strong>of</strong><br />
the content competencies (see table 5). In sum, it is shown, that:<br />
� the handling with the database systems turns from simple<br />
usage <strong>of</strong> the graphical user interface to formal statements,<br />
� the databases and queries turn from given databases and<br />
queries to independently developed databases and queries,<br />
� the use <strong>of</strong> databases turns from simple usage to evaluated and<br />
improved usage.<br />
However, there are only a few exercises concerning process<br />
standard 1 and 2. It is possible to compile a summary <strong>of</strong> process<br />
standards in exercises. Combining the answers <strong>of</strong> question 1 and 2<br />
results in a grading <strong>of</strong> the process competencies (see table 5). In<br />
sum, it is shown, that:<br />
� data modeling turns from modeling some elements <strong>of</strong> data<br />
worlds to modeling complete models <strong>of</strong> data worlds,<br />
� the transfer <strong>of</strong> data models into database systems turns from<br />
the transfer <strong>of</strong> given tables and relationships to the transfer <strong>of</strong><br />
independently created data models,<br />
� the use <strong>of</strong> data models turns from simple usage to evaluated<br />
and improved usage.
Table 5. Summarized results <strong>of</strong> question 1 and 2<br />
Content competencies in exercises<br />
In a given database learners...<br />
…insert and modify data by using the graphical user<br />
interface,<br />
…realize the used classes/entities and attributes,<br />
ABI<br />
…create forms by using the graphical user interface,<br />
…create possible queries as questions,<br />
…use simple operations (selection, sorting, summarizing)<br />
by using the graphical user interface.<br />
In a given database learners…<br />
…insert and modify data by using formal statements,<br />
…expand the database/data model,<br />
…define useful attributes and data types,<br />
…define, set, delete, and evaluate foreign and primary<br />
ABII keys,<br />
…define the cardinality <strong>of</strong> given relationships,<br />
…create formal statements (selection, projection,<br />
combination <strong>of</strong> both) to given questions,<br />
…read, run, interpret, evaluate, and improve given formal<br />
statements (selection, projection, combination <strong>of</strong> both).<br />
Learners...<br />
…define referential integrity,<br />
…check and improve the databases in compliance with<br />
simplified rules,<br />
…formulate and define independent useful queries,<br />
ABIII<br />
…independently create useful formal statements<br />
(selection, projection, combination <strong>of</strong> both, inner join),<br />
…read, run, interpret, evaluate, and improve formal<br />
statements (selection, projection, combination <strong>of</strong> both,<br />
inner join).<br />
Process competencies in exercises<br />
Learners...<br />
…find useful attributes for given classes/entities,<br />
ABI<br />
…transfer given tables and relationships into database<br />
systems by using the graphical user interface.<br />
Learners...<br />
…analyze and evaluate data worlds to find classes/entities<br />
and attributes for a data model,<br />
…create relationships between given classes/entities and<br />
define cardinalities,<br />
ABII<br />
…transfer simple data models into relational models and<br />
in a database management system (one-to-one<br />
relationship) in compliance with transfer rules,<br />
…evaluate simple data models (one-to-one relationship)<br />
according to the given problem in compliance with rules.<br />
Learners…<br />
…independently create complete data models in<br />
compliance with rules,<br />
ABIII …evaluate data models according to the given problem,<br />
…transfer complete data models (one-to-n and n-to-n<br />
relationship) into relational models and in a database<br />
management system in compliance with transfer rules.<br />
5. CONCLUSION<br />
With the use <strong>of</strong> recent teaching material in some parts it is<br />
possible to realize the requirements in chapter 1. Initial statements<br />
on criteria (contents and processes, requirement areas) for<br />
146<br />
exercises are emphasized in this study. Important contents and<br />
processes for teaching “data modeling and database systems” as<br />
general education are mentioned in chapter four. It is possible to<br />
reveal a first grading which allows to achieve the “educational<br />
standards <strong>of</strong> informatics SI” in several ways. Consi<strong>der</strong>ing these<br />
results it should be possible to turn the knowledge about “data<br />
modeling and database systems” from daily handling knowledge<br />
to independently evaluated and improved knowledge. With the<br />
investigated requirement areas the individual level <strong>of</strong> students can<br />
be included. This article presents a first proposal on criteria for<br />
teaching “data modeling and database systems” as part <strong>of</strong> general<br />
education in CSE. To complete and check the mentioned contents,<br />
processes and the grading these criteria need to be evaluated with<br />
further empirical studies.<br />
6. REFERENCES<br />
[1] Antonitsch, P. 2006. Databases as a tool <strong>of</strong> general education.<br />
In Proceedings <strong>of</strong> the 2006 international conference on Informatics<br />
in Secondary Schools - Evolution and Perspectives: the<br />
Bridge between Using and Un<strong>der</strong>standing Computer ISSEP<br />
2006. Springer. New York.<br />
[2] Baumann, R. 2008. Probleme <strong>der</strong> Aufgabenkonstruktion<br />
gemäß Bildungsstandards. In LOG IN Heft 153, 54-58.<br />
[3] Brinda, T.; Puhlmann, H.; Schulte, C. 2009. Bridging ICT and<br />
CS - Educational Standards for Computer Science in Lower<br />
Secondary Education. IN Proceedings <strong>of</strong> the 14th Annual<br />
SIGCSE Conference on Innovation and Technology in<br />
Computer Science Education, ITiCSE 2009. URL: http://<br />
www.inf.fu-berlin.de/inst/ag-ddi/docs/bridgingICTandCS.pdf.<br />
[4] CSTA Standards Task Force. 2011. K-12 Computer Science<br />
Standards – revised 2011. URL: http://www.csta.acm.org/<br />
Curriculum/sub/CurrFiles/CSTA_K-12_CSS.pdf.<br />
[5] Engelmann, L. 2010. Informatik SI – Informatische<br />
Grundbildung. Duden Paetec. Berlin.<br />
[6] Fothe, M. 2012. Modellieren, Programmieren, genetisches<br />
Prinzip und Informatik im Kontext - Was passt da wie zusammen?<br />
Lectured on GI-FIIB Berlin 2012, URL:<br />
http://hyfisch.de/Fachgruppe/tagung11/fothe.<br />
[7] Fothe, M. 2008. Bildungsstandards Informatik für die<br />
Sekundarstufe II – Vorüberlegungen zur Entwicklung. In<br />
Proceedings <strong>of</strong> <strong>Didaktik</strong> <strong>der</strong> Informatik – Aktuelle Forschungsergebnisse.<br />
URL: http://subs.emis.de/LNI/<br />
Proceedings/Proceedings135/gi-proc-135-010.pdf.<br />
[8] German Informatics society (GI) 2008. Grundsätze und<br />
Standards für die Informatik in <strong>der</strong> Schule. Bildungsstandards<br />
Informatik für die Sekundarstufe I. URL:<br />
http://www.informatikstandards.de/.<br />
[9] KMK 2004. Einheitliche Prüfungsanfor<strong>der</strong>ungen in <strong>der</strong><br />
Abiturprüfung Informatik. URL: http://www.kmk.org/<br />
fileadmin/veroeffentlichungen_beschluesse/1989/1989_12_01<br />
_EPA_Informatik.pdf.<br />
[10] Kollee, C. et al. 2009. Computer Science Education and Key<br />
Competencies. In Proceedings <strong>of</strong> 9th IFIP World Conference<br />
on Computers in Education - WCCE 2009. URL:<br />
http://www.die.informatik.uni-siegen.de/e-publikationen/<br />
Publikationen/2009/WCCE2009_pap147.pdf.<br />
[11] Mayring, P. 2003. Qualitative Inhaltsanalyse – Grundlagen<br />
und Techniken. Beltz-Verlag, Weinheim.<br />
[12] Nelles, W. et al. 2010. Entwicklung eines Kompetenzrahmenmodells<br />
– Informatische Modellieren und Systemverständinis.<br />
In: Informatik-Spektrum 33 (2010) 1, 45-35.<br />
URL: www.springerlink.com/content/128182n611u26557/<br />
fulltext.pdf.
The Mindstorm Effect: A Gen<strong>der</strong> Analysis on the Influence<br />
<strong>of</strong> LEGO Mindstorms in Computer Science Education ∗<br />
ABSTRACT<br />
Catherine Ball Faron Moller<br />
Director, Technocamps †<br />
In the UK, as elsewhere, the number <strong>of</strong> students choosing to<br />
study computer science is declining. This is especially true<br />
with female students. This paper explores the effectiveness<br />
<strong>of</strong> LEGO Mindstorms in a pedagogic context in education<br />
and its ability to attract female students to computer science.<br />
A mixed methods approach was used in this study, in<br />
which we looked at the FIRST LEGO League competition<br />
and how female students participate in these.<br />
The results demonstrate that, while young people enjoy<br />
using Mindstorms, they do little to influence young people<br />
to consi<strong>der</strong> computer science education. They can, however,<br />
be used effectively as an opportunity to engage young people<br />
in further computing skills and computational thinking. The<br />
lessons we learn from this research indicate that we need to<br />
use these tools as a foundation rather than as a solution to<br />
the problem for attracting more women into computing.<br />
1. INTRODUCTION<br />
Computer Science is an ever-growing field, however there is<br />
a noticeable lack <strong>of</strong> women choosing to study Computer Science<br />
or to enter the IT industry. There are many reasons<br />
that have been suggested for this. For instance it may be<br />
perceived to be too geeky or just for men. There have been<br />
various initiatives and tools proposed to help change the way<br />
that computing is perceived. One approach is through the<br />
use <strong>of</strong> LEGO Mindstorms. This has been used in several<br />
instances to change people’s perception <strong>of</strong> computer science<br />
and ICT in attempts to show people that these fields can be<br />
fun and exciting. Mindstorms can help teach children and<br />
young adults the importance <strong>of</strong> teamwork and problem solv-<br />
∗ This abstract is based on the first author’s BSc dissertation,<br />
Department <strong>of</strong> Computer Science, Swansea <strong>University</strong>, 2012.<br />
† Technocamps (www.technocamps.com) is a school outreach<br />
programme started in 2004 at Swansea <strong>University</strong>. In 2011,<br />
a three-year, £6 million EU project allowed Technocamps<br />
to expand its operation throughout Wales, with hubs at the<br />
Universities <strong>of</strong> Aberystwyth, Bangor and Glamorgan.<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WiPSCE 2012, Hamburg <strong>Germany</strong><br />
Copyright 2012 ACM ...$15.00.<br />
Swansea <strong>University</strong><br />
f.g.moller@swansea.ac.uk<br />
147<br />
Reena Pau<br />
ing and can encourage users to be creative and innovative<br />
in their design.<br />
2. LITERATURE REVIEW<br />
2.1 Women in Computing in the UK<br />
There are many reasons why women choose not to study or<br />
work in the technology sector. This has been reported as a<br />
weakening in the country’s economic position. The decline is<br />
apparent already at GCSE level (age 14-16), where there are<br />
ever fewer girls who choose to do a full course in ICT. The<br />
decline then continues right through to industry. Although<br />
there have been major national efforts to encourage females<br />
to study computing, the numbers remain low.<br />
2.2 Status <strong>of</strong> CSE in the UK<br />
Part <strong>of</strong> the above problem lies in the ICT national curriculum,<br />
and to this end in January 2012 it was announced that<br />
the ICT curriculum was to be dropped in September <strong>of</strong> the<br />
same year, and instead Computer Science would eventually<br />
be taught [2]. This has been supported by several industry<br />
experts and it is hoped that the new curriculum will give<br />
schools the freedom to create their own ICT and Computer<br />
Science curricula. Companies such as Google and Micros<strong>of</strong>t<br />
are currently working together and with technology education<br />
organisations, such as the British Computer Society<br />
(BCS), to produce free educational material for schools. In<br />
or<strong>der</strong> to encourage young people to study computing, tools<br />
such as Mindstorms are being used in various initiatives to<br />
enthuse young people about technology and creativity.<br />
2.3 The use <strong>of</strong> LEGO Mindstorms in schools<br />
Mindstorms kits are used in schools to provide an exciting<br />
way for children and young adults to learn about technology.<br />
They teach students various skills including teamwork<br />
and programming and can potentially change the way ICT<br />
is viewed. The teaching <strong>of</strong> ICT is <strong>of</strong>ten given as a reason for<br />
females being disinterested in pursuing ICT or computing<br />
in further education [3] because many females find it boring<br />
[6]. FIRST LEGO League (FLL) is a worldwide competition<br />
in which teams <strong>of</strong> 11-16 year olds design and program<br />
autonomous robots to complete tasks. With over 19,000<br />
teams in over 55 countries, FLL is constantly expanding.<br />
This type <strong>of</strong> initiative gives something like Mindstorms a<br />
purpose: pupils have something to aim for.<br />
In this study we looked at Mindstorms and whether it<br />
can attract females to the field <strong>of</strong> computing. From previous<br />
related studies, it is known that girls are willing to
engage in robotic activities [4]; one can no longer assume<br />
that robots are only for males. However, existing studies <strong>of</strong><br />
the effectiveness <strong>of</strong> Mindstorms in education (eg, [1, 5]) tend<br />
to concentrate on university students.<br />
3. METHODOLOGY<br />
A questionnaire was given to students who were taking part<br />
in a local version <strong>of</strong> the FLL; 93 questionnaires were returned.<br />
The questionnaire had four categories: Demographics,<br />
LEGO Mindstorms, Career Choices and Educational<br />
Choices. These were devised in or<strong>der</strong> to gather data about:<br />
how long the students had been using Mindstorms; if the<br />
students felt they had learnt any new skills by using them;<br />
and if this usage had influenced their education and career<br />
choices. These questions were asked in or<strong>der</strong> to un<strong>der</strong>stand<br />
if and how Mindstorms can change a young person’s perception<br />
<strong>of</strong> Computer Science and technology. Qualitative and<br />
quantitative data methods were used to collect data and<br />
there were a range <strong>of</strong> both open and closed questions. The<br />
data sample consisted <strong>of</strong> 32 females and 61 males and the<br />
students were aged between 9 and 14 years old.<br />
4. RESULTS<br />
Males typically used Mindstorms from a younger age than<br />
Females, though our sample is not representative <strong>of</strong> the UK<br />
population and this should be kept in mind. This study<br />
does not attempt to make population estimates; Tnor does<br />
it document or assume that every participant <strong>of</strong> the FLL in<br />
the UK will have the same experience or learning outcomes.<br />
However, we assume that the trends are an indicator <strong>of</strong> the<br />
effect <strong>of</strong> Mindstorms in education. It is recognised that this<br />
sample is limited and a wi<strong>der</strong> range <strong>of</strong> participants would<br />
have been needed for a more in depth study.<br />
4.1 Interest in Computers<br />
The aim <strong>of</strong> these questions was to gauge how interested<br />
pupils were in using LEGO Mindstorms. 81% <strong>of</strong> males and<br />
85% <strong>of</strong> females indicated that Mindstorms made them more<br />
interested in computers. The results <strong>of</strong> this section indicate<br />
that Mindstorms is popular amongst this cohort and<br />
increases the interest in computers for both male and female<br />
pupils. Using computers for something other than gaming<br />
was a common reason for this that came up on the surveys,<br />
as well as engen<strong>der</strong>ing quick thinking and creativity. All<br />
participants indicated that they wanted to continue to use<br />
Mindstorms during the next academic year.<br />
4.2 Careers<br />
The aim <strong>of</strong> these questions was to un<strong>der</strong>stand if LEGO<br />
Mindstorms would encourage pupils to consi<strong>der</strong> a career in<br />
computing. The results are interesting: even though the<br />
participants indicated that they really enjoyed using Mindstorms,<br />
their aspirations with regards to computing careers<br />
suggested the opposite. 44% <strong>of</strong> males and only 14% <strong>of</strong> females<br />
indicated that using LEGO Mindstorms made them<br />
interested in having a computing job. This suggests that<br />
Mindstorms can make male participants more interested in<br />
a computing career in some cases, but is not always successful<br />
in influencing people’s career choices. Male participants<br />
gave specific reasons stating that they enjoyed programming,<br />
whereas female participants gave reasons that were vague<br />
(such as finding it “interesting”). This data suggests that<br />
148<br />
although both males and females enjoy using Mindstorms<br />
and are more interested in technology after using the kits, it<br />
does not necessarily mean that Mindstorms can make someone<br />
more interested in pursuing a technical career.<br />
4.3 Skills<br />
The data we collected suggests that Mindstorms can teach<br />
people new skills. The students felt that they had learnt how<br />
to work as part <strong>of</strong> a team, how to communicate effectively<br />
with people, and problem solving skills. Additionally, both<br />
gen<strong>der</strong>s said they had learnt “new skills on the computer”,<br />
“how to handle problems”, and “how to be more creative”.<br />
However, 85% <strong>of</strong> females thought they had learnt new skills<br />
compared to 66% <strong>of</strong> males. This data could suggest that<br />
females have a more positive experience with Mindstorms<br />
than males, and feel that it is more rewarding.<br />
4.4 Educational Choices<br />
Almost half <strong>of</strong> the males said that they were more interested<br />
in taking GCSEs in the STEM (Science, Technology, Engineering<br />
and Mathematics) subjects after using Mindstorms.<br />
Half <strong>of</strong> the female participants said that Mindstorms had<br />
influenced which GCSEs to study and said that they were<br />
now going to study ICT, Design Technology, and Further<br />
Maths. This data suggests that Mindstorms can influence<br />
young adults’ educational choices in some cases. This also<br />
suggests that Mindstorms made them more aware <strong>of</strong> the<br />
STEM subjects and made the subjects seem more appealing<br />
and interesting.<br />
5. CONCLUSIONS<br />
LEGO Mindstorms can be used successfully in education;<br />
young people learn new skills and are keen to continue with<br />
it to widen their knowledge <strong>of</strong> programming and technology.<br />
Some females consi<strong>der</strong> programming to be the best<br />
part <strong>of</strong> using Mindstorms and taking part in the FLL. Although<br />
it does not seem to make young people more aware<br />
<strong>of</strong> Computer Science careers, it does enhance their interest<br />
in computers and their uses.<br />
6. REFERENCES<br />
[1] D. Auf<strong>der</strong>heide, W. Krybus, and U. Witkowski12.<br />
Experiences with LEGO Mindstorms as an embedded<br />
and robotics platform within the un<strong>der</strong>graduate<br />
curriculum. In FIRA-TAROS, volume 7429 <strong>of</strong> LNAI,<br />
pages 185–196. Springer, 2012.<br />
[2] Department <strong>of</strong> Education. Harmful ICT curriculum set<br />
to be dropped this September to make way for rigorous<br />
Computer Science. Press notice, 11 January 2012.<br />
[3] G. Lovegrove and W. Hall. Where have all the girls<br />
gone? <strong>University</strong> Computing, 9(4):207–210, 1987.<br />
[4] H. H. Lund and L. Pagliarini. Robocup Jr. with LEGO<br />
Mindstorms. In ICRA, pages 813–819. IEEE, 2000.<br />
[5] W. McWhorter and B. O’Connor. Do LEGO<br />
Mindstorms motivate programming students in CS1?<br />
In SIGCSE, pages 438–442. ACM, 2009.<br />
[6] R. Pau, M. Grace, and W. Hall. IT’s boring: A<br />
comparison <strong>of</strong> male and female students’ experiences <strong>of</strong><br />
ICT GCSE/A-level and Computing A-level lessons and<br />
their impact on student motivation. School Science<br />
Review, 92:89–94, 2011.
Exploring the processing <strong>of</strong> formatted texts by a<br />
kynesthetic approach ∗<br />
Carlo Bellettini Violetta Lonati Dario Malchiodi<br />
Mattia Monga Anna Morpurgo Mauro Torelli<br />
Dipartimento di Informatica<br />
Università degli Studi di Milano<br />
Milan, Italy<br />
{bellettini, lonati, malchiodi, monga, morpurgo, torelli}@di.unimi.it<br />
1. MOTIVATION<br />
One <strong>of</strong> the first experiences most pupils have with computing<br />
is through a word-processor and formatted texts. Indeed,<br />
computer literacy is frequently focused on acquiring dexterity<br />
with <strong>of</strong>fice automation tools, whereas the un<strong>der</strong>lying conceptual<br />
challenge <strong>of</strong> automatic processing <strong>of</strong> information is<br />
largely ignored [5, 4]. In fact, mastering the use <strong>of</strong> a word<br />
processor may increase one’s knowledge about typography<br />
or even the way one should organize ideas and discourse, but<br />
it gives very little insight about computing sciences without<br />
a specific emphasis on that topic.<br />
Recently, we started several activities [2, 3] aimed at introducing<br />
computing concepts to pupils, both striving to<br />
make boys and girls at ease with the intrinsic abstract nature<br />
<strong>of</strong> informatics and trying not to disappoint them with<br />
something not linked with the technology they are used to.<br />
In this paper we describe our experiments with a kind <strong>of</strong><br />
kinesthetic/tactile learning activity [1] we called algomotricity.<br />
In algomotricity the abstract symbolic manipulation is<br />
(partially) replaced by physical activities, which should help<br />
the pupils in developing their mental representation [6]. We<br />
tried to choose activities clearly linked to the acquaintance<br />
<strong>of</strong> students with computers and applications. In the following<br />
we report on a teaching activity about word-processors<br />
we proposed to a group <strong>of</strong> 25 pupils in 9th and 10th grades <strong>of</strong><br />
an Italian secondary school. The pupils had some familiarity<br />
with word-processor operations. The learning objective<br />
we had in mind was the un<strong>der</strong>standing <strong>of</strong> the challenges<br />
posed by the automatic elaboration <strong>of</strong> formatted text and<br />
we mainly focused on information representation techniques.<br />
2. DESCRIPTION OF THE EXPERIENCE<br />
The overall activity (8 hours in 4 non-consecutive days)<br />
∗ The authors would like to thank Giorgio Fattorelli and the<br />
Marie Curie IIS for giving the opportunity to experiment<br />
our ideas in their school.<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WiPSCE 2012, Hamburg, <strong>Germany</strong><br />
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$15.00.<br />
149<br />
was organized in four phases:<br />
1. A first approach to text formatting with a word processor;<br />
2. A dramatization <strong>of</strong> the process through the use <strong>of</strong> tangible<br />
objects;<br />
3. A game designed to force pupils to restructure their mental<br />
models and discover the power <strong>of</strong> symbolic meta-languages;<br />
4. A final use <strong>of</strong> special s<strong>of</strong>tware tools for formatting texts,<br />
also able to show the data structure used to record the<br />
meta-information.<br />
We started and ended in a computer lab, in or<strong>der</strong> to partially<br />
match pupils’ expectations about informatics. The<br />
risk we wanted to avoid was that <strong>of</strong> being perceived as unrealistic<br />
or fruitlessly “academic”. Thus, we designed the<br />
activity around two main ideas: (a) computers and s<strong>of</strong>tware<br />
tools should be <strong>of</strong> secondary importance, but the conceptual<br />
link with them should be clear; (b) the approach should be<br />
mostly allosteric [7]: the direct transmission <strong>of</strong> knowledge<br />
should be kept to a minimum, and pupils should be forced<br />
to reconsi<strong>der</strong> their mental models about text formatting by<br />
discovering themselves useful techniques. Moreover, the abstract<br />
nature <strong>of</strong> computing should be conveyed by concrete<br />
examples and physical activities. Pupils were requested to<br />
work in small groups to foster confrontation and almost every<br />
task was proposed together with an accompanying metacognitive<br />
reflection. In particular we <strong>of</strong>ten asked pupils to<br />
imagine how what they described would be un<strong>der</strong>stood by<br />
someone who ignored the context.<br />
Formatting transfers information.<br />
We started in a computer lab and the first objective was to<br />
convince pupils that formatting is not just an aesthetic issue,<br />
but it has also an important role in transferring information:<br />
indeed the meaning <strong>of</strong> a text is given by the words and their<br />
formatting. Pupils (working in couples) were requested to<br />
produce a formatted text <strong>of</strong> their choice. Then they had<br />
to answer some questions about their work: Which type <strong>of</strong><br />
formatting did you use? Why did you choose it? Did you<br />
use more than one formatting for the same piece <strong>of</strong> text?<br />
The activity took more than we thought and we were not<br />
able to introduce the second part, in which pupils would<br />
have been requested to produce a formatted version <strong>of</strong> a text<br />
read aloud by the conductor, where changes in voice tones
and emphasis should be ren<strong>der</strong>ed by italics, bold, colors,<br />
etc. Then students would have been divided into groups for<br />
reflecting on the reproducibility <strong>of</strong> this process by comparing<br />
results produced by different teams.<br />
How to record meta-information.<br />
The second phase was carried out in the gymnasium <strong>of</strong><br />
the school. Retrospectively this was a bad idea, since the<br />
big hall was rather chaotic and groups were too dispersed to<br />
foster a fruitful discussion. The proposed task was the reproduction<br />
<strong>of</strong> a formatted text on a big copy <strong>of</strong> the text put<br />
on the floor. The copy contained the same words, but no<br />
formatting (also the alignment was slightly different). Formatting<br />
had to be codified by using the objects available in<br />
the gym. Every group <strong>of</strong> pupils (6 persons) was requested to<br />
write down the rules they used in the codification. Another<br />
team had to interpret these rules in or<strong>der</strong> to decode the text<br />
formatted through physical objects and get back to the original<br />
formatted text. Most rules turned out to be ambiguous,<br />
especially when more than one kind <strong>of</strong> formatting had to be<br />
applied to the text. The objects were used mainly to mimic<br />
the formatting in the prototypical text. However, in some<br />
cases, pupils discovered a more abstract symbolic use as a<br />
means to cope with shortage <strong>of</strong> objects: for instance, two<br />
spoons were used to mark the beginning and the end <strong>of</strong> the<br />
word respectively, since there were not enough spoons to<br />
cover the whole word.<br />
Meta-language tricks.<br />
The third phase was carried out in a classroom and was<br />
organized around a game. Teams were again requested to<br />
reproduce a formatted text with objects and write down<br />
codification rules precise enough to be followed by another<br />
team. However, the game was made fun by the introduction<br />
<strong>of</strong> a “cost” for the objects. In fact, the more an object could<br />
be used to mimic a piece <strong>of</strong> formatting, the more it costed<br />
in or<strong>der</strong> to promote their symbolic use, e.g., since spaghetti<br />
pasta could be easily associated to the un<strong>der</strong>lying meaning,<br />
its cost was very high. The winner would be the team able<br />
to hand in an unambiguous codification with the lowest cost.<br />
The cost incentive was enough to let the pupils discover what<br />
is commonplace in mark-up languages: the use <strong>of</strong> tags at the<br />
beginning and at the end <strong>of</strong> (possibly overlapping) regions.<br />
A second round <strong>of</strong> the game was proposed without objects.<br />
Instead the pupils had only a multi-set <strong>of</strong> alphabetic characters<br />
on small pieces <strong>of</strong> paper. Some <strong>of</strong> the characters (for<br />
example, the letters that are not in the Italian alphabet: j, k,<br />
w, x, y) were not used in the words and could be easily used<br />
for weaving meta-information into the text. However, this<br />
trick was not suggested by the conductors but discovered by<br />
the pupils. Some <strong>of</strong> them tried to use the characters with a<br />
meta-meaning by placing them with a different orientation:<br />
a b for example, for meaning bold.<br />
Rediscovering formatting tools.<br />
The final phase was carried out again in a computer lab.<br />
Pupils were introduced to a special s<strong>of</strong>tware tool able to<br />
show a formatted text according to three different views:<br />
formatted and encoded either using a simplified mark-up<br />
language or a tree <strong>of</strong> objects. They were again requested to<br />
reproduce formatted texts by working on the other representations:<br />
they saw, however, the effect <strong>of</strong> their (syntactical)<br />
manipulation in the formatted view.<br />
150<br />
3. EVALUATION AND FUTURE WORKS<br />
We think the experience was successful. For example, all<br />
the pupils demonstrated to have grasped the idea <strong>of</strong> a metalanguage<br />
expressed in the same alphabet <strong>of</strong> the language<br />
itself. This is consi<strong>der</strong>ed quite an abstract concept, but<br />
it was found rather natural (even obvious) by the pupils.<br />
The link with word-processor and web technologies known<br />
to pupils was recognized. At the final recap we were also<br />
able to show that the same concepts are behind the scenes in<br />
several slightly different contexts, for example when editing<br />
Wikipedia entries. The pupils’ feedback was mostly positive:<br />
they did have fun and believe to have learned something.<br />
However, some <strong>of</strong> them found that the tasks were<br />
sometimes too easy and the part in the gym (see Section 2)<br />
was consi<strong>der</strong>ed boring by several participants.<br />
All in all, the proposed activity turned out to be a good<br />
way for conveying abstract computing concepts to pupils <strong>of</strong><br />
secondary schools. We are now working in refining the activity:<br />
as a mid term goal we aim at producing didactic<br />
material that should be self-contained enough to be used<br />
by independent teachers in their classes. As a first step we<br />
re-proposed the activity in an another school, with younger<br />
pupils (6th grade) un<strong>der</strong> the conduction <strong>of</strong> a math teacher<br />
who did not participate in the conception. We are now<br />
studying reports and videos <strong>of</strong> the experiences: the first<br />
impression is that it worked also in this different context.<br />
We found that the algomotricity approach can be effective<br />
in presenting abstract symbolic manipulations in very concrete<br />
ways. By choosing activities clearly connected with the<br />
acquaintance <strong>of</strong> pupils with computers and tools we believe<br />
this can be very successful in elaborating a fruitful un<strong>der</strong>standing<br />
<strong>of</strong> informatics concepts.<br />
4. REFERENCES<br />
[1] A. Begel, D. D. Garcia, and S. A. Wolfman. Kinesthetic<br />
learning in the classroom. In Proc. <strong>of</strong> the 35th SIGCSE<br />
TSCSE, pages 183–184, New York, USA, 2004. ACM.<br />
[2] A. Lissoni and V. Lonati and M. Monga and A.<br />
Morpurgo and M. Torelli. Working for a leap in the<br />
general perception <strong>of</strong> computing. In A. Cortesi and F.<br />
Luccio, editor, Proc. <strong>of</strong> informatics education europe<br />
III, pages 134–139. ACM, 2008.<br />
[3] V. Lonati and M. Monga and A. Morpurgo and M.<br />
Torelli. What’s the fun in informatics? Working to<br />
capture children and teachers into the pleasure <strong>of</strong><br />
computing. In Kalas and Mittermeir [8], pages 213–224.<br />
[4] M. Calzarossa, P. Ciancarini, L. Mich, and<br />
N. Scarabottolo. Informatics education in Italian high<br />
schools. In Kalas and Mittermeir [8], pages 31–42.<br />
[5] V. Dagien˙e. Informatics education for new millennium<br />
learners. In Kalas and Mittermeir [8], pages 9–20.<br />
[6] R. M. Fel<strong>der</strong> and L. K. Silverman. Learning and<br />
teaching styles in engineering education. J. <strong>of</strong><br />
Engineering Education, 78(7):674–681, 1988.<br />
[7] A. Giordan. From constructivisme to allosteric learning<br />
model. http://www.ldes.unige.ch/ang/publi/<br />
articles/unesco_AG_96/unesco96.htm, 1996.<br />
[8] I. Kalas and R. T. Mittermeir, editors. ISSEP 2011,<br />
volume 7013 <strong>of</strong> LNCS. Springer, 2011.
Teachersʼ Perceptions Of The<br />
Value Of Research-Based School Lectures<br />
Jonathan Black, Paul Curzon,<br />
Chrystie Myketiak, Peter W. McOwan<br />
Queen Mary <strong>University</strong> <strong>of</strong> London<br />
London<br />
{jonathanb, pc, chrystie,<br />
pmco}@eecs.qmul.ac.uk<br />
ABSTRACT<br />
A major challenge facing secondary schools is to encourage<br />
students to take computing courses. One approach is to invite<br />
external speakers from universities or industry to give lectures.<br />
The cs4fn project, a large UK-based initiative to enthuse students<br />
about computer science, includes this approach. Speakers from<br />
Queen Mary, <strong>University</strong> <strong>of</strong> London, visit schools to talk to<br />
students about computer science research. Our interactive talks<br />
tell engaging research-based stories on topics such as artificial<br />
intelligence and human-computer interaction as well as using<br />
magic tricks to illustrate computing principles. We asked teachers<br />
to complete post-talk surveys online; in particular we were<br />
interested in whether they believed students’ perceptions <strong>of</strong> the<br />
subject had changed. They reported that their students’ views <strong>of</strong><br />
computer science were improved, and that they felt students were<br />
more likely to take classes in computing in the future as a result <strong>of</strong><br />
the talk.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers and Education]: Computer and Information<br />
Science Education – Computer Science Education<br />
General Terms<br />
Human Factors.<br />
Keywords<br />
Public engagement, outreach, recruitment, K-12, schools,<br />
teachers, lectures, magic.<br />
1. INTRODUCTION<br />
A range <strong>of</strong> studies have looked at the effectiveness <strong>of</strong> lectures in<br />
higher education contexts, (e.g., Cooper & Foy, 1967) including<br />
those focusing on particular lecture styles such as PowerPoint<br />
lectures (e.g., Bartsch & Cobern, 2003) Despite the popularity <strong>of</strong><br />
the lecture approach in outreach activity, there is little published<br />
evidence about its effectiveness against the aims <strong>of</strong> public<br />
engagement activity and <strong>of</strong> its value in this context. This may be<br />
because organised computer science engagement projects are<br />
<strong>of</strong>ten workshop-style. Lecture-style approaches may be happening<br />
‘un<strong>der</strong> the radar’, missed out <strong>of</strong> evaluation work on workshopstyle<br />
programmes. However, the lecture style is widely used and<br />
warrants study.<br />
The cs4fn project (Curzon, 2007) highlights the importance <strong>of</strong><br />
flexible ‘take-away resources’ such as magazines and booklets<br />
rather than focusing just on workshops or lectures. We have found<br />
that this increases motivation and supports teachers (Myketiak et<br />
151<br />
Laura R. Meagher<br />
Technology Development Group<br />
Dairsie, Fife<br />
Laura.Meagher@ btinternet.com<br />
al., 2012). Curzon et al (2009) present evidence for the success <strong>of</strong><br />
this approach in engaging with potential students.<br />
In previous work we have surveyed audiences for our magic<br />
shows and showed their popularity with students (e.g., Curzon &<br />
McOwan, 2008). In this paper we evaluate the effectiveness <strong>of</strong> a<br />
lecture approach, surveying teachers about their own views and<br />
those <strong>of</strong> their students after seeing a talk. We focused on a range<br />
<strong>of</strong> indicators about the value <strong>of</strong> the talks but with a particular<br />
emphasis on their value in inspiring students to take their<br />
computing education further. We show that this kind <strong>of</strong> researchbased<br />
interactive lecture is highly valued by teachers and they<br />
believe it is effective in encouraging students to consi<strong>der</strong> taking<br />
the subject further both at school and university.<br />
2. METHOD<br />
We present feedback from two separate lectures given 19 times in<br />
the 2010-2011 and 2011-2012 academic years. We use audience<br />
volunteers in kinaesthetic activities to illustrate core concepts. The<br />
lectures also focus on telling engaging research stories. They do<br />
not attempt to cover curriculum topics directly, nor do they<br />
directly cover career choices or discuss university courses.<br />
Each time we delivered a talk at a school, we solicited a response<br />
to an online survey from the teacher. This meant the responses<br />
could be anonymous and gave the teacher an opportunity to talk to<br />
the students about the talk before filling out the survey. The<br />
survey consisted <strong>of</strong> 16 questions. We asked teachers to identify<br />
the lecture title, and which team member delivered it. The survey<br />
went on to ask general (non-identifying) details about the<br />
audience and the school that hosted the talk. There were then a<br />
series <strong>of</strong> Likert scale questions about the value <strong>of</strong> the approach,<br />
two yes/no questions followed by a series <strong>of</strong> open questions.<br />
3. TEACHERS’ PERCEPTION OF THE<br />
VALUE OF THE LECTURES<br />
3.1 Quantitative results<br />
The survey received 19 teacher responses. We asked the<br />
respondents to give an overall view <strong>of</strong> the lecture with options on<br />
a 5-point scale. All (100%) were positive. 89.5% (n=17) gave the<br />
highest rating (very good) while the remaining 2 teachers rated it<br />
as good. Respondents were also asked if the lecture met their<br />
needs and if they would recommend the lecture to other<br />
schools/teachers, with yes/no options in both cases. All 19<br />
teachers (100%) responded yes to both questions.<br />
When asked about students’ opinions, 84.2% (n=16) strongly<br />
agreed that their students had enjoyed the lectures with the<br />
remaining 3 agreeing. None were neutral or disagreed. 78.9%
(n=15) strongly agreed that their students found the lecture<br />
interesting while the remaining 4 agreed – a 100% positive<br />
response. 73.7% (n=14) strongly agreed that the lecture had<br />
improved the students’ un<strong>der</strong>standing <strong>of</strong> the subject with the<br />
remaining 5 agreeing.<br />
The last question in this section asked whether some students had<br />
changed their view <strong>of</strong> computer science in a positive way. 36.8%<br />
(n=7) strongly agreed. 42.1% (8) agreed and the remaining 21.1%<br />
were neutral. We asked whether, as a result <strong>of</strong> the lecture, one or<br />
more students were now more likely to consi<strong>der</strong> taking computing<br />
subjects further at school. 20% (n=3) strongly agreed with this<br />
statement, 53.3% (n=8) agreed (i.e. 73.3% <strong>of</strong> responses were<br />
positive) and the remaining 4 were neutral. Four teachers did not<br />
answer.<br />
We asked a similar question about students’ university choices. In<br />
response, 23.5% (n=4) strongly agreed that one or more students<br />
were more likely to go on to take the subject at university, 52.9%<br />
(n=9) agreed, meaning 76.4% <strong>of</strong> responses were positive. A<br />
further 17.6% (n=3) were neutral, 1 person strongly disagreed and<br />
1 did not answer the question. The teacher who strongly disagreed<br />
was otherwise very positive about the lecture, and did not suggest<br />
any improvements. It is unclear why they answered the question<br />
the way they did.<br />
3.2 Comments<br />
The comments in response suggest that teachers believe the form<br />
<strong>of</strong> a lecture affects student enjoyment as much as its content. A<br />
typical comment said, “The lecture was fast-paced, dynamic and<br />
really challenged the students [...] This was a very appealing way<br />
to promote the study <strong>of</strong> ICT.” Another teacher approved <strong>of</strong> “the<br />
mix <strong>of</strong> talking, video and practical examples.” Many others<br />
mentioned that the kinaesthetic activities were the students’<br />
favourite parts <strong>of</strong> the lectures. Comments by teachers on the<br />
benefits <strong>of</strong> the interactive elements followed a couple <strong>of</strong> themes.<br />
Some mentioned stylistic benefits (e.g., increasing student<br />
enjoyment and engagement), while others mentioned contentrelated<br />
benefits (e.g., modeling <strong>of</strong> computer functionality and<br />
opening students’ minds).<br />
4. DISCUSSION AND CONCLUSIONS<br />
We have evaluated teachers’ perceptions <strong>of</strong> the value <strong>of</strong> university<br />
lectures taking place in schools. This involves lectures that draw<br />
from research topics and involve interactivity using a kinaesthetic<br />
approach. In particular we focused on teachers’ perceptions about<br />
the immediate effect <strong>of</strong> the lectures and whether it had changed<br />
students’ attitudes to taking the subject further.<br />
Teachers were confident that their students had found the lectures<br />
interesting and now un<strong>der</strong>stood the subject better. They were less<br />
strong but still very positive in their agreement that the lectures<br />
had changed their students’ perception about computing, and that<br />
they had influenced students’ plans for the future. This may<br />
reflect the difficulty <strong>of</strong> one lecture to change overall perceptions<br />
or future plans, but it may also simply be a reflection <strong>of</strong> the<br />
difficulty teachers face in assessing and reporting students’ views.<br />
5. FUTURE WORK<br />
This pilot study examined teacher responses to 19 lectures about<br />
computing. We hope to gather more responses in a larger study to<br />
come. We also intend to gather more direct and in-depth data on<br />
student perceptions. Finally, we are in the process <strong>of</strong> collecting<br />
152<br />
and analysing recruitment data to examine the ultimate effect <strong>of</strong><br />
talks on recruitment at university. Whilst this study has focused<br />
on the value <strong>of</strong> such talks to teachers, this ongoing work will<br />
explore the value to universities.<br />
6. ACKNOWLEDGMENTS<br />
The cs4fn programme is funded by EPSRC research agreement<br />
(EP/F032641/1) with additional support from Google’s CS4HS<br />
programme. CHI+MED: Multidisciplinary Computer-Human<br />
Interaction research for the design and safe use <strong>of</strong> interactive<br />
medical devices project is funded by EPSRC research agreement<br />
EP/G059063/1.<br />
7. REFERENCES<br />
[1] Bartsch , R.A. & Cobern, K.M. 2003. Effectiveness <strong>of</strong><br />
PowerPoint presentations in lectures, Computers &<br />
Education, 41 (1) August, 77–86, Elsevier, DOI=<br />
http://dx.doi.org/10.1016/S0360-1315(03)00027-7<br />
[2] Cooper, B. & Foy, J.M. 1967. Evaluating the effectiveness <strong>of</strong><br />
lectures, Higher Education Quarterly, 21(2) 182-185. March.<br />
DOI: 10.1111/j.1468-2273.1967.tb00231.x<br />
[3] Curzon, P. 2007. "Serious Fun in Computer Science", 12th<br />
Annual Conference on Innovation and Technology in<br />
Computer Science Education, organised by the ACM Special<br />
Interest Group on Computer Science Education (SIGCSE),<br />
ACM SIGCSE Bulletin 39(3) p1, DOI:<br />
10.1145/1269900.1268785<br />
[4] Curzon, P., Black, J., Meagher, L.R., McOwan, P.W. 2009.<br />
“cs4fn.org: Enthusing students about Computer Science”,<br />
Proceedings <strong>of</strong> Informatics Education Europe IV, Christoph<br />
Hermann, Tobias Lauer, Thomas Ottmann and Martina<br />
Welte (Eds.), pp73-80, Freiburg, <strong>Germany</strong>, November.<br />
[5] Curzon, P. & McOwan, P.W. 2008. “Engaging with<br />
Computer Science through Magic Shows”, ITiCSE 2008, The<br />
13th ACM SIGCSE Annual Conference on Innovation and<br />
Technology in Computer Science Education, ACM SIGCSE<br />
Bulletin 40 (3), pp179-183. June 30-July 2, Madrid, Spain.<br />
DOI: 10.1145/1384271.1384320<br />
[6] Myketiak, C., Curzon, P., Black, J., McOwan, P.W., and<br />
Meagher, L.R. 2012. cs4fn: a flexible model for computer<br />
science outreach. In Proceedings <strong>of</strong> the 17th ACM annual<br />
conference on Innovation and technology in computer<br />
science education (ITiCSE '12). ACM, New York, NY,<br />
USA, 297-302. DOI: 10.1145/2325296.2325366
Technocamps: Bringing Computer Science to the far west<br />
ABSTRACT<br />
Roger D. Boyle<br />
Computer Science<br />
Aberystwyth <strong>University</strong><br />
Penglais<br />
Aberystwyth SY23 3DB<br />
Wales<br />
rob21@aber.ac.uk<br />
Technocamps is a 3-year EU funded project to bring an<br />
awareness <strong>of</strong> technical Informatics to the 11-19 age group in<br />
the ‘Convergence zone’ <strong>of</strong> Wales. The project is coordinated<br />
through four universities, with materials and activities being<br />
developed in each <strong>of</strong> the academic hubs. Projects <strong>of</strong> this<br />
scale are rare, both in terms <strong>of</strong> geographic area and financial<br />
backing, creating a new set <strong>of</strong> challenges and opportunities.<br />
We review the background to the project and its need,<br />
outline its activities during its first year and the challenges<br />
they have presented, and make observations about portability<br />
and <strong>der</strong>ived good practice that might inform similar<br />
projects elsewhere. Longer term evaluation and longitudinal<br />
study will develop during the next two years <strong>of</strong> the work.<br />
1. BACKGROUND<br />
Wales is a small country at the west <strong>of</strong> the UK mainland<br />
with a clear and illustrious culture, including a vibrant language.<br />
It is predominantly rural, although the south <strong>of</strong> the<br />
country became pre-eminent in [now defunct] coal mining<br />
and associated heavy industry. It is a victim <strong>of</strong> rural poverty,<br />
seasonal (un)employment and the collapse <strong>of</strong> industry with<br />
no compensatory re-investment.<br />
Informatics in UK schools is in a difficult state <strong>of</strong> transition<br />
[3]. 2012 saw very consequential announcements by<br />
government [5] and the Royal Society [6]. In particular the<br />
latter proposes the teaching <strong>of</strong> Informatics be partitioned<br />
three ways:<br />
Hannah M. Dee<br />
Computer Science<br />
Aberystwyth <strong>University</strong><br />
Penglais<br />
Aberystwyth SY23 3DB<br />
Wales<br />
hmd1@aber.ac.uk<br />
• Computer Science [CS]: The rigorous academic discipline,<br />
encompassing programming languages, data structures,<br />
algorithms, etc.<br />
• Information Technology [IT]: The use <strong>of</strong> computers,<br />
in industry, commerce, the arts and elsewhere, including<br />
aspects <strong>of</strong> IT systems architecture, human factors,<br />
project management, etc.<br />
• Digital Literacy: The general ability to use computers.<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WiPSCE 2012, Hamburg, <strong>Germany</strong><br />
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$15.00.<br />
153<br />
Frédéric Labrosse<br />
Computer Science<br />
Aberystwyth <strong>University</strong><br />
Penglais<br />
Aberystwyth SY23 3DB<br />
Wales<br />
ffl@aber.ac.uk<br />
In brief, it is widely agreed that Digital Literacy should<br />
be achieved by all but is a skill akin to the ability to read,<br />
unsuited to formal curriculum time or assessment in high<br />
schools. IT is a respectable and necessary academic activity<br />
but is distinct from CS. It is the very weak condition <strong>of</strong> CS<br />
that has provoked concern in influential quarters.<br />
2. IMPLEMENTATION<br />
2.1 Organisation<br />
Technocamps [7] is an EU Convergence European Social<br />
Fund project, specifically designed to address the CS gap<br />
within the 11-19 age group within the Welsh Convergence<br />
Zone. Its aims are: to encourage interest in Science and<br />
Technology; to provide opportunities for students to gain<br />
insight into the practical application <strong>of</strong> cognate STEM subjects<br />
in a work environment; to increase the number <strong>of</strong> girls<br />
taking up science and technology; to increase the number <strong>of</strong><br />
students who progress to computer science, technology and<br />
engineering at higher levels. The 3-year project attracted<br />
funding <strong>of</strong> £6M and involves 4 universities across the zone.<br />
By design, activities are where possible portable between<br />
sites. Significant transport problems and sparse population<br />
mean that in most cases schools are served by their closest<br />
HEI. After decades <strong>of</strong> decline, the Welsh language is enjoying<br />
a slow but steady recovery, which is strongest in many <strong>of</strong><br />
the more isolated areas the project serves [8]. Accordingly,<br />
materials and activities are being made bilingually.<br />
The project targets 11-19 year-olds, and especially females<br />
and those who are not in education, employment, or training.<br />
‘Engagement’ is defined not to be a passing contact,<br />
and should be a minimum <strong>of</strong> 6 hours <strong>of</strong> exposure. It is acknowledged<br />
that 6 consecutive hours would be unlikely, and<br />
the precise target is to meet an individual at least twice for<br />
at least 3.5 hours each time.<br />
2.2 Activities<br />
One day engagements: Most contact is via one-day contact;<br />
material is specifically designed to be distinct from<br />
the school curriculum. Sessions last about 4 hours, with<br />
topics ranging from general (Scratch programming) to the<br />
more rarefied (such as cryptography, robotics, or Arduino<br />
[1] based work).<br />
Bootcamps: ‘Bootcamps’ are 3-day engagements: robot<br />
navigation, and Arduino projects (sailing robots and (digitally<br />
ornamented clothing) have been topics. Clearly, three<br />
full days provide a depth <strong>of</strong> opportunity in excess <strong>of</strong> 3.5<br />
hours: this is an opportunity <strong>of</strong> great use to research-oriented
projects.<br />
Other: There are several other miscellaneous engagements:<br />
these include after-school ‘Technoclubs’, Science week [2],<br />
Home Schoolers, and other sundry public engagements such<br />
as the Aberystwyth ‘Beachlab’.<br />
3. EXPERIENCES<br />
Attempting to engage the schools is challenging; the country<br />
is sparsely populated with few top-quality roads – reaching<br />
parts <strong>of</strong> the population can be very time-consuming, and<br />
schools are reluctant to give up staff and pupil time to extracurricular<br />
activity. Schools prefer to devote time directly before<br />
or after holiday periods (to minimise disruption) which<br />
provokes serious congestion for university resource.<br />
One-day engagements are held either in the university or a<br />
school, and <strong>of</strong>ten present transport or scheduling difficulties.<br />
Subject matter is not hard to generate: programming, computer<br />
applications <strong>of</strong> music, robotics, AI are just some examples.<br />
We learn in delivery what may be obvious to trained<br />
schoolteachers: we have to take care to schedule breaks and<br />
variation into what is otherwise a very long experience.<br />
Bootcamps differ inasmuch as children have volunteered<br />
for the event and so the level <strong>of</strong> motivation is much higher.<br />
Not only is length <strong>of</strong> time appreciably greater, but the intensity<br />
and depth <strong>of</strong> the exercise allows an immersion that<br />
would be difficult to reproduce in the classroom. These activities<br />
have been ambitious and very successful: young participants<br />
have mastered programming for calibration and<br />
navigation (in C), full bottom-up construction, basic electronics,<br />
circuitry, and sewing.<br />
Miscellaneous engagements have productively exposed many<br />
hundreds <strong>of</strong> passers-by to robots, AI, UAVs, wearables, etc.,<br />
and were introduced to the project and its aims.<br />
4. OBSERVATIONS<br />
Resourcing: European funding has made it possible to focus<br />
on issues <strong>of</strong> logistics and liaison that are <strong>of</strong>ten victims<br />
<strong>of</strong> un<strong>der</strong>funded or volunteer outreach activity. There is no<br />
expectation <strong>of</strong> continuation resource and it is incumbent on<br />
the project to leave some lasting impact. We see lasting<br />
influence on teachers’ un<strong>der</strong>standing and enthusiasm as the<br />
best aim.<br />
Challenge <strong>of</strong> materials: We are delighted to witness that<br />
our workshops are all accessible, and that perhaps we can<br />
be more ambitious. We do not suggest that 100% <strong>of</strong> the<br />
class walk away with perfect un<strong>der</strong>standing, but are convinced<br />
that the great majority are having some aspect <strong>of</strong><br />
their un<strong>der</strong>standing <strong>of</strong> CS improved.<br />
Targets: The primary metric <strong>of</strong> the funding authority is<br />
pupil contact numbers; with hindsight, the numbers are not<br />
ambitious; all schools have been co-educational and it has<br />
not been difficult to engage with a good number <strong>of</strong> girls. We<br />
consi<strong>der</strong> it more fruitful to engage with the 11-14, rather<br />
than the 15-19 bracket. Geographic coverage has been uneven<br />
– many areas <strong>of</strong> the country are very rural with very<br />
small schools<br />
Evaluation: Short term evaluation has been achieved by<br />
crude questionnaires before and after formal engagements.<br />
While these are <strong>of</strong> questionable value, we can see a clear<br />
shift in view <strong>of</strong> science and technology toward the positive.<br />
Portability: Materials are routinely packaged for delivery<br />
across all sites, and in due course more generally. Complete<br />
154<br />
‘bundles’ are posted as freely web-available resource.<br />
Staffing: Very few university staff have the training or skills<br />
that go with full-time teaching and from time to time this<br />
has been an issue. Teachers will know classes well and will<br />
defuse both the over-eager or disruptive, and will have techniques<br />
to motivate the bored. Often, Technocamps staff<br />
had to learn quickly how to ‘crowd manage’, for example by<br />
having ready a repertoire <strong>of</strong> illustrative practical activities<br />
or videos that provide relevant variety.<br />
Welsh: Serious students <strong>of</strong> CS in most countries would benefit<br />
from working at least partly in English, and pursuing it<br />
in a minority language such as Welsh would be eccentric.<br />
Nevertheless, many <strong>of</strong> the target group operate in Welsh as<br />
a first language. It has not been easy to meet this need<br />
from the project personnel, and a solution has to be found<br />
by full engagement <strong>of</strong> Welsh speaking teachers who share<br />
the project aims.<br />
The project has been a success by all measurements: pupil<br />
and teacher reaction is positive and the targets are being hit<br />
with ease. It will continue to develop for the next 21 months,<br />
improving what we do, particularly in devising techniques<br />
for longitudinal monitoring. We aim to leave in place a suite<br />
<strong>of</strong> web-readable materials most <strong>of</strong> which will be applicable<br />
anywhere and would permit anyone so minded to bring ‘real<br />
CS’ into schools <strong>of</strong> their choice. Generally, we are accruing<br />
a range <strong>of</strong> first-hand experience <strong>of</strong> putting challenging CS in<br />
front <strong>of</strong> children much younger than those who normally see<br />
it, and doing so <strong>of</strong>ten in unusual and sometimes challenging<br />
environments (and languages!). We are developing conclusions<br />
on what represents best practice in these endeavours,<br />
some <strong>of</strong> which we have presented here. At project end, we<br />
will collate and publish these as a guide to efficient HE intervention<br />
in high school Informatics.<br />
5. REFERENCES<br />
[1] Arduino. 2012. http://www.arduino.cc/.<br />
[2] British Science Association. National Science and<br />
Engineering Week, 2012. http:<br />
//www.britishscienceassociation.org/web/nsew/.<br />
[3] Computing at School. Computing For the Next<br />
Generation . . . , 2012.<br />
http://www.computingatschool.org.uk/.<br />
[4] M. Gove MP. Speech given to BETT, January 11 th ,<br />
2012. http://www.guardian.co.uk/education/2012/<br />
jan/11/digital-literacy-michael-gove-speech.<br />
[5] Royal Society. Shut down or restart?, January 2012.<br />
Final report <strong>of</strong> the Computing in UK Schools group<br />
http://royalsociety.org/education/policy/<br />
computing-in-schools/report/.<br />
[6] Technocamps. Creating the Next Generation <strong>of</strong><br />
Technologists, 2012. http://www.technocamps.com/.<br />
[7] Wikipedia. History <strong>of</strong> the Welsh language, 2012.<br />
http://en.wikipedia.org/wiki/History_<strong>of</strong>_the_<br />
Welsh_language.
Abenteuer Informatik – Hands-on exhibits for learning<br />
about computational thinking<br />
Dr. Jens Gallenbacher<br />
<strong>Didaktik</strong> <strong>der</strong> Informatik,<br />
Technische Universität Darmstadt<br />
Hochschulstr. 10<br />
64289 Darmstadt<br />
+49 6151 16-2573<br />
jg@di.tu-darmstadt.de<br />
Abstract<br />
Computational thinking is one <strong>of</strong> the pillars <strong>of</strong> the ACM-CSTA<br />
standards for teaching computer science from kin<strong>der</strong>garten to<br />
college. Our approaches Abenteuer Informatik – Informatik<br />
begreifen (adventures in informatics – hands on computer<br />
science) and Abenteuer Technik are well established in the<br />
german-speaking community as means to connect computer<br />
science with other subjects and as means <strong>of</strong> clarifying some<br />
prejudices against computer science, especially problematic for<br />
establishing computer science as subject in schools.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computer and Information Science Education]:<br />
Curriculum<br />
General Terms<br />
Human Factors<br />
Keywords<br />
Computational thinking, curriculum, computer science education,<br />
Abenteuer Informatik, computer science unplugged<br />
1. Introduction<br />
Since the invention <strong>of</strong> commercial electronic computers in the<br />
middle <strong>of</strong> the last century the public perception <strong>of</strong> relevant<br />
aptitudes for computer-users shifted continuously: From<br />
scientificly trained expert over the well-informed programmer to<br />
the "just"-user today. In the same manner the computer-science<br />
education in schools changed.<br />
While in the beginning it has been obligatory to learn basics <strong>of</strong><br />
computer-technology and programming, because standardapplications<br />
were not available or to expensive for schools,<br />
today's focus <strong>of</strong> cs-education <strong>of</strong>ten is computer literacy, like the<br />
curriculum for european computer driving licence.<br />
The stereotypic public and political opinion about computer<br />
science also changed radically: On one hand the search for the<br />
correct buttons in <strong>of</strong>fice, on the other hand a cryptic science, that<br />
is everything about Devices and nothing about humans.<br />
Since 2006 Jeannette M. Wing, Pr<strong>of</strong>essor at CMU, gives<br />
distinction to the term "computational thinking" and elaborated,<br />
that computer science is not the science about computers. It<br />
particularly produces concepts <strong>of</strong> (human) thinking, that amongst<br />
others may be used to program machines, but mostly benefit other<br />
155<br />
sciences and dayly use. In short, Wing proved computer science<br />
provides general education (in europe <strong>of</strong>ten stated by the german<br />
term "allgemeinbildend"). She legitimized "true" computer<br />
science as subject in general schools, so computational thinking is<br />
positioned prominently in the standards <strong>of</strong> the CSTA and in the<br />
proposal for a computer science curriculum in Great Britain,<br />
which are likely to be implemented by many schools.<br />
Adventures in informatics – hands on computer science is a very<br />
practical approach with similar goals. The book [Gallenbacher<br />
2012] is sold over 10.000 times and in the exhibition over<br />
100.000 people played, puzzled to get their hands on computer<br />
science at 31 different locations, amongst others ars electronica<br />
center in Linz and Heinz Nixdorf Forum in Pa<strong>der</strong>born. It<br />
fascinated kids as well as their grandparents and many politicians.<br />
Reason enough to evaluate, how both approached could benefit<br />
from each other.<br />
2. Computational Thinking<br />
Wing describes “Computational Thinking is the thought<br />
processes involved in formulating problems and their solutions so<br />
that the solutions are represented in a form that can be effectively<br />
carried out by an information-processing agent Informally,<br />
computational thinking describes the mental activity in<br />
formulating a problem to admit a computational solution. The<br />
solution can be carried out by a human or machine, or more<br />
generally, by combinations <strong>of</strong> humans and machines.” [Wing<br />
2010]<br />
A position paper published by the german “Gesellschaft für<br />
Informatik” [Biundo 2006] states the situation in a very similar<br />
way.<br />
3. Abenteuer Informatik, Abenteuer Technik<br />
and Computational Thinking<br />
Abenteuer Informatik as concept to teach computer science<br />
without computers provides many approaches for teaching<br />
computer science with focus on computational thinking. The full<br />
exhibits referenced by the examples in this chapter may be<br />
downloaded from www.abenteuer-informatik.de - most <strong>of</strong> them<br />
are available in german, english and spanish language.
The Monkey Puzzle<br />
Sometimes you know exactly how to solve a problem but fail<br />
nonetheless, because it would cost too much time or is "not<br />
scalable". Problems <strong>of</strong> that kind exist in many disciplines.<br />
If you need a bridge over a ditch 1m wide, you only need to lay a<br />
metal or stone plate across it, and it‘s done ! However, a bridge<br />
across a gap <strong>of</strong> 10m cannot be made <strong>of</strong> ten such plates laid end to<br />
end; we need first a precise static analysis <strong>of</strong> the construction,<br />
with an outlay consi<strong>der</strong>ably greater than 10 times that <strong>of</strong> the first<br />
example. If you want a bridge over a gap <strong>of</strong> 100m , the necessary<br />
outlay will be many times greater. The longest bridge span<br />
achieved to date is about 2000m. Each additional meter requires<br />
an immeasurably greater planning effort and immense material<br />
and building costs.<br />
The peculiarity <strong>of</strong> the second set <strong>of</strong> monkey puzzles is not just<br />
that the computer would take a very long time to find a solution;<br />
it is fundamentally incapable <strong>of</strong> answering this problem! So<br />
computational thinking provides means <strong>of</strong> recognizing, that there<br />
are certain problems, which by their nature can never be<br />
completely solved algorithmically.<br />
Informatics Letter by Letter<br />
The formal term "information" is very important to computer<br />
science, as expressed by the also used "informatics", which is<br />
<strong>der</strong>ived from Informatik, build <strong>of</strong> information and automation.<br />
Information is important to all disciplines in our mo<strong>der</strong>n world -<br />
but what is information at all?<br />
An approach by Shannon to use language for defining information<br />
can easily be adapted for computational thinking. Please try to<br />
decipher the following text. It is from a well known story, but is<br />
missing all except the frequently used letters.<br />
hen aune as tee eas o, the enhantess shut he in a toe, hih a in a<br />
oest. e toe ha no oo, ut hih u as a ino. hen the enhantess ante to<br />
o in, she ae hese eo the ino, an ie „aune, aune, et on ou hai.“ aune<br />
ha aniient on hai, ine as sun o. hen she hea the oie o the enhantess<br />
she oun he hai oun a hoo o the ino. e hai e tent as on, an the<br />
enhantess ie u it.<br />
Interesting, isn‘t it? Here we have retained about 70% <strong>of</strong> the<br />
original letters, but the result is gibberish. Below is the same text,<br />
with almost the same number <strong>of</strong> letters as in the previous<br />
example, obtained by removing only the vowels from the original.<br />
Can you read the story now?<br />
Whn Rpnzl ws twlv yrs ld, th nchntrss sht hr n twr, whch ly n frst.<br />
twr hd n dr, bt hgh p ws wndw. Whn th nchntrss wntd t g n, sh<br />
plcd hrslf blw th wndw, nd crd „Rpnzl, Rpnzl, lt dwn yr hr.“<br />
Rpnzl hd mgnfcnt lng hr, fn s spn gld. Whn sh hrd th vc f th<br />
nchntrss sh wnd hr hr rnd hk f th wndw. hr � l twnty yrds dwn,<br />
nd th nchntrss clmbd p by t.<br />
Apparently the amount <strong>of</strong> information has to do something with<br />
the frequency: A text containing just common symbols contains<br />
no information or at least less information than the text with<br />
infrequent symbols.<br />
In computer science, this knowledge is used to encode texts and<br />
other information and save memory or bandwidth, e.g. by using a<br />
huffman-tree or just zipping files. In other disciplines, this is used<br />
as well - like in linguistics: In ancient sanskrit the symbol for T<br />
followed by A, which is a very common combination, is written<br />
as<br />
156<br />
If you want to use T without A, which is shorter, but less<br />
common, you have to write with more strokes:<br />
So computational thinking is not a new discipline, but known for<br />
some millennia...<br />
Abenteuer Technik<br />
In our science lab for children we implicitly use computational<br />
thinking too. We give out tasks like "building wings for an<br />
efficient wind generator out <strong>of</strong> some cardboard, wood and plastic<br />
film" or "building a water wheel generator out <strong>of</strong> milk carton".<br />
The degrees <strong>of</strong> freedom are intentionally set very high, so the<br />
children have to model the problem first and then make a<br />
conscious decision to focus on optimizing one or two out <strong>of</strong> more<br />
than 100 possible parameters.<br />
4. Conclusion<br />
Approaches to teach computer science and engineering without<br />
using complex and abstract technology, like [Bell 2006],<br />
[Gallenbacher 2008, 2012] are most suitable to implement<br />
computational thinking in computer science education.<br />
5. References<br />
[1] [Bell 2006] Tim Bell, Ian H. Witten, Mike Fellows:<br />
Computer Science Unplugged - Teachers edition,<br />
downloaded from http.//www.csunplugged.org/<br />
[2] [Biundo 2006] Susanne Biundo, Volker Claus, Heinrich C.<br />
Mayr: Was ist Informatik, unser Positionspapier,<br />
Gesellschaft für Informatik, 2006<br />
[3] [CSTA 2010] Cameron Wilson, Leigh Ann Sudol, Chris<br />
Stephenson, Mark Stehlik: Running On Empty: The Failure<br />
to Teach K–12 Computer Science in the Digital Age, Report<br />
<strong>of</strong> the ACM-CSTA, http://www.acm.org/Runningonempty/<br />
[4] [CSTA 2011] Deborah Seehorn, Stephen Carey, Brian<br />
Fuschetto, Irene Lee, Daniel Moix, Dianne O’Grady-<br />
Cunniff, Barbara Boucher Owens, Chris Stephenson, Anita<br />
Verno: CSTA K–12 Computer Science Standards Revised<br />
2011, Report <strong>of</strong> the ACM-CSTA, ISBN 978-1-4503-0881-6<br />
[5] [Gallenbacher 2008] Jens Gallenbacher: Abenteuer<br />
Informatik, Exhibition, http://www.abenteuer-informatik.de<br />
[6] [Gallenbacher 2012] Jens Gallenbacher: Abenteuer<br />
Informatik, Springer-Spektrum, ISBN 978-3827429650<br />
[7] [Wing 2006] Jeannette M. Wing: Computational thinking,<br />
COMMUNICATIONS OF THE ACM March 2006/Vol. 49,<br />
No. 3<br />
[8] [Wing 2008] Jeannette M. Wing: Computational thinking<br />
and thinking about computing, Philosophical Transactions <strong>of</strong><br />
the Royal Society A (2008) 366, 3717–3725<br />
[9] [Wing 2010] Jeannette M. Wing: Computational Thinking:<br />
What and why?, CMU, 17. November 2010
Gaming and Mathematics: A Cross Curricular Event<br />
Sharon Jones Ed.D<br />
Charlotte Mecklenburg Schools<br />
1430 Alleghany St.<br />
Charlotte, NC 28208<br />
1-980-343-5992<br />
sharont.jones@cms.k12.nc.us<br />
ABSTRACT<br />
Computer Science education can motivate students, through<br />
students’ interests and experiences but, it can be a challenge to<br />
create meaningful and engaging assignments that allow for both<br />
creativity and learning while also using mo<strong>der</strong>n technology<br />
practices. This poster will highlight the liaison between computer<br />
science and math as students use Build Your Own Blocks<br />
(BYOB) to build projects that will use traditional computation<br />
math skills with computer gaming processes. The poster will<br />
highlight an after school workshop success story.<br />
1. INTRODUCTION/BACKGROUND<br />
Computing or computer science as a scientific discipline predates<br />
the invention <strong>of</strong> computers. The first decades <strong>of</strong> the twentieth<br />
century saw development <strong>of</strong> fundamental concepts in this<br />
discipline (Gal-Ezer, Beer, Harel, Yehudai, 1995). More recently,<br />
fueled in part by the invention <strong>of</strong> computers and their widespread<br />
use, the study <strong>of</strong> computing has bloomed, and CS is now<br />
recognized as an autonomous scientific discipline (Gal-Ezer, et.<br />
al, 1995). The disciplines concepts include the analysis <strong>of</strong><br />
algorithmic processes as well as the design and implementation <strong>of</strong><br />
computing systems and these concepts influence work in other<br />
disciplines (Gal-Ezer, et. al, 1995). As the ideas <strong>of</strong> the discipline<br />
and the importance <strong>of</strong> technology innovation become assimilated<br />
in our everyday culture it becomes clear that computer science<br />
high school curriculums should be implemented to reflect the<br />
growing importance (Gal-Ezer, et. al, 1995). However, high<br />
school computer science education in the United States has shown<br />
significant declines in both the number <strong>of</strong> introductory CS courses<br />
being taught (CSTA, 2010), and the intention <strong>of</strong> students<br />
(especially those from un<strong>der</strong>represented populations) declaring<br />
computing as a major (HERI, 2009). However, the U.S.<br />
Department <strong>of</strong> Labor projects that between 2008 and 2018, 1.4<br />
million computing jobs will have opened in the U.S. If current<br />
graduation rates continue, only 61% <strong>of</strong> these jobs could be filled<br />
by U.S. computing degree-earners (NCWIT, 2011). When<br />
including only computing bachelor’s degrees, this percentage<br />
drops to 29% <strong>of</strong> projected job openings that could be filled.<br />
American students need a 21st-century computing education if we<br />
want a workforce that is innovative, competitive, and wellemployed<br />
(NCWIT, 2011). To try and change the educational<br />
landscape, the National Science Foundation (NSF) has put in<br />
place an initiative, the CS10K Initiative, to have 10,000 teachers<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that<br />
copies bear this notice and the full citation on the first page. To copy<br />
otherwise, or republish, to post on servers or to redistribute to lists,<br />
requires prior specific permission and/or a fee.<br />
Conference’10, Month 1–2, 2010, City, State, Country.<br />
Copyright 2010 ACM 1-58113-000-0/00/0010 …$15.00.<br />
Renada Poteat<br />
Charlotte Mecklenburg Schools<br />
11201 Old Statesville Road<br />
Huntersville, NC 28078<br />
1-980-343-3842<br />
renada.poteat@cms.k12.nc.us<br />
157<br />
Beth Frierson<br />
Charlotte Mecklenburg Schools<br />
1430 Alleghany St.<br />
Charlotte, NC 28208<br />
1-980-343-5992<br />
mary.frierson@cms.k12.nc.us<br />
teaching computer science by 2015.<br />
To support the NSF’s CS10K initiative and Computer Science<br />
Principles efforts to prepare both high school teachers and<br />
students to be creators in computing, the course Beauty and Joy <strong>of</strong><br />
Computing (BJC) was created. This course was developed to<br />
encourage students to experience Computer Science in a new way.<br />
The BJC course combines Moodle-based computer programming<br />
labs, lectures ranging from artificial intelligence and parallelism<br />
to the social implications <strong>of</strong> computing and technology, and small<br />
discussion sections. In the non-programming part <strong>of</strong> the course<br />
there is a balance <strong>of</strong> fundamental optimism about the future <strong>of</strong><br />
computer technology with an un<strong>der</strong>standing <strong>of</strong> its limitations and<br />
potential. The BJC class has been implemented in five major<br />
universities with great success. Therefore, to reach the CS10K<br />
goal, the BJC curriculum has begun to be adapted for the K-12<br />
curriculum. As selected members <strong>of</strong> the development team to<br />
bring BJC to the high schools, we have had the privilege <strong>of</strong> seeing<br />
firsthand the impact the course has on students. The course <strong>of</strong>fers<br />
insight into all aspects <strong>of</strong> computer science and helps to dispel<br />
misconceptions established about the computer science discipline<br />
and show cross curricular elements. The students are lead by the<br />
s<strong>of</strong>tware used in the class to build their own projects that link<br />
computer science and all other disciplines.<br />
Common Core standards are being implemented across the United<br />
States. Common Core standards define the knowledge and skills<br />
students should have within their K-12 education careers so that<br />
they will graduate high school able to succeed in entry-level,<br />
credit-bearing academic college courses and in workforce training<br />
programs (NCDPI, 2011). The standards stress not only<br />
procedural skill but also conceptual un<strong>der</strong>standing, to make sure<br />
students are learning and absorbing the critical information they<br />
need to succeed at higher levels - rather than the current practices<br />
by which many students learn enough to get by on the next test,<br />
but forget it shortly thereafter, only to review again the following<br />
year (NCDPI, 2011).<br />
Using the platform <strong>of</strong> common core and the computer science<br />
concepts, student exposure on how to integrate and un<strong>der</strong>stand<br />
how to apply subjects to multiple disciplines is significantly<br />
increased.<br />
2. WORKSHOP DETAILS<br />
2.1 Workshop Overview<br />
We created a cross departmental workshop, Get Your Game On,<br />
to assist students that were struggling with Algebra concepts and<br />
used computer gaming as a way differentiation strategy. The three<br />
day workshop after school taught students fundamental<br />
programming concepts (ie. variables, object, and looping) while<br />
infusing Algebra I questions provided by the schools math<br />
department. At the conclusion <strong>of</strong> the workshop students were able<br />
to learn basic programming concepts while reinforcing algebraic
concepts that they previously learned in class. The intentions were<br />
to introduce students to computer science concepts and how<br />
computer science is a part <strong>of</strong> every curriculum in a non-<br />
threatening environment and spark their interest to learn more<br />
about the discipline<br />
2.2 Workshop Data<br />
There were 15 (n = 15) students that participated in the workshop<br />
with 13 (86.7%) males and 2 (13.3%) females. The majority <strong>of</strong><br />
participants were in the 11 th grade (40%) and were 17 (33.3%).<br />
We implemented a limit on participants to allow evaluation and<br />
analysis. Prior to the workshop, students were asked to complete a<br />
survey including the following questions:<br />
1. I like to play games to learn educational material (PlayGames)<br />
2. I know what BYOB is and how to use it (BYOBis)<br />
3. I can see how math and gaming are connected (Connected)<br />
4. I am sure I can learn programming (LearnProg)<br />
5. I am interested in computer science (Interested)<br />
Results from the survey are represented in Table 1.<br />
Table 1. Workshop Pre Survey<br />
Variable Yes No Sometimes<br />
PlayGames 26.7 13.3 60.0<br />
BYOBis 0 100.0<br />
Connected 86.7 0 13.3<br />
LearnProg 100.0 0<br />
Interested 86.7 0 13.3<br />
On the last day <strong>of</strong> the workshop, students were asked to take a<br />
post survey to evaluate their experience and content. The<br />
following questions were included in the survey:<br />
1. Now that you know what BYOB is, would you use it in the<br />
future? (Future)<br />
2. Do you feel the game you created will help you learn a math<br />
concept? (LearnMath)<br />
3. After the workshop, I am more interested in Computer Science.<br />
(InterestedCS)<br />
4. Did you enjoy the workshop? (Enjoy)<br />
5. Would you recommend this workshop to someone?<br />
(Recommend)<br />
6. Any suggestions for future workshops? (Suggestions)<br />
Results from the survey are represented in Table 2.<br />
Table 2. Workshop Post Survey<br />
Variable Yes No Somewhat<br />
Future 90.0 10.0 0<br />
LearnMath 100.0 0 0<br />
InterestedCS 90.0 10.0 0<br />
Enjoy 100.0 0 0<br />
Recommend 80.0 0 20.0<br />
158<br />
The survey included a section for students to post suggestions to<br />
better the workshop. Below are a few responses:<br />
� I thought it was completely awesome and epic! it was<br />
great as is.<br />
� Try Gamemaker now that you have it. Other than that<br />
was an amazing course.<br />
� more time to make game<br />
� Tell more people :)<br />
� It was well put together, great job guys<br />
From the data collected, it was concluded students enjoyed the<br />
workshop.<br />
3. MOTIVATIONS<br />
Digital games, whether computer-, game console-, or handheldbased,<br />
are “Purposeful, goal-oriented, rule-based activity that the<br />
players perceive as fun” (Klopfer, 2008). Students are motivated<br />
by competition, challenge, and the ability to tell a story through<br />
interaction using the computer (Prenksy, 2001). The use <strong>of</strong> BYOB<br />
allowed students the interaction <strong>of</strong> creating the games along with<br />
the challenge <strong>of</strong> learning algebra in or<strong>der</strong> to make the game<br />
function properly.<br />
The workshop provided students the opportunity to work with<br />
educational gaming s<strong>of</strong>tware that they had not been privy to<br />
before. The actual learning <strong>of</strong> the s<strong>of</strong>tware motivated them to<br />
figure out the math problems in or<strong>der</strong> to have their games function<br />
properly. The use <strong>of</strong> the computer gaming s<strong>of</strong>tware was a catalyst<br />
to having students work on algebra problems that they would<br />
previously shied away from.<br />
4. SUMMARY<br />
Due to the positive feedback from students and teachers, we will<br />
use the format <strong>of</strong> this workshop as a template for a series <strong>of</strong><br />
computer science and math workshops. Through continuing to<br />
<strong>of</strong>fer opportunities for students to see the relevance <strong>of</strong> computer<br />
science and other disciplines, the outcome would predict<br />
increased interest and application.<br />
5. REFERENCES<br />
[1] (2005, January 1). Retrieved from Computer Science Teachers<br />
Association website: http://csta.acm.org/<br />
[2] (2012, January 1). Retrieved from National Center for Women<br />
and Information Technology website: http://www.ncwit.org/<br />
[3] Gal-Ezer, J., Beeri, C., Harel, D., & Yehudai, A. (1995). A<br />
high-school program in computer science. Computer Science,<br />
28(10), 73-80.<br />
[4] Klopfer, E. (2008). Augmented learning: Research and design<br />
<strong>of</strong> mobile educational games. Cambridge, MA: MIT Press.<br />
[5] Prensky, M. 2001. Digital game-based learning. New York:<br />
McGraw Hill.<br />
[6] Yu, Y. T. and Lau, M. F. (2006). A comparison <strong>of</strong> MC/DC,<br />
MUMCUT and several other coverage criteria for logical<br />
decisions. J. Syst. S<strong>of</strong>tw. 79, 5 (May. 2006), 577-590. DOI=<br />
http://dx.doi.org/10.1016/j.jss.2005.05.030.
Turi: Chatbot s<strong>of</strong>tware for schools in the Turing Centenary<br />
ABSTRACT<br />
Mathew Keegan<br />
Computer Science<br />
Aberystwyth <strong>University</strong><br />
Penglais<br />
Aberystwyth SY23 3DB<br />
Wales<br />
mtk8@aber.ac.uk<br />
We describe a workshop designed for 11-19 year-olds that<br />
consi<strong>der</strong>s the nature <strong>of</strong> intelligence and introduces the Turing<br />
test in various ways.<br />
Chatbots as mimics <strong>of</strong> intelligence are consi<strong>der</strong>ed at length.<br />
Pupils are invited to use our system Turi in which they can<br />
build and test their own chatbot.<br />
The materials are free, open source and available for all<br />
to download [1].<br />
Categories and Subject Descriptors<br />
K.3.1 [Computing Milieux]: COMPUTERS AND EDU-<br />
CATION Computer Uses in Education Miscellaneous<br />
General Terms<br />
Computer science education, artificial intelligence<br />
Keywords<br />
Alan Turing, Computer science education, Artificial Intelligence,<br />
Chatbots<br />
1. INTRODUCTION<br />
The Turing Test provides a readily un<strong>der</strong>stood means <strong>of</strong><br />
discussing the nature <strong>of</strong> intelligence; we have chosen the Turing<br />
Centenary year <strong>of</strong> 2012 to join the various schools outreach<br />
projects based on his work. We are doing this as part<br />
<strong>of</strong> the Technocamps project, a large, multi-site European<br />
funded project aimed at introducing technical informatics<br />
into schools across the convergence area <strong>of</strong> Wales [5].<br />
Students are introduced to the difficulty <strong>of</strong> defining intelligence;<br />
then with the addition <strong>of</strong> the Turi s<strong>of</strong>tware, they can<br />
get to grips with the ideas <strong>of</strong> pattern matching and callresponse<br />
conversations through practical interaction with<br />
∗ Corresponding author.<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
WiPSCE 2012, Hamburg, <strong>Germany</strong><br />
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$15.00.<br />
Roger D. Boyle<br />
Computer Science<br />
Aberystwyth <strong>University</strong><br />
Penglais<br />
Aberystwyth SY23 3DB<br />
Wales<br />
rob21@aber.ac.uk<br />
159<br />
Hannah M. Dee ∗<br />
Computer Science<br />
Aberystwyth <strong>University</strong><br />
Penglais<br />
Aberystwyth SY23 3DB<br />
Wales<br />
hmd1@aber.ac.uk<br />
a chatbot’s internal representation <strong>of</strong> sentences, written in<br />
AIML (Artificial Intelligence Markup Language).<br />
The activity we describe runs as a 4 hour workshop (but<br />
this would be easy to adjust). The s<strong>of</strong>tware is free, open<br />
source and available for all to download [1].<br />
2. BACKGROUND AND MOTIVATION<br />
AI chatbots have been widely used in teaching, but there<br />
is little if any use <strong>of</strong> them in a school context. We believe,<br />
however, that they are an excellent medium for introducing<br />
concepts <strong>of</strong> AI and for experimenting with simple pattern<br />
matching programs. Our s<strong>of</strong>tware facilitates this by providing<br />
a simple multi-user chatbot editor sitting on top <strong>of</strong> an<br />
AIML-based chatbot engine. AIML is a live AI Markup Language;<br />
it has the same visual appearance as many markup<br />
languages, with tags in angle brackets delimiting tokens.<br />
Thus students are exposed to a markup language, if only<br />
a subset <strong>of</strong> one.<br />
There are a number <strong>of</strong> excellent chatbot programs available<br />
to use for free on the Internet. In particular, we have<br />
used Cleverbot [3] and Jabberwacky [4] as examples <strong>of</strong> just<br />
how good a chatbot can be. This both motivates students<br />
and demonstrates what they should be doing with Turi.<br />
3. SOFTWARE IN CONTEXT<br />
Whilst it is possible to use the Turi s<strong>of</strong>tware as a standalone<br />
system, we have devised a workshop suited to mixedability<br />
pupils in the age range 11-19. We have found that<br />
with mixed-ability groups there is a real benefit in terms<br />
<strong>of</strong> student attention to embed computer use within a varied<br />
program <strong>of</strong> activities, including short video clips, discussions,<br />
pen-and-paper, and physical activities.<br />
The day proceeds initially by introducing students to the<br />
idea <strong>of</strong> the Turing test, exploring concepts surrounding Artificial<br />
Intelligence, and giving practical experience <strong>of</strong> two<br />
real chatbots:<br />
1. The Telephonic Turing Test: introducing the idea<br />
<strong>of</strong> the Turing test without actually using computers or<br />
referring to AI. A student and helper leave the room<br />
with a mobile phone. Students remaining in the class<br />
guess which <strong>of</strong> these two has the mobile phone, by<br />
asking questions using text messaging.<br />
2. Is it intelligent or not?: Printed pictures <strong>of</strong> various<br />
creatures and items are given to the class, one per student.<br />
These include some AIs from fiction and reality,<br />
some animals, and some wildcards (e.g. a rock, a tree,
Sherlock Holmes). Students sort themselves in or<strong>der</strong> <strong>of</strong><br />
intelligence, which can result in some illuminating and<br />
amusing conversations (“Is a washing machine more or<br />
less intelligent than a flea?”).<br />
Once or<strong>der</strong>ed, students indicate if they can do certain<br />
things (e.g., be creative). This shows that certain characteristics<br />
are more common at the intelligent end <strong>of</strong><br />
the line, triggering discussions on those features that<br />
are necessary/sufficient to call an agent intelligent.<br />
3. Chatbot comparison: We explore the idea <strong>of</strong> chatbots,<br />
and what kinds <strong>of</strong> conversations might expose<br />
the fact that they are programs rather than people.<br />
Students write down three questions to ask a chatbot,<br />
along with a sentence describing what the question is<br />
testing (e.g. “Does it have a memory?”). The questions<br />
are put to two chatbots and the answers recorded;<br />
the chatbots are then compared in discussion.<br />
4. TURI: SOFTWARE OVERVIEW<br />
The Turi s<strong>of</strong>tware provides an easy way for multiple instances<br />
<strong>of</strong> the open source chatbot Program O [2] to be constructed<br />
and allocated to individual students, with a simple<br />
tabular interface for students to build their chatbot by<br />
adding and editing AIML statements. It also enables students<br />
to chat with their own bot and those created<br />
Turi is web-based s<strong>of</strong>tware, written using MySQL and<br />
PHP within the CodeIgniter framework. Client side elements<br />
<strong>of</strong> the s<strong>of</strong>tware are written in HTML and JavaScript,<br />
and have been tested on all common web browsers and platforms.<br />
The server we currently use is a Virtual Private<br />
Server with 1GB <strong>of</strong> RAM and a dual-core processor. We<br />
have run side-by-side workshops with 30 children in each,<br />
meaning that 60 separate chatbots were being edited at the<br />
same time with no appreciable load visible in the server logs.<br />
The student interface is deliberately simple. The fundamental<br />
unit <strong>of</strong> the chatbot is the input-output pair, which<br />
defines an input to the chatbot and the output the chatbot<br />
will give upon receiving that specific input. There are<br />
three views for the student chatbot creator: New Phrase,<br />
Try Chatbot, and Chatbot Editor.<br />
New Phrase permits chatbot input-output pairs to be added<br />
or edited. Students type in input and the desired response –<br />
examples <strong>of</strong> the four main kinds <strong>of</strong> call-response supported<br />
are (see Table 1): simple call-response, *-response, *star<br />
and call-random. This permits users to copy-andpaste<br />
AIML templates for the common functions.<br />
Try Chatbot allows the user to test out phrases that have<br />
been entered, or pre-loaded. This provides a text box to<br />
type a conversation into, and responses are shown above the<br />
box scrolling upwards as the conversation continues.<br />
Chatbot Editor is the view in which users get to see all <strong>of</strong><br />
the input-output sentence pairs entered thus far, with the<br />
option to delete or to edit each one.<br />
All three views are minimalist in layout, with clear links<br />
and simple instructions. This simplicity is the result <strong>of</strong><br />
user testing early in the design phase, when we found that<br />
younger users were happiest with clear tabular layout.<br />
Administrative functions are hidden from the student view<br />
completely. Student chatbots are created by setting a passphrase<br />
for the session; students start Turi in a browser by<br />
typing the day’s passphrase into the login screen. Each time<br />
the passphrase is typed, Turi creates a new chatbot with a<br />
160<br />
I/O pair Description<br />
call-response Exact match <strong>of</strong> call generates response<br />
*-response Wildcard matches *, for a given response.<br />
*-star is replaced by text which<br />
matched * in the input.<br />
call-random Given the sentence which matches the input<br />
call, one <strong>of</strong> a random list is selected as<br />
response.<br />
combination It is possible to combine these.<br />
Table 1: The types <strong>of</strong> input-output pair covered by<br />
Turi in our sessions<br />
unique ID, and the student is ready to start. The instructor<br />
has no need to determine in advance exactly how many chatbots<br />
are required, and there is no need to give each student<br />
a login. At the end <strong>of</strong> the session the instructor can either<br />
drop the Turi instances for that class or can allow students<br />
to create a name and a password for their chatbot. In the<br />
second case students are able to continue working at home,<br />
or over multiple sessions.<br />
Turi has been written in English but this is easy to amend.<br />
Chatbots created by students can be in any language with a<br />
European alphabet – we have used chatbots successfully in<br />
classes where some students have been creating call-response<br />
pairs in French, German and Welsh.<br />
5. EVALUATION, EXPERIENCE, IMPROVE-<br />
MENTS<br />
We have run the workshop with several hundred pupils<br />
in the age range 12-19 (the number grows monthly). We<br />
re-iterate that the aim is to introduce them to ideas about<br />
AI and conversation, rather than to enable them to create<br />
conversational agents.<br />
The largest chatbot created during early engagements had<br />
27 input-output pairs, and the average size <strong>of</strong> the created<br />
chatbots had just un<strong>der</strong> 10. All response types were evident<br />
with the majority being simple. Later, we pre-seeded each<br />
chatbot with a set <strong>of</strong> phrases showing the kinds <strong>of</strong> things it<br />
is possible to do. This encouraged much greater creativity.<br />
Feedback from teachers has been very positive as the material<br />
is seen as novel and stimulating. We have found the<br />
module to be one <strong>of</strong> the more popular workshops we run,<br />
and disproportionately popular with girls. We have also deployed<br />
the material at open access events for the general<br />
public and provoked lengthy, involved interactions.<br />
We continue to develop: in particular an inbuilt spell<br />
checker has been found to be essential, and we will consi<strong>der</strong><br />
employing a speech synthesiser.<br />
6. REFERENCES<br />
[1] H. Dee. Technocamps AI module.<br />
http://users.aber.ac.uk/hmd1/ai.zip, 2012.<br />
[2] E. Perreau. Program-O, 2010.<br />
http://sourceforge.net/projects/program-o/.<br />
[3] Rollo Carpenter. Cleverbot, 2012.<br />
http://cleverbot.com/.<br />
[4] Rollo Carpenter. Jabberwacky, 2012.<br />
http://jabberwacky.com/.<br />
[5] Technocamps. Creating the Next Generation <strong>of</strong><br />
Technologists. <strong>University</strong> <strong>of</strong> Swansea, 2012.<br />
http://www.technocamps.com/.
ABSTRACT<br />
Learning Fields in Vocational IT Education –<br />
Why Teachers Refrain From Taking an Opportunity<br />
Simone Opel<br />
Didactics <strong>of</strong> Informatics<br />
<strong>University</strong> <strong>of</strong> Duisburg-Essen<br />
Schützenbahn 70<br />
45127 Essen, <strong>Germany</strong><br />
simone.opel@uni-due.de<br />
Vocational education in <strong>Germany</strong> is characterised by a learning<br />
venue cooperation between vocational schools and vocational<br />
training companies. For a better combination <strong>of</strong> theory<br />
and practice the curricula in the field <strong>of</strong> computer science<br />
(CS) and information and communication technologies (IT)<br />
are arranged in so-called “Lernfel<strong>der</strong>” (learning fields), which<br />
gives the teachers the leeway to develop different learning<br />
situations on their own. Unfortunately, it seems that teachers<br />
<strong>of</strong>ten do not put this idea into practice. The question<br />
is: Why would vocational IT school teachers willingly relinquish<br />
these benefits? The elicitation study described in this<br />
paper explores the IT teachers’ knowledge <strong>of</strong> and attitudes<br />
towards the concept <strong>of</strong> learning fields in or<strong>der</strong> to find an<br />
answer to that question. It is part <strong>of</strong> a project which aims<br />
to develop exemplary learning situations and helpful tools<br />
for several learning fields, which will make it easier for the<br />
teachers to create lessons in the context <strong>of</strong> learning fields.<br />
Categories and Subject Descriptors<br />
K.3.2 [Computers and Education]: Computer and Information<br />
Science Education—Computer science education,<br />
Curriculum, Information systems education<br />
General Terms<br />
Human Factors, Theory<br />
Keywords<br />
Vocational IT Education, Computer Science Education, Learning<br />
Fields, Learning Situations, Teachers’ Attitudes, Empirical<br />
Study, Elicitation Study<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that copies<br />
bear this notice and the full citation on the first page. To copy otherwise, to<br />
republish, to post on servers or to redistribute to lists, requires prior specific<br />
permission and/or a fee.<br />
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$15.00.<br />
161<br />
Torsten Brinda<br />
Didactics <strong>of</strong> Informatics<br />
<strong>University</strong> <strong>of</strong> Duisburg-Essen<br />
Schützenbahn 70<br />
45127 Essen, <strong>Germany</strong><br />
torsten.brinda@uni-due.de<br />
1. INTRODUCTION<br />
1.1 The Vocational School System in <strong>Germany</strong><br />
The German school system and especially the ways to<br />
pr<strong>of</strong>essional life differ from the systems <strong>of</strong> most other countries.<br />
Students can attend an upper secondary school, called<br />
“Gymnasium”, where they get a general qualification for university<br />
entrance. Another way is attending general or intermediate<br />
secondary schools, where students get general education<br />
and are prepared to take up training and education<br />
at a company and part time vocational schools (so-called<br />
“Duale Berufsausbildung” – dual vocational education and<br />
training). During this time students are employees <strong>of</strong> their<br />
companies. The students’ age is between 16 and 25 years<br />
and their previous skills are quite heteregenous. At the end<br />
<strong>of</strong> the training students receive a vocational certificate from<br />
the chamber <strong>of</strong> industry and commerce or the chamber <strong>of</strong><br />
trade. The heterogenity <strong>of</strong> the students implies different<br />
types <strong>of</strong> learners in each class, which need different types <strong>of</strong><br />
instruction to benefit from lessons [3]. For this reason the<br />
learning content in the curriculum is arranged in so-called<br />
“Lernfel<strong>der</strong>” (learning fields). Learning fields do not include<br />
specific aims to be reached or skills to be acquired, but they<br />
describe different competencies which students should gain.<br />
By alternating multidisciplinary theoretical and practical<br />
training, students should acquire the competencies to apply<br />
their skills in new pr<strong>of</strong>essional situations. The idea <strong>of</strong> cooperation<br />
between vocational schools and training companies<br />
and the integration <strong>of</strong> work and study spreads in vocationals<br />
systems <strong>of</strong> different countries [4] [6]. So it can be worthful<br />
to exchange the experience with these concepts in the field<br />
<strong>of</strong> IT.<br />
1.2 Significance <strong>of</strong> “Learning Fields” for<br />
IT/CS Education<br />
A learning field contains typical business and working areas,<br />
which are reflected and reconstructed pr<strong>of</strong>essional activities<br />
[2]. A learning field includes theoretical and pr<strong>of</strong>essional<br />
skills from different subjects as well als social and<br />
personal competencies. The concept was developed to better<br />
meet the requirements <strong>of</strong> all partners <strong>of</strong> the dual vocational<br />
education and training. The different learning fields<br />
are designed openly; it is up to the teachers to implement<br />
the learning fields into suitable learning situations. A learning<br />
situation is one didactically prepared working process.<br />
It contains theoretical knowledge, several working skills and<br />
different competencies to solve complex problems and can in-
clude difficulties from different subjects. All activity-oriented<br />
methods can be used. That concept may cost time in the<br />
beginning, but one part <strong>of</strong> the concept is to teach in teams,<br />
so teachers from different disciplines can share the workload<br />
and support each other. Concepts <strong>of</strong> interdisciplinary collaborative<br />
teaching and learning can be found in different scenarios<br />
<strong>of</strong> vocational training and education, e. g. in mechatronics<br />
[8] or in computer engineering education in Egypt<br />
[7]. In the area <strong>of</strong> IT the definition <strong>of</strong> the learning fields has<br />
not always succeeded. Most learning fields seem to be identical<br />
to the former subject [5]; it is tempting for IT teachers<br />
to work with the familiar subjects without thinking further<br />
about the specifics <strong>of</strong> the learning field. It appears that especially<br />
in IT-centered vocational schools the concept is hardly<br />
implemented by the teachers; we tried to find the problems<br />
teachers meet when translating the described competencies<br />
from several learning fields into adequate learning situations.<br />
The question was: why would vocational school teachers refrain<br />
from using the leeway given to them by the curriculum?<br />
Therefore we conducted an online survey with several vocational<br />
IT teachers 1 . In this paper we present the first part<br />
<strong>of</strong> the results <strong>of</strong> this elicitation study. It is part <strong>of</strong> a larger<br />
project with the purpose to convince vocational IT teachers<br />
to devise their classes according to the learning field concept<br />
by developing exemplary learning situations or helpful tools<br />
for several learning fields.<br />
2. METHODOLOGY<br />
We asked all vocational IT teachers in Bavaria (Fe<strong>der</strong>al<br />
State in <strong>Germany</strong>) to participate in the online survey. 28<br />
teachers answered (a response rate around 30 %). The online<br />
questionnaire consisted <strong>of</strong> three sections. First, all participants<br />
were asked for their age, sex, vocational discipline and<br />
their years <strong>of</strong> teaching (as range). The second section contained<br />
16 closed questions about the IT teachers’ knowledge<br />
<strong>of</strong> and attitudes towards the concept <strong>of</strong> learning fields. We<br />
used a 5-point scale answer format with options from “does<br />
not apply at all” (1) to “fully applies” (5). The third part<br />
consisted <strong>of</strong> open questions, which explored the teachers’<br />
attitudes toward the concept <strong>of</strong> learning fields following the<br />
theory <strong>of</strong> planned behaviour by Aizen and Fishbein [1] and<br />
the situation at different schools.<br />
3. RESULTS<br />
We got replies from 21 men and seven women with the<br />
average age <strong>of</strong> M = 47 years. The median <strong>of</strong> the years <strong>of</strong><br />
teaching is within the range <strong>of</strong> 11 to 15 years. 16 teachers<br />
are basically educated for teaching IT. The answers <strong>of</strong> the<br />
second part <strong>of</strong> the questionnaire were summarised following<br />
the theory <strong>of</strong> planned behaviour by Aizen and Fishbein [1].<br />
Only the first question about familiarity with the concept <strong>of</strong><br />
learning fields does not represent this theory, but shows an<br />
individual estimate <strong>of</strong> the personal skills on this topic. The<br />
participants reported a very high familiarity with the concept<br />
(M = 4.21). The results <strong>of</strong> the items about the teachers’s<br />
attitudes (M = 3.36), their self-effiacy (M = 3.21) and<br />
their subjective norms (M = 3.25) indicate that the teachers<br />
are principally open-minded to the concept <strong>of</strong> learning<br />
fields. Only the factor control belief (M = 2.83) shows a<br />
lower value, which could mean that the teachers see some<br />
difficulties to implement the concept. These results were<br />
1 IT teachers include also teachers for computer science<br />
162<br />
verified by a content analysis <strong>of</strong> the answers on the open<br />
questions. The responses to the first four questions were<br />
paraphrased and categorized [1]. We produced N = 151<br />
statements. The statements about the teachers’ attitudes<br />
(n = 47, 31.1 %) specified the information <strong>of</strong> the closed questions;<br />
the teachers described the advantages and difficulties<br />
they see in dealing with the concept. The statements about<br />
the teachers’ control belief (n = 102; 67.6 %) dealt with the<br />
school equipment, the strong heterogenity <strong>of</strong> the classes or<br />
the teaching staff composition. Only two statements about<br />
the subjective norm were found, which praise the colleagues’<br />
and the headmaste’s support. The examples <strong>of</strong> organization<br />
and learning situations are not analysed yet.<br />
4. DISCUSSION AND CONCLUSION<br />
The question <strong>of</strong> this study was: Why would vocational<br />
school teachers refrain from putting the concept <strong>of</strong> learning<br />
fields into practice? The results <strong>of</strong> this study show different<br />
aspects <strong>of</strong> the problems teachers have. The answers show<br />
that the teachers are motivated to implement learning fields<br />
into suitable learning situations, but it also seems that they<br />
need support to do that. In our opinion it seems to be<br />
helpful for the IT teachers to support them by developing<br />
guidelines with examples for learning situations and the related<br />
teaching material. It is also important to train the<br />
teachers on how to develop learning situations themselves,<br />
also in difficult environments, and how to evaluate criteria<br />
for appropriate learning situations.<br />
5. REFERENCES<br />
[1] M. Ajzen and I. Fishbein. Belief, attitude, intention,<br />
and behavior: An introduction to theory and research.<br />
Addison-Wesley, Reading, MA, 1975.<br />
[2] R. Ba<strong>der</strong>. Das Lernfeld-Konzept in den<br />
Rahmenlehrplänen (in German). Die berufsbildende<br />
Schule, 50(7–8):211–213, 1998.<br />
[3] A. Kluge, S. Ritzmann, D. Burkolter, and J. Sauer. The<br />
interaction <strong>of</strong> drill and practice and error training with<br />
individual differences. Cognition, Technology and Work,<br />
13(2):103–120, 2011.<br />
[4] C. Liang. The development research on higher<br />
vocational education curriculum based on the working<br />
process, volume 155 LNEE <strong>of</strong> Lecture Notes in Electrical<br />
Engineering. Springer, Berlin Heidelberg, 2012.<br />
[5] S. Opel. Lernfel<strong>der</strong> in <strong>der</strong> Praxis des IT-Unterrichts (in<br />
German). vlb-akzente Berufliche Bildung in Bayern,<br />
19(07):19–20, 2010.<br />
[6] Y. Ren and L. Zhao. Research and practice <strong>of</strong> ’teaching,<br />
learning, practice integration teaching model’ in higher<br />
vocational and technical education, volume 137 AISC <strong>of</strong><br />
Advances in Intelligent and S<strong>of</strong>t Computing. Springer,<br />
Berlin Heidelberg, 2012.<br />
[7] M. Salama and T. Thabet. A curricular reform<br />
proposal for egyptian computer engineering education<br />
ECEE. In 2010 IEEE Transforming Engineering<br />
Education: Creating Interdisciplinary Skills for<br />
Complex Global Environments, 2010.<br />
[8] S. Shooter and M. Mcneill. Interdisciplinary<br />
collaborative learning in mechatronics at bucknell<br />
university. Journal <strong>of</strong> Engineering Education,<br />
91(3):339–344, 2002.
ABSTRACT<br />
This study examines the potential for rethinking dated educational<br />
technologies. The mechanised Turtle Robot is taken as a test case<br />
to examine whether dated educational technologies can be renewed<br />
as a means <strong>of</strong> maximizing the tools and research <strong>of</strong> the<br />
past towards the new wave <strong>of</strong> interest in computing education.<br />
This paper will present research in progress and explore the Turtle<br />
Robot and other educational tools in the context <strong>of</strong> Technocamps,<br />
a Wales-based project aimed at inspiring young people 11-19<br />
years in computing.<br />
General Terms<br />
Experimentation, Human Factors, Languages<br />
Keywords<br />
Turtle Robot, Logo, Papert, Learning Tools, Programming, Electronics,<br />
Out-<strong>of</strong>-the-box, Teachers, Educators, Learners, Renew,<br />
Reuse, Recycle, Technocamps<br />
1. INTRODUCTION<br />
Keen not to contribute to the computing wasteland, this paper<br />
presents the findings <strong>of</strong> examining alternative approaches to<br />
learning from past educational tools. Using the Turtle Robot as a<br />
test case, the Turtle Robot is explored as a potential for contemporary<br />
educational engagement.<br />
A recent report reviewing the BBC’s Computer Literacy Programme<br />
examines the climate that led to the development <strong>of</strong> the<br />
BBC Micro [1] as a way <strong>of</strong> identifying the various influencing<br />
factors which bring about radical change in computing education.<br />
This study examines another seminal educational tool, the Turtle<br />
Robot, and specifically focuses on the mechanism <strong>of</strong> the robot as<br />
a potential contemporary teaching aid rather than the approach<br />
behind the robot.<br />
In exploring this outdated but very prevalent technology still<br />
available in classrooms today, the study examines whether the<br />
Permission to make digital or hard copies <strong>of</strong> all or part <strong>of</strong> this work for<br />
personal or classroom use is granted without fee provided that copies are<br />
not made or distributed for pr<strong>of</strong>it or commercial advantage and that<br />
copies bear this notice and the full citation on the first page. To copy<br />
otherwise, or republish, to post on servers or to redistribute to lists, requires<br />
prior specific permission and/or a fee.<br />
Conference’04, Month 1–2, 2004, City, State, Country.<br />
Copyright 2004 ACM 1-58113-000-0/00/0004…$5.00.<br />
Save Our Turtle Robots?<br />
Emma Posey<br />
Technocamps Computer Science<br />
College <strong>of</strong> Science Swansea <strong>University</strong><br />
Singleton Park, Swansea<br />
emma.posey@technocamps.com<br />
163<br />
robot mechanism has any relevancy, potentially through modification,<br />
as a learning tool.<br />
Comparisons are made with other robot simulators and their connection<br />
to real robots, especially through Technocamps, a project<br />
engaging young people in computing.<br />
2. TURTLE ROBOTS<br />
As part <strong>of</strong> the drive to encourage learning using computers, the<br />
Turtle Robot was developed using Logo, a computer language<br />
developed as a learning tool by Wally Feurzeig and Seymour<br />
Papert. Papert was a keen early advocate <strong>of</strong> learning via computers<br />
and he believed computers, and specifically Logo, could<br />
help young people plan, problem-solve, as well as develop critical<br />
thinking and logic.<br />
Papert initiated the Turtle Robots, a 3D robot in the form <strong>of</strong> a<br />
turtle. The turtle, Papert claimed, was useful as it was an object<br />
and therefore could be un<strong>der</strong>stood in real terms. Critically, Logo<br />
as a computer language, allows the user/programmer to introduce<br />
new words to the program – effectively developing a personalised<br />
vocabulary to define new procedures and commands. Papert<br />
identified the benefits <strong>of</strong> computational thinking [3] [4], a literacy<br />
which helps young people’s thinking to be ‘step-by-step, literal,<br />
mechanical’. An agreed definition for computation thinking and<br />
ways in which it can be taught continue to be a challenge [5].<br />
The ‘Turtle Graphics’ project [6] used in the educational tool<br />
Scratch [7] from MIT Media Lab and others such as RoboMind<br />
[8] use a simulated robot to engage young people in programming,<br />
one using a graphical drag and drop, the other syntax.<br />
AberBots, modified by Technocamps Aberystwyth, is a robot<br />
simulator whose program runs on real research robots, the Pioneer<br />
and IDRIS. The AberBots program is scripted and as with the<br />
Turtle Robot, there is a direct relationship between program, its<br />
simulation and a real robot. The study compares these various<br />
programs and other programmable toys and draws on findings to<br />
analyse their ability to engage young people in programming<br />
using simulated and real robots.<br />
With the ability these days to build and program one’s own creative<br />
machines using products such as LEGO/Logo and more recently<br />
Lego Mindstorms, the ready-made object <strong>of</strong> the Turtle<br />
Robot may seem too restrictive or prescriptive in today’s terms.
3. NEW WAVE EDUCATIONAL TOOLS<br />
There is a wave <strong>of</strong> interest and political will within the UK for<br />
‘real’ computing in education, (rather than the predominant ICT<br />
culture) [9] [10].<br />
The recent move to overhaul computer education has computational<br />
thinking at its heart [12]. There is a danger that ol<strong>der</strong> tools,<br />
resources and experience get swept aside for the new. Especially<br />
in computing, ol<strong>der</strong> technologies and learning tools become<br />
quickly disregarded and discarded. The computer wasteland is<br />
vast with an increasing rate <strong>of</strong> obsolescence in the industry.<br />
Yet there are Turtle Robots hibernating in schools, gathering dust<br />
whilst schools with ‘Green Flag’ initiatives and ‘Eco Gangs’ aim<br />
to apply the mantra <strong>of</strong> ‘Renew, Reuse, Recycle’ across all that<br />
they do. There are numerous sensors and motors available in discarded<br />
toys and computer hardware. In educational, environmental<br />
and resource terms the exploration <strong>of</strong> existing technology<br />
in schools makes real sense.<br />
Further, a key obstacle or ‘grand challenge’ in getting computing<br />
back in to schools is not the examining boards and curriculums,<br />
but the skill base <strong>of</strong> ICT teachers, some with little or no training<br />
in ICT let alone computing and many with little time or inclination.<br />
However, there are some teachers and educators who have had<br />
previous involvement with computing and they refer back to Logo<br />
and the Turtle Robot (as well as the BBC Micro) as a starting<br />
point in their own development. Those with some previous experience<br />
<strong>of</strong> Logo may be confronted with new programs to learn –<br />
revisiting a key reference point, the Turtle Robot, may be a useful<br />
starting point in their resumed computing in education journey.<br />
4. TECHNOCAMPS<br />
Technocamps is a European project aimed at inspiring 11 to 19<br />
year olds to engage with computer programming and electronics.<br />
The three-year project is based in Wales<br />
(www.technocamps.com). Technocamps has been exploring ways<br />
in which young people can be encouraged to think imaginatively<br />
about computing.<br />
At Aberystwyth Technocamps we are particularly keen on outside-the-box<br />
technology and most <strong>of</strong> the devised workshops<br />
closely follow Papert’s aim <strong>of</strong> encouraging young people to control<br />
and manipulate computers in the world rather than on the<br />
screen.<br />
164<br />
5. SAVE OUR TURTLE ROBOTS?<br />
The research project has explored the benefits <strong>of</strong> the ready-made<br />
Turtle Robot machine and whether working with existing equipment<br />
can save time and resources for schools. The study explores<br />
various ways to modify the design <strong>of</strong> the Turtle Robot so that is<br />
can be utilized as a relevant contemporary learning tool. It also<br />
investigates other, more recent, programming languages, environments<br />
and hardware with a view to unearthing any lasting<br />
benefits to rethinking this outdated technology.<br />
6. ACKNOWLEDGMENTS<br />
My thanks to Technocamps for allowing me to utilise the research<br />
and findings <strong>of</strong> the project through its modules and tracking with<br />
specific thanks to Dominic Roberts, Jonathan Roscoe, and Barry<br />
Thomas.<br />
7. REFERENCES<br />
1. Blyth. T., The Legacy <strong>of</strong> the BBC Micro, NESTA, May<br />
2012<br />
2. Papert. S., 1980. Mindstorms: Children, Computers,<br />
and Powerful Ideas, The Harvester Press<br />
3. Wing. J., Computational Thinking, Communication <strong>of</strong><br />
the ACM, Viewpoint, Vol. 49, No. 3, pp33-35, March<br />
2006<br />
4. Barr. V., and and Stephenson. C., Bridging Computational<br />
Thinking to K-12: What is Involved and What is<br />
the Role <strong>of</strong> the Computer Science Education Community?,<br />
ACM Inroads, Vol. 2(1), 2011, 48-54.<br />
5. diSessa. A., Changing Minds: computers, learning and<br />
literacy, Cambridge, MA: MIT Press, 2000.<br />
6. http://scratch.redware.com/project/turtle-graphics<br />
7. http://scratch.mit.edu<br />
8. http://www.technocamps.com/resources<br />
9. The Royal Society, Shut down or restart?, The way<br />
forward for computing in UK schools, January 2012<br />
10. Livingston. I., and Hope. A., Next Gen: Transforming<br />
the UK in to the World’s leading talent hub for the<br />
video games and visual effects industries, NESTA, February<br />
2011<br />
11. Computing: A curriculum for schools, Computing at<br />
Schools, Body <strong>of</strong> Knowledge Working Group, March<br />
2011