Maria Knobelsdorf, University of Dortmund, Germany

Fakultät für Mathematik, Informatik
und Naturwissenschaften
Maria Knobelsdorf, Technische Universität Dortmund, Germany
Ralf Romeike, University of Potsdam, Germany
(editors)
PRE-PROCEEDINGS
7 th Workshop in Primary and Secondary
Computing Education
WiPSCE 2012
November 7–9, 2012
Hamburg
Fachbereich Informatik

From the Conference Chairs
Welcome to Hamburg and the 7th Workshop in Primary and Secondary Computing Education
(WiPSCE) 2012.
You are looking at the workshop’s preproceedings that contain all papers and posters being
presented within the next two days. Because we give all authors the possibility to incorporate
into their manuscripts feedback and comments that they will receive during the workshop, the
overall final version of the proceedings will be officially published in the ACM digital library
after the workshop (approximately in December).
We hope you enjoy these contributions and their presentations during the workshop.
Maria Knobelsdorf and Ralf Romeike
WIPSCE 2012 Conference Chairs
2

Program Committee
Michal Armoni Weizmann Institute of Science, Israel
Tim Bell University of Canterbury, New Zealand
Yifat Ben-David Kolikant The Hebrew University of Jerusalem, Israel
Roger Boyle University of Leeds, UK
Torsten Brinda University of Duisburg-Essen, Germany
Michael E. Caspersen University of Aarhus, Denmark
Paul Curzon Queen Mary University of London, UK
Ira Diethelm University of Oldenburg, Germany
Judith Gal-Ezer The Open University of Israel, Israel
Mark Guzdial Georgia Institute of Technology, USA
Peter Hubwieser Technische Universität München, Germany
Maria Knobelsdorf (chair) Technische Universität Dortmund, Germany
Michael Kölling University of Kent, UK
Johannes Magenheim University of Paderborn, Germany
Orni Meerbaum-Salant The Weizmann Institute of Science, Israel
Ralf Romeike (chair) University of Potsdam, Germany
Ulrik Schroeder RWTH Aachen University, Germany
Carsten Schulte Freie Universität Berlin, Germany
Peer Stechert RBZ Technik Kiel, Germany
Chris Stephenson Computer Science Teachers Association (CSTA), USA
Leigh Ann Sudol Carnegie Mellon University, USA
Jan Vahrenhold University of Münster, Germany
Michael Weigend Fernuni Hagen, Germany
3

Table of Contents
On the Importance of Being Earnest: Challenges in Computer Science Education . . . . . . . . . 6
Jan Vahrenhold
Grand Challenges in Computing Education. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Judith Gal-Ezer
The introduction of Computer Science to NZ High Schools — an analysis of student work 8
Tim Bell, Heidi Newton, Peter Andreae and Anthony Robins
Is Self-Efficacy in Programming Decreasing with the Level of Programming Skills?. . . . . . . . 20
Michail N. Giannakos, Peter Hubwieser and Alexander Ruf
Concept of an Extracurricular Learning Environment for Computer Science . . . . . . . . . . . . . . 26
Nadine Bergner, Jan Holz and Ulrik Schroeder
The school experiment InTech - How to influence interest, self-concept of ability in
Informatics and vocational orientation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Claudia Hildebrandt and Ira Diethelm
Uncovering Structure behind Function the experiment as teaching method in computer
science education. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Carsten Schulte
Agile Projects in High School Computing Education - Emphasizing a Learners’
Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Ralf Romeike and Timo Göttel
Conceptual Change and Epistemological Belief Framework for Web Site Credibility
Instruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Marie Iding
How Teachers in Different Educational Systems Value Central Concepts of Computer
Science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Peter Hubwieser and Andreas Zendler
Ways of Planning Lessons on the Topic of Networks and the Internet . . . . . . . . . . . . . . . . . . . . . 75
Ana-Maria Mesaros and Ira Diethelm
Promoting Computational Thinking with Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Cynthia Selby
Preparing Teachers for Teaching Informatics: Theoretical Considerations and Practical
Implications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Vassilios Dagdilelis and Stelios Xinogalos
Grand challenges for the UK : Upskilling teachers to teach Computer Science within the
Secondary curriculum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Sue Sentance, Adam McNicol, Mark Dorling and Tom Crick
Articulating (Some) Grand Challenges of CSE at face of The Digital Age: A
Socio-Cultural Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Yifat Ben-David Kolikant
4

Challenge and Creativity: Students experiences of .NET Gadgeteer . . . . . . . . . . . . . . . . . . . . . . . 96
Sue Sentance and Scarlet Schwiderski-Grosche
eledSQL A New Web-Based Learning Environment for Teaching SQL at Secondary
School Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Andreas Grillenberger and Torsten Brinda
Bringing Contexts into the Classroom A Design-Based Approach . . . . . . . . . . . . . . . . . . . . . . . . 111
Detlef Rick, Marcel Morisse and Ingrid Schirmer
E-mail for You (only?) - Design and Implementation of a Context-based Learning
Process on Internetworking and Cryptography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Andreas Gramm, Malte Hornung and Helmut Witten
Comparing CSTA K-12 Computer Science Standards with Austrian Curricula . . . . . . . . . . . . 131
Daniel Egger, Sabrina M. Elsenbaumer and Peter Hubwieser
Information Theory on Czech Grammar Schools: First Findings . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Daniel Lessner
Data modeling and database systems as part of general education in CSE . . . . . . . . . . . . . . . . 143
Claudia Strödter
The Mindstorm Effect: A Gender Analysis on the Influence of Lego Mindstorms in
Computer Science Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Catherine Ball, Faron Moller and Reena Pau
An experience in exploring the processing of formatted texts by a kynesthetic approach. . . 149
Carlo Bellettini, Violetta Lonati, Dario Malchiodi, Mattia Monga, Anna Morpurgo
and Mauro Torelli
Teachers’ Perceptions of the Value of Research-Based School Lectures . . . . . . . . . . . . . . . . . . . . 151
Jonathan Black, Paul Curzon, Chrystie Myketiak, Laura R. Meagher and Peter W
McOwan
Technocamps: Dod â gwyddoniaeth Gyfrifiadurol ir gorllewin pell (Bringing Computer
Science to the far west) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Roger D. Boyle, Hannah M. Dee and Frederic Labrosse
Computational thinking in context of ”Abenteuer Informatik” . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Jens Gallenbacher
Gaming and Mathematics: A Cross Curricular Event . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Sharon Jones, Renada Poteat and Beth Frierson
Turi: Chatbot software for schools in the Turing Centenary year. . . . . . . . . . . . . . . . . . . . . . . . . . 159
Mathew J. Keegan, Hannah M. Dee and Roger D. Boyle
Learning Fields in Vocational IT Education - Why Teachers Refrain From Taking an
Opportunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Simone Opel and Torsten Brinda
Save our Turtle Robots? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Emma Posey
5

ABSTRACT
On the Importance of Being Earnest:
Challenges in Computer Science Education
Despite an increasing number of success reports from several
countries, establishing computer science as a subject
worth full curriculum credit is one of the most frequently
named goals in secondary computer science education. In
my keynote address, I will first present a personal perspective
on this issue and summarize challenges in research, recruitment,
and curriculum design that have to be met before
this goal can be reached in-breadth. Following up on the anticipated
success of our ambitions, I will then comment on
some probably even more pressing challenges in research, assessment,
and teacher training that we need to be prepared
to face once computer science has eventually been established
as a subject worth full curriculum credit.
Categories and Subject Descriptors
K.3.2 [Computers and Education]: Computer and Information
Science Education
Keywords
Secondary Computer Science Education
Acknowledgements
This work was partially funded by the Deutsche Telekom
Stiftung under the name dortMINT. Many of the thoughts
and observations discussed are based on work done together
with Holger Danielsiek, Rebecca Doherty, Arno Pasternak,
Wolfgang Paul, and Renate Thies.
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WiPSCE 2012 Hamburg, Germany
Copyright 2012 ACM X-XXXXX-XX-X/XX/XX ...$15.00.
Jan Vahrenhold
Westfälische Wilhelms-Universität Münster
Department of Computer Science
48149 Münster, Germany
jan.vahrenhold@uni-muenster.de
6

Challenges in Computer Science Education
ABSTRACT
Discussing challenges involves some knowledge of future trends.
Can one say anything about the future of computer science? Can
one predict the future of computing generally? In the September
issue of the CACM (Communication of the ACM) in his column
Peter Denning relates to this question and concludes by saying:
"Don't feel bad if you can't predict the future".
Thus, in my presentation I will not predict the future, but rather
describe the current situation of computer science education, and
the challenges we have to address at the moment, showing common
concerns related to computer science education we share
worldwide, and in particular Israel and the United States.
Categories and Subject Descriptors
K.3.2 [Computers and Education]: Computer Science Education
Keywords
Secondary computing education
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists, requires
prior specific permission and/or a fee.
Conference’10, Month 1–2, 2010, City, State, Country.
Copyright 2010 ACM 1-58113-000-0/00/0010…$10.00.
Judith Gal-Ezer
Computer Science Department
Open University of Israel
galezer@openu.ac.il
7

The introduction of Computer Science to NZ High Schools
— an analysis of student work
Tim Bell
Department of Computer
Science and Software
Engineering
University of Canterbury
Christchurch, New Zealand
tim.bell@canterbury.ac.nz
Heidi Newton
School of Engineering and
Computer Science
Victoria University of
Wellington
Wellington, New Zealand
heidi.newton@ecs.vuw.ac.nz
Peter Andreae
Anthony Robins
School of Engineering and Computer Science
Computer Science
Department
Victoria University of
University of Otago
Wellington
Wellington, New Zealand
Dunedin, New Zealand
anthony@cs.otago.ac.nz
peter.andreae@ecs.vuw.ac.nz
ABSTRACT
In 2011 New Zealand introduced computer science as a topic
that students could take as part of their studies in the last
three years of high school. The change was initiated in late
2008, so the new material was introduced with barely two
years of preparation and minimal teacher training. Despite
this tight timeline, many schools adopted the new topic,
and many students successfully completed assessment in it
in 2011. The format of the assessment was required to be
a report. In this paper we look carefully at the work that
students submitted by examining publicly available information
(statistics, markers’ comments and exemplars), and
performing a detailed analysis of a sample of 151 student
papers. We describe the nature of the assessment (which is
report-based with very flexible criteria for how students can
demonstrate their understanding), and examine the kind of
work that students submitted to meet the criteria, drawing
out good practices that enabled students to do well. A recurring
theme is the importance of students being able to use
personal authentic examples so that the examiner can hear
the “student’s voice” in their report work, which provides
evidence that the student has understood the topic rather
than paraphrased descriptions. The analysis also reveals the
value of prompting students effectively to get them engaged
properly with the concepts, and identifies successful ways
to achieve this in the three areas of the analysed standard
(algorithms, programming languages and usability).
8
Categories and Subject Descriptors
K.3.2 [Computer and Information Science Education]:
Computer Science education
General Terms
Measurement, Experimentation
Keywords
Computer Science education, report-based assessment, high
school, algorithms, programming languages, usability
1. INTRODUCTION
Over the period 2011 to 2013, new computer science standards
are being introduced to New Zealand (NZ) schools
that will give students the opportunity to explore topics
such as algorithms, human-computer interaction, cryptography,
AI, graphics, and other areas of computer science
that previously were normally encountered for the first time
in university courses. This change addresses a problem that
is common to many countries, where computing in schools
had become focussed on learning to use the computer as a
tool rather than supporting students interested in developing
new software and systems. For example, a January 2012
report from the UK Royal Society points out that “many
pupils are not inspired by what they are taught and gain
nothing beyond basic digital literacy skills such as how to use
a word-processor or a database” [8]. In the US, similar complaints
were being made; for example, Margolis and Goode
describe having to address the issue that “computer science
exists on the margins of many public school core requirements”
[10], and a 2009 report on the state of computing
in US high schools (surveyed in 2005 and 2007) points out
that “many schools and countries, policy-makers and administrators
are failing to provide students with access to the
key academic discipline of computer science”, and concludes
that “the results of these studies appear to indicate a continuing
and troubling decline in student interest in computer
science courses in high schools” [9].

These echo a 2008 report from the New Zealand Computer
Society that noted: “We suspect that there is a huge number
of potential computing professionals who have already
opted out of the discipline during secondary school, either
because of the lack of relevant achievement standards, or
because of the unpalatable offering of what they are told is
relevant for a future computing career” [11]. A report from a
group of NZ teachers also raised serious concerns: “Computing
is perceived to be second rate, is uncoordinated, under
resourced, unsupported, lacking in a professional body, and
in dire straits” [7]. Computer science has long existed as a
high school subject in Israel and South Korea [8], but the
above quotes illustrate how in the early 21st century serious
computing courses were essentially on the decline in the
US, UK and New Zealand. These concerns triggered a rapid
sequence of events beginning in 2008 that resulted in the
introduction of computer science as a subject in NZ schools
in 2011 [3, 4].
Part of the purpose of the new topics being introduced to
schools is to provide students with some grounding in CS
concepts, but the primary goal is giving them exposure to
the topic, providing the opportunity for them to find out
if it is something they might be passionate about, without
the confusion of it being presented as low-level digital literacy
skills. There are several related initiatives internationally
that have been started in recent years to provide
formal computer science courses in high schools, including:
(a) the “Georgia computes!” program “to increase interest
in computing at the pre-teen level, improve quality of
computing education at the high school level, draw students
into the undergraduate level, and make apparent to students
the opportunities for graduate study in computing” [6] (announced
in 2006); (b) the “Exploring Computer Science”
program in the Los Angeles Unified School District, which
aims to “make computer science knowledge more available
to and engaging for a broader segment of our student population”
[10] (started in 2008); (c) the US “Computer Science:
Principles”pilots to introduce CS as an Advanced Placement
course [2] (started in 2010); (d) the new German compulsory
standards introduced in some regions from 2008 [5]; and
(e) the 2012 report by the Royal Society in the UK advocating
sweeping changes to the ICT curriculum [8] that has
resulted in planned changes. The more general CSTA K-12
standards reflect a strong interest in changing how computing
is taught in schools 1 .
In the NZ high school context, computer science has been
introduced through three new “achievement standards” that
are typically taken in the last three years of high school.
The focus of this paper is the first standard, which was first
taught in January 2011 (the start of the school year in NZ),
and is described in Section 2. The paper looks in detail at
the actual work done by students under the new standard. A
sample of 151 student submissions were analysed in detail to
identify both good practice and issues such as student misconceptions.
From this we can learn better ways to deliver
computer science topics to students in this age group.
An important consideration for interpreting this analysis is
the context and the fast timeline with which the new ma-
1 http://csta.acm.org/Curriculum/sub/K12Standards.html
9
terial was introduced. The body of knowledge defining the
area was published in August 2009, just 16 months before
the standard was first taught. Between this point and the
beginning of 2011 the standard had to be written, teachers
prepared, and courses designed. This meant that the preparation
wasn’t as thorough as it could be, but it was seen
to be better to have a few schools adopt the new standards
quickly (they are not compulsory) than to delay for a year.
Effectively the first year was a pilot, although the scale of uptake
was remarkably high: in 2011 students from 49 schools
registered for the new computer science standard [4]. Adjusted
by population, this would be equivalent to 3,500 high
schools in the US, so the level of uptake would be a good
start on the 10,000 high school target of the CS10K project,
and was achieved with just two years lead time.
The formal training provided to teachers was minimal; some
had access to sessions at teacher events, and one-day workshops
were run around the country, but there was no formal
systematic training program and engagement was largely
driven by individual teachers’ motivation. A key role was
played by the newly-formed New Zealand Association of
Computing and Digital Technology Teachers (NZACDITT)
in sharing information and initiating events, and a network
of contacts was set up in the country’s seven universities that
teach computer science. A lot of the leadership (including
leaders of the NZACDITT) were in Christchurch, and these
people were impacted by the September 2010 and February
2011 Canterbury earthquakes, which occurred just at the
point that the standards were about to be taught, and had
a further impact on providing support for teachers making
the changes! An important component of teacher support in
the New Zealand system is exemplars of student work that
show what might be expected, and the computer science exemplar
didn’t become available until June 2011, well into
the year in which the new standard was being offered.
Despite the constraints on preparation time from introducing
the new standards so quickly, and adoption of the standards
being optional, the number of students attempting the
computer science standard in its first year of introduction
was still substantial (1,429 students in 49 schools registered
for the new CS standard [4], and 654 were submitted by the
deadline). To give a sense of scale, there are currently (in
2012) about 485 schools teaching at high school level in NZ,
although many of the schools are specialist or smaller rural
schools. In NZ there are about 62,500 students at “Year 11”
(the third to last year of high school, typically around 15–
16 years old), which is the year that most students would
attempt the new computer science standard. The first year
was essentially a pilot, but nevertheless had sufficient uptake
that it represents widespread national engagement.
In this paper we analyse the student submissions in detail
based on three sources: officially released statistics and commentary
about the grading, the published exemplars of student
work, and a detailed analysis of a sample of 151 student
submissions. Section 2 explains the new computer science
standard and how it is assessed; Section 3 reviews the publicly
available information about the submissions, Section 4
is a detailed analysis of the sample of student submissions,
and Section 5 reports some lessons learned based on the observations.

2. THE AS91074 STANDARD
During the final three years of high school in NZ students are
focused on the “National Certificate of Educational Achievement”(NCEA),
which involves completing a number of“standards”.
Details of how computing appears in these standards
are given in previous work [3, 4], and in this paper we focus
on one particular standard which was introduced to schools
in 2011, called“AS91074: Demonstrate understanding of basic
concepts from computer science” 2 . The standard is set
and regulated by the New Zealand Qualifications Authority
(NZQA).
The AS91074 standard is worth 3 credits, which translates
into approximately 30 hours of student work, so it can be
used as a relatively small component of a larger course, typically
combined with other standards on programming, web
design, digital media, information systems, electronics and
so on. It is a “level 1” standard, which means it is generally
used in the third to last year of high school; levels 2 and 3
apply to the final two years respectively.
The standard covers three main topics from computer science,
which can be summarised as:
algorithms: the difference between algorithms, programs,
and informal instructions; the kinds of steps that are
used to make an algorithm; and comparing the “cost”
of different algorithms for the same problem
programming languages: the kinds of programming languages
that exist (including high and low level languages)
and the roles of compilers and interpreters
usability: evaluating and comparing user interfaces in terms
of usability
These topics provide a broad exposure to ideas from computer
science, ranging from the mathematical analysis of
algorithms to the human-factors aspects of usability evaluation.
They are broad rather than deep — students don’t
have to implement algorithms, but they can run programs
that are provided; they don’t write compilers, but they need
to document how they use them; and they don’t design interfaces,
but instead learn to look at existing interfaces critically.
AS91074 is the first in in a series of computer science standards
(the level 2 and 3 standards are being phased in during
2012 and 2013) that together cover a wide range of topics
from the ACM curricula. The general area of “Programming
and Computer Science” includes two other standards at level
1, AS91075 (which is essentially about designing computer
programs) and AS91076 (implementing programs), with further
standards at levels 2 and 3. These are not analysed in
detail in this paper, but come up occasionally because most
students doing the CS standard are also attempting the programming
ones.
2 It is also known as “Digital Technologies 1.44”, which is
the number it had when in draft. AS91074 is available
publicly from http://www.nzqa.govt.nz/ncea/assessment/
search.do?query=Digital+Technologies&view=all .
10
AS91074 is an “external” standard, which means that it is
assessed at the national level; the majority of standards (including
the programming ones) are “internal” and assessed
at the school but moderated (i.e. audited) nationally. For
various reasons, AS91074 is assessed using a student report,
which means that the student submits a report on what is
essentially project work. The report is limited to 14 pages.
The “AS” in the title stands for “achievement standard”,
which means that it can be awarded with four levels of
achievement: Not achieved, Achieved, Merit and Excellence,
abbreviated as N/A/M/E respectively. Thus unlike many
other grading systems, “A” means a minimal pass, and “E”
is the best result; “N” means that the student gets no credit.
The standard itself is very brief — just over 2 pages, almost
half of which is boilerplate information such as review dates
and standard text. The main content of the standard is the
criteria for the A, M, and E levels of achievement, which are
given in just a few “bullet points” — 4, 5, and 4 points for
A, M, and E respectively. The bullet points for M and E
generally describe the higher standard of work required for
those levels rather than additional content. For example,
one of the Achieved bullet points is:
• describing an algorithm for a task, showing understanding
of the kinds of steps that can be in an algorithm,
and determining the cost of an algorithm for
a problem of a particular size.
The equivalent point at Merit is:
• showing understanding of the way steps in an algorithm
for a task can be combined in sequential, conditional,
and iterative structures and determining the
cost of an iterative algorithm for a problem of size n.
And the one at Excellence is:
• determining and comparing the costs of two different
iterative algorithms for the same problem of size n.
Each of the three criteria require that the student shows that
they can determine the “cost” of an algorithm (typically the
running time or number of steps/comparisons); if they simply
report the time from a single experiment then they have
reached the Achieved level; if they compare different values
of n (i.e. show the shape of the cost function) that would
reach the Merit level; and if they compare two algorithms
(typically a graph with two curves) then that would be Excellence.
Of course, the markers will be looking for more
than just numbers: the student needs to present the data
well and discuss the trends that it shows.
The brevity of the standard gives great flexibility to teachers
and students. In the algorithms example above, they can
explore whatever algorithms they choose as long as the chosen
algorithms can be used to demonstrate that the student
has met the criteria. Of course, this flexibility also puts a
lot of responsibility on teachers and students to choose good

examples, and in practice teachers share ideas and subject
experts publish advice to guide these choices.
Marking of the student submission is done “holistically”,
which means that if a student does very well in most areas
but has problems in just one point, they won’t get a
low grade just because they didn’t meet one basic criterion.
For example, in general a student will have met all 4 of
the Achieved bullet point criteria to be given the Achieved
grade. It may be that their work was weak in one of the
criteria, but if they had done particularly well in another
(perhaps at Merit or Excellence level) then they would be
given achieved overall.
The use of a report for assessment is challenging. To“demonstrate
understanding of basic concepts from computer science”,
a student report cannot just paraphrase definitions
or on-line articles about the topic. For example, a report
that describes selection sort and states that its cost is O(n 2 )
does not constitute a demonstration of understanding, since
this could be simply parroting text from a teacher or a paraphrase
of one of many descriptions available online. Similarly,
just listing the properties of a good user interface could
be paraphrased from a textbook, with no understanding. Instead,
students need to report on a personalised activity or
investigations in order to demonstrate that they have acquired
their own understanding of the concepts.
There are many ways to personalise the reports to meet
this expectation. For example, timing an implementation of
selection sort on their own computer using their own input
data reflects a personal experience with the issue of
quadratic behaviour as n increases. Given the number of
possible permutations of even a small sequence of values,
it is easy for students to example traces data that will be
unique to themselves. For the programming language topic,
a demonstration of a program written in some language,
and compiling and running it can be done use the local environment
in the student’s own account, using a program
they have written themselves (perhaps for the programming
standard, or even a simple modification of a “Hello World”
program that uses their name). For the HCI topic, a student
can apply the list of principles to the interface of a device
that they personally own, such as an alarm clock or an mp3
player.
3. STUDENT SUBMISSIONS
There is considerable public information available about the
student work for the AS91074 standard, which we have collated
and analysed in this section. The three areas we look
at are the student grade statistics from 2011, the published
exemplars, and the markers’ comments on the work.
3.1 Student grade statistics
The New Zealand Qualifications Authority publishes statistics
on student performance in each Achievement Standard 3 ,
and in Table 1 we show the number of candidates for the
three Programming and Computer Science standards, along
with the success rates of students (Not achieved, Achieved,
Merit and Excellence).
3 http://www.nzqa.govt.nz/studying-in-new-zealand/
secondary-school-and-ncea/secondary-school-statistics/
11
In 2011, the first year that the AS91074 achievement standard
was available, 654 student reports were submitted.
The programming standards were significantly more popular,
with nearly five times as many students taking those
standards. This isn’t surprising because the topic is better
understood, and is one that many teachers are familiar with
and would have been more confident to offer it.
To put the participation rates in perspective, there were
about 62,500 Year 11 students in school in NZ in 2011 4 ,
so participation in the new standards was relatively high
considering that at least 75% of schools didn’t offer them [4].
The student grade distribution for AS91074 is typical of
achievement standards, with relatively few students reaching
the Excellence level, but the majority getting credit for the
standard (Achieved or better). The programming standards
have a higher proportions of Excellence grades — in fact, the
results show similar numbers of students getting Excellence
and Merit, and reflects the bimodal distribution of performance
that has commonly been observed for programming
courses [13]. We also note that there is a difference in grades
between male and female students; fewer females attained
the Excellence grades, with the difference being particularly
noticeable for the programming standards.
It would appear that offering these standards early in high
school has attracted a higher proportion of female students
than advanced courses typically do. The proportion of female
students taking the computer science standard (AS91074)
was 27%, and the related programming standards (AS91075
and AS91076) had 30% and 36% respectively. While low,
this fraction, which is for students taking the standards
in their third to last year before university, is significantly
higher than the proportions who enrol in university computer
science courses — universities in NZ typically have
about 10 to 20% female students in first-year computer science
classes — so there is some hope that offering computer
science earlier in the education system may reduce the gender
imbalance.
3.2 Public exemplars
For privacy reasons our main analysis of student work in Section
4 is limited to broad statistics and observations about
the sample reports. However, there are some public samples
of student work: after the 2011 reports were marked, a
set of “exemplar” reports was selected (two or three at each
level of A/M/E) for publication 5 . The reader is encouraged
to view these to get an idea of the kind of work students
are submitting, and we will use some examples from these
in our discussions. The exemplars include annotations from
the markers to show how the criteria have been met.
The exemplars are provided to guide teachers on the standard
of work that is expected, but because they are public,
it has to be assumed that students will have access to them.
This has both benefits and problems; modelling good practice
is clearly helpful pedagogically, but it also provides the
4 http://www.nzqa.govt.nz/assets/About-us/Publications/
ncea-annualreport-2011.pdf (page 13)
5 These are available from http://www.nzqa.govt.nz/
assets/qualifications-and-standards/qualifications/ncea/
NCEA-subject-resources/Technology/91074-exp-2011.zip.

Table 1: Student success rates in Programming and Computer Science standards 2011
Candidates Gender N/A A M E
AS91074 Demonstrate understanding of basic
concepts from computer science
654 32.7% 37.2% 18.2% 11.9%
Male 475 72.6% 33.3% 36.0% 18.3% 12.4%
Female 179 27.4% 31.3% 40.2% 17.9% 10.6%
AS91075 Construct an algorithmic structure
for a basic task
2131 34.3% 31.5% 16.1% 18.1%
Male 1493 70.1% 34.3% 31.3% 15.0% 19.4%
Female 638 29.9% 34.3% 31.8% 18.7% 15.2%
AS91076 Construct a basic computer program
for a specified task
3245 31.8% 33.2% 18.0% 17.1%
Male 2082 64.2% 29.9% 33.4% 17.6% 19.2%
Female 1163 35.8% 35.3% 32.8% 18.7% 13.3%
temptation for students to simply paraphrase the model answers
in attempt to imitate the best practice, rather than
generate their own authentic report from scratch.
The first Excellence exemplar covers the criteria for the
standard, although it isn’t perfect. It draws a sound conclusion
about binary vs linear search, including analysing
the time needed for linear and binary search of 100 items;
linear search was estimated as “1-100” comparisons; binary
search is estimated at 7, calculated by repeatedly dividing
the number in half (i.e calculating log2100 without using logarithms).
The material about programming languages compares
Pascal and Scratch with binary code, which provide
a good contrast. The HCI report looks critically at some
aspects of two mobile touch-screen devices; it is a little adhoc,
but does provide a balanced view based on personal
experience. This report illustrates the holistic marking and
the focus on the criteria: despite being rated an Excellent
report, it does contain some confused reasoning; for example
the report claims “A binary search uses the yes/no, on/off,
0/1 concept” (perhaps confusing binary search with binary
numbers); and tells us that “[Pascal] cannot be used with a
mouse, and can only be controlled by the keyboard”. These
are reminders that students are encountering the terminology
and environments for the first time and are still making
sense of them.
The second Excellence example is longer and more carefully
presented. It compares two sorting algorithms using the
a balance scale activity from CS Unplugged (csunplugged.
org/sorting-algorithms), and includes notes and photos of
the activities (Figure 1) that provide good evidence that
the student has engaged in the process. The number of
comparisons used for n = 5, 10, 15 and 20 are given for a
single manual simulation of selection sort and quicksort using
a balance scale. The report provides a graph without
axes (Figure 2), which shows the difference between the two
algorithms, and a good conclusion is drawn, that the “lower
cost [of quicksort. . . ] will become more noticeable as the
number of items compared increases.” This is an articulation
of the difference between n 2 and n log n time; it does
have a slight imprecision in that the word“compared”should
be “sorted,” but it is reasonable evidence that the student
understands the relationship, which is the goal of the standard.
The HCI work in this exemplar mentions a visit to a
university Usability Laboratory, and this would have been
an excellent opportunity to see HCI evaluation happening
in an authentic context and no doubt contributed to the
12
Figure 1: Notes and photo of experiment using balance
scale activity from an Excellence exemplar
Figure 2: Graph from an Excellence exemplar
student’s appreciation of the topic.
At the other end of the scale, the Achieved exemplars show
work from students who have engaged with the topic, but
haven’t made in-depth observations. For example, the first
Achieved exemplar discusses bubble sort, and gives a trace
of the algorithm being applied to 8 names (Figure 3). At the
end of this example the writer gives the number of comparisons
and swaps for that example, but makes no comment
comparing this with different values of n, or with different
algorithms. The second Achieved exemplar uses a different
approach, showing a screenshot from an on-line visualisation
of bubble sort (Figure 4), and using the statistics from the
visualisation to determine the cost of the algorithm. In fact,
the exemplar goes on to compare bubble sort and selection
sort for different values on n, but there is no explanation of
how these were measured, and more significantly, the other
sections on programming languages and usability are weak,
and hence the overall grade is Achieved.
Exemplars are also available for work meeting the Merit criteria.
These show more advanced understanding and discussion
than the Achieved exemplars, but don’t have the
completeness of the Excellence ones.
Doing well in the report will be easier for students who are
good at writing, and because there are also elements of math

Figure 3: A trace of bubble sort from an Achieved
exemplar
Figure 4: Using a visualisation to demonstrate an
algorithm from an Achieved exemplar
and scientific experimentation, students with strengths in
science and math will also find it easier to do well. We consider
this to be a positive, since writing, math and general
science skills are valuable for computer scientists, but this
emphasis represents a significant change from the computing
classes before 2011 which were generally focussed more
on using computers and tended not to attract students with
good science and math skills.
It’s also important to note that these students are working
on topics that have traditionally been taught to university
students, and one has to avoid the trap of expecting them
to perform at the same level as more experienced students.
Inevitably the average level of experience in writing, science
and math will be lower, and it is important to calibrate
expectations appropriately.
3.3 Markers’ assessment report
After the grading of the reports was complete (January 2012),
the material was returned to the students and a report from
the markers was released 6 .
The markers’ report began by commenting on the quality of
the presentation. For example, the report notes that “small
screen shots, text too small to read, and graphs with unlabeled
axes did not contribute to a demonstration of understanding.”
Learning to conform to guidelines is good practice
for future CS practitioners, and being in the habit of presenting
data clearly is valuable! This is a benefit of requiring
a report from students rather than examinations, and helps
6 http://www.nzqa.govt.nz/nqfdocs/ncea-resource/reports/
2011/level1/technology.pdf
13
to emphasise the importance of communication skills. It
also reflects the value of students being aware of good experimental
method (particularly for reporting on algorithm
performance). Those who had a weaker science background
might not be familiar with the process of collecting, reporting
and discussing results in a precise manner.
Another issue raised by markers was the personalisation of
the work — the markers said that “candidates who clearly
demonstrated understanding of basic concepts from computer
science wrote in their own voice, providing evidence
from their own work and experience to support any factual
or referenced material.” The markers also pointed out that
reports that were simply paraphrasing other sources didn’t
make it clear that the student had understood the topic.
The public availability of exemplars will make it even more
important for the student’s “voice” to be heard i.e. to make
sure the report is from personal experience on examples that
are likely to be unique to the student.
Regarding the human-computer interaction work, the markers
observed the common problem of confusing “functionality
of devices with usability.” Students very quickly focused
on the physical form of an interface (buttons and menus),
or the functionality of the software, without identifying the
human factors that could make it difficult for a user to find
their way around the interface.
The markers’ report concludes with a summary of what was
required at each level, which is essentially the criteria of
the standard, although it includes typical problems found
in “Not Achieved” work, which (apart from not meeting the
criteria) were a lack of detail in descriptions, paraphrasing
without understanding, and not addressing all the requirements
when using a template.
The marking process is challenging because the standards
are very open-ended, and work can be submitted based on
a wide range of contexts. The standards are also new and
there is no past history to draw on. We note that markers
need to be well briefed, and provided with detailed guidelines.
In particular they need to be familiar with relevant
material from popular sources on the web (starting with but
not limited to Wikipedia) so as to recognise material that
has been used without understanding. This is particularly
important in some sub-areas (like programming languages).
4. ANALYSIS OF STUDENT PAPERS
To understand the student work better we manually analysed
in detail 151 sample reports from the 654 AS91074
submissions in 2011, noting the choice of examples that each
student used and how they presented their understanding of
the topic. The sample was selected by NZQA, who are responsible
for grading the student work. All of the sampled
papers were marked by one marker; the marking process
has checks in place to ensure consistency between markers
so this should not introduce a significant bias. The selection
process could not be rigorously randomised, but because
student submissions are essentially shuffled to prevent one
marker getting all the reports for one school, the sample is
reasonably broad. Table 2 shows the grades of the sampled
reports compared with those of all submissions.

Table 2: Student success rates in AS91074 2011 (all
submissions and sample analysed)
Source Number N/A A M E
All submissions 654 32.7% 37.2% 18.2% 11.9%
Sample analysed 151 36.4% 35.1% 18.5% 9.9%
Table 3: Number of pages in student submissions
for AS91074 in 2011 (sample of 151 submissions)
All Not Achv Achieved Merit Excellence
Submitted pages
Average 6.5 4.6 6.4 8.4 10.9
Min 1 1 2 4 8
Max 16 13 14 13 16
Student-written content
Average 5.4 3.4 5.2 7.3 9.8
Min 0.5 0.5 1.5 3.5 6.5
Max 16 12.5 11.5 12 16
We discuss the student work here first in terms of the length
of the document, then under the three main topics of the
standard: algorithms, programming languages and usability.
We conclude by discussing general issues with the writing
style, and the effect of teacher advice on the students’ work.
As would have been the case for grading the student work,
some of our analysis of the reports involved judgement calls
on what a student’s intention was, including dealing with
inaccurate descriptions and confusing text, so the figures
reported below are approximate, but indicative of the kind
of work that students submitted.
4.1 Report page length
The reports are limited to 14 pages (and markers will not
mark any additional pages). An important question is whether
this limit constrains the students. The number of pages submitted
by students in the sample of 151 analysed documents
is shown in Table 3. The first set of figures shows the number
of pages the students submitted, and the second set shows
our estimate of actual student-written content, which ignores
the teacher-generated rubrics (which is the most common
“padding” in the submissions), blank areas, tables of
contents, and reference lists (which were rare).
Only 4 of the 151 sampled submissions used the full 14 pages
(in fact, one of those 4 used more than 14 pages, but the
extra pages would have been ignored by the markers). Most
students did not seem to be unduly constrained by the limit,
and 91% of the submissions used 11 pages or fewer.
The average number of pages grows with the quality of the
work, which isn’t surprising since more information is needed
to cover all the criteria for Excellence. However, having a
long document doesn’t necessarily predict a good grade; one
13-page submission did not even reach the Achieved level
because most of the space was taken up with a lot of detail
of algorithm execution without discussing the bigger picture
such as the performance of the algorithm. One student
was awarded Excellence for just 6.5 pages of writing. In
this case the student didn’t use any images, but had clear
explanations and examples which showed understanding of
the topic; for example, binary and linear search were compared
for large values of n, providing a convincing contrast.
14
Two of the longest Excellence submissions used a lot of images,
including several screen shots of multiple interfaces.
The screen shots were relevant, although not essential, and
the work could have been presented in slightly less space.
Another 14-page Excellence submission was primarily text,
and contained considerably more detail and covered a wider
range of concepts than would be necessary to meet the Excellence
criteria.
Some of the longer submissions had extensive traces of algorithms
that went over multiple pages. Often these were
presented inefficiently — for example, one trace of a sorting
program listed every comparison made, one per line, which
was very difficult to read, and took four pages to describe
sorting eleven numbers. Other students managed to convey
similar traces in a fraction of the space by using better layout
which more clearly showed the big picture of what was
happening as well as making each step clear. Since layout
isn’t one of the criteria, both approaches should be accepted,
but teachers advising students on submission lengths would
need to take into account how compact a student’s representations
are.
Based on our observations, the limit of 14 pages seems suitable
as an absolute maximum, but students could be advised
that about 10 pages of their own writing is usually sufficient
to cover the Excellence criteria, and longer documents would
be required only if they rely heavily on images in their descriptions.
It is important that students focus on the criteria
of the standard and cover all the requirements.
Not all students addressed all three areas. Two students
did not attempt the algorithms topic; seven made submissions
that didn’t touch on programming languages, and nine
didn’t tackle usability. It would appear to indicate that these
few students simply ran out of time to complete all the criteria,
but submitted their work anyway (none of them received
a passing grade since there were criteria that were clearly not
met.) Most of these gave the impression of a hasty attempt,
but one student did a good job of two of the topics, and
would have done well if they had submitted something on
usability.
4.2 Algorithms
For the algorithms topic, the main criteria were to differentiate
an algorithm from a program and informal instructions,
and to discuss the “cost” of an algorithm. Almost any example
could be used to discuss the first criterion, but measuring
the cost required an algorithm that had interesting behavior
based on its input size n. This criterion requires a good
definition for the cost; generally it would be the running
time of a program or the number of steps taken by the algorithm.
In some cases memory requirements or disk accesses
might be considered as a cost, although this wasn’t used
in any of the reports analysed here. Unfortunately some
students misunderstood the cost to be the length of the program
or algorithm, which painted them into a corner since
this couldn’t be measured based on the size of the input, n.
About 63% of the students used sorting algorithms as their
example, and 19% used searching. Many students met the
Achieved criteria or better for the algorithms part using
these examples. About 9% of the students appeared to use

either their own program or another algorithm as an example,
and these were generally not very convincing for meeting
the criteria (most of them received “Not achieved” for
the standard). An example used by one student that was
(just) suitable for meeting the basic criteria was the Towers
of Hanoi solution, although it wasn’t feasible to compare
two different algorithms with different costs with this example,
and hence couldn’t be used for the Excellence criterion.
At least five students attempted to discuss the cost of an
algorithm without using an example at all — none of their
submissions met the criteria.
The most popular sorting algorithms were bubble sort and
quicksort (each was used in about 31% of the submissions).
The high use of bubble sort is likely to reflect its popularity
with teachers more than its pedagogical value [1]. Quicksort
provides a strongly contrasting example to compare with
quadratic time sorting algorithms. The other popular sorting
algorithms were selection sort (used in about 22% of
submissions) and insertion sort (20%).
To meet the Excellence criteria, students needed to choose
two algorithms to compare, and this was attempted by about
36% of the students. About 26% of them compared the
performance of two different sorting algorithms, with about
16% comparing an O(n log n) algorithm with an O(n 2 ) one,
giving the opportunity to observe a strong contrast in performance.
About 10% compared two O(n 2 ) sorting algorithms
(mainly two of bubble sort, selection sort and insertion sort),
which didn’t illustrate the contrasting performance differences
that can occur between different algorithms for the
same problem, although there were constant factor differences
observed. About 11% of the students used two different
searching algorithms for the comparison — all of these
compared linear search with binary search, which provide a
strong contrast in cost. One Excellence student also used
hashing (based on the CS Unplugged searching activity),
which led to a worthwhile discussion.
The values of n that were chosen for comparing algorithms
were often too small to make good observations. It wasn’t
unusual for student to report performance for searching or
sorting up to just 10 or 20 items, but it is at 100 or even 1000
items that some algorithms start to show compelling performance
differences. Very few students explicitly observed
that the relationship between two algorithms was non-linear,
although a few of the students who evaluated searching algorithms
did recognise the significant difference between their
measurements which reflected O(n) compared with O(log n)
time — these were mainly students who scored Excellence
overall. Note that students don’t have to implement binary
search (which is notoriously hard to get right [12]), but can
run a provided program that performs the search and counts
comparisons, or use a visualisation (several are available online)
that reports the time or number of comparisons made.
Similarly, they don’t have to do a formal analysis of the cost,
but could either plot data points and talk about trends, or
reason about it based on repeatedly halving n.
In general, students took a simplistic approach to analysing
their algorithms, and they need to be made aware of the
idea that algorithmic complexity can be non-linear, which
is a crucial factor surrounding the scalability of algorithms
15
(for example, a quadratic sorting algorithm given 10 times
as much input takes about 100 times as long to run). Some
weren’t able to observe the difference because the algorithms
they chose had the same complexity, and thus only showed
a small constant difference; others didn’t notice significant
differences because they only used small values of n. Only
three of the submissions analysed used a graph to show the
algorithm performance; others used tables, or simply presented
results in an uncollated fashion. The lack of graphs
would inevitably make it hard to see trends, although some
students were able to describe them from tables of figures.
Since the significant differences are likely to be unexpected,
students need to be guided to look for them, for example,
by encouraging them to experiment with very large values
of n, and to chart the results through a spreadsheet, which
makes it easier to deal with a larger number of observations.
The choice of algorithms can clearly make a difference to
the richness of the student report. The most effective choices
were the pairs of algorithms with significantly different asymptotic
complexities (binary search vs linear search, and quicksort
vs insertion/selection/bubble sort). Both the binary/linear
search and quicksort/quadratic sort pairs have costs that differ
by a factor of n/ log n, which provides a obvious contrast
for large values of n.
Students don’t need to implement the algorithms to meet the
criteria, and in fact the techniques needed (e.g. arrays/lists)
are beyond the requirements of the programming standards
at this level. For students who have more programming experience,
the simpler algorithms (linear search and quadratic
sorts) might serve as good programming exercises but the
difficulty is that a student must get the programming exercise
completely correct in order to be able to use them
for valid experiments (for example, one student gave results
that showed Bubble sort to be 73 times faster than Insertion
sort, which almost certainly reflects an issue with the
implementation given that Bubble sort is usually slower than
Insertion Sort). While quicksort is harder to implement, a
list-based version of it (where the list is duplicated at each
partition) is relatively easy to implement, and would still
demonstrate the performance improvement required. Binary
search is also easy to understand but difficult to get
exactly right [12]. Interestingly, some students explained in
their reports about why binary search is so hard to implement;
it is good to see them sensitised to issues surrounding
program correctness. In general we would recommend that
students experiment with programs they are provided with
(or visualisations), but interested students could be encouraged
to implement their own version as an extra exercise to
deepen their understanding, but not use those versions to
measure timing.
The use of bubble sort tended to be a distraction, as it is
very closely related to other quadratic sorting algorithms,
involves more steps for students to trace, and has a tradition
in algorithm pedagogy that paradoxically has made one of
the worst sorting algorithms one of the most well known [1].
As a contrasting algorithm to quicksort, selection sort is
useful because its performance is easily analysed using high
school math, and insertion sort is useful because it has a
contrasting best/average/worst case.

One complication with the algorithms topics is that a closely
related standard (AS91075) on designing a computer program
used the term “algorithmic structure” to describe the
design of the program. This appears to have been confused
with the use of “algorithm” in the AS91074 standard; some
online postings from teachers indicate that they regarded
them as describing the same thing, and at least one student
stated that an algorithm is a program plan, which is essentially
correct, but confusing in this context. Thus some
submissions on algorithm “cost” are based on a student’s
own design of a program, which typically doesn’t have any
interesting behaviour to analyse, and in fact may be O(1)
since the program does a very routine task. For example,
several programs simply read in a value, did a calculation
and printed the result. The use of the word “algorithm” has
since been removed from the AS91075 standard to avoid this
confusion.
4.3 Programming languages
The criteria for programming languages centre around the
role of programming languages, high and low level languages,
and how high level languages are translated to low level ones
(i.e. compiling and interpreting).
Most of the students described high and low level languages
and compilers abstractly, but at most a third attempted
some sort of concrete demonstration of a compiler being
used. A demonstration is much more convincing of student
understanding because the student has personally had the
experience of converting a program to a low level language
and running it. Even better than demonstrating the process
on a given program, students could have used their own
program (for example, one done for AS91076) as the example,
showing how it is compiled and/or interpreted. Because
the standard discusses both interpreted and compiled programs,
it’s unlikely that students would be familiar with
two different languages and thus only one of the examples
could reasonably be personalised, although some languages
have both compiled and interpreted versions (e.g. Basic)
and these could be contrasted. Another approach would be
for students to learn the bare minimum of a language e.g.
write a “Hello world” level program but substituting their
own name. None of the submissions seemed to have taken
this approach, and it would be interesting to evaluate, since
it forces students through the process of getting the program
to run. Using the student’s own program was risky for the
algorithms topic, but for the programming languages topic it
provides good personalisation and shows an awareness of the
process, and it isn’t sensitive to the quality of the program
(as long as it runs or compiles). Despite this, we observed
few examples in our sample where students had obviously
used their own program.
In the student work analysed, the main languages mentioned
as examples were Visual Basic/Basic (in about 18%
of reports), Scratch (13%), Pascal (5%), Flash/Actionscript
(4%), C (4%) and Alice (3%). Other languages used included
C++, Java, JavaScript, Picaxe, and Python. About
a half the reports didn’t use an example in a specific language,
and these generally didn’t reflect a lot of understanding
since the concepts were explained in very abstract terms.
Projects that compared a compiled and interpreted language
16
mainly used Visual Basic, C, Pascal and Java as examples
of compiled languages, and Scratch, VBA in Excel, Alice,
JavaScript and Python as examples of interpreted languages.
The situation isn’t always simple, as modern languages can
blur the distinction of compiling or interpreting (e.g. Java
compiles to byte code, which is then interpreted). One student
mentioned how Scratch can run as a Java applet itself,
so it is interpreted twice and compiled! If students can fully
understand these complexities then they will have a good
grasp on the issues intended, but it will be important for
teachers to provide guidance on the languages to explore.
Another example used in several reports was Visual Basic
(compiled) compared with Basic (VBA) macros in Excel (interpreted).
The contrast is effective because it is essentially
the same language and focuses the student on the different
way it is run rather than the difference in language; the examples
we observed using this contrast all appeared to meet
the criteria for Excellence in the programming languages requirements.
We would note that JavaScript is a useful language
to explore because it is definitely interpreted, readily
available, and is not confused by the drag-and-drop aspect
of Scratch, whereas it is problematic to use HTML as an example
of an interpreted language (as one project did) since
it is not at all clear that HTML is usefully viewed as a programming
language, even if it is a formal markup language.
Some students presented their work on programming language
as posters. This approach has potential since the
student is writing for a particular audience (probably other
students), and thus might be expected to focus the explanations
and examples more. It also seemed to discourage
paraphrasing, perhaps because they were thinking of it as
communicating to peers rather than trying to write something
that a teacher would recognise as correct.
4.4 Usability
The criteria in the standard emphasise factors of a user interface
that contribute to its usability. Unfortunately many
reports seemed to confuse usability with functionality, or focussed
on the physical layout of the interface. For example,
a student might comment that a cellphone can take photos
(the functionality), but miss the point that the typical task
that a user has (such as taking a photo and transferring it
to an online photo album) might involve pressing mysterious
key sequences, dealing with incompatible devices, and using
a time-consuming procedure that is difficult to remember
(usability). The better work focussed on typical tasks done
by the user, and any problems encountered completing those
tasks.
Using a personally owned device and talking about the student’s
own experience gives a clearer demonstration of understanding
than writing about interfaces abstractly. About
61% of the reports provided a suitably personalised report.
Those that didn’t discuss a personal device mainly gave general
usability principles that appeared to be paraphrased
from standard sources. In these cases, where illustrations
were given they tended to be standard examples of usability
issues, and once again it is hard to hear the “student voice”
when this approach is used.
The students who reported on experience with their own dig-

ital system used a variety of interfaces as examples: about
29% of the reports evaluated some sort of mobile phone (simple
phones, smart phones and iPhone related devices) or
some feature of one (such as sending a text message or taking
a photo). The main other commonly chosen devices were
an iPod/Touch device, used in about 7% of the reports, and
a video game controller (5%). Other interfaces appeared less
frequently, including a calculator, CD player, DVD player,
DVD recorder, email software, games, web browsers, web
sites, an mp3 player and a tape player. Even a toaster was
used to provide a reasonable evaluation.
Ten students reported on an interface that they had written
themselves. These generally weren’t good as examples:
few of them appeared to meet the criteria, and the remainder
were focussed on describing their interface without discussing
it from the user’s point of view. It would appear that
these students had tried to combine this standard with work
done for the design and programming standards. For user
interfaces this makes it difficult for the student to provide
an objective and critical evaluation.
An important step in usability assessment is to identify the
tasks that the device is to be used for, which helps to take
the focus away from features and layouts, and put it onto the
sequence of operations that a user would perform to achieve
their goals. It is not unusual for a device to be frustrating to
use because in practical situations the functionality of the
device can be difficult or slow to access, and this is revealed
if a task is evaluated instead of just a feature of the device.
Only about 36% of the submitted reports described the task
that a user might do.
Another useful approach that can pick up usability issues
is observing another person using the interface. Only 9 students
did this, and all these projects clearly met the criteria;
a third-person observation would apparently make it easier
to report on usability, although the sample is too small to
draw firm conclusions. Some of the students who observed
another person took notes of every step the person took.
This approach was particularly convincing when the student
described their observations, and it set the student up well
to meet the Merit criterion for usability, and Excellence if
they compared another device.
In general it seemed that the students had little teaching
on HCI and usability evaluation. Although not required at
this level, introducing usability heuristics (e.g. from useit.
com) provides a tool for students to analyse systems. Another
approach is the cognitive walkthrough, which is described
for high school students at www.cs4fn.org/usability/
cogwalkthrough.php. This encourages students to observe
someone else using the system, which makes the evaluator
more aware of the steps being made.
4.5 General writing issues
A key phrase in the title and wording of the standard is for
students to “demonstrate understanding” in order to meet
the criteria. It is tempting to respond to criteria such as
“explaining the need for programs to translate between high
and low level languages” with abstract textual explanations
that are paraphrased from books or online sources (and in
fact the instructions for the report explicitly say that para-
17
phrasing is acceptable). Although the general instructions
allow paraphrasing, a more compelling way to show understanding
would be for a student to provide a worked example
using a context unique to themselves (for example, analysing
the interface of their own mobile phone, or running experiments
on the speed of an algorithm on their own computer).
In many cases we found that “showing an understanding”
was interpreted as simply explaining the concept. In this
case the process could end up grading Wikipedia’s knowledge,
and any slips in the paraphrasing process only served
to show that the student didn’t understand the concept!
A common problem in the student reports is giving poor
explanations of what they had done. For example, in explaining
sorting, one student wrote that B is bigger than
A, and A is bigger than C, but we weren’t told what A, B
and C were — quite possibly they were sorting weights or
labeled envelopes, but this wasn’t mentioned at all. Another
problem was assuming that the marker would know things
that the teacher did (perhaps because students are used to
internal assessment); for example, referring to the “method
that was used in class” rather than describing it.
Many reports showed a lack of scientifically convincing reporting,
such as drawing conclusions from just one speed
measurement from each of two algorithms, or having results
spread around the document and not explicitly comparing
them. It was also not unusual for the student to assume that
the reader would go to the trouble of finding patterns in the
results for themselves; for example, one student had a table
of execution times that showed a clear difference between
selection sort, bubble sort and quicksort, but they never articulated
that they had noticed the difference, and thus got
credit for measuring the algorithms, but not for showing an
understanding of the different costs.
The work on the programming languages topic overwhelmingly
used paraphrased quotes. This is largely because it is
a lot harder to personalise (there are only so many compilers
and machine languages that students are likely to encounter,
and the role of interpreters and compilers is fairly well defined).
A similar situation arises with the requirement to
discuss the relationship between algorithms and programs;
there are only a few ways to describe this accurately. In
these cases the criteria aren’t ideal for a report and it would
be better if they prompted examples rather than factual information.
4.6 Guidance for students
The nature of the standards is such that teachers can direct
student work in a variety of ways to enable them to demonstrate
their understanding. Many of the projects reproduced
questions and rubrics from teachers, and the nature of these
instructions had an influence on how easy it would be for
students to produce work that met the criteria.
Several sets of the sampled reports had identical questions
heading each section, which either would have come from
the same teacher, or a group of teachers who had shared
teaching ideas.
Because the quality of student work depended a lot on the
tasks that were set for them, we have attempted to identify

groups of submissions that seem to have been set the same
tasks. Because teacher peer support means that teaching
material is shared, this does not necessarily mean that the
similar papers have the same teachers or are even at the
same school, although inevitably those who are in the same
class probably were given the same assignment.
For example, in one set of reports which had the same questions,
several students compared Insertion sort with Selection
sort, which will have a constant-factor difference. Many
of the students correctly concluded that Insertion sort is
twice as fast as Selection sort, but missed the opportunity
to explore the possibility of a difference that was more than
a constant factor.
Some of the questions were presumably intended to stimulate
discussion, but seemed to be treated by students as an
exam, resulting in very short answers. For example, a series
of questions about choices of programming languages in
one group resulted in very brief answers of just a few words.
Although the questions on programming languages may not
have had the intended effect, the questions on usability for
the same group resulted in generally good responses from
the students. For this the rubric prompted the students to
choose someone who isn’t a digital native (e.g. parent or
grandparent) and observe them using a device for about an
hour.
One group that did particularly well had prompting questions
that didn’t just mirror the criteria, but started off with
a creative situation for the students to get them thinking
beyond just checking off the criteria. For algorithms, the
students were prompted to get the names of favourite songs
from five friends, and then explain searching algorithms using
that list. This provided a familiar context (searching
for songs on an mp3 player), and a meaningful task (selecting
a favourite song). It also ensured unique lists for each
student, even though they were doing the same task. The
usability section for this group began by asking students to
describe the context the device is used in, and the role of the
interface. Once again, this prompt doesn’t directly address
the criteria, but it starts the students thinking of the big
picture, and resulted in good quality answers.
Now that the standard has been in use for a year, teachers
will be able to share best practices based on how well
students were able to respond to the various prompts.
5. CONCLUSION
The analysis above has provided some specific ideas on how
to teach and assess the three topics in the new standard. The
following general observations are based on the analysis of
student work.
Teacher professional development and support is essential:
Because the new standard was introduced very
quickly, and because computer science is a new subject in
New Zealand schools, teachers could not be expected to be
well prepared to teach the new material, either in terms of
subject knowledge or pedagogical knowledge. Unfortunately
this was reflected in the student work, particularly in the
guidance of topics they were given for their investigations
(e.g. comparing two algorithms that didn’t have contrast-
18
ing performance, or investigating only small values of n).
Teachers need to be guided on what the main messages are
that students should take away from this experience, and
be given recommendations of examples that illustrate these
in a compelling way — in this case, contrasting algorithms,
contrasting programming languages, and interfaces that are
small enough to evaluate but have interesting usability issues.
Students do better when they personalize their work:
When students report on their own experiences, whether it
is running an algorithm on their own computer or evaluating
a device they have chosen, their work carries a lot more
authenticity than abstract discussions of a concept. It also
moves their level of understanding up Bloom’s taxonomy:
descriptions of a concept often come across as a paraphrase
that at best reflects remembering facts, while discussions
based on a unique example reflect the student applying their
knowledge.
Assessment using reports: The AS91074 computer science
standard is assessed by a student report submission.
For most of the criteria this provides an opportunity for students
to report on a rich personal experience, such as experiments
with algorithms and evaluation of interface usability
with real users. Through project work they get to experience
the issues first-hand (such as a slow algorithm or a frustrated
user), and get experience communicating in writing about
computer science. However, some of the criteria (e.g. “describing...
the function of a compiler”) invite responses that
are just paraphrases of standard definitions, and would benefit
from clarification or re-wording (e.g. “demonstrate the
function of a compiler using an example”) to encourage students
to explore the concept in a context rather than simply
describe it.
Using provided systems rather than having students
build their own: The best student work was done when
they assessed the performance of a provided implementation
of an algorithm; students who implemented their own programs
tended to be distracted by the implementation, were
limited in the algorithms they could use, had potentially unreliable
performance results, and were biased in their usability
assessments. Because the standard is about algorithm
performance and not programming, those who focussed on
their own program were at risk of missing the big picture,
or else might implement algorithms that didn’t show a good
contrast. Likewise, students who discussed usability based
on interfaces they had implemented couldn’t see their weaknesses.
More advanced students might be expected to implement
their own systems and evaluate them, but at the
beginning high school level the messages about the key topics
are likely to come across more clearly if students start
with a working system.
The introduction of computer science to NZ schools has addressed
a need that was articulated by industry and teachers,
and has happened remarkably quickly. The work done
by students shows some pleasing levels of understanding of
computer science for some; others have at least had experiences
that indicate that they have had a taste of what kind
of topics might come up in computer science courses. With
very little lead-in time hundreds of students have studied

computer science, and with the lessons learned in the first
year, the quantity and quality of students is only likely to
increase in the future.
6. ACKNOWLEDGEMENTS
We are grateful to Scott Telfer (NZQA) for assisting with
access to key data.
7. REFERENCES
[1] O. Astrachan. Bubble sort: An archaeological
algorithmic analysis. ACM SIGCSE Bulletin,
35(1):1–5, 2003.
[2] O. Astrachan, J. Cuny, C. Stephenson, and C. Wilson.
The CS10K project: mobilizing the community to
transform high school computing. Proceedings of the
42nd ACM Technical Symposium on Computer
Science Education, SIGCSE 2011, pages 85–86, 2011.
[3] T. Bell, P. Andreae, and L. Lambert. Computer
Science in New Zealand High Schools. In T. Clear and
J. Hamer, editors, ACE ’10: Proceedings of the 12th
conference on Australasian Computing Education,
volume 32 of Australian Computer Science
Communications, pages 15–22, Brisbane, Australia,
Jan. 2010. Australian Computer Society, Inc.
[4] T. Bell, P. Andreae, and A. Robins. Computer Science
in NZ High Schools: The First Year of the New
Standards. In L. A. S. King, D. R. Musicant,
T. Camp, and P. Tymann, editors, Proceedings of the
43rd ACM technical symposium on Computer Science
Education, Raleigh, NC, USA, pages 343–348, New
York, 2012. ACM.
[5] T. Brinda, H. Puhlmann, and C. Schulte. Bridging
ICT and CS: Educational standards for computer
science in lower secondary education. ACM SIGCSE
Bulletin, 41(3):288–292, 2009.
[6] A. Bruckman, M. Biggers, B. Ericson, T. McKlin,
J. Dimond, B. DiSalvo, M. Hewner, L. Ni, and
S. Yardi. “Georgia computes!”: Improving the
Computing Education Pipeline. ACM SIGCSE
Bulletin, 41(1):86–90, Mar. 2009.
[7] T. Carrell, V. Gough-Jones, and K. Fahy. The future
of Computer Science and Digital Technologies in New
Zealand secondary schools: Issues of 21st teaching and
learning, senior courses and suitable assessments.
Technical report, 2008.
[8] S. Furber, editor. Shut down or restart? The way
forward for computing in UK schools. The Royal
Society, London, 2012.
[9] J. Gal-Ezer and C. Stephenson. The Current State of
Computer Science in U.S. High Schools : A Report
from Two National Surveys. Journal for Computing
Teachers, Spring, 2009.
[10] J. Goode and J. Margolis. Exploring Computer
Science. ACM Transactions on Computing Education,
11(2):1–16, July 2011.
[11] G. Grimsey and M. Phillipps. Evaluation of
technology achievement standards for use in New
Zealand secondary school computing education.
Technical report, New Zealand Computer Society
(NZCS), Wellington, 2008.
[12] R. E. Pattis. Textbook errors in binary searching.
ACM SIGCSE Bulletin, 20(1):190–194, Feb. 1988.
19
[13] A. Robins. Learning edge momentum: A new account
of outcomes. Computer Science Education, 20:37 – 71,
2010.

Is Self-Efficacy in Programming Decreasing with the Level
of Programming Skills?
Michail N. Giannakos*
Norwegian University of Science and
Technology (NTNU)
Dep. of Computer & Information Science
Trondheim, NO-7491, Norway
michail.giannakos@idi.ntnu.no
ABSTRACT
In this study, variables from the Unified Theory of Acceptance
and Use of Technology and Social Cognitive Theory were chosen
as important factors in students’ behavior and attitude towards
Computer Science Education (CSE). This hybrid framework aims
to measure the level of the selected key variables on CSE and
identify potential differences among our different groups. The
three different groups are consisted of students from: (1) programming
courses of Greek Lyceums, (2) computer science
courses at German Gymnasiums and (3) at the Department of
Informatics (freshmen) at the Ionian University, Greece. We
asked the three different groups of students to complete a questionnaire
that was derived from this combination of the prior
theories. The results revealed several differences in the measured
variables, e.g. the freshmen who should have the highest programming
competencies compared to the other two groups expressed
the lowest degree of self-efficacy. The overall outcomes
are expected to contribute to the understanding of students’ likelihood
to pursue computing related careers and promote the acceptance
of CSE.
Categories and Subject Descriptors
K.3.2 [Computer and Information Science Education]: Computer
Science Education, Curriculum.
General Terms
Measurement, Experimentation, Human Factors.
Keywords
ICT courses, Programming courses, Informatics, Secondary education,
Students’ beliefs, cross-cultural.
1. INTRODUCTION
As the success of Computer Science Education (CSE) might
depend on students’ perceptions, attitude and beliefs, we aim to
identify the differences in regard of these factors between students
of a) secondary schools in Germany, b) secondary schools in
Greece and c) freshmen of CS at a university in Greece. Based on
existing theories as well as on prior work [11, 17], we have chosen
variables related to students’ attitude and applied them to
students from two different countries that attended three different
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
Conference’10, Month 1–2, 2010, City, State, Country.
Copyright 2010 ACM 1-58113-000-0/00/0010…$10.00.
Peter Hubwieser
Technische Universität München
Fakultät für Informatik
Boltzmannstr. 3. D-85478 Garching
+49 89 289 17350
Peter.Hubwieser@tum.de
20
Alexander Ruf
Technische Universität München
Fakultät für Informatik
Boltzmannstr. 3. D-85478 Garching
+49 89 289 17350
rufa@in.tum.de
programming courses in secondary school and university, respectively.
The goal of this investigation was to measure students’
beliefs and to identify potential differences among different educational
contexts. As students’ beliefs and attitude are highly
correlated with their performance, students’ perceptions have an
impact on what they will learn as well as how they plan to continue
their course of study [15].
The study presented in this paper was conducted on a large sample
in Greece and Germany and seeks to provide insights for the
following research aspects:
a) To explore students’ perceptions regarding CS,
b) To investigate potential differences in perceptions regarding
CS amongst German and Greek upper secondary education
students,
c) To investigate potential differences in perceptions regarding
CS amongst the ending of secondary education and the beginning
of higher computing education students.
The study attempts to fill the gaps in the CSE literature regarding
students' perceptions in different countries and in different educational
levels. Based on the empirical results we could further our
understanding of students CS perception in the transition from
upper secondary education to higher computing education. The
study is significant in that it focuses on the upper secondary education
level, is situated in real conditions context collecting empirical
data, while at the same time addressing not students' career
choices with regard to CS, but their perceptions on several aspects
regarding CS. The outcomes of the study could further our understanding
of students participation in the field of computing and
could provide useful insight into the actions that should be devised
within the educational system with a view to encouraging
this participation.
2. THEORETICAL BACKGROUND AND
RELATED WORK
Students’ perceptions and intentions are considered as important
determinants of the learning success. Reluctance towards adoption
of informatics courses implies that research is needed to understand,
more comprehensively, how students can be engendered.
Although past research [3, 5] has empirically explained several
issues regarding students perceptions and beliefs regards CSE, it
is mostly focused on higher education and more specifically on
CS departments.
Several models and theories have been applied to address issues
of students’ attitude, perceptions and to identify the cause and the
* This work was carried out during the tenure of an ERCIM "Alain Bensoussan"
Fellowship programme. The research leading to these results has
received funding from the European Union Seventh Framework Programme
(FP7/2007-2013) under grant agreement no 246016.

effect of different factors on the adoption of science education [6].
For instance, the Unified Theory of Acceptance and Use of Technology
[18] and Social Cognitive Theory [2] are some of the most
widely applied theories in the context of students’ behavior [6]. In
addition, Confidence, Performance Expectancy, Social Influence,
Satisfaction, Self-Efficacy and Perceive Behavioral Control are
some of the most commonly used factors (i.e., [5, 6]) affecting
students’ intention to attend a respective course.
3. THE EDUCATIONAL CONTEXT
The empirical study was conducted in three Greek Lyceums (3 rd
grade), in classes of the 11 th and 12 th grade at two Bavarian (German)
Gymnasiums and in the Department of Informatics at Ionian
University in Corfu, Greece (1 st year). As there might be a substantial
influence of the specific educational context on our results,
we start with a short description of the school systems and
the differences of the CS education policies in the two regarded
countries. A very detailed description was presented recently by a
Working Group named: Computer Science/Informatics in Secondary
Education on the ITiCSE 2011 [13].
3.1 The Greek Lyceum
The 1 st grade of Greek Lyceum (Lykio), represents an orientation
year with a general education program. The 2 nd and 3 rd grades
offer three curricular directions: Theoretical, Scientific and Technological.
Students who follow the technological direction are
taking a specific course named Applications Development in a
Programming Environment (ADPE) that involves the development
of algorithms and programming. This course has been taught
for ten years. It focuses on the algorithmic approach and on the
development of problem-solving skills in a programming environment.
This subject is assigned to CS teachers.
The overall aim of 3 rd Lyceum programming courses is to develop
analytical and synthetic thinking, acquire methodological skills
and be able to solve simple problems within a programming environment.
This programming course has not been designed to
educate programmers, and for this reason it is not designed to
teach sophisticated programming techniques; it focuses on approaches
and techniques of problem solving with emphasis on
structured thinking. Many basic algorithmic and programming
concepts, such as conditions, expressions and logical reasoning,
are fundamentals of general knowledge and skills to be acquired
in general education.
The curriculum states that this subject must be taught (at least
partially) in a computer lab. The Ministry of Education has certified
specific Educational Software to support the lab work, especially
for the Lyceum programming course. The Educational
Software has been designed to support teaching, to complement
the subject's needs and IT use and to help students to consolidate
the material. The certified software includes an activity space, a
flow chart developer and a programming environment in accordance
with the textbook.
3.2 The Bavarian Gymnasium
The organization of the schools in Germany is different for each
of the 16 states. In the State of Bavaria, in which the survey was
conducted, secondary school starts with the 5 th grade. There are
three different types of secondary schools, which differ in their
orientation and their level of difficulty. The type of secondary
school that directly qualifies for the enrollment at universities is
called Gymnasium, where students spend a total of 8 years. Currently
about one third of all students in grade 8 attends the Gymnasium
[4].
21
CS is a compulsory subject for all students of 6 th and 7 th grade at
Gymnasium. During this time, the students learn to apply objectoriented
modeling to standard software applications (office,
graphics, text processing, file management, hypertext structures,
e-mail) as well as to describe an simulate simple algorithms. To
this purpose, the students use commercial standard software systems
as well as custom-made software tools and programming
systems to solve simple problems applying the algorithmic control
structures.
Starting with 8 th grade, there are offered four different educational
directions at Gymnasium. In the scientific-technological direction
CS continues to be a compulsory subject in the 9 th and 10 th grade.
While programming skills are not very relevant for the 9 th grade –
it deals primarily with functional modeling and relational database
systems – the students in 10 th grade learn OOM and OOP (incl.
inheritance). Nevertheless, the training programming skills is not
the main focus, which lies on modeling competencies instead. The
teachers are free to decide about the programming language and
development environment the classes use, but they mostly agree
to chose Java and BlueJ.
In the last two grades (11 and 12) of Gymnasium, the so-called
qualification stage, the students learn in courses that they can
choose themselves, disregarding some subjects that are compulsory
(e.g. Mathematics and German Language) and some additional
regulations. To choose a CS course, the student must have selected
the scientific-technological branch before. CS in the 11 th grade
is based strongly on the learning outcomes of the 10 th grade – the
learning content consists mainly of dynamic data structures and
software engineering – whereas the topics of the 12 th grade are
elements of theoretical and technical CS. Unfortunately, in 2009
only about 7% of all students of the qualification stage had chosen
a CS course (according to the Bavarian School Administration).
All German students who have completed our survey come from
this group, most from 11 th grade (19 of 29).
3.3 The Ionian University
The curriculum of the Department of Informatics of Ionian University
focuses on the broad area of CS/Informatics with two areas
of specialization namely: (a) Humanistic Informatics and (b)
Information Systems. The curriculum comprises a set of core
courses and a set of elective courses through which the student
specializes in one of the two aforementioned specializations. The
curriculum was established early at the inception of the Department
in 2004 and based on accepted international standards
(ACM IS 2002). Regarding the first semester of the department
which is in our interests, it has a general knowledge program (the
orientations start at 5 th semester) and the following courses are
being taught: 1) Introduction to Computer science, 2) Introduction
to Programming, 3) Mathematical Analysis, 4) Linear Algebra
and 5) Introduction to Information Society.
4. THE SURVEY
The questionnaire handed out to the students was divided into two
parts. The first included questions on the demographics of the
sample (age and gender) and the second part included measures of
the various factors identified in the literature from previous researches.
Table 4 in the Appendix lists the questionnaire factors,
their operational definition, their items and the source from the
literature review. In all cases, 7-point Likert scales were used.
4.1 Sampling
The survey was open in the middle of the school year 2011-2012
at the Greek Lyceums and the German Gymnasiums and at the
beginning of 2011-2012 year of study in the Department of In-

formatics. The final sample of respondents was comprised of 115
Students. From the total of students, 55 (47.8%) attended the 3 rd
of Greek Lyceum (16-17 years), 29 (25.2%) the 11 th or 12 th of a
German Gymnasium (16-18 years) and 31 (27.0%) the first year
of study at the Department of Informatics (17-18 years). 88 of the
students were males (76.5%) and 27 (23.5%) females.
4.2 Data Analysis and Results
As proposed by Fornell and Larcker [9], there are three procedures
to assess the convergent validity of any measure in a study:
1) Composite reliability of each construct, 2) Item reliability of
the measure and 3) The average variance extracted (AVE).
Therefore, we started with an analysis of composite reliability and
dimensionality to check the validity of the scale used in the questionnaire.
Concerning the reliability of the scales, Cronbach (CR)
α indicator was applied [7] and inter-item correlations statistics
for the items of the variable were calculated. As stated by to Fornell
& Larcker [9], CR α value greater than 0.7 indicates a high
reliability. Table 1 demonstrates the result of the test that revealed
acceptable indices of internal consistency in all the factors.
Table 1. Summary of Measurement Scales
Factors Items Mean S.D. CR Loads AVE
PE PE1 4.52 1.95 0.93 0.74 0.67
PE2 4.16 1.75 0.85
PE3 4.40 1.72 0.84
PE4 4.51 1.69 0.83
STF STF1 4.97 1.51 0.92 0.72 0.62
STF2 4.98 1.58 0.73
STF3 5.47 1.61 0.86
STF4 5.23 1.68 0.82
SI SN1 4.04 2.00 0.82 0.76 0.63
SN2 4.12 2.04 0.82
SEF SEF1 3.89 1.79 0.71 0.82 0.56
SEF2 4.19 1.73 0.70
SEF3 2.62 1.69 0.72
BI BI1 4.90 2.09 0.96 0.82 0.73
BI2 4.87 2.03 0.87
BI3 4.36 2.07 0.88
CPS CPS1 4.88 1.46 0.88 0.76 0.54
CPS2 4.89 1.44 0.75
CPS3 4.79 1.57 0.83
CPS4 4.36 1.62 0.58
C(CL)C CCC1 5.31 1.69 0.91 0.61 0.53
CCC2 5.13 1.60 0.85
CCC3 4.84 1.76 0.62
CLC1 5.50 1.59 0.71
CLC2 5.11 1.58 0.77
CLC3 5.03 1.55 0.78
CDS CDS1 4.74 1.68 0.90 0.80 0.59
CDS2 4.44 2.03 0.83
CDS3 4.82 1.84 0.82
CDS4 4.48 1.97 0.75
CDS5 4.07 1.92 0.63
The reliability of an item was assessed by measuring its factor
loading onto the underlying construct. Hair et al. [12] recommended
a factor loading of 0.5 to be good indicator of validity at
the item level. The factor analysis identified eight distinct factors
(Table 1): 1) Performance Expectancy (PE), 2) Satisfaction (STF),
3) Social Influence (SI), 4) Self-Efficacy (SEF), 5) Behavioral
Intention (BI), 6) Confidence with Problem Solving (CPS), 7)
Confidence for using Data Commands (Conditional-Loop)
(C(CL)C) and 8) Confidence for Data Structures (CDS).
The third step for assessing the convergent validity is the average
variance extracted (AVE); AVE measures the overall amount of
22
variance that is attributed to the construct in relation to the amount
of variance attributable to measurement error. Convergent validity
is found to be adequate when the average variance extracted is
equal or exceeds 0.50 [16].
To examine the research questions regarding the differences in
students’ perceptions among German Gymnasium, and CS freshmen,
we used an Analysis of Variances (ANOVA), including the
eight factors as dependent variables and the students’ group as
independent variable. As we can see from the outcome data in
Table 2, students’ group has a significant impact on students’ PE,
STF, SI, BI and CDS. On the other hand students’ group does not
exhibit significant difference on students’ SEF, CPS and C(CL)C.
Table 2 displays our results regarding the significance of the
differences, while Figure 3 shows the average results for each
factor over the groups. SEF, CPS and C(CL)C have no significance
difference among the three groups and from Figure 2 we
can notice that these factors are on the same levels at each group.
On the other hand PE, STF, SI, BI and CDS do have significant
differences among the groups and in some cases these difference
are quite remarkable (i.e., BI, PE).
Table 2. The differences among the students’ groups
Factor Mean (S.D.) F Result
Lyceum Gymnas. CS
(GR) (GE) Freshmen
PE 4.10 (1.58) 3.75 (1.70) 5.61 (0.79) 14.97** S.D.
STF
SI
5.03 (1.39)
3.54 (1.90)
4.65 (1.67)
3.93 (1.79)
5.90 (0.90)
5.19 (1.34)
6.86**
9.15**
S.D.
S.D.
SEF 3.58 (1.35) 3.77 (1.61) 3.34 (0.94) 0.78 I.D.
BI 3.94 (1.97) 4.48 (1.59) 6.24 (0.90) 17.11** S.D.
CPS 4.51 (1.46) 4.84 (1.44) 5.06 (0.79) 1.81 I.D.
C(CL)C 5.14 (1.37) 4.74 (1.52) 5.55 (1.10) 2.73 I.D.
CDS 4.10 (1.44) 4.66 (1.82) 5.08 (1.51) 4.03* S.D.
**p

(GE) and freshmen CS and Lyceum (GR). Table 3 summarizes
the significant results from the Games-Howell post hoc test.
Table 3. Games-Howell post hoc test
Freshmen CS
Gymnasium (GE) PE*, STF*, SI*, BI*
Lyceum (GR) PE*, STF*, SI*, BI*, CDS*
* The mean difference is significant at the 0.05 level.
5. CONCLUSION AND DISCUSSION
Looking at Figure 2, we can easily notice that the scores of the
Greek and German secondary education students are generally on
the same levels. Additionally, our Games-Howell test showed that
there is no significant difference in the perceptions of Greek and
German secondary education students.
On the other hand, compared to the school students, the Greek
higher education students had significantly higher scores in 4
(compared to German Gymnasium) respectively 5 (compared to
Lyceum) variables. Nevertheless, there are 3 factors, where is no
significant difference (SEF, CPS and C(CL)C). Notably, it seems
that these factors ranged on the same levels although CS freshmen
have more exposure on CS and Programming. This can be possibly
based to the fact that in secondary education educators are
using learner friendly environment (e.g. BlueJ, Alice or Scratch);
and these environments increasing students’ self-efficacy and
confidence regarding a CS and Programming [1, 8]. However, for
the case of data structures (CDS) we identified a significant difference
among the students of Lyceum (GR) and the CS freshmen,
as the latter look more confident for data structures.
The most astonishing result is the dramatic (relative) drop in Self-
Efficacy (SEF), which might indicate that those “future CS professionals”
are not as self-confident regarding their programming
abilities, as one would expect at the first glance, as this is the only
factor indicating a lower level or the CS freshmen, although this
difference is not significant, too. It might be due to the fact that
CS freshmen had to deal with substantially more demanding
problems that they found as hard to solve as the school students
found their more easy ones. In contrast, PE, SI and STF are also
indicating significance differences among the CS freshmen and
the secondary education students. This may be possible explained
to the wide enrolment and familiarity of freshmen with CS and
programming and shows that they are quite content with their
choice of career.
The most significant difference among the two courses is indicated
in students’ BI. This seems plausible, as the freshmen had
chosen their career at this point of time and regarded themselves
as specialists already.
As with any empirical study, there are some limitations. First, in
this study the respondents are Greek and German students, who
had attended the respective educational systems; this may limit
the extend of the generalization of the findings. However, this
study is one of the few so far which combines empirical data from
students’ beliefs in a cross-cultural framework; as such these
findings are expected to shed light on that direction. Secondly, the
data are based on self-reported method, other methods such as
depth interviews and observations could provide a complimentary
picture of the findings through data triangulation. Despite these
limitations, the findings of this study generate valuable insights,
which can be used as part of hypotheses for representative followup
studies in students’ beliefs for CS education and in cross-
country and cultural contexts of CS education research.
23
6. REFERENCES
[1] Anderson, M., et al., (2011). Affecting attitudes in first-year
computer science using syntaxfree robotics programming.
ACM Inroads 2, 3, 51-57. DOI=10.1145/2003616.2003635
[2] Bandura A. (1986). The explanatory and predictive scope of
self-effıcacy theory. J Clin Soc Psychol, 4, 359–73.
[3] Barker, LJ, McDowell, C, Kalahar. K. 2009. Exploring factors
that influence computer science introductory course students
to persist in the major. SIGCSE Bull. 41, 1, 153-157.
[4] Bayerisches Landesamt für Statistik und Datenverarbeitung
(2010). Verteilung der Schüler in der Jahrgangsstufe 8 nach
Schularten und Regierungsbezirken - Schuljahr 2010/11.
https://www.statistik.bayern.de/medien/statistik/bildungsozia
les/verteilung_der_sch__ler_2010.2011.pdf
[5] Biggers, M, Brauer, A and Yilmaz. T. (2008). Student perceptions
of computer science: a retention study comparing
graduating seniors with cs leavers. In SIGCSE 402-406.
[6] Chen, K., Razi, M. and Rienzo, T. (2011), Intrinsic Factors
for Continued ERP Learning: A Precursor to Interdisciplinary
ERP Curriculum Design. Decision Sciences Journal of
Innovative Education, 9, 149–176.
[7] Cronbach, L.J. (1951). Coefficient alpha and the internal
structure of tests. Psychometrika, 16(3), 297-334.
[8] Dillon, E., Anderson, M., and Brown. M. 2012. Comparing
feature assistance between programming environments and
their "effect" on novice programmers. J. Comput. Sci. Coll.
27, 5, 69-77.
[9] Fornell, C., Larcker, D.F. (1981): Evaluating structural equation
models with unobservable variables and measurement
error. Journal of Marketing Research, 48, 39--50.
[10] Games, P.A., Keselman, H.J., & Rogan, J.C. (1981). Simultaneous
pairwise multiple comparison procedures for means
when sample sizes are unequal. Psychological Bulletin, 90,
594-598.
[11] Giannakos, M. N et al. (2011). Programming in secondary
education: benefits and perspectives. In Proc. of the ITiCSE
'11, ACM, 349. DOI= 10.1145/1999747.1999863
[12] Hair, J.F., et al., (2006): Multivariate data analysis, 6th edn.
Upper saddle River, NJ: Prentice-Hall International
[13] Hubwieser, P, et al., (2011). Computer science/informatics in
secondary education. In Proc. of the ITiCSE-WGR '11,
ACM, 19-38. DOI=10.1145/2078856.2078859
[14] Lin, C. S., Wu, S. & Tsai, R. J. (2005). Integrated perceived
playfulness into expectation– confirmation model for web
portal context. Information & Management, 42(5), 683-693.
[15] Metcalfe, J., & Finn, B. (2008). Evidence that judgments of
learning are causally related to study choice. Psychonomic
Bulletin & Review, 15,174 –179.
[16] Segars, A.H. (1997). Assessing the unidimensionality of
measurement: A paradigm and illustration within the context
of information systems research. Omega, 25(1), 107-121.
[17] Shih, H. (2008). Using a cognitive-motivation-control view
to assess the adoption intention for Web-based learning.
Computer & Education, 50, (1), 327-337.
[18] Venkatesh, V. Morris, M.G. Davis, G.B. Davis, F.D. (2003)
User acceptance of information technology: toward a unified
view, MIS Quarterly 27(3), 425–478.

APPENDIX
Table 4. The Factors, their Definitions and their items
Factors Operational Definition Items Source
Performance The degree to which an Using programming improves my performance in a task. (PE1) [18]
Expectancy individual believes that Programming enhances my effectiveness in tasks progressing.
(PE)
attending the respective
course is useful for him/her.
(PE2)
Programming would make it easier to complete a task. (PE3)
Programming increases productivity in completing tasks. (PE4)
Satisfaction The degree to which a I am satisfied with the programming experience. (STF1)
[14]
(STF)
person positively feels with I am pleased with the programming experience. (STF2)
the respective course. My decision to use programming was a wise one. (STF3)
My feeling to use programming was good. (STF4)
Social Influ- The degree to which an People who are important to me think that Ι should learn pro- [14]
ence (SI) individual perceives that gramming. (SI1)
most people who are im- People who influence my behavior encourage me to learn proportant
to him think he
should or should not attend
the respective course.
gramming. (SI2)
Self-Efficacy The degree of conviction I could complete a programming task …
[17]
(SEF)
that one can successfully if there was no one around to tell me what to do. (SEF1)
execute the operation re- if I had only instructions for reference. (SEF2)
quired to produce the outcomes.
if I had never used it before. (SEF3)
Behavioral The degree of students’ I intend to continue learning programming in the future. (BI1) [18]
Intention (BI) willingness to attend the I will continue learning programming in the future. (BI2)
respective course
I will regularly learn programming in the future. (BI3)
Confidence I feel confident in …
Problem Solv- The degree of students’ understanding the problems presented to me. (CPS1)
[11]
ingConfidence (CPS)
confidence to successfully
cope with various problems
identifying the components of a problem. (CPS2)
analyzing a problem to other simpler ones. (CPS3)
posing a problem, formulating it accurately and completely.
(CPS4)
Confidence The degree of students’ formulating the forms of conditional statement if. (CCC1) [11]
for using Data
Commands
confidence to successfully
use data commands
discerning the differences of the forms of conditional statement
if. (CCC2)
(Conditional-
selecting the best form of conditional statement depending on the
Loop)
problem. (CCC3)
formulating the loop statement.(CLC1)
selecting the best loop statement. (CLC2)
using the appropriate loop statement. (CLC3)
Confidence The degree of students’ deciding whether it is necessary to use an array. (CDS1)
[11]
for Data Struc- confidence to successfully selecting the formula of array (one‐dimensional, two‐
tures (CDS) use data structures
dimensional, etc.). (CDS2)
entering, processing and printing the items of an array. (CDS3)
doing general exercises and exercises of searching and sorting
using the structure of the array. (CDS4)
defining the structures of the stack and queue with the correspondent
operations. (CDS5)
24

Performance
Expectancy
(PE)
Satisfaction
(STF)
Social Influence
(SI)
Self-Efficacy
(SEF)
Behavioral
Intention
(BI)
Problem
Solving
Confidence
(CPS)
Confidence
for using
Data Commands
(Conditional-
Loop)
Confidence
for Data
Structures
(CDS)
Table 5. Complete Results of the Games-Howell post hoc test
Independent Variables Mean Std.
(I) (J)
Differ.
(I-J)
Error
Sig.
3 rd Lyceum (GR) Gymnasium (GE) 0.35 0.38 0.63
Freshmen CS -1.51* 0.26 0.00
Gymnasium (GE) 3 rd Lyceum (GR) -0.35 0.38 0.63
Freshmen CS -1.86* 0.35 0.00
Freshmen CS 3 rd Lyceum (GR) 1.51* 0.26 0.00
Gymnasium (GE) 1.86* 0.35 0.00
3 rd Lyceum (GR) Gymnasium (GE) 0.38 0.36 0.55
Freshmen CS -0.87* 0.25 0.00
Gymnasium (GE) 3 rd Lyceum (GR) -0.38 0.36 0.55
Freshmen CS -1.25* 0.35 0.00
Freshmen CS 3 rd Lyceum (GR) 0.87* 0.25 0.00
Gymnasium (GE) 1.25* 0.35 0.00
3 rd Lyceum (GR) Gymnasium (GE) -0.39 0.42 0.62
Freshmen CS -1.66* 0.35 0.00
Gymnasium (GE) 3 rd Lyceum (GR) 0.39 0.42 0.62
Freshmen CS -1.26* 0.41 0.01
Freshmen CS 3 rd Lyceum (GR) 1.66* 0.35 0.00
Gymnasium (GE) 1.26* 0.41 0.01
3 rd Lyceum (GR) Gymnasium (GE) -0.19 0.35 0.86
Freshmen CS 0.24 0.25 0.61
Gymnasium (GE) 3 rd Lyceum (GR) 0.19 0.35 0.86
Freshmen CS 0.43 0.34 0.44
Freshmen CS 3 rd Lyceum (GR) -0.24 0.25 0.61
Gymnasium (GE) -0.43 0.34 0.44
3 rd Lyceum (GR) Gymnasium (GE) 0.54 0.46 0.47
Freshmen CS -2.29* 0.32 0.00
Gymnasium (GE) 3 rd Lyceum (GR) 0.54 0.46 0.47
Freshmen CS -1.75* 0.40 0.00
Freshmen CS 3 rd Lyceum (GR) 2.29* 0.32 0.00
Gymnasium (GE) 1.75* 0.40 0.00
3 rd Lyceum (GR) Gymnasium (GE) -0.32 0.33 0.60
Freshmen CS -0.54 0.24 0.07
Gymnasium (GE) 3 rd Lyceum (GR) 0.32 0.33 0.60
Freshmen CS -0.22 0.30 0.75
Freshmen CS 3 rd Lyceum (GR) 0.54 0.24 0.07
Gymnasium (GE) 0.22 0.30 0.75
3 rd Lyceum (GR) Gymnasium (GE) 0.40 0.34 0.47
Freshmen CS -0.41 0.27 0.29
Gymnasium (GE) 3 rd Lyceum (GR) -0.40 0.34 0.47
Freshmen CS -0.81 0.35 0.06
Freshmen CS 3 rd Lyceum (GR) 0.41 0.27 0.29
Gymnasium (GE) 0.81 0.35 0.06
3 rd Lyceum (GR) Gymnasium (GE) -0.56 0.39 0.33
Freshmen CS -0.98* 0.33 0.01
Gymnasium (GE) 3 rd Lyceum (GR) 0.56 0.39 0.33
Freshmen CS -0.42 0.43 0.61
Freshmen CS 3 rd Lyceum (GR) 0.98* 0.33 0.01
Gymnasium (GE) 0.42 0.43 0.61
*. The mean difference is significant at the 0.05 level.
25

InfoSphere: An Extracurricular Learning Environment for
Computer Science
Nadine Bergner
Lehr- und Forschungsgebiet
Informatik 9, RWTH Aachen
Ahornstr. 55
52074 Aachen
+49 241 8021933
bergner@cs.rwth-aachen.de
ABSTRACT
This paper describes one of our measures to raise students’
interest for Computer Science (CS) and to provide them with a
realistic idea of the field. We outline the underlying concepts and
theories of InfoSphere – the extracurricular learning environment
for CS at RWTH Aachen University. After explaining the
theoretical, organizational, and infrastructural foundations, we
provide an overview over the didactical concept of our 13
different CS workshops for school students of different ages.
Furthermore, we introduce the benefits of InfoSphere for our three
different target groups: school students, university students in
teacher training, and active CS teachers. Finally, we present first
results from our ongoing evaluation of the school students’
perception of the field of CS before and after visiting one of the
InfoSphere workshops.
Keywords
Extracurricular learning environment; didactical concept;
computer science workshops; school students; students in teacher
training; CS teachers.
1. MOTIVATION
There is a near wide agreement of economy and society that we
have a growing shortage of STEM (science, technology,
engineering and mathematics) graduates in general and
specifically a growing need for computer scientists. Despite the
fact of splendid job opportunities, too few students major in
Computer Science (CS). Main reasons for this mismatch are the
low interest and little previous knowledge in STEM topics among
school students [20]. For CS the situation is especially critical
because it is no compulsory school subject in most German states,
e.g. North Rhine-Westphalia. Therefore, mainly the social
environment is responsible for the children’s interest in CS [18].
There are a lot of prejudices and false images about computer
scientists, e.g. that they are loners and do not like to work in
teams [18]. As Engeser and Limbert showed it is specifically
important necessary to get young kids into contact with CS topics,
to prevent these misconceptions [8]. This is the only way to reach
the students before they decide against a CS course, which is only
an elective in school, and in the following against a CS study
program at university because of the negative social stereotypes.
In order to raise interest in CS and STEM in general and to
communicate a realistic idea about CS, the Learning Technology
Research Group of RWTH Aachen University established
Jan Holz
Lehr- und Forschungsgebiet
Informatik 9, RWTH Aachen
Ahornstr. 55
52074 Aachen
+49 241 8021935
holz@cs.rwth-aachen.de
26
Ulrik Schroeder
Lehr- und Forschungsgebiet
Informatik 9, RWTH Aachen
Ahornstr. 55
52074 Aachen
+49 241 8021930
schroeder@cs.rwth-aachen.de
InfoSphere 1 as an extracurricular learning environment (in
German Schülerlabor) for CS, which opened in 2010. The main
advantage of an extracurricular learning environment is its high
flexibility in time and space which allows for active and selfregulated
and explorative learning methodologies. Furthermore, it
offers extraordinary opportunities in educational technology and
specifically prepared learning materials which cannot be found in
normal school classes. Because of these possibilities various
topics of CS are presented in an attractive way and thus help to
represent an interesting and multifaceted image of CS in general
and thus also counteract social stereotypes. At the moment we
offer 13 workshops about a variety of CS topics for school
students from the age of eight up.
Additionally, InfoSphere provides opportunities for students in
teacher training for CS as well. They can gain first experiences in
teaching in a prepared and safe environment with pupils and
receive feedback from tutors. We let our CS teacher-students
design learning materials and have them supervise InfoSphere
workshops. The idea is to link the theoretical CS and teaching
knowledge acquired in university courses with practical
experiences in an early stage of their education for formative
evaluation.
Within InfoSphere we also address CS teachers of all school types
as our third target group. Their situation is often difficult because
CS is a novel school subject with too few learning materials
provided by established school books. Also, many current CS
teachers have not studied CS as a topic at university. Instead,
they only took part in a further training. Furthermore, CS
constantly changes and evolves. All this makes it is a complex
task to teach CS in school. With our workshop offerings we try to
support and inspire teachers, by providing them the possibility to
visit us with whole school classes. Moreover we make additional
school materials available free of charge and support external
teacher training programs.
In summary, InfoSphere is targeted at school students, university
students in teacher training, and active CS teachers. First of all,
we provide an opportunity for school students to find out about
various topics of CS and actively gain an own picture about what
CS is and if it might be an interesting field for their studies and
later profession. On the other hand, we utilize the flexible and
innovative learning environment to improve CS education for
university students in teacher training as well as further education
1 http://schuelerlabor.informatik.rwth-aachen.de

for active CS teachers. Furthermore, InfoSphere represents a
potent research environment, where we can explore and analyze
innovative learning methods, educational designs and media with
school students. To set up research experiments would be a lot
harder to achieve in regular school classes.
2. THEORETICAL BACKGROUND
There is ongoing research about extracurricular learning since
several years. In 1994 Griffin described how learning in informal
science settings would be most effective [11]. She looked into the
students’ view of learning in such informal settings and identified
orientation and preparation as the most important factors.
Otherwise the unknown learning environment attracts most of the
attention instead of the workshop itself. Thus, we connect our
workshops with topics of regular school lessons. More
specifically regarding extracurricular learning environments,
Euler scrutinized how school students can act as researchers and
what the resulting learning effects can be [9]. The main result is
that it is important for the school students to learn on their own,
which means they have to actively work with the contents instead
of just listening to a teacher. Based on these findings we design
our workshops with large proportions of action-oriented learning.
In addition, Guderian studied the learning effects of pupils in the
age of 10 to 14 after repeated visits of a physics lab [12]. In
accordance to his findings we offer additional workshops for
individual participants in order to keep up and raise students’
interest.
As a consequence of current findings about extracurricular
learning environments and with regard to research results
concerning learning concepts for STEM topics, we design our
workshops to facilitate explorative, action-oriented learning. So
far, there is little research about this didactical learning concept in
the field of CS, but results from other STEM fields can also be
applied to InfoSphere. For example, Bell has focused on
experimental learning in physics [4]. He worked out the following
activity model (translated by the authors):
Table 1. Activity model following Bell
1. defining
2. exploring
3. reflecting
orienting
formulating the problem
assuming
planning
searching for information
expressing knowledge / modeling
experimenting
analyzing / finding results
presenting
discussing / reflecting
applying
For InfoSphere workshops we distinguish three phases of defining
the problem, exploring by planning/modeling/experimenting and
reflecting by presenting/discussing learning and working results.
The first stage focusses on the students’ self-directed process of
finding and defining the problem. An authentic, all-day situation
27
is presented to the students and questions lead them to discover a
problem to be solved. The second stage is the main working phase
during the InfoSphere workshops. The participants work in
groups to solve tasks on different complexity levels. Thereby, we
follow a scheme leaning towards a software developing process of
planning, modeling, experimenting and realization. In the third
stage we encourage reflection and additionally promote students’
communication and presenting skills by letting them present their
findings to the other participants.
Aepkers et al. [1] studied exploratory, inquiry-based learning in
general. Many of their findings describe effects on the learning
process which are independent of the content and thus are applied
to InfoSphere workshops as well.
One specific goal of InfoSphere is to have participants develop a
more realistic concept of the field of CS. Studies about novices’
expectations and prior knowledge of CS e.g. [21] and [8] show
that the general perception of CS does not match well with reality.
Therefore, we consider it important to foster a more realistic
perception of CS, and to avoid stereotypes, and raise an interest
for elective CS courses. In detail, Knobelsdorf evaluated the
motives of successful CS university students and explored reasons
for dropping out. Her analyses show that the most mentioned
reason for problems in CS studies is “too challenging
requirements” followed by “too low mathematical prior
knowledge”, “lack of time” and “poor motivation” [16]. To avoid
students from enrolling into a CS university program with wrong
expectations we also inform about the contents and requirement of
CS study programs. This cannot be achieved only in InfoSphere
workshops. We rather educate (future) CS teachers as multipliers,
who observe the misconceptions concerning CS during
workshops and while preparing InfoSphere learning media. For
the other mentioned problems we also utilize InfoSphere for
freshman before they start their study program in prep courses.
Also in the first two terms we offer InfoSphere workshops for
students who experience difficulties in their normal courses.
These workshops provide them with an alternative access to the
current topics and give them a second chance catch up and pass
the exam. These uses are not further described in the following of
this paper; see [2] and [14] for more details.
3. INFOSPHERE CONCEPT
In order to enable building a broader perception of CS we select
exemplary topics from many different fields to demonstrate the
wide range of CS. The collection of workshop topics is based on
fundamental ideas of CS as described by Schwill [22]. In
addition, we took German CS school curricula as well as
educational standards [10] into account and link our workshops to
regular school lessons [6].
Furthermore, InfoSphere is targeted at a variety of target groups:
school students, university students in CS teacher training, CS
teachers, as well as other professional or non-professional people
with an interest in CS. The following subsection describes the
advantages for all target groups in more detail.
3.1 Target Groups
First of all, for primary and secondary school students InfoSphere
is the place to get in contact with CS. They can gain an insight
into the world of informatics without feeling school pressure. All
the workshops are held by students in teacher training of RWTH
Aachen University. So the participants’ performance does not
influence their school grades. Besides the possibility to visit

InfoSphere with their whole class as part of their school activity,
school students can take part in additional workshops on their
own. For this purpose we frequently offer special workshops
which allow an individual registration. Thus, we can first reach all
students of a class and try to raise an interest in CS topics and
then continue to support especially interested kids.
InfoSphere is integrated into our educational concept for CS
teacher training, which makes up our second target group.
Teacher students register for three consecutive courses in CS
didactics over three semesters. After learning the basics of CS
didactics during the first course, they create an InfoSphere
workshop in groups of two to three students during the second
course. By designing and preparing the learning materials and
then moderating InfoSphere workshops with school students they
gain practical experience about the didactical reduction of CS
topics, didactical design, utilizing different innovative learning
methodologies, and acting in front of a group of school students.
To merge the specialized knowledge of CS at university on the
one hand and the school students’ prior knowledge and skills on
the other hand is a challenging task when designing a workshop.
Especially the different challenges for different age groups are
addressed during the second semester course. The third course in
CS didactics is used to reflect the previous experiences under
consideration of various pedagogical approaches which can be
used to teach CS in school and in the InfoSphere.
In addition to students from school and university, professional
CS teachers are our third target group. The main value of
InfoSphere for this group consists of the extracurricular learning
units to various and novel CS topics. They can visit InfoSphere
with their class to give their students an enthusing insight into CS.
During the workshop the teachers experience their students from a
neutral and observing perspective, which is almost impossible
during regular lessons. Thereby they get to know much about the
students’ learning methods and social skills. Apart from this, the
teachers can download many additional learning materials for use
in their own school lessons from our website 2 for free.
3.2 Basic Conditions
In the following section we present the basic conditions of our
extracurricular learning environment. These specifications
comprise our public relations, the room concept, the available
technologies and the applied learning materials.
Public Relations
We have designed a website 2 , flyers and several posters. We
present InfoSphere at several conferences across Europe and we
regularly exhibit on events such as the science night of RWTH
Aachen University or the local forum for STEM education. In
addition, we present posters and spread out flyers at many
different activities for school students, e.g. the information day at
the RWTH Aachen University, Girl’s day, or go4IT!-workshops.
Our website helps us to get in contact with CS teachers, interested
school students and their parents. On the website you can find
information about all provided workshops. Teachers, parents or
older students can find out what the workshops are about, what
are the target groups, the duration of the workshops and the
expected prior knowledge. Accordingly, they can choose and
register to the best fitting workshop for their classes, their children
2 http://schuelerlabor.informatik.rwth-aachen.de or
http://www.schuelerlabor-informatik.de
28
or themselves. Two different monthly newsletters inform school
students and adults about current events. In the download area
especially teachers find a lot of material to enhance their regular
school lessons. For example we provide materials for self-study,
which allows teachers to promote their students’ talent without
neglecting the remaining class.
The equipment of an extracurricular learning environment also
plays an important role for its success; starting with the layout of
the rooms, the technological equipment up to the specifically
designed hands-on materials for the workshops.
Room Concept
The building consists of several separated rooms which are placed
on two different floors. The biggest one holds the “experimental
area” and is located in the center of the building over both floors.
The “teaching room” is located on level 2 and can be used for
microteaching units. The further rooms are a contest area, an
office and storerooms for all materials (see Figure 1).
Figure 1. Room concept
The students spend most of the time in the huge experimental
area, which also contains five big tables for six to eight students.
The specific designed furnishings support working in groups and
utilizing technology, which are both typical for CS and especially
software development. This room is ideal for all active parts of the
workshops, because the students are able to talk with each other
about problems and encouraged to experiment with possible
solutions and realize them in groups.
During microteaching phases or students’ presentations we use
the teaching room. The tables and chairs are flexible, so we can
arrange them in rows of chairs, a circle of chairs or small group
tables for four students. One big advantage of changing the
location is that the students stop to work on their tasks and thus
listen to the tutors or their classmates. During a workshop the

participants change between the rooms depending on the
didactical concept of a phase.
The contest area can be used by small groups of students to
discuss or test their work or to prepare presentations. Finally, the
storerooms are very important for us because we use a lot of selfdesigned
hands-on materials for the workshops (see section 4.2
Didactic Concept) which has to be stored.
Additionally to the available seating in the different rooms there is
a lot open space, which can be used to experiment with the handson-materials.
The overall idea is to encourage the students to
move around the building to find classmates to discuss their
questions.
Technologies
The collection of technologies available in InfoSphere also makes
a big difference to regular school classes. In brief, we can work
with laptops, interactive whiteboards, smartphones, tablets,
microcontrollers, multi-touch-tables, digital cameras, and several
more specialized devices such as barcode-scanners or traffic light
models.
We have 32 laptops with Wi-Fi connection. Although it is
theoretically possible to give each student one laptop we only do
so during the evaluation phase at the end of the workshops (see
section 5). During the workshops the participants share one laptop
by two or three students because we want them to work in groups.
This supports creative thinking and also counteracts the prejudice
that computer scientists always work alone. We prefer mobile
computers like laptops, tablets, or smartphones instead of desktop
computers so that participants are able to move around and
cooperate in flexible teams and working with varying technology
or learning materials.
Besides several flipcharts there are two interactive whiteboards,
one in the experimental area and one in the micro-teaching room.
They are height-adjustable and therefore also usable for smaller
kids. These whiteboards are used for presentations or classroom
discussions. A special advantage of InfoSphere over regular
schools are 12 smartphones and 27 tablets (of different types),
which run Android and can be programmed by the students. So
far they are utilized in two different workshops; another one is
currently in development. Furthermore, we are preparing two
workshops based on microcontrollers. With this technological
equipment we demonstrate that CS offers opportunities beyond
the user-side of systems and allows students to create and design
technology according to their own imagination. In addition, there
are three multi-touch-tables available. The big advantage of these
is the possibility to use them with up to eight students for
collaborative learning. Currently, we have only used them for
research prototypes with university students (see [13]) but not yet
during InfoSphere workshops with school students due to
technical problems with the tables. For more information about
our current workshop offering (as of June 2012) see section 4.
Materials
Because of our goal to present many different facets of CS to
primary and secondary school students of all ages it is
fundamental to design a plethora of learning materials. Many of
these are hands-on materials to address different senses. Because
there are no ready-to-use learning materials to all the different CS
topics most of them are especially designed for the corresponding
workshops. These materials consist of videos, simulations, handson
models, physical circuits or self-developed software programs.
29
Furthermore, we use the learning management system Moodle 3
for some workshops. This helps us to provide the digital materials
in a well-organized form. Within the system it is possible to
present texts, pictures, videos or simulations to the participants.
Sometimes the students’ have to answer question or solve little
tasks to proceed to the next section. As a result of this the learning
takes place in a self-regulated manner and there is no need for a
tutor to check the answers. Therefore the teacher has time to
observe her or his school students and the student instructors have
time to help participants with individual problems.
3.3 Workshop Organization
All workshops take place in the InfoSphere environment with its
special infrastructure and well prepared learning materials. It is
also a positive effect that students come to new surroundings with
a different atmosphere. The workshops are mentored by university
students; who help the school students without assessing them.
They are also close in their age. In this way InfoSphere also
targets school students who are not yet interested in CS and
allows them to experiment with different topics without being
under the pressure to perform.
The second significant difference from regular CS lessons in
school is the flexibility in time. There is no structure of 45 or 90
minute slots. The workshops last from three hours up to three
days. This is the reason why it is possible to use all kinds of
different learning methods (see 4.2).
4. WORKSHOP CONCEPT
4.1 Overview over the workshops
The current InfoSphere workshops with the corresponding target
groups are displayed in Table 2. There is one workshop for
primary school students, six workshops for middle school students
and six for high school students. Accordingly, the topics differ
from binary numbers to artificial intelligence.
Table 2. Overview over the existing workshops
Nr. Title Age of the
target group
(in years)
1 Magic of CS – A First Insight into CS
[3, 7]
8-9
2 Easily Programming with Scratch 10-12
3 Tour into the PC 10-12
4 What’s about the Zebra Crossing? –
EAN- & QR-Codes
5 Treasure Hunt with Cryptographic
Methods [5]
10-12
11-13
6 Searching for the Shortest Path 13-16
7 Object-oriented Programming with
Alice
15-17
8 Artificial Intelligence 16-18
9 Lego-Turing Machine 17-18
10 Calculating with DNA 17-19
11 GUI-Programming for Android 17-19
3 http://subprogra.informatik.rwth-aachen.de/moodle2/

12 Smartphone-Remote Control for Lego-
NXTs [13]
18-19
13 Insight into Computer Graphics 18-19
These workshops have been designed with learning materials and
detailed course plans by students in teacher training as seminar
papers or thesis projects. Furthermore, some students work on the
integration of the workshops into school lessons according to the
German CS curriculum [19] and to educational standard
recommendations of the German society for CS (GI) [10].
4.2 Didactic Concept
The fundamental concept of the InfoSphere workshops is
explorative learning. The learning environment supports students
to learn self-directed, which allows for differentiation between
individual learners. We explicitly aim at students who are not yet
interested in STEM topics and not only at highly motivated
students who come to the lab in their spare time. Therefore,
InfoSphere workshops contain open tasks which can be solved on
different complexity levels. Some workshops also offer a lot of
additional tasks, so very good students are challenged further.
Phases in lecture style are reduced to a minimum (microteaching).
Most tasks in InfoSphere are solved in teams with
hands-on materials.
The workshops can either be integrated into a sequence of school
lessons or visited as separate extracurricular learning event. In
future, we are going to design more learning materials to better
incorporate workshops in regular sequences of school lessons.
This should make it easier to integrate a study trip into the school
curriculum [11]. On the other hand, separate workshops are
necessary for all the students, who are not taking up CS in school
yet.
Our aim is to correct the social stereotype of computer scientists.
Thus, InfoSphere workshops demonstrate that CS is not only
attractive for “computer freaks”. We specifically address students
who did not choose to study CS in school (yet), only because of a
wrong CS perception. On the other hand, we also inform
InfoSphere visitors, who seem already interested in the field, what
it means to study CS at a university. As mentioned in section 2
many CS students admit that they had a different idea about
studying CS before enrolling into a program. Many professional
computer scientists develop innovative software which is a
creative process. That is why we explicitly value learning
materials which encourage participants to creative problem
solving. For example the workshop “Artificial Intelligence” ends
up with the open task to implement a chat bot. For most
InfoSphere tasks and projects there is not only one correct
solution. At the end of the workshops students are asked to
present their solutions and describe the problems they
encountered during the solution process. discussion phases
encourage reflecting what computational thinking can mean while
increasing the students’ communication skills by the way. The
workshop “Easily Programming with Scratch” is designed with
the jigsaw teaching technique where is it necessary to help each
other to achieve the goal within the whole group. Good
communication skills will enable the students to learn and work in
collaborative environments in their further life.
Besides the social skills mentioned above many university
students admit that they have not expected how much
mathematics is expected in CS studies. To encounter wrong
expectations, there are also workshops which explicitly make
30
mathematics as the foundation of many areas in CS visible. The
“Insight into Computer Graphics” workshop combines matrix
calculations with the corresponding visual effects on graphical
objects. The students get to know how important mathematic
foundations are to present images on a screen. They learn how
matrix computations allow the enlargement, the translation and
the rotation of any object in a three-dimensional coordinate
system and furthermore which mathematical calculations are
needed to represent a three-dimensional object on a twodimensional
screen.
In addition to choosing the right contents it is also important to
utilize the appropriate methodology and learning media.
Depending on the topic and the age of the target group the course
format, technologies, and materials are selected from the variety
of our educational pool. One of our important messages is that
technologies can not only be used as they are but also be designed
and adapted. For example in our two workshops “GUI-
Programming for Android” and “Smartphone-Remote Control for
Lego-NXTs” [15] the participants learn to program their own
Android-Apps, so they see smartphones not only as useful devices
but experience them also as platform which they can manipulate
and design according to their own imagination. Simultaneously,
we show them the significance of programming as a tool.
Closing up this section there are some didactical decisions which
are independent of the subject of the learning environment but
nevertheless very important for its success. To offer a learning
environment without school pressure the workshops are held by
university students. The unknown learning environment and the
new and young tutors enable a learning situation which is very
different from regular school lessons. That is one reason declared
by several school students who were not interested in CS courses
at school, but were quite motivated and interested when
participating in an InfoSphere workshop. To further support these
students we also offer workshops with individual registration.
These workshops have some other unique effects: the students
experience to learn in groups with unknown students. So they
learn to communicate with each other and to solve the problems in
teams. In addition the participants work in groups of students of
different ages and prior knowledge, e.g. the workshop “Treasure
hunt with cryptographic methods” [5] is open to kids at the age of
11 to 14. So they can learn from each other and combine their
different skills to reach the best possible result. Thereby we try to
support various competencies which are useful in school,
university and vocational training.
5. EVALUATION
In general it is very interesting to conduct research for all three
target groups to evaluate what are the main advantages of an
extracurricular learning environment for CS for them. Currently
we focus on the school students’ view. We explore how a visit of
the InfoSphere can change the students’ conception of CS. As
described above our aim is to counteract social stereotypes about
computer scientists and thus try to raise the number of school and
university students in CS and to reduce the dropout rate.
To examine the changes of the students’ conception of CS we
conduct a quantitative evaluation before and after the workshops.
For that purpose we developed two online questionnaires with
partly equal questions. The students fill out the first one in school
some days before their trip and the second one at the end of the
workshop. In order to correlate the two answers we integrate a
code into the questionnaires to pseudonymise the data sets.

Altogether 830 participants visited the InfoSphere since its first
workshops in January 2011. Figure 2 displays the distribution of
the 41 workshops to the different topics.
Figure 2. Overview over the number of workshops
After development and pretests the actual evaluation phase began
in January 2012. Since then 325 school students registered for an
InfoSphere workshop. Naturally some students take part in more
than one course, so the number of different students who visit the
InfoSphere is a bit lower. For our evaluation our sample consists
of N = 215 school students (301 valid pretests, 265 valid
posttests, 215 valid matches). In addition to this online-evaluation
we also ask primary school students for their opinion but they do
not have to fill out the extensive questionnaires. Therefore they
are not part of the presented evaluation.
More precisely in the first questionnaire we have evaluated the
personal background of the participants, their conception of CS,
their own interests, their performance at school, what they think
about computer scientists and their expectations for the workshop.
Afterwards we ask them the same questions about their
conception of CS and what they think about becoming a computer
scientist. In addition we want to hear their opinion about the
workshop itself.
Figure 3 displays the distribution of the participants by age and
gender. In contrast to the proportion of women in CS courses at
the RWTH Aachen University of 10.7 % (215 out of 2016
students) [17] the proportion of female students in our workshops
is relatively high with 33.5 % females (101 out of 301
participants). Students between 12 and 17 years form our main
target group. The group in the age of 12 to 14 has the chance to
select CS as their second elective at school. The older ones have
to choose their courses in the final two years in school (senior
classes – Oberstufe), select their advanced courses
(Leistungskurse), and make up their mind what to study at
university or during apprenticeship. For both groups it is very
important to be informed what CS really mean, to make a wellconsidered
decision.
We try to reach all types of schools, but as we have mostly contact
to grammar school teachers 80.3 % of our participants attend this
type of school. In addition we try to reach students who have not
been registered in a CS course so far. Up to now 17.8 % of the
participants belong to this group. For this group it is interesting to
analyze which workshops they are especially interested in and if
these workshops help them to decide, whether CS could be a right
topic to choose as an elective.
31
Figure 3. Distribution of the participants by age and gender
We try to provide good access to CS topics for all different kind
of school students. To evaluate what are their motives we ask the
participants for their personal interests in computers and other
technologies, discovering backgrounds, learning in groups and
presenting their results. Figure 4 shows that most of the
participants like working with a computer, develop programs on
their own and get to know novel technologies best.
Figure 4. School students' interests concerning CS (mean on a
scale from 1 to 6)
One specific goal of InfoSphere is to have participants develop a
more realistic concept of the field of CS and at least consider if it
would be the right choice to become an computer scientist.
Therefore we evaluate how much the students think they know
about being a computer scientist. Figure 5 gives an overview over
the results before and after the workshop. Both bar charts show a
peak at number 6, which means that most of the students think
they know quite much about being a computer scientist. The
bigger changes are at number 3 & 4 and number 9 & 10. This
shows that after the workshop more school students think that

they know what being a computer scientist means (67.5% chose a
score of 6 to 10) than before the workshop (58.5% chose 6 to 10).
Figure 5. How much students know about being a computer
scientist (before and after the workshop)
To get to know what the students think about being a computer
scientist we also ask some more detailed questions like “Do you
think CS is interesting?”, “Do you think computer scientists have
to be creative?” or “Do you think CS is something for boys
only?”. As a first analysis of the data Figure 6 gives an overview
over the school students’ idea of CS divided by gender.
Figure 6. School students' image of CS (mean on a scale from 1
to 10)
It becomes obvious that boys and girls differ in some points. In
particular, the differences in “CS is interesting and exciting” and
“I could imagine working with CS in future” are our motivation to
further support especially girls to get in contact with CS. In
contrast to that more girls said that they can imagine what a
computer scientist does. For further studies it is interesting what
the girls exactly think about being a computer scientist.
In conclusion, the evaluation shows that InfoSphere workshops
have an effect on the school students’ thinking about CS. To
research the exact effects we have to analyze the data in more
32
detail and develop follow up evaluations. Nevertheless we can
only evaluate the short-term effects of a single workshop.
Multiple measures are necessary to reach long-lasting effects.
That is the reason why we offer additional workshops with
individual registration. So the interested students can take part in
another workshop even if it is not possible to have second
excursion with the whole class.
6. SUMMARY & OUTLOOK
In conclusion, InfoSphere is an effective way to raise interest in
CS and to support interested school students. To communicate a
realistic conception of CS to school students of all ages, we
provide 13 workshops about different CS topics for kids from the
age of eight years up.
The flexibility in time and place in combination with innovative
technologies enables numerous extravagant learning possibilities.
This is a characteristic of our extracurricular learning environment
which allows us to use action-oriented and self-directed learning.
To further improve the effects of our workshops we are going to
design learning materials to prepare and follow up the courses in
school lessons. In addition we are going to extend our offerings
for single registration and evaluate the effect of repeated visits.
Apart from the advantages for school students InfoSphere offers
additional possibilities for students in teacher training and CS
teachers. Students in teacher training design the learning materials
and are responsible for the courses. Thereby they gain useful
experience for being a teacher. CS teachers can visit InfoSphere
with their classes to enable the school students to get an insight
into the vast field of CS. They have new possibilities to observe
their studnets and can learn about new CS topics and innovative
educational methodologies.
Furthermore, InfoSphere is a magnificent environment for
research in teaching methodology. Up to now we started with
evaluating in which ways InfoSphere workshops affect the school
students’ conception of CS but there are much more interesting
research questions, e.g. what is the effect of the practical training
for students in teacher training or how can extracurricular
workshops be integrated in regular school lessons. For the near
future our focus lies on further und detailed evaluation of the
school students. We are going to have a closer look at the
differences between boys and girls and how the results depend on
the content and structure of the visited workshop.
In addition to the 13 established workshops there are currently 8
more workshops in development (as of June 2012) (see Table 3).
Table 3. Overview over the planned workshops
Title Target age (in
years)
App-programming with the App Inventor 11-15
Algorithms 13-16
Modeling a Crossroad with Traffic Lights 15-16
Web Technologies 15-16
Greenfoot 17-18
Media Computation 17-18
Game Theory 18-19
House Automation 17-18

7. ACKNOWLEDGMENTS
The initial equipment of the InfoSphere was financed by the zdi -
Zukunft durch Innovation.NRW (future by innovation in North
Rhine-Westphalia). Furthermore, InfoSphere cooperates with and
is partly supported by the research project IGaDtools4MINT 4 .
REFERENCES
[1] Aepkers, M., Liebig, S., Bönsch, M., and Kaiser, A. 2002.
Entdeckendes, forschendes und genetisches Lernen.
Unterrichtskonzepte und -techniken 4. Schneider-Verl.
Hohengehren, Baltmannsweiler.
[2] Apel, R., Berg, T. Bergner, N., Chatti, M.A., Holz, J.,
Leicht-Scholten, C., and Schroeder, U. 2012. Ein
vierstufiges Förderkonzept für die Studieneingangsphase in
der Informatik. Accepted for publication in Proccedings of
5. Fachtagung zur Hochschuldidaktik der Informatik HDI
2012. (in preparation)
[3] Batur, F. and Bergner, N. 2012. Grundschulkinder
begeistern mit der Zauberschule Informatik. In Ideen und
Modelle. 5. Münsteraner Workshop zur Schulinformatik.
Books on Demand GmbH, Norderstedt, 87–94.
[4] Bell, T. 2006. Schülervorstellungen und Lernen von Physik.
Forschendes Lernen. Piko-Brief 6. IPN Kiel, Kiel.
[5] Bergner, N., Holz, J., and Schroeder, U. 2012.
Cryptography for Middle School Students in an
Extracurricular Learning Place. In CSEDU 2012 -
Proceedings of the 4th International Conference on
Computer Supported Education, Volume 2, Porto,
Portugal, 16-18 April, 2012. SciTePress, 265–270.
[6] Bergner, N., Holz, J., and Schroeder, U. 2012. Über
fundamentale Ideen hinaus: Informatik im InfoSphere -
Schülerlabor Informatik. In Ideen und Modelle. 5.
Münsteraner Workshop zur Schulinformatik. Books on
Demand GmbH, Norderstedt, 77–86.
[7] Bergner, N., Leonhardt, T., and Schroeder, U. 2011.
Zauberschule Informatik ‐ Einblick in die Welt der
Informatik für Kinder im Grundschulalter. In Weigend, M.;
Thomas, M.; Otte, F. (Hrsg.): Informatik mit Kopf, Herz
und Hand -- Praxisbeiträge zur 14. GI-Fachtagung
"Informatik und Schule" (INFOS2011) Münster: ZfL-
Verlag, in print
[8] Engeser, S., Limbert, N., and Kehr, H. 2010. Studienwahl
Informatik. Abschlussbericht zur Untersuchung, München.
[9] Euler, M. 2005. Schülerinnen und Schüler als Forscher:
Informelles Lernen im Schülerlabor. Naturwissenschaften
im Unterricht. Physik 16, 90, 4–12.
[10] Gesellschaft für Informatik (GI) e. V., Ed. Grundsätze und
Standards für die Informatik in der Schule.
Bildungsstandards Informatik für die Sekundarstufe I.
[11] Griffin, J. 1994. Learning to learn in informal science
settings. Research in Science Education 24, 24, 121–128.
[12] Guderian, P. 2007. Wirksamkeitsanalyse außerschulischer
Lernorte. Der Einfluss mehrmaliger Besuche eines
Schülerlabors auf die Entwicklung des Interesses an
Physik. Dissertation, Humboldt-Universität zu Berlin.
[13] Holz, J., Bergner, N., Schäfer, A., and Schroeder, U. 2012.
Serious Games on Multi Touch Tables for Computer
Science Students. In CSEDU 2012 - Proceedings of the 4th
International Conference on Computer Supported
4 www.igadtools4mint.de
33
Education, Volume 2, Porto, Portugal, 16-18 April, 2012.
SciTePress, 519–524.
[14] Holz, J., Bergner, N., and Schroeder, U. 2012.
Anwendungsorientierte Gestaltung eines Informatik-
Vorkurses als Studienmotivator. Accepted for publication
in Proccedings of 5. Fachtagung zur Hochschuldidaktik der
Informatik HDI 2012. (in preparation)
[15] Holz, J., Leonhardt, T., and Schroeder, U. 2011. Using
smartphones to motivate secondary school students for
informatics. In Proceedings of the 11th Koli Calling
International Conference on Computing Education
Research. Koli Calling ’11. ACM, New York, NY, USA,
89‐94.
[16] Knobelsdorf, M. 2011. Biographische Lern- und
Bildungsprozesse im Handlungskontext der
Computernutzung. Dissertation, Freie Universität Berlin.
[17] Kupferschläger, L. 2012. Zahlenspiegel der RWTH Aachen
2011. RWTH Aachen, Aachen.
[18] Lobbenmeier, D. 2003. Einstellungen von Schülerinnen
zum Informatikunterricht und zur Informatik. Eine
empirische Studie. Schriftliche Hausarbeit.
[19] Ministerium für Schule und Weiterbildung NRW.
Richtlinien und Lehrpläne für die Sekundarstufe II -
Gymnasium/Gesamtschule in Nordrhein-Westfalen
Informatik.
[20] Reher, J. 2010. Dossier über die Informatik in der
allgemeinbildenden Schule für die Bundesländer
Nordrhein-Westfalen und Niedersachsen. Dossier,
Universität Potsdam.
[21] Schulte, C. and Magenheim, J. 2005. Novices' expectations
and prior knowledge of software development. Results of a
Study with High School Students. In International Journal
of Environmental & Science Education, 143–153.
[22] Schwill, A. 2001. Fundamentale Ideen der Informatik.
http://ddi.cs.uni-potsdam.de/Forschung/Schriften/ZDM.pdf.
Accessed 21 February 2012.

The school experiment InTech – How to influence interest,
self-concept of ability in Informatics and vocational
orientation
ABSTRACT
Claudia Hildebrandt
University of Oldenburg
Computer Science Education
26111 Oldenburg, Germany
claudia.hildebrandt@uni-oldenburg.de
Engineers and professionals in Informatics are very important
nowadays but lacking in many European countries. The
school experiment InTech (Informatics with technical aspects)
wants to overcome this shortage. Therefore Informatics
was established as a regular subject in the grades 7 to 9
in 13 secondary schools in northern Germany. For 3 years
the teachers of theses classes met and developed studentand
context-oriented lessons in Informatics with technical
aspects to encourage technical thinking and practice and
influence students’ early career choices towards this direction.
This paper gives an overview of related research, the
organisation of the experiment and the research objectives
of our long-term study. We also describe first results which
show that this experiment influences interest, vocational orientation
and self-concept of ability in Informatics for most
students, especially for girls.
Keywords
Computer Science Education, interest of students in computer
science, self-concept of ability in Informatics, vocational
orientation, empirical study
1. INTRODUCTION
“Engineers are needed in this country.” This headline
marks an important turning point in discussions on educational
policy not only in Germany. And the demographic
change is going to increase the need for qualified trainees for
a lot of jobs in the fields of science, technology, engineering
and mathematics (STEM subjects) (see [21], 80). But in
most secondary schools in Germany technical thinking and
practice are not rated highly. Most students of grammar
schools (Gymnasium), which prepare for university, are not
introduced to technical subjects and procedures. As a consequence
too few students consider the technical field as a
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WiPSCE 2012 Hamburg, Germany
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$10.00.
34
Ira Diethelm
University of Oldenburg
Computer Science Education
26111 Oldenburg, Germany
ira.diethelm@uni-oldenburg.de
career option. In order to secure the country’s future economic
success it is important to increase the number of students
who want to take up a profession in computer science
or technology. The subject of Informatics is a good introduction
to a technical education in general. Therefore the
ministry of education encouraged some secondary schools in
northern Germany in joining a pilot study teaching Informatics
as a regular subject in grades 7 to 9, 2 to 3 lessons
a week. Some sponsors from the regional metal industry
support the project by providing additional teaching equipment.
The question is now in what ways teaching Informatics at
school can influence interest, the self-concept of ability in
Informatics and vocational orientation of students towards
a profession in computer science or technology.
We report on this pilot project with 13 schools in northern
Germany and describe implementation strategies and its
evaluation.
We begin by discussing the research results of related studies.
In section 3 we go on to describe the first and second
part of the school experiment InTech and the organisation
of the ongoing second InTech part. There we also present
the objectives of our project. Our concept is based on establishing
teams of Informatics teachers, so-called teacher sets.
Section 4 deals with research questions and hypotheses. Section
5 describes our research methods. Our first results will
be presented and explained in section 6. We close with a
final summary discussing important findings.
2. RELATED RESEARCH
2.1 Influence on vocational orientation
Vocational orientation can be defined as a process of approximation
and adjustment of young people to the world
of employment. On the one hand young people must get
to know their interests, wishes and competences in order
to orientate themselves. On the other hand there are the
requirements and needs of the world of employment and society
towards which young people need to be oriented (see
[1]).
According to Taskinen, Asseburg and Walter (see [21],
81f) the following characteristics of students can theoretically
influence the professional orientation of young people
with regard to the STEM-professions:
• scientific and mathematical literacy

• self-concept in science
• instrumental motivation
• interest in studying science
They analyzed within the framework of PISA 2006 to what
extent the mentioned characteristics are associated with professional
orientation. Analyzing questionnaires filled out by
about 7000 grade 9 students they came to primarily the following
conclusions:
36.6% of the young people answering the question on their
professional expectation, expect to pursue a STEM job at
age 30 (see [21], 90). In comparison with young people not
wanting to enter a STEM profession as adults these students
can be described as having a higher self-concept in
science, stronger instrumental motivation and greater interest
in studying science. The expected higher degree of scientific
and mathematical literacy, however, can only be found
in the partial sample of graduate professional expectation.
The girls’ self-concept in science is lower than the boys’.
Furthermore those girls with a graduate STEM professional
expectation are characterized by a lower mathematical and
scientific literacy as well as a lower interest in science than
the boys. These results correspond with the findings of numerous
other studies. In addition these girls also have a
lower scientific competence than the boys. The instrumental
motivation of girls with non-graduate STEM professional
expectation is significantly higher than that of the boys (see
[21], 95).
There are correlations between the investigated students’
characteristics (scientific and mathematical literacy, self-concept
in science, instrumental motivation and interest in studying
science).
In several analyses of STEM-professional groups the correlations
between the investigated students’ characteristics
differ according to the examined professional group. In some
professional groups a correlation between students’ characteristics
and their professional expectations could not be
observed. The results of the analyses show that the professional
expectations of the students can only partially be
explained by the theoretically deduced influencing factors of
the students’ choice of career.
Consequently, professional choice must be considered as
an interplay of numerous factors additionally influenced by
gender-specific mechanisms. The results are explained in
detail by Taskinen, Asseburg and Walter in ([21], 90-103).
According to them, instrumental motivation and interest
can, among other things, affect young people’s vocational
orientation. In order to positively influence these students’
characteristics teaching must be aware of the importance of
certain conditions of the learning environment.
2.2 Influence on students’ motivation and interest
in learning
Prenzel et al. [17] explore the question to what extent the
conditions of the learning environment, as perceived by the
trainees, can assist or interfere with the development of selfdetermined
motivation and interest. This question, along
with some others, is researched in a longitudinal study which
accompanied a small group of trainees (office administrators)
in their training. Influencing factors which according
to the state of research have a positive influence on the variants
of self-determined motivation are assigned to six sets of
theoretical conditions ([17], 111):
35
• Contextual relevance of the learning material (e.g. references
to applicability; relevance in real life; interdisciplinary
approach)
• Quality of instruction (e.g. precise structure; intelligibility)
• Communicated interest of the lecturer (e.g. commitment;
enthusiasm)
• Social integration (e.g. cooperative style of working;
friendly learning environment)
• Support of competence (e.g. direct, pertinent and
helpful feedback; individual reference standard)
• Support of autonomy (e.g. offer of alternatives; scope
to experiment; support in exploring, planning, acting,
learning independently)
These factors are also taken into account by Parchmann et
al. ([16], 17) in the ChiK project (see also below, section
2.4).
The trainees are questioned at intervals of about two months
on that particular phase of their training and on the
actual last day of that particular phase. The study gives an
account of the results of the first ten months (five measuring
points). There are 18 complete data sets (see [17], 111f).
The theoretically justified assumption that there are systematic
connections between the conditions relevant for motivation
and the variants of learning motivation are confirmed
by the findings (see [17], 124).
Take, for example, the variation of the motivation “interested”
1 . It indicates that the six supportive conditions are
positively correlated with this variation at the 0,01 level of
significance. The highest correlation exists with the contextual
relevance and the support of autonomy (the value of
the coefficient of correlation r = 0.62) (see [17], 118).
2.3 Influence on students’ plans to take CIT
Downes und Looker ([5], 179) explore the following factors
that influence the participation rate in computer and
information technology (CIT) subjects in senior secondary
schools:
• gender
• parental education
• access to and use of computers at home and at school
• self-perceived ability in nine different IT home-related
tasks
• self-perceived ability in CIT subjects in school
• attitudes towards CIT subjects and other school subjects
(Mathematics and English) and schooling in general
from which measures of “value” were constructed
1 Apart from the factual incentive, the personal and general
importance of the topic makes the student work hard to try
and understand. Interested learning also means that the
student wants to find out more about the topic and wants
to study it (see [17], 109f).

Data in this study are drawn from 11 NSW (state of Australia)
schools between 2005 and 2007. It includes 722 surveys.
The male students present higher levels of home-based
and school-related abilities than the females. Only for male
students there is a direct connection between home-related
confidence in one’s ability and plans to take CIT subjects
(see [5], 193).
Downes and Looker also find out that there are three key
factors which influence plans to take CIT subjects in senior
secondary years: gender, amount of use at school and the
value students place on CIT subjects (see [5], 194). The
latter two factors are inter-related either directly or indirectly
to home use and home- and school-related confidence
in one’s ability. Consequently the authors conclude that
“any school-based interventions that focus on increasing use
of IT at school, and increasing the ’value’ of CIT subjects,
need to also address increasing home use and self-perceived
skill levels in tasks associated with both home and school
use” ([5], 194f).
2.4 Results from ChiK
The project “Chemie im Kontext (short: ChiK)” is introduced
representatively for the context-based projects biology,
chemistry and physics. Contexts are chosen as starting
points and structuring elements of teaching units. These
contexts are either based on the environment of the learner
or are relevant through social references or later vocational
perspectives. The chemistry lessons are planned as practical
training and a means to advance the learners’ development
of competence (see [11] and [13]). Moreover, the
symbiotic implementation strategy is employed (see [10]).
So-called ”learning communities”, consisting of 8 - 12 teachers
from different schools in cooperation with scientists and
representatives of the education administration develop, test
and improve teaching units on the basis of a context-based
approach to teaching and learning chemistry (see [7], 54-
64). The learning communities are named “sets”. Sets 2 are
follow-up projects of sets 1 with low organisational changes.
Two fundamental questions of this research are ([7], 65):
• Do the students perceive changes in the teaching approach?
Do the students of the sets 1- and sets 2teachers
differ from each other in their perception of a
change in teaching approaches? In order to see the
teaching practice from the students’ point of view,
they were asked to describe their chemistry lessons
using four characteristic features of teaching: systematic
teaching of the subject, practical knowledge, crosscurricular
competences and self-controlled learning.
• Do motivation and interest develop equally in students
of sets 1 and sets 2?
The data collection for the sets 1 took place at the beginning
of the school year 2002/2003 (t0) and at the end of the
school years 2002/2003 (t1) and 2003/2004 (t2). The survey
of the sets 2 was carried out at the beginning of the school
year 2003/2004 and at the end of the school years 2003/2004
and 2004/2005. The students had to indicate their level of
agreement or disagreement for a series of statements. The
four-level answer scale ranged from “1” meaning “strongly
disagree” to “4” meaning “strongly agree” (see [7], 66-69).
The following results were obtained:
The students did perceive changes in their chemistry teaching
approach (see table 1). The increased perception of prac-
36
tical knowledge and self-controlled learning and at the same
time the reduction of the dominance of systematic learning
leads to the conclusion that students do notice the essential
criteria of the ChiK concept.
The motivation of the ChiK students in the chemistry
lessons shows a declining development in the course of the
second project year (see table 2). In another comparative
study it could be shown that the motivation of the ChiKstudents
declined significantly less than in classes which received
traditional chemistry teaching (see [7], 75). A positive
result is that in the two years under observation there
was no decline of interest (see table 2), whereas other scientific
studies reviewing subjects like physics or biology show
a continuous decline in interest from grade 5 to grade 9 (see
[22], 367). The development concerning sustainable interest
is not significant. Fussangel et al. describe the results in
([7], 72-76).
2.5 Summary of important results
The studies discussed show that vocational orientation
must be seen as an interplay of a lot of factors also influenced
by gender. Among other things interests and the selfconcept
play a crucial role. Todt, who analyzes the importance
of school for the development of childrens’ interests,
also maintains that interests tend to be an important condition
for the choice of profession or of academic study (see
[22], 373). Structures of interest are also closely connected
to children’s general estimations and orientations. Specific
and general interests are integrated into the individual selfconcept,
especially towards the end of adolescence, i.e. the
end of school education. There are a lot of indications, however,
that interests can develop during the whole span of
life, provided they do not contradict the self-image and the
activities involved can be mastered and lead to some satisfaction
(see [22], 375). Certain conditions of teaching practices
noticed by the students can influence the students’ characteristics
”self-determined motivation” and ”interest”. These
characteristics are interrelated with each other. On the basis
of our research project InTech the teaching characteristics
student- and context-orientation as well as the student
characteristics interest, self-concept of ability in Informatics
and vocational orientation will be explored. In the following
chapters we give an overview of the history and the organisation
of our experiment, the evaluating research and finally
the first results.
3. THE INTECH EXPERIMENT
3.1 Description of part I
For three school years, from 2005 to 2008, six secondary
schools in Lower Saxony, Germany, tested ways of teaching
Informatics in grades 7 to 9. Lessons in participating
classes were held according to “Time Schedule I”, which allocated
three hours per week in grade 7 and four hours in
grades 8 and 9. Different organisational concepts and teaching
units were developed, e.g. some schools cooperated with
the subjects of arts and economy and others cooperated with
astronomy or physics. All participating teachers met at regular
intervals to exchange experiences, at least 4 times a
year. Due to the fact that all schools decided to use robots
in their classes and most schools in Germany cannot afford
to buy 8 or more sets of robots at the same time, the project
was sponsored by the “Stiftung NiedersachsenMetall”.

mean at t0 mean at t2 mean at t0 mean at t2
(sets 1, N = 549) (sets 1)
(sets 2, N = 237) (sets 2)
systematic
subject
teaching of the 3.28 3.16 3.09 2.90
practical knowledge 2.37 2.45 2.33 2.50
cross-curricular competences 2.42 2.20 2.33 2.50
self-controlled learning 2.29 2.63 2.10 2.48
Table 1: Description of teaching approach from a student-oriented point of view (see [7], 73)
mean at t0 mean at t2 mean at t0 mean at t2
(sets 1, N = 59) (sets 1)
(sets 2, N = 238) (sets 2)
motivation 3.08 2.81 3.00 2.74
sustainable interest 1.98 2.04 2.09 2.08
Table 2: Perception of motivation and interest from a student-oriented point of view (see [7], 74)
Although there was no scientific evaluation during this
first phase some results can yet be reported by Modrow and
Reineke (see [14]):
• At all participating schools the offer was so much in
demand that selection procedures had to be introduced
in the following years.
• Boys and girls were attracted to the offer in equal numbers.
• Motivation exceeded that of “normal” compulsory lessons.
• Teachers were positively surprised at the students’ performances.
• At some schools, groups could be motivated to participate
in national and even international competitions
for robotics. Two students from Intech even became
world champions.
• All schools developed teaching units involving robots
(mostly using LEGO-Mindstorms). Boys and girls were
equally enthusiastic about constructing and programming
these.
• All lessons were action/activity-oriented.
• Teamwork was encouraged by the lessons’ project-oriented
layout.
• The participating teachers found it easy to introduce
interdisciplinary questions and problems in Informatics
classes.
3.2 InTech - part II
In order to use and expand the experiences attained and
to enable others to benefit from them as well, technicaloriented
Informatics classes will be introduced or continued
to be developed at 23 additional schools for three more
years, beginning in grade 7. 13 of these schools are coordinated
by us from the Carl von Ossietzky Universität Oldenburg
and financed by “Arbeitgeberverband Nordmetall”
and the “Stiftung des Verbands der Metall- und Elektroindustrie”.
The other 10 schools are once more supported by
the “Stiftung NiedersachsenMetall” and coordinated by colleagues
from the Georg-August-Universität Göttingen.
37
Schools wanting to participate must possess facilities that
enable them to teach Informatics from the 7th grade onwards.
A further requirement is the presence of two appropriately
qualified teachers.
The central concern of InTech in relation to the students
is the following: Beginning with the 7th grade, the contextand
student-oriented lessons in Informatics with technical
aspect should encourage technical thinking and practice as
well as introduce students to the functional principles and
utilization of information and information systems. By doing
so we want to support students in choosing a career in
the Informatics business.
In order to ensure sustainability by the distribution of the
results, we pursue the following arrangements regarding the
teachers:
• to establish a network of Informatics teachers,
• to develop, test and upgrade student- and contextoriented
teaching material,
• to make the teaching material available via the Internet
in order to be used and developed further and
• to distribute the teaching concepts by further professional
training.
Finally, one goal of this second phase is also to conduct
new research and to develop and expand further education
opportunities for teachers.
3.3 The concept of InTech
The described goals will be realized by using a “symbiotic
strategy of implementation” as explained by Frey et al. (see
[6], 242-244) on the basis of the guideline“Technical thinking
and action in meaningful contexts”, developed by our team.
It contains the following strategy: Informatics teachers and
researchers must establish a “working team” (so-called set of
teachers) which gets together on a regular basis, participates
in different training courses and discusses, develops and realizes
technical-based teaching units co-operatively. The set
of teachers consists of about 24 teachers from 13 schools who
participate in the project InTech (see [3], 97).
The involvement with technical-related contents and working
methods from different fields leads to the development
of project-oriented and activity-based lessons. Options for

interdisciplinary lessons may be worked out. After testing
the developed teaching material it will be improved cooperatively,
if necessary. Afterwards, the teachers upload
their material on the InTech-Portal.
It is important to consider that the success of the implementation
process depends on numerous factors, such
as the teachers’ attitudes, beliefs and skills as well as the
co-operation and communication culture at the schools and
the quality of the learning communities (see [10], 208-209).
These aspects and the special conditions of Informatics teachers
are taken into consideration in order to develop and distribute
innovative Informatics teaching concepts. These conditions
and relevant questions are presented in [3].
3.4 Teaching approach
In order to develop context-based and student-oriented
lessons, Diethelm et al. offer a framework (see ([4], 103f))
and an exemplary unit (see [2]). The regular meetings of the
teachers and researchers as well as the continuing work and
the exchange of experiences and teaching units should support
the teachers to change their teaching towards contextand
student-orientation. But the way of teaching at school
is determined by each teacher and could be very different.
4. RESEARCH QUESTIONS
The central intention of the InTech-project is to inspire
and increase the interest in Informatics and engineering. By
doing so we want to encourage students to choose a career
in this field. In order to judge the success of InTech, our
research has to answer the following questions:
1. In which ways did the teaching of Informatics change
during the first year of the project from the students’
point of view?
We think that the co-operation in learning communities
(the sets of teachers), different training courses
in the framework of InTech, and the development of
teaching material improve the teaching competence of
the participating teachers and we expect the students
to observe these changes. It will be analyzed if the students
notice this student-oriented and context-oriented
approach.
2. In which ways do interest in Informatics, self-concept
of ability and vocational orientation towards Informatics
and technology change in the course of this first
year?
Keeping the results of ChiK (see section 2.4) and [22]
in mind, we expect that the interest in Informatics
will either be consistent or decrease a little. If the
boys, as may be expected, overestimate their competence
in Informatics, their self-perceived ability should
decrease during the first project year. Since girls generally
have a less strongly developed self-concept in
science than the boys (see [21]), their self-assessment
should increase between the two measuring points. If
the teaching of Informatics is considered to be more
student- and context-oriented the average value on the
scale of vocational orientation towards Informatics and
technology should rise.
The following research questions concern the correlation of
the different characteristics.
38
3. To what extent do the teaching characteristics studentand
context-orientation influence vocational orientation,
interest and the self-concept of ability in Informatics
and technology during the first project year?
Based on ChiK (see section 2.4) and the study on the
office manager trainees (see section 2.2) we will analyze
the correlations between the teaching characteristics
student- and context-orientation and the student
characteristics interest in Informatics, self-concept of
ability in Informatics and technology and vocational
orientation. We expect that the teaching of Informatics
will have a positive influence on vocational orientation,
interest and self-concept.
4. To what extent do vocational orientation, interest and
self-concept of ability of the first and second surveys
correlate?
The studies discussed in 2 show that interests play a
crucial role in the choice of professional or academic careers
and are closely connected to general estimations
and orientations (see also [22]). We therefore examine
the correlation of the three student characteristics.
As described in ([21], 101) vocational orientation can be seen
as an interplay of many factors, additionally influenced by
gender. It must be kept in mind that our results cannot
do more than describe certain tendencies, since the samples
are too small and only those students participated who were
members of the InTech project. It can therefore be expected
that they already have a basic interest in Informatics, having
chosen the subject voluntarily.
5. RESEARCH METHODS
5.1 Timing and planning of the survey
To find answers to the questions above we decided to use
online-questionnaires at the beginning of the school year
2009/2010 and again at the end of each school year for the
next three years. Due to the fact that most of the teachers
don’t teach a learning group for the whole three years and
there also are fluctuations in the learning groups, especially
in the sets, we decided to additionally question the students
at the beginning of the last project year (see table 3) in order
to notice time-related changes during the last year of
the project. We developed questionnaires for teachers and
students based on the concept of the ChiK-project 2 .
To adapt the concept to our needs, we added some items
accommodating special requirements of our project.
5.2 Participants and rate of response
Initially, 24 teachers (four female teachers and 20 male
teachers) and about 300 students from 13 different schools
participated in the project. About 70% of the participants
in the students’ survey are male, most of them were 12 to
16 years old and were grammar school students (Gymnasiasten).
Only a few of the participants came from general
secondary schools (Realschule, Hauptschule, Gesamtschule).
2 Note that the questionnaire of the ChiK-concept is also
including parts of another project – the SINUS-project.
Therefore, further information can be found in Ostermeier
[15] and in unpublished scale handbooks about teachers and
students of C. Gräsel, K. Fussangel and J. Schellenbach-Zell.

survey at the
time t0
time of survey beginning of
the school year
2009/2010
survey at the
time t1
end of the school
year 2009/2010
survey at the
time t2
end of the school
year 2010/2011
survey at the
time t3
beginning of
the school year
2011/2012
survey at the
time t4
end of the school
year 2011/2012
number of teachers 24 19 19 - in process
number of students 215 221 160 235 in process
Table 3: Overview of the times and the numbers of participants of the surveys
Table 3 provides an overview of the times as well as the various
numbers of participants in the surveys.
59 students, who are involved in the InTech-project, participated
in the first two surveys. This was identifiable by
means of an individual code. In order to observe the development
during the time of the project, the study of the
research questions (see section 6) is based on the data sets
of these 59 students. Of these 59 participants 45 were boys
and 12 were girls (two students gave contradicting information
about gender). Most of them were 13 and 14 years old
at the time of survey t0 and 14 and 15 years at the time of
survey t1, and attended the 8th grade, about 85% of them
at a grammar school (Gymnasium).
5.3 Definitions of constructs with regards to
the students
In order to gain information about the students’ perception
of the teaching approach we use the following constructs:
Student-orientation: The construct means that the choice
of goals, the content settings and choice of methods
have to be adjusted to the students’ needs. It includes
the criterion “support of autonomy” from [17] (see section
2.2).
Context-orientation: This construct represents the idea
that the given contexts in the lessons are oriented at
real-world contexts which are of interest to the students.
It corresponds with the criterion “contextual
relevance” from [17] (see section 2.2).
The following student-oriented constructs were also used in
the questionnaires:
Interest in Informatics: This construct includes aspects
like enjoying Informatics-related topics and having an
interest in gaining more knowledge in Informatics (for
natural sciences – in the PISA studies 2006 – see ([20],
128); for specific analyses see [18]).
Self-concept of ability in Informatics: The construct includes
self-awareness of competence.
Vocational orientation: This construct reflects the motivation
to learn a specific job in the field of Informatics.
An overview of the constructs, their reliability and exemplary
items can be found in table 4.
5.4 Instruments
The examination of the constructs (see above) takes place
via items. Most of the items are answered on a scale system,
ranging from “disagreement” to “agreement”, where “1”
39
means “I strongly disagree” and “6” means “I strongly agree”
(Likert scale).
A certain number of items form a construct. Using Cronbach’s
Alpha-coefficient we can see how far a group of test
items can be used as measurement of individual variables
(here: constructs). The Cronbach’s Alpha value should lie
between 0.7 and 1. The Cronbach’s Alpha value of the construct
“student-orientation” lies slightly under this value at
the time t0 and at the time t1 the rounded-off value is 0.7
exactly. Therefore we include this construct into the analysis.
In table 4 the constructs, exemplary items and the
Cronbach’s Alpha value can be found.
Answers on a Likert scale are typically ordinally-scaled.
Because the answers on our Likert scale are equidistantly
displayed, test participants should recognize the various possible
answers as being equidistant. Consequently, for the
analysis we use the Likert scale as an interval scale.
6. FIRST RESULTS
The first research question deals with the changes in
the Informatics lessons based on the perception of the students.
The constructs “student”- and “context-orientation”
will be analyzed separately, according to their gender, at
the times of survey t0 and t1. The scale mean values and
standard deviations of these constructs are shown in table
5. The scale mean values of the construct “student-orientation”
show a substantial increase between the beginning and
the end of the first project year independently of gender.
The scale mean values of the construct“context-orientation”,
however, have only risen slightly.
The scale mean difference of the construct “student-orientation”
is statistically significant only with regard to the
boys. The girls’ scale mean difference of this construct is
not statistically significant, the reason possibly being that
too little data could be collected.
It can be seen from table 5 that the scale mean values of
both constructs describing the Informatics lessons increase.
Therefore it can be assumed that the students’ impression is
that their teachers changed to a more student- and contextoriented
approach towards the end of the first project year.
This student perception agrees with the InTech-teachers’
own assessment of their teaching. The results of the teachers’
surveys show that the 18 teachers who participated in
the first two questionnaires assess their teaching as being
more student- and context-oriented and as using more variety
of teaching and learning methods towards the end of
the first year than at the beginning of the project (studentorientation:
scale mean value at the time of survey t0: 3.40,
at the time of survey t1: 3.58; context-orientation: scale
mean value at the time of survey t0: 3.46, at the time of
survey t1: 3.69; variety of teaching and learning methods:
scale mean value at the time of survey t0: 3.90, at the time

characteristic number
of
items
exemplary items reliability
(Cronbach’s
Alpha)
student-orientation 5 In the Informatics lessons we get the scope to set
ourselves tasks and goals.
context-orientation 5 In the Informatics lessons we are confronted with
topics which we also like dealing with in our spare
time.
interest in Informatics 9 Informatics lessons are interesting to me because
of the topics.
self-concept of ability in Infor- 11 I believe that I am able to solve new problems in
matics
Informatics.
vocational orientation 4 I attend Informatics lessons because it establishes
a basis for my professional choice.
characteristics mean m
(regarding the
data
boys)
of the
student-orientation t0: 3.60
t1: 4.03
context-orientation t0: 3.23
t1: 3.34
Table 4: Registration of characteristics of students
difference mean m
m1-m0 (regarding the
data
girls)
of the
+0.43 t0: 3.40
t1: 3.98
+0.11 t0: 2.73
t1: 2.83
t0
difference StdDev
m1-m0 (regarding the
data
boys)
of the
+0.58 t0: 0.92
t1: 0.88
+0.10 t0: 1.30
t1: 1.25
reliability
(Cronbach’s
Alpha)
t1
0.641 0.699
0.830 0.825
0.860 0.862
0.911 0.926
0.844 0.844
StdDev
(regarding the
data of the
girls)
t0: 1.09
t1: 1.15
t0: 1.05
t1: 1.14
Table 5: Means and standard deviations regarding students’ perception of Informatics lessons
of survey t1: 4.07; the mean value differences are not statistically
significant). These tendencies could be attributed
to the constructive work in the InTech-project, because the
teachers profited from the teaching materials and the experience
of colleagues, and it can be taken to indicate a success
of the implementation strategies of InTech.
Looking at the construct “context-orientation” it stands
out that the scale mean value of the boys is much higher than
that of the girls. This mean value difference is, however,
not statistically significant. The reason for this tendency
could be the girls’ lower interest in Informatics (see table
6). Therefore the girls become less aware of the contextorientation
than the boys.
For the second research question we examine the constructs
”interest in Informatics”, ”self-concept of ability in
Informatics” and ”vocational orientation”. The differences
between the scale mean values at time t1 and time t0 and
the standard deviations of the constructs are calculated (see
table 6). Only the mean difference of the construct “interest
in Informatics”regarding the result of the girls is statistically
significant (based on the significance level of 5 %).
As expected the interest in Informatics of the boys decreases
a little (see section 4). The self-concept of ability,
however, increases. This may mean that the teachers encouraged
the students to embrace technical thinking and
practice as well as introduced them to the functional principles
and utilisation of information and information systems.
The boys’ enthusiasm for careers in the Informatics business
could not be increased. But the scale mean value of “vocational
orientation towards Informatics” is already high with
a value of 4. Moreover the table 6 shows that the scale mean
values of the constructs regarding the girls increase. In con-
40
trast to the development in other natural-science subjects
we find that there is a statistically significant increase of the
scale mean value of the construct “interest in Informatics”
during the first project year. Consequently the Informatics
lessons represent an important instrument to make informatical
and technical aspects accessible to the girls. They were
able to find an interest in Informatics, increase their selfconcept
of ability and their vocational orientation regarding
Informatics.
Looking at table 6 we also recognize that girls show much
less interest in technical professions and in Informatics. Their
“self-concept of ability in Informatics” turns out to be lower
than that of the boys. Comparing the three student characteristics
the scale mean differences between the boys and
girls are statistically significant at the time of survey t0 (level
of significance 5 %). We notice comparable results for the
scientific concepts of competence of the girls. There are indications
that the scientific concept of competence is less
developed in girls than in boys, see [12] and [19]. Possible
reasons regarding the natural sciences are discussed in ([20],
126f) and may be applied to Informatics.
At the time of survey t1 the observed scale mean differences
are not statistically significant any more (level of
significance 5 %). This may indicate that the differences
between the boys and girls decrease in the course of the Informatics
lessons, as far as the students’ characteristics are
concerned which would be a good sign.
The third research question is concerned with the correlation
between the constructs “student-orientation” and
“context-orientation”on the one hand and the constructs“interest
in Informatics”, “self-concept of ability in Informatics”
and “vocational orientation” on the other hand (see table 7

characteristics mean m
(regarding the
data
boys)
of the
interest in Informatics t0: 3.79
t1: 3.65
self-concept of ability in In- t0: 3.82
formatics
t1: 4.10
vocational orientation t0: 4.07
t1: 4.00
difference mean m
m1-m0 (regarding the
data
girls)
of the
-0.14 t0: 2.92
t1: 3.36
+0.28 t0: 3.11
t1: 3.35
-0.07 t0: 2.95
t1: 3.44
difference StdDev
m1-m0 (regarding the
data
boys)
of the
+0.44 t0: 0.80
t1: 1.08
+0.24 t0: 1.08
t1: 0.97
+0.49 t0: 1.17
t1: 1.24
Table 6: Scale mean values and standard deviations regarding student characteristics
and 8). The correlation coefficients are Spearman’s Rho coefficients
and provide a measure of the relationship between
two sets of data.
The correlation tables 7 and 8 show that apart from one
correlation between “context-orientation” and “interest” for
the boys there are no statistically significant correlations
to be found at the time of the first survey, the explanation
being that hardly any of the students had any teaching
in Informatics at the time of the survey and therefore
could not give an assessment. This is supported by free text
comments of some students. At the time of survey t1, all
students characteristics correlate with the teaching characteristic
“student-orientation”. Gender differences are to be
noticed only in the level of the correlation coefficients and
the level of significance. The correlations between the surveyed
characteristics are stronger regarding the data of the
girls. It must be taken into account, however, that there are
only 12 data sets of girls as opposed to 45 of boys. 12 data
sets do not allow any general statements.
studentorientationcontextorientation
interest in
Informatics
t0: 0.306
t1: 0.673*
t0: 0.439**
t1: 0.298**
selfconcept
of ability
t0: 0.297
t1: 0.543*
t0: 0.229
t1: 0.290
vocational
orientation
t0: - 0.145
t1: 0.347**
t0: 0.162
t1: 0.128
Table 7: Correlations regarding student characteristics
(*p < 0.01, **p < 0.05, other correlations are
non-significant). Only the data of the boys are taken
into account.
studentorientationcontextorientation
interest in
Informatics
t0: 0.500
t1: 0.773*
t0: -0.008
t1: 0.534
self-concept
of ability
t0: 0.172
t1: 0.874*
t0: -0.239
t1: 0.391
vocational
orientation
t0: 0.305
t1: 0.577**
t0: -0.136
t1: 0.330
Table 8: Correlations regarding student characteristics
(*p < 0.01, **p < 0.05, other correlations are
non-significant). Only the data of the girls are taken
into account.
Contrary to our expectations, only the boys’ data present
a low correlation between “context-orientation” and “interest
in Informatics”. Consequently these results show that
“context-orientation” has only little or no influence on the
41
StdDev
(regarding the
data of the
girls)
t0: 0.84
t1: 1.20
t0: 1.07
t1: 1.23
t0: 1.20
t1: 1.20
students’ characteristics. This outcome contradicts the results
of [17] (see section 2.2). Further investigation must
show if these findings will be confirmed.
Our next research question referred to the correlations
between the ascertained student characteristics. According
to analyses in natural sciences some correlations were to be
expected (see section 2.1). The correlations of the students’
characteristics are shown in table 9 and 10, differentiated by
gender. The correlations are calculated by means of the correlation
measure Spearmans Rho, which allows to calculate
the relationship of ordinal scaled data.
interest in
Informatics
self-concept
of ability
vocational
orientation
interest in
Informatics
1.00
t0: 0.204
t1: 0.534*
t0: 0.537*
t1: 0.599*
self-concept
of ability
1.00
t0: 0.365**
t1: 0.230
vocational
orientation
1.00
Table 9: Correlations regarding student characteristics
(*p < 0.01, **p < 0.05, other correlations are
non-significant). Only the data of the boys are taken
into account.
interest in
Informatics
self-concept
of ability
vocational
orientation
interest in
Informatics
1.00
t0: 0.763*
t1: 0.792*
t0: -0.023
t1: 0.332
self-concept
of ability
1.00
t0: 0.156
t1: 0.767*
vocational
orientation
1.00
Table 10: Correlations regarding student characteristics
(*p < 0.01, other correlations are nonsignificant).
Only the data of the girls are taken
into account.
Table 9 and 10 present that for the boys there are moderate
correlations between “interest in Informatics” and “vocational
orientation” (regardless of whether or not Informatics
lessons take place) as well as a moderate correlation between
“interest in Informatics” and “self-concept of ability in Informatics”
(after one year of Informatics lessons) and a low
correlation between “self-concept of ability in Informatics”
and “vocational orientation” (only at the beginning of the
project). For the girls there are strong correlations between

“self-concept of ability in Informatics” and “vocational orientation”
(after one year of Informatics lessons) and between
“self-concept of ability in Informatics” and “interest in Informatics”
(regardless of whether or not Informatics lessons
take place).
Contrary to our assumptions no correlations can be determined
mathematically for the girls between “interest in
Informatics” and “vocational orientation” . We’ll have to
wait and see whether this phenomenon can be confirmed in
the surveys at the beginning and the end of the last project
year. If the answer is positive we’ll have to search for the
reasons.
A strong correlation for the girls and a moderate correlation
for the boys between “interest in Informatics” and
“self-concept of ability in Informatics” exists at the level of
significance of p < 1%. The results demonstrate that the
“interest in Informatics” plays a crucial role. In addition the
“self-concept of ability in Informatics” is important for the
girls and correlates strongly with the “vocational orientation”.
Figure 1: Statistically significant correlations between
the examined characteristics after the first
project year
The statistically significant correlations between the examined
characteristics after the first project year are presented
in figure 1. The continuous lines in this figure show
the relationships between the characteristics regarding all
students, the dashed lines only regarding the boys and the
dotted lines display the results regarding only the girls.
The correlation results can not give answers to the question
of whether variable A influences variable B or vice versa.
Therefore further research is required, for example observation
of the teaching-learning situations from the perspectives
of teachers and learners. For the verification of the positive
outcomes of the InTech-project we have to compare classes
that have Informatics lessons with classes that do not.
7. CONCLUSIONS
We can summarize the following results after the first
project year:
• The Informatics lessons are perceived to be more student-oriented.
Consequently the competence of the
teachers and thereby the quality of the Informatics
lessons seem to increase.
• “Student-orientation” correlates positively and significantly
with the “interest in Informatics”, the “selfconcept
of ability in Informatics” and the “vocational
orientation regarding Informatics”after the first project
year. These and further correlations are presented in
figure 1.
42
• Regarding the girls, the “interest in Informatics”, the
“vocational orientation regarding Informatics” and the
“self-concept of ability in Informatics” increase.
It is surprising and of course encouraging that the girls’
“interest in Informatics” significantly increased contrary to
the trend in science subjects like chemistry. Consequently
they should get the opportunity to learn Informatics in the
middle grades to enable them to discover possible interests
in technical fields.
According to the findings of Downes and Looker (2011,
see [5], 191) home use and home-based ability belief regarding
Informatics are indirectly related to the variable to take
up CIT subjects (the second factor for the boys directly).
Therefore the acquisition of knowledge about the functional
principles and utilisation of information systems should not
depend on coincidences such as social or regional origin, but
should instead be part of general education. The subject
of Informatics is well-suited as an introduction to a general
technical education, as the required equipment and facilities
are already in place at schools. With the help of the published
experiences of the InTech-schools explaining how to
introduce the subject Informatics and offering the existing
teaching material, more schools can start offering the subject
Informatics in the middle grades. And thus the number of
highly qualified graduates in technical professions and study
courses could increase. So far we can call the implementation
of InTech a success.
8. ACKNOWLEDGMENTS
The realization of the project InTech was only possible
with the support of many different people. At this point we
would like to thank Eckart Modrow (Georg-August-Universität
Göttingen), Vera Reineke, Astrid Tengen and Gudrun
Köppen-Castrop (Niedersächsisches Kultusministerium), Peter
Golinski (Arbeitgeberverband NORDMETALL), Natascha
Clasen and Sabine Stöhr (VME-Stiftung Osnabrück-
Emsland), Andreas Breiter (ifib - Institut für Informationsmanagement
Bremen GmbH), Ilka Parchmann (IPN - Leibnitz-Institut
für die Pädagogik der Naturwissenschaften und
Mathematik in Kiel), and last but not least all participating
teachers and their schools.
9. REFERENCES
[1] Bundesinstitut für Berufsbildung (Hrsg.).
Berufsorientierung. Version: 2012. demand:
30.04.2012. http://www.good-practice.de/3337.php.
[2] I. Diethelm. Wie forschend-entdeckendes Lernen
gelingen kann - Forschendes und entdeckendes Lernen
im Kontext zum Thema Datenschutz, Internet und
Urheberrecht. LOG IN, (168):28–34, 2011.
[3] I. Diethelm, C. Hildebrandt, and L. Krekeler.
Implementation of computer science in context - a
research perspective regarding teacher-training. 9th
Koli Calling International Conference on Computing
Education Research, Finland, pages 96–99, November
2009.
[4] I. Diethelm, J. Koubek, and H. Witten. IniK -
Informatik im Kontext - Entwicklungen, Merkmale
und Perspektiven. LOG IN, (169/170):97–105, 2011.
[5] T. Downes and D. Looker. Factors that influence
students’ plans to take computing and information

technology subjects in senior secondary schools.
Comuter Science Education, 21(2):175–199, June 2011.
[6] A. Frey, C. Gräsel, T. Puhl, and I. Parchmann.
Implementation einer kontextorientierten
Unterrichtskonzeption für den Chemieunterricht.
Unterrichtswissenschaft, 32(3):238–256, 2004.
[7] F. Fussangel, J. Schellenbach-Zell, and C. Gräsel. Die
Verbreitung von Chemie im Kontext: Entwicklung der
symbiotischen Implementationsstrategie. In Chemie
im Kontext. Von der Innovation zur nachhaltigen
Verbreitung eines Unterrichtskonzepts, pages 49–82.
Waxmann, 2008.
[8] C. Gräsel, K. Fussangel, and J. Schellenbach-Zell.
Skalenhandbuch Lehrer/-innen. (unpublished).
[9] C. Gräsel, K. Fussangel, and J. Schellenbach-Zell.
Skalenhandbuch Schüler/-innen. (unpublished).
[10] C. Gräsel and I. Parchmann.
Implementationsforschung - oder: der steinige Weg,
Unterricht zu verändern. Unterrichtswissenschaft,
32(3):196–214, 2004.
[11] C. Hössle, I. Parchmann, and M. Komorek.
Naturwissenschaften im Kontext. demand: 30.04.2012.
http://www.mathematik.uni-dortmund.de/ieem/
BzMU/BzMU2009/Beitraege/Hauptvortraege/
HOESSLE PARCHMANN KOMOREK 2009
Naturwissenschaften.pdf.
[12] P. Häussler and L. Hoffmann. A curricular frame for
physics education: Development, comparison with
students’ interests, and impact on students’
achievement and self-concept. Science Education,
84:689 – 705, 2000.
[13] I. Parchmann, C. Gräsel, A. Baer, P. Nentwig, R.
Demuth, B. Ralle, and the ChiK Project Group.
”Chemie im Kontext”: A symbiotic implementation of
a context-based teaching and learning approach.
International Jounal of Science Education, 28(9):1041
– 1062, 2006.
[14] E. Modrow and V. Reineke. Der Modellversuch
Intech08 - Informatikunterricht mit technischen
Aspekten. information brochure, Stiftung
NiedersachsenMetall, 2008. http://www.stiftungniedersachsenmetall.de/docs/SNM
intech web.pdf.
[15] C. Ostermeier. Kooperative Qualitätsentwicklung in
Schulnetzwerken. Eine empirische Studie am Beispiel
des BLK-Modellversuchsprogramms ”Steigerung der
Effizienz des mathematisch-naturwissenschaftlichen
Unterrichts” (SINUS). Waxmann, Münster, 2004.
[16] I. Parchmann, B. Ralle, and D.-S. D. Fuccia.
Entwicklung und Struktur der Unterrichtskonzeption.
Chemie im Kontex. In Chemie im Kontext. Von der
Innovation zur nachhaltigen Verbreitung eines
Unterrichtskonzepts, pages 9–47. Waxmann, 2008.
[17] M. Prenzel, A. Kristen, R. Dengler, P. Ettle, and
T. Beer. Selbstbestimmt motiviertes und interessiertes
Lernen in der kaufmännischen Erstausbildung. In
Lehr-Lernprozesse in der kaufmännischen
Erstausbildung. Wissenserwerb,
Motivierungsgeschehen und Handlungskompetenzen,
pages 108–127. Steiner, 1996.
[18] M. Prenzel, K. Schütte, and O. Walter. Interesse an
den Naturwissenschaften. In PISA 06. Die Ergebnisse
der dritten internationalen Vergleichsstudie, pages
43
107–124. Waxmann, 2007.
[19] S. Reis and S. Park. Gender differences in
high-achieving students in math and science. Journal
for the Education of the Gifted, 25:52 – 73, 2001.
[20] K. Schütte, A. C. Frenzel, R. Asseburg, and
R. Pekrun. Schülermerkmale, naturwissenschaftliche
Kompetenz und Berufserwartung. In PISA 06. Die
Ergebnisse der dritten internationalen
Vergleichsstudie, pages 125–146. Waxmann, 2007.
[21] P. Taskinen, R. Asseburg, and O. Walter. Wer möchte
später einen naturwissenschaftsbezogenen oder
technischen Beruf ergreifen? - Kompetenzen,
Selbstkonzept und Motivationen als Prädiktoren der
Berufserwartungen in PISA 2006. In: Prenzel, M,
Baumert (Hrsg.): Vertiefende Analysen zu PISA
2006. Zeitschrift für Erziehungswissenschaft,
Sonderheft, pages 79–105, October 2008.
[22] E. Todt. Die Bedeutung der Schule für die
Entwicklung der Interessen von Kindern und
Jugendlichen. Unterrichtswissenschaft, (4):362–376,
1985.

Uncovering Structure behind Function – the experiment as
teaching method in computer science education
ABSTRACT
There are lots of reports about activities that aim at fostering and
maintaining interest in computing. These activities rely on different
ideas that should help to involve novices and young learners.
In this article examples are given and explained from a disciplinespecific
perspective based on the notion of a dual nature of digital
artifacts, which has to be reconstructed and integrated during the
process of computer science education. The claim is that learners
either focus on the internal computational properties: the structure,
or on the more external reasons to use them: the function –
but rarely are able to develop an integrated perspective on both
sides. Therefore learning experiences should be designed to
bridge these perspectives and enable to perceive the dual nature of
digital artifacts. This idea of bridges between structure and function
is used to design and analyze learning activities with regard
to their potential in fostering and maintaining interest in computer
science, especially to those not already interested. This article
presents and discusses three examples for such experiments.
Based on this, guidelines for experiments as teaching method, and
questions for further research are derived.
Categories and Subject Descriptors
K3.2 [Computers & Education]: Computer and Information
Science Education – computer science education, information
systems education.
General Terms
Experimentation, Human Factors.
Keywords
Secondary CS Education, K-12, Didactics, CS Ed Research,
Pedagogy, Gender, Teaching Methods, Dual Nature, Duality
Reconstruction, Experiment
1. INTRODUCTION
Teaching computer science (CS) often faces problems like low
enrollments, low retention, and low ratio of female computer
science students. As these are urgent concerns, a lot of – often
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
Conference’12, Month 1–2, 2010, City, State, Country.
Copyright 2010 ACM 1-58113-000-0/00/0010…$10.00.
Carsten Schulte
Freie Universität Berlin
Computing Education Research
Königin-Luise Str. 24
14195 Berlin Germany
schulte@inf.fu-berlin.de
44
locally successful – initiatives and activities take place, but still, in
general the problems remain unsolved.
Here only a fraction of such activities can be listed (and of course
sometimes there are overlaps between the different variants):
CS for fun / magical computer science, CS unplugged [6, 7,
9]: These activities are introducing motivational examples to
foster interest and motivation without the use of computers,
in order to refocus attention from the ‘technical or usage
skills’ to the ‘thinking skills’ involved and useful in everyday
life.
Summer Schools, working groups, and weekend activities:
Their aim is to enable a gentle introduction without a steep
learning curve, without grades, and to help getting in contact
with peers (e.g. courses designed specifically for female novices),
in order to lower the entry barriers to computer science.
Competitions in different organizational forms, local, national
or international, or related to specific tools like pedagogical
IDE’s, robotics or others (see [5]): Competitions foster
high engagement in the activities, and provide feedback from
the engagement itself, but also from comparing results to
others.
Digital communities, like e.g. Scratch [22], build around a
tool/topic: In general, they are aiming at similar aspects like
the above mentioned summer schools, but provide more ongoing
support.
Innovation in the computer science classroom: This is often
a mix of the above mentioned activities, but also specific
ideas and interventions to foster interest. Like e.g. media
computing [13], or computer science in context [29].
Overall, there are lots of activities reported, that aim at fostering
and maintaining interest in the computing disciplines (see e.g.. [1,
10, 35, 36] for a comparison and evaluation of such types of
activities). These activities rely on different ideas that should help
to engage novices and young learners. While these ideas often
denote theories from other domains—like psychology with regard
to motivation or attribution—in this article, a simple idea is presented
which can be used to analyze and describe such activities
and predict effects on learners with different attitudes, and – as the
given examples should demonstrate – is powerful enough to drive
the development of new activities.
2. THE SITUATION IN CS ED
The idea of introducing experiments draws on prevalent topics
debated in CS Ed (section 2.1 ). Although a myriad of activities

aim to react on these topics, experiments focus on a new explanation
of these issues, which is outlined in section 2.2 .
2.1 Misconceptions of Computing
The situation in CS education is often characterized by limited
interest in computing, due to e.g. misconceptions, attitudes, gender-bias,
and similar issues [14, 20, 24, 28, 33, 35].
For CS unplugged activities are some evaluative reports available:
Feaster et. al. ([10], p.252) conclude that “the program had no
statistically significant impact on student attitudes toward computer
science or perceived content understanding.” Taub et.al.
[35] yield similar results (p.24) and found the following explanation
(similar to [36]): “only some of the objectives were addressed
in the activities, […] the activities do not engage with the students’
prior knowledge and […] most of the activities are not
explicitly linked to central concepts in CS” (p.1) and thus modifications
to the activities are needed (A proposal for such modifications
can be found in [2]).
Basically, programs to increase interest (see section 1. ) aim at
demonstrating a ‘true’ image of the discipline, showing that it is
accessible for outsiders or minorities, motivating by highlighting
(different) career choices, and so on. In essence, these activities
are trying to minimize misconceptions of computing. The – often
implicit – approach is to exchange limited conceptions by more
suitable ones.
What, if students who are skeptical and don’t see the usefulness
for their daily lives are right? Probably the activities described
above are successful in triggering situational interest, but not in
maintaining and developing individual interest [25]? What if their
so-called misconceptions indeed are suitable for their current
personal situations, in which they are mostly users of digital artifacts,
and not programmers or computer scientists?
Consequently, the approach presented here does not aim for (immediately)
changing the perception of the discipline, but to
change the perception of digital artifacts, and of suitable interaction
patterns with digital artifacts. And this then is the starting
point for being able to perceive the discipline within a different
framework.
2.2 Analysis of Insider and Outsider
The claim that the perception of digital artifacts is of more immediate
concern is based on the research in computing biographies;
its results will be briefly presented in this section.
Crutzen [8] deconstructed the opposition of use and design in
computer science. According to this analysis, “the symbolic
meaning of use and design is constructed as an opposition in
which design is active and virtuous and use is passive and not
creative” – and in addition users are constructed as ‘outsiders’,
whereas designers are ‘Insiders’.
In Schulte and Knobelsdorf [32] the effects of prior experiences
on attitudes and enrollment in computer science were analyzed.
The dichotomy of ‘use’ (utilizing pre-given applications) and
‘design’ (creating new programs and applications), based on [8],
was used. Due to this dichotomy, Outsiders perceive themselves
as capable only of interacting with a computer in terms of using,
whereas Insiders perceive themselves as more competent due to
the skill to design new digital artifacts.
In some contrast to the original assumption of such a gap between
‘use’ and ‘design’, the empirical results of the study revealed
another biographical process of unaffiliated students. This lead to
a third concept: ‘professional use’. It refers to the ability to administrate
digital devices, and a focus on interaction with the comput-
45
er in terms of installing, configuring, and solving (usage) problems
(see [32]).
This is connected with a misconception of the computing disciplines
by Outsiders, who perceive a computing professional as a
kind of ‘digital caretaker’: A person who takes care of and maintains
digital infrastructures and helps when something malfunctions.
Typical activities resemble cleaning, repairing, exchanging
and updating some aspects of the whole, but not activities like
designing, creating, or problem solving on a larger scale.
(Aside: There is a popular distinction between digital immigrants
and digital natives, where everybody born after 1980 is regarded
as being a native. In our study, only digital natives participated.
So the difference between use, design, professional use, and of
Insider and Outsider is within the group of so-called Insiders.)
The overall conclusion of this is that Outsiders are not discouraged
by those affordances educators may see (difficulties in learning
programming, abstraction, complexity), but by problems
related to the everyday use of digital technologies: coping with
errors, configuring, installing, and so forth. In other words, the socalled
group of Outsiders is not discouraged by perceiving an
insurmountable gap between them as users and the goal to become
a designer. Instead, they perceive a gap between use and being
able to become a digital caretaker.
Moreover, it seems as if this group perceives the digital world and
its myriad artifacts more or less as a given, natural environment -
like nature itself. This perception completely neglects the role of
human effort, creativity and engagement in developing and shaping
the digital world. (Aside: In personal communication with
other researchers’ they approved this diagnosis, but it still is – in
my opinion - so dramatic that more effort should be done in researching
this finding. See also section 4)
Based on the above description, what can be done to foster interest
for this group of outsiders? And, moreover, how can this be
done without deterring novices who are in the group of insiders,
and are already interested in topics like programming?
In order to find the answer, we need to explore the above outlined
model a little further. First of all, both groups (Insiders and Outsiders)
start as inexperienced users, e.g. as gamers or web surfers.
But somehow several (the Insiders) gain confidence, interest and
are motivated to explore digital artifacts further, and eventually
discover the possibilities of designing by adapting, configuring or
developing web pages and by programming. In contrast the Outsiders
experience that they barely cope with the affordances of
everyday interaction with digital devices. The reason for this
difficulty might be due to attributing an irrational behavior to
digital devices, and/or a lack of specific ‘digital skills’ – as if only
people with special abilities are able to persuade digital artifacts to
do as they are supposed to.
Interesting is the anthropomorphism in attributing ‘irrational
behavior’ to digital devices. It shows a lack of knowledge of the
internal principles behind the perceived functions; and this
knowledge gap is filled by construing anthropomorphisms in the
conception of digital devices; in the hope of being able to understand
one’s own usage problems.
And here we find two handles to foster interest: First, there is still
some interest in understanding ‘what’s going on’ internally. This
should be used and deepened. Second, there is some immediate
need to understand the structure of digital devices.

3. DUAL NATURE CONCEPT
The concept (and term) ‘structure’ is borrowed from the philosophy
of technology (see [31] for details). Due to the dual nature of
technical artifacts, there is the need to understand the function of a
technical artifact as well as its internal design, the structure (structure),
in order to fully understand it.
Function captures the perspective of the use, and the purpose of
the artifact: What should it be used for? What can be done with it?
structure describes the internal mechanics: How it is made, how it
works and which concepts are used in its (internal) design.
Note, that ‘structure’ – in difference to the use of this term in
computing – includes (data) structures as well as algorithms and
processes.
The need to understand the internal ‘mechanics’ or ‘principles’
underlying the perceived function (functionality) is not restricted
to use. It is also acknowledged as a problem for novices learning
programming, in [3], [23], and [34] similar dual conceptions are
proposed. From now on and the rest of the article we use the terms
‘structure’ and ‘function’ to refer to this philosophical meaning.
The concept proposed here is to uncover structure behind function
(by experimenting); and thus bridging the perceived gap between
function and structure for Outsiders. But at the same time, structure
is still in focus, so that such teaching units should be of interest
and motivating for Insiders, too.
The key idea is the perception of the dual nature of digital artifacts,
which can be described by structure and function. Function
is implemented by structure, and structure is designed with a
purpose in mind; hence both are closely connected. Ideally, there
is a ‘harmony’ between structure and function.
Based on structure-knowledge, it should be easier to predict (and
expect) certain functionalities in digital artifacts, to predict and
expect certain ruptures, and to understand and memorize tactics
and strategies in the use of digital artifacts (for such examples, see
e.g. [3]).
Vice-versa, based on function-knowledge, it should be easier to
evaluate such an artifact. If a person truly understands the purpose
of an artifact, she is better able to assess whether the chosen and
implemented structure (the technical ‘solution’) fits the user expectations.
Without such an integrated understanding one is only
able to state being unable to carry out the desired purpose with the
tool – but not whether such usage problems are due to features of
the structure.
Of course, this duality is an analytical concept; in ‘real life’ there
exist only ‘complete’ digital artifacts – but the duality highlights
typical pattern of perceiving and understanding such artifacts.
A last issue is the notion that immediate ‘jumping into’ discussing
structure by a teacher might slow down the learning process,
because learners are missing important aspects (namely the function)
needed to really understand and memorize the teaching
content. Even worse, if they don’t grasp the function related to
structure, they aren’t able to see the usefulness of the learning
content at all.
3.1 Discussion of possible misconceptions
In this section some important aspects of the concept are rephrased,
in order to discuss some misconception the author noticed
when presenting the concept.
These misconceptions might be due to the education of the typical
audience of computing engineers. As Kroes [19] discuss, engineering
education can be interpreted as learning to bridge struc-
46
ture and function – so to speak automatically and unconsciously.
Therefore engineers are trained to overlook the gap and therefore
are likely to regard the concept as not really important, as the
main problems of the engineer are to refine and produce a suitable
technical solution (the structure) for a given description of the
function.
1) The first misconception is therefore the notion that duality
refers to the gap between the outside and the inside: The
function is perceived by the (external) user, whereas the
structure is in focus of the engineer, who focusses on the inside.
In fact, the duality can be observed for or within every
part of a digital system. Think e.g. about abstraction and the
layered model of network protocols. Each layer is used for
the next one, and itself hides the internal structure. Another
example is abstract data types: Again, we see abstraction
from structure, so that the engineer can focus on using the
data type. (Note, however, that e.g. the java library gives
hints about the internal structure when naming classes like
ArrayList or LinkedList – so that the user of the class has
some information whether it is better to use ArrayList of
LinkedList, also the functionality is the same)
2) The second misconception, related to the first one is that
people have problems declaring an aspect as structure, and
another as function. It seems easy when looking at inside/outside
the whole system (see misconception one). But
is an ArrayList a function or a structure? The answer is – of
course – that it is both. Every part of a digital artifact as well
as the whole can be perceived from either the viewpoint of
function, or from the viewpoint of structure. Because the artifact(s)
embody a dual nature: The need to be understood in
terms of function, of structure and in terms of their dual nature,
there specific link between both sides. (The crucial observation
from philosophy of technology is that a) it is quite
hard to see both sides of the coin at the same time and b) that
the two sides are based on quite different ontological and
epistemological grounds)
3) Misconceptions regarding structure: structure seen as in data
structure, and missing e.g. algorithms or any other dynamic
aspect of the implementation. Of course, structure is not a
term from computing (as it is used here), but from technical
philosophy and as such encompasses the mentioned aspects.
A similar term used in computing might be mechanism.
4) Misconception regarding function: as function as it is used in
computing: the functionality, or as the use of the artifact, its
design goals, requirements. To refer to the example used
above, the dichotomy was mechanisms and goals. Similarly
intention could be used. However, such an impression is too
narrow, as function also includes the impact which was not
intended (or as an engineer might say: mal-function).
5) Another misconception might be triggered by the use of the
terms insider and outsider, which associate the superiority or
preference for insiders. While it is a legitimate goal to foster
interest for computing, the main goal of this approach is fostering
self-determination - which includes of course the legitimate
decision not to study computing. But such decisions, as
well as a so-called Outsiders attitude towards digital artifacts
should be based on a somewhat truthful perception of the
discipline and the nature of digital artifacts.
4. Changing computing education
In the end, this approach strives for a change in computing education
on the school level. As can be seen in natural science or math,

the educational programs and rationales are strongly linked to the
academic disciplines, but nevertheless are having a strong selfview
of their unique role and importance for secondary education.
(However, readers interested only in the actual method might skip
this section. It gives an additional explanation and motivation for
experiments as a method and why they are suggested in this specific
form)
So far, the role of computing in school is still somewhat unclear.
Based on the discussion above – the duality of structure and function,
and associated perceptions of a dichotomy between user and
designer, and of Insider-Outsider – we define this role as supporting
learners to develop along this continuum from Users~Outsiders~function
to Designers~Insiders~duality. Learners
should explore different steps between use and design. (And,
experiments are likely to be one of the teaching methods supporting
this development.)
Interestingly, we can find several accounts within our discipline
which – at least partially – refer to similar goals. Some of them
will be briefly outlined within our framework of duality.
First of all, the abstraction and theorizing of the work in computing
biographies suggest a pattern of four different roles, which
can be used to describe different biographical stages and development
processes. The following is based on the thesis of
Knobelsdorf [17] p. 135ff. The empirically and theoretically
derived four biographical steps (or roles), are:
1) checking out or trial; 2) use or apply; 3) configure or modify;
and 4) create or produce.
1. Trial: Focus is on getting to know the digital artifact (DA),
trying out typical applications. Often motivated by curiosity.
DA is perceived in a playful fashion, as a toy to tinker with.
2. Use: Focus is on function; that is interacting with the DA in a
way that is useful for a certain task or within a given context;
and is therefore motivated by this need (which is external or
unrelated to the artifact itself). The DA is consequently perceived
as a medium, a communication device or a tool.
3. Configure: Focus is on adapting the DA to the user's need, by
changing parts of the hardware or software. Motivation is
based either on the desire to adapt the system to the individual
need, or by coping with errors. Corresponding to these different
motivations, the DA can be perceived as mysterious
and incomprehensible, or as a kind of magical artifact with
potentially unlimited possibilities.
4. Create: Focus is on extending the DA by individually or selfbuilt
items. Motivated by a need induced from use-context,
or by the desire or joy of being creative. DA is perceived as a
tool for creativity (a computer scientist might say, as a universal
machine).
Note, together with a change of the interaction pattern, the perception
of the artifact changes as well; but this correlation is not
causality. However, it seems possible that inducing a change in
interaction might cause a change in perception, too.
The second related area is design / end user programming. For
example Fischer and Giaccardi [12] propose Meta-Design. Metadesign
recommends a different role of design and designers,
because at design-time not all user issues and needs occurring at
use-time could be foreseen and fulfilled in a suitable manner.
Therefor users need to be designers, too – at least in part. They
also describe a continuum between end user usage and professional
design: 1) Passive consumer, 2) well-informed consumer,
3) end-user, 4) power user, 5) domain designer, 6) meta-designer
47
[12]. Note this continuum (again) develops as an increase of
awareness, understanding, and changing structure.
They also argue: “Cultures are substantially defined by their
media and their tools for thinking, working, learning, and collaborating.
[…] The importance of meta-design rests on the fundamental
belief that humans (not all of them, not at all times, not in all
contexts) want to be and act as designers in personally meaningful
activities. Meta-design encourages users to be actively engaged in
generating creative extensions to the artifacts given to them and
has the potential to break down the strict counterproductive barriers
between consumers and designers” ([12], section 6.2). Reframing
these arguments in terms of computer science education
suggests that it is not only about avoiding usage problems or
having proper tools at hand, but also about being actively engaged,
creative, being able to shape one’s own (immediate) environment,
and being able to develop oneself. The difference here is
that the authors may believe that meta-designers as professionals
need an excellent education; while maybe they believe that users
can act as designers without being (digitally) educated. But based
on e.g. the above described biographical results, such a selfeducation
process seems to be possible only for some people.
A third example puts these ideas discussed even further. It was
already postulated in the 1970es by Kay and Goldberg [16]. The
Dynabook should allow people to change the DA they are using,
and this would produce: “a metamedium, whose content would be
a wide range of already-existing and not-yet-invented media.”
This requires that the “burden of system design and specification
is transferred to the user. This approach will only work if we do a
very careful and comprehensive job of providing a general medium
of communication which will allow ordinary users to casually
and easily describe their desires for a specific tool.” In terms of
our framework such a description would mean to create a suitable
idea of structure from understanding the need for new or changed
function. Note, this is a design process, and no direct translation
from function to structure would be possible. So while from the
perspective of the discipline (or the authors of [16] at that time)
the idea was that ‘ordinary users’ will be capable of doing so - I
think in light of current developments such users would need to be
educated to be able to engage in such interaction with DA. And,
they would need to be educated in CS.
The authors also contrast two types of media or tools: Those with
fixed purposes, and those with the ability to be adapted to new
needs and ideas. While the first type, like cars and TV’s aim at
anticipating users’ needs, the other type (like paper and pencil)
“offer[s] many dimensions of possibility and high resolution;
these can be used in an unanticipated way by many, though tools
need to be made or obtained to stir some of the medium’s possibilities
while constraining others” [16].
While the aforementioned ideas (users as designers) focused on
personal development and self-fulfillment, here now the perspective
is widening towards development of the society and culture as
such. The vision is – so to speak– to unleash the full power of the
universal machine. From an educational point of view, it’s not
only about personal development, but also on participation: being
able to participate in the discourse about advancement of society,
and being able to take part in these developments.
Of course there are different degrees of such participation – again
from user to – maybe – meta-designer. Nevertheless, an educated
person should be able to understand what is going on.
While [12] and [16] the role of tool development and professional
design was considered, now the necessity of people, and their
education should be considered. In summary, the grand vision

ehind the small suggestion to employ experiments in CS education
is meant as a step towards this direction. The educational goal
is to emancipate and educate all, so that they can use and modify
or even create digital artifacts according to their own creativity
and needs. And on a general level, to support well-being and
development of society as such, by being a responsible and participating
member of (the digital) society.
4.1 Goals
In summary, the duality of function and structure is an analytical
perspective that should frame the perception of topics to learn for
teachers and learners. It highlights some aspects, and is therefore
connected to certain goals:
Link computer science to ICT-experiences from everyday
life.
Foster interest and motivation.
Foster integration in long term memory by linking new content
(mainly from structure) to already known topics from
everyday life (mainly from function).
Develop / teach competencies to analyze different kinds of
digital artifacts (and thus foster self-efficacy and maybe
change attribution).
Demonstrate the relevance of CS for future development; in
personal life as well as with regard to the wellbeing of society.
This has three facets:
Getting help for usage problems (reattribution, get problem
solving ideas based on structure knowledge) and
hence acknowledge the (immediate) value of the learning
content,
Understanding the internal mechanics of digital artifacts,
as a) a prerequisite or aspect of learning programming
as well as b) an aspect of understanding the
dual nature of digital artifacts.
Understanding the need (and role) of design (structure);
and therefore be able to alter or widen conceptions of
computing disciplines and the role of computing professionals.
Teaching and learning methods are needed, that provide bridges
between structure and function. Experiments - understood similarly
to science education - are promising candidates to build such
bridges in the CS classroom. In science, experiments are rigorous
methods developed to test (and sometimes develop) hypotheses
about phenomena in nature. They are different from the everyday
meaning in which experimenting is often associated with tinkering
or trial and error. In the next section, examples of experiments as
bridges between structure and function are presented and briefly
discussed.
5. DUALITY EXPERIMENTS
In this section some examples are given, which show a variety of
experiments as teaching method in CS education. In doing so, the
rationale behind and the desired effects are discussed.
Note, however, that the examples are more or less based on literature.
All of them were enacted, at least in basic form - but not yet
empirically evaluated within the above outlined approach; therefore
the intended effects are described on account of the described
theory, and are not empirical results.
5.1 The cell phone network
Using cell phones / smart-phones is an everyday activity of school
children. E.g. in Germany 96% of all teenagers (age 12-19) own a
mobile phone [26].
48
The teaching unit focusses on location based data produced by
mobile phones. In a first step, students are uncovering structure
and function of the cellular network.
A small experiment can be used as a motivating introduction[30]:
Two cell phones and a metal box are needed. In the first experiment
one phone is placed inside the box, which is closed. What
happens if one tries to make a phone call to that phone? It
wouldn’t ring because it is not available – no connection from the
other phone is possible. In the second version of this experiment,
the second phone is – after dialing the number of the first – also
put inside the box. What happens now?
This starting point raises the question how the cell phone network
works – what is its structure? A part of it is the need to localize
the mobile stations (the cell phones). And therefore, in the example
above, both phones in the metal box cannot connect to the
base station, and the called phone wouldn’t ring.
In a second step – which also may be called an experiment, but
more correctly is a role play – learners encounter the basic structure
of a cellular network (mobile stations, base stations, and the
central switching station with home location register) by enacting
(or experimenting with) use cases like connecting to another
phone. By enacting several scenarios learners will discover that
besides the content data (what they are talking over the phone)
also traffic data is transmitted and collected. Of these traffic data
the location data is in focus of the following step.
In [4] an example of location data collected by the phone service
provider is shown, including the raw data as a spreadsheet, an
interactive visualization and an interpretation of the data.
The learners will visualize the location data themselves, by using
a predefined framework. This framework allows inserting a
graphical representation (e.g. a dot) at the geographical coordinates
in one of the typical online maps (like OpenStreetMap).
Experiments with the data are made by changing the visualization,
e.g. adding a counter for each position and/or filtering for time
slots. For example, showing the position of the cell phone only
during night time indicates the residence of the cell phone owner.
These activities (producing different visualizations of the same
data) demonstrate the information that can be revealed, and what
possible (future) use of that information may be possible. This
includes a demonstration and discussion of location based services
like Google traffic, which predicts traffic conditions, based on
accumulated location data of cell phone users [27].
Students can learn – among other aspects – how computing is
changing and devolving, and eventually reflect on their conception
of computing disciplines.
In summary, the above outlined experiment (or, if you will experiments
(box, role play, and interactive visualization)) is aiming at
visualizing structure, namely effects of the receiving signal
strength, important parts of the cell phone network, and location
data. The main role of the experiment is to build the above mentioned
bridge from use experiences to perceiving the structure of
the cell phone network. It should do so by making the learners
curious and start to ask, what will happen and how does it work.
In other words, it should raise awareness for aspects of the otherwise
hidden structure of the cell phone network. The playful
approach should be motivating by making learners’ curious and
asking themselves, what will happen.
The immediate benefit for users is small, however. It may only a
better understanding of connection errors. In addition, the experiments
might raise the awareness of differences with regard to the
precision of one’s own location data, and that it can be influenced

y choice of handset, choice of network provider, and of course
by the individual pattern (also named habits, see [31]) in using the
mobile phone.
5.2 Algorithms in text-processing
Typically text processing is associated with ICT, and with teaching
usage. Sometimes, e.g. the object oriented analysis of text
processing is part of teaching Computer science, but predominantly
only function is analyzed (see e.g. [15]). But text processing
can also be used to demonstrate structure so that such knowledge
helps understanding the sometimes ‘irrational’ or ‘unpredictable’
behavior of text layout. A typical question of users might be:
“Why aren’t my pictures where I wanted them to be?”
In order to answer such a question, a course can start with the
historical example of Donald Knuth’s endeavor to capture the art
of computer programming, and his task to develop TeX in order to
be able to present it in a suitable layout.
One issue involved in developing such a text processing system is
line breaking. Paragraphs can be automatically aligned, too. We
are considering justification (as in this article) as the desired style
for paragraph alignment.
In its simplest form, full justification is rendered by filling a line
word by word until a word hasn't enough space, and then the
spaces between words are widened until the line is in full justification
(first-fit algorithm). A more advanced algorithm allows
white space between words not only to grow, but also to shrink
(best-fit algorithm).
Here different line breaks are possible, and thus different possibilities
have to be taken into account and evaluated in order to decide
whether shrinking or growing spaces leads to a better full justification.
The most advanced algorithm takes into account not only
the current line, but the whole paragraph. E.g. by omitting a local
maximum the global maximum for the paragraph can be optimized.
This is the total-fit algorithm developed together with TeX
by Donald Knuth and Michael Plass [18].
Thinking about these possibilities alone can help students to be
able to perceive the structure of text processing. To do this, an
experiment can be conducted in which students try to figure out
whether the local text processor (like e.g. Open Office or Microsoft
Word) uses total-fit.
If so, within a paragraph line breaks can change, if the last line of
the paragraph is altered. This should be observable in WYSIWYG
(“what you see is what you get”) text processors. Usually – at least
in our experiments with students – they don’t see such changes, so
the conclusion is that modern WYSIWIG processors do not use
the best algorithm available. However, maybe the data used in the
experiment didn’t reveal any changes, but with other text input it
would. On the other hands students might argue that changing line
breaks in already finished lines while typing would be confusing.
A web search can be added to check if producers of word processors
provide any information about the algorithms used.
However, the main learning goal of the experiment is not to figure
out which algorithm is used, but to understand that algorithms are
used, and that there are differences, and a lot of things happening
while the function experience suggests that simply characters
from the keyboard are ‘typed onto the screen’.
Does this help to better use text processing and e.g. prevent pictures
from being placed by random within a document? Short
answer: No (A longer answer will follow below). Is CS Education
(for Outsiders) therefore useless? Again: No. Let’s compare this
to science: What is the ‘use’ of knowledge about the solar sys-
49
tems, and planets circling? A sunset is still a sunset – yet many
people would agree that knowledge about the world is worthwhile
in its own. Similarly, bridging function to structure enables learners
to perceive structure, to get a better perception and understanding
of the digital world.
In summary, there are three intended effects of examples like this:
First, some learners should become intrigued to explore this world
further and e.g. learn what an algorithm is in detail.
Second, this understanding should be transferred to other aspects
of word processing, namely the example mentioned at the beginning:
Why can it be that pictures are seemingly moved randomly
around the text? A closer look here reveals that (usually) the user
has the option to change which algorithm is used to place the
picture on the page. So in a specific manner, this example might
also reduce difficulties in using.
Third, self-efficacy should be increased. As this experiment introduces
algorithms, but in the context of everyday use of ICT, it
builds a bridge to structure: Outsiders should experience and
notice that they have the ability to understand and gather such
structure-knowledge.
5.3 Search engine results
In this example, the notion of ‘personalized web search’ is experimentally
analyzed, using a method based on [11]. The question is
how the result-lists of Google are compiled, especially whether
everybody gets the same results.
Freuz et al. [11] designed an experiment to analyze this. They
opened three Google accounts (Kant, Nietzsche, and Foucault),
and performed individually different searches, and so allowed
Google to collect a search history of the accounts.
Afterwards they compared results of the three accounts to some
selected common searches. It became clear that within the first
page of presented results, the three accounts got different answers
to the same search term.
The interesting aspect here is the methodology used: Focusing on
the user experience (function), a scientific experiment is devised:
Using a hypothesis, a methodological setup, and the gathering of
data which has to be analyzed and interpreted. The remarkable
thing about this experiment is, that mainly operating within the
usual user paradigm (within the framework of function), the experiment
reveals novel insights in the function but at the same
time also insights in the internal mechanics (structure).
Students can learn that by engaging in thorough observation it is
possible to infer structural aspects relying on normal usage only.
6. DISCUSSION
In science teaching, experiments are a well-known teaching method.
E.g. science labs offer courses for interested schools, where
learners can experience interesting experiments. Although quite
successful, such activities are sometimes labeled as “hands-on,
minds off”-approaches.
While these activities - that’s the essence of the saying - are very
good in triggering situational interest, as they provide most often a
quite spectacular experience, they are less good in developing and
maintaining individual interest.
In research on interest, situational and individual interest is differentiated.
Situational interest is conceptualized as being due to
external causes, by catch facets like e.g. new approach to teaching
or an interesting puzzle to solve [25]. Thus, the teacher can catch
the attention, and situational interest is triggered. Individual inter-

est is developed by repeated experience of situational interest.
According to Mitchell [25], meaningfulness and involvement are
the two most important aspects to foster individual interest. Meaningfulness
is the perception of a learner of the content as being
valuable for her current life. Involvement is due to active engagement
in the learning process.
Based on this theoretical account, the following aspects are important
affordances for developing and analyzing experiments as
interest developing approach to teaching:
If learners feel or believe they can influence the process (e.g. the
learning process, or the process of interacting with a digital artifact)
they are more likely to be involved.
Teachers and learners often have different perceptions of what is
of most importance in the learning process. How can Outsiders be
enabled or supported to see value (=meaningfulness) in learning
computer science?
Self-efficacy (or similar: self-confidence, self-view) should be
preserved in order to prevent learners from avoiding situations
which are threatening to their self-esteem. And, vice-versa, learners
should have opportunities to succeed, because “as achievement
enhances self-image and confidence in an upward spiral
in which increased levels of achievement enhance motivation
which in turn leads to further increases in achievement” [21].
The thesis is that building bridges between function and structure
supports the perception of the value in learning structure (and
hence CS, too).
There are two issues connected to experiments: One is that such
activities are interesting due to external factors (hands on), which
are not transferred to the discipline. The other is that embedding
the disciplines’ content in engaging activities can be seen only as
one first step, and in addition to involvement also meaningfulness
needs to be supported. And here we are talking about meaningfulness
for Outsiders, who are most likely are conceptualizing CS as
related to computers and a kind of mysterious professional pattern
in interaction with such devices.
Thus the main point is to shift attention from function to structure,
while also preserving the link to the original function and context
of an Outsider as a user. When and how will computing become
meaningful and valuable for such a person?
In other words, the above described activities might be labeled as
experiments, but only if they are building a bridge between function
and structure. Most likely such a bridge can be built by starting
from everyday use experiences with digital artifacts.
Here we also see a difference to e.g. CS unplugged. Both approaches
have some commonalities: They are aiming at triggering
situational interest (of Outsiders) by engaging them in unusual
and intriguing activities, and both shift attention away from using
the PC. The difference is that CS unplugged simply removes the
distracting content and plugs out the PC, duality experiments start
with use experiences and some kind of digital artifact from an
unusual perspective that should raise curiosity in the usually
hidden structural aspect. In terms of the briefly outlined theory of
interest, CS unplugged provides a good catch facet for situational
interest, while a duality experiment aims at providing some hold
facet in order to develop individual interest by raising the impression
of personal meaningfulness and value.
From the above discussed experiments some conclusions can be
drawn, in order to highlight specific features and intended effects
of experiments in computer science education.
50
Firstly, they do not aim to (immediately) change the perception of
the discipline, but to change the perception of digital artifacts, and
the perception of suitable interaction patterns with digital artifacts.
They introduce learners to the ‘internal mechanics’ (structure) of
digital artifacts so that they can reattribute the causes for usage
problems.
They should raise awareness for possible variants of the internal
mechanics, so that learners can experience and explore a ‘design
space’, in which one can understand that (within constraints)
different structure-variants can lead to (sometimes subtle but
important) differences in function.(see e.g. example in 5.2 ).
Experiments should take into account different usage approaches,
so to include different levels of use-competence and habits for
interaction with a digital artifact (see e.g. 5.1 ).
In order to support a general change, and not only a local change
in interacting with the currently analyzed digital artifact, transfer
should be included; e.g. from the actual experiment to other digital
artifacts, or other aspects of the same digital artifact (like e.g.
from ‘text alignment’ to ‘picture alignment’ in section 5.2 ).
7. CONCLUSIONS
Overall, the examples described here are based on the idea of
duality reconstruction [31]. Nevertheless empirical research is
needed to discern what in detail do students learn from such experiments.
Also empirical data could help to understand which
experiments really foster bridging structure and function and
support learners to see the dual nature of digital artifacts. And of
course, it will be interesting to conduct empirical research on the
question if learners change with regard to self-efficacy, motivation
or interest as predicted.
Structure and function are somewhat connected to Insider (-
>structure) and Outsider (->function). Providing bridges should
support the construction of learning-units that are appealing to
both groups and their different interests (and perspectives on the
topics).
And, most important, the experiment should provide opportunities
to perceive structure and function as essential. Especially using
bridges between structure and function should increase the sense
of value as it mutually supports the perspective that is not in focus
of the learner.
Questions yet to solve arise in the following aspects:
The experiments outlined above allow learners to control parts of
the experiments. The actual interaction with an artifact as part of
the experiment usually can vary and is not as strictly defined as in
‘real’ scientific experiments. More work should be done to discern
important aspects of the teaching method, including e.g. link to
learning theory and interaction of outcome and learner attributes.
The above described experiments are fulfilling more or less the
just mentioned affordances. In order to further develop the teaching
method, our discipline can learn from experiments in natural
science - both positively and negatively. For example, we know
from science education that it is possible to engage in contraproductive
trial-and error, so called ‘Hands-on, minds-off’ experiments.
It is also an open question whether the intended transfer of a
changed perception in the interaction with DA triggers a change in
perception of the computing disciplines, too. Remember e.g. the
digital caretaker, discussed in section 2.2 . Such a misconception
of the discipline should be changed – but will it?
These questions are answerable only by (empirical) research.

8. REFERENCES
[1] Bell, T. et al. 2011. Overcoming obstacles to CS education
by using non-programming outreach programmes. Proceedings
of the 5th international conference on Informatics
in Schools: situation, Evolution and Perspectives (Berlin,
Heidelberg, 2011), 71–81.
[2] Bell, T. and Newton, H. 2013. Unplugging Computer
Science. Improving Computer Science Education. D.
Kadijevich et al., eds. Routledge. 66–82.
[3] Bhavnani, S.K. et al. 2008. Strategy-Based Instruction:
Lessons Learned in Teaching the Effective and Efficient
Use of Computer Applications. ACM Trans. Comput.-
Hum. Interact. 15, 1 (May. 2008), 2:1–2:43.
[4] Biermann, K. 2011. Data Protection: Betrayed by our own
data. Die Zeit. http://www.zeit.de/digital/datenschutz/2011-
03/data-protection-malte-spitz. Accessed: 2012-01-13
[5] Competition Workshop 2006 - Accepted Papers:
http://www.bwinf.de/competition-workshop/papers.html.
Accessed: 2012-01-13.
[6] Computer Science for Fun - cs4fn: HOME:
http://www.cs4fn.org/. Accessed: 2012-01-13.
[7] Computer Science Unplugged |: http://csunplugged.org/.
Accessed: 2012-01-13.
[8] Crutzen, C.K.M. 2000. Interaction, a world of differences.
A vision on informatics from the perspective of gender
studies. Open Universiteit Nederland.
[9] Curzon, P. et al. 2009. Enthusing & inspiring with reusable
kinaesthetic activities. Proceedings of the 14th annual
ACM SIGCSE conference on Innovation and technology in
computer science education (New York, NY, USA, 2009),
94–98.
[10] Feaster, Y. et al. 2011. Teaching CS unplugged in the high
school (with limited success). Proceedings of the 16th annual
joint conference on Innovation and technology in
computer science education (New York, NY, USA, 2011),
248–252.
[11] Feuz, Martin et al. 2011. Personal Web searching in the
age of semantic capitalism: Diagnosing the mechanisms of
personalisation. First Monday. 16, 2 (2011).
[12] Fischer, G. and Giaccardi, E. 2004. Meta-design: A
Framework for the Future of End-User Development
(Online-Version). End User Development. H. Lieberman et
al., eds. Kluwer Academic Publishers, Dordrecht, The
Netherlands. 1–26.
[13] Guzdial, M. 2003. A media computation course for nonmajors.
Proceedings of the 8th annual conference on Innovation
and technology in computer science education (New
York, NY, USA, 2003), 104–108.
[14] Heersink, D. and Moskal, B.M. 2010. Measuring high
school students’ attitudes toward computing. Proceedings
of the 41st ACM technical symposium on Computer science
education (New York, NY, USA, 2010), 446–450.
[15] Hubwieser, P. 2006. Functions, Objects and States: Teaching
Informatics in Secondary Schools. Informatics Education
– The Bridge between Using and Understanding Computers.
R.T. Mittermeir, ed. Springer Berlin Heidelberg.
104–116.
[16] Kay, A. and Goldberg, A. 1977. Personal Dynamic Media.
Computer. 10, 3 (1977), 31–41.
51
[17] Knobelsdorf, M. 2011. Biographische Lern- und Bildungsprozesse
im Handlungskontext der Computernutzung. Freie
Universität Berlin.
[18] Knuth, D.E. and Plass, M.F. 1981. Breaking paragraphs
into lines. Software: Practice and Experience. 11, 11 (Nov.
1981), 1119–1184.
[19] Kroes, P. 2002. Design methodology and the nature of
technical artefacts. Design Studies. 23, 3 (May. 2002),
287–302.
[20] Leiviskä, K. and Siponen, M. 2010. Attitudes of sixth form
female students toward the IT field. SIGCAS Comput. Soc.
40, 1 (Mar. 2010), 34–49.
[21] Littlejohn, A. 2008. The Tip of the Iceberg: Factors Affecting
Learner Motivation. RELC Journal. 39, 2 (Aug. 2008),
214–225.
[22] Maloney, J. et al. 2010. The Scratch Programming Language
and Environment. Trans. Comput. Educ. 10, 4
(2010), 1–15.
[23] Mayer, R.E. 1981. The Psychology of How Novices Learn
Computer Programming. ACM Comput. Surv. 13, 1 (1981),
121–141.
[24] McLachlan, C. et al. 2010. Student perceptions of ICT: a
gendered analysis. Proceedings of the Twelfth Australasian
Conference on Computing Education - Volume 103 (Darlinghurst,
Australia, Australia, 2010), 127–136.
[25] Mitchell, M. 1993. Situational interest: Its multifaceted
structure in the secondary school mathematics classroom.
Journal of Educational Psychology. 85, 3 (1993), 424.
[26] MPFS 2011. JIM-Studie 2011: Jugend, Information, (Multi-)Media
; Basisuntersuchung zum Medienumgang 12-
bis 19jähriger. Medienpädagogischer Forschungsverbund
Südwest.
[27] Official Google Blog: The bright side of sitting in traffic:
Crowdsourcing road congestion data:
http://googleblog.blogspot.com/2009/08/bright-side-ofsitting-in-traffic.html.
Accessed: 2012-01-05.
[28] Papastergiou, M. 2008. Are Computer Science and Information
Technology still masculine fields? High school students’
perceptions and career choices. Computers & Education.
51, 2 (Sep. 2008), 594–608.
[29] Pasternak, A. and Vahrenhold, J. 2010. Braided teaching in
secondary CS education: contexts, continuity, and the role
of programming. Proceedings of the 41st ACM technical
symposium on Computer science education (New York,
NY, USA, 2010), 204–208.
[30] Redaktion Schulprojekt Mobilfunk 2008. Mobilfunk und
Technik. Fächerübergreifende Sachinformationen für projektorientiertes
Lernen. Klassen 5–10 sowie gymnasiale
Oberstufe. Informationszentrum Mobilfunk e. V.
[31] Schulte, C. 2008. Duality Reconstruction - Teaching Digital
Artifacts from a Socio-technical Perspective. Proceedings
of the 3rd international conference on Informatics in
Secondary Schools - Evolution and Perspectives: Informatics
Education - Supporting Computational Thinking (Torun,
Poland, 2008), 110–121.
[32] Schulte, C. and Knobelsdorf, M. 2007. Attitudes towards
computer science-computing experiences as a starting
point and barrier to computer science. Proceedings of the
third international workshop on Computing education research
(Atlanta, Georgia, USA, 2007), 27–38.

[33] Schulte, C. and Magenheim, J. 2005. Novices’ expectations
and prior knowledge of software development: results
of a study with high school students. ICER 2005: Proceedings
of the 1st International Computing Education Research
Workshop; Seattle Washington USA October 1 - 2
2005 ([New York, NY], 2005), 143–153.
[34] Soloway, E. 1986. Learning to program = learning to construct
mechanisms and explanations. Commun. ACM. 29, 9
(1986), 850–858.
[35] Taub, R. et al. 2012. CS Unplugged and Middle-School
Students’ Views, Attitudes, and Intentions Regarding CS.
Trans. Comput. Educ. 12, 2 (Apr. 2012), 8:1–8:29.
[36] Thies, R. and Vahrenhold, J. 2012. Reflections on outreach
programs in CS classes: learning objectives for “unplugged”
activities. Proceedings of the 43rd ACM technical
symposium on Computer Science Education (New
York, NY, USA, 2012), 487–492.
52

Agile Projects in High School Computing Education –
Emphasizing a Learners’ Perspective
Ralf Romeike
University of Potsdam
August-Bebel-Str. 89
14482 Potsdam, Germany
romeike@cs.uni-potsdam.de
ABSTRACT
Software projects are seen as a methodology for secondary computing
education which is highly appropriate and meets the demands
and goals of Computer Science (CS). Yet the majority of
models and examples for project-based lessons rely on a traditional
software development approach: the waterfall model. In
this paper such models are analyzed for their strength, problems,
and deficiencies. Based on the results of the analysis a new approach
to projects in secondary computing education is presented
which uses the concept of didactic transposition to adapt agile
software development methods for project organization, management,
and implementation in class. The resulting model applies
valuable practices of eXtreme Programming and Scrum and provides
a set of tools that allow high school software projects to
benefit from modern software development methods. By emphasizing
dynamic processes and a clear course of action an attractive
perspective on CS is promoted.
Categories and Subject Descriptors
K.3.2 [Computers and Education]: Computer Science Education
General Terms
Human Factors, Theory.
Keywords
Secondary computing education, agile methods, project-based
learning.
1. INTRODUCTION
In secondary computing education, software projects are promoted
to provide an appropriate and student-oriented approach to
Computer Science (CS) [19, 23, 32, 39]. Yet, most projects in this
context are mainly focused on sequential project layouts that resemble
traditional software development (SD) methodologies
such as the waterfall model. In recent years it became apparent in
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists, requires
prior specific permission and/or a fee.
Conference’10, Month 1–2, 2010, City, State, Country.
Copyright 2010 ACM 1-58113-000-0/00/0010…$10.00.
53
Timo Göttel
University of Hamburg
Vogt-Kölln-Str. 30
22527 Hamburg, Germany
tgoettel@acm.org
professional SE that such methodologies often fail to produce
high quality products, bring forward delays in delivery, and insufficiently
consider customers’ needs (e.g. [26]). Analogue issues
can be found in school projects: unfinished projects, missing time
and motivation for testing, neglected documentation, and teachers’
difficulties in managing software projects are just some of the
problems reported (cp. [19, 24, 32]). In modern SD, agile methods
are promoted to provide a dynamic project management that relies
on interaction and short design iterations. Agile methods build
upon values and provide practices that are also highly expedient
in high school contexts. Therefore, we present a new approach to
projects in secondary computing education, which implements the
theory of didactic transposition to adapt agile methods for project
organization, management and implementation in classroom.
Valuable agile practices of eXtreme Programming (XP) [2] and
Scrum [39] will provide a set of tools allowing software projects
in high schools to reference modern SD by highlighting dynamic
processes that help to focus on good results, a clear course of
action, and an attractive perspective on professional CS by addressing
common problems at the same time. In section 2, research
on projects in computing education will be discussed and
problems with prevalent models will be analyzed. The findings
suggest a missing consideration of the learners’ perspective in
project models. In section 3, agile methods in professional and
educational settings are discussed for their potential of supporting
a learners’ perspective. By describing the agile model for school
projects in computing education common agile practices are characterized
and adapted. Finally, the model is discussed in the context
of existing models, its potential, and issues in computing
education.
2. PROJECTS IN EDUCATION
2.1 Project Based Learning
Project-based learning (PBL) is an approach to teaching and
learning in the classroom aiming for engaging students in explorative
and problem-solving activities in authentic contexts. The
concept is described to originate from teaching in Zurich and
Paris in the 19th century. Since then, the idea has been picked up
by various teachers and researchers and enriched with psychological,
pedagogical and sociopolitical aspects, e.g. by Dewey und
Kilpatrick [11]. Projects are understood as learning processes that
draw on interests and demands of the students by striving for a
complex result, often a product. This includes planning, problemsolving,
analysis of different solutions and the evaluation of the
process and its product. PBL is known for increasing students’
motivation, for strengthening self confidence, and for fostering
satisfaction in process and outcome (cp. [5]). Therefore, it is pro-

moted to foster high-level thinking skills including problemsolving
and analysis skills. Thus, it helps to gain a deep understanding
of topics and processes (cp. [1, 25]). PBL is known to
encourage peer interaction (cp. [25]).
In Germany, Frey [16] elaborated PBL by describing the process
along the following steps: A project starts with a project idea,
which should be based on the interests of the students. Subsequently
the idea is specified in order to find agreement on what
the class will be trying to achieve. After planning the necessary
activities the project plan is carried out. Milestones and metacommunication
serve as tools for supporting the process. This
model is commonly accepted within Germany and was fundamental
for an adapted model in German computing education (cp. 2.3)
Summarizing, projects are characterized by being problemfocused
and interdisciplinary, by allowing students a choice of
topic and foster personal responsibility. Projects are generally
carried out over a longer time span and are often graded “differently”,
e.g. by focusing more on processes and applying less pressure.
2.2 PBL in Secondary Computing Education
Because of the above mentioned benefits, PBL is widely used in
higher computing education (cp. [6, 13]) and secondary computing
education, especially in the context of software projects [39,
41]. Consequently, PBL is recommended as an appropriate approach
to computing education in the majority of German curricula.
However, there is limited research focusing on methodologies
for SD projects in secondary education. Meerbaum-Salant and
Hazzan [33] constitute “as far as we know, no general methodology
has been developed for software development projects in the
high school“. Consequently, they propose an approach which
focuses on a mentoring model for teachers (cp. 2.4).
In Germany, a model for high school software projects was proposed
by Frey [15] and elaborated by Schubert and Schwill [39].
The majority of published school SD projects can be attributed to
this model. Therefore the model will be analyzed in the following.
2.3 A Professional Perspective
A majority of publications concerning the use of projects in secondary
computing education 1 stems from the 1980s and 1990s,
generally analyzing and adopting the professional approach to SD
and adopting the waterfall model for the classroom. The idea was
that a professional model for SD would provide an appropriate
framework for school projects: it offers students a structured
learning process and gives insights into professional SD processes
at the same time. In comparison to other subjects using PBL only
as a teaching method without a connection to the subject matter
itself, projects are scientifically anchored in CS [39]. Consequently,
there is a large body of practice reports on SD projects.
However, by analyzing publications of the major German conferences
in computing education of the last 10 years we could not
find any publication concerning methodologies for school projects
in general secondary education. However, research shows that
teachers consider it as important that students get familiar with
SD processes, e.g. by running “through the workflows of the waterfall
model” [30].
1 Research on this topic is rare. The statement reflects the situation
in Germany. However, we are not aware of comparable
publications from other countries.
54
Figure 1 illustrates the project steps of Schubert and Schwill’s
model [39] with the corresponding output of each phase. The
proposed model describes student activities along the software life
cycle. However, it does not provide methods or practices of how
students can reach the expected outcome of each phase. In the
problem analysis phase all important environmental conditions
need to be gathered clearly and completely. Furthermore, this
phase includes planning activities for time, team and equipment.
The resulting requirements definition serves as a contract between
teacher and students. In the subsequent system design process a
model of the system is specified by dividing the “overall system”
into modules. Until here, the activities shall be performed by the
full team, which is allowed to split up into subgroups for minor
tasks. Then, smaller groups handle a module each under their own
responsibility. The remaining phases follow the software life cycle.
Fig. 1. Project model by Schubert and Schwill [39] .
Concerning the team structure, Schubert and Schwill recognize
that the organization of the team cannot follow hierarchical structures
as typically used in professional SD teams. Instead, they
suggest equal status and responsibility among team members and
a “force” for communication and common goals. This goal is
addressed by assigning eight different roles for student positions
within the project (computer responsible, project supervisor, interface
responsible, tester, documentation responsible, butler, session
chair, secretary).
Criticism of the methodology points out problems with a perceived
bureaucratic overhead: “It is always the same: Students
refuse to first plan on paper. Because only small programs are
written, these are not documented. The taught principles of software
development are hardly noticed” [19]. Often, testing is omitted
in the project realization due to a lack of time and a lack of
perceived importance, especially if the software does not have a
practical use after the project (cp. [24]). Other problems may result
from the structure of the project: The time span that needs to
be scheduled is up to half a year (Schubert and Schwill [39] suggest
to perform one project per semester). This is difficult to plan
for, especially if students lack project experience and supportive
practices. Additionally, a sequential project layout collides with
“project-unfriendly” circumstances of formal lessons such as lim-

ited time, heterogeneous student abilities and lessons spread over
several weeks. Humbert [23] even summarizes that the pedagogical
dimensions of PBL are not sufficiently considered in such a
project model. Additional problems of conducting school SD
projects are outlined in the following by pursuing a teachers’
perspective.
2.4 A Teachers’ Perspective
Meerbaum-Salant and Hazzan [32] analyzed difficulties encountered
by teachers in mentoring SD projects in Israeli high schools.
Even though the curricular background and objectives are somewhat
different 2 than for projects in general educational settings as
described in this paper, the results are similar to the problems
reported from German teachers. Additional problems were identified
in the contexts of scheduling the project, CS expertise of the
teachers, considering students’ individual performances, and
evaluation of the project. Teachers describe mentoring of SD
projects as a more complex task compared to traditional teaching.
Meerbaum-Salant and Hazzan [33] express the need for a general
methodology for SD projects in high school and address the previously
identified problems in a mentoring methodology
(ACMM). It is intended to support teachers, who are expected to
be confident in a variety of knowledge types [32]. Therefore it
describes a set of practices (Pedagogical Class Management Aspect,
Social Aspect, Project Management Aspect) that shall be
considered by teachers while mentoring a project. The ACMM
takes into account the principles of agile software (such as communication,
simplicity, feedback, respect), which basically are
reflected in the teacher-student interaction.
2.5 A Learners’ Perspective
For learning settings, where a team of students is working cooperatively
on projects, we see potential for taking the idea of applying
agile methods further than it is described in the ACMM: project
management can be done by the student team. This can be
supported by straightforward and easy to use methods adopted
from modern SD. Additionally, students may benefit from experiencing
a SD process which also includes management aspects in
addition to activities like analysis, designing, coding, and testing,
as described in the other models. In the following we demonstrate
how problems identified by Meerbaum-Salant and Hazzan [32]
may be addressed in such a project by considering agile methods
as presented in section 3.
Schedule: Teachers may need to catch up with teaching of material
during the project. The sequential project approach does not
allow for such a teacher’s intervention without disturbing the
process. However, in an iterative project design, issues and success
can be discussed in class regularly.
Required CS knowledge: Some teachers admit a lack of project
development knowledge. Students will need help with CS knowledge
while solving problems. It meets the ideas of PBL if student
2 German curricula emphasize computer science concepts in the
context of general education which only partly includes algorithmic
thinking and programming. In comparison, the underlying
curriculum of the study emphasizes foundations of algorithmic
thinking and programming [17]. Additionally, we understand
projects as teamwork where several students or the
whole class are working on the same goal, Meerbaum-Salant
and Hazzan [31] describe projects where students work individually
and “each student has his or her own project subject”.
55
teams would be empowered to manage projects themselves. This
can be supported by easy to follow practices and strategies which
make teacher involvement almost unnecessary. Additionally,
clearly defined practices may support teachers’ confidence. Heterogeneous
student teams stimulate mutual assistance before requiring
the help of a teacher.
Students’ individual work: Teachers see a need for personal supervision
in order to achieve a timely completion of the project
and meeting of the requirements. Agile methods allow for transferring
this responsibility to the project team, hence relieving the
teacher.
Evaluation of project outcomes: Agile practices naturally lead to a
variety of documents which can be considered for project evaluation
(e.g. estimates in planning poker or burn-down charts). Furthermore
mutual assessment within teams may be performed.
All suggested practices require a change of the project perspective
from the teacher to the learner. In section 3 the mentioned agile
practices are discussed in more detail and transferred to classroom
settings by the use of didactic transposition.
2.6 Didactic Transposition for Project Methodologies
Didactic transposition describes the process of adapting professional
knowledge of a domain for teaching scenarios based on a
didactic intent [9]. Hazzan et. al. [21] applied didactic transposition
on agile SD methods with the intent to create a teaching
framework and a mentoring methodology for software projects.
The model for school SD projects discussed in 2.1 can be attributed
to didactic transposition as well. Here, the professional process
was adapted under consideration of the underlying principles
of PBL. Schubert and Schwill [39] emphasize the advantage of a
method which is learning activity and learning content at the same
time. However, we see potential for shifting the focus from professional
process knowledge to modern professional methods
which may be adapted in a way that they address previously outlined
problems in school SD projects.
In the following, we discuss agile methods in professional SD and
in education. We applied didactical transposition for developing
an agile approach to projects in computing education which emphasizes
a learners’ perspective.
3. AGILE PROJECTS IN COMPUTING
EDUCATION
3.1 Agile Methods in Professional and Educational
Settings
Agile methods are popular amongst researchers and practitioners
for enabling software developers to create systems that are more
likely to be accurate in meeting customers requests, finishing in
time, building robust systems, and creating usable/readable code
(cp. [22]). Therefore, in industry agile methods are currently replacing
waterfall or other linear methodologies that are known for
shortcomings in the above mentioned goals of SD. Agile methods
are focused on social interactions and dynamic creative processes.
Hence, developers in agile teams often report on a strong satisfaction
in their work experience and strong confidence in their outcomes
(cp. [27, 31]).
The agile manifesto of 2001 [3] clearly presents values contrasting
traditional linear methodologies and underlying understandings:

1. Individuals and interactions over processes and tools
2. Working software over comprehensive documentation
3. Customer collaboration over contract negotiation
4. Responding to change over following a plan
Agile methods are implemented in various frameworks. XP [2]
and Scrum [40] are the most prominent implementations applied
in industry and academia. Those methodologies define practices
to assure compliance with the agile values as described in the
agile manifesto. Agile methods are mostly understood to support a
development process comprehensively from start to the final
stage. However, the individual practices can be classified according
to their main targets. Some practices are designed to structure
team processes and customer collaboration while other practices
focus on the quality of code and outcome.
Recently, several authors report on using agile methods in CS
education at university level. Braught et al. [7] promote the use of
agile methods because it helps female students to engage in programming
tasks through interaction with peers. Nagappan et al.
[34] highlight social experiences, learning processes, and quality
of code when using agile methods in CS1 courses. However, literature
shows that there may be possible barriers hindering an
implementation of agile methods in university scenarios. Rico and
Sayani [37] present a study where they found that students had
already established their own approaches and habits for SD that
were opposing practices of agile methods. According to Rico and
Sayani it was almost impossible to convince the students to adhere
to the introduced practices of agile methods. In this connection,
they recommend to introduce agile methods as early as possible.
Consequently, a benefit is assumed in the use of agile methods
in school contexts. Literature on agile methods at university
levels is still discussing the possibility of conducting a fullyfledged
project management according to an implementation as
XP or Scrum. Some authors promote an almost complete implementation
of XP (e.g. [28]), some recommend to use and adopt a
subset of practices (e.g. [8]), while others exclusively use one
practice (e.g. pair programming) to support computing education
(e.g. [7]). Schneider and Johnson [38] reviewed agile methods in
computing education and highlight the importance of applying
suitable practices according to their goals instead of fulfilling
complete implementation of agile methods. Accordingly, Hazzan
and Dubinsky [20] present ten reasons to consider agile methods
in computing education.
Yet, literature on agile methods in secondary computing education
is rare. Weigend [42] introduced elements of XP (user stories,
spikes, test driven development, refactoring, and big visible
charts) to provide a project-based iterative infrastructure that supports
writing of high quality code. However, a sound methodology
for connecting the presented elements in projects was not
provided.
The work in hand is based on encouraging classroom experiences
presented by Göttel [18]. In various educational CS projects, agile
methods, such as pair programming, standup meetings, informative
workspaces, and user stories supported students in their project
work and additionally helped students to discover social aspects
of CS. This success provided our basis for developing a
comprehensive agile model for school projects in computing education.
56
3.2 An Agile Model for Projects in
Computing Education (AMoPCE)
As discussed above, PBL represents a common teaching and
learning method in computing education. However, even if common
models suggest a structure and requirements for school software
projects, methods are described insufficiently for the individual
phases. Agile methods and modern SD principles provide a
set of clearly described strategies that seems well suited for an
implementation in school contexts. As described in the agile
manifesto, they emphasize communication, visualization, teamwork
and common goals. In the following, we want to introduce a
model for school software projects that builds on the character of
agile methods in order to address the problems outlined in section
2. It follows the agile manifesto by focusing on
1. Students and their interactions
2. Rapid success and working software
3. Collaboration in order to strive for a common goal over fulfilling
a contract
4. Responding to change and learning progress over following a
plan
The individual strategies and tools are illustrated in fig. 7 and will
be described below in an agile model for projects in computing
education (AMoPCE).
In this description we focus on processes and methods that are
central for the agile SD process. Additional pedagogical aspects
such as triggering students’ motivation or finding agreement in
choosing a project topic are not covered.
The process contains various techniques adapted from professional
SD practices. They provide clear lines of action that can be
followed by the students (e.g. generating user stories, planning
poker, defining tasks). However, before applying them in a project
we suggest introducing and practicing each method. On the
other hand, an explorative learning approach is possible: The
methods describe the processes in such detail that appropriate
material can be created which allows students to learn and perform
the processes independently.
The methods will be first discussed from a professional perspective
3 and subsequently transferred in a way that they can be applied
in classroom (italic text). Examples will illustrate the process.
3.2.1 Preparation
Creating software requires competencies in programming and
using tools. It is the responsibility of the teacher to make sure that
the students have acquired the competencies needed for the software
project ahead or that they will be able to acquire them during
the project (e.g. with provided teaching material). Another
aspect considers establishing an appropriate infrastructure (see
[33] for further elaboration).
3.2.2 Ideas in
The initial steps of a SD process usually are devoted to requirements
analysis listing possible features, approaches, and needs of
3 In order to maintain a consistent presentation, methods and practices
described in this paper are based on [35], which may be
referred to for elaboration and additional information on modern
SD practices.

the target audience. This phase is based on interviews, observations,
and brainstorming sessions with the customers.
Building upon ideas and interests of students is a central characteristic
of PBL. However, experience shows that students may not
easily come up with ideas that can be implemented in such a project,
especially at the first time. Therefore, an initial presentation
of possible projects and the use of creativity techniques (e.g.
brainstorming) are suggested. Resulting ideas shall be written
down on individual Post-Its or cards for each activity that the
software needs to provide (cp. fig. 2).
Fig 2. Post-its for recording ideas.
3.2.3 User Stories
User stories briefly describe features of a product that should be
available to the actual user. Each user story addresses a specific
activity of a user and is derived from the ideas of the requirements
analysis. They are written from the perspective of a customer.
User stories should easily fit on index cards and also be understood
by non-developers. They should provide additional space
for an estimation of the work effort.
In combination, user stories specify the entire intended product. A
final state and amount of user stories has to be accomplished in
agreement with the customer. Thereafter, stories are prioritized
together with the customer by sorting user stories according to the
importance of each story. Priorities are presented using incrementing
numbers by powers of ten from 10 (most important) to
50 (least important).
Furthermore, the customer is asked to pick those stories that
should be available in the initially delivered outcome or rather
first major release. Consequently, discussion and reprioritizing
stories may be necessary considering the basic features wanted for
the first release.
User stories are created using various brainstorming techniques
and take account of domain specific needs, knowledge, and approaches
of the actual users specified in the requirements analysis.
A user story
- covers one activity that needs to be addressed
- represents the perspective of the customer
- is short, i.e. contains no more than three sentences
- does not use technical terms
- does not specify technology or tools
57
Fig. 3. Cards for user stories holding a title, a description, an
estimate for workload, and a priority. Estimates will be added
after completing the planning poker.
In professional SD projects one of the most important (and often
unsuccessful) tasks is to find out what the customer wants. User
stories provide a helpful way for achieving this goal. Since the
students are going to implement their own ideas in their software
projects, this goal does not apply. However, the team needs to
find an agreement on the requirements for the software. These
will be represented from a user’s perspective: User stories briefly
describe how a user interacts with the software. They can be developed
from the previously recorded ideas. These need to be
analyzed to find out, which interaction is really going to happen.
This will be achieved by role playing and observation. Role playing
is suggested as an attractive method for computing education
to understand processes (cp. [4, 12, 14]). However, some of these
examples use role plays in awkward contexts. In contrast, by role
playing user interaction with the desired software system, students
use a method which is anchored in modern SD processes
and helps to identify relevant processes. The rules of the role play
are simple: One student pretends to be the software and reacts
accordingly. A sheet of paper may be used to illustrate the display.
Another student takes the role of the user and instructs the
software about what he or she wants to do, according to the previously
obtained ideas. The remaining students observe the situation
carefully to understand details and constraints of the desired
product. The role play should be repeated several times with
changing actors until no more new requirements arise. With this
experience it should be easy to formulate the requirements from a
user’s perspective and write down the corresponding user stories
(cp. fig. 3). Finally, the user stories receive a value for their priority.
Priorities can be determined as a team, since generally
agreement is quickly found.
Communication is an essential element of agile SD. Even if a set
of user stories will now describe the final goal, questions and
changes will appear in the following process. Since there is no
customer who can answer questions and make decisions in order
to clarify yet open questions, a group member needs to take over
this special role: the product owner. This position may be passed
around with each iteration.
3.2.4 Planning Poker
Planning poker is a hands-on method helping participants to estimate
time needed for the work packages and guarantees a fair and
comprehensible approach amongst all team members. Each participant
holds a deck of cards to estimate the workload of a user
story. There should be cards representing estimates in comprehensible
units (e.g. developer-days) and special cards allowing players
to indicate a lack of information, a need for a break, and already
finished functionalities as shown in fig. 4.

Fig. 4. Deck of cards used for the planning poker.
Each play round is devoted to one user story. A user story is
placed in the middle of the table by the dealer and all participants
place a card specifying their estimate face down on the table. All
played cards are turned at the same time. The dealer collects the
played cards and sums up the estimates trying to set up an average
estimate. The dealer should address outliers by asking for reasons
explaining fundamental differences in the estimates. Furthermore,
the dealer should reflect on average estimates referencing the
differences in the played cards. After each round the acquired
estimate is written on the card of the user story. Additionally, the
individual estimates are written on the back of the user story card
to keep track of the decision process.
For a student team it is one of the most difficult tasks to estimate
the workload and time demands of a given project due to a lack of
project experience. Additionally, very likely not all planning relevant
aspects are known at this point, learning processes will happen
and changes may be necessary. Planning poker describes a
playful way for challenging all students of the team to engage in
the planning process by analyzing user stories and tasks, relating,
estimating, explaining and defending their calculations, thus
practicing their communication skills and ability to give and receive
criticism.
For school software projects the same card values can be used as
in professional SD. However, since these projects comprise a
shorter working time, instead of days, 15 minute-periods seem to
be appropriate. Each student estimates the time he or she believes
he or she would need to implement the user story in focus. User
stories should be presented with decreasing priority. Discussion
of very divergent estimates will help resolving unspecified requirements
and assumptions. After the planning poker is finished,
the total workload for all user stories is divided by the number of
programmers or programming teams (if pair programming is
used) and compared with the time available. Unlike in professional
SD, the priority is not to fulfill all requirements of the software
but it is indispensable to keep the time available. If there is
a major difference, e.g. more than 20 per cent, amount or complexity
of the user stories need to be modified.
Again, for planning poker one team member needs to be the
dealer. We suggest for this special role the position of a “teammaster”,
who leads the planning poker as well as team meetings.
In the subsequent process this position also may be passed around
with each iteration.
3.2.5 Tasks
After the initial planning poker, user stories are broken down into
tasks. Usually each user story can be seen as a collection of tasks.
A task is a rough description of a work package that should be
done by a single developer. A task should have a unique selfexplanatory
name and should indicate its priority. The workload
58
of each task should also be estimated by planning poker. Afterwards,
task estimates are summed up to double check them with
the initial estimates on the corresponding user stories. Thus, these
estimates should not differ tremendously.
Fig. 5. Tasks for User Story “Control Lives”.
While the user stories describe project goals from the perspective
of the user, the students now need to change their perspective and
look at them from a developer’s viewpoint. By dividing user stories
into tasks, various design decisions need to be made. Experienced
programmers will now benefit from their competencies and
known best practices, less confident students at this point can
benefit and learn from team members, processes and team discussion.
3.2.6 Iterative Development, Prototypes, and Milestones
Agile processes are designed to provide short iterations that constantly
come up with working prototypes that can be used and
discussed with users or customers. This allows for rapid feedback
loops that help to uncover misunderstandings, to detect issues in
using the interface, and to adapt to new requests. An iteration is
supposed to be short and has to be balanced according to implementing
new features, fixing bugs, responding to change, and
considering group dynamics or individual demands. In professional
contexts, iterations vary between one week (5 working
days) to one month (approximately 20 working days).
Iterations are planned in a team meeting by considering user stories’
priorities, estimates, and the intended duration of an iteration.
After deciding on the user stories to work on for the next
iteration, they are pinned on a project board including associated
tasks.
In school software projects, the planning of long development
processes is reported to be difficult. Also, teachers report issues
maintaining student motivation while they are not getting a grasp
of the product until the whole project is assembled. The learning
theory of constructionism emphasizes that learning happens especially
felicitously in a context where learners are engaged with
creating and investigating a personal relevant product (cp. [35]).
Iterative development allows for creating a series of prototypes
that can be analyzed, examined and played with in a constructionist
sense. Also, such a design of the development processes allows
for a higher flexibility in team organization and diversification
due to more frequently changing tasks. Hence, each iteration is a
mini-project containing each phase of the SD processes (requirements,
design, code, test), but is easier to handle. This gives students
the opportunity to perform the whole process several times
within one project, to learn from and reflect on previous experiences
and to take over several tasks in the team (in comparison to
other project models, where tasks are more strictly divided
amongst students). Besides the iterations, which should not be

longer than one to two weeks (which equals 2-4 lessons), milestones
are used to structure the process and point out major
achievements within the project. We suggest identifying 2-3 milestones
for each project, representing versions of the final product
with increasing value. However, only the achievement for the next
milestone in the development is determined at a time. Milestones
can be used for presenting the project progress for the rest of the
class or teachers. Also, milestones should be positioned at times
when the project pauses and teacher input is planned. Goals for
the next milestone and the project progress are visualized at the
project board.
3.2.7 Project Board
Project boards visualize goals and status of a current iteration and
support target-oriented discussions. They present user stories and
tasks in different status areas. Project boards are updated and
discussed throughout the entire process. Thereby, it helps team
members to keep track of the progress of the design process: the
different areas of the board are used to present goals and accomplishments
to the whole team. There are three main status areas:
to-do user stories with associated tasks for the current iteration,
tasks that are in progress, and completed tasks. Furthermore, there
is an area to store user stories that need to be reconsidered in a
future iteration. To provide a clear view, another area is reserved
for finished user stories, allowing to take off corresponding task
cards. Figure 6 presents an accordant project board.
Additionally, a burn-down chart is available on the project board.
The chart visualizes the working time left in an iteration and work
that needs to be done according to the task estimates. The chart is
constantly keeping track of the progress by plotting the remaining
sum of tasks at the end of a working unit.
Likewise, in a school software project all user stories with corresponding
tasks are collected and presented at the team’s project
board. The project board is the central organizational and informative
workspace for the entire project and should be available
at all times, e.g. by placement at the classroom wall. It is also the
meeting point for the regular standup meetings.
Fig. 6. Project Board including a burn-down chart.
3.2.8 Standup Meetings
Standup meetings provide a recurring fast and short update of the
efforts of the team: Each team member has to report on accomplished
tasks, possible issues in accomplishing certain goals, and
59
a plan for the work day. Meetings are done while standing to
guarantee a fast and goal oriented session kicking off a workday
and should not exceed 5 to 15 minutes.
In school projects, standup meetings can provide an elegant way
for starting off a lesson or working day within a project by encouraging
team communication, sustaining motivation and identifying
problems. The team gathers around the project board and
recalls the project status, success and problems of the last working
session and the goals for the day. After each team member has
given a short statement the burn down chart is updated. If nonminor
problems are identified, a longer meeting may be scheduled.
3.2.9 Pair Programming
Pair programming ensures an elaborated coding style: A pair of
programmers uses a single programming environment for coding.
The person using the keyboard and mouse is adopting the role of
the “driver”. The driver is actually coding and asked to present his
or her ideas to the second programmer (the “navigator”) verbally.
Meanwhile, the navigator questions the coding outcomes, discusses
possible misinterpretations, and seeks for alternative solutions
that are more straightforward by keeping in focus the overall
goals. The roles of driver and navigator are changed repeatedly
during a workday. Programming in pairs helps to detect possible
slips in the design and architecture of the code at early stages.
Furthermore, it helps programmers to build upon social interaction
uncovering misinterpretations of relations and intentions of
code parts.
In many schools, two students share one computer due to limited
hardware availability and hence often program in pairs. The agile
method of pair programming supports this practice and adds a
framework that encourages attention from both students, mutual
learning and a notion of programming as a social activity.
Fig. 7 illustrates the organization of a school software project
based on AMoPCE as described above. Other agile methods and
ideas may be included in such a school project as well, e.g. test
driven development, refactoring or “keep it simple”.
3.3 Focusing on Programming Style and Outcome
Several agile practices may be applied in the process to bring
forward a high quality outcome. However, since these practices
are optional for project organization and partly depend on programming
environments, in the following they are only summarized
but not adapted for school software projects.
3.3.1 Test Driven Development
Test driven development replaces documentation and provides
criteria to evaluate the code solutions: Programmers define intended
functionalities by writing automatic tests covering all
states and the correctness of the accordant results. First reports of
using this method in secondary education have been published
(e.g. [10]).
3.3.2 Refactoring
Refactoring introduces the idea that every part of the code should
be reconsidered and changed if a more accurate solution can be
found: Developers should rewrite parts of code without adding
functionality when there is a more straightforward solution available
that passes all automatic tests. Emphasizing this aspect may
raise students’ awareness of efficiency and for evaluating different
solutions.

3.3.3 Keep It Simple
This claim refers to code minimalism: each function should be
solved by minimal and straightforward code snippets to ensure an
elegant and readable code that is easy to maintain.
4. DISCUSSION
The proposed agile model for SD projects AMoPCE addresses a
majority of the previously identified problems. Agile practices fill
the learners’ gap between requirements and outcome by providing
clearly defined strategies for handling difficult planning tasks.
Based on the perception of PBL as a team activity, which is in
line with modern SD, a team size of 4-6 students is recommended.
Dividing the class into several teams, in comparison to having the
full class working on one project as proposed e.g. by Schubert and
Schwill [39], allows for addressing a broader range of topics,
hence it is easier to meet the interest of more students. Also, it
should be much easier to find agreement within smaller teams.
Adopting an iterative project design matches the formal circumstances
of school projects. In most cases, projects will be worked
at along the regular school timetable, sometimes in a so-called
project week on several days. In both settings, pedagogical aspects
of working in group settings, such as giving curricular input,
intermediate project presentations or discussing of common problem
solving strategies, can be taken into account easier, since
meetings in plenum are part of the model. Correspondingly, projects
are divided into mini-projects, which are easier to overview,
plan and understand; bureaucratic overhead is reduced. Classroom-management
aspects are addressed with professional practices
such as standup meetings. This includes recalling of the
project status at the beginning of each lesson and quickly planning
the individual and group activities for the day.
Fig. 7. Agile Model for Projects in Computing Education (AMoPCE).
60
Within projects, ideas, students’ motivation, and skills change
over time. Due to the limited time available to work on the projects
per week, this is especially very likely for school software
projects. Agile methods welcome changes and provide mechanisms
to adapt to them with often changing tasks and a straightforward
implementation of PBL. Students’ confidence is addressed
by increasing familiarity stemming from the iterative
character of the process.
In comparison to the ACMM, which provides solutions for teachers’
difficulties that are grounded in teaching situations, the proposed
agile model provides solutions for students’ difficulties in
learning situations. This in return is expected to relieve the
teacher. Giving students clearly defined practices to manage their
development processes allows teachers to focus on supporting
elaboration and implementation of students’ ideas, thus changing
the teacher’s role from instructor to coach. This better meets the
demands of PBL. Additionally, it allows teachers to highlight
creativity and social aspects that are rarely seen in connection
with computer science (cp. [8, 29]). Applying an adapted SD
methodology in school that is also implemented by well known
large scale companies may help teachers and students to build an
adequate and appealing understanding of computer science relying
on creativity, dynamic change, feedback, and soft skills.
These attributes may support an attractive notion of computer
science, as found in our first experimental settings with agile
methods in school projects [18] .
In summary, AMoPCE is suited for supporting the objectives of
PBL, for maintaining a professional orientation and for easing the
mentoring of software projects. However, in this model curricular
aspects such as content and size of a project are not explicitly
considered. Nevertheless, it provides the flexibility to fit into a
variety of possible scenarios. Evaluation and assessment aspects,

e.g. assessing individual achievement in comparison to group
achievement are not determined by the model. However, again
practices of SD appear to be adaptable and can be considered:
Frequently, agile development teams use self assessments and
subsequent peer reviews to verify individual workload and commitment.
As outlined in section 3, there are further practices of agile methods
that focus on the quality of the outcome: test driven development,
collective code ownership, refactoring, and keep it simple.
We acknowledge these practices to be also useful in educational
settings but we understand them to be highly dependant on features
and methodologies of the used programming environments
and tools (e.g. code repositories, automatic testing environments).
The use of agile practices in school SD projects has the potential
for replacing the so far predominantly used sequential model.
From discussions with teachers we know of the high interest, but
a lack of knowledge and resources concerning the use of agile
methods. This includes the demand for a revised project model.
With AMoPCE, as outlined in this article, we present a model
which explains ideas and realization possibilities of agile practices
and which can be used as blueprint. In a next step of our
research the model will be applied in classroom settings and it
will be investigated, to which extend the expected benefits and
problem solutions will be approved in practice.
5. REFERENCES
[1] Barak, M., Waks, S. and Doppelt, Y. (2000). Majoring in
technology studies at high school and fostering learning.
Learning Environments Research 3(2): 135-158.
[2] Beck, K. and Andres, C. (2004). Extreme Programming Explained:
Embrace Change (2nd Edition).
[3] Beck, K., Beedle, M., et al. (2001). Manifesto for agile software
development. The Agile Alliance: 2002-04.
[4] Bergin, J. (2000). The Object Game. An Exercise for Studying
Objects. Online at:
http://csis.pace.edu/~bergin/patterns/objectgame.html (visited
01.07.2012).
[5] Blumenfeld, P. C., Soloway, E., et al. (1991). Motivating
Project-Based Learning: Sustaining the Doing, Supporting
the Learning. Educational Psychologist 26: 369-398.
[6] Blumenfeld, P. C., Soloway, E., et al. (1991). Motivating
project-based learning: Sustaining the doing, supporting the
learning. Educational Psychologist 26(3-4): 369-398.
[7] Braught, G., Wahls, T. and Eby, L. M. (2011). The Case for
Pair Programming in the Computer Science Classroom.
Transactions on Computing Education 11: 2:1--2:21.
[8] Caspersen, M. E. and Kölling, M. (2009). STREAM: A First
Programming Process. Transactions on Computing Education
9: 4:1--4:29.
[9] Chevallard, Y. (1988). On didactic transposition theory:
Some introductory notes. In Proceedings of the International
Symposium on Research and Development in Mathematics,
Bratislava, Czechoslavakia.
[10] Christopher, G. J. (2004). Test-driven development goes to
school. J. Comput. Small Coll. 20(1): 220-231.
[11] Dewey, J. and Kilpatrick, W. H. (1935). Der Projekt-Plan:
Grundlegung und Praxis, Hermann Böhlaus Nachfolger.
[12] Diethelm, I. (2007). Strictly models and objects first - Unterrichtskonzept
und -methodik für objektorientierte Modellierung
im Informatikunterricht. Kassel, Universität Kassel.
61
[13] Fincher, S. and Petre, M. (1998). Project-based learning practices
in computer science education. In Proceedings of the
IEEE Frontiers in Education Conference (Tempe, Arizona).
[14] Fothe, M. (2007). Algorithmen in spielerischer Form. In
Proceedings of the Praxisband der GI-Fachtagung Informatik
und Schule.
[15] Frey, K. (1983). Die sieben Komponenten der Projektmethode
- mit Beispielen aus dem Schulfach Informatik. LOG IN
3(2): 16-20.
[16] Frey, K. and Schäfer, U. (1982). Die Projektmethode. Weinheim
[a.o.] Beltz.
[17] Gal-Ezer, J., Beeri, C., et al. (1995). A high school program
in computer science. Computer 28(10): 73-80.
[18] Göttel, T. (2012). The image of Computer Science and social
aspects of modern software development processes in school
contexts. Hamburg, University of Hamburg.
[19] Hartmann, W., Näf, M. and Reichert, R. (2006). Informatikunterricht
planen und durchführen, Springer.
[20] Hazzan, O. and Dubinsky, Y. (2007). Why Software Engineering
Programs Should Teach Agile Software Development.
SIGSOFT Software Engineering Notes 32: 1-3.
[21] Hazzan, O., Dubinsky, Y. and Meerbaum-Salant, O. (2010).
Didactic transposition in computer science education. ACM
Inroads 1(4): 33-37.
[22] Highsmith, J. and Cockburn, A. (2001). Agile Software Development:
The Business of Innovation. Computer 34: 120-
122.
[23] Humbert, L. (2005). Didaktik der Informatik, Teubner.
[24] Koerber, B. (1992). Die Angst des Lehrers vorm Projektunterricht.
LOG IN 12(5/6): 3.
[25] Krajcik, J. S., Czerniak, C. and Berger, C. (1999). Teaching
children science: A project-based approach, McGraw-Hill
Boston, MA.
[26] Larman, C. and Basili, V. R. (2003). Iterative and Incremental
Development: A Brief History. Computer 36: 47-56.
[27] Layman, L., Williams, L. and Cunningham, L. (2004). Exploring
Extreme Programming in Context: An Industrial Case
Study. Proceedings of the Agile Development Conference.
Washington, DC, USA, IEEE Computer Society: 32-41.
[28] Loftus, C. and Ratcliffe, M. (2005). Extreme Programming
Promotes Extreme Learning? Proceedings of the 10th annual
SIGCSE conference on Innovation and technology in computer
science education. New York, NY, USA, ACM: 311-
315.
[29] Maass, S. and Wiesner, H. (2006). Programmieren, Mathe
und ein bisschen Hardware... Wen lockt dies Bild der Informatik?
Informatik-Spektrum 29(2): 125-132.
[30] Magenheim, J., Nelles, W., et al. Competencies for informatics
systems and modeling: Results of qualitative content
analysis of expert interviews. In Proceedings of. IEEE.
[31] Mann, C. and Maurer, F. (2005). A Case Study on the Impact
of Scrum on Overtime and Customer Satisfaction. Proceedings
of the Agile Development Conference. Washington, DC,
USA, IEEE Computer Society: 70-79.
[32] Meerbaum-Salant, O. and Hazzan, O. (2009). Challenges in
Mentoring Software Development Projects in the High
School: Analysis According to Shulman. Journal of Computers
in Mathematics and Science Teaching 28(1): 21.
[33] Meerbaum-Salant, O. and Hazzan, O. (2010). An Agile Constructionist
Mentoring Methodology for Software Projects in
the High School. Transactions on Computing Education 9:
21:1--21:29.

[34] Nagappan, N., Williams, L., et al. (2003). Improving the CS1
Experience with Pair Programming. ACM SIGCSE Bulletin
35: 359-362.
[35] Papert, S. and Harel, I. (1991). Situating Constructionism.
Constructionism. S. Papert and I. Harel. Norwood, N.J.,
Ablex Publishing.
[36] Pilone, D. and Miles, R. (2008). Head first software development,
O'Reilly Media, Inc.
[37] Rico, D. F. and Sayani, H. H. (2009). Use of Agile Methods
in Software Engineering Education. Proceedings of the 2009
Agile Conference. Washington, DC, USA, IEEE Computer
Society: 174-179.
[38] Schneider, J.-G. and Johnston, L. (2005). eXtreme Programming:
Helpful or Harmful in Educating Undergraduates?
Journal of Systems and Software 74: 121-132.
62
[39] Schubert, S. and Schwill, A. (2011). Didaktik der Informatik,
Springer.
[40] Schwaber, K. and Beedle, M. (2001). Agile Software Development
with Scrum.
[41] Tucker, A., Deek, F., et al. (2003). A model curriculum for
K-12 computer science: Final report of the ACM K-12 task
force curriculum committee. New York, NY: The Association
for Computing Machinery.
[42] Weigend, M. (2005). Extreme Programming im Klassenraum.
In Proceedings of the INFOS 2005 - 11. GI-
Fachtagung Informatik und Schule (Dresden, Germany,
2005). GI-Edition - Lecture Notes in Informatics (LNI).

Conceptual Change and Epistemological Belief
Framework for Web Site Credibility Instruction
Abstract—This paper describes instructional needs regarding
Web-based instruction and reviews a line of research
examining students’ determinations of Web site credibility. It
proposes a framework for instructional interventions that
draws upon two research traditions - scientific conceptual
change research and epistemological beliefs about Web-based
information. Recommendations for instruction and research
based upon this model are delineated.
Keywords-misconceptions, student engagement with
technology, beliefs, conceptualizations of computing, teaching
approaches, teaching methods, teaching with educational
technology
I. INTRODUCTION
Is Web site credibility an important consideration for
students doing research and using Web-based materials or are
most students already knowledgeable in this area? This paper
argues that considerations of Web site credibility are even
more important than in the past and strongly tied to today’s
critical thinking skills, both for computing education and for
learning in general. Further, it is erroneous to assume that
students are knowledgeable in this area. To illustrate, I
describe research that investigates students’ and pre-service
teachers (i.e., teachers-in-training) determination of Web site
credibility. This research underscores the need for
instructional interventions at elementary, secondary and postsecondary
(university levels), and provides the impetus for
developing effective instructional interventions. Thus, the
main focus of this paper is in the development of a theoretical
framework that can guide instructional interventions. In
developing this framework, we borrow upon research
frameworks provided in the conceptual change research
related to science learning and epistemological belief research
related to the Internet.
II. CREDIBILITY RESEARCH
Do students make determinations of Web site credibility
that differ from determinations of experts (professionals in
various fields and doctoral students)? What sorts of
determinations do they make? What kinds of instructional
interventions are effective for students?
Marie Iding, University of Hawaii
College of Education
1776 University Avenue
Honolulu, HI, 96822, U.S.A.
miding@hawaii.edu
63
In an initial study, Klemm, Iding and Speitel [1]
compared scientists’ and pre-service elementary and
secondary teachers’ determination of information credibility
of various science resources, including information from the
Web. They found that secondary pre-service teachers (who
had more science training and required coursework in the
area) were more like scientists in their determinations.
However, pre-service teachers were more likely to rate
popular information sources such as newsmagazines and
popular news television shows as credible information
sources. In contrast, scientists rated these information sources
as low in credibility. There was some agreement on museum,
aquarium and nature center resources, which were rated
uniformly highly by all 3 groups.
In subsequent research investigating tenth graders
credibility determinations regarding biology information on
Web sites, an instructional intervention was employed [2].
Over a four-day period of time, a teacher provided instruction
in examining specific areas relevant to making credibility
determinations – Web site authors and institutional
affiliations, information veracity, and organization of Web
sites. Students’ lists of criteria affecting their own credibility
determinations were more extensive after the intervention, and
students reported that they would spend more time evaluating
Web resources for credibility in the future.
In other research Iding, Crosby, Auernheimer and Klemm
[3] investigated university students’ credibility determinations
regarding content related to topics that they had covered in
class (i.e., computer science or education topics). Findings
indicated that students exhibited awareness of and suspicion
regarding vested interests of Web site authors with
commercial interests, although they did not as critically regard
vested interests of Web site authors whose purposes were
viewed as solely educational in nature.
In work with Norwegian undergraduate and graduatelevel
university students enrolled in film studies classes, Iding,
Nordbotten and Singh [4] noted that the more educated
students were, the less confident they were about their Web

site credibility determinations. This finding makes sense
intuitively, as it underscores the difficulties inherent in
information credibility determinations.
In other research, credibility determinations were affected
by literacy levels and socioeconomic status (SES) [5]. For
example, with respect to health care posted on Web sites, HIV
positive patients with lower educational levels and literacy
skills were less critical of fraudulent claims regarding
purported “cures” for AIDS than were their more educated
counterparts from higher SES levels.
Furthermore, Fogg, Soohoo, Danielson, Marable,
Stanford and Taber [6] found that members of the general
public could describe adequate bases upon which to make
credibility determinations. However, when faced with actual
Web sites, they abandoned these criteria in favor of
determinations based on Web site appearance (or “design
look”).
Anecdotally, the author has found that while teaching in a
Pacific Island US territory in June, 2012, university level
students appeared to be less aware of Web site credibility
issues in citing sources than were their US counterparts. It
should be noted that these students are from a developing part
of the Pacific and the majority would fit US definitions of
living below the poverty level. Thus, many do not have
computers at home or regular easily obtainable access to the
Internet. Additionally, the university-level courses in which
they were enrolled were only for several weeks, not allowing
extensive work with resources or time to make effective
credibility determinations.
These examples underscore the ongoing need for critical
Web credibility evaluation instruction in educational contexts,
at all levels, or wherever and whenever computers are
introduced into instruction. While it is clearly possible to
teach Web and information evaluation without reference to a
theoretical basis underlying instruction, I propose that it may
be valuable to develop instruction that is consistent with
frameworks known to be associated with the growth of critical
learning skills. Because determinations about information and
credibility implicitly involve epistemological considerations, I
draw briefly on research that investigates epistemological
beliefs regarding information on the Internet. Secondly, since
science and scientific learning is an area where students must
counter their initial conceptions or preconceptions in order to
learn effectively, and because science-related Web sites are
replete with misinformation, I discuss conceptual change
research and how it might be drawn on analogically to
facilitate the development of aspects of an instructional model
for effective instruction in credibility evaluation.
III. EPISTEMOLOGICAL BELIEFS AND THE INTERNET
People implicitly make epistemological determinations or
exercise epistemological beliefs in selecting information from
the Web; reading, watching or listening to it; and finally in
64
citing it or using that information to make decisions or to
influence their thinking in some way. Braten, Stromso, and
Samuelstuen [7] conducted a relevant study in this area and
adopted the definition of personal epistemology originally
proposed by Hofer as, “how knowing occurs, what counts as
knowledge and where it resides, and how knowledge is
constructed and evaluated” (p. 1).
Such a definition provides a useful basis for instruction,
especially in different content areas. Braten et al. [7] carried
out research investigating Norwegian university students’
epistemological beliefs around information on the Internet,
using dimensions of epistemological beliefs proposed by
Hofer and Pintrich [8]. Results indicated that two dimensions
of epistemological beliefs were confirmed via factor analyses.
As Braten et al. explained, “The first dimension, general
Internet epistemology, ranged from the integrated view that
the Internet is an essential source of true, specific facts to
doubt about the Internet as a good source of true factual
knowledge. The other dimension, justification for knowing,
ranged from the view that Internet-based knowledge claims
can be accepted without critical evaluation to the view that
knowledge claims encountered on the Internet should be
checked against other sources, reason, and prior knowledge”
(p. 141).
Clearly, effective instruction for credibility
determinations should ideally include an introduction to
epistemological considerations that can be general or content
specific. Considerations such as those raised by Hofer in the
definition above can guide the discussion, and inspire
questions such as: What is knowledge? Who creates it? Who
determines whether it is accepted in a field? What are the
characteristics of credible knowledge? How does one
personally determine whether knowledge expressed on the
Web is credible?
IV. CONCEPTUAL CHANGE RESEARCH
A classic means of countering science misconceptions
was summarized by Chinn and Brewer [9]. Their instructional
model consisted of the following points:
1) “Consider a physical scenario whose outcome is not
known” (p. 38).
2) “Predict the outcome” (p. 38).
3) “Construct competing theoretical explanations to
support the predictions” (p. 38).
4) “Observe the outcome (anomalous data)” (p. 38).
5) “Modify competing theoretical explanations, if
necessary” (p. 38)
6) “Evaluate competing explanations” (p. 38)
7) “Reiterate the preceding steps with different data.” (p.
38)
Chinn and Brewer [9] described various factors that can
influence the likelihood of conceptual change, including how
anomalous data is handled – whether it is accepted, ignored,

einterpreted, or excluded, for example. They explained that
individuals’ background knowledge, competing theories,
characteristics of the anomalous data, and strategies used also
affect the likelihood of conceptual change.
Chinn and Brewer [9] argued that only by considering
one’s preconceptions in light of anomalous data deeply, (often
prompted by explaining one’s thoughts to others) can true
conceptual change occur. They also convincingly explain
that, “In order to learn epistemological commitments
appropriate to evaluating evidence and theories, students may
need to participate in a community that regularly debates
alternative theories discusses responses to anomalous data,
and evaluates evidence and theories. During this process of
enculturation, students are like apprentices learning the craft
of scientific reasoning and teachers can use strategies of
modeling (by thinking out loud in front of students), coaching,
providing scaffolding and gradually withdrawing it, and
reflecting upon the cognitive strategies used (A. Collins,
Brown & Newman, 1989)” (p. 33).
Pintrich, Marx and Boyle [10] also addressed the
conceptual change research, emphasizing that countering
preconceptions at the logical, cognitive level may leave out
important non-rational, motivational and affective elements of
the determinations. These elements include motivational
aspects of conceptual change, including goals, interests,
values, and self-efficacy beliefs. They called for further
research in this area. Clearly, affective and motivational
aspects also play major roles in people’s credibility
determinations, as the work of Fogg et al. [6], mentioned
earlier, illustrates. To what specific extent and how remain
potentially fruitful questions for further research in Web site
credibility determinations as well.
Thus, if we borrow from conceptual change research in
science learning, we need to consider students’ goals in
credibility determinations and possible motives (or nonmotives).
Further, we need to consider whether there is an
analogous event that can trigger conceptual change in much the
same way conceptual change is trigger by anomalous events or
information that does not fit with preconceptions. Within
context of the conceptual change research, perhaps an
analogous triggering event could be confrontation of Webbased
information on a Web site that has a very credible
appearance. Exercises with Web sites from different content
areas and with different levels of information quality should
facilitate learning.
V. GENERAL INSTRUCTIONAL RECOMMENDATIONS
Based upon this summary of research on Web credibility,
and syntheses of suggestions that can be drawn from the
research in epistemological beliefs related to the Internet, and
conceptual change in science, the following suggestions are
proposed for inclusion in effective Web information
credibility instruction:
65
• Consideration of content specific epistemological
approaches and beliefs. Discussion of the questions
such as the following can be profitable: What is
knowledge? Who evaluates it and determines whether
it is accepted?
• Discussion of specific aspects of Web site credibility,
including sources, possible vested interests of Web site
authors, corroboration of information with other
sources
• Teachers should model and think aloud as they
evaluate the credibility of actual Web sites of varying
degrees of credibility [10,11]
• Students should work collaboratively to evaluate actual
sites of varying credibility, and be encouraged to
develop own strategies.
• Students should be encouraged to articulate own
explanations for their determinations to others.
• Self-regulation should be central. Thus, if we consider
Web site credibility determinations to involve to some
extent a process of conceptual change, students should
be encouraged to address the following questions when
considering a Web site as an information source:
Firstly, what is my initial conception or credibility
determination regarding this Web site? Secondly, why?
What characteristics caused me to make this
determination – Web site appearance, author’s title,
educational or company affiliation? Lastly, when
comparing my initial conception to criteria I have
learned, do I still consider this information/this Web
site to be credible? Why or why not?
• Teachers should provide opportunities for continued
discussions of successes or failures and suggestions for
improving credibility determinations. Simple
questions such as the following can elicit this
information: What’s working well? What isn’t? What
suggestions do you have for improvement?
VI. CONCLUSION
This paper has summarized a line of research
investigating students’ Web site credibility determinations. It
has argued for the need for critical Web credibility instruction
at all levels and has summarized research in epistemological
beliefs about information from the Internet, and conceptual
change in science learning. By drawing on these perspectives,
instructional recommendations have been synthesized. These
recommendations provide a foundation for further
development in instruction and research that addresses the
need for critical information credibility determinations when
using Web-based information.
REFERENCES
[1] E.B. Klemm and M. Iding, Do scientists amd teachers agree on the
credibility of media information sources, International Journal of
Instructional Media, vol. 28, 2001, pp. 83-91.

[2] M. Iding, R. Landsman, and T. Nguyen. Critical evaluation of scientific
websites by high school students. In D. Watson & J. Anderson (Eds.),
Networking the Learner: Computers in Education: Seventh IFIP World
Conference on Computers in Education Conference Proceedings.
Dordrecht, Netherlands: Kluwer Academic Publishers, pp. 373-382,
2002.
[3] M. Iding, M.E. Crosby, B. Auernheimer, and E. B. Klemm, E.B. Web
site credibility: Why do people believe what they believe? Instructional
Science, 37, 2009, pp. 43-63
[4] M. Iding, J. Nordbotten, and J.M. Singh, Credibility: Norwegian
students evaluate media studies Web sites. In D. Kuma & J. Turner
(Eds.), Education for the 21 st century: Impact of ICT and digital
resources: WCC 2006 Santiago, Chile, (pp. 327-331). New York:
Springer, 2006.
[5] E. G. Benotsch, S. Kalichman, and L.S. Weinhardt. HIV–AIDS patients’
evaluation of health information on the Internet: The digital divide and
vulnerability to fraudulent claims. Journal of Consulting and Clinical
Psychology, 72(6), 2004, pp. 1004-1011.
[6] Fogg, B.J., Soohoo, C., Danielson, D., Marable, L., Stanford, J. &
Tauber, E. R. 2002 How do people evaluate a Web site’s credibility:
Results from a large study.
http://www.consumerwebwatch.org/dynamic/web-credibility-reportevaluate.cfm
66
[7] I. Braten, H. I. Stromso, and M. Samuelstuen. The relationship between
Internet-specific epistemological beliefs and learning within Internet
technologies. J. Educational Computing Research, vol. 33(2), 2005, pp.
141-171.
[8] B. K. Hofer, and P. R. Pintrich. The development of epistemological
theories: Beliefs about knowledge and knowing and their relationship to
learning. Review of Educational Research, vol. 67, 1997, pp. 88-140.
[9] C.A. Chinn, and W. F. Brewer. The role of anomalous data in
knowledge acquisition: A theoretical framework and implications for
science instruction. Review of Educational Research, vol. 63, 1993, pp.
1-49
[10] P.R. Pintrich, R.W. Marx, and R. A Boyle. Beyond cold conceptual
change: The role of motivational beliefs and clssroom contextual factors
in the process of conceptual change. Review of Educational Research,
vol. 63, 1993, pp. 167-199.
[11] D. Schunk. Social cognitive theory and sefl-regulated learning. In B.
Zimmerman and D. Schunk (Eds.), Self-regulated learning and academic
achievement: Theory, research and practice, New York: Springer, 1989,
pp. 83-110.

How Teachers in Different Educational Systems Value
Central Concepts of Computer Science
Peter Hubwieser
Technische Universität München
TUM School of Education
Fakultät für Informatik
Boltzmannstr. 3. D-85478 Garching
+49 89 289 17350
Peter.Hubwieser@tum.de
ABSTRACT
The 16 German states exhibit substantial differences regarding the
organization as well as the substantial focus of computer science
education at their schools. This empirical study investigates how
teachers from two German states with different educational systems
assess the value of central concepts of computer science. We
asked 120 teachers in each country to complete our questionnaire,
received 38 responses and applied a specific split-plot design to
evaluate the results. The findings show that the assessments by the
two groups differ regarding the content concepts model, system,
computer, and information. Additionally, we detected differences
in the rating of some individual process concepts (analyzing,
classifying, finding relationships, generalizing, comparing, and
ordering) in relation to the content concept model. These results
are consistent with the differences in the focus of the curricula as
well as with the content of the teacher education programs in the
two states.
Categories and Subject Descriptors
K.3.2 [Computer and Information Science Education ]: Computer
science education.
General Terms
Human Factors
Keywords
computer science education, teacher education, subject domain
knowledge, central concepts.
1. INTRODUCTION
The educational systems of the 16 German states exhibit substantial
structural differences. This applies to the implementation of
computer science education in schools as well as to the teacher
education programs. This empirical study investigates how teachers
from two German states differ in their evaluations of central
concepts of computer science that are given particular emphasis in
the curricula respectively in the teacher training programs of these
states.
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
Conference’10, Month 1–2, 2010, City, State, Country.
Copyright 2010 ACM 1-58113-000-0/00/0010 …$15.00.
Andreas Zendler
University of Education Ludwigsburg
Institute of Mathematics and Computer
Science
Reuteallee 46 D-71634 Ludwigsburg
+49 7141 140 683
zendler@ph-ludwigsburg.de
67
The provision of effective computer science education (CSE) in
schools requires computer science (CS) teachers who are able to
draw on three types of professional knowledge [41], [48], [7, 26],
[1]:
(1) disciplinary knowledge of the subject taught (content
knowledge),
(2) general pedagogical knowledge (pedagogical knowledge),
and
(3) knowledge of how to teach specific subject matter (pedagogical
content knowledge).
A crucial tasks for teacher training programs is to define what the
subject of computer science is and what it is not: “One of the
challenges we face when discussing computer science education is
that the field of computer science seems to progress so quickly
that is difficult even for computer scientists to clearly define its
contents and proscribe its boundaries” [12].
This study builds on different approaches to teacher training
programs (See sections “Theoretical Background” and “Related
Work”). Specifically, we investigated whether and how computer
science teachers differ from the two different German states Baden-Württemberg
(BW) and Bavaria (BY) in their evaluations of
the relationships between central content concepts and central
process concepts of the discipline. We also examine whether
Bavarian computer science teachers evaluate the concepts given in
their training programs (data, model, process, structure, information)
[43] as being of particular importance. By addressing
these research questions, we hope to provide insights that can
inform the pre- and in-service training of computer science teachers.
2. THEORETICAL BACKGROUND
In the last decade the research-based approach to teacher education
has been gaining ground [24], [35], [12], [11] . It is characterized
by three main desiderata:
(1) teachers need in-depth knowledge of recent research advances
in the subject they teach,
(2) teacher training should itself be an object of research,
(3) teachers need to internalize a research-based approach to
their work.
In implementing curricular reforms, moreover, it is important to
consider a bottom-up approach [10], [14] that explicitly draws on
teachers’ knowledge and perspectives: “To understand the role of
teachers with respect to educational reform, it has been suggested
that their beliefs and views (…), or their practical knowledge (…)
be analyzed” (see [10], p. 138). Fincher and Tenenberg [13] as
well as Ni [34] have discussed the application of the bottom-up
approach to training programs for computer science teachers.

Deciding on the subject matter to be taught is a key challenge for
those involved in the design of computer science education [44].
The spectrum currently ranges from training the use of computer
office software to programming courses or to the theoretical solution
of algorithmic problems. Teachers’ uncertainty is evident in
the fact that course content tends to reflect current trends and
developments and to draw on short-lived product knowledge [16].
Given the rapid pace of development in the field of information
technology, however, this knowledge soon becomes obsolete.
Instead, computer science education should equip students with
knowledge and skills that will remain relevant in the longer term,
that they can use in their everyday lives, and that are to some
extent representative of the subject [33].
The contents to be covered in computer science education have
previously been discussed primarily in the context of fundamental
ideas [39], first introduced by [5]. According to [39], a fundamental
idea is a scheme of thinking, action, description, or explanation
that satisfies four criteria: It must be relevant in multiple domains
of a discipline (horizontal criterion). It must be teachable on every
intellectual level (vertical criterion). It must remain relevant in the
longer term (time criterion). And it must be related to everyday
language and/or thinking (sense criterion). Arguments based on
these criteria have been applied also to justify the learning content
covered in the new school subject of computer science in Bavaria.
Additionally, however, it has been postulated that the modeling of
real or planned systems using special representation techniques
from software engineering (e.g., object and class diagrams or data
flow charts) is of particular importance in the context of general
education [19]. Computer science instruction is expected to explain
the basic functioning of computer systems, and the abstract
representation of these systems in the form of such models is
expected to further this aim. Moreover, computer science instruction
is expected to develop students’ ability to structure complex
systems, which is clearly also the goal of such modeling activities
[4].
3. RELATED WORK
Several scientists have proposed catalogues of the basic concepts
or fundamental ideas of computer science [36], [39]; [2]; [8], [27],
[17], [47].
However, these catalogues have a number of shortcomings in
terms of validity:
(1) they are based on the subjective judgments of a single author
or small group of authors,
(2) they lack empirical verification,
(3) they relate only to subdomains of computer science,
(4) their validity has been established only for the national context
in which they were developed.
We have addressed points (1) to (3) in previous research. Specifically,
the results of our three studies [49], [50], [51] were based
on the judgments of a larger sample of experts, were empirically
derived, and related to various subdomains of computer science.
In the current discussion on curriculum development in computer
science education, the combination of two scientifically informed
oriented approaches is considered crucial: first, the structure of the
discipline approach introduced by [5]; second, the process as
content approach, based on the work of Parker and Rubin [37],
which has more recently enjoyed a renaissance thanks to Costa
and Liebmann [6].
In 2003, Meyer and Land introduced the idea of threshold concepts,
core concepts in a discipline that are transformative, irre-
68
versible and integrative [28]. Recently Sorva presented a comprehensive
overview of research about this issue [42], while Sanders
et al. investigated the relations between threshold concepts and
threshold skills in computing [38].
In the study by Zendler, Spannagel, and Klaudt [51], which drew
on these two approaches and on a constructivist theory of learning
[3], [29]. For the study 24 German computer science professors
rated the relevance of 15 content concepts (e.g., algorithm, problem,
and model) with respect to 16 process concepts (e.g., analyzing,
categorizing, and classifying) on a 6-point scale from (“no
importance”) to 5 (“great importance”). The main finding of the
study was that there are specific groups of content concepts that
should be taught in combination with specific groups of process
concepts. In total, 15 blocks of content and process concepts were
identified as being particularly relevant (e.g., the blocks of the
content concepts algorithm, data, information, problem, model,
and structure in combination with the process concepts categorizing,
classifying, finding cause-and-effect relationships, and generalizing).
In 2009 we have investigated the view of active teachers on the
recently installed compulsory subject of informatics in the state of
Bavaria [32], [22]. Regarding the valuation of CS concepts we
found that the teachers could be assigned to one of three different
clusters as far as their valuation of the CS concepts is concerned
[32]:
(1) “office users”, concentrating on the application of software
systems,
(2) “fans of the curriculum“ and “anti-programmers”, preferring
object-oriented modeling (OOM) instead of programming
and algorithmic concepts,
(3) “traditional computer scientists” that focus on traditional
algorithmic views instead of OOM.
4. COMPUTER SCIENCE EDUCATION
4.1 General Education in BY and BW
The educational systems of BY and BW are quite similar, at least
as far as the structure of the general education is concerned (see
Fig. 1). In both states the 4-yeared primary education takes place
in the Grundschule. Following this, the students split according to
their performance level in 3 types of secondary schools:
(1) Gymnasium (8 years),
(2) Realschule (6 years),
(3) Hauptschule (5-6 years).
Figure 1. School System in BW and BY
As our data were gathered among teachers at Gymnasium, we will
restrict the following explanations to this school type, which is

attended by about a 33-40 % of all school students in grade 5
currently.
4.2 CSE in BY and BW
By the occasion of the extensive structural reform project of the
Bavarian Gymnasium in 2004 (from G9 to G8) a new (at least
partly) compulsory subject of CS was incorporated. Fig. 2 displays
the organization of the new subject and the years it has
started in the different grades. In grade 6 and 7, CS is incorporated
formally in the subject combination Nature and Technology
(NuT) that comprises the Biology (in grade 5 and 6), Physics (in
grade 7) and CS (in grade 6 and 7). Nevertheless, all three subjects
are taught separated in a prescribed number of lessons per
week by teachers that need a university degree in the respective
subject.
Figure 2. The subject of CS in Bavarian Gymnasiums
In the Science & Technology direction of study, the students have
to attend CS as a compulsory subject in grade 9 and 10. In average
about 50% of the students choose this direction. In grade 11 and
12, an elective CS course might be chosen instead of a second
natural science or a second foreign language. In the final examination,
CS can be chosen for written as well as for oral examination.
The learning topics of the subject of CS are prescribed in specific
curricula, which we have described in several publications, e.g.
[18], [19], [20].
In contrary to the situation in Bavaria, there is no compulsory
subject of CS in BW. Despite there are working groups beside the
regular schedule up to grade 10. From this point of time, the
students can choose an elective course that is designed according
to the proposed educational standards of the German Gesellschaft
für Informatik [15].
5. TEACHER EDUCATION
Regarding the curricula of regular teacher education programs,
there aren’t big differences between the two states, at least as far
as the teachers at Gymnasiums are concerned. The education takes
place at the Universities in contrary to the teachers for the other
school types in BW, who are educated at specific pedagogical
colleges (Pädagogische Hochschulen). The content of the teacher
education in CS was standardized in 2008 by the German Kultusministerkonferenz
(KMK), the national board of the 16 secretary
of states that are in charge for the schools, which worked out
Standards for Teacher Education [40]. Thus we can assume that
the subject knowledge content of the regular teacher education
courses is not much different in the two states.
Nevertheless, in Bavaria, many of the teachers that are active
teaching currently were educated in specific courses that were
installed in order to support the introduction of the new compulsory
subject. At this time, the dilemma associated with all new
subjects has to be addressed: On the one hand, if teachers are
69
trained before a new subject is introduced, there is a risk that they
will not find employment after graduation. On the other hand, a
new subject can be successfully introduced only if sufficient
numbers of appropriately trained teachers are available. When
computer science was first implemented as a compulsory subject
in academic-track schools (called Gymnasia) in the southern
German state of Bavaria, this dilemma was resolved by providing
specific in-service training programs for teachers who had already
been appointed to teach two other subjects before [43].
These programs emphasized the importance of modeling techniques
in computer science. In particular, the modeling of processes
[21] and object-oriented modeling formed the core of the
SIGNAL and FLIEG in-service-courses attended by the majority
of the currently active computer science teachers in Bavaria, who
have received further training to date. Specifically, the courses
contained the following modules M1 to M8:
- M1: Data modeling and database systems
- M2: Modeling processes
- M3: Object-oriented modeling and programming
- M4: Algorithms and data structures
- M5: Software technology
- M6: Technological computer science, data security
- M7: Theoretical computer science
- M8: The teaching of computer science.
As the content of these specific courses is well-tailored to the
curriculum of the Bavarian subject of CS, there might exist substantial
differences to the regular teacher education in BW.
6. METHODOLOGY
6.1 Study design
The hypothesis was tested in a 2•15×16 split-plot design (3-factor
design with repeated measures of factors B and C, see Fig. 3; [46];
[25]. Factor A comprised the p = 2 groups surveyed, with factor
level a1 representing group G1 of n1 Baden-Württemberg (BW)
teachers of computer science and factor level a2 representing
group G2 of n2 Bavarian (BY) teachers of computer science teachers.
Factor B represented the q = 15 content concepts (CC) b 1, ...,
b15: problem, information, model, algorithm, data, structure,
system, computation, process, software, program, test, communication,
language, and computer. Factor C consisted of the r = 16
process concepts (PC) c 1, ..., c 16: analyzing, classifying, problem
solving and problem posing, categorizing, investigating, finding
relationships, generalizing, creating and inventing, comparing,
finding cause-and-effect relationships, questioning, transferring,
communicating, presenting, collaborating, and ordering.
…
…
…
…
…
…
…
Figure 3. Layout of the 2•15×16 split-plot design

While the factors A, B and C were regarded as independent variables,
the dependent variable was the respondents’ evaluation of
the importance of a specific process concept for a specific content
concept. Ratings were given on a 6-point scale from 0 (“no importance”)
to 5 (“great importance”).
6.2 Power analysis
A power calculation of type II, N being a function of power (1–β),
Δ, and α, was used to determine the necessary sample size for the
2•15×16 split-plot design (see [31]): With a power (1–β) of 0.99,
only large effects (Δ = 0.80) on the dependent variable being
considered significant, and a significance level of α = 0.05, a total
sample of approximately N*= 30 ( n1* = 15 Baden-Württemberg
teachers of computer science teachers, n 2* = 15 Bavarian teachers
of computer science teachers), would be required, based on the
power computations of Mueller and Barton [30] or Mueller et al.
[31] for ε-corrected F tests.
6.3 Operational hypothesis
Given the study design and the above specification of the independent
and dependent variables, the operational hypothesis of the
study can be formulated as follows: “CS teachers from BW differ
from CS teachers from BY in their evaluations of the relations
between central content concepts of computer science (CC, see
section 2.1) and central process concepts of computer science
(PC, see section 2.1), as operationalized by their rating on a sixpoint
scale of the importance of a specific process concept for a
specific content concept."
6.4 Sampling
A total of 120 CS teachers in BW and 120 CS teachers in BY
were contacted and invited to complete a questionnaire pertaining
to computer science concepts. The questionnaire began with a
short introduction, in which the 15 central content concepts and
the 16 central process concepts were listed in tabular form in
alphabetical order. Following this, the q = 15 content concepts and
the r = 16 process concepts were presented in alphabetical order
in a matrix, with the content concepts in the rows and the process
concepts in the columns. Participants were asked to rate the following
statement for each of the 15×16 = 240 cells of the matrix:
Each cell represents a combination of a concept and a process and
requires an integer from 0 (no importance) to 5 (great importance)
indicating the relevance of the combination. Participants filled in
each cell on a 6-point scale from 0 (“no importance”) to 5 (“great
importance”).
To maximize the return rate, we mailed both samples the questionnaires
in sealed, personalized envelopes, enclosing a preaddressed
return envelope franked with stamps showing flower
designs (see [9] for recommendations on increasing return rates).
The return rate for the BW teachers was 14.2% (n1 = 17 valid
questionnaires), which can be considered reasonable for a postal
survey (see [45]). The return rate for the BY teachers was 17.5%
(n2 = 21 valid questionnaires).
6.5 Data Analysis
The procedure used to analyze the experimental data was as follows:
First, we performed a descriptive analysis (7.1-7.2), focusing
on the content concepts. Second, we conducted a three-factor
analysis of variance with repeated measures (7.3) in accordance
with the SPF-2•15×16 split-plot design (see Winer [46], chapter
7). Third, we conducted a posteriori comparisons of means to test
for significant effects of the A × B interaction (7.4) and the A × B
× C interaction (7.5). The process concepts were included in
analyses (2) to (4).
70
Data analyses were conducted using SPSS 17.0; the power analysis
was computed with PASS 8.0.9.
7. RESULTS
Fig. 6 in the appendix displays the original data of the mean ratings
obtained from the BW teachers (a 1) and the BY teachers (a 2)
for each of the 15 × 16 combinations of content concepts × process
concepts (repeated measures factors B × C).
7.1 Means
The four content concepts with the highest averages (see Appendix)
are the same for the two groups of teachers: problem, information,
model and algorithm. The concept with the lowest average
is also the same, namely computer. Major differences in the
assessment of content concepts between the two groups of teachers
can be found for the content concepts information (2.79 vs.
3.11), model (2.62 vs. 3.30), system (1.90 vs. 2.44) and computer
(1.62 vs. 2.04).
7.2 Process-related coverage
To determine differences in the assessment of content and process
concepts by the two groups of teachers, the process-related coverage
can be used: A content concept has high process-related coverage
if it is rated as highly important (> 2.50) for many of the
process concepts; it has lower process-related coverage if it is
rated as less important (≤ 2.50) for many of the process concepts.
It is striking that the content concepts information, model and
system are rated highly in relation to more process concepts by the
BY teachers compared to the BW teachers. On the other hand,
teachers from Bavaria rate more process concepts highly in combination
with the content concepts information (14 vs. 11), model
(15 vs. 12) and system (6 vs. 1) compared to the BW teachers,
while the latter rate more process concepts highly in combination
with computation (8 vs. 3) and program (8 vs. 5).
7.3 Analysis of Variance
To examine whether the BW teachers differed from the BY teachers
in their evaluations of the relationships between the content
concepts and the process concepts, we formulated three statistical
hypotheses, which were tested at the significance level of α =
0.05. The three null hypotheses were chosen as follows:
i) The means of the content concepts µ1 under factor level a1
(BW teachers) and µ2 under factor level a2 (BY teachers) are
equal, such that:
H 0: µ 1 = µ 2.
ii) The means of the content concepts µ11, µ12, ..., µ215 under
the 2 15 levels of the factor combinations A × B are equal,
such that:
H0: µ11 = µ12 = ... = µ215.
iii) The means of the content concepts µ11×1, µ11×2, ..., µ215×16
under the 2 15 ×16 levels of the factor combinations A × B
× C are equal, such that:
H0: µ11×1 = µ11×2 = ... = µ215×16.
For an analysis of variance of a split-plot design, the data must
satisfy the condition of sphericity. This assumption was tested
using Mauchly’s W test for sphericity, with the test statistic W
being compared to a chi-square distribution to assess the adequacy
of the sphericity assumption.

The assumption of sphericity was not met for either the content
concepts (W = 0.003, χ 2 104 = 178.35, p < 0.001) or the process
concepts (W < 0.001, χ 2 119 = 248.60, p < 0.001) at the α level of
0.05. In the further analyses, we therefore applied the ε correction
of degrees of freedom proposed by [23], as presented in table 1.
Table 1. Results of the ANOVA with Huynh–Feldt ε correction
of degrees of freedom
The main effect A (BW vs. BY teachers) was not significant at the
α level of 0.05 (F1, 36 = 0.24, p < 0.63). The corresponding H0 was
therefore not rejected: The teachers from BW respectively BY did
not differ in their global evaluations of the content concepts.
The interaction effect A × B (group × content concept) was significant
at the α level of 0.05 (F10, 360 = 2.26, p < 0.02). The corresponding
H0 was therefore rejected: The teachers from BW respectively
BY differed significantly in their evaluations of individual
content concepts.
The interaction effect A × B × C (group × content concept ×
process concept) was not significant at the α level of 0.05 (F78, 2809
= 1.09, p < 0.29). The corresponding H0 was therefore not rejected:
The teachers from BW respectively BY did not differ in their
evaluations of the relationships between individual content concepts
and individual process concepts.
7.4 Individual Comparisons for the A × B Interactions
The global test of the A × B interaction revealed a significant
overall effect of group × content concept. Therefore we evaluated,
which concepts were rated differently by comparing the mean
values, applying t-tests in order to test simple AB effects for
p•q×r split-plot designs (see [46], pp. 535–536), concerning the ε
correction of the degrees of freedom (see section 7.3). Fig. 4
visualizes the means of the A × B interaction. As the figure shows,
the content concept model was rated significantly different by the
two groups of teachers (a1, a2) at the α level of 0.05 (t396 = 2.23, p
< 0.027). The differences for system, computer, and information
are remarkable, but not significant.
7.5 Individual Comparisons for the A × B × C
Interaction
The global test of the A × B × C interaction did not reveal a significant
overall effect of group × content concept × process concept.
Taking into account the significant difference regarding the
content concept model (see section 7.4), it makes sense to compare
only this concept with respect to the process concepts a
posteriori. Fig. 5 displays the comparisons of the means on the
concept model regarding the different process concepts.
71
These were calculated by applying 16 t-tests to analyze simple
AC effects for SPF-p•q×r experimental designs ([46] pp. 535-
536); An ε correction of degrees of freedom was taken into account
again (see above). The t-tests were calculated at an adjusted
α level of 0.05/16 = 0.0031. It turned out that the ratings from BW
teachers (a 1) respectively from BY teachers (a 2) differed significantly
regarding the content concept model related to the following
process concepts: classifying, finding relationships, generalizing,
comparing, questioning, and ordering.
Content concepts
b1 = problem
b2 = information
b3 = model
b4 = algorithm
b5 = data
b6 = structure
b7 = system
b8 = computation
b9 = process
b10 = software
b11 = program
b12 = test
b13 = communication
b14 = language
b15 = computer
0.0
1.75
Rating
2.0 2.25 2.5 2.75 3.0 3.25 3.5
Groups
a1= BW teachers of computer science
a2= BY teachers of computer science
a1
a2
__
simple AB effects
0.85-
0.63-0.84
0.53-0.62
0.31-0.52
0.16-0.30
0.00-0.15
Figure 4. Comparisons for the factor level combinations A × B
Rating
4.0 4.5
4.25
3.75
2.0 2.25 2.5 2.75 3.0 3.25 3.5
1.75
c1 = analyzing
c2 = classifying
c3 = problem solving and posing
c4 = categorizing
c6 = finding relationships
c5 = investigating
c7 = generalizing
c8 = creating and inventing
c9 = comparing
c10 = finding cause-and-effect r.
c11 = questioning
c12 = transferring
c13 = communicating
c14 = presenting
c15 = collaborating
c16 = ordering
p < .0006
p < .003
p < .006
1.05-
0.91-1.04
0.67-0.90
0.45-0.66
0.23-0.44
0.00-0.22
p < .01
p < .05
p < .10
Figure 5. Comparisons for the content concept model
8. DISCUSSION
The results of the performed evaluations support the research
hypothesis that computer science teachers from Baden-
Württemberg differ from computer science teachers from Bavaria
in the assessment of key content concepts of computer science
related to central process concepts of computer science. Already
from the descriptive evaluation, it has become clear that there are
differences in the assessment of content concepts by the two
groups of computer science teachers. There were differences for
the concepts of content model, system, computer, and information.
The analysis of variance and the individual comparisons showed
that the two groups rated the individual process concepts classifying,
finding relationships, generalizing, comparing, questioning,
and ordering differently with respect to the content concept model.
As the regular teacher education programs in the German states
are standardized by the KMK (see section 5), it is not likely that
those differences would be caused by the courses of lessons in CS
that the teachers had attended at their universities. Therefore, the

eason has to be sought outside the regular programs. In Bavaria
this might have been a specific in-service program like SIGNAL
or FLIEG (see section 5), attended by many Bavarian teachers
instead of regular programs, a specific didactical training programs
for the new Bavarian subject of CS or the daily teaching
environment at schools. It would be plausible to assume one of
those three reasons because all those courses or environments
focus on modeling and thus might cause a quite special view on
the concept model as well as on the salient process concepts, e.g.
classifying or finding relationship.
On the other hand, there are certainly some threats to the validity
of our results. Firstly, the question of the interview could be more
precise, e.g. asking to assess the value of the concepts for CS
courses in school or as background of the teachers. A stimulating
text passage before the question might help also. Secondly, the
teachers had not much time to think about the 240 combinations
of content and process concepts.
9. CONCLUSION AND FUTURE WORK
Suggested by this comparison of two different states, we aim to
conduct a larger study, interviewing teachers in Germany about
their valuation of content and process concepts in the future.
Our results show that there are significant differences in the valuation
of certain concepts between the teachers of those two states.
It would be very interesting to investigate the perceptions of those
concepts by the teachers, trying to detect if the differences concern
only the degree of valuation or even different subject
knowledge about these aspects.
Further, it should be considered to incorporate our results in the
teacher education programs. The first step could be to analyze the
coverage of the salient concepts by those programs. In the case
that there are differences, those should be removed. Additionally,
there might be some fuzziness in the German standards for teacher
education o the KMK regarding those concepts.
10. REFERENCES
[1] Abell, S. K. 2007. Research on science teacher knowledge.
In Handbook of research on science, S. K. Abell and N. G.
Lederman, Eds. Lawrence Erlbaum, London, 1105–1149.
[2] Baumann, R. 1998. Fundamentale Ideen der Informatik –
gibt es das? In Informatische Bildung in Deutschland. Perspektiven
für das 21. Jahrhundert ; Festschrift für Professor
Dr. Wolfgang Arlt aus Anlass seines Eintritts in den Ruhestand,
B. Koerber and W. Arlt, Eds. Log-In-Verl., Berlin.
[3] Ben-Ari, M. 1998. Constructivism in computer science
education. ACM SIGCSE Bulletin 30, 1, 257–261.
[4] Breier, N. and Hubwieser, P. 2002. An Information-
Oriented Approach to Informatical Education. Informatics
in Education 1, 31–42.
[5] Bruner, J. S. 1960. The process of education. [a searching
discussion of school education opening new paths to learning
and teaching]. A Vintage book V-234. Vintage Books,
New York NY.
[6] Costa, A. L. and Liebmann, R. M. 1997. The Processcentered
school. Sustaning a renaissance community. Corwin
Press, Thousand Oaks, Calif.
[7] Davis, E. A. 2004. Knowledge integration in science teaching:
Analyzing teachers’ knowledge development. Research
in Science Education, 34 (2004), 21–53.
[8] Denning, P. J. 2003. Great principles of computing. Communications
of the ACM, 46(11) (2003), 15–20.
72
[9] Dillman, D. A. 2000. Mail and Internet surveys: the tailored
design method. Wiley, New York.
[10] Driel van, J. H., Beijaard, D., and Verloop, N. 2001. Professional
development and reform in science education: the
role of teachers' practical knowledge. Journal of Research
in Science Teaching, 38(2) (2001), 137–158.
[11] EC – European Commission. 2009. Common european
principles for teacher competences and qualifications. Accessed
10 August 2009.
[12] Ericson, B. 2008. Ensuring exemplary teaching in an essential
discipline. Addressing the crisis in computer science
teacher certification. ACM, New York.
[13] Fincher, S. and Tenenberg, J. 2007. Warren's questions. In
Proccedings of Third International Computing Education
Research Workshop, ICER, Ed. ACM, New York, 51–60.
[14] Fishman, B. and Davis, E. 2006. Teacher learning research
and the learning sci-ences. In The Cambridge handbook of
the learning sciences, R. K. Sawyer, Ed. Cambridge University
Press., New York, 535–550.
[15] Gesellschaft für Informatik e. V., Ed. 2008. Grundsätze und
Standards für die Informatik in der Schule. Bildungsstandards
Informatik für die Sekundarstufe I. Empfehlungen der
Gesellschaft für Informatik e. V., erarbeitet vom Arbeitskreis
»Bildungsstandards«, Bonn.
[16] Hartmann, W. and Nievergelt, J. 2002. Informatik und
Bildung zwischen Wandel und Beständigkeit. Informatik
Spektrum 25, 6, 465–476.
[17] Hartmann, W., Näf, M., and Reichert, R. 2006. Informatikunterricht
planen und durchführen. eXamen.press. Springer,
Berlin.
[18] Hubwieser, P. 2003. Object Models of IT-Systems: Supporting
Cognitive Structures in Novice Courses of Informatics.
In Informatics and the digital society: Social, ethical
and cognitive issues ; IFIP TC3/WG 3.1 & 3.2 Open
Conference on Social, Ethical, and Cognitive Issues of Informatics
and ICT, July 22 - 26, 2002, Dortmund, Germany.
IFIP / International Federation for Information Processing.
Kluwer Academic Publishers, Boston, Mass.,
129‐140.
[19] Hubwieser, P. 2006. Functions, Objects and States: Teaching
Informatics in Secondary Schools: Invited talk. In Informatics
Education - The Bridge between Using and Understanding
Computers. Lecture notes in computer science.
Springer.
[20] Hubwieser, P. 2007. A smooth way towards object oriented
programming in secondary schools. In Informatics, Mathematics
and ICT: A golden triangle: Proceedings of the
Working Joint IFIP Conference: WG3.1 Secondary Education,
WG3.5 Primary Education; College of Computer and
Information Science, Northeastern University Boston, Massachusetts,
USA 27th ‐ 29th June 2007, D. Benzie and M.
Iding, Eds., Boston,MA.
[21] Hubwieser, P. and Aiglstorfer, G. 2004. Fundamente der
Informatik: Ablaufmodellierung, Algorithmen und Datenstrukturen.
Oldenbourg, München.
[22] Hubwieser, P. and Mühling, A. 2009. Ausgewählte Ergebnisse
einer Befragung von Lehrkräften zum Unterrichtsfach
Informatik in Bayern. durchgeführt im November/Dezember
2009 als Online-Umfrage unter Lehrkräften, die an bayerischen
Gymnasien Informatik unterrichten.
http://www.ddi.edu.tum.de/fileadmin/tueds10/www/materia
l/Forschung/Lehrerbefragung_2009/Lehrerbefragung-2009-
Abschlussbericht.pdf. Accessed 29 February 2012.

[23] Huynh, H. and Feldt, L. S. 1976. Estimation of the Box
correction for degrees of freedom from sample data in randomised
block and split-plot designs. Journal of Educational
Statistics, 1 (1976), 69–82.
[24] Kansanen, P. 2006. Constructing a research-based program
in teacher education. In Competence oriented teacher training,
F. Oser, Achtenhagen F. and U. Renold, Eds. Sense
Publishers, Rotterdam, 11–22.
[25] Kirk, E. 1994. Experimental design. Wadsworth, Belmont.
[26] Linn, M. C., Eylon, B.-S., and Davis, E. A. 2004. The
knowledge integration perspective on learning. In Internet
environments for science education, M. C. Linn, E. A. Davis
and P. Bell, Eds. Erlbaum, Mahwah, NJ, 29–46.
[27] Loidl, S., Mühlbacher, J. R., and Schauer, H. 2005. Preparatoy
knowledge: propaedeutic in informatics. In From computer
literacy to informatics fundamentals, Proceedings of
international conference on informat-ics in secondary
schools—evolution and perspectives, R. T. Mittermeir, Ed.
Springer, New York, 104–115.
[28] Meyer, J. H. F. and Land, R. 2003. Threshold concepts and
troublesome knowledge 1: Linkages to ways of thinking
and practising within the disciplines. In Improving Student
Learning – Ten Years On., C. Rust, Ed. OCSLD, Oxford.
[29] Moreno, L., González, C., Castilla, I., González, E. J., and
Sigut, I. 2007. Use of constructivism and collaborative
teaching in an ILP processors course. IEEE Transactions on
Education 50, 2, 101–111.
[30] Mueller, K. E. and Barton, C. N. 1989. Approximate power
for repeated-measures ANOVA lacking sphericity. Journal
of the American Statistical Association, 84(406) (1989),
549–555.
[31] Mueller, K. E., LaVange, L. E., Ramey, S. L., and Ramey,
C. T. 1992. Power calcu-lations for general linear multivariate
models including repeated measures appli-cations.
Journal of the American Statistical Association, 87(420)
(1992), 1209–1226.
[32] Mühling, A., Hubwieser, P., and Brinda, T. 2010. Exploring
Teachers‘ Attitudes Towards Object Oriented Modelling
and Programming in Secondary Schools. In ICER ‘10: Proceedings
of the Sixth international workshop on Computing
education research. ACM, New York, NY, USA, 59–68.
[33] Müller, P. and Hubwieser, P. 2000. Informatics as a Mandatory
Subject at Secondary Schools in Bavaria. In Open
Classrooms in the Digital Age, EDEN, Ed., 129‐140.
[34] Ni, L. 2009. What makes CS teachers change? Factors
influencing CS teachers' adoption of curriculum innovations.
SIGCSE Bulletin, 41(1) (2009), 544–548.
[35] Niemi, H. and Jakku-Sihvonen, R. 2006. Research-based
teacher education. In Research-based teacher education in
Finland, R. Jakku-Sihvonen and H. Niemi, Eds. Finnish
Educational Research Association, Turku, 31–50.
[36] Nievergelt, J. 1990. Computer Science for Teachers: A
quest for classics and how to present them. In Computer Assisted
Learning, D. H. Norrie and Six H. W., Eds. Lecture
notes in computer science 438. Springer, Berlin, 2–15.
[37] Parker, J. C. and Rubin, L. J. 1966. Process as content.
Curriculum design and the application of knowledge. Rand
McNally, Chicago.
[38] Sanders, K., Boustedt, J., Eckerdal, A., McCartney, R.,
Moström, J. E., Thomas, L., and Zander, C. 2012. Threshold
concepts and threshold skills in computing. In Proceed-
73
ings of the ninth annual international conference on International
computing education research. ICER ’12. ACM,
New York, NY, USA, 23‐30.
[39] Schwill, A. 1994. Fundamental ideas of computer science.
Bulletin European Association for Theoretical Computer
Science 53, 274–274.
[40] Sekretariat der ständigen Konferenz der Kultusminister der
Länder in der Bundesrepublik Deutschland, Ed. 2008. Ländergemeinsame
inhaltliche Anforderungen für die Fachwissenschaften
und Fachdidaktiken in der Lehrerbildung. Beschluss
der Kultusministerkonferenz vom 16.10.2008 i. d. F.
vom 16.09.2010, Berlin.
[41] Shulman, L. S. 1986. Those who understand: Knowledge
growth in teaching. Educational Researcher, 15(2) (1986),
4–14.
[42] Sorva, J. 2010. Reflections on threshold concepts in computer
programming and beyond. In Proceedings of the 10th
Koli Calling International Conference on Computing Education
Research. Koli Calling ’10. ACM, New York, NY,
USA, 21‐30.
[43] Spohrer, M. 2009. Konzeption und Analyse neuer Maßnahmen
in der Fort- und Weiterbildung von Informatiklehrkräften.
Dissertation, Technische Universität.
[44] Stephenson, C., Gal-Ezer, J., Haberman, B., and Verno, A.
2005. The New Educational Imperative: Improving High
School Computer Science Education. Using worldwide research
and professional experience to improve U. S.
Schools. Final Report of the CSTA Curriculum Improvement
Task Force. ACM, CSTA, New York.
[45] Vaux, A. and Briggs, C. S. 2005. Conducting mail and
Internet surveys. In The psychology research handbook, F.
T. L. Leong and J. T. Austin, Eds. Sage, Thousand Oaks,
CA, 186–209.
[46] Winer, B. J., Brown, D. R., and Michels, K. M. 1991. Statistical
principles in ex-perimental design. McGraw-Hill,
Boston.
[47] Wursthorn, B. 2005. Fundamental concepts of computer
science in a Logo-environment. In Digital tools for lifelong
learning. Proceedings of the tenth European Logo Conference
(pp. 219-227). Warsaw, Poland Centre for Informatics
and Technology Education inDigital tools for lifelong
learning., G. Gregorczyk, A. Walat, W. Kranas and M.
Borowiecki, Eds., Warsaw, 219–227.
[48] Zeidler, D. L. 2002. Dancing with maggots and saints:
visions for subject matter knowledge, pedagogical
knowledge, and pedagogical content knowledge in science
teacher education reform. Journal of Science Teacher Education,
13(1) (2002), 27–42.
[49] Zendler, A. and Spannagel, C. 2008. Empirical foundation
of central concepts for computer science education. ACM
Journal on Educational Resources in Computing, 8(2)
(2008), Article No. 6.
[50] Zendler, A., Spannagel, C., and Klaudt, D. 2008. Process as
content in computer science education: empirical determination
of central processes. Computer Science Education,
18(4) (2008), 231–245.
[51] Zendler, A., Spannagel, C., and Klaudt, D. 2011. Marrying
content and process in computer science education. IEEE
Transactions on Education 54, 3, 387–397.

APPENDIX
Groups
a1 = BW teachers of computer science
a2 = BY teachers of computer science
a1
a2
Process concepts
c1 = analyzing
= finding relationships
Content concepts c6
b1 = problem
b2 = information
b3 = model
b4 = algorithm
b5 = data
b6 = structure
b7 = system
b8 = computation
b9 = process
b10 = software
b11 = program
b12 = test
b13 = communication
b14 = language
b15 = computer
grand means
b1 = problem
b2 = information
b3 = model
b4 = algorithm
b5 = data
b6 = structure
b7 = system
b8 = computation
b9 = process
b10 = software
b11 = program
b12 = test
b13 = communication
b14 = language
b15 = computer
grand means
c2 = classiyfing
c3 = problem solving and posing
c4 = categorizing
c5 = investigating
c7 = generalizing
c8 = creating and inventing
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
c9 = comparing
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
Figure 6. Means for the 2•15×16 split-plot design (n1 = 17; n2 = 21)
74
c10 = finding cause-and-effect r.
c11 = questioning
c12 = transferring
c13 = communicating
c14 = presenting
c15 = collaborating
c16 = ordering
grand means
3.21
2.79
2.62
2.92
2.56
2.74
1.90
2.38
2.40
2.18
2.57
2.20
2.55
2.22.
1.62
3.09
3.11
3.30
2.70
2.61
3.03
2.45
2.20
2.41
2.17
2.40
2.35
2.48
2.22
2.04
2.57

Ways of Planning Lessons on the Topic of Networks and
the Internet
ABSTRACT
Ana-Maria Mesaro¸s
University of Oldenburg
Computer Science Education
26111 Oldenburg, Germany
ana.maria.mesaros@uni-oldenburg.de
Changing conditions and different educational concepts are
wide-spread problems of Computer Science education. Due
to this fact suitable teacher training can only be developed
if the preconditions of CS teachers are being considered. In
order to find out about teachers’ subjective theories on planning
lessons we asked them how they would plan lessons on
the specific topic of networks and the Internet. In this paper
we show the research framework for examining this question
and describe the organization of our study and first results.
The different ways of planning lessons for this specific topic
we found and describe here differ enormously.
Keywords
subjective theories, educational reconstruction, planning lessons
1. INTRODUCTION
Without a basis of generally accepted standards and with
varying qualifications Computer Science (CS) teachers make
their own decisions on what topics to teach and how to teach
them. This leads to a situation in which the planning of
lessons on a specific topic differs greatly from teacher to
teacher. The reasons for this are different subjective theories
about CS and different ideas about teaching approaches.
For those who want to provide effective teacher training
like us those subjective theories are very important to know.
To gain this basic knowledge, we conducted a survey on the
question on how teachers would plan a lesson on the specific
topic of networks and the Internet.
Parts of this research have already been introduced in [9]
where we reported on a pre-test. We therefore confine ourselves
here to giving only a short introduction to the research
framework. The focus of this paper is on the different possibilities
to plan lessons on the topic of networks and the
Internet. Therefore, our next step is to describe this framework
shortly, the design of the interview and the sample. We
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WiPSCE 2012 Hamburg, Germany
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$10.00.
75
Ira Diethelm
University of Oldenburg
Computer Science Education
26111 Oldenburg, Germany
ira.diethelm@uni-oldenburg.de
then present the results achieved so far: How the planning
of lessons defers in the reasons for teaching this topic, in
the contents of the topic, in the structuring the content, in
the learning objectives pursued with it and the used teaching
methods and teaching material. We also report on the
different ways of integrating students’ perspectives in the
planning of lessons. Finally we draw our conclusions and
give a preview of our future work.
2. EDUCATIONAL RECONSTRUCTION
FOR TEACHER EDUCATION
The research questions presented above are embedded in
the framework of Educational Reconstruction for Teacher
Education which shows the significance of teachers’ subjective
theories on the planning of their lessons.
This model is based on the model of Educational Reconstruction.
The main idea is that to plan lessons means to
structure a content considering the students’ perspectives
[6]. Therefore the scientific content has to be broken down
and rearranged, so that the students’ perspectives are taken
into account. This study is based on the adaptation of the
model of Educational Reconstruction for Teacher Education.
This adaptation, see fig. 2, contains some changes in
Figure 1: Educational Reconstruction for Teacher
Education [6]
comparison with the original one. The content for teacher
training is defined by the scientific view of a topic and by
the domain-specific educational concepts. As in the original
model of Educational Reconstruction the content has to

include learners’ perspectives. While talking about Teacher
Education, the teacher himself becomes the learner. This
implies that teachers’ perspectives have to supplement the
content given by the educational concepts. Teachers’ perspectives
of planning lessons in CS are based on their subjective
theories. Based on Kelly [5] and Groeben [4] we define
subjective theories as individual cognitive structures of
self- and worldviews that have the function to explain and
to predict.
This model must be understood as an iterative one. The
educational concepts have to be structured anew once the
subjective theories of teachers are known. Similarly the subjective
theories can be understood better, if the educational
concepts are used as an aid. When developing guidelines for
teacher training one has to make sure that these educational
concepts and subjective theories are taken into account.
Therefore, there are several steps to achieve the goal of
creating guidelines for Teacher Education in CS. In the following
chapter we show by the design of our investigation
how we will achieve it.
3. EXPLORING TEACHERS’ SUBJECTIVE
THEORIES
The intention of this study is to explore subjective theories
of CS teachers on planning CS lessons. The method we used
was a semi-structured guided interview. For our aim it is
important to detect a wide range of subjective theories of
CS teachers, which calls for a qualitative research approach.
In order to increase the quality of the answers and for a
better comparison we decided to do the investigation on a
specific topic. For several reasons we decided on the topic
of networks and the Internet. This topic does not only have
high relevance in students’ everyday life [8], but the idea
of communication is one of Dennings’ principles [1] and it is
also mentioned in the Education Standards for CS published
by the German Association of CS [3]. Additionally, students’
perspectives have been investigated lately on this topic [2]
and we can use these results for our investigation.
In the following chapter we will describe the structure of
the interview and the sample.
3.1 The Structure of the Interview
For the collection of data we chose a semi-structured guided
interview. This means that there are set questions, but
they may be asked in a different order or put in a different
way. During the interview a communicative confirmation
took place. The interviewer repeated the given answer in
his own words and asked the teacher if he had understood
him correctly.
As an introduction we asked the teacher how he or she
had planned the past school year. This question does
not only provide an overview of the topics the teacher has
taught in the past year, but it directs the focus on the (professional)
expertise of the teacher. In addition, this question
shows how the teacher plans other topics in CS and thereby
helps to generalize the results of this study.
In the following, questions were asked about the planning
of lessons on the Internet. Therefore we asked
if this topic has been taught, and how relevant this topic
is. Afterwards we asked in which grade they would teach
this topic and which contents it would contain. Questions
on the learning objectives, the used methods and materials
76
followed. We will come to the answers of this section later.
The next set of questions concerned students’ perspectives.
Teachers were asked for their assessment of their
students’ understanding of the functioning of the Internet.
They were also asked if they considered it important to
know their students’ perspectives and which part this aspect
played in planning the lessons. Since during a previous pretest
most teachers had not been able to imagine students’
perspectives, the interviewer read a student’s statement on
how the internet works [2] to them.
The following questions focus on the other subjects of
teachers because a teacher in Germany normally teaches
two subjects and not just one like in many other countries.
With these answers we understand better in which ways CS
differs from other subjects in the teachers’ perception. This
might be important for the concepts for teacher training.
The interview ends with questions about the teachers’
biography. After describing the methods for the data collection
we will go on to give some information about the
sample.
3.2 The Sample
For the aim of discovering different subjective theories
of CS teachers on planning lessons, not the frequency is
important but the variety of theories. Our focus was on
questioning teachers with possibly different subjective theories.
Therefore, we selected teachers with varying experience,
those who had been teaching for thirty years and
others who had just started a few weeks before.
We also interviewed teachers with different qualifications.
Some of them had an official certificate for teaching CS,
others just taught the subject because they were interested
in it or because they have been asked to. Not all teachers
had studied CS at university.
We also took the different school types in Germany into
account and chose teachers from all types of secondary schools
in which CS is being taught. All these characteristics led
to a heterogeneous group of fifteen teachers whose different
subjective theories we set out to examine.
3.3 Evaluation of the Interviews
The interviews took place in school and were recorded.
All interviews were transcribed. The transcription only included
the spoken word without showing differences in intonation.
We considered this way of transcription sufficient
for our objective. The transcribed interviews were analyzed
in several steps. First they were coded with the method
of qualitative content analysis [7]. In a first step the categories
of codings were built in a deductive way from the
interview structure. The second coding was an inductive
one. Therefore the given categories from the interview were
supplemented by new categories, taken from the data.
In the following, the results of this two first coding steps
are shown.
4. DIFFERENT WAYS OF PLANNING
LESSONS
In the following we describe the different approaches to
teaching the topic networks and the Internet.
4.1 Reasons for the Topic
Most of the interviewed teachers did not handle this topic
in class or only included some aspects of it in lessons on some

other topic. The reasons given varied considerably. Some of
the teachers who did teach this topic considered it to be
important because it is very interesting for the students.
Other teachers gave various reasons for not teaching this
topic. Some of them thought that the students were not
ready for this difficult topic yet or that they were not interested.
Others considered it more important to do some
programming or teach the topics needed for the university
entrance qualification. Some teachers just said that they did
not plan to teach this topic in their school and others explained
that their knowledge of this topic was not sufficient.
But most of them thought that the topic of networks and
the Internet was relevant for students, because they use it
daily and it is a part of general knowledge to know how the
Internet works in their opinion.
An explanation for that might be that teachers find this
topic important but their own knowledge is not sufficient.
When we asked them how they felt about their knowledge
of networks and the Internet most of them expressed some
uncertainty. They did have ideas about how to teach the
topic even though they had not done so, yet.
The answers on how many lessons should be dedicated
to this topic showed that it might be a topic with lots of
contents. Nearly all of the teachers considered a lot of time
necessary for teaching this topic. The answers varied from
nine periods of 90 minutes to half a school year.
The grade in which they would teach this topic differs as
well. Some teachers would teach networks and the Internet
in the sixth or seventh grade, but would focus on using
the Internet and learning about the hazards. Others would
teach it from the ninth grade onwards and consider different
contents as important.
4.2 Contents of the Topic
The answers on which topics teachers think should be
taught in lessons about network and the Internet can be
classified in four groups. The first group consists of all the
topics concerning the function or the history of the Internet.
Included in this category is the question on how the Internet
works. Almost all teachers would include the function of the
Internet.
Another topical group is programming/creating something
in the Internet. Teachers would teach how to create a personal
homepage. They would teach HTML or PHP, as an
method how to create an individual part of the Internet.
Another theme would be different network technologies,
like emails. This group also contains cryptology. Here teachers
would describe how an email gets from the sender to the
recipient and what kind of encryption can be used.
Yet another topic is the utilization of the Internet. Themes
like social networks and privacy are part of that. This content
seems to be important to CS teachers, because nearly
all of them mentioned it.
4.3 Structuring the Content
The different possibilities of introducing the subject depend
on the topics the teacher has decided to teach: like the
function of the Internet, the creation of a personal homepage,
social networks, data security or privacy. Some of the
teachers would want their introduction to be problem-based.
Therefore they would present a real problem to be solved by
the students during the lessons. Another possibility would
be to start with a problem of current interest.
77
Other teachers have no plans for structuring lessons, they
are very flexible and prepared to structure the topics spontaneously.
They are guided by the students’ interests and
decide together on the content of the lessons. Some of them
would introduce the topic of networks and the Internet by
asking students how they imagine the Internet works, or by
asking them what contents they think are important to learn
about this topic. Other teachers are convinced that students
do not have a perspective on this topic, so they would begin
the lessons without taking students’ perspectives into account.
If they should later realize that the students do have
some perspectives on this topic, they would consider them.
4.4 Learning Objectives
We also asked about the learning objectives teachers pursued
with their lessons. We found out that the objectives
named can be classified in four groups. The first group is
about content knowledge. Students should know how the
Internet works, they should know something about the history
of the Internet and they should know some cryptology
algorithms.
Another group is about comprehension. Students should
understand how the Internet works so that they can use it
better, and they should also understand what was wrong
about their perspectives. This group of learning objectives
can be characterized very well by a statement one teacher
made: ”‘Students should understand what they use”’.
Another group is about using the Internet/social networks
correctly. The last group of learning objectives is about soft
skills, like teamwork or presentation skills.
4.5 Teaching Methods
Teachers use different methods to achieve their learning
objectives. Some of them prepare the content and just tell
the students how everything works. Other teachers inquire
about students’ interests and plan their lessons based on
that information. In this case topics and methods are based
on students’ interests.
Most of the teachers use the method of projects. Students,
mostly in groups, have to do their own research on a topic,
understand it, prepare it for presentation and of course do
the presentation. This is the opposite of teachers’ input,
because it is a very student-based method.
Other teachers explained that visualization is very important
for this topic. By visualization teachers mean different
things. It might be a kind of game, or a simulation where
the students act as part of a network and can see how the
data gets from one computer to another. Visualization can
also mean watching a movie in which the pathway of the
data through the network is shown. Furthermore visualization
can also mean to have a look at the real network, such
as the school network or the network at home.
4.6 Teaching Materials
On the question about the materials used in the lessons
the interviewed teachers did not only give examples of what
kind of materials they would use, but also mentioned the
sources of the material. Some teachers use their own knowledge
and creativity as a source and produce their material
themselves. Others fall back on prepared materials and just
use them. Some of the interviewed teachers had no idea
where they would get the materials for the topic of networks
and the Internet.

Some teachers would use applets or software to do a simulation
of a network or to show the pathway of the data
through the Internet. Others would use films as a kind of
visualization. Some would use newspaper articles to describe
a problem that could then be handled in class. Worksheets
are also popular materials.
4.7 Integrating Students’ Perspectives
Most teachers agreed that students’ perspectives are important
in planning the lessons. Only a few said that students’
perspectives are not relevant. They are convinced
that students are not interested in finding out how the Internet
works. One teacher compared the students with fish
which do not need to know what water is.
Those teachers who agree that students’ perspectives have
to be taken into account, claim to observe students’ perspectives
during classes and to take them into consideration.
About half of the teachers said that the lessons they plan
would change the students’ conceptions of the Internet.
We also asked teachers to describe their students’ conceptions
of the Internet. Teachers think that students imagine
the Internet...
• to be like a Super-Computer.
• to be like a cloud that surrounds the whole world, and
you can connect everywhere with this cloud.
• to be like a cupboard filled with information
• to consist of unstructured connections between the computer
to have something like rays to a radio tower or
a satellite
• to function like a phone call
5. CONCLUSIONS AND FUTURE WORK
In the previous chapter we discussed the different teaching
methods and learning objectives which we discovered in
our survey. These answers only refer to that part of the interviews
dealing with the question of the planning of lessons
on the topic of networks and the Internet.
An interesting result is that most teachers think that networks
and the Internet is an important topic to teach, but
for various reasons they do not do it. One widespread reason
seems to be the problem of not having enough content
knowledge about this topic. This problem can fortunately
be solved easily by developing teacher training courses.
An obvious result is that teachers have a clear idea on how
to plan lessons, although they feel a lack of knowledge about
this topic. Their pedagogical knowhow seems to help them
in overcoming this lack of knowledge. This might be important
in developing teacher training. It shows that teachers
rather need training on content knowledge than on pedagogical
knowhow.
An astonishing result for us was the fact that in CS it
seems to be usual that students were allowed to choose the
topics they wanted to learn. This can probably be explained
by the lack of generally accepted standards and by the teachers’
insufficient confidence in their own knowledge. But some
teachers also mentioned that they thought their students
might know more about CS than they did themselves and
that might explain why students were allowed to make decisions
on the content of their lessons.
78
Contrary to our expectation teachers claim to include students’
perspectives in planning lessons. It is not unlikely
that they may just have given the expected answer. Another
explanation is a presentation about students’ perspectives
that some of the teachers of our sample attended.
The different ways of planning lessons we discussed have
an internal structure, a kind of a pattern which expands over
the other topics taught by the teachers. These patterns are
formed by the subjective theories of teachers on planning CS
lessons. Therefore our next step will be to discover the types
of planning CS lessons, based on this subjective theories.
These types of planning lessons determine not only the topic
networks and the Internet but any kind of lesson planning.
By using these types, our final objective is the development
of guidelines for teacher training. We do not only want
to show in which different ways teachers plan their lessons
but also how to include this knowledge into teacher trainingfor
the benefit of all teachers. Therefore, the differing ways
of planning presented here are a big step towards this goal.
6. REFERENCES
[1] P. Denning. Great principles of computing.
Communications of the ACM, 46(11):15–20, November
2003.
[2] I. Diethelm and S. Zumbrägel. An investigation of
secondary school students’ conceptions on how the
internet works. In Proceedings of the 12th Koli Calling
International Conference on Computing Education
Research, Koli, Finland, 2012. (accepted).
[3] Gesellschaft für Informatik (GI) e. V. Grundsätze und
Standards für die Informatik in der Schule. Number
28,150/151 in LOG IN. Berlin, 2008.
[4] N. Groeben, D. Wahl, J. Schlee, and B. Scheele.
Forschungsprogramm Subjektive Theorien. Eine
Einführung in die Psychologie des reflexiven Subjekts.
A. Francke Verlag, Tübingen, 1988.
[5] G. A. Kelly. The psychology of personal constructs.
Routledge, London, 1991.
[6] M. Komorek and U. Kattmann. The model of
educational reconstruction. In S. Mikelskis-Seifert,
U. Ringelband, and M. B. (Eds.), editors, Four Decades
of Research in Science Education - from Curriculum
Development to Quality Improvement, chapter 7, pages
171–188. Waxmann, 2008.
[7] P. Mayring. Qualitative content analysis. Forum
Qualitative Sozialforschung / Forum: Qualitative Social
Research, 1(2):10, 2000.
[8] Medienpädagogischer Forschungsverbund Südwest. JIM
2010 Jugend, Information, (Multi-)Media.
Forschungsberichte. Baden-Baden, 2010.
[9] A.-M. Mesaros and I. Diethelm. Exploring computer
science teachers’ subjective theories on designing their
lessons. In Proceedings of the 5th International
Conference on Informatics in Schools: Situation,
Evolution and Perspectives ISSEP, Selected Papers,
Bratislava, Slovenia, 2011.

Promoting Computational Thinking with Programming
Cynthia C. Selby
University of Southampton
Highfield
Southampton UK
44 (0) 2380 593475
ABSTRACT
The term computational thinking has received some discussion in
the field of computer science education research. The term is
defined as the concept of thinking about problems in a way that
can be implemented in a computing device. Of course, after
having thought about a problem using computational thinking
skills, the next step should be to use programming skills to
implement the solution. This work in progress is exploring ways
in which programming can be employed as a tool to teach
computational thinking and problem solving. Data is collected
from teachers, academics, and professionals from various
industries. They are purposively selected because of their
knowledge of or interest in the topics of problem solving,
computational thinking, and the teaching of programming. This
data is analyzed within the paradigm of the grounded theory
approach. The results of an initial analysis imply an ordering of
complexity associated with computational thinking skills, imply
connections between computational thinking skills and
programming activities, and imply a relationship between
computational thinking skills and other taxonomies of learning.
Categories and Subject Descriptors
K.3.2 [Computers and Education]: Computers and Education,
Curriculum
General Terms
Design, Experimentation, Theory
Keywords
Computational thinking, pedagogy of programming, problem
solving
1. INTRODUCTION
Problem solving skills are used in developing or implementing
strategies to solve problems in many domains. These skills are
often expressed as heuristics [12], appropriate and plausible
approaches to a problem. Effective problem solving has been
promoted by the use of strategies including means-ends analysis,
schema acquisition, algorithmic approaches, and targeted
frameworks [19, 12, 9].
C.Selby@soton.ac.uk
79
However, in the domain of computer science, some research [11]
has found that learners do not naturally solve problems in ways
that can be translated to computing devices by the use of
programming. This disparity is highlighted in the Lister study
[6], where it is suggested that ineffective problem solving skills,
including the ability to work through lines of logic, may be the
cause of ineffective programming skills. Additional studies [15,
13] suggest that problems in learning to program are exacerbated
by a lack of strategic tools.
The strategic tools, identified as useful for those attempting to
solve problems with the aid of computational devices in various
domains, include, but are not limited to, decomposition,
abstraction, simulation, and generalization [10]. The name given
to these specialized mental skills, resulting in solutions to
problems directly translatable to a computing device, is
computational thinking [21]. Actually implementing these
solutions requires a different set of skills.
Programming skills are the specific technical skills needed to
produce specific solutions using a set of defined digital tools,
often associated with a programming language [9]. Research
often reports that learners struggle with programming skills such
as tracing [3, 6] and understanding a model of the machine [2, 9].
Having recognized this issue, other researchers [7, 20] highlight
the need for a defined hierarchy of programming skills. One
study [16] attempted to provide such a hierarchy for objectoriented
programming. The researchers found that teachers
interpreted the hierarchy as a capability hierarchy.
Given that a hierarchy of generic programming skills could be
developed and interpreted as capability, the levels could be
aligned with existing hierarchies, such as the cognitive domain of
Bloom’s Taxonomy. These same programming skills could be
mapped to the higher-level computational thinking skills that they
evidence, thereby defining a hierarchy of computational thinking
skills. This setting provides the context for an ongoing
investigation into the relationship between the teaching of
programming and its effect on the acquisition of computational
thinking skills by learners.
2. STUDY METHOD
This study is based on a grounded theory approach employing
qualitative data collection methods and qualitative data analysis
techniques. The first activity is the administration of an Internet
based questionnaire. The second activity is the collection of data
from an Internet based community of practice forum. The third
activity is the administration of a face-to-face, audio recorded,
semi-structured interview schedule for respondents previously
identified by an analysis of the questionnaire results and
community of practice discussions. All data is iteratively
augmented and analyzed guided by Strauss and Corbin’s [18]
grounded theory procedures and techniques until theoretical

saturation. It is anticipated that a product of the theory generation
may be a model of relationships between problem solving skills,
computational thinking skills, and programming skills.
2.1 Participants and Sampling
The participants in this research all have some interest in the
teaching of programming, computational thinking, problem
solving, or any combination of the three. Not all participants are
teachers. Participants may be employed in industries where
computational thinking skills and programming skills are useful
or required. Other participants may be members of professional
communities of practice, representing industry, academia, or
education. They are still perceived to have an interest in and
appropriate knowledge of the research context.
An individual participant may not engage with every data
collection instrument. Participants are matched to instruments. In
the case of the first instrument, an online questionnaire, the
targeted sample consists of members of organizations whose
ideologies promote the teaching of programming or
computational thinking skills. In the case of the second, the
online community of practice, conversation threads are chosen
purposively for their applicability to the context of this research,
without regard to the identity of the poster. From the
questionnaire responses and the community of practice
conversations, a further purposive selection is made to identify
participants for the interviews. This purposive sampling is
supported by Strauss and Corbin who affirm that theoretical
sampling is a foundation stone of grounded theory which, “…
enables the researcher to choose those avenues of sampling that
can bring about the greatest theoretical return” ([18], p. 202).
2.2 Data Collection
The questionnaire and interview schedule have been designed
specifically to elicit responses applicable to the topics of problem
solving, computational thinking, and the teaching of
programming. To ensure the same level of appropriateness of
response, a set of keyword criteria has been developed on which
the community of practice messages are searched.
2.2.1 Online Questionnaires
The questionnaire makes use of some closed questions but the
majority of questions are open-ended to allow participants to
respond as they wish. The ordering of the questions is from
general to specific, divided into major sections. Results are
submitted one screen or page at a time. In this way, even the
results of abandoned questionnaires have the potential to be used.
Personal information is requested early in the response process to
identify participants. This provides a mechanism for contacting
the participant, should he or she be selected for an interview. The
design of the resulting questionnaire aims to be as open as
possible to facilitate depth of response, while controlling for
researcher and question bias.
2.2.2 Community of Practice
The community of practice, whose discussions and opinions are
of interest in this study, is computer-mediated. Simply by
contributing, the members signify some interest in the topics that
overlap with this study. Although computer-mediated, some
individuals share collaborative practices in the classroom. There
are also face-to-face meetings, of varying scale, held throughout
the year.
In order to identify the most appropriate threads for inclusion in
the dataset, discussions are keyword searched. The keywords
80
have been chosen to correspond to the terminology used in the
initial research literature and early questionnaire responses.
These terms include computational thinking, abstraction,
decomposition, algorithm, and problem solving. Discussions,
composed of individual and related messages, are considered as a
whole [17]. Regardless of the age of a discussion, once it has
been identified as pertinent, every individual message in that
discussion is read and coded, in line with the questionnaire and
interview data.
2.2.3 Interviews
The design of the interview schedule used in this research is based
on a semi-structured approach, as defined by Cohen, Manion, and
Morrison [1]. In particular, the question wording and sequences
are specified in advance of the interview. The interviewer is
granted the flexibility to provide additional questions in order to
elicit greater depth in the responses. The interviewer is also
granted the flexibility to record non-verbal indicators, such as
body language or gestures. This semi-structured approach should
provide sufficient control to ensure comparability of results,
sufficient flexibility to ensure depth of responses, and sufficient
consistency to support the simultaneous collection and analysis of
data indicated by the grounded theory paradigm.
3. FINDINGS
The current, non-saturated dataset is being analyzed in line with
grounded theory, first as conceptual free nodes, then as
categories. These categories and concepts may change, as more
data is added and processed. Three of the categories presented
here, problem solving skills, computational thinking skills, and
programming skills have been introduced above.
3.1 Problem Solving Skills
In the context of this study, problem solving skills are not specific
to programming, but are a wider skill set applicable in many
domains. Participants have highlighted problem understanding
and persistence as important concepts in this category. Analysis
of the data indicates that a common key first step in both learning
to solve problems and learning to program is being able to
understand the problem and its constraints. This observation
agrees with Pólya’s problem solving approach [12]. The theme of
persistence is often linked with the idea of “not giving up”.
Puzzles and games are named as appropriate activities to promote
persistence. They are identified as providing sustained and
lengthy problem solving with discrimination of useful data, back
tracking, and constant evaluation. Many participants recognize
that the opportunity to problem solve is relevant in many different
contexts.
3.2 Computational Thinking Skills
Participants in this study identified explicit examples of
computational thinking skills, as defined by the National Research
Council [10], and related them specifically to problem solving.
Recognizable computational thinking skills such as
decomposition, modeling, and algorithm design are found in the
responses, along with other skills such as planning, justifying, and
evaluating.
Decomposition, the skill to break problems down, is viewed as
being taken for granted. However, for some students, this is
reported as being very difficult and requiring explicit teaching.
Modeling is identified in the sense of high-level systems that are
decomposed into smaller parts, with each individual part
modeling behavior of a subsystem. The act of planning an

algorithm or a product is viewed as a high-level computational
thinking skill. Algorithm design is expressly tied to problem
solving by the participants. It is defining the steps, using some
accepted convention, to solve a problem. This is viewed
differently to program design, which is seen as the translation of
an algorithm into automation understandable by a computing
device. Unexpectedly, participants also included learning to ask
questions about alternatives, identifying trade-offs, justifying
decisions, identifying limitations, refining solutions, and
evaluating results. These are frequently complemented by the
term analytical thinking, which is perceived to involve comparing
alternatives, precisely describing, explaining how, and criticizing
weaknesses.
In general, participants in this study agree with the National
Research Council [10] and Wing [21] concerning the broad
definition of computational thinking and none limited the use of
the term or the skill set identified by the use of the term only to
the domain of computer science.
3.3 Teaching Programming Skills
The concepts in this category represent high-level concerns for
the participants. Included here are the concepts of logical
thinking, programming as a tool, and collaboration as a pedagogic
strategy.
The term logical thinking occurs prolifically in the dataset and
appears to be associated closely with programming constructs
such as sequence, selection, and iteration. This association is
anticipated and parallels that of Saeli [9] who reports that the
most identified big idea of programming is control structures.
The idea of programming as a vehicle for teaching computational
thinking crosses boundaries between respondents, encompassing
academics, teachers, and industry professionals. The components
of computational thinking, such as decomposition and
generalization, are also reported by Saeli [9] to be a big idea of
programming. Collaboration is identified, by participants, as an
effective strategy for teaching computational thinking. This is
usually described as paired or group work, most commonly
involving discussions at the analysis or design phases of software
development. Notably, there are currently no responses
indicating provision for group implementation or paired
programming.
While it is not surprising that participants associate the teaching
of programming constructs, decomposition, and generalization
with computational thinking, it is surprising that an established
pedagogic technique, collaboration, has not been extended to
opportunities for paired programming.
4. CONCLUSION
4.1 Preliminary Model
Although the dataset has not yet been shown to be saturated, as
proscribed by grounded theory, it can form the basis for
preliminary theory generation. As indicated in the introduction,
participants’ responses are used directly to derive a model of the
relationships between computational thinking skills, programming
skills, and the cognitive domain of Bloom’s Taxonomy. Figure 1,
a preliminary model, has been derived to illustrate some of these
relationships.
81
Figure 1: Preliminary Model
The computational thinking skills, reflecting the terminology [10,
21] introduced previously, are represented by an increasing level
of complexity. This hierarchy can be discerned from the
participants’ responses and the reported order of introduction in
the classroom. For example, breaking problems down,
decomposition, is one technique introduced early in the teaching
of both problem solving and programming. The programming
activities column represents those activities that participants view
as promoting computational thinking. For example, the
collaborative work, reported by participants and described in 3.3,
usually takes place during the analysis or design phase where a
problem is broken down into subcomponents. Interestingly, the
terminology used in Bloom’s Taxonomy, the last column, is
represented directly in the participants’ responses. The terms
analyze and understand are also used to describe activities found
in the initial stages phases of problem solving or programming
task. Although the dataset on which the model is based will
change and grow, possible relationships can already be discerned
4.2 Implications
This study assumes, in line with Isbell and colleagues [5], that
computational thinking skills are a requirement of 21 st century
society and that these skills must be taught. This research
contributes to the body of knowledge that may be used to inform
the issue of effective teaching strategies for both programming
and computational thinking. By more explicitly defining the
relationship between computational thinking and programming,
educators may be motivated to move the focus of activities from
the production of an artifact to the acquisition of computational
thinking skills. In the context of the current educational
requirements to include more computer science at all key stages,

the results of this study could influence the design of curriculums
aiming to incorporate the development of computational thinking
skills. In addition, this research responds directly to Guzdial’s
call [4] for more research into how to teach computing in a way
that enforces computational thinking.
4.3 Future Work
Although the current study has not yet reached its conclusion,
areas for further study have already been exposed by analysis of
the data. These include:
How do learners move from the specifics of programming,
such as language constructs or blocks, to more abstract
concepts, such as sorting an array or finding an average,
which aid higher-level problem solving?
How could the explicit teaching of general problem solving
skills and high-level problem solving strategies influence the
development of computational thinking skills?
More work into the relationships between problem solving,
computational thinking, and programming could lead to improved
classroom lessons, improved curriculums, and an improved
understanding of the skills required in 21 st century society.
5. REFERENCES
[1] Cohen, L., Manion, L. & Morrision, K. 2007. Research
Methods in Education, Abingdon, England, Routledge.
[2] Du Boulay, B. 1989. Some difficulties of learning to
program. In: SOLOWAY, E. & SPOHRER, J. C. (eds.)
Studying the novice programmer. Hillsdale, NJ: Lawrence
Erlbaum.
[3] Fitzgerald, S., Simon, B. & Thomas, L. 2005. Strategies that
students use to trace code: an analysis based in grounded
theory. Proceedings of the first international workshop on
Computing education research. Seattle, WA, USA: ACM.
[4] Guzdial, M. 2008. Education: Paving the way for
computational thinking. Commun. ACM, 51, 25-27.
[5] Isbell, C. L., Stein, L. A., Cutler, R., Forbes, J., Fraser, L.,
Impagliazzo, J., Proulx, V., Russ, S., Thomas, R. & Xu, Y.
2010. (Re)defining computing curricula by (re)defining
computing. SIGCSE Bull., 41, 195-207.
[6] Lister, R., Adams, E. S., Fitzgerald, S., Fone, W., Hamer, J.,
Lindholm, M., McCartney, R., Moström, J. E., Sanders, K.,
Seppälä, O., Simon, B. & Thomas, L. 2004. A multi-national
study of reading and tracing skills in novice programmers.
Working group reports from ITiCSE on Innovation and
technology in computer science education. Leeds, United
Kingdom: ACM.
[7] Lopez, M., Whalley, J., Robbins, P. & Lister, R. 2008.
Relationships between reading, tracing and writing skills in
introductory programming. Proceeding of the Fourth
international Workshop on Computing Education Research.
Sydney, Australia: ACM.
[8] Ma, L., Ferguson, J., Roper, M. & Wood, M. 2011.
Investigating and improving the models of programming
82
concepts held by novice programmers. Computer Science
Education, 21, 57 - 80.
[9] McCracken, M., Almstrum, V., Diaz, D., Guzdial, M.,
Hagan, D., Kolikant, Y., Laxer, C., Thomas, L., Utting, I. &
Wilusz, T. 2001. A multi-national, multi-institutional study
of assessment of programming skills of first-year CS
students. Working group reports from ITiCSE on Innovation
and technology in computer science education. Canterbury,
UK: ACM.
[10] National Research Council. 2010. Report of a Workshop on
the Scope and Nature of Computational Thinking. Available:
http://www.nap.edu/catalog.php?record_id=12840 [Accessed
10-05-2011].
[11] Pane, J. F., Ratanamahatana, C. A. & MYERS, B. A. 2001.
Studying the language and structure in non-programmers'
solutions to programming problems. International Journal of
Human-Computer Studies, 54, 237-264.
[12] Pólya, G. 1985. How To Solve It, London, Penguin.
[13] Robins, A., Rountree, J. & Rountree, N. 2003. Learning and
Teaching Programming: A Review and Discussion.
Computer Science Education, 13, 137 - 172.
[14] Saeli, M. 2012. Teaching Programming for Secondary
School: a Pedagogical Content Knowledge Based Approach.
PhD, Eindhoven University of Technology.
[15] Saknini, V. & Hazzan, O. 2008. Reducing Abstraction in
High School Computer Science Education: The Case of
Definition, Implementation, and Use of Abstract Data Types.
J. Educ. Resour. Comput., 8, 1-13.
[16] Schulte, C. & Bennedsen, J. 2006. What do teachers teach in
introductory programming? Proceedings of the second
international workshop on Computing education research.
Canterbury, United Kingdom: ACM.
[17] Sixsmith, J. & Murray, C. 2001. Ethical issues in the
documentary analysis of e-mail posts and archives.
Qualitative Health Research, 11, 423-432.
[18] Strauss, A. & Corban, J. 1998. Basics of qualitative
research: Techniques and procedures for developing
grounded theory, London, Sage Publications Ltd.
[19] Sweller, J. 1988. Cognitive Load During Problem Solving:
Effects on learning. Cognitive Science, 12, 257-285.
[20] Venables, A., Tan, G. & Lister, R. 2009. A closer look at
tracing, explaining and code writing skills in the novice
programmer. Proceedings of the fifth international workshop
on Computing education research workshop. Berkeley, CA,
USA: ACM.
[21] Wing, J. 2011. Research Notebook: Computational Thinking
- What and Why? The Link. Pittsburgh, PA: Carneige
Mellon.

Preparing Teachers for Teaching Informatics: Theoretical
Considerations and Practical Implications
Vassilios Dagdilelis
Department of Educational and Social Policy
University of Macedonia
Thessaloniki, Greece
+30 2310 891 336
dagdil@uom.gr
ABSTRACT
The integration of ICT in Primary and Secondary Education is
considered of great importance for the enhancement of both
teaching and learning. However, a successful integration of ICT in
the teaching practice requires a well-organized training of
teachers. In this paper, we present qualitative results from a
national-scale program aiming at training Secondary Education
teachers in the usage of ICT. Specifically, we focus on the training
of Informatics and Computer Science teachers that were for the
first time included in this training program taking place the last 12
years in Greece. These trainee-teachers will act as trainers of
teachers after the successful completion of their own training.
Several important aspects of organizing the training of trainers are
examined, while some findings concerning the major problems of
preparing teachers for teaching Informatics are analysed.
Categories and Subject Descriptors
K.3.2 [Computers and Education]: Computer and Information
Science Education – Computer science education, literacy.
General Terms
Teachers of Computer Science.
Keywords
Secondary Computing Education, Teacher’s Training.
1. INTRODUCTION
Over these last 12 years in Greece, a very large program has
gradually been implemented whose aim is to integrate ICT into
education generally, and into teaching practice specifically.
Recently, Informatics and Computer Science teachers were also
included. Training of teachers is an important part of this project
– especially Informatics’ teachers. Preparation of the trainers is
also an important part of the project. It is obvious that the success
of the whole program lies heavily in the successful training of the
relatively small number of teachers that are going to train a great
number of teachers in the next face of the program. In this paper,
several important aspects of this preparation of trainers are
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
Conference’10, Month 1–2, 2010, City, State, Country.
Copyright 2010 ACM 1-58113-000-0/00/0010 …$15.00.
83
Stelios Xinogalos
Department of Technology Management
University of Macedonia
Naoussa, Greece
+30 23320 52469
stelios@uom.gr
examined and some findings, concerning the major problems of
preparing teachers for teaching Informatics are.
2. ORGANIZING THE TRAINING OF
TEACHERS
The preparation of trainers takes place in training schools named
with the Greek acronym “PAKE” which stands for
PAnepistimiaka Kendra Ekpedefsis, meaning "University Centers
for Educating (Trainers)". These centers periodically operate as
schools preparing trainers. In other words, they are a very special
kind of educational establishment where trainers of teachers are
instructed by highly specialized educators.
Integrating effectively technology into teaching is a very
demanding task. Mishra και Koehler ([8], [9]) proposed the
TPACK framework. Accordingly to this framework, effective
technology integration for teaching specific content or subject
matter requires a deep understanding and a negotiation between
three components: Technology, Pedagogy, and Content. Their
combination is necessary, as a form of a specific expertise of the
teacher. This framework, eventually, should be considered in a
broader sense: in many cases the social dimension of a specific
subject matter is also important – and Informatics/Computer
Science belongs to this category.
Preparing trainers is an even more demanding goal, from two
points of view:
(a) in the international literature there is a relatively sufficient
number of research findings on the preparation of teachers in
various subjects. There are however, far less data regarding
the trainers of teachers and trainers of secondary professors.
(b) the trainers, both during their preparation, and when actually
training other teachers, have to operate almost simultaneously
at different levels: they have to prepare teachers who, in turn,
will teach students. Consequently, they must learn how to get
prepared, in order to teach teachers who will teach students.
The following diagram, with each arrow meaning “educates”,
may better explain the role of PAKE:
Figure 1. Educating trainers in “PAKE”.

Thus, training of trainers in PAKE requires negotiation on three
distinct levels:
Level 1: teachers teach programming to students. Several
research findings point out the difficulties faced by students
when dealing with programming concepts and students’
perceptions when attempting to solve problems [2], [7]. The
concept of teaching scenario was created for the teaching of
programming. A teaching scenario is a detailed description of
a teaching module which can last more than one hour. Apart
from the description of the learning objectives and the lesson
itself, each teaching scenario includes information concerning
classroom organization, the learning theories which the course
is founded on, and when necessary, elements on epistemology.
The teaching scenario attempts to provide a broad and indepth
analysis of the course the teacher has prepared. So, in
this context the teacher must know how to create teaching
scenarios, which means that he/she must have knowledge and
skills related to the Didactics of the specific subject (i.e. of
Informatics): teaching methods, issues associated with
learning and the problems faced by students, assessment
techniques and a wide range of other knowledge and skills.
Level 2: trainers instructing teachers. Besides the general
principles of adult education, there is no substantial research
on findings specifically on the training of Informatics and
Computer Science teachers. The training scenario - similar to
the teaching scenario- is the central idea of these courses. A
training scenario describes in detail a unit (a piece of
knowledge) which is taught to teachers. As is the case with
the teaching scenario, where a very important element is
students’ misconceptions, a significant part of the training
scenarios makes reference to the basic problems that may
arise in the training of teachers. A characteristic problem is
Greek teachers’ awareness (or lack of) the necessity of such
training. A typical response is: what can 140 training hours of
seminars offer to teachers with 10-15 years of teaching
experience? Teachers also tend to pose specific objections:
why, for instance, they have to be trained in the teaching of
Logo and Logo-like environments (such as Scratch and the
Greek environment Xelonokosmos, the StarLogo and Turtle
Art) – environments, specifically constructed for teaching
purposes (and thus not really helpful)? The program tries to
include all the necessary elements in order to provide
complete training to Informatics/ Computer Science teachers
(CS/ICT teachers) and to provide satisfactory answers to this
kind of questions as well [3].
Level 3: training trainers (educators at PAKE teach trainers).
Few research data exist in this area – if any. In this case there
is no particular scenario to follow, as is the case with levels 1
& 2 above, as the content and purpose of training trainers is
too broad to fit into a single scenario model. A key element in
the training of trainers was the distinction between levels 1
and 2. For instance, the trainers themselves many times
questioned the need to actually create training seminars. The
training of trainers does not only include preparing them to
deal with new technologies that are known but have not yet
been introduced in education - such as tablets and
smartphones, but also to prepare them for technologies that
are completely unknown. How does one prepare trainers in
technologies that are unknown? Among other things, certain
recent changes in the Computer Science course at school, has
resulted in added obstacles to training. In Greek schools, the
interdisciplinary approach to the various subjects has recently
been introduced along with a ‘projects’ approach (usually
84
students working in groups on a specific project topic over the
semester). Even in IT, there is an evident orientation to digital
literacy or computers/information literacy. These changes are
creating a new framework, a new "school ecology" in which
the IT course is a part of. As is perhaps to be expected, these
changes were not immediately accepted by all teachers thus
compounding the task of training [4], [5], [6].
3. DATA
Teachers of CS/ICT followed an intensive program in order to
become trainers of other teachers of CS/ICT. This program lasted
380 hours and almost 60 selected teachers followed the program.
The recording of data related to all levels of teachers’ preparation,
is part of a larger system for recording and evaluating the progress
of the whole project, a project related to the pedagogical and
didactical use of ICT. These data are of course a useful guide for
the further development of the whole project.
Along with this material, we had access to the material trainers
display themselves at various sites - with open access for all. This
material either falls within the framework of their training or not -
for instance, can be personal blogs. Finally, other data, such as
official documents and papers written and presented by the
trainers, were gathered from various sources.
In this paper we refer briefly to some of our findings that emerged
from the qualitative analysis of the data. The data analysis has not
yet completed, but the first important results are already
emerging. The time range, the volume (quantity) and the diversity
of the data, even if they do not guarantee the reliability of our
results, certainly are an important factor of this reliability.
4. SOME FINDINGS
Among the subjects that are taught in the context of trainers’
training for Informatics/Computer Science, programming
occupies a dominant position. Taking into account the fact that
the training focuses on secondary education, it is obvious that the
subjects of Information Technology that are taught, do not cover
the whole field of science. Programming stands out among these
subjects, since it is considered to be more closely connected with
mental processes, such as problem solving, in comparison to other
subjects that occur in sciences. Thus, the value of programming
and algorithm design is often considered to be more general, and
as such, surpasses the boundaries of the school course and has a
more general educational value. Most of the findings mentioned
below refer to programming activities, since in this field the
didactical phenomena we present become clearer.
The transformation of teachers from instructors to trainers of
teachers requires getting familiar with a variety of new theoretical
concepts, as well as acquiring knowledge of various theories
related to the learning and teaching of IT issues. This field is
generally known as Didactics of Informatics. It is obvious that
these theoretical concepts have no value per se, but as tools with
which the teacher can better understand didactical phenomena and
act accordingly. However, these concepts are often related to
views on education and learning, which do not get easily accepted
by teachers. Typical examples of such concepts are the concepts
of constructivism by J. Piaget and the theories of L. Vygotsky.
Modern theories on the organization of a course (such as the
project method or the so-called student-centered teaching) are
based on the socio-constructivism theories. Consequently, these
theories have a great practical usefulness.

However, even when such learning theories get accepted by
teachers, they are sometimes superficially interpreted and
applied. For example, adopting a Logo or Logo-like environment
for teaching programming, does not necessarily mean that a
constructivism teaching approach has also been "automatically"
adopted. The teaching scenario plays an important role too in
achieving good learning results. Asking students to print 10 or
more times a “Hello world” message in Scratch, in order to
comprehend the “repeat N times” programming construct, cannot
be considered as a good example of constructivism.
Comprehending the potential benefits of various learning theories
and using them as tools for devising successful didactical
interventions cannot be easily taught to trainers and mainly
applied by them. The theories that seem to be more popular
among trainers are collaborative learning, constructivism,
exploratory/discovery learning, social development theory and
cultural context. The most referenced (by trainers) proponents of
such theories are Piaget, Vygotsky and Bruner. Unfortunately,
this "popularity" does not seem to be applied into teaching
actions: in most cases, trainers are using typical teaching methods
in their practice, claiming however that these methods are
constructivist ones. In fact, the example of learning theories to
which we refer is of great importance, because it is part of the socalled
"theoretical topics" for which, too often, teachers and
consequently their trainers, express strong objections.
More precisely, an important obstacle of the training process lies
in the unwillingness of some trainers to participate actively in
training seminars or courses on topics they believe that have a
theoretical nature and/or they already have experience in. It is
not that trainers don't believe Vygotsky's Theory or Piaget's
points of view: they believe that it is useless to learn any learning
theory - such theories have nothing to do with "real" teaching and
real classes. A typical example is the teaching of programming
itself, a topic that most Informatics teachers have more or less
experienced. The personal experience of trainers is definitely
important, but the experience that has been gathered from decades
of rigorous research on the teaching and learning of programming
is equally, or even more, important. Several teachers have little or
no knowledge of the results of such research and unfortunately
some of them do not get easily persuaded that these results can
truly help them in teaching programming more successfully. It is
not a matter of theory, it is a matter of practice. The educator has
to work heavily on making such a seminar interesting for trainers.
The educator has to present research results along with specific
examples for utilizing them in order to devise examples and
assignments with the aim of avoiding common misconceptions
and difficulties that are the source of students' more severe errors.
So, subjects such as "learning theories" and didactic concepts, for
example "misconceptions of students", are not considered as
really useful. The integration of programming into the school
curriculum requires a specific didactic transposition: scientific
knowledge and professional practice is transferred to the school
and included in the educational context [1]. Thus, the concepts are
simplified to make them more easily understood; the course is
organized into 45-50-minute sessions, which is the duration of a
teaching period in Greek schools; the subject matter has been
organized into introductory units, exercises, questions, problems.
However, during this transposition of concepts into the school
environment, another transformation takes place: namely, the
school education of an entire scientific field favours certain types
of problems, while marginalizing other aspects of the key
concepts, such as algorithms, data structures, and programming
85
generally. For example, data structures as a means of encoding
entities of the external world, such as images, text, etc., are rarely
referred to in the school environment. Trainee-trainers in many
cases did not fully acknowledge the importance of teaching or the
activities on such issues, considering them of no use, since they
were not directly related to current concepts of programming – at
least the “school version” of programming.
Concepts, such as the didactic contract and the didactic
transposition, research data related to concepts such as variable,
repetition and selection structures, and data about teaching
recursion are included in the curriculum of PAKE. Empirical
research on the different programming environments is included
as well. This approach allowed a clear distinction between the
teaching scenarios and the training scenarios, as the theoretical
scheme of reciprocal interactions and "observations" of a teaching
system (see Figure 1) that clearly defines the framework within
which these teaching interactions take place. The theoretical
approach also (was supposed that) helped the understanding of
didactic phenomena as phenomena, i.e. as observable events that
are not random, but have causes producing them and therefore can
be studied. As we stated earlier, theory of Didactics of
Informatics and especially of programming was accepted by the
trainers - but it is almost sure that they are not all convinced for
the value and usefulness of such "theoretical" subjects.
It is also worth noting that, in a general way, when trainers are
invited to invent new problems and situations for teaching new
concepts, they often produce very questionable examples and
scenarios. For instance, as mentioned earlier, when they were
asked to invent situations for the introduction of the repetitive
structure in Logo-like environments, usually they adopted a
typical approach: draw a "regular" geometrical figure - such as a
regular polygon - with the help of the "turtle" and show step-bystep
how this figure could be constructed directly with a repetitive
statement. However, sometimes the proposed figures are very
complex and practically impossible to be constructed by young
pupils. In other cases, the proposed initial activities are totally
inappropriate for an introduction to repetitive statements, since
they deal with complex arithmetic operations.
Teachers in general are enthusiastic with educational
programming environments that have adopted more or less the
notion of game-based learning, such as Kodu, GameMaker and
especially Scratch. These environments are considered to
motivate students and engage them in the cognitive demanding
process of programming. Environments that offer some kind of
structure editor for developing programs and avoiding syntax
errors are considered by teachers even more effective for
introducing young students to programming. It is possible that
trainers are very positive for this kind of environments, because
they believe that programming is boring for young pupils. So, the
adoption of "cool" environments and programming goals like the
construction of games could be a motivation for the pupils.
In spite of that, a significant number of teachers do not decide to
use such environments in cases where programming is part of the
written exams that take place at the end of the school year, in the
3 rd grade of Gymnasium for example. Their main argument is that
there is no way to examine in paper students' knowledge of
programming, since they were not obliged to learn the syntax of
the language. Of course, this argument is not valid since there are
several ways of examining students’ knowledge, such as
providing them pictures of statements, giving them segments of
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
problems in an algorithmic way and not to learn a specific
programming language. Things are worst in the 3 rd Grade of
Greek Lyceum where students are examined in Pan-Hellenic level
in a subset of a Pascal-like pseudolanguage translated in Greek in
order to enter University. Although specialized tools have been
developed, which offer software visualisation features that have
proven valuable for students, such as step by step execution and
visualisation of variables' state both for programs and structured
flowcharts, several teachers decide not to use them in class. It is
surprising that there are teachers teaching the specific course that
have never used such tools, or at least have not encouraged their
students to use them at home for their assignments, in order to
execute them step by step for testing their assumptions and facing
their difficulties with flow of control. Their main arguments for
not using such tools are the limited number of teaching hours and
the large proportion of students per computer. It is obvious that
such facts of the educational system act as great obstacles for
persuading teachers towards the incorporation of valuable tools
in the educational process. The problem, of course, remains
identical when teachers follow courses in PAKE in order to
become trainers.
A number of trainees, fortunately small, are moreover negative in
getting trained in subjects that are not currently taught at Greek
schools. Such a subject is object-oriented programming (OOP)
that can be taught only in an elective basis in Secondary
education. Although trainees recognize the fact that OOP is the
dominant programming technique nowadays, and they accept the
fact that several educational environments exist that can be used
for a successful teaching of OOP to students, they still have
objections. However, most of the trainees believe that sooner or
later OOP should be included in the curriculum and are positive
towards utilizing programming microworlds, such as Jeroo or
objectKarel, for younger students and educational programming
environments, such as BlueJ and Greenfoot for technologicaloriented
Lyceum students. Especially trainees with no experience
on OOP seem to appreciate even more the importance of
programming microworlds for teaching basic OOP concepts.
5. FINAL REMARKS AND CONCLUSIONS
The exploitation of ICT in Secondary Education is considered
necessary, as it is expected to enhance both teaching and learning.
In Greece, a large-scale program is being carried out over the last
12 years with the aim of integrating ICT into Secondary
Education teaching practice. However, teachers of Informatics
were not included in this program until recently. The training of
the very first Informatics' teachers, that are going to train
massively other teachers in the usage of ICT in Secondary
Computing Education, has just finished. As expected, this training
of teachers carried out in order to prepare them for training other
teachers in the usage of ICT was not easy.
Teachers of Informatics tend to give more emphasis on practice,
rather than theory. Furthermore, they consider themselves experts
in technological issues and do not get easily, at least as easily as
teachers in other disciplines, enthusiastic with ICT. Their
transformation from instructors to trainers of teachers requires
getting familiar with a variety of new theoretical concepts, as well
as acquiring knowledge of various theories related to Didactics of
Informatics. As mentioned, several factors make this process
difficult: unwillingness to participate actively in seminars with a
theoretical nature, or relevant to issues they consider they have
experience in; learning theories and didactic concepts are not
86
considered useful; superficial adoption of learning theories and
misapplication of acquired knowledge; objections in getting
trained in subjects that are not currently taught at schools;
teachers' adherence in school books and their gathered experience
do not let them innovate in their classrooms; adherence to
intricacies of the educational system prevents them from utilizing
ICT tools that they consider useful. Of course, the aforementioned
limitations are true for some of the trainees only.
It is obvious that in order to face these difficulties special
attention must be paid in the material that is used during the
preparation of trainers and the way it is presented. Theoretical
issues have to be presented in conjunction with good examples of
their application in teaching specific IT issues in the classroom
and not in general. An abundance of examples and practical ideas,
as well as material that trainees can easily work on/modify and
use in classroom has to be offered. The experience gathered from
the first cycle of preparing trainers for teachers of Informatics
provides great help towards this direction. Our plans for further
research include the rigorous investigation of the vast amount of
data that has been collected, in order to shed more light on and
enhance the training of Secondary Computing Education teachers.
6. REFERENCES
[1] Chevallard Y. 1994. Les processus de transposition
didactique et leur théorisation, Contribution à l’ouvrage
dirigé par G. Arsac, Y. Chevallard, J.-L. Martinand, Andrée
Tiberghien (éds), La transposition didactique à l’épreuve, La
Pensée sauvage, Grenoble, 135-180.
[2] Dagdilelis V., Satratzemi M., Evangelidis G., 2004.
Introducing Secondary Education Students to Algorithms
and Programming. Education and Information Technologies,
vol.9, no.2, 159-173.
[3] Demetriadis et. Al. 2003. Cultures in negotiation: teachers’
acceptance/resistance attitudes considering the infusion of
technology into schools, Computers & Education, 41(3), 19-
37.
[4] Drent, M., and Meelisen, M. 2008. Which factors obstruct or
stimulate teacher educators to use ICT innovatively?
Computers & Education, 51, 187-199.
[5] Jung, I. 2005. ICT-Pedagogy Integration in Teacher
Training: Application Cases Worldwide. Educational
Technology & Society, 8 (2), 94-101.
[6] Khe Foon Hew, and Brush T. 2007. Integrating technology
into K-12 teaching and learning: current knowledge gaps and
recommendations for future research, Educational
Technology Research and Development, 55 (3), 223-252.
[7] Schwill A. 1997 Computer science education based on
fundamental ideas, In: Passey, D.; Samways, B. (eds.):
Information Technology. Supporting change through teacher
education. Chapman Hall, 285–291.
[8] Mishra, P. & Koehler, M. J. 2006. Technological
Pedagogical Content Knowledge: A new framework for
teacher knowledge. Teachers College Record, 108 (6), 1017-
1054.
[9] Koehler, M. J. & Mishra, P. 2008. Introducing TPCK. In J.
A. Colbert, K. E. Boyd, K. A. Clark, S. Guan, J. B. Harris,
M. A. Kelly & A. D. Thompson (Eds.), Handbook of
Technological Pedagogical Content Knowledge for
Educators (pp. 1–29). NewYork: Routledge.

Grand challenges for the UK: Upskilling teachers to teach
Computer Science within the Secondary Curriculum
ABSTRACT
Sue Sentance
Anglia Ruskin University
Chelmsford, Essex, UK
sue.sentance@anglia.ac.uk
Mark Dorling
Langley Grammar School
Langley, Slough, UK
markdorling@lgs.slough.sch.uk
Recent changes in UK education policy with respect to ICT
and Computer Science (CS) have meant that more teachers
need the skills and knowledge to teach CS in schools.
This paper reports on work in progress in the UK researching
models of continuing professional development (CPD)
for such teachers. We work with many teachers who either
do not have an appropriate academic background to teach
Computer Science, or who do and have not utilised it in
the classroom due to the curriculum in place for the last
fifteen years. In this paper we outline how educational policy
changes are affecting teachers in the area of ICT and
Computer Science; we describe a range of models of CPD
and discuss the role that local and national initiatives can
play in developing a hybrid model of transformational CPD,
briefly reporting on our initial findings to date.
Categories and Subject Descriptors
K.3.2 [Computers & Education ]: Computers and Information
Science Education- Computer Science Education
Keywords
Computer Science Education, Continuing Professional Development,
CPD, K-12 Curriculum, High School
1. INTRODUCTION
Recent curriculum changes in the UK have meant that
there is a huge demand for continuing professional development
(CPD) 1 in Computer Science (CS) and for more focus
1 In this paper the term continuing professional development
will be used as equivalent to the term in-service training
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WIPSCE 2012 Hamburg, Germany
Copyright 2012 ACM X-XXXXX-XX-X/XX/XX ...$15.00.
87
Adam McNicol
Long Road Sixth Form College
Cambridge, UK
amcnicol@longroad.ac.uk
Tom Crick
Cardiff Metropolitan University
Cardiff, Wales, UK
tcrick@cardiffmet.ac.uk
on CS in pre-service training (initial teacher training) for
ICT teachers. Finding ways to support the re-skilling of
many teachers in the UK to be able to teach CS in school is
of interest to universities, industry, government departments
and teacher associations. This paper reports on the work being
done to investigate models of CPD that will meet the
needs of teachers who wish, or are required to, teach GCSE 2
CS in their schools, and do not have sufficient up-to-date
subject knowledge. Much research has been done on a variety
of models for building and sustaining an effective and
confident teaching workforce. In this paper, we consider the
context of CS Education CPD in the UK, and draw on the
framework of CPD outlined by Aileen Kennedy [9], proposing
a range of strategies that would lead to transformative
CPD in the area of CS in schools.
2. BACKGROUND TO CS EDUCATION IN
SCHOOL IN THE UK
There have been significant changes to compulsory education
in the UK over the last 25 years, which resulted over
time in CS essentially disappearing as a curriculum subject
for under-16 year olds and being replaced by Information &
Communication Technology (ICT) [4]. However this situation
in the UK is currently being reversed. Computing At
School (CAS) is an organisation formed in 2009, to promote
CS education in the UK and support CS teachers. Its efforts
have been augmented by the effect of a lecture by Eric
Schmidt, Executive Chairman of Google, criticising the lack
of computer science education in UK schools [13], and also a
report by The Royal Society describing the teaching of computer
science in many schools as “highly unsatisfactory” [14,
p.1]. The Royal Society report’s recommendations included:
• increasing the number of teachers trained to teach Computer
Science
• improving in-service training for teachers
• providing more technical resources for schools.
The UK government subsequently announced that the National
Curriculum for ICT in England was to be disapplied
from September 2012, removing a prescriptive programme of
2 GCSE is a qualification in the UK that is taken at age 16.
GCSE Computing and CS have just been re-introduced in
the UK after approximately15 years

study and facilitating the teaching of more Computer Science
in school. Whilst this is a hugely positive shift in education
policy for the UK, and it is apparent that there are
many schools willing to offer CS to students, there are many
teachers who feel they do not have the skills and an urgent
need for more CPD for schools in this area.
3. MODELS OF CPD
There is a considerable corpus of research on CPD spanning
several decades, including hundreds of individual studies
of different types of CPD and the evaluation of CPD.
Teachers can participate in many hours of training or other
prescribed CPD in school or college and take up external
courses in order to improve their own skills in teaching and
learning but sometimes actual change is difficult to achieve,
particularly as “change is a gradual and difficult process for
teachers”[7]. As Bell & Gilbert report[2], sometimes even
the most well-intentioned efforts to change do not succeed:
“. . . many teachers, even after attending an inservice
course, for example, feel unable to use
the new teaching activities, curriculum materials
or content knowledge to improve the learning
of their students . . . Many teachers are aware of
this pattern and feel frustrated in their attempts
to change” [2, p.9]).
A particularly effective form of CPD is collaborative, which
can be defined as “teachers working together on a sustained
basis and/or teachers working with LEA or HEI or other
professional colleagues” [3] . In all but one of 266 studies
of collaborative CPD reviewed by Cordingley, Bell, Rundell
& Evans [3] there was a definite teacher improvement as a
result.
Kennedy [9] considers a wide range of models of CPD and
proposes a framework through which they can be analysed.
She defines nine different categories of CPD and places them
on a spectrum in terms of their “potential capacity for transformative
practice and professional autonomy” [9, p.236].
Kennedy’s nine categories of CPD are as follows:
• Transmissional
– training
– award-bearing
– deficit
– cascade
• Transitional
– standards-based
– coaching/mentoring
– community of practice
• Transformative
– action research
– transformative
There is a notion that the transmissional type of CPD
is less successful in facilitating teacher change than those
that offer a degree of teacher autonomy. Fraser, Kennedy,
Reid & McKinney posit that “Formal planned opportunities,
which are essentially transmissive, are unlikely to result
in transformative professional learning for teachers, because
they attend primarily to occupational aspects of professional
learning” [6, p165]. In contrast, they consider that transformational
learning is more likely to take place where the
opportunities for learning attend to the personal and social
aspect of professional learning [6]. This will be an important
factor in establishing an effective model of CPD in the UK.
88
Another category of CPD is the community of practice,
where there is a joint enterprise, mutuality and a shared
repertoire of communal resources [15]. In practice, teachers
working together towards a common goal, for example,
implementing a new strategy, who share their experiences,
talk the same language, and are willing to learn from one
another, can be said to be a community of practice.
Being mentored or coached is another way that a teacher
can develop professionally. With a peer coaching model [8],
teachers of equal status work together; in contrast, mentoring
assumes that the mentor has a higher level of expertise
than the mentee [5]. Mentoring may be less likely to be
transformative than coaching, whereas the coaching experience
is designed such that the coachee is able to solve their
own problems and thus become empowered to be able to
effect change.
As Lipowski, Jorde, Prenzel & Seidel report [10], there is
also a need for institutional support within CPD. In a recent
study, experts from a range of countries report an urgent
need to “reform existing insititutional conditions, including
existing cooperation or coordination structures between institutions
involved in the TPD system” [10, p694]. The impact
of effective CPD can be directly linked to school improvement.
Opfer and Pedder report [12] that teachers in the
highest performing schools reported participating in professional
learning activities with higher levels of effectiveness:
they were of longer duration, were more active, and teachers
shared what they had learned with colleagues more often.
This demonstrates that achieving good-quality CPD can affect
the performance of the school. An Ofsted [11] 3 report
on professional development (PD) supports this by stating
that the weakest link in the chain is the way the schools
evaluate the effectiveness of their PD activities.
In terms of the particular case of CPD for CS teachers,
particularly relevant is the balance between subject matter
knowledge (SMK) and pedagogical content knowledge
(PCK) as described in a recent review by Armoni [1]. Armoni
emphasises the importance of learning how to teach
CS as well as a teacher’s own understanding of the subject
and we hope to incorporate this work in our study, as well
as the focus on a constructivist approach to the preparation
of teachers in this area. However, our study will consider
preparation of in-service teachers as well as pre-service
teachers in the rapidly changing curriculum in the UK.
4. RESEARCH QUESTIONS
This paper reports on the beginning of a study to investigate
the nature of the CPD required to facilitate more Computer
Science in school in the UK (particularly England and
Wales). The research questions addressed at the beginning
of this study are as follows:
• To what extent teachers without CS-related degrees
can be trained and given the confidence to be able to
teach CS in the curriculum up to age 16 (GCSE level)?
• What are the effective models of teachers’ CPD and to
what extent they can be applied to CS?
• How does CPD for teachers in CS impact on improved
provision of courses in England and Wales, and correspondingly
on pupil learning in this area?
There are many emerging initiatives in the UK around the
3 Office for Standards in Education, Children’s Services and
Skills

first two questions. We maintain that what is needed is to
be able to address the third question, the impact on pupils.
In aiming for an upskilling of teachers, the transformation
will only be achieved if pupil learning is enhanced and there
are increased opportunities for pupils.
5. TRANSFORMATIVE CPD FOR
SECONDARY TEACHERS IN THE UK
A successful model of CPD can be bottom-up, top-down,
or a combination of both. In the UK at present there are several
initiatives arising from individual institutions becoming
aware of local need for professional development. In addition
subject associations can have a role to play in establishing
a national framework for upskilling the teaching workforce.
5.1 Local initiatives: training
An initial study has been carried out using two models
of training for ICT teachers wishing to include more CS in
their teaching. These courses are taking place at a university
in the east of England. There are two approaches to
university based training that have been trialled at Anglia
Ruskin University by the first and second authors: (i) twilight
sessions over a period of 10 weeks, and (ii) intensive
courses over a number of days.
The twilight model has been so far aimed at existing teachers
who may find it difficult to attend an intensive course as
it requires so much time away from school, and the intensive
course at trainee teachers who have just finished their
initial teacher education. In both approaches the delivery
has been by current teachers/teacher educators who have
experience in teaching CS in a school environment. It has
been assumed that current practitioners are best placed to
explain how to deliver this content to students and to avoid
overcomplicating the process with unnecessary detail.
5.1.1 Twilight Model
The twilight model involves a 2.5-hour session each week
over the course of a 10-week period. Each session is split
into 1 hour of CS theory and 1.5 hours of programming.
There are no expectations of existing knowledge or skills
in either component but there is flexibility to react to the
requirements of a particular cohort.
Each theory session focuses on a different part of the CS
specification each week. This is delivered through a mix
of lectures, practical activities and discussion. Each theory
session stands alone and therefore it is possible for teachers
to miss some weeks and still gain benefit from attending further
sessions without having to make significant investments
in time covering missed material. Programming concepts
are delivered through a mix of demonstration and paired
working. The focus is on getting to the practical programming
quickly so that the instructors can support the attendees.
Both theory and programming are backed by websites,
which contain the material taught during the sessions and
supplementary material including video tutorials.
5.1.2 Intensive Courses
Intensive courses are provided over a period of five days
where attendees will focus completely on developing the
knowledge and skills required to deliver the CS specification.
The content is similar to the twilight model. The intensive
model has focused solely on developing programming skills
89
in attendees as this is an area that many trainees have no
experience in. The intensive nature of the course gives attendees
the opportunity to focus and practice without other
competing demands on their time.
5.1.3 Initial Findngs
At the end of the intensive programme, 90% said that they
had the skills to teach introductory programming. Teachers
commented that: “The training was brilliant. I feel that I
have learnt heaps and this has definitely sparked an interest
in computing for me. I’m keen to learn more!”. Four months
later, the attendees were asked if they had used what they
had learned on the course. There was a low response but
four out of six reported that they were applying the training
and one teacher reported: “ It has given me the impetus to
drive forward with introduction of GCSE Computing in my
school.”
After the twilight model, the respondents gave similar enthusiastic
feedback on their experience of the course, for
example: “My expectations have been exceeded. I’ve learned
more about Python than I thought or hoped and the computer
science lessons have been very thorough.”. Teachers
were asked to rate their confidence levels before and after
the course and the average confidence level rose from 2.9
to 7.7 from the 10-week program. Some teachers did find
the course very challenging: “ [I would like] . . . more time
programming or do this first as by the later time I was really
tired and found it more difficult to focus.” . Attending
training after a long teaching day is demanding on teachers.
Whilst there has been very positive feedback from teachers
themselves, the trainers‘ observations are that some teachers
are finding it difficult to fully engage with the course,
particularly in the twilight model. Many teachers appear to
find it difficult to set aside time to practise programming
between sessions and unfortunately this significantly hampers
their ability to become competent programmers. Like
learning any new skill if it is not practised the knowledge
quickly fades.
In terms of the intensive model, the focus is also on getting
to practical activity as soon as possible and the concepts are
again conveyed through a mix of demonstration and paired
working. Since there is no delay between sessions there is little
opportunity for knowledge to fade so there is not the same
imperative to practise between sessions. Once the course is
over the same dangers relating to practice still apply.
Other models of CPD will now be considered which are
being initiated by CAS.
5.2 Local initiatives: community of practice
Local hub meetings are held after school for groups of
teachers in the areas across the UK to discuss CS teaching
issues. Guest speakers are invited to share their own areas
of expertise. Typically, hub meetings take place two or
three times per year with about 20 to 30 attendees, although
this varies. Hubs provide a community of practice for participating
teachers where they can discuss issues relating to
teaching Computer Science in school and find out about new
developments and resources.
5.3 National initiatives: the Network
of Excellence
The Network of Computer Science Teaching Excellence is
an initiative that has been set up by CAS and BCS Academy

of Computing, the learned society which is dedicated to advancing
Computing (CS & IT) as an academic discipline. It
is designed to utilise and formalise the hub system set up
within CAS, with schools and universities across the country
registering to support one another. The ambitious aim
of the network is to establish CS teaching in at least 1000
schools by 2015. It is planned that initially one university
will support twenty-five secondary schools. Using the university
as a central point of reference it is hoped that they
will be able to better identify and adapt their support for the
needs of the local schools. To make this model sustainable,
schools will then support at least one other school.
5.4 National initiatives: the Master Teacher
To support the universities with developing and delivering
teacher training materials that meet the needs of local
schools, CAS is recruiting Master Teachers to form a local
provider team. Master Teachers will work with universities
in the Network of Excellence. In the medium term it is
hoped that this will create CS departments who are more
aware of the needs of local schools and how to meet them
as well as a national network of advanced skills teachers
in CS. These ‘CAS Master teachers’ will be responsible for
the delivery of CPD to schools in their region working in
association with HE and industry. The structure and content
of the courses can be determined by the local provider
team but will be influenced by the CAS Curriculum, and
will point to suitable resources on the CAS Community site.
Each resource would be mapped to the points of study in
the CAS Curriculum and in the long term ensure curriculum
coverage.
6. DISCUSSION/NEXT STEPS
Using Kennedy (2005) classification we have a series of
different models that are planned to be used in the UK:
• National
– Cascade – the Network of Excellence
– Coaching/mentoring – the Master Teacher model
• Local
– Training/deficit courses – subject-knowledge and
pedagogy provided by universities
– Community of Practice – local CAS hubs supporting
a network of teachers
Together, as a hybrid collection, we posit that these form
a transformational model of CPD.
This range of models of CPD has been planned to tackle
a national emerging situation in teacher education in England
and Wales. An online questionnaire has been designed
to collate the perceived needs of ICT teachers in terms of
their preparation to teach CS in secondary school. This
will enable us to answer the first research question above.
The next focus in this research study will be on the third
research question and on developing a research instrument
to measure the impact of the professional development on
both pupils, teachers and schools. The initiatives outlined
in this paper and the impact of the various models will be
evaluated.
7. CONCLUSION
It cannot be assumed that providing training or facilitating
professional development activities for teachers will necessarily
bring about the transformation of Computer Science
90
education in the UK that we require. There is a need in the
UK for teachers to have confidence at an academic level to
teach Computer Science; however professional development
relating to pedagogy must not be ignored. The question as
to who is responsible for upskilling the teachers is increasingly
important. Local initiatives may be most valued by
teachers as they create networks and interpersonal relationships
to support teachers. A local approach, however, can
be ad-hoc and areas of the country will be neglected.
8. REFERENCES
[1] M. Armoni. Looking at secondary teacher preparation
through the lens of computer science.
Trans.Comput.Educ., 11(4):23:1–23:38, nov 2011.
[2] B. Bell and J. Gilbert. Teacher development: a model
from science education. Falmer Press, London, 1996.
[3] P. Cordingley, M. Bell, B. Rundell, and D. Evans.
The impact of collaborative CPD on classroom
teaching and learning: how does collaborative
continuing professional development (CPD) for
teachers of the 5-16 age range affect teaching and
learning? Technical report, Social Research Unit,
Institute of Education, 2003.
[4] T. Crick and S. Sentance. Computing at school:
Stimulating Computing Education in the UK. In
Proceedings of the 11th Koli Calling International
Conference on Computing Education Research, Koli
Calling ’11, pages 122–123, NY, USA, 2011. ACM.
[5] Department for Education and Skills. National
framework for mentoring and coaching. Technical
report, CUREE, 2005.
[6] C. Fraser, A. Kennedy, L. Reid, and L. McKinney.
Teachers‘ continuing professional development:
contested concepts, understanding and models.
Journal of In-Service Education, 33(22):153–169, 2007.
[7] T. R. Guskey. Professional development and teacher
change. Teachers and Teaching, 8(3):381–391, 08/01;
2012/06 2002.
[8] B. Joyce and B. Showers. The evolution of peer
coaching. Educational Leadership, 53(6):12–16, 1996.
[9] A. Kennedy. Models of continuing professional
development: a framework for analysis. Journal of
In-Service Education, 31(2):235–250, 2005.
[10] K. Lipowski, D. Jorde, M. Prenzel, and T. Seidel.
Expert views on the implementation of teacher
professional development in European countries.
Professional Development in Education,
37(5):685–700, 11/01; 2012/06 2011.
[11] Ofsted. The logical chain: continuing professional
development in effective schools. Technical report,
Ofsted, 2006.
[12] V. D. Opfer. The lost promise of teacher professional
development in England. European Journal of Teacher
Education, 34(1):3–24, 2011.
[13] E. Schmidt. Eric Schmidt‘s MacTaggart lecture - full
text, Friday 26th August 2011 2011.
[14] The Royal Society. Shut Down or Restart? The way
forward for Computing in UK Schools. Technical
Report January 2012, DES 2448, 2012.
[15] E. Wenger. Communities of Practice and Learning
Systems. Organization, 7(2):225–246, 2000.

(Some) Grand Challenges of Computer Science Education
in the Digital Age: A Socio-Cultural Perspective
ABSTRACT
The goal of this paper is to articulate (some of) the grand
challenges that computer science education (CSE) at the school
level faces in the digital age. Based on the socio-cultural
theoretical idea that learning means entering a culture, I suggest
viewing schooling as an encounter between intertwined cultures.
Computer science (CS) students can be viewed as members of
many intertwined cultures: (a) they are newcomers to the
professional computing culture, but (b) most are also old timers
in a “user” culture, living in a world surrounded by informationcommunication
technologies (ICT), and also have informal
learning experience (and values) within ICT, mostly from out-ofschool
experience; and finally, (c) they are members of the
school culture which itself is currently in a process of
transformation due to the digital age). Using this framework, I
discuss two interrelated grand challenges of CSE in K-12 school
levels: (1) the need to adjust the CS curriculum to better overlap
with life-long learning skills; and (2) the need to better learn the
characteristics of the “digital” generation and attune education to
address these needs.
Categories and Subject Descriptors
K.3.2 [Computers and Education]: Computer and Information
Science Education – computer science education.
General Terms
Human Factors
Keywords
Digital age, sociocultural theories, cultural encounter, challenges
in CSE, K-12
1. INTRODUCTION
Due to the accelerating processes of globalization and
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, or
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
Conference’04, Month 1–2, 2004, City, State, Country.
Copyright 2004 ACM 1-58113-000-0/00/0004…$5.00.
Yifat Ben-David Kolikant
School of Education
Hebrew University of Jerusalem, Jerusalem
Israel 91905
Tel: 972-2-5882056
Yifat.Kolikant@mail.huji.ac.il
91
digitalization, educational goals in many countries have changed.
More emphasis is put on innovation and knowledge creation in
order to maintain or increase a country's competitiveness in a
global economy. To this end, the new primary objective of
schooling is to provide opportunities for students to develop lifelong
learning skills. Moreover, much work has been invested in
transforming the school agenda to move beyond information and
teaching practices and from being teacher-centered to being more
student-centered. This creates a great challenge to education, and
as I argue in this paper, CSE is no exception.
This paper offers a socio-cultural perspective to discuss (some
of) the grand challenges that the digital age poses to CSE. The
socio-cultural perspective is described in Section 2. Section 3
presents a socio-culturally inspired viewpoint on CSE as a
cultural encounter. Section 4 presents the challenges to education
in the digital age. Section 5 discusses the challenges and
opportunities CSE deals with due to the digital age and finally in
Section 6, conclusions and implications are provided.
2. THE SOCIO-CULTURAL
THEORETICAL FRAMEWORK
According to Lave and Wegner [10], learning is a process of
enculturation into a culture or a community of practice by
newcomers' participation in peripheral yet legitimate, genuine
activities of the target culture. The newcomers gradually, through
continuous negotiation on the meaning of actions achieve a
holistic viewpoint of professional practice and increase their
participation, ultimately becoming full-fledged participants.
Instructional models that rely on this theoretical framework
emulate "real world" participation in a culture in school. In these
emulations, the teacher usually represents an old-timer in the
culture whereas students are viewed as newcomers to the culture
(e.g.,[4]). Figure 1 is a schematic representation of the social
fabric as pictured by this approach. This power relationship
faithfully reproduces the power relationship between newcomers
and old-timers described by Lave and Wenger [10] in craft
apprenticeship.
Nevertheless, I claim that in certain domains, the metaphorical
view of the educational milieu as a "busy tailoring shop" [4]
where the expert tailor is the teacher and the students are merely
apprentices (or newcomers) does not capture the complicated
social fabric of the milieu and is therefore limited in its ability to

explain and anticipate learning difficulties. Today's students
come
Figure 1. A schematic representation of a cognitiveapprenticeship
viewpoint on schooling
to school with a wealth of informal knowledge that includes
learning practices which have proved useful in out-of-school
learning experiences, such as their interaction with the Internet
as consumers and producers of knowledge as well as their
gaming experiences (for further discussion of the impact of
students' computer experience on their cognitive skills, see Gee,
[6]. As our expectations of schooling sociology shift so that we
no longer assume that the teacher is the “knowledge authority”
and instead expect students to assume responsibility for creating
knowledge, our metaphors for educational milieus should reflect
this change.
Figure 2. A schematic representation of a cultural encounter
viewpoint on schooling
I suggest using the metaphor of a cultural encounter. This
metaphor faithfully maintains students' “newcomerness” to the
culture represented by the instructional setting but, at the same
time, regards students as longstanding members of other,
intertwined cultures. Moreover, this metaphor also enables us to
regard school goals and practices as being driven from multiple,
not necessarily overlapping cultures, thereby adding to the
picture by emphasizing the multicultural nature of the sociology
of schooling today. Figure 2 is a schematic representation of this
viewpoint.
92
3. COMPUTER SCIENCE EDUCATION AS
A CULTURAL ENCOUNTER
In any domain, the teacher and the student are active in two
cultures simultaneously, or metaphorically, each of them wears
two hats simultaneously. The first hat is the hat of the target
community of practice. The teacher, like the entire instructional
setting, represents the community of the studied practice, and the
student is the newcomer to that community of practice. Indeed,
the CS curriculum is oriented toward the academic community
whose understanding of the computer world involves the
abstraction, solution, and proof of algorithmic problems.
The second hat is the hat of participants in a school milieu.
School is a cultural environment in itself. Brousseau [3] defines
school activity as creating and playing didactical milieus,
designed by the teachers according to the knowledge they want to
devolve, with consideration of the students they teach. Learning
is achieved when the student develops strategies to make sense
of the milieu.
CS students wear a third hat, that of a computer user or local
developer. Their viewpoint regarding the CS world—what
constitutes a good problem, an accountable approach to a
solution, and a satisfactory solution of the problem—were shaped
without much interaction with the other two "professional" CS
cultures: the CS academia and industry [1]. A user is defined by
Turkle [18] as someone who is “involved in the machine in a
hands-on way, but is not interested in the technology except as it
enables an application” [18, p.32]. Many students, thus, are
veterans in the world of computers; this shapes their
understanding of and interests in this world. Furthermore, as oldtimers
in this world, they might delegitimize the CS curriculum
as relevant to their CS life, which ultimately leads to a culture
clash.
Often the borderline between users and programmers is not
clear-cut [7, 12]. In fact, it is all a matter of the context. Students
might mistakenly mis-contextualize educational computer-related
situations and approach them “wearing” the user hat rather than
the programmer’s hat. For example, in previous work, it was
evident that both the user culture and school culture nurtured
students (mis)conceptions of correctness as relative and partial.
“Relatively correct” meant that that the program “worked,” i.e.
produced the desired output with additional irrelevant output.
Relative correctness meant that there was a “grain of
correctness” in the program, that is, something in the code was
written correctly. In fact, students referred to program correctness
as the sum of the correctness of the code constituents, as teachers
would grade their programs. Moreover, some attributed the
teacher’s judgment that the program was incorrect to
‘‘pettiness’’ [1, 2].
4. EDUCATIONAL CHALLENGES IN THE
DIGITAL AGE
4.1 The change in school culture
Traditional schooling is characterized by its focus on information
and teacher-centered practices. Resnick [16] and other scholars
contend that these schooling foci are unsuitable for current times
and should therefore change, moving beyond information and

ecoming more student-centered. Indeed, the primary educational
goals in many countries have changed. More emphasis is put on
innovation and knowledge creation. In other words, the school
culture is undergoing process of transformation.
The partnership for 21 st century skills (P21), a broad coalition of
education, business, government and foundations, has advanced a
powerful learning framework that captures the three essential
learning goals most needed in our times. P21 indicates major sets
of skills needed by tomorrow’s citizens: (a) learning and
innovation, with emphasis on team-work skills, (b) information,
media, and ICT skills, and (c) life and career skills since
students need to develop the “ability to navigate the complex life
and work environments in the globally competitive information
age” (www.p21.org). P21 suggests reducing the number of core
domains taught and then use these core domains as a basis to
help students develop those skills, and moreover, weaving 21st
century interdisciplinary themes into core subjects, such as
financial, economic, business and entrepreneurial literacy and
global awareness.
4.2 The change in students’ capital
Another body of work is devoted to understanding the nature of
this generation who grew up surrounded by ICT, and adjust
education accordingly. Some claim that digital natives are
different and better learners than the so-called digital
immigrants, those who were already adults when these
technologies were introduced (e.g., [14, 15]). In contrast, others
are concerned that this generation poses many challenges to
educators because it has many attributes that hinder learning in
traditional settings. According to Twenge [19], the generation
born after 1970, which she terms Generation Me, is more selffocused,
to the point of narcissism, and less motivated to learn
something unless its immediate benefits are clear. Furthermore,
while more assertive, students in this generation are less selfreliant.
5. THE GRAND CHALLENGES TO CSE
5.1 The rapid technological changes: a
catalyst of a cultural clash
The rapid technological changes in the world might intensify the
culture clash between students who hold user-oriented
viewpoints and educators aiming at facilitating students’
understanding of the professional CS culture. Educators might
find themselves racing to catch up with new innovative
programming environments, chosen not purely due to their
pedagogical potential but rather as a means to be judged as less
“backward” by their students. Moreover, since ICT and learningby-doing,
hands-on ICT incrementally touches on all school
domains, students would be less likely to elect CS only because
they get hands-on computer time. Educators thus face the need to
balance between the need to maintain the students' sense of
relevance by adjusting the CS educational environment to the
outside world while at the same time maintaining pedagogical
and curricular merit. This challenge is not new, only the
increment in the frequency of changes together with students’
expectations intensifies it.
93
5.2 The ongoing transformation of schooling
to a 21 st century education
Some would assume that CS need not consider a change since it
is a high-tech related domain and hence is already aligned with
the new goals of school, i.e. preparing students for life-long
learning with ICT. They might argue that, programming, as well
as the rest of CS, are important computer-related skills and
knowledge. Moreover, CSE has always been embedded with
hands-on student-centered activities. Most CS curricula go
beyond information, aiming at nurturing problem-solving
capabilities by means of algorithmic thinking and programming,
an important asset for life-long learning.
However, other 21st century skills, such as collaboration,
communication, and career skills are not necessarily emphasized
in these CS curricula. Principals, parents, and students might
doubt the benefit of studying CS since it loses its unique
importance as nurturing computing-oriented problem-solving
skills. Some might argue that CS should be omitted from school
programs in favor of other domains which emphasize more 21 st
century skills. This, in itself, is a legitimate idea given the
tendency to reduce the number of domains taught. Therefore,
CSE goals should be extended to involve, in addition to
acculturation into the professional CS culture, the preparation of
students for life-long learning in the 21 st century.
CSE can be re-designed to leverage 21 st century skills without
compromising the knowledge, values, and practices of the
professional culture. Obviously, as any domain, more ICT tools
can be used to support students’ learning, starting with
simulations (of algorithms, scripts, the machine, and so forth).
More importantly, CS prides itself on nurturing the required
skills. It encourages creativity and critical thinking. It can also
encourage communication and collaboration. For example,
documentation can be taught as a way for communicating
effectively with others at work. Additionally, CS can and should
be used to facilitate cross-disciplinary instruction. For example,
programming environments, such as Scratch [17] and Alice [5]
enable students to express ideas and tell stories, and therefore
can be used in humanities’ educational environments. In fact, CS
has a unique advantage over other domains for advancing the
goals of knowledge creation, creativity, and entrepreneurial skills
because it enables students to create new technological tools. CS
can also serve as a rich arena or teaching life and career skills.
The young history of CS is replete with people and companies
who have thought out of the box, made bold decisions,
sometimes failed and sometimes succeeded and managed from
there on. This type of career action resembles the future of these
students: tomorrow’s adults in a rapidly changing world.
Students can benefit from the exposure to the life paths of people
and products in this field.
5.3 Teaching "generation me"
The fact that ICT has become more and more prevalent in
students' lives in and out of school might intensify their useroriented
computer-related capital (see, for example the report of
the OECD (2006) on the growth of student use of ICT,[13]). This
computer-related capital influences students’ judgment as to
what is important and where to invest their efforts to learn.
Moreover, according to Twenge [19], today's students are less

motivated to learn something unless the immediate benefits are
clear. Hence, CS teachers might experience even greater
difficulties in introducing the professional CS culture as
legitimate, let alone the desired one. Simple introductory
problems previously used to teach programming with gradual
levels of difficulty, such as a program that prints "hello world"
on the screen, might influence students' judgment for the worse
about the worthiness of taking CS classes given that they can
(or know someone who can) achieve more attractive outcomes in
the use of the computer. Students' high level of frustration and
their low self-reliance is also an obstacle to CS education.
Debugging, for example, can be frustrating. Hence, educators
face the tension between the need to provide students with
attractive problems and less “hello-world”-like problems, so that
the students will become engaged in their solution, while taking
into consideration students’ “newcomerness” programming
capabilities.
This concern leads educators to offer students problems the
students might value as worth of their efforts, such as the
development of computer games [11] and the processing of
media [8, 9]. Some teach introductory courses using visual
programming environments that enable work processes more
reminiscent of students’ experience with the user interface. For
example, Alice 3 [5] enables students to program 3-D characters
using a graphical drag-and- drop programming. Similarly,
Scratch was designed to motivate all students to program [17].
These environments enabled teachers to provide problems that
students perceive as attractive while at the same time avoiding or
postponing frustrating actions, such as debugging syntax.
Both approaches rely on students' computer-related capital, as
users, in an attempt to meet students where they are, provide
them with an immediate sense of the relevance of school
experience to their lives, and at the same time introduce them to
the principles in programming in a way that the students might
value and will be motivated to further their experience.
Educators, however, should keep in mind that meeting students
where they are should not aspire to leave them in their current
cultural horizons, but rather to help the students expand their
cultural perspective and cross the boundaries towards the
professional culture.
6. CONCLUSIONS AND IMPLICATIONS
The challenges articulated in this paper emerge from the
recognized need to transform schooling so that it better prepares
students for life within a digitalized and globalized world. The
cultural-encounter metaphor is useful to detect challenges as
interaction between two (or more) intertwined cultures. Part of
the resolution of these challenges includes the awareness of
educators of the cultural capital of their students, accumulated
through the students’ in and out-of school experience with ICT.
The other part is to bridge between students’ current
perspectives and the desired target practices and values. To this
end, in previous studies [1], it has been suggested that
educational situations should be designed as fertile zones of
cultural encounter (FZCE), a metaphorical zone that enables
using the students’ cultural capital as a bridge to the perspective
represented by the teachers. FZCE occurs when the educational
milieu clearly reflects the distinction between the intertwined
94
cultures, while making the cultural tools represented by the
instructional setting accessible and relevant to the students.
Creating FZCEs for today’s students is a great challenge. First,
the target is complicated: it is about enculturation within the CS
professional culture while, at the same time, nurturing students’
life-long learning capacities. Moreover, as part of the general
transformation in school, instructional milieus should be
designed with more student-centeredness, allowing for and
encouraging students to manage their learning actions. However,
CS has great potential to serve as an arena to teach 21 st century
skills in ways students perceive as relevant and therefore are
more motivated to invest mental effort.
7. ACKNOWLEDGEMENT
I thank M. Ben-Ari and S. Pollack for their insightful comments.
8. REFERENCES
[1] Ben-David Kolikant, Y., & Ben Ari, M. Fertile zones of
cultural encounter in computer science education, Journal of
the Learning Science, 2008, 1,17,1-32.
[2] Ben-David Kolikant, Y., & Mussai, M. 'So my program does
not run': definition, origins, and practical expressions of
students' (mis)conceptions of correctness, CSE, 18, 2, 131-
151.
[3] Brousseau, G. Theory of didactical situations in
mathematics, Dordrecht: Kluwer Academic, 1997.
[4] Collins, A., Brown, J. S., & Holum, A. (1991). Cognitive
apprenticeship: Making thinking visible. American
Educator, 12(6), 38-47.
[5] Dann, W., & Cooper, S. Alice 3: concrete to abstract.
Communications of the ACM, 2009, 52(8), 27–29.
[6] Gee, P. J. (2003).What video games have to teach us about
learning and literacy. New York: Palgrave Macmillan.
[7] Goodell, H., Maulsby, D., Kuhn, S., & Traynor, C. End-user
programming / informal programming. A workshop
conducted at the ACM CHI 1999, SIGCHI Bulletin, 31 (4),
1999, 17-21.
[8] Guzdial, M. (2003). A media computation course for nonmajors.
In Proceeding of the ITiCSE (pp. 104–108).
[9] Guzdial, M. & Tew, A. E. Imagineering inauthentic
legitimate peripheral participation: an instructional design
approach for motivating computing education. ICER 2006,
ACM Press, New York, NY. (pp. 51–58).
[10] Lave, J., & Wenger, E. (1991). Situated learning:
Legitimate peripheral participation. Cambridge, UK:
Cambridge University Press.
[11] Leutenegger, S. and Edgington, J. 2007. A games first
approach to teaching introductory programming. In SIGCSE
'07. ACM, New York, NY, 115-118.
[12] Nardi, B. A. (1993). A small matter of programming:
Perspectives on end user computing. Cambridge: MIT Press.
Organisation for Economic Co-operation and Development

[13] OECD. (2006). Are students ready for a technology-rich
world? What PISA studies tell us.
http://www.oecd.org/document/31/0,3343,en_32252351_32
236173_35995743_1_1_1_1,00.html. Retrieved 09.08.09.
[14] Papert, S. (1996). The connected family: Bridging the
digital generation gap. Atlanta, GA: Longstreet Press.
[15] Prensky, M. (2001). Digital natives, digital immigrants. On
the Horizon, 9(5), 1–6.
[16] Resnick, M. 2002. Rethinking learning in the digital age. In
The global information technology report: Readiness for the
networked world, ed. G. Kirkman. New York: Oxford
University Press.
95
[17] Resnick, M., Maloney, J. , Monroy-Hernández, A., Rusk,
N., Eastmond, E., Brennan, K., Millner, A., Rosenbaum, E.,
Silver, J., Silverman, B. Kafai., Y. (2009). Scratch:
programming for all. Commun. ACM 52, 11, 60-67.
[18] Turkle, S. (1995). Life on the screen: Identity in the age of
the Internet. New York: Simon and Schuster.
[19] Twenge, J. (2006). Generation Me: Why today’s young
Americans are more confident, assertive, entitled – and
more miserable than ever before. New York: Free Press.

ABSTRACT
Challenge and Creativity: Using .NET Gadgeteer In
Schools
Sue Sentance
Anglia Ruskin University
Chelmsford, UK
sue.sentance@anglia.ac.uk
This paper reports on a study carried out in secondary
schools in the UK with students learning to use .NET Gadgeteer,
a rapid prototyping platform for building small electronic
devices [32]. A case study methodology has been
used. Some of the students involved in this four-monthlong
project had some prior background in computer programming
whereas for others this was completely new. The
teaching materials provided a two-phase model of learning:
an instruction phase followed by a creative phase, the latter
utilising a bricolage approach to learning programming [30].
The aim of the pilot was to generate an interest in building
devices and stimulate creativity. The research found that
the tangible nature of the .NET Gadgeteer modules helped
to engage the students in becoming creative, and that students
valued challenges with which they were not usually
presented within the curriculum.
Categories and Subject Descriptors
K.3.2 [Computers & Education]: Computers and Information
Science Education - Computer Science Education
Keywords
Secondary education, .NET Gadgeteer, high schools, computer
programming, creativity, bricolage
1. INTRODUCTION
Microsoft .NET Gadgeteer 1 is a platform that enables
rapid prototyping of small electronic gadgets and embedded
hardware devices. It combines the advantages of objectoriented
programming, solderless assembly of electronics using
a kit of hardware modules, and the quick fabrication of
physical enclosures using computer-aided design. The fact
that .NET Gadgeteer covers a variety of sophisticated computer
science and engineering skills, but requires minimal
1 http://netmf.com/gadgeteer
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WIPSCE 2012 Hamburg, Germany
Copyright 2012 ACM X-XXXXX-XX-X/XX/XX ...$15.00.
96
Scarlet Schwiderski-Grosche
Microsoft Research Cambridge
Cambridge, UK
scarlets@microsoft.com
prior knowledge, makes it especially suitable for education.
.NET Gadgeteer has great potential in schools as it can be
used to teach students simple electronics and computer programming
as well as computer-aided design. Moreover, it is
very motivating for young people to be able to build their
own gadgets.
A research project was initiated to investigate whether
.NET Gadgeteer has the potential to be used in schools.
This was designed to elicit students’ and teachers’ perceptions
of the platform, and suggest directions for further research
and development. A case study methodology was
used, as defined by Stake [26], whereby different sources were
drawn together to give an accurate description of how the
pilot use of this platform was perceived by both teachers,
students and researchers.
This paper reports on the findings of this first study. In
the paper, we will give an outline of the current situation
with regard to Computer Science (CS) education in schools
in the UK which provides some motivation for the current
project. A brief description of .NET Gadgeteer will help to
set the scene and the next section of the paper will focus on
the methodology and findings of the research. The findings
will be discussed and then further work suggested.
2. COMPUTER SCIENCE EDUCATION IN
SCHOOLS
2.1 Background
A recent report by the Royal Society, an influential academic
body in the UK, begins with the statement “The current
delivery of Computing education in many UK schools
is highly unsatisfactory” [29, p.1]. Prior to this report, there
had been an increasing awareness of the need for more Computer
Science in schools in England and Wales. The Computing
At School group 2 (CAS) has been very active since
2008 in advocating the need for more Computer Science in
the curriculum and supporting teachers on the ground [6].
Since the publication of the Royal Society report, the government
has started to implement change in the curriculum;
more qualifications are being launched to feed the need
of schools and pupils to study this subject within the curriculum.
The Royal Society states that “Computer Science
is sufficiently important and foundational that it should be
recognised as a high status subject in schools, like mathematics,
physics or history” [29, p.34]. These developments
in the UK mirror those that have been happening elsewhere
2 http://www.computingatschool.org.uk

in the world, for example, as described by Wilson et al. in
the USA’s Running on Empty report produced by the CSTA
[33], and as implemented in Israel [10].
It is apparent that there are many schools keen to offer
Computing-related content to students aged 11-14 and
also school qualifications in Computing/Computer Science
to students aged 14-18. Within this climate, providing engaging
resources and vehicles for learning Computer Science
is very timely.
As the emphasis grows on secondary Computing education,
more research in this area is needed. It is important
to look at the pedagogical approaches to bridge the gap between
computer literacy and Computer Science as well as
what should be taught within the curriculum [3]. Constructivist
learning theories applied to Computer Science
emphasise the active, subjective and constructive character
of knowledge, placing students at the centre of the learning
process [13]. Specifically, constructivist learning is based on
students’ active participation in problem-solving and critical
thinking regarding a learning activity which they found
relevant and engaging [11]. As a learning theory, it has profoundly
influenced the teaching of programming [2]. Experiential
learning stems from constructivism and is a term
which can be used to describe the design of activities which
engage learners in a very direct way. It describes the process
of engaging learners in an authentic experience in which
they can make discoveries and experiment with knowledge
at first hand. Through reflection, students construct new
knowledge and ways of thinking about themselves, leading
to deeper learning.
Research to date in secondary computing education has
investigated a range of approaches to structuring the curriculum
[27, 10, 1], whilst other research has focused on
tools and environments that may motivate and engage young
people in the classroom such as Scratch[17], Alice[5], Greenfoot[15],
and Kodu[16]. It is important, however, to address
the approach to learning that underlies these tools or environments.
Scratch, for example, builds on work in Logo and
on the constructionist ideas of Papert [17]. Papert used the
term constructionism which is a combination of ‘constructivism’
and ‘construction’ [21]. This is particularly relevant
for .NET Gadgeteer which involves physically constructing
devices in an exploratory way.
2.2 Engaging students
There have been many discussions in the literature about
how to engage students with Computer Science and suggestions
for reasons why students do not have a positive
attitude to the subject, for example [25, 9, 34, 4]. Downes
and Looker suggest that the more IT students use at school,
the more they are likely to take up computing-related subjects
when they are given a choice [8]. Brinda, Puhlmann
and Schulte discuss how to introduce Computer Science by
working from what students already know from their ICT
education and making Computer Science relevant to their
own experience [3]. Pollock and Harvey integrated a range of
pedagogical approaches in order to engage students more effectively
and demonstrated the effectiveness of collaborative
work and reflection on learning [22]. Collaborative working
is especially important in our .NET Gadgeteer trials. Cutts,
Esper and Simon discuss how to offer all students Computing
education in relation to “what a computer can do and
how one can interact with it” [7] by giving students an un-
97
derstanding that computers are deterministic, precise and
comprehensible. This understanding can be gained without
necessarily learning to program. With .NET Gadgeteer, we
hope that we can engage all students in terms of a better
understanding of how the devices and technology all around
us works.
Schulte and Knobelsdorf look at attitudes towards Computer
Science using a biographical approach, and note the
differences between those that regard themselves as insiders
and outsiders [25]. They recommend that teachers “should
intertwine introduction to CS (e.g. learning programming
- a design activity) with learning professional use . . . a major
problem is to teach another world-image of CS” [p.37].
The interest of students in real devices and current technology
makes .NET Gadgeteer quite engaging in this regard,
as users can relate it to the professional world of developing
devices.
2.3 Using tangible environments
Besides .NET Gadgeteer, other tangible devices are available
which enable students to write programs using actual
hardware components. These include Lego Mindstorms 3 ,
the Scratch Pico Board 4 , Open University‘s SenseBoard [23]
and Arduino 5 . All these hardware kits have different features
but offer the same experience of hardware in addition
to software, thus broadening the exposure to the way that
technology works. The Raspberry Pi 6 is now available which
also has the appeal of being tactile and exposed; this has generated
a lot of enthusiasm, although it is a computer rather
than a means to build devices such as the other devices described
here.
Marshall [18] and Horn et al. [12] both describe how tangible
environments can have a very positive effect on collaborative
and active learning, as they enable students to
share work together in a very visible way. They also utilise
concrete physical manipulation which can facilitate more effective
or natural learning. Working with physical devices
can encourage an exploratory or bricolage approach, as discussed
next.
2.4 The bricolage approach
The concept of bricolage was introduced by Levi-Strauss
in The Savage Mind [14]. It refers to a science of concrete
development as an alternative to abstract planning, and was
applied to the area of computer programming by Turkle and
Papert [30]. It represents a mode of learning based on ‘tryit-and-see-what-happens’
[2], of which Ben-Ari is rather critical
claiming that “A student who exclusively uses such techniques
is ultimately not qualified to work on the software of
embedded and operating system, which requires the ability to
create and test abstract hypotheses” [2]. However, Stiller [28]
successfully used a bricolage approach when teaching programming
by encouraging students to build on previous programs,
based on a pedagogy of incremental problem-solving.
The intention of this project was to use a bricolage approach
when introducing .NET Gadgeteer to students.
3 http://mindstorms.lego.com
4 http://wiki.scratch.mit.edu/wiki/PicoBoard
5 http://www.arduino.cc/
6 http://www.raspberrypi.org/

Figure 1: .NET Gadgeteer modules
Figure 2: Designer view of a .NET Gadgeteer device
3. .NET GADGETEER
.NET Gadgeteer is a platform for creating your own electronic
devices using a wide variety of hardware and a powerful
programming environment [32]. .NET Gadgeteer hardware
consists of an ever increasing range of mainboards and
modules. The environment is open source and therefore,
new modules can be developed by any enthusiast. Students
with little or no Computing background can build robot-like
devices made up of components that sense and react to their
environment using switches, displays, motor controllers, and
more. Components are plugged into a mainboard and subsequently
programmed to make them work together (see Figure
1).
.NET Gadgeteer originated at Microsoft Research Cambridge,
UK. It was designed as a tool for researchers to make
it faster and easier to prototype new kinds of devices. For
example, a digital camera can be built in about half an hour.
One of the motivations for the Gadgeteer project has been
to provide “both a low threshold for entry – allowing non
expert developers and designers to quickly sketch and construct
functional devices – together with a high ceiling that
allows experienced users to create sophisticated and capable
devices that can be used in practice” [31]. This is ideal
for education where there is a need to stretch and challenge
youngsters to achieve their highest potential. Since
then, the platform has proven to be of interest to hobbyists
and for secondary and tertiary education. Microsoft
Research has launched .NET Gadgeteer as open source software/hardware,
and .NET Gadgeteer kits are now available
from a variety of hardware vendors.
A starter .NET Gadgeteer kit consists of a mainboard,
and various modules including a camera, joystick, buttons,
LEDs, potentiometer, Ethernet port, and touch-sensitive
display. In this study, the FEZ Spider Starter Kit from
98
Figure 3: Example Visual C# code
Figure 4: Aspects of learning with .NET Gadgeteer
GHI Electronics 7 was used. In addition, there are many
other sensors and other modules that can be added separately.
Devices can be constructed by connecting modules
with cables and then programming it with respect to the
events triggered when using the device (for example, a ButtonPressed
or PictureCaptured event in case of a digital
camera). The programming language used is Visual C#
with support for Visual Basic .NET coming soon. Figure
1 shows the .NET Gadgeteer hardware with a range of
modules connected together. Figure 2 shows the graphical
Gadgeteer Designer that is used to generate code for
each module being used. Figure 3 shows the programming
environment in Visual C#.
Programming a .NET Gadgeteer device involves learning
to program in Visual C#, but there are other skills involved
in designing and implementing a gadget. Connecting the
input and output modules to the mainboard will give students
more of an understanding of hardware, and there is
the option to teach students the underlying electronics of
the board. A so-called extender board can be used to attach
third-party hardware. In addition, the design of the case
and the interaction of the user with the device will involve
an understanding of user-interface design and physical form
factor. Being able to program the screen will also teach
students an understanding of graphics objects. Overall, the
experience of using .NET Gadgeteer will cover a variety of
aspects of the CS school curriculum, as shown in Figure 4.
4. CASE STUDY
The case study was designed as an observation of how
students and teachers used .NET Gadgeteer and their experiences.
4.1 Aims and objectives
7 http://www.ghielectronics.com/

The aims of the case study were to investigate the potential
of .NET Gadgeteer to consider which age group, which
type of lessons, and which type of activities would be suitable
for learning with the use of this tool. Specifically, the
following research questions were addressed:
1. Does .NET Gadgeteer yield an engaging and motivating
environment to work with in schools?
2. Are the initial teaching materials sufficient for students
in lower and upper secondary schools to build .NET
Gadgeteer devices?
3. Could .NET Gadgeteer be used to support student
learning in Computer Science in school?
4. Is .NET Gadgeteer most suitable in schools as an extracurricular
activity or could it have a place in the main
curriculum (in England and Wales)?
4.2 Participants
In this study, .NET Gadgeteer was used for the first time
with secondary school students. Eight local schools in Cambridgeshire
(UK) volunteered to be involved. The schools
were a mixture of age 11-16 and age 11-18 schools, state
schools (seven) and private schools (one), mixed schools
(six), girls’ schools (one) and boys’ schools (one). The teachers
were initially trained in the use of .NET Gadgeteer and
provided with lesson plans and example projects. Munson
et al. [20] point out that in order to engage students, it is
also necessary to support teachers. We devised this pilot
project with an initial learning session for teachers at Microsoft
Research Cambridge, for them to cascade to pupils.
Teachers introduced .NET Gadgeteer to their schools in the
form of after-school or lunchtime clubs. The ages of the students
attending the club ranged from 11 to 15 year, with one
school choosing to use .NET Gadgeteer with an older group
of 17 year olds. The students worked in groups of three to
four with one .NET Gadgeteer kit per group. There was
no requirement, given the existing ICT curriculum, for the
schools to have taught programming before and it was acknowledged
that it would not be possible to teach significant
amounts of C# programming in the course of the pilot.
The materials developed consisted of eight lesson plans
with approximately one hour’s teaching material in each.
The materials included learning objectives, aims of the session,
a starter session as well as the main session in several
steps, and extension material. The lesson plan contained all
the instructions for carrying out the tasks. The session plans
had aims and outcomes, in terms of programming skills to be
learned, but the approach taken was actually to get devices
working by following through instructions.
The students were loaned .NET Gadgeteer kits for four
month. Teachers planned to run either one after-school club
or one lunchtime club each school week. The plan was for
them to run around ten hours of activity.
4.3 Methodology
The students’ progress with .NET Gadgeteer was monitored
by observation visits and their experiences evaluated
at the end of the project. A case study methodology was
used [24], taking the form of a collective case study,
whereby a variety of sources were used to give an overall
picture of how .NET Gadgeteer could be used in schools.
An online questionnaire was designed to give teachers the
99
Figure 5: Extract from teachers’ questionnaire
opportunity to give feedback in their own time. This was
available via a web link and had a mixture of open and closed
questions. Teachers were also interviewed informally about
their experiences of the project. An extract of the teachers’
questionnaire is shown in Figure 5.
To collect verbal responses from students about their experiences,
short interviews were held with students who attended
at an event associated with the pilot. This represented
a sample of 16 out of 84 students who took part in
the pilot. The students volunteered to be interviewed, and
had the option of speaking to the researcher individually or
in pairs. This made it less stressful for the participants, although
the sampling was therefore not of a representative
group of the participating students. The interviewer asked
each of the students the following four questions:
1. What sort of things have you worked on in your .NET
Gadgeteer sessions?
2. What has been your favourite part of working with
.NET Gadgeteer and why?
3. Would you continue to use .NET Gadgeteer if you had
a kit at school or home?
4. Has working with .NET Gadgeteer increased your understanding
of how computers and electronic devices
work?
The researcher filmed a demonstration of their gadget if
they had made one. These interviews were transcribed and
coded and key themes drawn out of the resultant data. The
questions were designed to generate a range of data about
engagement with the project.
The short interviews were followed by an in-depth focus
group at one of the pilot schools. Two girls and two boys
were invited to participate in the focus group which was
conducted by the first author with a teacher present at all
times. A focus group has many advantages when seeking
to find out attitudes as the one-to-many dialogue, perhaps
including a difference of opinion, may allow other issues to
be raised which may not arise in an interview where there

Figure 6: Pedagogical model
is no opposing view. The focus group was planned such
that a teacher was present and the students all knew each
other. It included more in-depth questions about the students’
engagement with technology and computer programming
more generally, with the intention of providing an elaboration
of the points made in the short interviews and in the
teacher questionnaires. The questions were based around
the following themes: experiences of the .NET Gadgeteer
project, learning new technologies, and aptitudes and difficulties
with learning to program. The focus group discussion
was carried out by the first author, and recorded, transcribed
and then coded using TAMS analyser 8 .
4.4 Pedagogical approach
The pedagogical approach taken with the teaching materials
was based on a two-phase model whereby the first
phase involved instruction and the second phase involved
presenting students with a challenge. The students needed
some instruction on how to use .NET Gadgeteer and the
instructions led them through the implementation of three
devices. Teachers utilised the materials differently, some setting
student tasks based on the materials, and others letting
the students work through the materials at their own pace.
Each teacher had their own different style of teaching and
this was noted in observations made by the first author at
visits to each of the schools. Despite the teacher-differences,
there was a common approach to teaching which led into the
challenge phase of the project. Students were encouraged to
create their own device for the end of the project. In order
to investigate how students were able to learn with the ‘tryit-and-see’
bricolage approach, students were encouraged to
be as inventive as they could in coming up with their own
ideas for devices. Figure 6 shows the rationale behind the
simple two-phase pedagogical model adopted.
5. RESULTS
Table 1 shows the schools, numbers of students attending
sessions, and the number of session plans they managed to
complete, although some schools ran more sessions and some
students used some of their own time to complete projects.
After completing the session plans, students then worked
on their own projects in Phase 2 of the pilot, applying what
they had learned to their own ideas. Students worked in
teams of between two and five students and developed a
range of small devices. Students also enjoyed building the
housing for the gadgets, which in some cases was quite sophisticated
using moulded plastic, and for other gadgets,
just as effective, using polystyrene or cardboard. Examples
8 http://tamsys.sourceforge.net/
100
Figure 7: Examples of devices created by students
with .NET Gadgeteer: “Robber Gadget” takes a picture
of a burglar when a precious item is taken off
the sensor platform; “Rainbow Press” is a reaction
game; “Alien Invasion” is a shooting game; “Gadgea-sketch”
is a drawing device.
of the gadgets developed by students are shown in Figure 7.
Table 1: Participants in the project
School Number Number Sessions Lowest Highest
sessions students (out of 8) year year
group group
A 10 8 5 Y9 Y10
B 4 6 4 Y7 Y13
C 4 24 3 Y7 Y8
D 5 6 6 Y11 Y13
E >10 10 6 Y8 Y11
F >10 3 7 Y9 Y10
G >10 10 6 Y7 Y11
H 7 18 4 Y9 Y9
(Y7 = age 11/12 (equivalent to 6 in USA);
Y13 = age 17/18 (equivalent to 12 in USA)
Several schools ran more than ten sessions although engagement
between the eight schools varied. The number of
students attending at each school varied from three to 24
students. None of the schools completed all of the eight session
plans provided, indicating that the material took longer
to cover and that the difficulty level may have been underestimated.
We originally aimed the project at Y9 students
(aged 13-14) but gave teachers the freedom to select appropriate
students as school environments differ. This had the
benefit of giving us some examples of projects from students
from Y7 to Y13, demonstrating the wide potential of .NET
Gadgeteer.
Teachers were asked to rate how they thought that the students
had mastered key programming concepts. It was not
expected that novice programmers would be able to achieve
a thorough understanding of Visual C# programming in just
10 hours of sessions, and we had no prior hypothesis about
how much students in such a mixed age group at different

Figure 8: Acquisition of programming concepts (average
across 8 schools)
schools with different teaching styles would learn. However,
it was encouraging to see that most schools felt that their
students had an understanding of assignment, data types,
variables and selection as a result of the pilot. Figure 8
shows the relative acquisition of programming concepts as
rated by their teachers.
Teachers were broadly positive about the achievements
of their pupils and the motivation provided by .NET Gadgeteer.
One teacher commented that “It was fun and really
nice to have things to touch and build. Some students were
very engaged in the whole thing” (School H, teacher).
There were a few technical problems experienced by teachers
as the platform had never been tested before on school
networks. These are discussed in Section 5.4.
5.1 Interviews with students
Approximately 85 students used .NET Gadgeteer in school
as part of this pilot and 50 attended the end-of-school pilot
event. Of these, 16 were interviewed during the day (14
boys, two girls). The comments from the student interviews
and the focus group clearly identified four emerging themes.
These were around the following features of the programme:
• Challenge and difficulty
• Creativity and freedom
• The tactile nature of .NET Gadgeteer
• The concept of “real programming”
These will be discussed in turn.
5.1.1 Challenge and difficulty
Without being prompted, students commented on the fact
that they had found working with .NET Gadgeteer quite
challenging at times. For example, one student said “I enjoyed
it but it was hard. And it was a challenge”(School C,
male), and another found it “. . . a big learning curve really”
(School G, male). One student tried to explain why some
other students from his school found the programming difficult:
“. . . but it took a great deal of effort, and not everyone’s
ready to . . . you know, actually step up to the challenge”
(School B, male). However, the level of challenge seemed
to be popular with several students: “. . . on Gadgeteer it’s
101
so much better because it’s harder, but that’s the good of it
you know” (School B, male). One young student with some
programming experience commented that it was “. . . not too
simple, so that you’re not really not learning that much, but
it’s not too complex, that you‘re not getting any help with
debugging” (School G, male).
There were fewer comments to the extent that it was not
too difficult, and these were from two students who had had
extensive programming experience from working on projects
at home and self-teaching. There was recognition from several
students that they had learned a lot from their experience:
“. . . we never learnt code before at all, so it most of
the things, well everything we learnt for Gadgeteer was stuff
that was new” (School A, male).
5.1.2 Creativity and freedom
The pilot was designed with a second phase whereby students
were to design a gadget or device of their choice, with
freedom limited only by the modules that were available in
the particular .NET Gadgeteer kit they were working with.
Students referred to the fact that they liked the freedom
to be creative and that they liked the tactile nature of the
modules: “You’re in control you can take an idea anywhere
and use the hardware that’s available to make it and without
limited . . . so the key it’s versatile” (School A, male). Another
student particularly liked the camera project, which
could be extended in many ways: “What I enjoyed most
about .NET Gadgeteer is the creativity you can have and the
challenge it poses . . . especially with the camera, I really enjoyed
that. Also trying to build your own sort of gadgets, and
that was, you could really use your imagination” (School G,
male). Another student commented on the freedom given to
use .NET Gadgeteer as they wanted: “You’re allowed to be
sort of creative and sort of like make anything so you weren’t
really limited to what you can make” (School A, male). An
older student, with experience of programming and the Arduino
platform, was very impressed by the flexibility it gave
him: “. . . you just plug it together and it works, you don’t
have to figure out the circuits for yourself . . . , but it’s brilliant,
really fun” (School D, male).
5.1.3 The tactile nature of .NET Gadgeteer
Six students made comments to the extent that the tangible
nature of the modules made it exciting to build and
program devices. For example, one of the boys commented
that “. . . you don’t just like have it all as pictures on the
screen, you actually have the stuff that you can put together
. . . ” (School C, male). During the event, the students were
proud to be able to demonstrate the devices that they had
made to other students, and their achievement was more
visible in its physical form. Another student commented
that developing a physical gadget meant that they could
build something that had a purpose: “You actually have
something you could hold and manipulate, that’s, it’s sort
of feeling a practical use for your programs, instead of just
doing it for the sake of it” (School G, male). Having both
aspects of development, with the hardware and the software,
also appealed to another student: “You need to program it,
you need to put it together, and it just makes it a whole lot
better” (School B, male).
5.1.4 Doing “real programming”
Students made reference to the fact that they knew that

Visual C# was a programming language used in industry by
professionals, and that they were using a tool that felt very
‘adult’. For example, one student commented that “This
was much better because it was more professional and, you
could do a lot more with it” (School G, male). The comment
made was in comparison to Scratch, which is the environment
with which many of the students were already familiar.
Another student compared .NET Gadgeteer to his experience
with Scratch also: “.NET Gadgeteer . . . sort of like
got proper programming if you know what I mean, . . . with
Scratch you’ve got words already set out for you and you can
sort of like make building blocks . . . but with .NET Gadgeteer
you’ve definitely got to have that more adult aspect of it and
the programming . . . ” (School G, male).
The pilot had encouraged students to consider doing more
programming in the future. Some students reflected that
their views on Computer Science as a subject had changed
as a result of the engagement with .NET Gadgeteer: “ . . . I
definitely want to take Computing and I’m looking to take
a job inside Computing as well when I’m older” (School G,
male). A young student mused about the possibilities that
might open for him if he pursued a career in Computer Science:
“It would be a nice career to take on . . . and so I’d
really like to be working at a place like Silicon Valley, it
would be a really nice opportunity” (School G, male).
Overall, the interviews elicited mostly positive comments
about .NET Gadgeteer. The results of the focus group discussion
will be reported next.
5.2 Focus group findings
This group of students included three from Y10 and one
from Y9 (this translates to the beginning upper secondary
school in Europe and the beginning of high school in the
USA). The focus group was held in School A with two boys
and two girls who had participated in the project. They were
asked similar questions to the short interviews initially and
similar themes emerged, for example, relating to freedom
and creativity: “I liked designing for the competition, the
fact that we got to design our own product”(School A, male).
The students also reported that they liked the tactile nature
of the kits, and that the first project was quite accessible:
“ . . . like the ease of it, it was quite easy to integrate the bits,
for instance making the camera was quite simple” (School
A, male). Other comments were made that mirrored the
content of the short interviews already reported on. They
commented that .NET Gadgeteer made the concept of the
technologies they use more accessible “because essentially we
had many of the bits of an iPod Touch, a camera and a touch
screen, and you can get the audio jack and stuff, to give you
an idea that you could perhaps make something like that and
that these technologies weren’t far away” (School A, male).
Students also discussed some of the key aspects that they
felt appealed about Computer Science to young people, what
helped them to learn, what qualities were useful in learning
programming and why other students might not succeed.
The first quality to be suggested was to be open-minded:
“Open-minded . . . because it’s like some people if they don’t
want to learn something then they most probably won’t listen
to it” (School A, female). The student went on to explain
that the stereotype of being “weird and nerds” could put
people off doing Computer Science if they were not openminded.
The students agreed between them a set of qualities that
102
might make students good at programming, or working with
.NET Gadgeteer, as follows:
• Open-mindedness
• Common sense
• Problem-solving skills
• Patience
• Imagination
• Intuition
Common sense was defined by the student as “just being
able to look at something, and then work out, kind of maybe
use your brain to go through what the computer is doing, the
way the computer is looking at it. There’s a lot of people
can’t do that. I think they find it a lot more difficult to
program” (School A, male). From the student’s point of
view (this student had taught himself some programming
already) this was common sense, whereas another might see
this as logical thinking. This was an interesting example of a
student with experience perhaps underestimating what was
difficult to a non-programmer [25]. The two girls were more
able to empathise with the issues that other students might
have in learning to program.
The notion of patience being a key to success in programming
was mentioned at other times in the focus group and
in one of the short interviews: “Problem-solving skills and
patience, because when you’re learning it, you have to have
a lot of patience in you because it might take some time to
learn everything, you can’t exactly rush through it otherwise
you won’t learn it properly” (School A, male).
Students also reported on the sources that they used to
assist them when they had problems and needed help with
their work in computing programming. They demonstrated
that they knew where they could look for help on the Internet:
“Looking at other people’s code” (School A, male), and
also: “I read tutorials and find the answer on the Internet”
(School A, male). The two girls were more reliant on teachers
and friends to help them with their coding: “Mainly the
teachers who help me try and learn and help me go through it
so it sticks in my head a bit more” (School A, female) and:
“Just having people explain it to you and if you go wrong
having people explaining exactly where you went wrong and
how you’re supposed to do it right, and things like that would
be quite helpful. Having someone that can really explain it
to you” (School A, female).
This illustrates some of the gender differences, as discussed
briefly below.
5.3 Girls and .NET Gadgeteer
In this study, we had not particularly focused on comparing
the responses from different gender groups in our design,
and we had a much smaller number of girls in the project
than boys. However, the comments from girls (two in the
short interviews and two more in the focus group) suggested
some surprise that they had managed to be successful with
.NET Gadgeteer. Girls reported that prior to using .NET
Gadgteer “I really thought I wouldn’t be able to do anything”
(School G, female). Another girl admitted that it hadn’t really
appealed to her: “. . . just thought it was all a bit, sort of
hard, just like why would I need to know” (School E, female).

The focus group girls were more confident: “Well, maybe a
little bit overwhelming at first but once you understand what
it all means it’s not that difficult once you know what you’re
doing” (School A, female).
The girls reported more confidence after they had finished
creating their gadgets with .NET Gadgeteer: “Yeah, it’s definitely
opened up more about what I know about programming”
(School E, female). Another of the girls said that she
was more confident since building her (highly innovative)
gadget: “I’ve really surprised myself, and I’d have to say,
I’ve grown in confidence as well” (School G, female).
One of the teachers (School C), in an interview, commented
on what girls enjoyed about the programme: “The
girls, I think they were just more interested in getting the
camera working, and they all wanted to be able to take pictures
of themselves, which kind of gives them that sense of
ownership”.
With such a small number of girls, we can only note these
gender differences; intuitively, though, it seems as if working
with this platform could appeal widely to both genders, as
gadgets with a socially useful function could be developed
over time that are more appealing to girls. The next section
addresses some of the limitations and problems we found
with .NET Gadgeteer on this first use in UK schools.
5.4 Problems with .NET Gadgeteer
In terms of their dissatisfaction with the project the main
complaint from students was about lack of time to complete
their projects: “ . . . we had after-school clubs every Wednesday,
but towards the end we ended up doing it at lunchtimes
as well” (School G, male). Another student reported spending
much time on his .NET Gadgeteer trying to complete a
project: “ . . . mainly just in lunchtimes, then Thursday and
Wednesday I’ve been going after school, and every opportunity
possible really . . . I’ve been trying to get things finished”
(School E, male).
Teachers also would have liked more time to complete
projects, indicating that working with .NET Gadgeteer can
encourage longer-lasting project work than we had originally
envisaged: “I would have loved to take it further if we had
had more time, and to see the possibilities and how far our
students would be able to take it. What could our students
create if they had another term? How advanced could our
students take it? What new systems, concepts could they interpret
and understand to produce bigger and better systems?
Just more time for students to play and to see the real potential
of the gadgets they create” (School A, teacher).
An issue for teachers was some of the technical difficulties
they had had at the outset when the pilot started. These
schools were some of the first public users ever of .NET
Gadgeteer, so some locked-down school networks posed a
challenge initially. This links to the fact that the students
felt that they did not have enough time to build what they
wanted to. For our research project, we had set these time
constraints. For future work with .NET Gadgeteer it is likely
that a full academic year would be needed to really benefit
and learn, as with any new skill. .NET Gadgeteer was
very new when the project started, as the teachers received
some of the first kits that were shipped over to the UK. The
students were using the .NET Gadgeteer environment before
the schools had had time to test it properly on their systems:
“I still have not been able to load the required software onto
the school system, and have had to work using a teacher
103
laptop that I had to battle 8 weeks for to get admin rights
to load the programs onto it” (School C, teacher). These
teething problems have now been eradicated as indicated
by this teacher: “ . . . there were some minor glitches with a
mismatch of software and hardware at the start” (School G,
teacher).
The teachers also wanted more materials, as identified
here: “Some form of ‘student book’ to allow them to track
and take notes might help” (School H, teacher). Another
teacher wanted more model projects to inspire students: “I
would have quite liked a very advanced gadget to have been
created and code provided, not for students to see code, but so
we could easily show some of the high potential that could be
achieved using Gadgeteer systems” (School A, teacher). Finally,
one teacher felt that their Y7 students were too young
and that the older students got more out of the experience:
“Great ideas and all the students were enthusiastic despite
often making very little progress. Our mistake - we opened
it to all age groups and while the younger students were very
keen, they lacked the concentration and basic programming
experience to get as much from it as the older students did”
(School G, teacher).
6. DISCUSSION
The results of this study of .NET Gadgeteer in secondary
schools demonstrate that students were engaged by .NET
Gadgeteer, although they did find it difficult. This is backed
up by their teachers’ comments. Challenge is seen as a positive
feature as reported by the students in the short interviews;
many students referred to this although it was not
specifically elicited. There were some very able students
within the pilot groups for whom the opportunity to use
.NET Gadgeteer provided them with a stimulating resource
that they very much enjoyed. However, these students had
volunteered to be part of the pilot groups, and more investigation
is needed to see how less motivated students experience
the platform.
Students automatically seemed to compare .NET Gadgeteer
with the Scratch programming environment, although
again this was not mentioned by the interviewer, presumably
because that is the experience that students had had so far
and as such the only experience with which they were able
to compare .NET Gadgeteer. Their comments implied that,
after using Scratch, they felt that .NET Gadgeteer allowed
them to write their own functions and have more freedom to
customise their own programs. What .NET Gadgeteer offers
after having learned Scratch is an introduction to a real
programming language and this was popular with students.
The introduction to programming provided by Scratch, if
incorporated in a teaching programme at school, will help
them prior to moving forward with .NET Gadgeteer.
Although a challenge, students seem to have enjoyed working
together in groups and creating something of their own.
The production of a physical device to demonstrate was
a motivating factor. Observations of students working revealed
that they worked well in groups in the main, with
teachers being flexible about sizes and make-up of groups.
Within groups, students assigned themselves different roles,
with some having more responsibility for the design of the
case or user-interface, with others writing more of the code.
This is one of the outcomes of the project that we had not
expected. To develop their own gadget, students were necessarily
involved in both problem-solving and some inde-

pendent research. They had to experiment with different
modules to find out exactly how they worked. The wider
skills they gain through engagement in these activities have
a positive effect on developing their personal, learning and
thinking skills.
The pedagogical model that we used encouraged a bricolage
approach. We wanted to provide some instruction, but
then to inspire children to experiment with modules and devices
on their own and come up with their own devices. For
some of the children, though, the gap between our Phase 1
and Phase 2 was too great and the teachers did not obviously
have the background or sufficient training in the platform to
support them. We need to build on the incremental problem
solving as used successfully by Stiller [28] and provide
smaller gaps in future materials, to maintain both teacher
and student confidence.
We have not specifically tested the students’ understanding
of programming principles in this project. Meerbaum-
Salant carried out a study whereby she measured the understanding
of key programming concepts of students who
had participated in a course using the Scratch programming
environment [19]. She reports that “the bottom-up programming
habit is clearly encouraged by the characteristics of the
Scratch environment and is in line with Papert’s philosophy
of constructionism . . . and with bricolage” [p.171] and
concludes that “. . . we are disturbed by the habits of programming
that we uncovered. These habits are not at all what one
expects as the outcome of learning computer science” [p.172].
It is not clear whether she is blaming the Scratch environment
in being too “easy” or the constructionist and bricolage
approach in her concerns about long-term effects of an exploratory
approach to programming. We would hope to be
able to show that an exploratory, and bricolage-centred, pedagogical
approach can be both motivating and teach good
programming habits, in the context of a real programming
language that works behind a motivating hardware environment.
As the platform is still new and materials are still under
development, we have not yet been able to replicate this
level of analysis of learning for .NET Gadgeteer. However,
although a bricolage approach seems implicit in ‘gadgetbuilding’,
it may not be as bottom-up as the Scratch programming
that Meerbaum-Salant describes. Where students
assemble their gadget before programming, they have put
together all the modules they need for the project and then
can break down the programming into events. This provides
more of a top-down, but also very concrete experience
of program development. More research is obviously
needed to discover if this is effective in the long-term acquisition
of key concepts. This will involve the development
of a suitable research instrument to measure learning with
.NET Gadgeteer.
Figure 9 adds the development of the wider personal,
learning and thinking skills to that shown in Figure 4. This
illustrates some of the potential wider skills that can be
gained by students working on .NET Gadgeteer in groups
on their own projects.
7. CONCLUSIONS AND FURTHER WORK
In this paper, we have presented our experiences of using
.NET Gadgeteer in schools with students of various ages
and backgrounds over a four-month period. Our findings
have revealed that students are very engaged by it, and are
104
Figure 9: Wider skills development with .NET Gadgeteer
inspired to build devices for which they can see a real purpose.
Through analysis of data from students and teachers,
we have uncovered some key features of .NET Gadgeteer
that provide motivation and interest: challenge, freedom to
explore, tangibility and exposure to the real world of Computing.
We are partway towards answering our research questions:
1. Is .NET Gadgeteer an engaging and motivating environment
to work with in schools? We have found that
students reported positively about using .NET Gadgeteer
for a range of reasons, as described in this paper.
2. Can students in lower and upper secondary schools use
.NET Gadgeteer to build devices with the initial teaching
materials developed? Students have been successful
in building devices; however, the teachers indicated
that the materials need to be developed further.
3. Could .NET Gadgeteer be used to support student learning
in Computer Science in school? With the changes
in the curriculum in the UK, involving the introduction
of more Computer Science in secondary schools, .NET
Gadgeteer provides an environment that is both engaging
and encourages experiential learning. Working
with .NET Gadgeteer has the potential to develop students’
ability to work collaboratively. Programming
skills can be developed through a mixture of instruction
and an exploratory approach to learning.
4. Is .NET Gadgeteer most suitable in schools as an extracurricular
activity or could it have a place in the curriculum
(in England and Wales)? The research revealed
that working through the materials and creating
the devices was quite intensive and required more time
than available in an after-school club. Further work is
needed to write longer courses that would work with
curricular in UK schools and within examined courses.
The .NET Gadgeteer platform is still in its infancy as it
was only launched in August 2011. At the time of writing,
more modules are being developed by an increasing number
of manufacturers.
The pilot project was successful in its aims and objectives
which were to establish that .NET Gadgeteer can be a useful

tool for teachers and students in the classroom. Its tangible
nature engenders curiosity and creativity, and motivates
students to explore, bricolage-fashion, how to program their
device. The physical nature of the platform encourages a
learning-by-doing experiential approach to learning and in
addition lends itself to collaborative working, whereby students
with a range of skills and abilities can support and
learn from each other.
As an initial pilot project with a new platform, this research
has highlighted particular areas of further work. In
future studies using .NET Gadgeteer, we would like to investigate
how students’ perception of the importance of challenge
and creativity links to the development of secure and
robust programming understanding. It is thus proposed that
further studies with .NET Gadgeteer focus on the facilitation
of the acquisition of certain programming concepts. The
two-phase pedagogical model utilised in the pilot project will
be developed further to provide more staged support, whilst
retaining an exploratory and experiential approach.
8. ACKNOWLEDGMENTS
With thanks to Steve Hodges and Steven Johnston, Microsoft
Research, and to all the participating teachers and
students for their assistance and enthusiasm during the pilot
project. The first author would like to thank Microsoft
Research for supporting her work.
9. REFERENCES
[1] O. Astrachan, J. Cuny, C. Stephenson, and C. Wilson.
The CS10K Project: Mobilizing the Community to
Transform High School Computing. In Proceedings of
the 42nd ACM Technical Symposium on Computer
Science Education, SIGCSE ’11, pages 85–86, New
York, NY, USA, 2011. ACM.
[2] M. Ben-Ari. Constructivism in Computer Science
Education. In Proceedings of the 29th SIGCSE
Technical Symposium on Computer Science Education,
pages 257–261. ACM, 1998.
[3] T. Brinda, H. Puhlmann, and C. Schulte. Bridging
ICT and CS - Educational Standards for Computer
Science in Lower Secondary Education. Proceedings of
ITICSE’09, Paris, France, July 6-9 2009.
[4] L. Carter. Why Students with an Apparent Aptitude
for Computer Science Don’t Choose to Major in
Computer Science. In Proceedings of the 37th SIGCSE
Technical Symposium on Computer Science Education,
pages 27–31. ACM, 2006.
[5] S. Cooper, W. Dann, and R. Pausch. Alice: a 3-D
Tool for Introductory Programming Concepts.
J.Comput.Sci.Coll., 15(5):107–116, Apr. 2000.
[6] T. Crick and S. Sentance. Computing at School:
Stimulating Computing Education in the UK. In
Proceedings of the 11th Koli Calling International
Conference on Computing Education Research, Koli
Calling ’11, pages 122–123, New York, USA, 2011.
ACM.
[7] Q. Cutts, S. Esper, and B. Simon. Computing as the
4th R: a General Education Approach to Computing
Education. In Proceedings of the 7th International
Workshop on Computing Education Research, ICER
’11, pages 133–138, New York, NY, USA, 2011. ACM.
105
[8] T. Downes and D. Looker. Factors that Influence
Students’ Plans to Take Computing and Information
Technology Subjects in Senior Secondary School.
Computer Science Education, 21(2):175–199, 2011.
[9] A. Fisher and J. Margolis. Unlocking the Clubhouse:
the Carnegie Mellon Experience. SIGCSE Bulletin,
34(2):79–83, June 2002.
[10] O. Hazzan, J. Gal-Ezer, and L. Blum. A Model for
High School Computer Science Education: the Four
Key Elements that Make It! In Proceedings of the 39th
SIGCSE Technical Symposium on Computer Science
Education, pages 281–285. ACM, 2008.
[11] F. T. Hofstetter. Multimedia Literacy.
Irwin/McGraw-Hill, Boston, 2nd edition, 1997.
[12] M. S. Horn, R. J. Crouser, and M. U. Bers. Tangible
Interaction and Learning: the Case for a Hybrid
Approach. Personal Ubiquitous Computing,
16(4):379–389, Apr. 2012.
[13] M. Kordaki, M. Miatidis, and G. Kapsampelis. A
Computer Environment for Beginners’ Learning of
Sorting Algorithms: Design and Pilot Evaluation.
Computers and Education, 51:708–723, 2008.
[14] C. Lévi-Strauss, J. Weightman, and D. Weightman.
The Savage Mind. University of Chicago Press, 1966.
[15] M. Kolling. The Greenfoot Programming
Environment. Transactions on Computing Education,
10(4):14:1–14:21, Nov. 2010.
[16] M. MacLaurin. Kodu: End-User Programming and
Design for Games. In Proceedings of the 4th
International Conference on Foundations of Digital
Games, FDG ’09, New York, USA, 2009. ACM.
[17] J. Maloney, M. Resnick, N. Rusk, B. Silverman, and
E. Eastmond. The Scratch Programming Language
and Environment. Transactions on Computing
Education, 10(4):16:1–16:15, Nov. 2010.
[18] P. Marshall. Do Tangible Interfaces Enhance
Learning? In Proceedings of the 1st International
Conference on Tangible and Embedded Interaction,
TEI ’07, pages 163–170, New York, NY, USA, 2007.
ACM.
[19] O. Meerbaum-Salant, M. Armoni, and M. Ben-Ari.
Habits of Programming in Scratch. In Proceedings of
the 16th Annual Joint Conference on Innovation and
Technology in Computer Science Education, ITiCSE
’11, pages 168–172, New York, NY, USA, 2011. ACM.
[20] A. Munson, B. Moskal, A. Harriger,
T. Lauriski-Karriker, and D. Heersink. Computing at
the High School Level: Changing What Teachers and
Students Know and Believe. Computing Education,
57(2):1836–1849, Sept. 2011.
[21] S. Papert. Mindstorms: Children, Computers, And
Powerful Ideas. Basic Books, 1993. 92053249.
[22] L. Pollock and T. Harvey. Combining Multiple
Pedagogies to Boost Learning and Enthusiasm. In
Proceedings of the 16th Annual Joint Conference on
Innovation and Technology in Computer Science
Education, ITiCSE ’11, pages 258–262, New York,
NY, USA, 2011. ACM.
[23] M. Richards, M. Petre, and A. K. Bandara. Starting
with Ubicomp: Using the Senseboard to Introduce
Computing. In Proceedings of the 43rd ACM Technical
Symposium on Computer Science Education, SIGCSE

’12, pages 583–588, New York, NY, USA, 2012. ACM.
[24] R.Stake. The Art of Case Study Research. London,
Sage, 1995.
[25] C. Schulte and M. Knobelsdorf. Attitudes Towards
Computer Science-Computing Experiences as a
Starting Point and Barrier to Computer Science. In
Proceedings of the 3rd International Workshop on
Computing Education Research, ICER ’07, pages
27–38, New York, NY, USA, 2007. ACM.
[26] R. E. Stake. The Art of Case Study Research. Sage
Publications, 1995. 95004979.
[27] C. W. Starr, D. Bergman, and P. Zaubi. The
Development and Implementation of a Context-Based
Curricular Framework for Computer Science
Education in High Schools. In Proceedings of the 14th
Annual ACM SIGCSE Conference on Innovation and
Technology in Computer Science Education, pages
283–287. ACM, 2009.
[28] E. Stiller. Teaching Programming Using Bricolage.
J.Comput.Sci.Coll., 24(6):35–42, June 2009.
[29] The Royal Society. Shut Down or Restart? The Way
Forward for Computing in UK Schools. Technical
Report January 2012 DES2448, The Royal Society,
2012.
[30] S. Turkle and S. Paper. Epistemological Pluralism and
the Revaluation of the Concrete. Journal of
Mathematical Behaviour, 11(1):pp 3 – 33, 1992.
[31] N. Villar, J. Scott, and S. Hodges. Prototyping with
Microsoft .NET Gadgeteer. In Proceedings of the 5th
International Conference on Tangible, Embedded, and
Embodied Interaction, TEI ’11, pages 377–380, New
York, NY, USA, 2011. ACM.
[32] N. Villar, J. Scott, S. Hodges, K. Hammill, and
C. Miller. .NET Gadgeteer: A Platform for Custom
Devices. Proceedings of Pervasive 2012, Lecture Notes
in Computer Science, 2012.
[33] C. Wilson, L. A. Sudol, C. Stephenson, and
M. Stehlik. Running on Empty: The Failure to Teach
K–12 Computer Science in the Digital Age. Technical
report, CSTA, 2010.
[34] S. Yarosh and M. Guzdial. Narrating Data Structures:
the Role of Context in CS2. J.Educ.Resour.Comput.,
7(4):6:1–6:20, Jan. 2008.
106

eledSQL – A New Web-Based Learning Environment for
Teaching Databases and SQL at Secondary School Level
ABSTRACT
Andreas Grillenberger
University of Erlangen-Nuremberg
Didactics of Informatics
Martensstr. 3
91058 Erlangen
mail@andreas-grillenberger.de
Data modeling using databases and SQL is a fundamental
part of the curriculum of secondary computing education
in Germany. Professional database tools like HeidiSQL,
phpMyAdmin or Microsoft Access are often used in class as
a “learning software”, although these tools have been developed
for managing complex databases, often for companies,
and not for educational purposes. Such tools offer a wide
range of functions of which only a small part is required by
secondary computing education. At the beginning of such
instruction, students can hardly work independently with
these programs, as they do not know the database language
SQL by then. This often leads to theory-loaded introductory
phases of such classes or alternatively to a usage of such
tools like spreadsheet programs. To address this problem,
a new web-based learning environment for databases and
SQL (named eledSQL), also suitable for mobile devices and
only with the functionality needed for secondary computing
education, was developed. The basic idea was to initially
allow students to make database queries using natural language
and then gradually introduce them to the use of SQL.
Starting with a problem analysis and a discussion of related
work in the field of teaching databases and SQL, in this paper
the conception of eledSQL, its implementation and first
experiences with its practical use are described.
Categories and Subject Descriptors
K.3.2 [Computers and Education]: Computer and Information
Science Education—Computer science education
General Terms
Design, Human Factors, Languages
Keywords
Databases, SQL, Secondary Education, Learning Environment,
Web-based Learning, Mobile Learning, Tools
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WiPCSE 2012 Hamburg, Germany
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$15.00.
107
Torsten Brinda
University of Erlangen-Nuremberg
Didactics of Informatics
Martensstr. 3
91058 Erlangen
brinda@cs.fau.de
1. MOTIVATION
In secondary computing education in Germany data modeling
using databases and SQL is a fundamental part of the
curriculum [1]. As “learning software” professional database
tools such as HeidiSQL 1 , phpMyAdmin 2 or Microsoft Access
are used in class – tools, which were not designed for educational
purposes but for the professional management of
databases. In supplementary material to the Bavarian computer
science high school curriculum [16], for example, either
the use of the XAMPP package 3 (with web server, MySQL
database server, phpMyAdmin) or of Microsoft Access is recommended
by the authors. These professional tools have in
common that on the one hand they offer a wide range of
functions, of which only a small part is needed for secondary
computing education, and on the other they require previous
knowledge about databases and the SQL language for
its competent use. The use of such programs in introductory
lessons on databases is therefore hardly possible. As a
consequence, often either theory-loaded introduction phases
can be observed in class or the database software is used
like a spreadsheet program there. The second option entails
the risk that the students – because of the similarities of the
user interfaces of for example Microsoft Excel and the table
view of Microsoft Access – which supports their orientation
in the beginning – often have difficulties in recognizing the
differences between the two system categories and do not or
not fully understand the difference between a spreadsheet
and a database. Other problems arise, when the learners
find out how to use the QBE function (query by example)
within Microsoft Access, where the underlying SQL code is
generated automatically [5].
The use of XAMPP has the advantage that with the appropriate
user rights configuration the learners can experience
the advantages of multi-user access to a database.
With Microsoft Access this is potentially also possible but
considerably more complicated. Another advantage of the
client-server architecture is that the learners can access the
database from outside the school (e.g. to do homework),
if the database server is accessible on the Internet at all.
However, secondary school servers are often configured very
restrictively and often the installation of a database server
is not allowed there for security reasons. In this case, client
and server are sometimes installed in the computer room(s)
of the school on every single workstation, which is very la-
1 www.heidisql.com (Jun 20th, 2012)
2 www.phpmyadmin.net (Jun 20th, 2012)
3 www.apachefriends.org/en/xampp.html (Jun 20th, 2012)

orious, contrary to the client-server idea and excludes the
experience of multi-user access.
Summing up, although these systems cover all concepts to
be taught in class by corresponding functions, they appear,
mainly because of the required prior knowledge and the variety
of functions contained in them, not very recommendable
for a use in introduction phases into databases and SQL in
computer science classes in secondary schools. So the question
is: how should a system better suitable for this purpose
be designed?
2. TEACHWARE FOR DATABASES / SQL
As a next step published research on learning environments
in the area of databases and SQL was analyzed. In
the literature a number of such environments are described
which, however, refer mostly to higher education. Most of
these environments can be categorized as systems, which visualize
interactive SQL examples using multimedia or those
that provide SQL problems and evaluate student solutions
[2].
The first category includes, e.g. the desktop application
eSQL [10], which visualizes stepwise the selection of output
data instead of the presentation of a result for given SQL
queries, the QueryViz system [6] with a special focus on the
visualization of nested queries, the web-based SAVI system
(SQL Advanced Visualization [4]), which allows navigation
of the animations in both directions, and ADbC (Animated
Database Courseware) [13], which provides a large number
of interactive components for the visualization of concepts
of the fields of databases and database security. The ADbC
collection is extensive, sees itself as a blended learning material
and provides interactive visualizations of e.g. various
notations in the area of entity relationship (ER) modeling or
of prepared questions, e.g. the transfer of given ER diagrams
into tables with a visualization of the consequences of incorrect
answers. In addition, visualizations of the execution of
predefined SQL queries are provided. Other systems focus
on the visualization of specific aspects such as on the process
of ER modeling [9] or the normalization of a database [11].
The second category includes systems such as SQLTutor
[12] and SQLator [15]. The learners are confronted with
problems of everyday life and must create respective SQL
queries to solve these problems. The solutions are then analyzed
by these systems using heuristics, correct solutions are
presented and any errors explained as far as possible. Elements
of both categories can be found in web-based systems
such as SQLzoo.net and Teradata SQL Assistant, which provide
a set of predefined and not modifiable tables and queries
plus a detailed feedback upon execution [5].
The most far-reaching integrated approach finally comes
from Brusilovsky et al. [2, 3], who combined interactive
examples, an SQL knowledge tester, adaptivity and SQL
problems in a self-developed LMS using an underlying SQL
ontology.
All these systems are pure learning environments. Some
include hard-wired examples that are not changeable, others
focus only on parts of the school-relevant content but all
emphasize the explanation of concepts and neglect productive
work. A system suitable for secondary schools should
combine learning and working. This also requires an adjustment
to the special needs of schools with teachers, classes
and students.
108
3. REQUIREMENTS ANALYSIS
Based on the results of the analysis of database tools
and learning environments, we propose a new system better
suited to the needs of introductory school classes. Such
a system should be accessible with any web browser over
the Internet and offer software-ergonomic user interfaces for
both desktop pcs and mobile devices of all kinds in accordance
with the currently changing media use in schools [14].
The system’s user interface should be easily adaptable to the
corporate design of an educational institution. Installation
on a school server or an external web space (in the case of
school safety concerns) should be as simple as possible in order
to keep the resulting workload for teachers low. A common
weakness of previously published learning environments
(see sect. 2) is that they provide many different useful visualizations
and problems but that their integration into basic
learning management system functionality adapted to the
school needs is missing. The roles and user rights concept
to be implemented allows differentiated functions and rights
for teachers, school classes, learners and groups of learners
as well as an administrator. This provides the possibility of
adapting such a system to meet the specific needs of different
computer science courses of a school and its students and to
enable their simultaneous work with the system. Concerning
content, the students should be introduced to working with
databases and SQL with the system. This results in the
requirement that the system must be customizable to the
learning progress of learners and groups of learners. This
could be realized by three different modes to interact with
databases. At the lowest level the formulation of queries
is partly based on forms and partly on natural language.
Prior SQL knowledge is not required there. On the next
higher level queries can be entered as on the lowest level, in
addition, however, the underlying SQL syntax appears automatically
(see fig. 1). On the third level the interaction with
the database should be only possible with SQL. The system
should provide only the functionality needed in class and try
not to compete with professional database tools. The implemented
functionality should cover typical secondary schools’
database curricula like the Bavarian one [16].
Users of the role of teachers should be able to create student
accounts or import accounts from a file and manage
their user rights. Students should be assignable to a class
or a course by them. Furthermore, groups of students for
group work should be definable. Another point of criticism
of existing learning environments was a lack of adaptability
to the specific target group (see sect. 2). The teacher should
be able to set the query mode (see above) for entire classes.
Many learning environments provide predefined tables and
queries, which can not be changed and adapted to a specific
course. Therefore, a teacher should be able to create own
table templates, import such templates from files in CSV
format and share them with classes, groups or learners as a
copy. By this, tables can easily be recovered when needed
again. To focus the educational processes on the essential,
SQL functions, such as select, update, insert, delete,
show (or their natural language representation), should be
activatable and also deactivatable for specific tables. At various
locations in the system, the teacher should be able to
adjust teaching texts and provide stepwise assistance for the
SQL queries. A teacher should finally be able to switch to
the perspective of any learner and use all his or her features.
A user in the administrator should additionally be able to

create and manage teacher accounts.
Users in the role of students should be able to perform all
activated operations on the tables provided by the teacher.
As the students work only on table copies all activated functions
can be executed without any risk of destruction. This
also supports exploratory learning. There is also the possibility
that students, groups or classes can create, edit and
delete tables together, if this feature has been enabled by
the teacher. The aim is to make the learners experience the
multi-user mode of a database. The tables provided by the
teacher can only be deleted though, if this has explicitely
been approved by him or her. Finally, the learners should
be able to see only those materials provided to them, their
group or class and not those provided to any others.
4. DESIGN AND IMPLEMENTATION
The development process of the specified system, the Erlangen
learning environment for databases and SQL – eled-
SQL, with requirements analysis, use cases, database design,
class diagrams and implementation of the software is
described in detail in [7]. In the following, we sketch only
some important aspects. Due to the requirements that the
system should be web-based and that the installation efforts
should be kept low for teachers, the decision for a serverbased
system with browser-based access was made. To meet
school conditions, it was taken into account that on the one
hand, the components needed on an internal school server to
install eledSQL are standard components free of charge and
that the software can also be run via an external web hoster,
if the school safety conception is contrary to the installation
on site. External web hosters usually offer pre-structured
packages where, for example, usable programming languages
and databases are predefined. Because the language PHP
and MySQL databases are considerably more widespread
than their alternatives, both have been chosen for the development
of eledSQL. PHP is very well suited for the development
of web applications and also supports object-oriented
software development, so that the system was developed in
an object-oriented way because of the resulting structuring
benefits, where possible and appropriate. Conceptually, the
software was structured in three parts following the modelview-controller
pattern. The view contains templates that
correspond to the user interface (see fig. 1) and can be
Figure 1: User interface of eledSQL system
109
easily adapted to the corporate design of a school without
deep PHP knowledge. In order to support different language
versions of the software, all texts of the user interface
have been moved to a language file. The model contains
functional classes that generate content based on user input
and database access. The controller mediates between
these levels (see fig. 2). In order to use potentially other
Template
Controller
Functions
MySQL
database
Login
Single column
Database
Learner
Database
access
MySQL any
PDO-compatible
database
Two columns
Teacher
Invocation of the responsible controller
Role-dependant
functions
Invocation
of the template
Invocation
of functions
Initialisation of function classes
Initialisation of the used database
PDO
database
Legend
Template
index file
main class
PDO
database
Architectural level
User
Index file
Main class
Registry
(Procedural) Index file
(Object-oriented) Classes
(Object-oriented) Database
classes
Figure 2: Architecture of the eledSQL system
databases like PostgreSQL an additional abstraction layer
from the database queries using PHP Data Objects (PDO)
was integrated into the system. This makes the system even
more flexible. To implement the security requirements first
separate databases for system data, classes, groups and individuals
were considered. Since external web hosters typically
limit the number of databases this option was dismissed.
For this reason, the system uses a single database
in which the table names reflect the intended addressees.
System tables (marked with “sys”, such as user accounts)
are processable only by means of the management function
built-in into eledSQL or an external program. Template tables
(marked with “tpl”) can be created, filled with data
and then be made available to the learners as a copy by
a teacher. Tables marked with an identifier of the respective
class, group or learner are available only for the referred
learners. Before executing an SQL statement it can be easily
verified by checking the label of each involved table whether
the permission to do so is given. Furthermore, it is easy to
check whether a student is allowed to perform the respective
functions on the tables involved. Also, SQL injections are
therewith preventable.
The web-based user interface of the system of eledSQL
was tested with major browsers of the operating systems
Windows, Linux and MacOS. Furthermore, its usability was
checked with some probands. Identified problems were eliminated.

5. FIRST CLASSROOM EXPERIENCES
An early version of eledSQL was used within a teaching
internship of the first author at a secondary school in Erlangen
(Germany), where a 9 th grade class (learners 14 to
15 years old) was introduced into the field of databases. Before
the teaching internship the class had learned to deal
with spreadsheets. The experiences within this course were
throughout positive – the learners could use the tool without
a detailled introduction into the software itself. Thereby
they were able to discover the differences between the wellknown
spreadsheets and the new databases by dealing with
exercises prepared by the teacher. For example, they were to
find the address of a person in a spreadsheet and a database
with about 1.000 entries each or all persons with names including
a defined string. While dealing with these exercises,
they understood the different fields of application of
databases and spreadsheets. This is a huge adavantage over
the conventional way of teaching – which is in many cases
to tell the differences to the pupils, because they cannot
discover them theirselves due to missing SQL knowledge.
This sometimes results in motivational problems, especially
when the pupils cannot realize why “the same application
with another name” should have these pros and cons, when
only using the tabular view at the beginning.
6. SUMMARY AND OUTLOOK
In this paper we motivated the need for a new learning
environment on databases and SQL for the computing education
at secondary level. We analyzed database software
and learning environments used in education and derived
from the analysis results and curricular needs requirements
for the new learning environment, whose conception, design,
implementation and first experiences of the educational use
we described afterwards. The current version of eledSQL
is available for free (Open Source License) at Sourceforge 4 .
The development of this software is part of a bigger student
project that is continued as part of another bachelor thesis,
where a detailed analysis of the classroom use in two parallel
classes is made. A further development of the system
is also planned. In a next release assessments and teaching
material should be integrated.
7. REFERENCES
[1] T. Brinda, H. Puhlmann, and C. Schulte. Bridging
ICT and CS: educational standards for computer
science in lower secondary education. In Proceedings of
the 14th annual ACM SIGCSE conference on
Innovation and technology in computer science
education, ITiCSE ’09, pages 288–292, New York, NY,
2009. ACM.
[2] P. Brusilovsky, S. Sosnovsky, D. H. Lee, M. Yudelson,
V. Zadorozhny, and X. Zhou. An open integrated
exploratorium for database courses. In Proceedings of
the 13th annual conference on Innovation and
technology in computer science education, ITiCSE ’08,
pages 22–26, New York, NY, 2008. ACM.
[3] P. Brusilovsky, S. Sosnovsky, M. V. Yudelson, D. H.
Lee, V. Zadorozhny, and X. Zhou. Learning SQL
programming with interactive tools: From integration
to personalization. Transactions on Computing
Education, 9(4):19:1–19:15, Jan. 2010.
4 http://eledsql.sourceforge.net/
110
[4] M. Cembalo, A. De Santis, and U. Ferraro Petrillo.
SAVI: a new system for advanced SQL visualization.
In Proceedings of the 2011 conference on Information
technology education, SIGITE ’11, pages 165–170, New
York, NY, 2011. ACM.
[5] J. Cigas and B. Kushan. Experiences with online SQL
environments. Journal of Computing Sciences in
Colleges, 25(5):251–257, May 2010.
[6] J. Danaparamita and W. Gatterbauer. Queryviz:
helping users understand SQL queries and their
patterns. In Proceedings of the 14th International
Conference on Extending Database Technology,
EDBT/ICDT ’11, pages 558–561, New York, NY,
2011. ACM.
[7] A. Grillenberger. Entwurf und Implementierung einer
Lernumgebung für den Informatikunterricht zur
Unterrichtssequenz Datenbanksysteme (in German).
Bachelor’s thesis, University of Erlangen-Nuremberg,
Didactics of Informatics, Erlangen, Germany, 2012.
[8] M. Guimaraes and M. Murray. Using animation
courseware in the teaching of database security. In
Proceedings of the 8th ACM SIGITE conference on
Information technology education, SIGITE ’07, pages
253–258, New York, NY, USA, 2007. ACM.
[9] L. Hall and A. Gordon. A virtual learning
environment for entity relationship modelling. In
Proceedings of the 29th ACM SIGCSE technical
symposium on Computer science education, SIGCSE
’98, pages 345–349, New York, NY, 1998. ACM.
[10] R. Kearns, S. Shead, and A. Fekete. A teaching system
for SQL. In Proceedings of the 2nd Australasian
conference on Computer science education, ACSE ’97,
pages 224–231, New York, NY, 1996. ACM.
[11] H.-J. Kung and H.-L. Tung. A web-based tool for
teaching data modeling. Journal of Computing
Sciences in Colleges, 26(2):231–237, Dec. 2010.
[12] A. Mitrovic. Learning sql with a computerized tutor.
In Proceedings of the twenty-ninth SIGCSE technical
symposium on Computer science education, SIGCSE
’98, pages 307–311, New York, NY, 1998. ACM.
[13] M. Murray and M. Guimaraes. Animated database
courseware: using animations to extend conceptual
understanding of database concepts. Journal of
Computing Sciences in Colleges, 24(2):144–150, Dec.
2008.
[14] C. Piech and E. Roberts. Informatics education using
nothing but a browser. Proceedings of the IFIP
Conference on ICT and Informatics in a Globalised
World of Education, Mombasa, Kenya, August 2011,
2011. http://cs.stanford.edu/˜eroberts/papers/
NothingButABrowser.pdf.
[15] S. Sadiq, M. Orlowska, W. Sadiq, and J. Lin.
SQLator: an online SQL learning workbench. In
Proceedings of the 9th annual ACM SIGCSE
conference on Innovation and technology in computer
science education, ITiCSE ’04, pages 223–227, New
York, NY, 2004. ACM.
[16] Staatsinstitut für Schulqualität und
Bildungsforschung. Informatik am
Naturwissenschaftlich-technologischen Gymnasium,
Jahrgangsstufe 9 (Handreichung, in German), 2004.

ABSTRACT
Bringing Contexts into the Classroom –
A Design-Based Approach
Detlef Rick, Marcel Morisse, Ingrid Schirmer
Universität Hamburg, Department of Informatics
Vogt-Kölln-Straße 30
D-22527 Hamburg, Germany
{rick | morisse | schirmer}@informatik.uni-hamburg.de
Project-based learning and application-oriented approaches
drawing on real-world issues and examples, have always been
widely used in both secondary and higher Computer Science
(CS) education. Recent demands for teaching CS or Informatics
1 in context and building courses on coherent application
areas outside CS go even further [4, 23]. In Germany the
open working group IniK (Informatik im Kontext) compiles
educational material and concepts for teaching Informatics
in context [42, 24, 10, 8]. However, there is only little theoretical
foundation for IniK as an educational approach and
its implementation at schools and universities [7].
In this paper we suggest a process model for bringing
‘real-world’ application contexts into the Informatics classroom.
The proposed model is grounded on a design-based
study and has evolved from theoretical and practical insights
gained in five university-level project courses and five associated
school projects. The model focuses on the development
of context artifacts which originate from the real-world context,
and teaching units building on the context artifacts.
When teaching units are instantiated in the classroom, context
artifacts are recontextualized in an educational context
and are employed as learning equipment fostering the students’
imagination, and thus anchoring the real-world application
context within the classroom context.
In the tradition of classical German Didaktik models [18,
28], the process model aims at providing a tool and a terminology
for the preparation and reflection of instruction. On
the basis of the model we discuss both the potential value
and the drawbacks of teaching Informatics in context.
Categories and Subject Descriptors
K.3.2 [Computers and Education]: Computer and Information
Science Education
1 We use the term Informatics in a broad sense with a view
to describe the disciplines of Computer Engineering, Computer
Science, Software Engineering, Information Systems
and others.
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WiPSCE 2012 Hamburg, Germany
Copyright 2012 ACM X-XXXXX-XX-X/XX/XX ...$10.00.
111
General Terms
Design, Human Factors
Keywords
Informatics in context, learning environment, process model,
greenfoot, design-based research
1. INTRODUCTION
The Great Hall at Hogwarts School of Witchcraft and
Wizardry is equipped with a ceiling that has been bewitched
to look like the sky outside. So for J. K. Rowling’s famous
first-year student Harry Potter 2 it was “hard to believe there
was a ceiling there at all, and that the Great Hall didn’t
simply open on to the heavens” (p. 129). Apparently the
founders of Hogwarts for some reason deemed it necessary
to let the school’s inside resemble the outside world in some
degree.
Recent demands for teaching Computer Science and Informatics
in context [4, 7, 23, 24, 39] request educators to
contextualize their teaching and to bring the outside world
into the classroom. Although CS “would seem to want to
connect to the real world” and although for some projectbased
and application-oriented courses it is not at all uncommon
to build on a consistent real-world context, Cooper and
Cunningham [4] criticize that “computer science curriculum
recommendations are silent about the opportunity to do so”
and that “most widely used texts do not seem to have much
content addressing real-world questions” (p. 5).
Guzdial [15] qualifies this critique, remarking that “it is
reasonable to fear that students who learn computing within
a context might over-specialize that knowledge” and that
“the additional knowledge taught about the context might be
a distraction, especially for the lower-ability students” (p. 5).
Guzdial comes to the conclusion that context can provide
additional relevance if there is evidence that students can’t
see meaning and usefulness of decontextualized content.
Instead of trying to answer the question of whether contextoriented
teaching approaches actually result in higher student
motivation and commitment and thus in better learning
and deeper understanding [39], in the following we propose
a process model for bringing ‘real-word’ contexts into the
classroom environment. The model has evolved from five
cycles of our university-level project course for Informatics
students majoring in Computer Science (CS), Software
2 J. K. Rowling. Harry Potter and the Philosopher’s Stone.
Bloomsbury, London, 1997.

Engineering (SE), Information Systems (IS), or Computer
Science Education (CSEd).
Our students were assigned to develop school projects in
the context of our department’s outreach programs conducted
in cooperation with schools and local software development
enterprises. Thus our project course followed a
contextualized approach in that our students were to develop
project management and software development skills
in the context of preparing Informatics teaching units and
designing computational artifacts for the classroom, i. e. in
an educational context outside Informatics. In the project
our students implemented Greenfoot [41] scenarios simulating
some technical and non-technical issues of ‘real-world’
contexts, namely the context of airport baggage handling
and the context of inventory control in retail stores.
The proposed process model is design-based in two respects.
On the one hand the development of teaching units
and the design of artifacts fitting the classroom context
are considered to be main activities in preparing contextoriented
teaching. Thus design activities are an integral part
of the proposed process. And on the other hand the model
itself is the result of an iterative design process.
Albeit grounded on our specific project experience the
model is reasonably abstract to present a reference framework
within which different conceptions of teaching Informatics
in context can be discussed.
2. DESIGN-BASED METHODOLOGY
In 1969 Simon introduced the concept of design sciences
as the sciences of the artificial [36]. According to Simon the
artificial is investigated not only in terms of what is, but also
in terms of what ought to be. Design studies therefore aim at
both the development of artifacts – e. g. software, processes,
institutions, or interventions [20, 40] – and the generation of
significant theories and usable knowledge, e. g. design principles
or patterns [1, 29]. Moreover design-based approaches
are characterized by an iterative design process, where the
artifacts are developed in a process of ‘progressive refinement.’
Thus, for instance, IS and SE can be characterized
as design sciences [17].
In education research, design-based research methodology
(DBR) is being increasingly utilized [1]. According to
Collins et al. [3] design experiments “bring together two critical
pieces in order to guide us to better educational refinement:
a design focus and assessment of critical design elements”
(p. 21). While design experiments are conducted
in specific educational contexts, they are aiming at general
design principles.
The artifact we developed in the course of our design experiments
was the process model for bringing contexts into
the classroom, described in section 4. The process model
was instantiated in five university-level software development
project courses and five associated week-long school
projects. In the project courses the students assumed diverse
roles. They developed (educational) software, prepared
school projects, and finally taught classes in the ‘classroom
context.’ Thus the assessment of the process model
was primarily based on the evaluation of the project courses,
grounded on:
• assessment of the software and other artifacts developed
by the students with respect to requirements,
• students’ project reports and course achievements,
112
• student feedback (oral, questionnaire),
• feedback from the boys and girls who attended the
associated school projects (oral, questionnaire), and
• feedback from the teachers of our cooperation school
(oral).
In the following section we describe how the process model
evolved from our understanding of software development as
decontextualization and recontextualization, in combination
with the seminal design of Greenfoot, and the idea of teaching
Informatics ‘in context.’
3. ORIGINATION OF THE MODEL
Our process model evolved in a multi-institutional project
setting including multiple arenas. In a project-based software
development course for Informatics students, first offered
in 2009/10, our students were assigned to develop an
Informatics school project. In the following we describe the
prerequisites that informed the design of our first project
course and the corresponding process model, as well as the
subsequent iterations of progressive refinement.
3.1 Phase 0: Prerequisites
Our objectives for the project course were twofold. On
the one hand our Informatics students were to develop, for
instance, project management and social skills, software development
methods and so forth. In this project we had external
clients – the cooperating schools – and a firm project
deadline. On the other hand we defined requirements concerning
the school project – the product.
With the school project we wanted to rouse the boys’ and
girls’ interest in Informatics and thus induce them to choose
optional Informatics courses at school. According to Knobelsdorf
and Schulte [23] context-oriented approaches are
auspicious with regard to increasing participation in Computing.
Therefore we were positive that a context-oriented
approach was motivating not only for our students but also
for the boys and girls who participated in the school project.
In order to present a broad image of our discipline we aimed
to expose the youngsters to a complex real-world context
where information technology can make a difference. Moreover
we wanted to provide insight into a few CS core concepts,
like variables, control structures and object orientation,
as well as into more application-oriented issues. So our
students were expected to develop multiple views of Computing
and to share these with the pupils, drawing examples
from one coherent application context. In cooperation with
a local software development enterprise, we chose the context
of airport baggage handling [31] and made it accessible
to the students by arranging an on-site inspection at the
Cologne Bonn Airport, and talks with domain experts from
both the airport and the software producer PSI Logistics
GmbH. Thus our students were provided insight into the
context they had to analyze and to understand in order to
present it to the class later on in the school project.
As a technological framework we used the educational
Java programming environment Greenfoot [41] which is also
a framework for the development of microworlds, i. e. 2D
games and simulations, or scenarios [16, 26]. Hence the software
our students were to develop, essentially was Greenfoot
scenarios.

3.1.1 Greenfoot as a framework for designing
learning environments
The idea of using Greenfoot for the project course arose
from the fact that Greenfoot is not just the programming environment
itself but features an easy-to-use Java API for the
development of 2D-graphical games and simulations. Since
the Greenfoot framework is based on Java, it facilitates the
development of quite complex scenarios: “While it is possible
to create simple games quickly and easily in Greenfoot, it
is equally possible to build highly sophisticated simulations
of complex systems [...]” (Kölling [27], p. 2).
Thus Greenfoot is, in principle, well suited to provide both
an introductory development environment and meaningful
contexts. Henriksen and Kölling suggest that Greenfoot
would open new possibilities for course design because it allows
for using multiple scenarios of varied complexity and
from different application areas [16]. Moreover scenarios
can, in principle, be adapted to different learning theories
and to different learning objectives related to both object
orientation and other informatical issues beyond programming.
These new possibilities are promising. But obviously the
freedom to develop and adapt scenarios, or maybe even just
to select them to match specific course objectives, involves
additional time and effort for course preparation (figure 1).
Traditional microworlds Greenfoot
Implement framework
(with fixed scenario)
Design exercise
Solve exercise
Framework
implementor
Teacher
Student
Implement framework
Design scenario
Design exercise
Solve exercise
Framework
implementor
Student
Teacher
Figure 1: Roles in creation and use of traditional
microworlds and Greenfoot scenarios, according to
Henriksen and Kölling [16]. Greenfoot scenarios –
games and simulations – can be shared in the Greenfoot
gallery or maybe along with some ‘nifty assignments’
[11] in the Greenroom. (Adapted from [16])
Our project course aimed for fathoming the possibilities
provided by the Greenfoot framework, and the limitations.
In consideration of Nygaard’s advice that teaching objectorientation
must start with a “sufficiently complex example”
[30], we focused on developing scenarios that were conspicuously
more complex than conventional microworlds. The
challenge was to design these complex scenarios in such a
way that they would fit well into the framework and at the
same time match the classroom requirements so as to arrange
learning environments around them.
While Greenfoot as an IDE is aimed at young students
[43], for advanced students it’s well-designed API gives a
good and concise example of an object-oriented framework.
In developing against the Greenfoot framework (and yet using
a professional development environment) our Informatics
students learn how to elegantly handle and extend a given
framework.
113
3.1.2 Software development as decontextualization
and recontextualization
Our department has a strong tradition in questioning the
“separability of production from the use of products in social
contexts” ([14], p. 49), that is associated with e. g. Floyd
and colleagues [12, 14], and that has been influenced by
the so-called Scandinavian Approach [13]. Typically, when
our students enroll for the project course they are already
acquainted with the ideas of agility [2] and the perspective
on software development as an evolutionary process that
intertwines software production, embedment and use.
Therefore another starting point for the development of
our process model is our understanding of software development
as a design process and a social activity incorporating
alternate decontextualization and recontextualization
activities. Instead of bringing up Dijkstra’s ‘firewall’ between
the realm of the ‘correctness problem’ and the realm
of the ‘pleasantness problem’ [9], the dialectic of decontextualization
and recontextualization emphasizes that software
is situated in human contexts.
Decontextualization refers to the analysis and translation
of contextualized meaning and ‘situated action’ [38] into
models, algorithms, programs and other computational artifacts.
On the other hand recontextualization is the transfer
and embedment of formalized decontextualized activities,
programs etc. into a human context and results in transforming
it. Figure 2 illustrates this ‘socio-technical core’ on
the basis of Rolf and colleagues [25, 33, 44].
Along with the ‘roles in creation and use of Greenfoot
scenarios’ shown in figure 1, we adopted this model as a
central pattern for the design of our process model.
Context
Recontextualization
Decontextualization
Artifacts
Figure 2: The ‘socio-technical core’ (cf. [25, 33, 44])
of alternate decontextualization and recontextualization
processes in software development activities
serves us as a starting point and as a pattern for our
model design.
3.2 Phase 1: First project course
The outline of our first project course offered in winter
2009/10 can be considered as our first prototype process
model. Our course of action featured many aspects we later
elaborated in our model. A central activity was developing
computational artifacts like models, Greenfoot simulations
and other programs and documents that would be employed
as learning equipment in the classroom context of a school
project. But we also gave attention on theoretical and transdisciplinary
foundations like learning styles, gender and diversity
issues [32].
The most important result that emerged from this first
project course was a terminology related to our course setting.
In contrast to other software development projects in
our project the context that was to be analyzed often was not

the same as the context the software was being developed
for. The artifacts were never to be recontextualized into the
original context they were derived from. Moreover conversations
with our cooperation partners from different domains
occasionally caused confusion about what ‘the project’, ‘the
requirements’ or ‘the application context’ was.
For example, in our first project we had a Requirements
team who analyzed the context of airport baggage handling
and who held interviews with domain experts. In the course
of the project it became apparent that requirements came
from multiple different contexts (see figure 3) and that the
software had to meet not only the requirements from the application
context but also the above-mentioned requirements
that were defined by the course instructors for the school
project. We therefore accustomed ourselves and our students
to distinguish, for instance, between the ‘real-world’
application context and the classroom or school context.
Evaluation of phase 1.
In their feedback our students confirmed that the topic
of the project course (Knowledge acquisition, Gender and
Diversity in Informatics and IT ) was interesting and important.
They also liked trying things out on their own.
” gutes, wichtiges Thema; gute Einbindung der
Theorie in die Praxis; Möglichkeit des Ausprobierens“
[good, important topic; good integration
of theory in practice; opportunity to try things
out]
” Interessantes Thema, viel eigenständige Arbeit“
[interesting topic, a good deal of independent
work]
On the other hand students criticized that project organization,
assignments and purpose were not clear enough.
” nicht immer gut organisiert, klare Zielsetzung
hat gefehlt“ [not always well organized, clear
purpose was missing]
Also the students’ suggestions for improvement focused on
project organization and clarity of assignments:
” klare Aufgaben / Ziele“ [clear mission / purpose]
” aktive Organisationsgruppe, um die Interaktion
zwischen den Gruppen zu verbessern“ [establish
an active organization team in order to improve
interaction between teams]
Therefore in the next iteration we concentrated on structuring
the aspects and tasks we identified during phase one.
The development of Greenfoot scenarios and other artifacts
became the focus of all subsequent project courses while
theoretical, technological, contextual and educational perspectives
were conditioning requirements and design aims.
Figure 3 represents the structure of this task, thus extending
the model on the right side of figure 1.
Not only the university-level project course but also the
school projects during phase one were very successful. The
Greenfoot Baggage Handling System (GBHS) developed by
the students was used in three school projects with pupils
at different ages (13–15; 18–20; 17–18). As Informatics-incontext
projects these hands-on classes aimed at bringing
a real-world context into the classroom, not only programming.
Therefore our students included issues from IS in the
114
Application
context,
domain
knowledge
Greenfoot
framework
Design of
Greenfoot
scenario
Teaching- learning
situation, classroom
environment
Different
views on
Computing
Figure 3: The requirements for the design of Greenfoot
scenarios come from multiple contexts.
school project [31]. This was extremely motivating. Especially
the young pupils (13–15) were very focused when
they calculated key indicators in order to optimize the layout
of their conveyor systems (see section 5). The cooperating
teacher came to the conclusion:
” Die Kinder haben spielerisch gelernt, ohne
es wirklich zu merken, wir konnten für etwas mehr
Informatikinhalte in der Schule werben und einige
Eltern haben mich sogar angesprochen, was
wir denn mit den Kindern gemacht hätten, so
überschwänglich hätten sie noch nie von der Schule
berichtet.“ [The children learned through play
without really noticing it. We succeeded in promoting
Informatics in schools: some parents even
asked what we’ve done to the kids because they’ve
never before been so rhapsodic about school!]
3.3 Phase 2: Design aims for context artifacts
Subsequent to our first iteration we developed a process
model by combining the tasks identified in phase one (figure
3) with a modified ‘socio-technical core’ where decontextualization
and recontextualization concern two different
contexts (figure 2). Furthermore in phase two we abstracted
from the concrete endeavor of designing Greenfoot scenarios.
In fact from the beginning of our first project course our
students have also created other artifacts, e. g. models and
diagrams. Thus in figure 4 the Greenfoot scenarios, along
with the other artifacts, are depicted by the centric trapezoid
shapes. The Greenfoot framework now is part of the
technological platform the artifacts are usually based on.
In order to make clear that the artifacts are related to
a consistent context and that not any and every artifact
can be included in the Informatics-in-context classroom, we
suggest the notion of context artifacts. With this term we
refer to both artifacts that have been decontextualized from
their original context where they have a situated meaning,
and newly designed artifacts objectifying situated meaning,
e. g. formal models, algorithms and software.
As a result from phase two we present the following design
aims for context artifacts (figure 4):
Technological framework. Developers need to know their
tools. Context artifacts should be designed to elegantly
build on and seamlessly fit into the technological framework.
The technological framework also includes, for instance, the

Real-world
context
Views
of
Computing
Context
Artifacts
Technological framework;
tools and media
Classroom
context
Figure 4: When context artifacts (e. g. Greenfoot
scenarios) are developed as a vehicle for bringing
context into the classroom, they need to be designed
in such a way that they integrate not only with the
technological framework and the educational context
but also with the ‘real-world’ context and the
different perspectives on Computing.
school network infrastructure and learning management system.
Educational context. Special attention must be paid to
the educational context the artifacts are designed for, i. e.
subject-matter, learning objectives, curriculum and educational
principles like learning theories, gender and diversity
aspects etc. Often context artifacts need to be specifically
adapted to the classroom context.
Real-world application context. Moreover, in an Informatics-in-context
learning environment, context artifacts serve
as vehicles bringing the real-world context into the classroom.
They are not just learning objects but represent the
‘real world’ and translate meaning from the real-world application
context into the classroom context. Therefore they
must be designed to bear a meaning when contextualized in
or related with the real-world context.
Informatical perspective. Finally Informatics-in-context
classes are Informatics classes. Hence context artifacts must
be selected or designed on the basis of their relevance from
an informatical point of view. On the one hand curricula define
what the body of knowledge of Informatics is and what
is relevant for educational purposes. On the other hand, as
we see it, Informatics in context aims at presenting a broad
image of Informatics. Therefore the design of context artifacts
should integrate multiple views on Computing and
Informatics.
Evaluation of phase 2.
The phase-two students’ feedback proved that clarity and
project organization had improved. At the same time we
succeeded in maintaining the strengths of the project:
” Guter Themenbereich, eher ungewöhnlich für
FB Informatik“
[Good topic, quite unusual for the department of
Informatics]
” Die Veranstalter waren immer gut vorbereitet.
Die Aufgabenstellungen waren größtenteils klar
115
definiert. Durch die Projektwoche war die Arbeit
nicht so einseitig.“
[The lecturers were always well prepared. The assignments
were mostly clear. The school project
offered a diversion from the other activities.]
Moreover many of the students appreciated the guidance by
the lecturers. But there were still also many critical voices:
” teilweise unübersichtliche Organisation, schwammig
gestellte Aufgaben, was zu Unklarheiten führte“
[sometimes unclear organization, vaguely formulated
assignments resulting in confusion]
Other objections were related to the composition of teams.
Some students complained about “unequal” teams and demanded
“fair group forming” for future projects. Therefore
in phase three we gave attention to roles.
3.4 Phase 3: Pre-educational stage
and educational stage
At the beginning of phase three we started to ask which
of the proposed tasks teachers can effectively assume when
preparing lessons. While in the first cycle of our project
course there were many students studying IS (14 in 21),
the students in the second course had a very diverse background,
majoring in CS, SE, IS or CSEd, so that they focused
on different aspects of the project. Students from CS,
SE and IS concentrated mostly on different aspects of decontextualization
(analysis of context and construction of
context artifacts), whereas CSEd students were interested
in the recontextualization (preparation and implementation
of teaching units).
Thus the idea to discern a pre-educational and an educational
stage seemed appropriate. The pre-educational stage
is independent from a concrete educational context though
it nevertheless assumes an educational purpose. Because of
its independence the first stage can be treated by external
experts, or educators can be supported by them in doing so,
e. g. in cooperation with universities. The educational stage
needs to be conducted by educators.
The distinction of pre-educational and educational stage
brought about another distinction. As a result from this
iteration we also distinguish between context artifacts and
teaching units. Teaching units are a class of artifacts augmented
with context artifacts. In contrast to context artifacts,
teaching units are not decontextualized from the realworld
application context but they provide plans for classroom
activities and teaching-learning situations. They can
describe a single lesson, a whole course or any meaningful
chapter of it.
Evaluation of phase 3.
In the first iterations of the project we focused on the
design of the university-level project course. Now we broadened
our scope to include the adoption of Informatics-incontext
in schools.
The context artifacts the students developed were not only
used in the school projects but also in the Informatics Profilkurs
(advanced placement course) of one of the cooperating
schools. Profiles – implemented only recently and only in
part at German secondary schools – are meant to promote
competition among schools and to encourage the formation

of partnerships and cooperation models in order to diversify
educational opportunities. Since educational standards
must be met by all schools profiles are closely related to
context-oriented approaches.
Therefore for the evaluation of phase three we paid more
attention to the two profile teachers’ feedback. They reported
that the profile is very attractive for pupils from both
cooperating schools and even from other schools. Since 2010
when the profile was implemented, the number of pupils who
change schools because of the offered profile, has increased.
Both teachers explain the profile’s success by their cooperation
with external partners and the school projects:
” Ohne Euch hätten wir das nie geschafft.“
[Without you we would never have accomplished
this.]
” Unser Informatikprofil ist erfolgreich, es kommt
wieder eins zustande, auch Dank der tollen Angebote,
die wir durch Euch machen können.“ [Our
Profilkurs is successful: we can offer it again this
year thanks to the great opportunities we can
give due to your support.]
” Für uns ist es wichtig, dass die Projekte klein
genug sind für die Schule, aber dennoch mit praxisbezogenem
Hintergrund. [Das Projekt] hat die
Schüler unheimlich motiviert und es war in unser
Semester eingebettet.“ [For us it is important
that the school projects are small enough to be
practical at school, but at the same time reflect
a professional background. The pupils were very
motivated, and the project was well integrated in
the Profilkurs this school semester.]
However, the success of school courses doesn’t end with
the motivation of the pupils but with learning outcomes
which have to meet educational standards.
3.5 Phase 4: From input to outcome
For the first three project courses our process of bringing
contexts into the classroom essentially resembled a waterfall
model. It was only after the decontextualization was
completed that the recontextualization started: the school
project took place not before end of term. Our students
who developed context artifacts and teaching units had no
second opportunity to improve their products.
However, during the school projects we observed how the
pupils got along with the Greenfoot scenarios and gained
first-hand insights into how the context artifacts and the
teaching units fitted into the classroom context. The teachers
from the cooperation school who were temporarily replaced
by our students also gave valuable feedback. Moreover
the boys and girls developed their own ideas on how
to improve the software. Some of them even engaged in actively
implementing new features and thus created context
artifacts themselves. These results informed the requirements
for our next project course.
An additional aspect to be considered in the model is the
role of educational standards. In our school project educational
standards have been playing a minor role. But by
now a cooperation school had integrated some of our course
material into their curriculum. Since current educational
standards regulate content only to a minor degree and instead
focus on learning outcome, it was perfectly possible for
them to adapt our material to their courses. However in this
116
case courses need to be aligned to educational standards.
Evaluation of phase 4.
The evaluation of phases four and five is object of further
research because it requires a comparison of learning outcomes
of both traditional Informatics courses and Informaticsin-context
courses, which is beyond our scope. But our
school projects are already producing results. There are
pupils who attended the first school projects in 2010 at the
age of 14 and who now chose the Profilkurs offered by our
cooperating school. And by now there are also pupils who
successfully finished the Profilkurs and enrolled for Informatics
or engineering degree programs.
3.6 Phase 5: Consolidation
Within the current fifth phase we have consolidated the
graphical representation of the suggested model in order to
communicate it to other researchers and practitioners. Furthermore
we want to discuss both the relevance of the model
and the choice of context. The current version is described
in the following section and illustrated in figure 5.
4. THE PROCESS MODEL
In the following we will describe our process model for
bringing contexts into the classroom along the lines of the
distinction of the pre-educational and the educational stages
(see 3.4). The model is shown in figure 5.
4.1 Pre-Educational Stage
At the pre-educational stage the real-world context is analyzed
with an essentially informatical stance. This stage is
characterized by the decontextualization of context artifacts
from their original context.
4.1.1 Decontextualization of Context Artifacts
The simplest instance of a context artifact is when a physical
artifact – e. g. a punch card – has lost its original meaning,
yet is still associated with a context. We call these
kind of context artifacts aboriginal or primary. Most artifacts
cannot simply be removed from their original context
though, and therefore must be mediated. Examples of
mediated or secondary context artifacts are photographs or
videos. However in order to understand the situated meaning
of activities and artifacts in a complex system, they need
to be fully decontextualized and reassembled. Instances for
these tertiary newly designed context artifacts are models or
simulations, and other objectifications.
As described above (3.1.2) decontextualization, i. e. formalization
and constructing artifacts, is usually one important
aspect of software development. The other is recontextualization,
i. e. implementation of the artifacts into the
context. In our model two contexts must be analyzed: the
real-world context and the classroom context. Thus the preeducational
stage is pre-educational in that educational purposes
are assumed in the decontextualization of context artifacts.
Yet for this activity an informatical stance might be
predominant.
4.1.2 Guiding Informatical Principles (GP)
The decontextualization is guided by fundamental informatical
principles because the artifacts are selected and developed
from an informatical point of view. Guiding principles
(GP) may be found in the Great Principles of Comput-

Real-world
context
GP
Context
Artifacts
EP
Classroom
context
Technological framework;
tools and media
Teaching
units
Learning
outcomes
Educational
standards
Figure 5: The process model for bringing real-world contexts into the classroom focuses on the design of
context artifacts and teaching units. In the process two stages can be identified: at the pre-educational
stage context artifacts are decontextualized from their original context; at the educational stage they are
recontextualized into the classroom context.
ing [5] or Fundamental Ideas in Informatics [35] frameworks.
The notion of Computational Thinking [45] might be helpful
as well. At this stage the perspective should not be too
narrow though, because the artifacts should be designed to
translate many aspects of the real-world context into the
classroom context and open up a broad image of Informatics.
In figure 5 the guiding informatical principles are represented
by triangles side by side with the educational principles.
4.1.3 Technological Framework
Often information and communication technology (ICT)
are needed in order to bring real-world contexts into the
classroom. The technological framework provides the infrastructure,
tools, media etc. needed for developing artifacts
and recontextualizing them into the classroom context.
Thus both the artifacts and the classroom context build on
the technological platform. This belongs to both the preeducational
stage and the educational stage.
4.2 Educational Stage
During the educational stage an educational stance is predominant.
The guiding principles for the recontextualization
of context artifacts into the classroom context therefore are
first and foremost educational.
4.2.1 Recontextualization of Context Artifacts
Artifacts are meaningless unless they are contextualized.
Thus recontextualizing formalized action or computational
artifacts is as important for software development as decontextualization.
In our process, context artifacts are to be
implemented into the classroom context. According to figure
1 this requires designing exercises. That is by far not all.
Learning objectives have to be defined; the artifacts must be
adapted to the concrete classroom context and possibly augmented
with additional material; the learning environment
must be arranged around the artifacts; methods of instruc-
117
tion are to be selected and prepared etc.
Recontextualized into the classroom context, context artifacts
become part of the learning environment. Moreover
they foster the students’ imagination and thus anchor the
real-world context in the classroom context.
4.2.2 Guiding Educational Principles (EP)
The development of context artifacts is guided by informatical
principles as well as educational principles (EP). In
the present version of our model we distinguish between educational
principles and educational standards. Educational
standards regulate which learning outcomes students must
meet. Principles guide how this is to be achieved. These
include learning theories, teaching methodology, gender and
diversity aspects, etc.
4.2.3 Outcome-oriented Standards and
Learning Outcomes
Current educational standards are outcome-oriented, i. e.
they describe what the students must learn in terms of outcome
like competencies. Since the outcome artifacts produced
in the classroom should match the descriptions in the
standards, the teaching units must be adjusted in such a
way that they address the required competencies and skills.
5. EXAMPLE: PREPARATION OF THE
SCHOOL PROJECT
In this section we will illustrate the application of the
model described above. The context used in several of our
project courses was baggage handling at airports. We started
with an important airport component, the baggage handling
system (BHS). Using the example of baggage handling, the
different informatical principles, organizational as well as
socio-technical concepts could be covered in teaching units.
The initial difficulties with the BHS at Heathrow Airport’s
Terminal 5 in 2008 gave a vivid example and were great
motivation for both our students and the youngsters.

Figure 6: Two pupils presenting their conception of
airport baggage handling after they were introduced
to the context by a YouTube video.
Starting with the first stage of the proposed model, the
pre-educational stage, the students began to analyze the
context of baggage handling. First, they searched the Internet
to find diverse artifacts describing the context. The students
found pictures, videos and technical documents some
of which could also be used as (secondary) context artifacts.
Furthermore, they had the chance to visit an international
airport and to inspect the baggage handling system on site.
They took notes and additional pictures and had interviews
with domain experts.
Moreover, since we used the Greenfoot as a framework
for the development of a “micro baggage handling world,”
the students had to learn how to develop Greenfoot scenarios
with professional development tools 3 . For the school
project the students implemented a Greenfoot simulation of
an airport baggage handling system. They modelled animated
passengers who check in their luggage and go to their
flights while their luggage is transported to the correct airplane.
The context artifacts Greenfoot scenario, the interviews,
photos, videos and further documents were the result
of the first stage.
At the educational stage the context artifacts and educational
standards were used to develop teaching units. The
developed teaching units consisted of a timetable, descriptions
of basic definitions and concepts and different exercises.
For example, the pupils were introduced to the context
of airport baggage handling by videos and prosa text.
After that they were to draw a schematic of the BHS (figure
6).
In another teaching unit the boys and girls started to investigate
a Greenfoot simulation of a BHS, that intentionally
exhibited some serious problems. After a few minutes the
baggage cumulated on the conveyor (figure 7). The first exercise
was to figure out the problem and to rearrange the
layout of the conveyor system. This was done by drag-anddrop
and interactively exploring the objects’ interfaces.
3 As a didactical programming environment Greenfoot is not
well suited for professional teamwork or code refactoring.
Therefore we use both Greenfoot and Eclipse for the development
of the Greenfoot scenarios.
118
Figure 7: Greenfoot simulation of an airport baggage
handling system, used in the school projects of
2010 [31]. In this scenario the BHS layout is flawed
so that after a few minutes the conveyor system is
jammed with luggage.
In this section we have illustrated how we have used the
model in order to bring the context of baggage handling into
the classroom. We have described the decontextualization of
context artifacts based on informatical principles and supported
by the Java programming environment Greenfoot.
In the second stage of the proposed model we have developed
teaching units using the context artifacts. The designed
teaching units have been brought into the classroom
during different school projects. The feedback, we received
from the pupils, were mainly positive because they enjoyed
to get insight into complex real world context like baggage
handling.
6. DISCUSSION
The proposed process model has been developed in a specific
course setting with multiple subprojects. The different
roles and the division of work established in our project
are obviously reflected in our model. Thus our model may
not hold universally. Nevertheless we argue that the division
of work in our project results from the fact that the
different contexts involved are essentially separated in the
first place. After the discussion of the our methodology, we
therefore consider other possible constellations and a fundamental
dilemma for Informatics teachers – especially at
school.
6.1 Model Validity
Although the process model for bringing contexts into the
classroom is so abstract that it can only serve as a reference
framework, there is good reason to challenge it. In our
projects it has been proven to be viable, but further inquiries
may remain necessary. Nevertheless we claim ‘validity by design’
for our model arguing along the following principles for
design-based research in IS, that were introduced by Hevner
et al. [17]:
Design as an Artifact. The goal of design-based approaches
is the creation of appropriate and reliable artifacts,

i. e. theories, models, frameworks, technology or tools. In
our research project it is the process model for bringing contexts
into the classroom.
Problem relevance. A design-based research project must
deliver a solution to a relevant and significant problem. In
the complexity and amount of task during analysis of realworld
contexts, the translation into the classroom context,
as well as in the additional workload for teachers we see the
major problems of CS in context. At the moment there is
little research addressing these problems. Our model doesn’t
provide a solution to them, but it can help identifying them
(see 6.2 and 6.3).
Design evaluation. The relevance, quality and viability
of designed artifacts have to be shown via elaborate evaluation
methods. We tested our process model during each
iteration in at least two classroom contexts: at least one
school project and the project course itself which can also
be described by the model. Here the context of the school
project was the ‘real-world’ context. Furthermore we found
our model to be coherent with other proposed models (e. g.
[6]) as well as practical (e. g. anchored instruction [19]) and
theoretical (e. g. boundary objects [37]) conceptions. The
evaluations of each iterations produced new aspects, which
were addressed in the next iteration.
Research Contribution. A concrete contribution to educational
research has to be provided by the design-science
research. Our contribution is a process model which serves
as a reference framework with its own terminology based on
approaches to software development.
Research rigor. A very important aspect of design science
is “the application of rigorous methods in both the construction
and evaluation of the designed artifact” (Hevner et
al. [17], p. 87). In our design we used a methodology similar
to software development methods. We also used prospective
and reflective approaches supporting our research process
[29]. In the current phase we want to communicate our
model to other researchers and practitioners and, given the
model, to survey their own experience.
Design as a Search Process. Design science must be conducted
via an iterative process with construction and evaluation
periods in order to mature [36]. In our research
project, four consecutive research phases (both construction
and evaluation) have been accomplished. Currently a fifth
iteration is ongoing.
Communication of results. The results will be communicated
to practitioners and researchers enabling to benefit
from the constructed artifacts as well as to discuss and evaluate
the results. On the one hand, we have introduced our
proposed process model to students and teachers, on the
other hand, we will bring the model to the research community
via the present publication.
6.2 Contexts
As described above, the ‘real-world’ context normally is
different from classroom context. Therefore new challenges
for teachers and Computing education experts occur in the
decontextualization and recontextualization processes of Informatics-in-context
course development and preparation.
But how can an appropriate real-world context be chosen?
First, the real-world context should be in the same sociocultural
frame as the classroom context. Otherwise too
much time and effort must be spent on understanding the
context instead of learning Informatics, and the context might
119
turn out to be unattractive for the pupils. For example
in low-income neighborhoods baggage handling systems or
inventory control systems might be unattractive contexts.
Also in rural areas information systems in animal husbandry
might be more motivating than the baggage handling system.
Second, contexts can arise from Informatics itself. For
example Informatics contexts like artificial intelligence or
compiler construction provide rich contexts offering the possibility
to include many Informatics aspects and principles
(provided that the first principle is taken into consideration
and the students understand the practical purposes).
Finally, often technological frameworks like specific hardware
or tools constitute contexts that are attractive at least
for technophiles. For example robotics classes often don’t
get beyond playful experimentation toward serious application
contexts.
6.3 CS Teachers’ Working Conditions
As outlined by Diethelm et al. [7], the teachers’ role and
demands on CS teachers have changed in the last years.
In addition to teaching, new contexts have to be analyzed,
teaching units must be designed and transformed for the
actual classroom setting as well as the outcome has to be
evaluated. All tasks are highly communicative and cooperative
and require a set of various skills and competencies CS
teachers do not always have [7].
The pre-educational examination of contexts is a timeconsuming
endeavor that requires a broad background in
Informatics and the real-world context. Most teachers don’t
have time and background to develop contexts for Informaticsin-context
courses.
These kind of problems are not new and have already
been discussed in German Didaktik discourse in the 1950s
by Roth [34] and Klafki [21]:
“We believe that it would be demanding too
much of teachers in terms of time and mental
energy to expect them to ‘rationalize’ about the
contents in a pre-pedagogical context [or stance]
whenever they set out to prepare themselves for
teaching. This would involve, for example, adopting
the role of a scientist who sees the contents in
question as a research exercise in a specific field.
We are of the opinion that this applies not only to
teachers at primary, junior secondary and vocational
level, but also to those at senior secondary
level!” (Klafki, [22], p. 17)
Although the problems remain the same today’s schools
look different from the Volksschulen of the 1950s. Teachers
work in teams and can collaborate even if they are distributed
all over the globe.
Our supposed model tries to support teachers in their
work. It names necessary steps and requirements for bringing
real world contexts into the classroom as well as introduces
a pre-educational stages allowing to “outsource” certain
steps (esp. analysis of contexts) to other experts. Thus
we hope to strengthen the link between theory and practice
in teaching Informatics in context.
7. CONCLUSION
Learning always is situated in contexts. But usually the
classroom context is not the ‘real-world’ context where ac-

tivities, artifacts, and concepts have a situated meaning implying
real consequences. In the classroom things bear a
special meaning according to the ‘hidden curriculum’. Hence
teaching is a performative act in that it constitutes and reinforces
its own prerequisites. Moreover teaching in context
requires the classroom context to resemble aspects of a realworld
context but is still relying on both the teacher’s and
the students’ imagination and play-acting. Therefore the
development, planning and preparation of context-oriented
teaching units is an intricate and time-comsuming endeavor.
The ceiling of the Great Hall at Hogwarts was supposedly
intentionally designed to bring the outside world into the
school. Unlike the teachers at Hogwarts, Informatics teachers
at our schools cannot make use of magic in order to bring
the outside world into the classroom. They can, however,
use computing technology and develop context artifacts so
as to arrange the learning environment around them.
Though our process model for bringing contexts into the
classroom emerged from a special project constellation, it
reveals that contextualized teaching requires not only more
careful preparation but also some preparation steps that
possibly are beyond the means of most teachers. Educators
and most notably primary and secondary school teachers
who want to contextualize their teaching would therefore be
well advised to draw on the strengths of their institution
and to take advantage of cooperations with other educational
and research facilities as well as local enterprises.
8. REFERENCES
[1] T. Anderson and J. Shattuck. Design-based research:
A decade of progress in education research?
Educational Researcher, 41(1):16–25, 2012.
[2] K. Beck et al. Manifesto for agile software
development. URL http://agilemanifesto.org/
(08/2012), 2001.
[3] A. Collins, D. Joseph, and K. Bielaczyc. Design
research: Theoretical and methodological issues. The
Journal of the Learning Sciences, 13(1):15–42, 2004.
[4] S. Cooper and S. Cunningham. Teaching computer
science in context. ACM Inroads, 1(1):5–8, 2010.
[5] P. J. Denning and C. Martell. Great principles of
computing; official web site, 2005 ff. URL
http://cs.gmu.edu/cne/pjd/GP/ (April 2012).
[6] I. Diethelm and C. Dörge. From context to
competencies. In N. Reynolds and
M. Turcsányi-Szabó, editors, Key Competencies in the
Knowledge Society, IFIP TC 3 International
Conference, KCKS 2010, pages 67–77, Berlin, 2010.
Springer.
[7] I. Diethelm, C. Hildebrandt, and L. Krekeler.
Implementation of Computer Science in Context – a
research perspective regarding teacher-training. In
Koli Calling 2009 : 9th International Conference on
Computing Education Research, pages 93–99, 2010.
[8] I. Diethelm, J. Koubek, and H. Witten. IniK –
Informatik im Kontext : Entwicklungen, Merkmale
und Perspektiven. LOG IN, (169/170):97–105, 2011.
[9] E. W. Dijkstra. On the cruelty of really teaching
computing science [a debate on teaching computer
science]. Communications of the ACM,
32(12):1397–1414, 1989.
[10] D. Engbring and A. Pasternak. Einige anmerkungen
120
zum begriff inik. In I. Diethelm, C. Dörge,
C. Hildebrandt, and C. Schulte, editors, Didaktik der
Informatik : Möglichkeiten empirischer
Forschungsmethoden und Perspektiven der
Fachdidaktik [6. Workshop der GI-Fachgruppe DDI],
number P-168 in Lecture Notes in Informatics (LNI) –
Proceedings, pages 119–124, Bonn, 2010. Gesellschaft
für Informatik.
[11] S. Fincher, M. Kölling, I. Utting, N. Brown, and
P. Stevens. Repositories of teaching material and
communities of use: Nifty assignments and the
Greenroom. In ICER ’10: Proceedings of the 6th
International Computing Education Research
Workshop, pages 107–114, 2010.
[12] C. Floyd. Steps – a methodical approach to pd
[paricipatory design]. Communications of the ACM,
36(4):83, June 1993.
[13] C. Floyd, W.-M. Mehl, F.-M. Reisin, G. Schmidt, and
G. Wolf. Out of scandinavia: Alternative approaches
to software design and system development.
Human-Computer Interaction, 4(4):253–350, 1989.
[14] C. Floyd, F.-M. Reisin, and G. Schmidt. Steps to
software development with users. In Proceedings of the
2nd European Software Engineering Conference, ESEC
’89, pages 48–64, London, UK, 1989. Springer-Verlag.
[15] M. Guzdial. Does contextualized computing education
help? ACM Inroads, 1(4):4–6, 2010.
[16] P. Henriksen and M. Kölling. Greenfoot: Combining
object visualisation with interaction. In OOPSLA ’04:
Proceedings of the 19th Annual ACM Conference on
Object-Oriented Programming, Systems, Languages,
and Applications, pages 73–82, 2004.
[17] A. R. Hevner, S. T. March, J. Park, and S. Ram.
Design science in information systems research. MIS
Quarterly, 28(1):75–105, 2004.
[18] W. Jank and H. Meyer. Didaktische Modelle.
Cornelsen, Berlin, 10th edition, 2002.
[19] M. Kariuki and M. Duran. Using anchored instruction
to teach preservice teachers to integrate technology in
the curriculum. Journal of Technology and Teacher
Education, 12(3):431–445, 2004.
[20] A. E. Kelly. Design research in education: Yes, but is
it methodological? The Journal of the Learning
Sciences, 13(1):115–128, 2004.
[21] W. Klafki. Didaktische Analyse als Kern der
Unterrichtsvorbereitung. Die Deutsche Schule,
50(10):450–471, 1958.
[22] W. Klafki. Didactic analysis as the core of preparation
of instruction. Journal of Curriculum Studies,
27(1):13–30, 1995. Translated and abridged version of
[21].
[23] M. Knobelsdorf and C. Schulte. Computer science in
context – pathways to computer science. In Koli
Calling 2007 – Proceedings of the Seventh Baltic Sea
Conference on Computing Education Research, pages
65–76, 2008.
[24] J. Koubek, C. Schulte, P. Schulze, and H. Witten.
Informatik im kontext (inik) : Ein integratives
unterrichtskonzept für den informatikunterricht. In
B. Koerber, editor, Zukunft braucht Herkunft : 25
Jahre ” INFOS – Informatik und Schule“ [INFOS
2009, 13. GI-Fachtagung Informatik und Schule],

number P-156 in Lecture Notes in Informatics (LNI) –
Proceedings, pages 268–279, Bonn, 2009. Gesellschaft
für Informatik.
[25] D. Krause, A. Rolf, M. Christ, and E. Simon. Wissen,
wie alles zusammenhängt. Informatik-Spektrum,
29(4):263–273, 2006.
[26] M. Kölling. The greenfoot programming environment.
ACM Transactions on Computing Education,
10(4):14:1–14:21, 2010.
[27] M. Kölling. Introduction to Programming with
Greenfoot : Object-Oriented Programming in Java with
Games and Simulations. Pearson, Boston, MA, 2010.
[28] R. Künzli. German didaktik: Models of
re-presentation, of intercourse, and of experience. In
I. Westbury et al., editors, Teaching as a Reflective
Practice, pages 41–54. Lawrence Erlbaum, Mahwah,
NJ, 1999.
[29] Y. Mor and N. Winters. Design approaches in
technology-enhanced learning. Interactive Learning
Environments, 15(1):61–75, 2007.
[30] K. Nygaard. A process-oriented conceptual platform
for informatics. URL
http://hci.stanford.edu/courses/cs547/abstracts/01-
02/011019-nygaard.html (July 2012),
2001.
[31] D. Rick, J. Ludwig, S. Meyer, C. Rehder, and
I. Schirmer. Introduction to business informatics with
greenfoot using the example of airport baggage
handling. In Koli Calling ’10: Proceedings of the 10th
Koli Calling International Conference on Computing
Education Research, pages 68–69, 2010.
[32] D. Rick and I. Schirmer. ” IT-Aneignung, Gender und
Diversity“: Ein transdisziplinäres Projektseminar für
Informatik-Studierende. In M. Thomas and
M. Weigend, editors, Informatik und Kultur : 4.
Münsteraner Workshop zur Schulinformatik, pages
64–77, Münster, Germany, 2010. Zentrum für
Lehrerbildung. URL http://miami.unimuenster.de/servlets/DocumentServlet?id=5304
(June
2010).
[33] A. Rolf. Mikropolis 2010 : Menschen, Computer,
Internet in der globalen Gesellschaft.
Metropolis-Verlag, Marburg, 2008.
[34] H. Roth. Pädagogische Psychologie des Lehrens und
Lernens. Schroedel, Hannover, 7th edition, 1963
[1957].
121
[35] A. Schwill. Computer science education based on
fundamental ideas. In Proceedings of the IFIP TC3
WG3.1/3.5 joint working conference on Information
technology : supporting change through teacher
education: supporting change through teacher
education, pages 285–291. Chapman & Hall, Ltd.,
1997.
[36] H. A. Simon. The Sciences of the Artificial.
MIT Press, Cambridge, MA, first edition, 1969.
[37] S. L. Star and J. R. Griesemer. Institutional ecology,
‘translations’ and boundary objects: Amateurs and
professionals in berkeley’s museum of vertebrate
zoology, 1907-39. Social Studies of Science,
19(3):387–420, 1989.
[38] L. A. Suchman. Plans and situated actions : the
problem of human-machine communication.
Cambridge University Press, 1987.
[39] A. E. Tew, B. Dorn, W. D. Leahy, Jr., and
M. Guzdial. Context as support for learning computer
organization. ACM Journal on Educational Resources
in Computing, 8(3):8:1–8:18, 2008.
[40] The Design-Based Research Collective. Design-based
research: An emerging paradigm for educational
inquiry. Educational Researcher, 32(1):5–8, 2003.
[41] The Greenfoot Team. Greenfoot. Free download,
2006ff. URL http://www.greenfoot.org/ (April 2012).
[42] The IniK Group. Informatik im Kontext – IniK für
alle. IniK Website, 2007ff. URL
http://www.informatik-im-kontext.de/ (July 2012).
[43] I. Utting, S. Cooper, M. Kölling, J. Maloney, and
M. Resnick. Alice, greenfoot, and scratch – a
discussion. Transactions on Computing Education,
10(4):17:1–17:11, November 2010.
DOI: 10.1145/1868358.1868364.
[44] J.-H. Wahoff, J. P. de Albuquerque, and A. Rolf. The
mikropolis model: A framework for transdisciplinary
research of information systems in society. In Y. K.
Dwivedi, M. R. Wade, and S. L. Schneberger, editors,
Information Systems Theory, volume 29 of Integrated
Series in Information Systems, pages 367–386.
Springer New York, 2012.
[45] J. M. Wing. Computational thinking.
Communications of the ACM, 49(3):33–35, 2006.

Email for You (only?) – Design and Implementation
of a Context-based Learning Process on
Internetworking and Cryptography
Andreas Gramm
1. Schulpraktisches Seminar
Charlottenburg-Wilmersdorf
Otto-Suhr-Allee 100
10585 Berlin, Germany
+49 (0)30 90 29 13 280
gramm@gymnasium-tiergarten.de
ABSTRACT
The didactical approach of teaching computer science in context
aims at enabling learners to understand concepts of computer
science better through the help of concrete illustration and
meaning. This paper describes a learning arrangement in which
students in lower secondary school education are motivated to
engage with cryptographic algorithms ranging from Caesar to
RSA by making them discover the challenges of a private and
trustable communication over public networks. We also describe
experiences we made by developing and testing the context-based
learning process in several classes of different age. While
designing and implementing context-based teaching material
proved to be demanding, we were rewarded with a high level of
both interest and understanding on the part of the students
suggesting that context-based learning will prove a promising
didactical tool for computer science teachers.
Categories and Subject Descriptors
K.3.2 [Computer and Information Science Education]: Computer
science education
General Terms
Algorithms, Reliability, Security, Human Factors.
Keywords
Context-based computer science education, computer science in
context, IniK, communication protocol, email protocol,
communication security, cryptology, RSA
1. COMPUTER SCIENCE IN CONTEXT
In the light of new insights into learning through neurobiological
research and with regard to a constructivist understanding of
learning, it has recently been argued that while educators wish to
implement largely transferable, abstract knowledge, concrete
contexts are necessary for learners to understand the nature,
meaning and use of an abstract idea or concept. Cooper and
Cunningham showed how this approach has been applied to the
design of introductory courses at university level using 3D
animations (Alice), media computation, computer graphics or
robotics as contexts [6]. According to their analysis, suitable
Malte Hornung
Freie Universität Berlin
Didactics of Informatics
Königin-Luise-Straße 24-26
14195 Berlin, Germany
+49 (0)30 83 87 51 87
malte.hornung@fu-berlin.de
122
Helmut Witten
Gesellschaft für Informatik (GI)
Working group “CS education in
Berlin and Brandenburg (IBBB)”
Brandenburgische Straße 23
10797 Berlin, Germany
+49 (0)30 88 68 25 69
helmut@witten-berlin.de
contexts are a source of illustrative examples, project ideas,
motivation and meaning. Guzdial points out that the use of
contexts improves retention of students: Students are more likely
to continue CS studies when they understand the usefulness and
value of what they are learning [17]. But he also warns that
learners might overgeneralise a concept when only seen in a
single context. He therefore calls for concepts to be presented in
various different contexts to allow for an adequate
decontextualisation.
While the international debate of learning CS in context primarily
discusses introductory courses to computer science at universities,
the discussion in Germany focuses on secondary school
education. An open project group called “Informatik im Kontext –
IniK” (translates to “computer science in context”) was established
to develop and promote the approach of context-based
computer science secondary school education. The project’s
website [21] describes the concept of IniK and provides teaching
material and didactic instructions for teachers for a number of
contexts and possible teaching units.
IniK is partly modelled on projects promoting context-based
learning of sciences (chemistry, physics and biology), which have
received substantial funding from both the Federal Government of
Germany as well as the different federal states (known as
Bundesländer). Out of these three projects “Chemie im Kontext -
ChiK” [19] was the first and has been the project most influential
for IniK (cf. [10]). In contrast to the three science projects, IniK is
more of the character of a grass root development with almost no
funding. To our knowledge, Berlin is the only federal state to
grant a small amount of resources to a group of teachers
committed to the IniK approach. Apart from teachers interested in
modern teaching and learning approaches it is mainly professors
and students from several universities who further develop the
ideas and concepts of IniK.
While the idea of a context-based approach to computer science
has been discussed in Germany before (see e.g. [12] and [7]), the
IniK approach was first fully described in 2009 by Koubek,
Schulte, Schulze and Witten in [24]. They identify three
prerequisites for a learning process to qualify as being contextbased
in the field of secondary school computer science
education:

1. Context-based education in computer science makes use
of a context that is relevant to students. The context is a
concrete situation, in which aspects of different
dimensions are significant to understand the situation
and to find adequate solutions.
2. The competencies acquired through the learning process
are compliant with standards for teaching computer
science in secondary school education. The Society for
Informatics in Germany Gesellschaft für Informatik (GI)
has published principles and minimal standards for
computer science in lower secondary school education
[1] (for an English summary of these standards see [4]
and [25]). Referring to the definition of Weinert, the
competencies described in the standards are to be
understood as “the cognitive abilities and skills
possessed by or able to be learned by individuals that
enable them to solve particular problems, as well as the
motivational, volitional and social readiness and capacity
to use the solutions successfully and responsibly in
variable situations” ([20], p. 65).
3. Context-based education in computer science shows a
variety of teaching and learning methods to activate
different types of learners and enhance cooperative
student-student interaction in class.
Since the founding of IniK, several further ideas on how to design
and implement context-based learning of computer science have
been developed. Diethelm, Borowski and Weber suggested a way
to find contexts which are relevant to learners by indirectly asking
them what interests they have in a certain informatics devices [8].
Furthermore, Diethelm and Dörge described a way to develop
competencies using a context [9]. In 2011, Diethelm, Koubek and
Witten summarised the development, features and perspectives of
IniK and state criteria for the selection of suitable contexts and as
to how units could best be documented [11].
The goal of the IniK-working group was to design several
context-based learning environments with a high permeability to
everyday school practice and reflect on its implementation in
secondary schools both to prove the feasibility of the concept to
other CS teachers as well as to gain experience and further
develop the concept. The method used for design and reflection
has an alignment to research methods like Action Research [2]
and Design Based Research [3]. Note that therefore this report’s
conclusions are mostly based on in-service reflection, not on a
systematic empirical approach. “Email for You (only?)” is the title
of a context-based learning environment on internetworking and
cryptography which is presented and discussed in this report.
2. DESIGN AND IMPLEMENTATION OF
THE LERANING PROCESS
We have developed a context-based learning process on security
issues of internetwork communication (for a more detailed
documentation of the learning process in German see [22], [16],
[14] and [13]). In doing so, we intended to show the feasibility of
the IniK concept for a context-based learning of computer science
in everyday computer science classrooms as well as to gather
experience that can be helpful for the further development of the
concept and other context-based learning processes.
The indented age group is from year 9 to year 12 (puplis aged 14
to 18). From prior experience we can say that students often enjoy
123
the topic cryptography, because transmitting secrets is associated
with adventure and decrypting or even cracking a cipher
resembles a strategic game. However, just discussing cryptographic
algorithms such as Caesar doesn’t help raising awareness
for the fact, that we as users make use, or at least are well-advised
to make use of cryptography whenever communicating over
public networks such as the Internet. Most students communicate
over the Internet on a daily basis, be it via instant messengers,
social networks or email. The recent annual survey among
German teenagers “Jugend, Information, (Multi-) Media (JIM)
2011” [18] shows that they spend almost half of their online time
communicating. The context of security issues in private communication
over public networks is therefore part of students’ daily
life, only that in most cases they are not aware of it. We believe
that rising awareness of both advantages and challenges of using
computer systems is a central aim of secondary school computer
science education.
The context can be explored with regard to different dimensions.
On the one hand, assessing security risks requires a thorough
understanding of the underlying network technology and the
communication protocols employed. On the other hand, security
always requires additional resources. In the case of private
communication, it requires the exchange of public keys. Today,
many users don’t use encryption even though the technology is
available for free. This clearly indicates to human factors such as
too little awareness of security issues and possible consequences.
We structured the learning process by five questions:
1. How is an email communicated from my computer to
the computer of the addressee?
2. What dangers challenge a private communication over a
public network, such as the Internet?
3. How can I achieve privacy?
4. How can I assure the integrity of a message and the
authenticity of a sender?
5. Why should I communicate privately?
The answers to questions 2 to 4 respectively build on the
knowledge students acquire to answer the questions asked before.
This way, they shall experience this knowledge as being useful to
solve tasks such as providing security in an unsafe environment.
The last question could alternatively be asked at the beginning of
the learning process, but in our view it is more rewarding to
discuss it when students have got clear ideas of the challenges of
communication over public networks, which they might not have
at the beginning of the learning process. It also seems more
attractive to us that students discover the possible dangers in a
hands-on activity rather than in an academic discussion.
Of course, showing students ways to disturb email communication
raises ethical questions. In fact, the Bundesverfassungsgericht
(Germany’s High Court) has discussed the question whether using
a network analysing tool is in breach of German law. The court
decided that the law in question only applies when such tools are
used with malevolent intentions. We prefer that our students know
about existing dangers and take protective measures rather than
leaving this knowledge to possibly more malevolent experts who
might use their lack of knowledge to infect their computer with
malware. However, we took the discussion as an occasion to look
for a tool which restricts the accessed network traffic to those

packets that are directly addressed to the client computer, which
suits our purposes sufficiently.
Table 1 shows the underlying model of competencies on three
different levels. It is up to teachers, the age of the learners and the
time available to decide on how deeply the insights into security
provided by cryptography should be developed. It is possible to
just do parts of the activities in one year and continue with a more
profound examination of e.g. the RSA key generation algorithm in
a following year. By pointing out these levels of competency, we
enable teachers to adjust the material to their needs more flexibly.
Table 1. Competency model for the understanding of the level
of security provided for communication by cryptography.
Level 1 Students describe dangers of communicating via
public networks (eavesdropping, manipulation of
messages, false sender’s identity) and state
prerequisites for a secure communication (privacy,
integrity of the message, authenticity of the sender).
Students use tools to create a pair of keys, exchange
public keys, encrypt and digitally sign emails and
verify incoming emails.
Students explain on a general level how a computer
verifies the integrity of a message and the
authenticity of its sender.
Level 2 Students assess the level of security provided by
RSA using keys of different length based on
computer experiments for a reconstruction of a
private key and an internet research on the RSA
Challenge.
Level 3 Students create pairs of keys using the RSA key
generation algorithm with small prime numbers and
use them to encrypt and decrypt messages manually.
Students reconstruct private keys with small prime
numbers manually, explore algorithms to find large
prime numbers and define the relation between the
length of keys and the level of security provided
using the problem to factor large semi-prime
numbers as a mathematical argument. They assess
the level of security provided by RSA using keys of
different length based on mathematical reasoning.
In the German documentation of the learning process, links to
relevant sections of the GI standards for teaching computer
science are provided for each section of the learning process.
Relevant competencies are mostly from the content standards
informatics systems and informatics, man and society as well as
the process standards reason and evaluate, for example:
- Students understand the basic structure and functionality of
informatics systems.
- Students react appropriately to risks arising from the use of
informatics systems.
- Students ask questions and state hypotheses on matters of
informatics.
- Students measure different criteria and assess their adequacy
for their own actions.
124
Relevant links to the K–12 Computer Science Standards
published by the Computer Science Teachers Association (CSTA)
in 2011 [5] are not pointed out but can be easily found, for
example:
- Students explain the multiple levels of hardware and
software that support program execution (e.g., […]
networks), […] describe how the Internet facilitates global
communication. [Computers and Communications Devices]
- Students exhibit legal and ethical behaviors when using
information and technology and discuss the consequences of
misuse, analyze the positive and negative impacts of
computing on human culture. [Community, Global, and
Ethical Impacts]
- Students explain the principles of security by examining
encryption, cryptography, and authentication techniques
[Computing Practice and Programming].
While the material is in German, we have translated selected
extracts of the material into English, to enable readers to
understand how the materials are intended to work. The full
material as well as a didactic comment can be found at [22].
In the following sections, we describe the activities and the
material developed to stimulate and enable the activities in the
context-based learning process.
2.1 Discovering Email Protocols
The first question asked is: “How is an email communicated from
my computer to the computer of the addressee?” To understand
internetwork communication, students need to find out about the
structure of interconnected computernetworks such as the Internet
as well as end-to-end communication protocols between client
computer and email server such as the Simple Mail Transfer
Protocol (SMTP) for sending email and the Post Office Protocol
(POP) for retrieving email from a mail server.
To understand the characteristics and function of communication
protocols, students are first asked to communicate a single word
through a door by pulling a cord that passes underneath a door.
Students will soon come up with ideas for coding letters as long
or short, strong or weak pulls or pulls to one of two possible
sides. Inevitably from time to time, students will make mistakes in
their coding or decoding and thus will agree on signs to cancel the
transmission of a letter and start again from the beginning. They
will also decide for a sign to show the end of a letter, e.g. a long
break. Some groups agree on a sign for the receiver to signal an
acknowledgement. This way, without any explicit knowledge of
what a communication protocol is, they have defined such a
protocol including the separation of operational and data signals.
When comparing the protocols of different groups they will find
different quality of service criteria such as speed and reliability.
The next step will be to have an insight into real life email
network traffic. To this end, we set up a new email server in the
classrooms. We use the email server software Hamster which
needs no installation and runs from a USB flash drive, that the
teacher provides during the computer science classes. Students
will sign up for an individual account one after another setting a
secret password, which should not be one they are already using
for another account.

Figure 1. Network traffic for authentification via POP.
After configuring their email account in the email client
application Mozilla Thunderbird, they start sending each other
emails and watch the generated email network traffic with a
network traffic analyzer called Socket Sniff (cf. Figure 1).
Figure 2. Pupils are to rearrange the messages in the
correct order using the network traffic analyzed.
Students are asked to analyze their network traffic to reconstruct
either the Simple Mail Transfer Protocol or the Post Office
Protocol. Figure 2 shows a scrambled collection of generalized
messages. Using the example of their concrete network
connection, students will one by one identify these messages in
their concrete network traffic and thus bring the general messages
into the order specified in the emailing protocol in question. The
left side shows terms for some more general steps, which are to be
assigned to a number of messages exchanged to identify typical
phases of communication such as initiating a communication
125
channel, authorizing the communication partners, and closing the
communication channel.
In a following cooperative exchange phase students will explain
the protocol they have reconstructed to a student who has
reconstructed the other protocol and then discuss similarities and
differences of the two communication protocols.
2.2 Discovering possible Dangers
The second question structuring the learning process is: “What
dangers challenge a private communication over a public network,
such as the Internet?” This question is not explicitly asked in the
class room. Instead, while students are still working on
reconstructing the emailing protocols they experience the teacher
exploiting the dangers of an unencrypted, plain-text communication
over a public network in the didactic environment of the
class room’s local area network, the different dangers are then
collected in a discussion based on the students’ experiences. This
is actually the reason why we use a separate mail server for the
computer science classes and not the student’s private email
accounts. Therefore, students should be advised at the beginning
to choose a password other than any passwords they already use.
Figure 3. Unsolicited mail from the German Chancellor?
First, the teacher sends an email with a false sender address. We
used an email which claims, that the German Chancellor Angela
Merkel suggests buying stocks of the Russian gas company
Gazprom (see Figure 3). While many email providers check the
correctness of a sender’s address, this is not required by the
SMTP protocol or the Hamster email server, The email shows
several features that identify it as spam: It shows frequent spelling
mistakes, shows no specific knowledge about the addressee (Be
honest, who is really friend with a head of state?), shows an
inappropriate choice of colloquial language and the link leads to a
Url that differs from the text shown for the link. The last aspect is
also the reason why modern versions of email client applications
identify the mail as a possible fraud mail such as phishing mails.
These features can be collected in the discussion to raise the
students’ awareness of spam and phishing mails.

While spam and phishing mails can be easily detected when users
are aware of the features discussed above, users are more likely to
react to emails that seem more personal. If we go back to Figure 1,
we can see that anyone between the client computer and the mail
server can read the email. With this knowledge, a malevolent
communication participant could send a more individual email, in
which he or she relates to something personal he or she knows
from reading intercepted email. Not only can such a person
retrieve personal information. The Post Office Protocol requires
users to send their password in plain text. Intercepting password
and username, a malevolent communication participant could
access the mailbox and send mail from this mailbox as if it were
his or her own. To demonstrate intercepting email on the
communication path between client an mail server, the server is
connected to the school network via a computer which is extended
by a second network card and configured to bridge the two
network cards. While this computer is technically a network
bridge, it is perfectly fit to simulate an inter-network router. We
have decided to provide students with a network traffic analyzer
with very limited potential. It only reads the traffic assigned to a
process running on the machine that the analyzer is running on.
To intercept messages, teachers need to make use of a more
powerful analyzer tool such as Wireshark.
Thirdly, a person with access to the mail server’s memory could
modify the content of emails before they are downloaded by the
user. Here, the teacher manipulates emails by opening them from
the computer’s file system using a simple text editor, changes
some facts such as the date of a suggested appointment and saves
the changes made to the file. Students follow the manipulation on
a video projector attached to the computer on which the mail
server is running.
The discussion of the dangers to a private communication over a
public network should lead to describing the requirements for
secure communication over public networks shown in Table 2:
Table 2. Dangers to and requirements for secure
communication over public networks.
danger requirement
intercepting messages privacy
manipulating messages integrity of received messages
faking the sender’s address authenticity of senders
These three requirements structure the steps striving to find
mechanisms that provide the desired security feature as described
in the next section.
2.3 Achieving Privacy through Encryption
By retracing the genesis of cryptographic algorithms, students
should not only learn of certain algorithms but develop an
awareness of important criteria for a cryptographic algorithm and
learn to question the level of security provided by a certain
security feature. They should also learn that security always
requires a certain amount of trust into people and technology and
can never be fully guaranteed.
The desire to hide information on the communication path to
prevent it from being intercepted by enemies is not at all a new
requirement. The development of cryptographic algorithms dates
back to ancient times. The most prominent historic algorithm is
126
probably the Caesar algorithm, where Roman emperors
substituted the characters of a message by other characters,
retrieved by shifting an alphabet by a certain amount of letters.
The key here is how far the two alphabets are shifted. This
algorithm can easily be applied by children of all ages by shifting
two stripes showing two alphabets each, as shown in Figure 4.
Figure 4. Alphabets to apply the Caesar algorithm.
In order to remind students of the overall aim to achieve means
for secure emailing, they are asked to exchange messages
encrypted with the Caesar algorithm via email. They are then also
challenged to crack a Caesar cipher of which the key is unknown.
By providing a tool called Krypto 1.5 by Michael Kühn to de- and
encrypt Caesar ciphers, we enable students to quickly test the 25
possible keys for a Caesar encryption.
This way it becomes clear that despite having been in use for
centuries the Caesar algorithm is not safe at all. One could argue
that if the letters in the target alphabet were to be distributed to
arbitrary positions and not in the alphabetic order, there would be
26! = 403.291.461.126.605.635.584.000.000 possible keys. While
it is true, that it seems impossible to test all possible keys, longer
texts can still be cracked by comparing the frequency of signs to
that of the typical frequency of letters in the assumed language of
the original message. This mechanism is well explained in the
story “The Gold Bug” by Edgar Allen Poe.
A more promising way seems to work with multiple alphabets, as
the Vigenère algorithm does, probably the best known symmetric
polyalphabetic algorithm. Here several keys are employed,
described by the letters of a keyword. The letter at the current
position of the keyword determines how far the alphabet for a
Caesar encryption of this letter is shifted.
Figure 5. Animation of Vigenère with Krypto 1.5.
The tool Krypto 1.5 provides an excellent animation of the
algorithm. A screenshot of this animation is shown in Figure 5.
With this animation students can watch the program slowly
encrypting a text.

A weak point of the Vigenère algorithm is that when the key is
used repeatedly, typical frequent short sequences of letters such as
“and”, “in”, and “but” are likely to translate to the same sequences
in the cipher. The least common multiple of the positions of such
parallel sequences hints to the length of the key. If the length of
the key is guessed, a frequency analysis can be conducted and the
key can be reconstructed. To secure 100% security, the key used
must be of the same length as the message. This lead to the
creation of code books called “one time pads”.
While the cipher generated from an arbitrarily generated one-time
pad cannot be cracked, sender and receiver still need to hold a
copy of the same one-time-pad. The exchange of secret keys can
only be avoided by means of asymmetric cryptography such as the
RSA algorithm. While the underlying mathematical concepts of
RSA are not trivial, the idea of using pairs of keys where each
communication participant holds a secret and a public key can be
understood even by young learners.
The image of a padlock and a key can be used to a certain extend.
We could imagine handing out a number of unlocked padlocks to
communication partners, who can now lock a box by closing the
lock. If we keep the key, we are the only person able to unlock the
box. Of course, we cannot communicate padlocks digitally, so we
need a function on numbers to do a comparable job. Here the
modulo function comes into play.
Figure 6. Animation of asymmetric cryptography.
To introduce students to the idea of asymmetric cryptography we
have developed an online form, where students can exchange
messages and encrypt them with public and private keys. Figure 6
shows the form in use. Very small numbers are used as keys so
that students can better follow the flow of information. The form
is accompanied by instructions on how to apply the keys. If they
understand the instructions correctly, they will be able to
successfully decrypt messages encrypted with their public key.
Once the idea of asymmetric cryptography has been established,
the algorithm for generating RSA keys is presented and students
manually en- and decrypt their birthday using very small keys.
Students are then asked to challenge the encryption with small
numbers by trying to reconstruct a possible private key for a given
public key. They will find out, that using larger keys makes it
more difficult to reconstruct a private key. This raises the question
127
how large a key must be so that modern computers will not be
able to crack an RSA cipher in a realistic time span. We use the
tool CrypTool to challenge larger RSA ciphers and ask students to
gather information on the RSA Factoring Challenge.
2.4 Verifying a digital Signature
Figure 6 shows, that from the texts a hash value is generated, that
is unique for a given sequence of characters. When the sender
sends the hash value as well, the receiver can detect changes made
to the message, because the hash value calculated from the
received text does not match the one sent by the sender.
Unfortunately, a person could manipulate both the message and
the hash value. This means, that the hash value has to be
protected. The sender encrypts the hash value using his or her
own private key. Now anyone can decrypt the signature using the
sender’s public key to verify the signature. But nobody except the
sender can produce a signature that can be correctly decrypted
with the sender’s public key. This way, both the message’s
integrity and the sender’s authenticity can be verified.
Students are often confused that now the order in which the keys
are employed differs from the one used to encrypt messages to
achieve privacy. Teachers are best advised to point out and
discuss this difference and stress the fact, that this mechanism
suits a very different purpose.
To acquire also practical competencies in using digital signatures
students should first encrypt and sign emails using the animation
from the project website. Then, teachers should motivate them to
configure the PGP-plugin Enigmail for Mozilla Thunderbird,
generate a set of keys, exchange public keys with their fellow
students and encrypt and sign, decrypt and verify emails and
signatures with Enigmail.
2.5 Motivation for secure Communication
In a jigsaw puzzle students first study texts to become experts for
one of the following four topics:
1. Freedom of communication as a request in countries
under autocratic rule such as Iran or China.
2. The Echelon project as an example of an intelligence
system analyzing email traffic.
3. De-Mail as an example of attempts for commercial
email cryptography with trust centres.
4. Pretty Good Privacy (PGP) as an example of royalty
free cryptography with a web of trust.
Students then exchange the expert knowledge they have acquired
in mixed groups.
3. LESSONS LEARNED
One question frequently discussed is: Should teachers find subject
matter that suits an interesting context or should they find a
context that is suitable to teach a certain subject matter?
Traditionally, curricula point out fields of subject matter and
assign them to certain years of learning. Teachers therefore are
used to decide on the subject matter first, and then select a
suitable contextualisation or application of an abstract scientific
concept. More recently, curricula also point out fields of
competences as well as fields of content. Teachers now have to
decide on which competences they want their students to acquire,
which content from the curriculum is suitable to develop these
competences and which context can be used to show the concept’s

elevance. Within the IniK working group, we have started out
designing context-based units both from a context, brainstorming
on subject matter that could best be acquired, as well as from a
certain subject matter, brainstorming on what context could best
demonstrate the concept's relevance to students. In the case of
“Email for You (only?)” we first grouped and rearranged parts of
activities developed earlier and then modified and added material
to bridge gaps where needed. In practise, both ways seem
possible, practical and worth exploring.
Figure 7. Fundamental and optional modules.
In the student-oriented learning process students spend a lot of
time actively and individually engaging with the different topics.
It is at times challenging to select just some of the many possible
aspects and dimensions of a topic. Several very different units are
possible for a single context. If a learning process turns out to
require a lot of time, it makes sense to mark several aspects as
optional and others as mandatory parts, to encourage teachers to
first try out parts of the material. We have grouped the different
learning activities into modules as shown in Figure 7.
While some modules are necessary for an understanding of
following modules, others only lead to a more thorough
understanding of aspects such as the concrete implementation of
algorithms or the level of security they provide.
The proposed learning arrangement involves a lot of software. All
the software proposed can be used free of royalties. While most of
them are available for Windows, Mac and Linux operating system,
some are only available for the Windows platform. Here,
alternative applications should be found for teachers working with
other platforms. For a simple start we have gathered portable
versions of the software into a Zip-File, which can be extracted to
a USB flash drive. Programs can then be started directly from the
USB flash drive without any need for setup routines. This is
extremely practical for teachers working in an environment, where
software installation on clients is laborious. A disadvantage of
providing the Zip-File is, that the applications are quickly
outdated, especially in the case of Mozilla Thunderbird. This is
also true for the directions given to students. Several times, these
directions had to be adjusted after the dialogue for setting up an
email account in Thunderbird was changed. (As a matter of fact, it
128
has been improved in the sense that users are now warned of
connecting to a mail server without SSL or TLS encryption. We
have found a way to work around the warning so that students can
still discover the dangers of plain text communication with an
email server.)
While we believe we have made good use of the context for
purposes of providing illustration and meaning, we must admit
that with using the material as it is little is done to decontextualise
newly gained knowledge by applying it to similar problems in
other fields of application, e. g. designing a protocol for the
communication of different parts or layers of services within a
single device or discussing security issues in other situations, e. g.
large permanent storages such as data bases. It is our impression
that most of the other learning processes proposed at the projects
website [21] lack opportunities for decontextualisation. Since the
curricula only give orientation and concrete courses in computer
science vary a lot between different schools and teachers, it is
difficult to suggest connections to what students have learned
before. Thus, it is left to teachers to realize these connections – so
far no guidance is given on how to achieve decontextualisation.
Finally, designing, implementing, testing and readjusting “Email
for You (only?)” has taken a lot of time. If the project is to be
successful in establishing the concept in everyday CS classes, it
will be necessary to acquire further resources to give teachers and
university staff the time needed to develop such material. A
promising approach could be to include university students and
teacher trainees in the design and reflection of context-based
learning processes.
4. FUTURE WORK
Further contexts should be explored to estimate their potential for
teaching computer science in secondary school. The IniK project
group has started a list of contexts [23] which is grouped into
ideas for contexts with a potential but need for developing
suitable teaching materials, ideas for contexts where it is unclear,
which concrete aspects of the contexts are useful to the teaching
of computer science in secondary school education, and ideas for
contexts which at a first glance seem attractive but after a more
thorough examination don’t seem useful to the teaching of
computer science in secondary school education. It would be
helpful to have a more detailed understanding as to what potential
different contexts have. A suggested learning process will always
discuss only a selection of possible aspects, so eventually we
could have different units exploring the same contexts with
different focus. More teaching materials for further topics should
be developed. Today, there is e. g. no unit that motivates the
design and use of databases – visualization using open data seems
an interesting approach here.
But next to the development of further material, experiences made
in the design and implementation of context-based computer
science courses should be collected, shared and discussed in order
to develop guidelines and advice for the development of contextbased
learning processes in secondary school computer science
education. An example would be concepts for a sustainable
decontextualisation, which are missing so far.
5. CONCLUSION
“Email four You (only?)” shows that a context-based learning of
computer science is possible and helps students understand the
relevance of knowledge they acquire in computer science classes.

We have learnt that we can both look for contexts where a certain
topic is of relevance as well as screen an interesting context as to
which topics of computer science education are relevant for it.
Either way, it is a long way of selecting suitable aspects out of a
wide choice of possible aspects and developing material which
makes sense at a certain point of the learning process. If in doubt,
one should decide on whether aspects are necessary for the
following sections of the learning process. If this is not the case,
they can be marked as optional. Whenever software is involved,
the software should be easy to use, if possible without installation.
We have by now presented our proposed learning arrangement in
several hands-on workshops for computer science teachers
throughout Germany and also in the journal LOG IN [16], which
is widely distributed amongst computer science teachers
throughout Germany. Since then, we have received a lot of
positive feedback from colleagues stating that their students are
highly motivated to understand both the underlying technology of
email as well as the different cryptographic algorithms. While we
are happy to see our own impressions affirmed, we are indeed
aware that this is only a first but nevertheless promising hint as to
the potential of context-based learning of computer science in
secondary school education. Reaching a stage where we now have
several learning processes readily prepared and testes in class, it
now seems important to validate our theses on context-based
learning of computer science in a more through, scientific
evaluation.
6. REFERENCES
[1] Arbeitskreis »Bildungsstandards« der Gesellschaft für
Informatik (ed.), 2008. Grundsätze und Standards für die
Informatik in der Schule – Bildungsstandards Informatik
für die Sekundarstufe I. Empfehlungen der Gesellschaft
für Informatik e. V. vom 24. Januar 2008. In LOG IN ,
supplementary to vol. 150/151 (2008). Retrieved
October 10, 2012. http://informatikstandards.de
[2] Altrichter, H., Posch, P., Somekh, B. 1993. Teachers
Investigate their Work: Introduction to the Method of
Action Research. Routledge, London, New York 1993.
[3] Barab, S. and Squire, K. 2004. Introduction: Design-
Based Research: Putting a Stake in the Ground. The
Journal of learning sciences. 13,1 (2004), 1-14.
[4] Brinda, T., Puhlmann, H., and Schulte, C., 2009.
Bridging ICT and CS – Educational Standards for
Computer Science in Lower Secondary Education. In
ITiCSE '09: Proceedings of the 14th annual ACM
SIGCSE conference on Innovation and technology in
computer science education, 288-292.
[5] Computer Science Teachers Association (CSTA) 2011.
K-12 Computer Science Standards. Revised 2011.
Retrieved October 10, 2012. http://csta.acm.org/
Curriculum/sub/CurrFiles/CSTA_K-12_CSS.pdf
[6] Cooper, S. and Cunningham, S. 2010. Teaching
computer science in context. ACM Inroads 1, 5‐8.
[7] Coy, W. 2005. Informatik … im Großen und Ganzen. In
LOG IN 136 (2005), pp.17-23.
[8] Diethelm, I., Borowski, C., and Weber, T. 2010.
Identifying relevant CS contexts using the miracle
question. In Proceedings of the 10th Koli Calling
International Conference on Computing Education
129
Research. Koli Calling 10. ACM, New York, NY, USA,
74‐75.
[9] Diethelm, I. and Dörge, C. 2010. From Context to
Competencies. In Key Competencies in the Knowledge
Society, N. Reynolds and M. Turcsányi-Szabó, Eds. IFIP
Advances in Information and Communication
Technology. Springer Boston, 67–77.
[10] Diethelm, I., Hildebrandt, C., and Krekeler, L. 2009.
Implementation of Computer Science in Context - a
research perspective regarding teacher-training. In Koli
Calling 2009. 9th International Conference on
Computing Education Research, A. Pears and C. Schulte,
Eds., 96–99.
[11] Diethelm, I., Schulte, C. and Witten, H. 2011. Informatik
im Kontext (IniK) – Entwicklungen, Merkmale und
Perspektiven. In Praxisberichte zur 14. GI-Fachtagung
Informatik und Schule (INFOS), Münster 2011.
[12] Engbring, D. 2005. Informatik im Kontext. In LOG IN,
vol. 136 (2005), pp.28-33.
[13] Esslinger, B., Gramm, A., Hornung, M. and Witten, H.
2011. Asymmetrische Kryptographie für die Sek I - RSA
(fast) ohne Mathematik? In Informatik mit Kopf, Herz
und Hand. Praxisbeiträge zur 14. GI-Fachtagung
Informatik und Schule (INFOS 2011) (Münster,
Germany, September 13 - 15, 2011) Zentrum für Lehrerbildung
der Universität Münster 2011, pp. 225-235.
[14] Esslinger, B., Gramm, A., Hornung, M. and Witten, H.
2012. Kann man RSA vertrauen? In LOG IN vol.
172/173 (2012), in print.
[15] Gramm, A. 2012. Animation of Asymmetric
Cryptography. Retrieved October 10, 2012
http://it-lehren.de/asymcrypt
[16] Gramm, A., Hornung, M. and Witten, H. 2011. E-Mail
(nur?) für Dich - Eine Unterrichtsreihe des Projekts
Informatik im Kontext. In LOG IN , supplementary to
vol. 169/170 (2011).
[17] Guizdal, M. 2010. Does contextualized computing
education help? ACM Inroads 1, 4, 4‐6.
[18] mpfs – Medienpädagogischer Forschungsverbund
Südwest (ed.), 2011. JIM-Studie 2011 – Jugend,
Information, (Multi-)Media. Basisuntersuchung zum
Medienumgang 12- bis 19-Jähriger. Stuttgart:
Medienpädagogischer Forschungsverbund Südwest,
2011, p. 32. Retrieved October 10, 2012.
http://www.mpfs.de/fileadmin/JIM-pdf11/JIM2011.pdf
[19] Parchmann, I., Gräsel, C., Baer, A., Demuth, R. and
Ralle, B. 2007. Chemie im Kontext - a symbiotic
implementation of a context-based teaching and learning
approach. In International Journal of Science Education
28, 09 (2007) pp. 1041-1062. DOI=
http://doi.acm.org/10.1080/09500690600702512
[20] Klieme, E. et al. 2004. The Development of National
Educational Standards. An Expertise. Bundesministerium
für Bildung und Forschung, Berlin.
[21] Koubek, J. 2012. Website of the “Informatik im Kontext”
project. Retrieved October 10, 2012.
http://informatik-im-kontext.de

[22] Koubek, J. 2012. Website of the “Informatik im Kontext”
project – section “Entwürfe » Email (nur?) für Dich”.
Retrieved October 10, 2012. http://informatik-imkontext.de/index.php/entwuerfe/email-nur-fuer-dich/
[23] Koubek, J. 2012. Website of the “Informatik im Kontext“
project – section “Kontexte und Ideen”. Retrieved
October 10, 2012.
http://informatik-im-kontext.de/index.php/kontextideen/
[24] Koubek, J.; Schulte, C.; Schulze, P. and Witten, H. 2009.
Informatik im Kontext (IniK) – Ein integratives
130
Unterrichtskonzept für den Informatikunterricht. In GI-
Edition LNI – Lecture Notes in Informatics, vol. P156.
Zukunft braucht Herkunft – 25 Jahre »INFOS –
Informatik und Schule«. INFOS 2009 – 13. GI-
Fachtagung Informatik und Schule 21.–24. September
2009 in Berlin, pp. 268–279.
[25] Saeli, M., Schulte, S. To be published 2013. Applying
Standards to Computer Science Education. In Improving
Computer Science Education (by Kadijevic, D., Angeli,
C., Schulte, C.), 117-131.

Comparing CSTA K-12 Computer Science Standards
with Austrian Curricula
Daniel L. Egger
Alpen-Adria-Universität Klagenfurt
Universitätsstraße 65-67
A- 9020 Klagenfurt
+43 650 519 7333
egger.it@edu.uni-klu.ac.at
ABSTRACT
In 2011, the CSTA has published a review of its K-12 Computer
Science Standards. Aiming to assess the situation of Computer
Science Education in Austria, we used these standards as an external
level rod. For this purpose, we investigated to which degree
a well-chosen subset of these standards is implemented in Austrian
curricula for computer science education. We analyzed curricula
of several types of secondary schools (AHS, HTL for
Chemistry, HTL for Informatics, HLW). Our findings reveal that
the CSTA standards are quite poorly implemented at this point of
time in Austria.
Categories and Subject Descriptors
K.3.2 [Computer and Information Science Education]: Computer
Science Education, Curriculum.
General Terms
Human Factors, Standardization
Keywords
Computer Science Education, Secondary School, Educational
Standards, Curricula
1. INTRODUCTION
During the last years, more and more countries have come to the
conclusion that it would be necessary to incorporate serious programs
for Computer Science Education (CSE) in their school
systems. As stated in 2011 by the US Computer Science Teacher
Association (CSTA): “To function in society, every citizen in the
21 st century must understand at least the principles of computer
science. […] Elementary and secondary schools have a unique
opportunity and responsibility to address this need.” [10]. Nevertheless,
as the education systems of the countries differ substantially,
the implementation and organization of CSE has to be also
very diverse from country to country. In order to compare the
outcomes of such different education systems, suitable educational
standards have to be defined and applied.
This paper, which was produced mainly by the students of a
teacher education course at our university, aims to demonstrate
how such a comparison might look like. We investigated, to
which degree a suitable subset of the K-12 standards that have
been most recently published by the CSTA [10] is adopted by the
Austrian education system. Our work was inspired by the CSTA
study “Running On Empty” that compared the outcomes of CSE
in the 50 US states. The comparison will show to which degree
the current Austrian programs for CSE fulfill the US standards.
Sabrina M. Elsenbaumer
Alpen-Adria-Universität Klagenfurt
Universitätsstraße 65-67
A- 9020 Klagenfurt
+43 650 830 3302
selsenba@edu.uni-klu.ac.at
131
Peter Hubwieser
Technische Universität München
Fakultät für Informatik
Boltzmannstr. 3, D-85478 Garching
+49 89 289 17350
Peter.Hubwieser@tum.de
Although there are other educational standards for CSE, this
paper will not deal with any standards in detail other than the
CSTA standards. It is also not the purpose of this paper to discuss
or argue for the appropriateness of the CSTA standards in general.
We will simply take them for granted and for a reference of comparison
for the means of this paper.
As we do not intend to discuss the process of standardization for
CSE in general, we refer to [5] and [3], where the current situation
regarding standardization in CSE in connection with curriculum
topics is discussed systematically.
We will start with the explanation of the theoretical background
and of the Austrian school system and continue with a short overview
of the CSTA standards. A final comparison of the individual
curricula will detect the differences in implementation, which will
give reason for a general conclusion.
2. THEORETICAL BACKGROUND
Klieme et al. (see [4] p. 42) define educational standards as “a
clear and concise statement of what matters in [a] school system”.
However, educational standards shall not be confused with regulations
for grading and examination. They do not substitute for
whole curricula, but rather specify fundamental themes in key
learning fields (see [4], pp. 85-86). It seems clear that standards
can only be effective if they mirror parents’, teachers’ and students’
expectations of what should be taught/learned and if they
only serve a purpose other than intensification of the already
predominant (see [7], pp. 203). Also, it seems as if educational
standards could provide an instrument to foster system-wide
educational equality and comparability.
Yet, some negative aspects have to be considered. First, it seems
obvious that some teachers may perceive educational standards as
an annoyance, irritation or burden limiting their flexibility and
freedom in teaching. Furthermore, the teaching staff will have to
agree upon how set standards can be met, which definitely can
lead to major arguments among the staff (see [4],pp. 42-49).
In addition, there is a strong disagreement about whether standards
in education lead to standardization rather than to diversification
of learning processes. Reigeluth [7] warns that we have to
discriminate between two uses/types of standards: standards that
foster uniformity and standards that support customization,
whereby the latter is definitely to be given priority. Uniform
standards will make the education system increasingly inflexible
and in “the broadest sense, [such standards can be seen as] an
expression of professional values in the context of a test’s purpose”
(see [6], p. 464) thus they could have a huge impact on
business and economy (see [1], p. 266).

On the contrary, customizing standards allow for a greater degree
of flexibility in teaching as they respect that different students
learn at different rates. This means standards must specify minimal
standards on the one hand and extended standards, in which
students can pursue their personal interests further, on the other.
In doing so, Reigeluth suggests establishing measurable, crossgradual
standard levels without relation to timetables, permitting
students to choose specifications freely within certain limits. Such
standards work against the sorting-out of students and promote
motivation building in reaching the respective standards [7].
Though educational standards affect everyone dealing with (public)
education ([7], p. 202), special attention should be given to
their implications on school curricula. “The experiences […] with
educational standards show that the implementation of this new
form of output-driven management is certain to prompt changes
in the work of schools, but the implications for the curricula are
not as clear-cut; rather, the options here are diverse and openended”,
state Klieme et al. [4] on p. 82.
3. RELATED WORK
The publication that was certainly the most influential for this
paper is the report Running on Empty: The Failure to Teach K–12
Computer Science in the Digital Age by Wilson et al. [12]. It
describes how computer science education is represented in the
curricula of the United States. Wilson et al. apply a methodology
that is very similar to ours, providing evidence for the implementation
of the 55 CSTA K-12 Standards in the version of 2006 [8]
in the mentioned curricula: “The researchers were very liberal in
the analysis; that is, they required minimal evidence to consider a
particular ACM/ CSTA standard as adopted. If the state standards
made any reference to the general idea detailed in the
ACM/CSTA standards, it was marked as adopted“ see [12], p. 34.
The report covers all 50 states of the USA and all 3 levels of the
CSTA standards from 2006.
Unfortunately, it provides a disillusioning result: It finds “that
roughly two-thirds of the entire country has few computer science
standards for secondary school education, K–8 computer science
standards are deeply confused, few states count computer science
as a core academic subject for graduation, and computer science
teacher certification is deeply flawed” [11].
The individual results vary largely from one state to another,
since there are states with hardly any standards implemented
while some conform to (almost) all of them. A neatly arranged,
interactive overview of the state-by-state results can be found at
[2] . The overall results are displayed in Fig. 1.
Figure 1. Overall results of the CSTA survey [2] .
As our investigation refers to the revised version of the CSTA
standards from 2011 ([10], see below), which is different from the
2006 version in many respects (e.g. the overall number of stand-
132
ards, the definition of the levels and the strands) it is , unfortunately,
not possible to compare our results with those of Wilson et
al. [12].
4. THE EDUCATIONAL CONTEXT
Since the CSTA Standards were developed in the context of the
US educational system, whereas we refer to the Austrian system,
it seems appropriate to outline the major differences and similarities
of these two systems (see Fig. 2). The relevance of national
policies and educational systems to CSE was discussed closely by
a recent ITiCSE working group [3].
In the Austrian school system, children are required to attend
school for 9 years (respectively grades). This section of compulsory
schooling includes 4 years of primary school education, 4
years of lower secondary school education and one additional
year from the upper secondary education level. For the upper
secondary education (up to the final ‘Reifeprüfungen’ that qualify
for enrollment at universities) there are various types of schools
with various specifications and also different numbers of years
that can be chosen by the students after they have passed grade 8,
e.g.
- AHS (“Allgemeinbildende höhere Schule”, aiming to general
education),
- HAK (“Handelsakademie”), focused on trade and business,
- HLW (“Höhere Lehranstalt für wirtschaftliche Berufe”),
mainly educating for tourism and business,
- HTL (“Höhere Technische Lehranstalt”), mainly technically
and business-oriented),
- BMS (“Berufsbildende mittlere Schulen”) that lead to a
vocational qualification.
Besides these types, there are specific schools for Vocational
Education and Training, specializing in one specific profession.
Figure 2: School Systems of Austria and the USA
In the US school system, the number of years (respectively
grades) which students are required to attend school differs from
one state to the next. Elementary schooling includes nursery
schools and kindergartens on the lowest level, which are followed
by elementary schools. The primary education may end with

grade 4, 5, 6 or 8, depending on choice of the parents (and the
system of the respective state). After primary schools, the children
may attend a middle school, a 4-year high school a junior high
school, followed by a senior high school or a combined juniorsenior
high school up to the 12th grade. After graduation from
high school, it is possible to attend various forms of postsecondary
education such as colleges, universities and vocational institutions.
5. CSTA STANDARDS
In addition to a revision of the ACM K-12 curriculum from 2003
[9] the CSTA published the first version of the K-12 CSE standards
in 2006 [8]. To keep track with the current flow of technological
development, the standards were revised again in 2011
[10]. This included also a substantial raise of the number of standards
(from 55 in 2006 to 174 in 2011).
5.1 Levels
The CSTA developed a three-level-model for K-12 computer
science, which means that each of the three levels aims to represent
different age- and knowledge-levels for students from Kindergartens
(K) up to grade 12. Level one is valid for K to grade 6,
level two from grade 6 to 9 and finally level three for grades 9 to
12.
Fig. 3 visualizes the three levels. Level 1 is called “Computer
Science and Me”, is designed for young elementary school students
and aims to support those with a basic knowledge in technology.
It demands simple concepts of computational thinking.
Furthermore, the students in K-6 should develop a first understanding
of the importance of computer science in an engaging,
creative and explorative atmosphere. Often these first ideas are
embedded within suitable contexts from social science, language,
arts or mathematics.
Figure 3. Three-Level-Model for CSTA Standards [10].
Level 2, titled “Computer Science and Community”, demands
deeper knowledge in order to understand computational thinking
as a problem solving tool. Moreover, grade 6-9 students should
apply computers as devices which facilitate communication and
collaboration within the computer science classroom and also in
other curricular areas.
Finally, according to level 3, the students in grades 9 to 12 should
be able to apply learned concepts and create real-world solutions.
The students are advanced learners of computer science, who
should already be able to contribute to the development of new
technologies. Level 3 is divided into three different categories,
each focusing on a diverse field of computer science. Level 3A,
called “Computer Science in the Modern World”, represents basic
abilities that should be acquired in grade 9 or 10. It addresses all
133
students, demanding consolidated knowledge and abilities. The
students should be able to make competent decisions regarding
computer systems and computational techniques and transfer
these abilities to whatever job they might be working in later.
Apart from making use of their knowledge, they should also be
aware of and appreciate the benefits of computers but also understand
the social and ethical impact of modern technologies in
today’s society.
Level 3B and 3C are intended for more advanced studies. In
grades 10 or 11, according to Level 3B (“Computer Science concepts
and practices”), students should learn algorithmic thinking
and problem-solving strategies, they should deepen their
knowledge of the principles of computer science and become
experts in working collaboratively, using collaboration tools when
solving problems. Level 3C, represents in-depth studies in one
particular field. It is recommended for grade 11 or 12 students
who wish to gain professional computing certification or equivalent
education, helping them to carve out their career (see [10],
pp. 7-9).
5.2 Strands
To take the complexity of computer science into account, the
CSTA standards distinguish five “complementary and essential
strands” (see [10], p. 9), displayed by Fig. 4. In the following we
give a short description of these five strands.
Computational Thinking (CT) is one of the core elements of
computer science. The main idea of CT is the problem-solving
methodology. Concepts like abstraction, recursion or analysis not
only target for educational consumers but also for producers of
new technological devices.
Collaboration (CO) starts at school, when students work cooperatively
together, gathering information or using a variety of communication
tools, to name just a few examples, and leads to an
effective collaboration on important projects with colleagues in
future careers.
Figure 4. Strands according to CSTA Standards [10].
Computing Practice & Programming (CP). As though programming
is an essential part of computer science, it is not the only
one. Other abilities, like developing homepages, using software
tools, handling databases, organizing files and folders, dealing
with computational problems, etc. are good examples of what
computing practice might also include.

Computers and Communications Devices (CC): the students
should understand the elements of modern computers and communication
devices and be able to use them competently.
Community, global and ethical impacts (CG) deals with ethical
norms, responsibility and general awareness. The students need to
be informed about privacy and security topics, and develop a
certain sense for ethically correct choices and appropriate social
networking behavior. It is important to understand the positive but
also negative impacts computers and technology might have on
everyone’s life.
6. METHODOLOGY
6.1 Selection of Curricula
As already mentioned in the introduction, the results that are
presented here were produced by students during a seminar that is
part of a teacher education program. Under these circumstances it
was not possible to investigate the adoption of all 174 standards.
Also, it would have taken too much time to analyze the curricula
of all Austrian school types (which are numerous, see section 4).
Thus we had to select several curricula and an appropriate subset
of standards for our research. As computer science education in
Austria does not take place before secondary education, we decided
to focus on the curricula of several important types of secondary
schools, which are described in the following subsections.
6.1.1 AHS
The “Allgemeinbildende Höhere Schule” is the school type with
the broadest form of higher education. Schools that belong to this
category include the so-called “Gymnasium” and the “Bundesoberstufenrealgymnasium”
(BORG). An AHS is a school with
focus on a comprehensive and broad general education that lays
the foundations for the studies at a university. The AHS is divided
into an “Unterstufe”, comprising the grades 5 to 8, and an “Oberstufe”
that consists of grades 9 to 12. After successfully passing
the “Unterstufe”, students may subsequently attend the “Oberstufe”,
but instead they might also choose to attend a different
school like a BHS, which will be explained in the chapter below,
or might switch to vocational education and training. Furthermore,
the AHS distinguishes between three kinds of branches:
The “Gymnasium”, the “Realgymnasium” and the
“wirtschaftskundliche Realgymnasium”. After the sixth grade,
students have to decide for one of these three branches. All of
them follow almost the same curriculum, except a few specializations
for each branch. The main focus of the “Gymnasium” is on
languages, whereas the “Realgymnasium” goes deeper into mathematics
and natural sciences. The “wirtschaftkundliche Realgymnasium”,
specializes in economics, chemistry, psychology and
philosophy. However, these three types are still targeted on general
education and cannot be compared to e.g. the BHS, which are
much more specialized schools aiming for a certain career path.
6.1.2 HTLs for Informatics and for Chemistry
This type of school (“Höhere Technische Lehranstalt”) is mostly
associated with a strong technical and scientific education. However,
there are numerous fields of specialization like Computer
Science, Business Administrations, Chemistry, IT, Networking,
Mechanics, Design and many more. We analyzed the curricula of
two branches that might represent the opposite end of the CSE
spectrum: First, the branch dedicated specifically to computer
sciences, and second, the branch for Chemistry that has probably
the minimal CSE implementation. These two extreme values
should delimit the range of CSE in all the HTL curricula.
134
Because of the large variety of specializations, each branch of
HTL has its own specialized curriculum. However, there is also a
basic/core curriculum for all of the branches which describes the
respective competencies that the students should have by the end
of their studies. The specialized branch-curricula then make
changes and amendments to the basic/core curriculum in order to
put importance on the respective focus of the branch.
6.1.3 HLW
The “Höhere Lehranstalt für Wirtschaftliche Berufe” has its focus
clearly on business administrations and tourism, although other
specializations exist. This curriculum should serve as a counterexample
to the very technical HTL curricula.
Other schools like HAK (Handelsakademien) or Tourism schools
have different specializations, though they cover pretty much the
same computer science topics, as HLWs. For this, and timesaving
reasons, only the curriculum for HLWs will be sampled
out. Since 2009, the curriculum regarding computer science
courses got revised and adapted. Actuality is very important and,
therefore, the revised curriculum of 2009 will serve as the basis
for comparison. The new curriculum proposes two mandatory
computer science courses for all students attending a HLW. These
two are titled Office- and Information-Management and Applied
Informatics.
6.2 Selection of Standards
When looking at the subset of selected curricula, it becomes
obvious that all of them are located at the higher secondary level,
from the 9 th to the 12 th /13 th grade. Comparing this to the CSTA
standards framework, it seems suitable to limit the regarded set of
standards to level 3A as, according to the CSTA, the standards of
this level should be implemented throughout the whole range of
upper secondary schools. Level 3B and 3C already represent indepth
specializations that are not to be intended by schools of all
types, but only at the ones that focus on computer sciences. In
consequence, as the viewpoint of this paper lies on general education,
we will restrict our investigations to the competencies subset
of level 3A.
The table 8 in the appendix lists the standards of CSTA Level 3A,
which we have chosen as a reference for our research about the
adoption by the curricula of the selected school types listed
above.
6.3 The Measure of Adoption
To measure the degree of the adoption of the standards by the
curricula, we had to define, under which condition a standard can
be considered as implemented.
It has to be mentioned that it is not possible to look for literal
incorporations, since, first, Austrian curricula are not written or
available in English and, secondly, this would not detect all forms
of implementations. Thus it seems more suitable to look for
equivalents in content, meaning that the described competencies
have to be featured in the respective curricula in some respect.
For this, it seems appropriate to not only rate if standards were
fully implemented or not, but also whether the standard has been
implemented partially. For this, we applied the scoring scale that
is shown in table 1.

Table 1. Scoring Scale for the Implementation of a Standard
Rating Description
The specific competence is implemented in the
1 respective curriculum more or less totally (i.e. the
specific competence is described in the curriculum).
The specific competence is implemented partially
0.5 in the respective curriculum (i.e. something similar
bot not equivalent can be found in the curriculum).
The specific competence is not at all implemented
in the respective curriculum (i.e. nothing similar to
0
the specific competence can be found in the curriculum).
6.4 The Rating Process
To ensure a fair amount of objectivity and reliability, all ratings
were cross-examined and double checked by the two student
authors of this paper, who generally worked in parallel on separate
parts of the curricula. Thereby, the rating itself happened
exclusively on the basis of the statements featured in the curricula.
Due to the sometimes very universal statements in the curricula,
it was occasionally difficult to decide which rating to apply.
As an example we will draw on the standard 4.2 Develop criteria
for purchasing or upgrading computer system hardware (cf. table
8 in appendix). The HTL core curriculum states “assess a PC
configuration and make purchasing decisions” 1 which is basically
the same as specified by the respective standard. Consequently,
the rating 1 was applied. The HLW curriculum, however, only
states “hardware and software requirements” 2 . We decided that,
although one could infer the ability to make purchasing decisions
from someone’s knowledge of the hardware requirements, this
was too general and that therefore only the rating 0.5 should be
applied. By and large, we judged strictly and usually opted for
giving the worse rating in borderline cases.
7. RESULTS AND DISCUSSION
First, we present the results of our investigation for the selected
school types. This is followed by a comparison and summary of
the results of all types.
7.1 Results for the Selected School Types
7.1.1 AHS
For the purpose of further elaborations, only the curricula of the
“Oberstufe” of AHS will be investigated in order to compare and
contrast them with the standards of level 3A, as they are recommended
approximately for the same age levels and grades. However,
computer science courses are only mandatory in the ninth
grade only. Students in grade 10 to 12 are offered elective computer
science courses, which is referred to as “Wahlpflichtfach”
and follows a separate curriculum, which is included in the analysis
as well.
Table 2. Selected Standards in the AHS Curriculum
Standard
No.
Strand Strand Strand Strand
CT CO CP CC
x=1 x=2 x=3 x=4 x=5
x.1 0 0 0 0 0
x.2 0 0.5 0 0 0
x.3 0 0.5 0 0.5 0
Strand
CG
1
German: “eine PC-Konfiguration bewerten und Anschaffungsentscheidungen
treffen“
2
German: „Hard- und Softwareanforderungen“
135
x.4 0.5 0 0 0 1
x.5 0
0 0 0
x.6 0.5 0.5 0 0.5
x.7 0 0 0 0
x.8 0.5 0 0.5 0
x.9 0.5 0 0 0
x.10 0 1 0.5 0.5
x.11 0 0.5
0
x.12 0
Sum 2 / 11 1 / 4 2 / 12 1.5 / 10 2 / 11
Strand (18%) (25%) (17%) (15%) (18%)
Overall
Sum
8.5 / 48 (18%)
The results are shown in table 2, where e.g. 2/11 represents ”score
2 of possible 11”. The abbreviations of the strands and the numbers
of the standards are listed in table 8 in the appendix. It turned
out that only a small number (18%) of recommended CSTA level
3A standards can be regarded as implemented in the Austrian
AHS curriculum. As the AHS curriculum is kept quite short and
generally formulated, it is difficult to compare it to the standards
that are described very precisely. However, the curriculum frequently
mentions basic skills and general concepts the students
should gain, without going too much into depth. The main focus
of AHS lies on developing abstract thinking, making use of information
technology and creating a general picture of computation
and its impacts on society (http://www.bmukk.gv.at/).
7.1.2 HTLs for Informatics and for Chemistry
Initially we will rate the implementation of the CSTA level 3A
standards in the basic/core curriculum, before rating the specialized
curricula for Chemistry and Informatics. The curricula give
rather detailed information on the required competencies and
skills and, therefore, were rather straightforward to compare with
the standards. The results of this process are depicted in table 3,
table 4 and table 5.
Table 3. Selected Standards in the Core HTL Curriculum
Stand- Strand Strand Strand Strand Strand
ard No. CT CO CP CC CG
x.1 0.5 0 0.5 0 0
x.2 0 0 0 1 0.5
x.3 0.5 0 0 1 0
x.4 0 0 0 0 1
x.5 0.5
0 0 0
x.6 0 0 0.5 0.5
x.7 0.5 0.5 0 1
x.8 0 0.5 1 0
x.9 0 0 0.5 0.5
x.10 0 0 0 0.5
x.11 0 0
0
x.12 0
Sum 2 / 11 0 / 4 1.5 / 12 4 / 10 4 / 11
Strand (18%) (0%) (13%) (40%) (36%)
Overall
Sum
11.5 / 48 (24%)
Table 4. Selected Standards in the HTL for Chemistry
Stand- Strand Strand Strand Strand Strand
ard No. CT CO CP CC CG
x.1 0.5 0 0.5 0 0
x.2 0 0 0 1 0.5
x.3 0.5 0 0 1 0

x.4 0 0 0 0 1
x.5 0.5
0 0 0
x.6 0 0 0.5 0.5
x.7 0.5 0.5 0 1
x.8 0 0.5 1 0
x.9 0 0 0.5 0.5
x.10 0 0 0 0.5
x.11 0 0
0
x.12 0
Sum 2 / 11 0 / 4 1.5 / 12 4 / 10 4 / 11
Strand (18%) (0%) (13%) (40%) (36%)
Overall
Sum
11.5 / 48 (24%)
Table 5. Selected Standards in the in the HTL for Informatics
Stand- Strand Strand Strand Strand Strand
ard No. CT CO CP CC CG
x.1 1 0 1 0.5 0
x.2 1 0.5 0 1 0.5
x.3 1 0 1 1 0
x.4 0.5 0 1 0.5 1
x.5 1
0.5 1 0
x.6 1 1 0.5 0.5
x.7 1 0.5 1 1
x.8 0.5 0.5 1 0
x.9 0.5 1 0.5 0.5
x.10 1 0 0 1
x.11 0 0.5
0
x.12 0
Sum 8.5 / 11 0.5 / 4 7 / 12 7 / 10 4.5 / 11
Strand (77%) (13%) (58%) (70%) (41%)
Overall
Sum
27.5 / 48 (57%)
As the tables 3, 4 and 5 show, there is no difference at all between
the basic/core curriculum (25%) and the specialization in Chemistry.
However, the Informatics curriculum specifies a lot more
skills and competencies in terms of computer science education
than the other two curricula do. Consequently, the degree of
implementation rises from 25% to 56%. Since Chemistry and
Informatics certainly represent two extreme points on a continuum,
one can interpolate these scores in order to estimate the implementation
percentage of an average HTL curriculum.
7.1.3 HLW
Finally we analyze the curriculum of the HLW. The results are
displayed in table 6.
Table 6. Selected Standards in the HLW Curriculum
Stand- Strand Strand Strand Strand Strand
ard No. CT CO CP CC CG
x.1 0 0 0.5 0 0
x.2 0 0.5 0 0.5 0.5
x.3 0 0 0 0.5 0
x.4 0.5 0 0 0 1
x.5 0
0 0 0.5
x.6 0.5 0 0.5 0.5
x.7 0 0 0.5 0.5
x.8 0 0 1 0
x.9 0 0 0 0
x.10 0 0 0 0.5
x.11 0 0
0
x.12 0
136
Sum
Strand
Overall
Sum
1 / 11
(9%)
0.5 / 4
(13%)
0.5 / 12
(4%)
8.5 / 48 (18%)
3 / 10
(30%)
3.5 / 11
(32%)
As table 6 shows, the implementation of the CSTA standards in
the HLW is quite poor. It only scores an overall sum of 16% over
all five strands. Similar to the AHS, programming skills are ignored
in the HLW curriculum. It mainly focuses on office- and
information-management and user skills, which students may
need for future professions in the field of economics, tourism
and/or commerce. Moreover, the HLW’s curriculum demands a
lot of general concepts and ideas on computer science from students,
which are, in some cases, quite difficult to interpret. According
to the curriculum, students should, for example, be able to
master a current operating system, which is very complex to
define. Knowing or mastering an operating system might imply to
understand the general concepts and how to use it, but might also
go far deeper into other subject matters. Personal experience and
the overall structure of the curriculum, led to a very general interpretation
of this and similar statements. Therefore, it is assumed
that certain declarations, such as the one mentioned above, do not
purport to focus on depth but on giving overviews related to
practice. However, this fact might have also diminished the results,
as the CSTA standards for level 3A, are partly a lot more
specialized and ask for in-depth knowledge as well.
7.2 Comparison of the Curricula
Following the isolated inspection of the selected curricula, it
seems interesting to compare the scores of the curricula. Table 8
clearly depicts the differences in implementation of the CSTA
standards in the respective curricula.
Table 7. CSTA Standards in Curricula Selection
Strands AHS
(1) CT
(2) CO
(3) CP
(4) CC
(5) CG
Sum
Average
2 / 11
(18%)
1 / 4
(25%)
2 / 12
(17%)
1.5 / 10
(15%)
2 / 11
(18%)
8.5 / 48
(18%)
HTL
HTL
Chemistry Informatics
2 / 11 8.5 / 11
(18%) (77%)
0 / 4
0.5 / 4
(0%)
(13%)
1.5 / 12 7 / 12
(13%) (58%)
4 / 10 7 / 10
(40%) (70%)
4 / 11 4.5 / 11
(36%) (41%)
11.5 / 48 27.5 / 48
(24%) (57%)
14 / 48
(29%)
HLW
1 / 11
(9%)
0.5 / 4
(13%)
0.5 / 12
(4%)
3 / 10
(30%)
3.5 / 11
(32%)
8.5 / 48
(18%)
Note: Since the basic/core HTL curriculum is the same as
the one for the specialization in Chemistry, only the latter
is featured in the table above.
As expected beforehand, the AHS, being the school with the most
general background and aim in education, also features the lowest
score (8.5/48). Likewise, the HLW, since specialized in business
and tourism, does not exceed this score (8.5/48). The HTL curriculum
for Chemistry, which has the same features as the basic/core
HTL curriculum, tops them by a few points (11.5/48) which might
be due to the more technical focus of the school. However, the
HTL for Informatics is the only specimen that manages to break
through the 50% barrier (27.5/48) and is thus the school type with

the highest rate of implementation of the CSTA Level 3A standards.
Calculating an average score from these four curricula gives
us a mean rating of 14/48 (29%).
Figure 5. Overview of results.
Looking at the relative adoption of the 5 strands, it is manifest
that the HTL for Informatics has percentages over 50% regarding
most of the strands except Collaboration (only 13%) and Community,
global and ethical impacts (41%). Oppositely, at the AHS
Collaboration has the highest scores and therefore seems to be
encouraged there much more. On the other hand, at the HTL for
Chemistry and the HLW, the strands Computers and Communications
Devices (CC) and Community, global and ethical impacts
(CG) produced the highest percentages, indicating the very different
focuses of these schools compared to AHS. As a whole, the
results reflect the schools’ top priorities (or profiles) that are
generally associated with them (cf. section 4), i.e. the HTL has a
very strong technical markedness, the AHS a stronger emphasis
on collaborative and social aspects, and the HLW focusses on
communication, international and global aspects.
8. CONCLUSION AND FUTURE WORK
The purpose of this paper was to demonstrate to which degree a
subset of CSTA Level 3 standards (i.e. level 3A) is implemented
in a selection of Austrian school curricula, which lead to a clear
and unambiguous result: The incorporation of CSTA standards
into Austrian school curricula is, to a very large degree, unsatisfying.
Even the HTL for informatics, with its special focus on computer
sciences, does not reach more than 57% (27.5/8) adoption
rate. This seems even more disappointing when considering that
only the Level 3A standards were compared to the Austrian curricula
and the more elaborate standards of Level 3B and 3C,
which feature more in-depth competencies, were not taken into
account at all.
However, it has to be considered that the Austrian curricula do
not go into much detail when describing the required competencies
and skills, but use rather abstract terms. This makes it quite
difficult to compare the (English) CSTA standards to the (German)
Austrian curricula since a lot of freedom, in what specifically
to teach, is still present.
Prospective future research in this area could involve a more
exhaustive elaboration of curricula from other school types or
countries. In doing so, it would be necessary to broaden the range
of curricula to be examined in the course of the rating process.
Furthermore, the rating scale/key should be redefined more precisely
in order to assure a more objective scoring. Also, the scoring
should be done by more people to ensure the objectiveness
and reliability of the whole process. Last, but not least, it seems
137
somewhat obvious that also the rest of the CSTA standards (1, 2,
3B, 3C) could/should be compared to the respective curricula of
other levels of education. To our regret, it was not possible to
elaborate on these aspects in more detail due to the limited time
and resources under which this paper has been produced.
9. REFERENCES
[1] Betts, J. 1998. The Impact of Educational Standards on the
Level and Distribution of Earnings. The American Economic
Review 88, 1, 266–275.
[2] CSTA – Computer Science Teachers Association. 2010.
Running On Empty. State-by-State Results.
http://www.acm.org/runningonempty/roemap.html. Accessed
22 June 2012.
[3] Hubwieser, P., Armoni, M., Brinda, T., Dagiene, V., Diethelm,
I., Giannakos, M. N., Knobelsdorf, M., Magenheim, J.,
Mittermeir, R., and Schubert, S. 2011. Computer science/informatics
in secondary education. In Proceedings of
the 16th annual conference reports on Innovation and technology
in computer science education - working group reports.
ITiCSE-WGR ’11. ACM, New York, NY, USA,
19‐38.
[4] Klieme, E., Avenarius, H., Blum, W., Döbrich, P., Gruber,
H., Prenzel, M., Reiss, K., Riquarts, K., Rost, J., Tenorth, H.-
E., and Vollmer, H. J. 2004. The Development of National
Educational Standards. An Expertise. Bundesministerium für
Bildung und Forschung, Berlin.
[5] Micheuz, P. 2008. Harmonization of Informatics Education -
Science Fiction or Prospective Reality? In Informatics Education
- Supporting Computational Thinking, Third International
Conference on Informatics in Secondary Schools -
Evolution and Perspectives, ISSEP 2008, Torun, Poland, July
1-4, 2008. Lecture notes in computer science. Springer,
317–326.
[6] Norcini, J. J. 2003. Setting standards on educational tests.
Medical Education 37, 5, 464‐469.
[7] Reigeluth, C. M. 1997. Educational Standards: To Standardize
or to Customize Learning? Phi Delta Kappan 78, 3, 202–
206.
[8] Tucker, A., Deek, F., Jones, J., McCowan, D., Stephenson,
C., and Verno, A. 2006. A . Final Report of the ACM K–12
Task Force Curriculum Committee, New York.
[9] Tucker, A., Ed. 2003. A . Final Report of the ACM K–12
Task Force Curriculum Committee October 2003, New
York.
[10] Tucker, A., Seehorn, D., Carey, S., Moix, D., Fuschetto, B.,
Lee, I., O’Grady-Cuniff, D., Stephenson, C., and Verno, A.
2011. CSTA K-12 Computer Science Standards. Revised
2011. CSTA Standards Task Force.
[11] Wilson, C., Sudol, L. A., Stephenson, C., and Stehlik, M.
2010. Running on Empty. Executive Summary.
http://csta.acm.org/runningonempty/fullreport.pdf. Accessed
21 June 2011.
[12] Wilson, C., Sudol, L. A., Stephenson, C., and Stehlik, M.
2010. Running on Empty.
http://csta.acm.org/runningonempty/fullreport.pdf. Accessed
21 June 2011

APPENDIX
Table 8. CSTA Level 3A Standards [10]
No. Level 3A Standards
Strand 1: Computational Thinking (CT)
1.1 Use predefined functions and parameters, classes and
methods to divide a complex problem into simpler
parts.
1.2 Describe a software development process used to solve
software problems (e.g., design, coding, testing, verification).
1.3 Explain how sequence, selection, iteration, and recursion
are building blocks of algorithms.
1.4 Compare techniques for analyzing massive data collections.
1.5 Describe the relationship between binary and hexadecimal
representations.
1.6 Analyze the representation and trade-offs among
various forms of digital information.
1.7 Describe how various types of data are stored in a
computer system.
1.8 Use modeling and simulation to represent and understand
natural phenomena.
1.9 Discuss the value of abstraction to manage problem
complexity.
1.10 Describe the concept of parallel processing as a strategy
to solve large problems.
1.11 Describe how computation shares features with art
and music by translating human intention into an
artifact.
Strand 2: Collaboration (CO)
2.1 Work in a team to design and develop a software
artifact.
2.2 Use collaborative tools to communicate with project
team members (e.g., discussion threads, wikis, blogs,
version control, etc.).
2.3 Describe how computing enhances traditional forms
and enables new forms of experience, expression,
communication, and collaboration.
2.4 Identify how collaboration influences the design and
development of software products.
Strand 3: Computing Practice and Programming (CP)
3.1 Crate and organize Web pages through the use of a
variety of web programming design tools.
3.2 Use mobile devices/emulators to design, develop, and
implement mobile computing applications.
3.3 Use various debugging and testing methods to ensure
program correctness (e.g., test cases, unit testing,
whit box, block box, integration testing)
3.4 Apply analysis, design, and implementation techniques
to solve problems (e.g., use one or more software
lifecycle models).
3.5 Use Application Program Interfaces (APIs) and libraries
to facilitate programming solutions.
3.6 Select appropriate file formats for various types and
uses of data.
3.7 Describe a variety of programming languages available
to solve problems and develop systems.
3.8 Explain the program execution process.
3.9 Explain the principles of security by examining encryption,
cryptography, and authentication techniques.
3.10 Explore a variety of careers to which computing is
138
No. Level 3A Standards
central.
3.11 Describe techniques for locating and collecting small
and large-scale data sets.
3.12 Describe how mathematical and statistical functions,
sets, and logic are used in computation.
Strand 4: Computers and Communications Devices (CC)
4.1 Describe the unique features of computers embedded
in mobile devices and vehicles (e.g., cell phones, automobiles,
airplanes).
4.2 Develop criteria for purchasing or upgrading computer
system hardware.
4.3 Describe the principal components of computer organization
(e.g., input, output, processing, and storage).
4.4 Compare various forms of input and output.
4.5 Explain the multiple levels of hardware and software
that support program execution (e.g., compilers, interpreters,
operating systems, networks).
4.6 Apply strategies for identifying and solving routine
hardware and software problems that occur in everyday
life.
4.7 Compare and contrast client-server and peer-to-peer
network strategies.
4.8 Explain the basic components of computer networks
(e.g., servers, file protection, routing, spoolers and
queues, shared resources, and fault-tolerance).
4.9 Describe how the Internet facilitates global communication.
4.10 Describe the major applications of artificial intelligence
and robotics.
Strand 5: Community, Global, and Ethical Impacts (CG)
5.1 Compare appropriate and inappropriate social networking
behaviors.
5.2 Discuss the impact of computing technology on business
and commerce (e.g., automated tracking of goods,
automated financial transactions, e-commerce, cloud
computing).
5.3 Describe the role that adaptive technology can play in
the lives of people with special needs.
5.4 Compare the positive and negative impacts of technology
on culture (e.g., social networking, delivery of
news and other public media, and intercultural communication).
5.5 Describe strategies for determining the reliability of
information found on the Internet.
5.6 Differentiate between information access and information
distribution rights.
5.7 Describe how different kinds of software licenses can
be used to share and protect intellectual property.
5.8 Discuss the social and economic implications associated
with hacking and software privacy.
5.9 Describe different ways in which software is created
and shared and their benefits and drawbacks (commercial
software, public domain software, open source
development).
5.10 Describe security and privacy issues that relate to
computer networks.
5.11 Explain the impact of the digital divide on access to
critical information.

Information Theory on Czech Grammar Schools: First
Findings
ABSTRACT
Daniel Lessner
Department of Software and Computer Science Education,
Faculty of Mathematics and Physics,
Prague, Czech Republic
lessner@ksvi.mff.cuni.cz
Computing science is not a part of standard grammar school
(secondary) education in the Czech Republic. As a part of
an effort to change this, we have developed an introductory
computing science course.
It consists of multiple month long modules. The first module
deals with the notion of information. Students are expected
to understand it sufficiently for comparing informational
gain of statements or questions. Along with binary
numeral system, they are expected to understand the relation
between data length and the amount of information
stored.
In this paper we describe the module itself and present the
results from the first run of experimental teaching. The aim
was to find out the limits of such topic on grammar schools
and suitability of chosen teaching methods.
It turned out that no matter how abstract is the term
information itself, students are capable of acquiring it deep
enough so that it would be beneficial for them.
Categories and Subject Descriptors
K.3.2 [Computers and education]: Computer and Information
Science Education—Curriculum, computer science
education
General Terms
Human factors, Experimentation
Keywords
Secondary education, computing science, information theory
1. INTRODUCTION
In this paper we first briefly discuss the unfortunate situation
of computing science (CS) education on Czech grammar
schools 1 . Then we describe a basic CS course developed with
1 General secondary education, students aged 15 – 19.
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WiPSCE ’12 Hamburg, Germany
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$10.00.
139
an intent to change this state. This course is tested with two
seminar groups. The aim of this experimental teaching is to
find out whether it is even possible to teach CS on a grammar
school level in Czechia.
We have decided to focus this paper on the first module,
which helps students study information. We describe
its specific objectives and content. The necessary aspects
such as the course organization and students characteristics
are described in a separate section. The last section finally
shows the results and findings and their interpretation.
Grammar schools in the Czech Republic are one of the
secondary education branches. The curriculum is defined in
the Framework Education Programme for Secondary General
Education (FEP) [1]. From it the individual school programmes
are derived. According to FEP, grammar schools
shall prepare students for university studies. This shall
be accomplished through development of key competences.
FEP defines six of them: to solve problems, to communicate,
to learn, civic, entrepreneurial and personal and social
competence. Classical subjects are means to develop these
key competences 2 .
We use the term computing science for the general area
of efficient (i.e. automated) information processing. It has
very little to do with digital literacy. It is also not programming.
Computer is merely the tool, not the subject.
A similar view can be seen in Computer Science Unplugged
from New Zealand [2]. It does not employ computers by its
definition, what certainly says enough about its relation to
digital literacy or programming.
The K-12 CS curriculum is not widely applied [5], but
serves as a good prototype and reference. Given a much
longer time frame, the programme can afford to be much
more complex then the other examples here. The goals are
similar to ours: to introduce basics of CS on an appropriate
level. All topics included in our course except for a few
details are covered in this curriculum.
FEP casually mentions a few CS topics. It includes so
called informatics 3 as a compulsory subject, which is focused
on using digital technology. There is no systematic approach
to efficient information processing.However, considering the
grammar school educational goals and the concept of key
competences, CS seems to be very promising. Moreover, it
2 Yet they are not unimportant – to communicate we need
languages, to solve problem sciences, to be good citizen history
etc.
3 Informatics in Czech is an umbrella term, which can mean
anything related to computers or digital technology, depending
on the current context.

is a part of general knowledge. CS is shaping our world
significantly and most of grammar school graduates are not
even aware of its existence.
We consider the absence of CS in grammar schools a mistake.
We believe we are missing here a fruitful opportunity
to achieve secondary education goals easier. In an imaginary
ideal case, CS is a regular school subject, such as physics or
history. More thorough reasons are given in [3]. We intend
to develop, carry out and evaluate a basic CS course to have
a solid base for a possible future discussion about including
CS into grammar schools.
This paper presents results from one course module. The
question is, whether it is possible to teach information and
to what extent. We have planned the module (objectives,
content, activities, assignments, final test) and carried it out
with two seminar groups.
2. INTRODUCTORY CS COURSE
In an effort to move towards our above mentioned ideas,
we have developed a program for a basic CS course. It covers
what we think any grammar school student should know.
General goals of the course are derived from secondary education
goals. We must enhance students problem solving
skills along with other key competences and thus prepare
students for studying university, for their career and for a
life in our society [1].
Further, as with other sciences, the aim of their study
on grammar schools is to get familiar with their subjects,
methods, connections to other areas, fundamental results
and their implications and applications. Last but not least,
students shall be aware of some fundamental limits to CS,
such as the existence of practically or computationally unsolvable
problems. Students shall be able to actually use the
basics of CS in everyday life.
Based on the usual situation on our grammar schools, we
have planned the course to last one school year, 90 minutes
a week. It is structured into approximately one month (four
lessons) lasting modules. Each module deals explicitly with
one fundamental idea, while other ideas are present in the
background. The main topics chosen for individual modules
are information, graph, problem, algorithm and efficiency.
More detailed information can be found in [4].
3. INFORMATION: MODULE PLAN
Information is the first module out of 8 in the course plan.
The word informatics used in Czech is based on the word information,
so it seems logical to deal with it in the beginning
of the course. But most importantly, this module contains
many notions that will come later (such as algorithm, efficiency
or recursion) in a simple and conceivable form.
3.1 Objectives
The basic objective to achieve is understanding of the term
information on an appropriate level and understanding of
what is it good for.Along with the term information itself,
there are others to be understood: communication. message,
data, knowledge, encoding, language, reliability, bit.
Further, students shall be able to calculate the amount of
information contained in a message.
By a step higher is efficient querying in a given situation.
This goes along with the concept of halving the set
of possibilities. Students shall understand the concept of
140
decision tree, use it properly, judge its efficiency and optimize
it in simple cases (including non-uniform distribution
of answers)Students shall be able to construct binary codes
for objects and work correctly with the relation between the
number of codes and their length.
As the CS course opens with this module, we use it to prepare
and develop concepts for upcoming topics, such as decomposition,
algorithm (a reliable workflow of simple steps,
e.g. a decision tree) or efficiency (related to defined criteria).
3.2 Process
Here we describe briefly the educational activities and related
content included in the module. They result from the
above discussed objectives and methods. There is some slack
to vary the activities between each study group.
Students are at first introduced to the course and to the
content and objectives of the first module. Then the teacher
helps them realize the importance of the term information
and the fact that they actually can not explain what it is.
They are asked to search their resources and come up with
an explanation themselves. Discussing their various results
and opinions, key features and links emerge.
Guessing games
To make things less abstract and vague, the teacher proposes
an animal guessing game. He thinks of an animal,
students ask him yes or no questions in an effort to find the
animal as fast as possible. Students intuitively review the efficiency
of each question after the game and formulate their
first hypotheses on the optimal strategy. Frequent finding
is for example that direct guessing is useless in the beginning,
negated ”reassuring“ question is useless completely in
noise-free environment, making the teacher answer yes may
be emotionally satisfying, though it is meaningless for the
game.
However, animals turn out to be a rather too complex
model. They are not a well bounded domain, success in
the game depends on some knowledge of animals, and it is
surprisingly tricky to answer some questions (e.g. elephants
do live in Africa, as well as in India, and in captivity all over
the world).
That is why we switch to a much better arranged domain –
natural numbers. They still model the situation well enough
to give us important results. Numbers allow students to
clear their minds from animals and focus on the problem
itself. The task is to formulate a good strategy, i.e. find
a set of rules to create efficient questions. Some heuristics
emerge eventually in most of the pairs. Then we play one
more game together so that each pair can see their results
in practice.
Decision trees and information definitions
One of the ways to describe a strategy is a decision tree
scheme. It shows the efficiency also visually. Looking at
a tree scheme, one may work out the formula to the relate
number of possibilities and and number of questions asked
to isolate one. This finally allow us to quantify information.
Considering the scheme, it is practical to measure uncertainty
by number of questions necessary to remove it, i.e.
⌈log(|S|)⌉, where S denotes the set of possible outcomes.
Base of the logarithm is two in case of binary questions.
Amount of information is then the decrease of such uncertainty
after getting the answer. More formally, the amount

of information in a message M, represented as a set of eliminated
outcomes, is I(M) = log(|S|) − log(|S \ M|). This
approach is a special case of the well known approach using
entropy, using uniform probability distribution. This makes
the basic idea of reduction of uncertainty more accessible as
reduction of possible outcomes. This simplified definition is
sufficiently useful to lead to the main ideas behind.
The second definition of information we use says that information
consists of data and their interpretation. This is
a fruitful starting point to further observations too. It is in
accordance with students intuitive concepts. It can be illustrated
using the decision tree scheme too. Another closer
look to the decision tree for guessing numbers reveals the
tight connection to binary numbers. It also opens the opportunity
to think of an informational value of an individual
character, e.g. a bit.
Students realize sooner or later that not every animal have
the same chance to take part in the game. The teacher
provides them with some more examples where non-uniform
distribution is important, e.g. assessing stroke risk factor
or other medical diagnosis tasks. This topic leands to a
fundamental heuristic: what comes often shall be at hand.
Additional activities
The objectives of this module are covered with activities
above. Below suggested activities are optional and utilized
mostly based on students interests. They serve to review the
matter and to investigate it from different points of view.
They also imply some real world applications. Not all students
in the group must work on the same task of these.
The guessing game can be extended for real numbers. It
of course deepens the understanding of mathematics. For
CS, it opens the question of a process finiteness. Another
domain for guessing game may be subsets of a given finite
set. Halving the possibilities leads to the simplest encoding
by characteristic function. And again, students reach deeper
understanding of sets and subsets, which will be useful later
in the course. An even more advanced strategy is required
to play mastermind.
To remind students that binary questions are just a convention,
we let them solve problems with balance scale. One
kind of problem is to find a heavier item in otherwise indistinguishable
group. Another kind is about designing somehow
optimal set of masses.
Popular TV shows, such as Who Wants to Be a Millionaire?
can be a good resource. Students can think of heuristics
on which kind of help shall be used in which situation. It
is easier to explain these heuristics using the new knowledge
of information theory. A little bit similar and very openly
defined project deals with designing cheat sheets, where efficiency
is of course crucial.
Efficiency of information encoding in natural languages
can be investigated by comparing same texts in different
languages. Another question of this field is the efficiency
of Morse code for different languages, and to find a way to
enhance it, if possible. Also DNA encoding amino acids is
an interesting system to be investigated from the informational
point of view. The last one to mention is the START
method (Simple triage and rapid treatment) utilized in mass
disasters.
3.3 Feedback and evaluation
Students work mostly in groups, what allows them to re-
141
ceive continuous feedback on their thoughts from their peers
as well as from other groups. They also consult the teacher
when needed during their work, what allows him to provide
some more informal feedback. Another source of feedback
are submitted home assignments, commented by the teacher.
They are usually problematic or require some more intensive
work, which can be done outside the classroom.
Students write a test at the end of each module. These
tests are rather difficult. The matter itself is quite abstract,
the tests last the whole lesson (90 minutes) and contain some
unusual tasks. Some aspects on the other hand make the
tests easier. The required skills are stated very clearly in
advance. Further, students may use their lecture notes and
internet resources.
4. EXPERIMENTAL LESSONS
Here we describe the actual teaching to allow the reader to
better understand the below presented results. We worked
with two groups, denoted A and B. Both groups are of 12
students. Majority of students does not intend to continue
their carrier in CS anyhow. Their motivation is only to
pass the subject as effortlessly as possible. Almost a half of
them have signed up for the seminar as for ”the best of bad
options“, because they just had to pick one. This fact of
course hinders teaching CS. On the other hand, it increases
the validity of our research. Our target is a regular grammar
school student. The less interested students we work with,
the more valuable results we obtain.
All twelve students from group A are in their last year.
Only one of them intends to pass the school-leaving exam
in informatics. Their school specializes on sports. This determines
their attitude to sciences well enough (and again,
it increases the value of our results). They all have been
through a one year lasting optional seminar on programming
in Pascal. The reason they took the seminar was the
usual, i.e. picking ”the best of bad options“ and the result
corresponds this attitude.
Group B is on an average grammar school with no specific
specialization. Four students are in their last year and
they all intend to pass the school-leaving exam in informatics.
The others (eight students) are one year younger and
their decision is still to be made. The students have various
programming experience.
The module on information took six weeks (90 minutes
teaching time each) instead of the originally planned four.
This is mostly due to building in the necessary mathematics
reviews (e.g. numeral systems and logarithms).
5. RESULTS
Students accepted the module surprisingly well (considering
it being quite deep and abstract and out of the official
curriculum). Some of them realized the importance of the
topic and were glad to understand it better, some were challenged
by the proposed problems.
5.1 Observation during lessons
An unexpectedly fruitful discussion arose from ”Is it even?“
question during numbers guessing. It is informatically optimal.
Some students therefore insist to use it as the first
question. Some realize that they can ask it also at the end
in a simpler way, when they are choosing between the last
two alternatives. This shows an important fact: difficulty

of answers management is important in practice. The question
does not allow us to construct consequent questions
straightforwardly with the same principle. Yet, after being
reminded of binary system, students are able to realize the
connection. They are doing almost the same as bisecting,
only reversed. The value of these thoughts lay in the fact
that they were driven mostly by students themselves.
Disturbing (however expected) finding is low level of understanding
of certain parts of mathematics among students.
Logarithms are taught intensively, yet no (!) student have
seen a direction to calculate the depth of a tree. Our guidance
had to be very explicit. A few months later, most of
the students actively knew that it is related to exponential
growth and therefore also to logarithms.
Some of the ideas encountered in the first module were
needed again during the school year. More than a half of
students (in both groups) recalled them actively. This includes
bisection, decision trees, binary numeral system and
efficiency modeled using some kind of elementary step.
5.2 Test results
Further explanations and remarks are based on collected
test sheets examination and our own notes from the lessons
and consecutive discussions with students. We have not used
any formal pre-test. It was more than clear from the initial
discussions held for this purpose that students had difficulties
to even understand the given questions. Student can
get a certain amount of points for each task. Furhther given
percentages indicate the average ratio of acquired points,
illustrating the success on each task.
Brief explanation.
Students were to explain briefly the meaning of the term
information. Any meaningful answer was rewarded with
points. Definition or equally complete description got all the
points, less precise expressions or examples got only partial
score. The resulting success is lower than expected, given
that it is a question of knowledge and outer resources were
available. We see the reason in the lack of communication
skills. Most of the students understood the term well, as we
have seen during other tasks. To express it was however too
difficult.
Information and logarithm.
This was a double task. First was to determine how many
yes/no questions does one need to reliably identify a Czech
deputy (they are 200). The second was to determine the
minimal sufficient length for their hypothetical binary ID
code. Very few students have used logarithm for any of the
tasks. Majority of them chose to simulate the known process
(sequential halving and doubling, respectively. Only two
students have realized that the result is the same for both
questions and the second calculation is not needed.
Suboptimal decision tree.
There was a small (7 leaves, 4 levels) decision tree with
given rates per leaf. First task was to calculate the expected
number of questions asked. The second was to lower
that number by adjusting the tree.Those who made some
calculations usually got both parts right. Less than a quarter
of students (in both groups) passed the calculation, as
it was still too mysterious for them. The bigger surprise is
then that some of them understood the situation well enough
142
to optimize the tree (swapping frequent nodes towards the
root).
Inquiries comparison.
Students were to sort seven given questions (such as “Is it
a prime?”) according to the expected amount of information
provided by the answers. All the questions were bound to
the same situation, identifying an integer number between
0 and 255. This task was very dependent on mathematical
skills, therefore the results are rather poor.
Statements comparison.
This last task was not included in the same test with the
others. We put it into another test six months later. There
were four different smiley-like crime suspect faces and four
statements regarding the offender, such as “The offender has
no hair”. The task was to sort these statements according
to the amount of information included. Unlike with the
previous task, here it was obvious enough that those who had
a mistake in their resulted did not understand the matter
at all. Hence we gave all the points or none. The success
rates were unexpectedly high, considering the time passed.
We can conclude that the educational results hold well.
6. CONCLUSION
Based on the results above, we may conclude that students
are capable of understanding the notion of information and
its basic applications. The goals of the module were set
surprisingly close to the real need. The limiting factors lay
mostly in students mathematics skills. The teaching methods
are found useful and efficient by the students, although
they have recommended some tweaks, mostly in activities
distribution in time.
We described the concept of an introductory CS course
for Czech grammar schools and gave more detailed specification
about the module about information. Then we could
proceed to the results obtained during the first testing year
of the course.
We intended to find out whether the module is feasible in
the given constraints and conditions. The result shown here
is that the proposed concept works, although it needs a few
adjustments. The most distinctive feature of the proposed
course is the shift in goals from CS itself to key competences
and connections to other grammar school subjects.
7. REFERENCES
[1] Framework Education Programme for Secondary
General Education (Grammar Schools). V´yzkumn´y
ústav pedagogick´y v Praze, 2007.
[2] M. R. Fellows, T. Bell, and I. Witten. Computer
Science Unplugged... offline activities and games for all
ages: Original Activities Book. Computer Science
Unplugged, 1996.
[3] J. Hromkovič and B. Steffen. Why teaching informatics
in schools is as important as teaching mathematics and
natural sciences. In Proceedings of ISSEP, ISSEP’11,
pages 21–30, Berlin, Heidelberg, 2011. Springer-Verlag.
[4] D. Lessner. Computer science curriculum proposal for
czech grammar schools. In Zborník príspevkov, ITAT
2011, pages 99–104, Seňa, Slovakia, 2011. PONT s. r. o.
[5] A. Verno. Supporting k-12 computer science education.
J. Comput. Sci. Coll., 23(3):145–146, jan 2008.

Data modeling and database systems
as part of general education in CSE
Claudia Strödter
University of Jena
Didactics of mathematics and computer science
Ernst-Abbe-Platz 2
07743 Jena
0049/3641946202
claudia.stroedter@uni-jena.de
ABSTRACT
Everyday we unconsciously send queries and actions to databases.
The used applications are part of school and working life and in
specific areas of leisure. In order to teach a responsible handling
with such applications it will be necessary that “data modeling
and database systems” will be a part of computer science
education. By analyzing selected teaching material it is
investigated if and how it is possible to teach this topic as a
general education. The study shows that it´s possible to teach
general contents and processes with selected material.
Furthermore, a first proposal for a grading of the topic “data
modeling and database systems” reveals.
Categories and Subject Descriptors
K3.2 [Computers & Education]: Computer and Information
Science Education – computer science education, information
systems education.
General Terms
Documentation, Experimentation, Human Factors, Theory,
Standardization, Verification
Keywords
teaching in heterogeneous classes, education standards in lower
secondary education, curricular aspects, didactical approaches,
data modeling and database systems
1. MOTIVATION
Computing systems have influence on several parts of our
everyday life. Activities in our working life and in specific areas
of leisure (e.g. credit card payment, shopping, seeing medical
professionals) are supported by informatics applications.
Unconsciously we initiate several procedures to read and process
data and send queries to databases. Young people or students
spend much time a day using the internet. In addition, they use
smartphones, chip cards, and several vending machines. In sum,
the handling of computing systems is an important element in
young people’s life. Mostly, they are unaware of the associated
informatics contents and processes. In order to ensure a
responsible use of computing systems and personal data this
knowledge should be a part of school education, especially of
computer science education (CSE). Therefore, the CSTA K-12
Standards for Computer Science (CS) focus on the following
goals. Students should:
143
(1) understand the nature of CS and its place in the modern
world,
(2) understand that CS interweaves concepts and skills,
(3) be able to use CS skills (especially computational thinking)
in their problem-solving activities in other subjects. […] (see
[4] page 7).
Similar demands are used in the basic experiences by Fothe.
“CSE is a general education that allows students the following
basic experiences:
(1) to discover and to understand computing systems
(informatics systems) in several parts of life,
(2) to recognize, that done or planned activities can be formulated
as an algorithm and transferred into a computer
program, that modeling allows to transfer parts of reality into
a computer-compatible form and that computing systems
(informatics systems) are designed by humans,
(3) to acquire problem-solving skills in dealing with exercises,
which are useful at school and real life (“modeling
techniques” means as techniques of mental work).
These experiences are connected to each other in several ways. 1 ”
[6] In summary, to understand CSE as a general education it is
necessary to understand general informatics contents and
processes and to use this knowledge in everyday life.
As mentioned earlier many activities in our everyday life access
databases. In order to ensure a general education the topic “data
modeling and database systems” should be a part of CSE. In the
CSTA K-12 Standards for CS is required that students should be
able to select appropriate database formats for a particular
computational problem and to form abstract ideas about specific
components (e.g., input, output, processors, and databases) and
their roles in the computational spectrum (see [4] page 11). There
exist many ideas and concepts to teach “data modeling and
database systems” for university students. Other concepts focus
only on the use of specific database tools. Antonitsch uses another
approach. He suggests to develop structure awareness by starting
with analyzing and querying ready-to-use databases provided by
the teacher [1].
Nevertheless, it remains unclear which general informatics
contents and processes are necessary to teach “data modeling and
database systems”. In German school this topic is a part of CSE in
lower secondary education. The objective of this article is to
provide criteria for exercises (informatics contents and processes,
requirement areas) in order to teach “data modeling and database
systems” in CSE in lower secondary education. In this study,
existing teaching material is used to get an idea of teaching this
topic as a general education.
1 All translations (German – English) were done by the author.

2. DATA MODELING AND DATABASE
SYSTEMS IN GERMAN SCHOOLS
In consequence of the results of international comparative studies
(e.g. International student assessment - PISA 2000) in Germany
the “Principles and Standards for School Informatics –
Educational Standards of Informatics in Lower Secondary
Education” (translation in accordance with: [8], abbreviated:
educational standards of informatics SI) were developed. These
are output-orientated minimum standards and specify the
competencies which should be acquired when completing 7th and
10th grade. They consist of process standards and content
standards (see Figure 1) and were proposed to bridge the
knowledge about information and communication technology
(ICT) and CS (see [3]). Therefore, in this study these competencies
are used to evaluate exercises.
Figure 1. Process and content standards 2
“Data modeling and database systems” influences several process
and content standards. For this study it is reasonable to focus on a
few main competencies. Most contents and processes associated
with “data modeling and database systems” are summarized in the
content standard Information and data and the process standard
Model and implementation (see table 1).
Table 1. Associated competencies of the “educational
standards of informatics SI”
Standards Competencies
Content
Standard 3
Information
and data
Process
Standard
Model and
implementation
Learners of all age should …
…understand the connection between data and information
and different forms of representations of data,
…understand operations on data, and interpret these
with regard to the represented information,
…be able to trigger suitable operations on data.
Learners of all age should …
…create informatics models based on given facts,
…implement models with suitable applications,
…reflect models and their implementation.
Currently, in Germany no educational standards for upper
secondary education of informatics exist. The “Einheitliche
Prüfungsanforderungen Informatik” [9] (translation: standardized
requirements for final examinations in CSE, abbreviated: EPA)
were developed to ensure transparency, comparability, and
standardization of final examinations. They define examination
areas and assist in compiling exercises (especially requirement
areas). “Data modeling and database systems” is included in all
examination areas. The requirement areas are often used for
grading of competencies and exercises (e.g. see [10]). In this
2 [8], translated by [3].
3 The translations of the content standards are taken over by [3].
144
study the levels reproduction performance, transfer performance
and constructive performance are also used to evaluate the
cognitive complexity of exercises. On the first level students are
able to account known facts, to explain and present methods and
principles of CS. Transfer performances allow the independent
use of known facts, methods and principals in order to solve new
problems. Learners on the last level independently decide on
methods in new and complex problem situations (see [12]).
The “educational standards of informatics SI” and the EPA are
structured in several ways. Nevertheless, it is possible to assign
the competencies and contents of the EPA to the process
standards and content standards of the “educational standards of
informatics SI” [7]. Therefore, in this study parts of both
educational principals are used to evaluate exercises.
3. RESEARCH QUESTION AND DESIGN
Education is influenced by many details. There are, for example:
the teaching methods, the scientific contents and processes, the
asked questions, and the used exercises. This study focusses on
the contents and processes in exercises.
Material
By comparing five recent German schoolbooks, it was shown that
the topic “data modeling and database systems” is divided into
exercises about storage of large data volumes, data modeling,
implementation and queries. Developing schoolbooks is a longterm
process, in which the authors’ entire expert-knowledge about
education and science is used and discussed in several ways.
Therefore, it is assumed that exercises in schoolbooks are typical
exercises. This selection enables a nonreactive research, which
excludes experimenter effects, test or other answer falsifications.
In this study, 61 exercises taken out of one frequently used
schoolbook (see [5]) for lower secondary education are analyzed.
The book is permitted in all funeral states of Germany and allows
to achieve the existing curricula. The chosen exercises include the
whole topic “data modeling and database systems” and are similar
to the exercises in the other four schoolbooks.
Analysis method
To analyze the teaching material the qualitative content analysis
by Mayring [11], more precisely the structuring analysis is used.
Table 2. Schematic representation of the analysis method
Operation Explanation
Selection of material 61 exercises taken from one school book
Determination of
analyzing units
Deductive development
of categories
Ensuring of the interrater
reliability
One exercise is one unit
Developing of categories, anchors, and
coding rules
The categories were discussed with
members of the department of didactics
of mathematics and computer science.
Rating All exercises were rated by two Raters.
Interpretation Structuring analysis
In order to provide criteria for exercises in “data modeling and
database systems”, the following questions are posed to the
material:
Question 1: Which contents and processes are necessary to
achieve the competencies of the “educational standards of
informatics SI”?
The “educational standards of informatics SI” allow an outputorientated
teaching, simultaneously they were proposed to bridge

the knowledge about ICT and CS (see [3]). CSE as a general
education in German school should comply with these standards.
Question 2: Which contents and processes are used to teach “data
modeling and database systems” according to the individual level
of students (means as different existing knowledge)?
The knowledge about handling computing systems and the
experiences students have in their everyday life are mostly
different. In order to understand general informatics contents and
processes and to use this knowledge in everyday life it is
necessary to use exercises according to the individual level of
students. Doing so, it will be also possible to combine the already
existing knowledge about handling computing systems with
knowledge about general informatics contents and processes.
Category systems
The first question focusses on the “educational standards of
informatics SI”. The competencies relating to “data modeling and
database systems” are used as categories to evaluate the exercises
(see table 3, CS1-PS3). It is investigated which informatics
contents and processes enable the achievement of the competencies.
The presented competencies are connected to each other in
several ways. It was possible to assign more than one category for
each exercise.
The individual level of students can be considered in different
ways (e.g. information representation, teaching arrangement, time
to work on exercises, context of exercises, complexity of
exercises). By using the method applied in this study, it is
possible to investigate the cognitive complexity 4 . In order to
evaluate the exercises three categories (see table 3, ABI-ABIII)
are used. Only one category can be assigned for each exercise.
Table 3. Category systems for question 1 and 2
Code Category and Anchor
Learners understand the connection between data and
information and different forms of representations of data.
CS1 “Choose a problem area where large data volumes are
stored and connected to the internet. What information on
objects, persons, or events are stored in the database?”
Learners understand operations on data, and interpret
these with regard to the represented information.
CS2 “Send the following described queries to the given
database mail-order business.
Give an overview of all clients of a specific city.”
Learners are able to trigger suitable operations on data.
“The following SQL statement is given
CS3 SELECT Artikel.ANr, Artikel.Bezeichnung, Lieferant.LNr,
Lieferant.Name FROM Artikel, Lieferant;
Send a query to the database and interpret the results.”
Learners create informatics models based on given facts.
“Decide which columns are necessary to acquire data in
PS1
the table CD. Complete the current schema with the new
columns.”
Learners implement models with suitable applications.
PS2 “Transfer your current data model into a relational model.
Use the necessary rules and specify the developed tables.”
Learners reflect models and their implementation.
“Check your database mail-order business in accordance
PS3
with the given rules. If necessary, optimize your
database.”
4 Cognitive complexity involves the requirement, the level of
difficultly etc. (see [2]).
145
Reproduction performance
ABI “Which information can you receive from the tables in the
database CD? Formulate five questions.”
Transfer performance
“In the last exercise you verbally formulated possible
ABII queries to your database. Describe your queries as
selection, projection or as a combination of both. Express
your queries as SQL-statements.”
Constructive performance
ABIII “Does your database mail-order business guarantee
referential integrity? If necessary, update your database.”
Rating
All exercises were evaluated by one teacher and one student
teacher of CSE. For all categories the inter-rater reliability was
higher than 0.65 (Cohen-Kappa). To get an idea of the rating
method an example for rating an exercise is given first:
Exercise: “A part of the data model CD rental is given. Find other
useful classes/entities for this data model. Complete the data
model with appropriate attributes. Define the cardinality of the
relationships.”
Rating of Rater 1 (n1) and 2 (n2): PS1, PS2, CS1, ABII
All equal rated exercises (N=45) were used to receive a common
grading of the content and process competencies. The grading is
presented in the next chapter.
4. RESULTS
The “educational standards of informatics S1” are developed as
minimum standards. Nevertheless, there are informatics contents
and processes above and below these standards. The answers of
question 1 and 2 emphasize contents and processes to achieve the
standards in three requirement areas.
In order to get an overview of the used material some quantitative
data (see table 4) are given first.
Table 4. Quantitative data of the analysis
Rater CS1 CS2 CS3 PS1 PS2 PS3 ABI ABII ABIII
n1 30 30 26 8 6 21 13 32 16
n2 31 29 24 8 6 23 11 31 19
In regard to answering the first part (contents) of question 1 a
summary of the content standards in exercises is shown in table 5.
Combining the answers of question 1 and 2 results in a grading of
the content competencies (see table 5). In sum, it is shown, that:
the handling with the database systems turns from simple
usage of the graphical user interface to formal statements,
the databases and queries turn from given databases and
queries to independently developed databases and queries,
the use of databases turns from simple usage to evaluated and
improved usage.
However, there are only a few exercises concerning process
standard 1 and 2. It is possible to compile a summary of process
standards in exercises. Combining the answers of question 1 and 2
results in a grading of the process competencies (see table 5). In
sum, it is shown, that:
data modeling turns from modeling some elements of data
worlds to modeling complete models of data worlds,
the transfer of data models into database systems turns from
the transfer of given tables and relationships to the transfer of
independently created data models,
the use of data models turns from simple usage to evaluated
and improved usage.

Table 5. Summarized results of question 1 and 2
Content competencies in exercises
In a given database learners...
…insert and modify data by using the graphical user
interface,
…realize the used classes/entities and attributes,
ABI
…create forms by using the graphical user interface,
…create possible queries as questions,
…use simple operations (selection, sorting, summarizing)
by using the graphical user interface.
In a given database learners…
…insert and modify data by using formal statements,
…expand the database/data model,
…define useful attributes and data types,
…define, set, delete, and evaluate foreign and primary
ABII keys,
…define the cardinality of given relationships,
…create formal statements (selection, projection,
combination of both) to given questions,
…read, run, interpret, evaluate, and improve given formal
statements (selection, projection, combination of both).
Learners...
…define referential integrity,
…check and improve the databases in compliance with
simplified rules,
…formulate and define independent useful queries,
ABIII
…independently create useful formal statements
(selection, projection, combination of both, inner join),
…read, run, interpret, evaluate, and improve formal
statements (selection, projection, combination of both,
inner join).
Process competencies in exercises
Learners...
…find useful attributes for given classes/entities,
ABI
…transfer given tables and relationships into database
systems by using the graphical user interface.
Learners...
…analyze and evaluate data worlds to find classes/entities
and attributes for a data model,
…create relationships between given classes/entities and
define cardinalities,
ABII
…transfer simple data models into relational models and
in a database management system (one-to-one
relationship) in compliance with transfer rules,
…evaluate simple data models (one-to-one relationship)
according to the given problem in compliance with rules.
Learners…
…independently create complete data models in
compliance with rules,
ABIII …evaluate data models according to the given problem,
…transfer complete data models (one-to-n and n-to-n
relationship) into relational models and in a database
management system in compliance with transfer rules.
5. CONCLUSION
With the use of recent teaching material in some parts it is
possible to realize the requirements in chapter 1. Initial statements
on criteria (contents and processes, requirement areas) for
146
exercises are emphasized in this study. Important contents and
processes for teaching “data modeling and database systems” as
general education are mentioned in chapter four. It is possible to
reveal a first grading which allows to achieve the “educational
standards of informatics SI” in several ways. Considering these
results it should be possible to turn the knowledge about “data
modeling and database systems” from daily handling knowledge
to independently evaluated and improved knowledge. With the
investigated requirement areas the individual level of students can
be included. This article presents a first proposal on criteria for
teaching “data modeling and database systems” as part of general
education in CSE. To complete and check the mentioned contents,
processes and the grading these criteria need to be evaluated with
further empirical studies.
6. REFERENCES
[1] Antonitsch, P. 2006. Databases as a tool of general education.
In Proceedings of the 2006 international conference on Informatics
in Secondary Schools - Evolution and Perspectives: the
Bridge between Using and Understanding Computer ISSEP
2006. Springer. New York.
[2] Baumann, R. 2008. Probleme der Aufgabenkonstruktion
gemäß Bildungsstandards. In LOG IN Heft 153, 54-58.
[3] Brinda, T.; Puhlmann, H.; Schulte, C. 2009. Bridging ICT and
CS - Educational Standards for Computer Science in Lower
Secondary Education. IN Proceedings of the 14th Annual
SIGCSE Conference on Innovation and Technology in
Computer Science Education, ITiCSE 2009. URL: http://
www.inf.fu-berlin.de/inst/ag-ddi/docs/bridgingICTandCS.pdf.
[4] CSTA Standards Task Force. 2011. K-12 Computer Science
Standards – revised 2011. URL: http://www.csta.acm.org/
Curriculum/sub/CurrFiles/CSTA_K-12_CSS.pdf.
[5] Engelmann, L. 2010. Informatik SI – Informatische
Grundbildung. Duden Paetec. Berlin.
[6] Fothe, M. 2012. Modellieren, Programmieren, genetisches
Prinzip und Informatik im Kontext - Was passt da wie zusammen?
Lectured on GI-FIIB Berlin 2012, URL:
http://hyfisch.de/Fachgruppe/tagung11/fothe.
[7] Fothe, M. 2008. Bildungsstandards Informatik für die
Sekundarstufe II – Vorüberlegungen zur Entwicklung. In
Proceedings of Didaktik der Informatik – Aktuelle Forschungsergebnisse.
URL: http://subs.emis.de/LNI/
Proceedings/Proceedings135/gi-proc-135-010.pdf.
[8] German Informatics society (GI) 2008. Grundsätze und
Standards für die Informatik in der Schule. Bildungsstandards
Informatik für die Sekundarstufe I. URL:
http://www.informatikstandards.de/.
[9] KMK 2004. Einheitliche Prüfungsanforderungen in der
Abiturprüfung Informatik. URL: http://www.kmk.org/
fileadmin/veroeffentlichungen_beschluesse/1989/1989_12_01
_EPA_Informatik.pdf.
[10] Kollee, C. et al. 2009. Computer Science Education and Key
Competencies. In Proceedings of 9th IFIP World Conference
on Computers in Education - WCCE 2009. URL:
http://www.die.informatik.uni-siegen.de/e-publikationen/
Publikationen/2009/WCCE2009_pap147.pdf.
[11] Mayring, P. 2003. Qualitative Inhaltsanalyse – Grundlagen
und Techniken. Beltz-Verlag, Weinheim.
[12] Nelles, W. et al. 2010. Entwicklung eines Kompetenzrahmenmodells
– Informatische Modellieren und Systemverständinis.
In: Informatik-Spektrum 33 (2010) 1, 45-35.
URL: www.springerlink.com/content/128182n611u26557/
fulltext.pdf.

The Mindstorm Effect: A Gender Analysis on the Influence
of LEGO Mindstorms in Computer Science Education ∗
ABSTRACT
Catherine Ball Faron Moller
Director, Technocamps †
In the UK, as elsewhere, the number of students choosing to
study computer science is declining. This is especially true
with female students. This paper explores the effectiveness
of LEGO Mindstorms in a pedagogic context in education
and its ability to attract female students to computer science.
A mixed methods approach was used in this study, in
which we looked at the FIRST LEGO League competition
and how female students participate in these.
The results demonstrate that, while young people enjoy
using Mindstorms, they do little to influence young people
to consider computer science education. They can, however,
be used effectively as an opportunity to engage young people
in further computing skills and computational thinking. The
lessons we learn from this research indicate that we need to
use these tools as a foundation rather than as a solution to
the problem for attracting more women into computing.
1. INTRODUCTION
Computer Science is an ever-growing field, however there is
a noticeable lack of women choosing to study Computer Science
or to enter the IT industry. There are many reasons
that have been suggested for this. For instance it may be
perceived to be too geeky or just for men. There have been
various initiatives and tools proposed to help change the way
that computing is perceived. One approach is through the
use of LEGO Mindstorms. This has been used in several
instances to change people’s perception of computer science
and ICT in attempts to show people that these fields can be
fun and exciting. Mindstorms can help teach children and
young adults the importance of teamwork and problem solv-
∗ This abstract is based on the first author’s BSc dissertation,
Department of Computer Science, Swansea University, 2012.
† Technocamps (www.technocamps.com) is a school outreach
programme started in 2004 at Swansea University. In 2011,
a three-year, £6 million EU project allowed Technocamps
to expand its operation throughout Wales, with hubs at the
Universities of Aberystwyth, Bangor and Glamorgan.
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WiPSCE 2012, Hamburg Germany
Copyright 2012 ACM ...$15.00.
Swansea University
f.g.moller@swansea.ac.uk
147
Reena Pau
ing and can encourage users to be creative and innovative
in their design.
2. LITERATURE REVIEW
2.1 Women in Computing in the UK
There are many reasons why women choose not to study or
work in the technology sector. This has been reported as a
weakening in the country’s economic position. The decline is
apparent already at GCSE level (age 14-16), where there are
ever fewer girls who choose to do a full course in ICT. The
decline then continues right through to industry. Although
there have been major national efforts to encourage females
to study computing, the numbers remain low.
2.2 Status of CSE in the UK
Part of the above problem lies in the ICT national curriculum,
and to this end in January 2012 it was announced that
the ICT curriculum was to be dropped in September of the
same year, and instead Computer Science would eventually
be taught [2]. This has been supported by several industry
experts and it is hoped that the new curriculum will give
schools the freedom to create their own ICT and Computer
Science curricula. Companies such as Google and Microsoft
are currently working together and with technology education
organisations, such as the British Computer Society
(BCS), to produce free educational material for schools. In
order to encourage young people to study computing, tools
such as Mindstorms are being used in various initiatives to
enthuse young people about technology and creativity.
2.3 The use of LEGO Mindstorms in schools
Mindstorms kits are used in schools to provide an exciting
way for children and young adults to learn about technology.
They teach students various skills including teamwork
and programming and can potentially change the way ICT
is viewed. The teaching of ICT is often given as a reason for
females being disinterested in pursuing ICT or computing
in further education [3] because many females find it boring
[6]. FIRST LEGO League (FLL) is a worldwide competition
in which teams of 11-16 year olds design and program
autonomous robots to complete tasks. With over 19,000
teams in over 55 countries, FLL is constantly expanding.
This type of initiative gives something like Mindstorms a
purpose: pupils have something to aim for.
In this study we looked at Mindstorms and whether it
can attract females to the field of computing. From previous
related studies, it is known that girls are willing to

engage in robotic activities [4]; one can no longer assume
that robots are only for males. However, existing studies of
the effectiveness of Mindstorms in education (eg, [1, 5]) tend
to concentrate on university students.
3. METHODOLOGY
A questionnaire was given to students who were taking part
in a local version of the FLL; 93 questionnaires were returned.
The questionnaire had four categories: Demographics,
LEGO Mindstorms, Career Choices and Educational
Choices. These were devised in order to gather data about:
how long the students had been using Mindstorms; if the
students felt they had learnt any new skills by using them;
and if this usage had influenced their education and career
choices. These questions were asked in order to understand
if and how Mindstorms can change a young person’s perception
of Computer Science and technology. Qualitative and
quantitative data methods were used to collect data and
there were a range of both open and closed questions. The
data sample consisted of 32 females and 61 males and the
students were aged between 9 and 14 years old.
4. RESULTS
Males typically used Mindstorms from a younger age than
Females, though our sample is not representative of the UK
population and this should be kept in mind. This study
does not attempt to make population estimates; Tnor does
it document or assume that every participant of the FLL in
the UK will have the same experience or learning outcomes.
However, we assume that the trends are an indicator of the
effect of Mindstorms in education. It is recognised that this
sample is limited and a wider range of participants would
have been needed for a more in depth study.
4.1 Interest in Computers
The aim of these questions was to gauge how interested
pupils were in using LEGO Mindstorms. 81% of males and
85% of females indicated that Mindstorms made them more
interested in computers. The results of this section indicate
that Mindstorms is popular amongst this cohort and
increases the interest in computers for both male and female
pupils. Using computers for something other than gaming
was a common reason for this that came up on the surveys,
as well as engendering quick thinking and creativity. All
participants indicated that they wanted to continue to use
Mindstorms during the next academic year.
4.2 Careers
The aim of these questions was to understand if LEGO
Mindstorms would encourage pupils to consider a career in
computing. The results are interesting: even though the
participants indicated that they really enjoyed using Mindstorms,
their aspirations with regards to computing careers
suggested the opposite. 44% of males and only 14% of females
indicated that using LEGO Mindstorms made them
interested in having a computing job. This suggests that
Mindstorms can make male participants more interested in
a computing career in some cases, but is not always successful
in influencing people’s career choices. Male participants
gave specific reasons stating that they enjoyed programming,
whereas female participants gave reasons that were vague
(such as finding it “interesting”). This data suggests that
148
although both males and females enjoy using Mindstorms
and are more interested in technology after using the kits, it
does not necessarily mean that Mindstorms can make someone
more interested in pursuing a technical career.
4.3 Skills
The data we collected suggests that Mindstorms can teach
people new skills. The students felt that they had learnt how
to work as part of a team, how to communicate effectively
with people, and problem solving skills. Additionally, both
genders said they had learnt “new skills on the computer”,
“how to handle problems”, and “how to be more creative”.
However, 85% of females thought they had learnt new skills
compared to 66% of males. This data could suggest that
females have a more positive experience with Mindstorms
than males, and feel that it is more rewarding.
4.4 Educational Choices
Almost half of the males said that they were more interested
in taking GCSEs in the STEM (Science, Technology, Engineering
and Mathematics) subjects after using Mindstorms.
Half of the female participants said that Mindstorms had
influenced which GCSEs to study and said that they were
now going to study ICT, Design Technology, and Further
Maths. This data suggests that Mindstorms can influence
young adults’ educational choices in some cases. This also
suggests that Mindstorms made them more aware of the
STEM subjects and made the subjects seem more appealing
and interesting.
5. CONCLUSIONS
LEGO Mindstorms can be used successfully in education;
young people learn new skills and are keen to continue with
it to widen their knowledge of programming and technology.
Some females consider programming to be the best
part of using Mindstorms and taking part in the FLL. Although
it does not seem to make young people more aware
of Computer Science careers, it does enhance their interest
in computers and their uses.
6. REFERENCES
[1] D. Aufderheide, W. Krybus, and U. Witkowski12.
Experiences with LEGO Mindstorms as an embedded
and robotics platform within the undergraduate
curriculum. In FIRA-TAROS, volume 7429 of LNAI,
pages 185–196. Springer, 2012.
[2] Department of Education. Harmful ICT curriculum set
to be dropped this September to make way for rigorous
Computer Science. Press notice, 11 January 2012.
[3] G. Lovegrove and W. Hall. Where have all the girls
gone? University Computing, 9(4):207–210, 1987.
[4] H. H. Lund and L. Pagliarini. Robocup Jr. with LEGO
Mindstorms. In ICRA, pages 813–819. IEEE, 2000.
[5] W. McWhorter and B. O’Connor. Do LEGO
Mindstorms motivate programming students in CS1?
In SIGCSE, pages 438–442. ACM, 2009.
[6] R. Pau, M. Grace, and W. Hall. IT’s boring: A
comparison of male and female students’ experiences of
ICT GCSE/A-level and Computing A-level lessons and
their impact on student motivation. School Science
Review, 92:89–94, 2011.

Exploring the processing of formatted texts by a
kynesthetic approach ∗
Carlo Bellettini Violetta Lonati Dario Malchiodi
Mattia Monga Anna Morpurgo Mauro Torelli
Dipartimento di Informatica
Università degli Studi di Milano
Milan, Italy
{bellettini, lonati, malchiodi, monga, morpurgo, torelli}@di.unimi.it
1. MOTIVATION
One of the first experiences most pupils have with computing
is through a word-processor and formatted texts. Indeed,
computer literacy is frequently focused on acquiring dexterity
with office automation tools, whereas the underlying conceptual
challenge of automatic processing of information is
largely ignored [5, 4]. In fact, mastering the use of a word
processor may increase one’s knowledge about typography
or even the way one should organize ideas and discourse, but
it gives very little insight about computing sciences without
a specific emphasis on that topic.
Recently, we started several activities [2, 3] aimed at introducing
computing concepts to pupils, both striving to
make boys and girls at ease with the intrinsic abstract nature
of informatics and trying not to disappoint them with
something not linked with the technology they are used to.
In this paper we describe our experiments with a kind of
kinesthetic/tactile learning activity [1] we called algomotricity.
In algomotricity the abstract symbolic manipulation is
(partially) replaced by physical activities, which should help
the pupils in developing their mental representation [6]. We
tried to choose activities clearly linked to the acquaintance
of students with computers and applications. In the following
we report on a teaching activity about word-processors
we proposed to a group of 25 pupils in 9th and 10th grades of
an Italian secondary school. The pupils had some familiarity
with word-processor operations. The learning objective
we had in mind was the understanding of the challenges
posed by the automatic elaboration of formatted text and
we mainly focused on information representation techniques.
2. DESCRIPTION OF THE EXPERIENCE
The overall activity (8 hours in 4 non-consecutive days)
∗ The authors would like to thank Giorgio Fattorelli and the
Marie Curie IIS for giving the opportunity to experiment
our ideas in their school.
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WiPSCE 2012, Hamburg, Germany
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$15.00.
149
was organized in four phases:
1. A first approach to text formatting with a word processor;
2. A dramatization of the process through the use of tangible
objects;
3. A game designed to force pupils to restructure their mental
models and discover the power of symbolic meta-languages;
4. A final use of special software tools for formatting texts,
also able to show the data structure used to record the
meta-information.
We started and ended in a computer lab, in order to partially
match pupils’ expectations about informatics. The
risk we wanted to avoid was that of being perceived as unrealistic
or fruitlessly “academic”. Thus, we designed the
activity around two main ideas: (a) computers and software
tools should be of secondary importance, but the conceptual
link with them should be clear; (b) the approach should be
mostly allosteric [7]: the direct transmission of knowledge
should be kept to a minimum, and pupils should be forced
to reconsider their mental models about text formatting by
discovering themselves useful techniques. Moreover, the abstract
nature of computing should be conveyed by concrete
examples and physical activities. Pupils were requested to
work in small groups to foster confrontation and almost every
task was proposed together with an accompanying metacognitive
reflection. In particular we often asked pupils to
imagine how what they described would be understood by
someone who ignored the context.
Formatting transfers information.
We started in a computer lab and the first objective was to
convince pupils that formatting is not just an aesthetic issue,
but it has also an important role in transferring information:
indeed the meaning of a text is given by the words and their
formatting. Pupils (working in couples) were requested to
produce a formatted text of their choice. Then they had
to answer some questions about their work: Which type of
formatting did you use? Why did you choose it? Did you
use more than one formatting for the same piece of text?
The activity took more than we thought and we were not
able to introduce the second part, in which pupils would
have been requested to produce a formatted version of a text
read aloud by the conductor, where changes in voice tones

and emphasis should be rendered by italics, bold, colors,
etc. Then students would have been divided into groups for
reflecting on the reproducibility of this process by comparing
results produced by different teams.
How to record meta-information.
The second phase was carried out in the gymnasium of
the school. Retrospectively this was a bad idea, since the
big hall was rather chaotic and groups were too dispersed to
foster a fruitful discussion. The proposed task was the reproduction
of a formatted text on a big copy of the text put
on the floor. The copy contained the same words, but no
formatting (also the alignment was slightly different). Formatting
had to be codified by using the objects available in
the gym. Every group of pupils (6 persons) was requested to
write down the rules they used in the codification. Another
team had to interpret these rules in order to decode the text
formatted through physical objects and get back to the original
formatted text. Most rules turned out to be ambiguous,
especially when more than one kind of formatting had to be
applied to the text. The objects were used mainly to mimic
the formatting in the prototypical text. However, in some
cases, pupils discovered a more abstract symbolic use as a
means to cope with shortage of objects: for instance, two
spoons were used to mark the beginning and the end of the
word respectively, since there were not enough spoons to
cover the whole word.
Meta-language tricks.
The third phase was carried out in a classroom and was
organized around a game. Teams were again requested to
reproduce a formatted text with objects and write down
codification rules precise enough to be followed by another
team. However, the game was made fun by the introduction
of a “cost” for the objects. In fact, the more an object could
be used to mimic a piece of formatting, the more it costed
in order to promote their symbolic use, e.g., since spaghetti
pasta could be easily associated to the underlying meaning,
its cost was very high. The winner would be the team able
to hand in an unambiguous codification with the lowest cost.
The cost incentive was enough to let the pupils discover what
is commonplace in mark-up languages: the use of tags at the
beginning and at the end of (possibly overlapping) regions.
A second round of the game was proposed without objects.
Instead the pupils had only a multi-set of alphabetic characters
on small pieces of paper. Some of the characters (for
example, the letters that are not in the Italian alphabet: j, k,
w, x, y) were not used in the words and could be easily used
for weaving meta-information into the text. However, this
trick was not suggested by the conductors but discovered by
the pupils. Some of them tried to use the characters with a
meta-meaning by placing them with a different orientation:
a b for example, for meaning bold.
Rediscovering formatting tools.
The final phase was carried out again in a computer lab.
Pupils were introduced to a special software tool able to
show a formatted text according to three different views:
formatted and encoded either using a simplified mark-up
language or a tree of objects. They were again requested to
reproduce formatted texts by working on the other representations:
they saw, however, the effect of their (syntactical)
manipulation in the formatted view.
150
3. EVALUATION AND FUTURE WORKS
We think the experience was successful. For example, all
the pupils demonstrated to have grasped the idea of a metalanguage
expressed in the same alphabet of the language
itself. This is considered quite an abstract concept, but
it was found rather natural (even obvious) by the pupils.
The link with word-processor and web technologies known
to pupils was recognized. At the final recap we were also
able to show that the same concepts are behind the scenes in
several slightly different contexts, for example when editing
Wikipedia entries. The pupils’ feedback was mostly positive:
they did have fun and believe to have learned something.
However, some of them found that the tasks were
sometimes too easy and the part in the gym (see Section 2)
was considered boring by several participants.
All in all, the proposed activity turned out to be a good
way for conveying abstract computing concepts to pupils of
secondary schools. We are now working in refining the activity:
as a mid term goal we aim at producing didactic
material that should be self-contained enough to be used
by independent teachers in their classes. As a first step we
re-proposed the activity in an another school, with younger
pupils (6th grade) under the conduction of a math teacher
who did not participate in the conception. We are now
studying reports and videos of the experiences: the first
impression is that it worked also in this different context.
We found that the algomotricity approach can be effective
in presenting abstract symbolic manipulations in very concrete
ways. By choosing activities clearly connected with the
acquaintance of pupils with computers and tools we believe
this can be very successful in elaborating a fruitful understanding
of informatics concepts.
4. REFERENCES
[1] A. Begel, D. D. Garcia, and S. A. Wolfman. Kinesthetic
learning in the classroom. In Proc. of the 35th SIGCSE
TSCSE, pages 183–184, New York, USA, 2004. ACM.
[2] A. Lissoni and V. Lonati and M. Monga and A.
Morpurgo and M. Torelli. Working for a leap in the
general perception of computing. In A. Cortesi and F.
Luccio, editor, Proc. of informatics education europe
III, pages 134–139. ACM, 2008.
[3] V. Lonati and M. Monga and A. Morpurgo and M.
Torelli. What’s the fun in informatics? Working to
capture children and teachers into the pleasure of
computing. In Kalas and Mittermeir [8], pages 213–224.
[4] M. Calzarossa, P. Ciancarini, L. Mich, and
N. Scarabottolo. Informatics education in Italian high
schools. In Kalas and Mittermeir [8], pages 31–42.
[5] V. Dagien˙e. Informatics education for new millennium
learners. In Kalas and Mittermeir [8], pages 9–20.
[6] R. M. Felder and L. K. Silverman. Learning and
teaching styles in engineering education. J. of
Engineering Education, 78(7):674–681, 1988.
[7] A. Giordan. From constructivisme to allosteric learning
model. http://www.ldes.unige.ch/ang/publi/
articles/unesco_AG_96/unesco96.htm, 1996.
[8] I. Kalas and R. T. Mittermeir, editors. ISSEP 2011,
volume 7013 of LNCS. Springer, 2011.

Teachersʼ Perceptions Of The
Value Of Research-Based School Lectures
Jonathan Black, Paul Curzon,
Chrystie Myketiak, Peter W. McOwan
Queen Mary University of London
London
{jonathanb, pc, chrystie,
pmco}@eecs.qmul.ac.uk
ABSTRACT
A major challenge facing secondary schools is to encourage
students to take computing courses. One approach is to invite
external speakers from universities or industry to give lectures.
The cs4fn project, a large UK-based initiative to enthuse students
about computer science, includes this approach. Speakers from
Queen Mary, University of London, visit schools to talk to
students about computer science research. Our interactive talks
tell engaging research-based stories on topics such as artificial
intelligence and human-computer interaction as well as using
magic tricks to illustrate computing principles. We asked teachers
to complete post-talk surveys online; in particular we were
interested in whether they believed students’ perceptions of the
subject had changed. They reported that their students’ views of
computer science were improved, and that they felt students were
more likely to take classes in computing in the future as a result of
the talk.
Categories and Subject Descriptors
K.3.2 [Computers and Education]: Computer and Information
Science Education – Computer Science Education
General Terms
Human Factors.
Keywords
Public engagement, outreach, recruitment, K-12, schools,
teachers, lectures, magic.
1. INTRODUCTION
A range of studies have looked at the effectiveness of lectures in
higher education contexts, (e.g., Cooper & Foy, 1967) including
those focusing on particular lecture styles such as PowerPoint
lectures (e.g., Bartsch & Cobern, 2003) Despite the popularity of
the lecture approach in outreach activity, there is little published
evidence about its effectiveness against the aims of public
engagement activity and of its value in this context. This may be
because organised computer science engagement projects are
often workshop-style. Lecture-style approaches may be happening
‘under the radar’, missed out of evaluation work on workshopstyle
programmes. However, the lecture style is widely used and
warrants study.
The cs4fn project (Curzon, 2007) highlights the importance of
flexible ‘take-away resources’ such as magazines and booklets
rather than focusing just on workshops or lectures. We have found
that this increases motivation and supports teachers (Myketiak et
151
Laura R. Meagher
Technology Development Group
Dairsie, Fife
Laura.Meagher@ btinternet.com
al., 2012). Curzon et al (2009) present evidence for the success of
this approach in engaging with potential students.
In previous work we have surveyed audiences for our magic
shows and showed their popularity with students (e.g., Curzon &
McOwan, 2008). In this paper we evaluate the effectiveness of a
lecture approach, surveying teachers about their own views and
those of their students after seeing a talk. We focused on a range
of indicators about the value of the talks but with a particular
emphasis on their value in inspiring students to take their
computing education further. We show that this kind of researchbased
interactive lecture is highly valued by teachers and they
believe it is effective in encouraging students to consider taking
the subject further both at school and university.
2. METHOD
We present feedback from two separate lectures given 19 times in
the 2010-2011 and 2011-2012 academic years. We use audience
volunteers in kinaesthetic activities to illustrate core concepts. The
lectures also focus on telling engaging research stories. They do
not attempt to cover curriculum topics directly, nor do they
directly cover career choices or discuss university courses.
Each time we delivered a talk at a school, we solicited a response
to an online survey from the teacher. This meant the responses
could be anonymous and gave the teacher an opportunity to talk to
the students about the talk before filling out the survey. The
survey consisted of 16 questions. We asked teachers to identify
the lecture title, and which team member delivered it. The survey
went on to ask general (non-identifying) details about the
audience and the school that hosted the talk. There were then a
series of Likert scale questions about the value of the approach,
two yes/no questions followed by a series of open questions.
3. TEACHERS’ PERCEPTION OF THE
VALUE OF THE LECTURES
3.1 Quantitative results
The survey received 19 teacher responses. We asked the
respondents to give an overall view of the lecture with options on
a 5-point scale. All (100%) were positive. 89.5% (n=17) gave the
highest rating (very good) while the remaining 2 teachers rated it
as good. Respondents were also asked if the lecture met their
needs and if they would recommend the lecture to other
schools/teachers, with yes/no options in both cases. All 19
teachers (100%) responded yes to both questions.
When asked about students’ opinions, 84.2% (n=16) strongly
agreed that their students had enjoyed the lectures with the
remaining 3 agreeing. None were neutral or disagreed. 78.9%

(n=15) strongly agreed that their students found the lecture
interesting while the remaining 4 agreed – a 100% positive
response. 73.7% (n=14) strongly agreed that the lecture had
improved the students’ understanding of the subject with the
remaining 5 agreeing.
The last question in this section asked whether some students had
changed their view of computer science in a positive way. 36.8%
(n=7) strongly agreed. 42.1% (8) agreed and the remaining 21.1%
were neutral. We asked whether, as a result of the lecture, one or
more students were now more likely to consider taking computing
subjects further at school. 20% (n=3) strongly agreed with this
statement, 53.3% (n=8) agreed (i.e. 73.3% of responses were
positive) and the remaining 4 were neutral. Four teachers did not
answer.
We asked a similar question about students’ university choices. In
response, 23.5% (n=4) strongly agreed that one or more students
were more likely to go on to take the subject at university, 52.9%
(n=9) agreed, meaning 76.4% of responses were positive. A
further 17.6% (n=3) were neutral, 1 person strongly disagreed and
1 did not answer the question. The teacher who strongly disagreed
was otherwise very positive about the lecture, and did not suggest
any improvements. It is unclear why they answered the question
the way they did.
3.2 Comments
The comments in response suggest that teachers believe the form
of a lecture affects student enjoyment as much as its content. A
typical comment said, “The lecture was fast-paced, dynamic and
really challenged the students [...] This was a very appealing way
to promote the study of ICT.” Another teacher approved of “the
mix of talking, video and practical examples.” Many others
mentioned that the kinaesthetic activities were the students’
favourite parts of the lectures. Comments by teachers on the
benefits of the interactive elements followed a couple of themes.
Some mentioned stylistic benefits (e.g., increasing student
enjoyment and engagement), while others mentioned contentrelated
benefits (e.g., modeling of computer functionality and
opening students’ minds).
4. DISCUSSION AND CONCLUSIONS
We have evaluated teachers’ perceptions of the value of university
lectures taking place in schools. This involves lectures that draw
from research topics and involve interactivity using a kinaesthetic
approach. In particular we focused on teachers’ perceptions about
the immediate effect of the lectures and whether it had changed
students’ attitudes to taking the subject further.
Teachers were confident that their students had found the lectures
interesting and now understood the subject better. They were less
strong but still very positive in their agreement that the lectures
had changed their students’ perception about computing, and that
they had influenced students’ plans for the future. This may
reflect the difficulty of one lecture to change overall perceptions
or future plans, but it may also simply be a reflection of the
difficulty teachers face in assessing and reporting students’ views.
5. FUTURE WORK
This pilot study examined teacher responses to 19 lectures about
computing. We hope to gather more responses in a larger study to
come. We also intend to gather more direct and in-depth data on
student perceptions. Finally, we are in the process of collecting
152
and analysing recruitment data to examine the ultimate effect of
talks on recruitment at university. Whilst this study has focused
on the value of such talks to teachers, this ongoing work will
explore the value to universities.
6. ACKNOWLEDGMENTS
The cs4fn programme is funded by EPSRC research agreement
(EP/F032641/1) with additional support from Google’s CS4HS
programme. CHI+MED: Multidisciplinary Computer-Human
Interaction research for the design and safe use of interactive
medical devices project is funded by EPSRC research agreement
EP/G059063/1.
7. REFERENCES
[1] Bartsch , R.A. & Cobern, K.M. 2003. Effectiveness of
PowerPoint presentations in lectures, Computers &
Education, 41 (1) August, 77–86, Elsevier, DOI=
http://dx.doi.org/10.1016/S0360-1315(03)00027-7
[2] Cooper, B. & Foy, J.M. 1967. Evaluating the effectiveness of
lectures, Higher Education Quarterly, 21(2) 182-185. March.
DOI: 10.1111/j.1468-2273.1967.tb00231.x
[3] Curzon, P. 2007. "Serious Fun in Computer Science", 12th
Annual Conference on Innovation and Technology in
Computer Science Education, organised by the ACM Special
Interest Group on Computer Science Education (SIGCSE),
ACM SIGCSE Bulletin 39(3) p1, DOI:
10.1145/1269900.1268785
[4] Curzon, P., Black, J., Meagher, L.R., McOwan, P.W. 2009.
“cs4fn.org: Enthusing students about Computer Science”,
Proceedings of Informatics Education Europe IV, Christoph
Hermann, Tobias Lauer, Thomas Ottmann and Martina
Welte (Eds.), pp73-80, Freiburg, Germany, November.
[5] Curzon, P. & McOwan, P.W. 2008. “Engaging with
Computer Science through Magic Shows”, ITiCSE 2008, The
13th ACM SIGCSE Annual Conference on Innovation and
Technology in Computer Science Education, ACM SIGCSE
Bulletin 40 (3), pp179-183. June 30-July 2, Madrid, Spain.
DOI: 10.1145/1384271.1384320
[6] Myketiak, C., Curzon, P., Black, J., McOwan, P.W., and
Meagher, L.R. 2012. cs4fn: a flexible model for computer
science outreach. In Proceedings of the 17th ACM annual
conference on Innovation and technology in computer
science education (ITiCSE '12). ACM, New York, NY,
USA, 297-302. DOI: 10.1145/2325296.2325366

Technocamps: Bringing Computer Science to the far west
ABSTRACT
Roger D. Boyle
Computer Science
Aberystwyth University
Penglais
Aberystwyth SY23 3DB
Wales
rob21@aber.ac.uk
Technocamps is a 3-year EU funded project to bring an
awareness of technical Informatics to the 11-19 age group in
the ‘Convergence zone’ of Wales. The project is coordinated
through four universities, with materials and activities being
developed in each of the academic hubs. Projects of this
scale are rare, both in terms of geographic area and financial
backing, creating a new set of challenges and opportunities.
We review the background to the project and its need,
outline its activities during its first year and the challenges
they have presented, and make observations about portability
and derived good practice that might inform similar
projects elsewhere. Longer term evaluation and longitudinal
study will develop during the next two years of the work.
1. BACKGROUND
Wales is a small country at the west of the UK mainland
with a clear and illustrious culture, including a vibrant language.
It is predominantly rural, although the south of the
country became pre-eminent in [now defunct] coal mining
and associated heavy industry. It is a victim of rural poverty,
seasonal (un)employment and the collapse of industry with
no compensatory re-investment.
Informatics in UK schools is in a difficult state of transition
[3]. 2012 saw very consequential announcements by
government [5] and the Royal Society [6]. In particular the
latter proposes the teaching of Informatics be partitioned
three ways:
Hannah M. Dee
Computer Science
Aberystwyth University
Penglais
Aberystwyth SY23 3DB
Wales
hmd1@aber.ac.uk
• Computer Science [CS]: The rigorous academic discipline,
encompassing programming languages, data structures,
algorithms, etc.
• Information Technology [IT]: The use of computers,
in industry, commerce, the arts and elsewhere, including
aspects of IT systems architecture, human factors,
project management, etc.
• Digital Literacy: The general ability to use computers.
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WiPSCE 2012, Hamburg, Germany
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$15.00.
153
Frédéric Labrosse
Computer Science
Aberystwyth University
Penglais
Aberystwyth SY23 3DB
Wales
ffl@aber.ac.uk
In brief, it is widely agreed that Digital Literacy should
be achieved by all but is a skill akin to the ability to read,
unsuited to formal curriculum time or assessment in high
schools. IT is a respectable and necessary academic activity
but is distinct from CS. It is the very weak condition of CS
that has provoked concern in influential quarters.
2. IMPLEMENTATION
2.1 Organisation
Technocamps [7] is an EU Convergence European Social
Fund project, specifically designed to address the CS gap
within the 11-19 age group within the Welsh Convergence
Zone. Its aims are: to encourage interest in Science and
Technology; to provide opportunities for students to gain
insight into the practical application of cognate STEM subjects
in a work environment; to increase the number of girls
taking up science and technology; to increase the number of
students who progress to computer science, technology and
engineering at higher levels. The 3-year project attracted
funding of £6M and involves 4 universities across the zone.
By design, activities are where possible portable between
sites. Significant transport problems and sparse population
mean that in most cases schools are served by their closest
HEI. After decades of decline, the Welsh language is enjoying
a slow but steady recovery, which is strongest in many of
the more isolated areas the project serves [8]. Accordingly,
materials and activities are being made bilingually.
The project targets 11-19 year-olds, and especially females
and those who are not in education, employment, or training.
‘Engagement’ is defined not to be a passing contact,
and should be a minimum of 6 hours of exposure. It is acknowledged
that 6 consecutive hours would be unlikely, and
the precise target is to meet an individual at least twice for
at least 3.5 hours each time.
2.2 Activities
One day engagements: Most contact is via one-day contact;
material is specifically designed to be distinct from
the school curriculum. Sessions last about 4 hours, with
topics ranging from general (Scratch programming) to the
more rarefied (such as cryptography, robotics, or Arduino
[1] based work).
Bootcamps: ‘Bootcamps’ are 3-day engagements: robot
navigation, and Arduino projects (sailing robots and (digitally
ornamented clothing) have been topics. Clearly, three
full days provide a depth of opportunity in excess of 3.5
hours: this is an opportunity of great use to research-oriented

projects.
Other: There are several other miscellaneous engagements:
these include after-school ‘Technoclubs’, Science week [2],
Home Schoolers, and other sundry public engagements such
as the Aberystwyth ‘Beachlab’.
3. EXPERIENCES
Attempting to engage the schools is challenging; the country
is sparsely populated with few top-quality roads – reaching
parts of the population can be very time-consuming, and
schools are reluctant to give up staff and pupil time to extracurricular
activity. Schools prefer to devote time directly before
or after holiday periods (to minimise disruption) which
provokes serious congestion for university resource.
One-day engagements are held either in the university or a
school, and often present transport or scheduling difficulties.
Subject matter is not hard to generate: programming, computer
applications of music, robotics, AI are just some examples.
We learn in delivery what may be obvious to trained
schoolteachers: we have to take care to schedule breaks and
variation into what is otherwise a very long experience.
Bootcamps differ inasmuch as children have volunteered
for the event and so the level of motivation is much higher.
Not only is length of time appreciably greater, but the intensity
and depth of the exercise allows an immersion that
would be difficult to reproduce in the classroom. These activities
have been ambitious and very successful: young participants
have mastered programming for calibration and
navigation (in C), full bottom-up construction, basic electronics,
circuitry, and sewing.
Miscellaneous engagements have productively exposed many
hundreds of passers-by to robots, AI, UAVs, wearables, etc.,
and were introduced to the project and its aims.
4. OBSERVATIONS
Resourcing: European funding has made it possible to focus
on issues of logistics and liaison that are often victims
of underfunded or volunteer outreach activity. There is no
expectation of continuation resource and it is incumbent on
the project to leave some lasting impact. We see lasting
influence on teachers’ understanding and enthusiasm as the
best aim.
Challenge of materials: We are delighted to witness that
our workshops are all accessible, and that perhaps we can
be more ambitious. We do not suggest that 100% of the
class walk away with perfect understanding, but are convinced
that the great majority are having some aspect of
their understanding of CS improved.
Targets: The primary metric of the funding authority is
pupil contact numbers; with hindsight, the numbers are not
ambitious; all schools have been co-educational and it has
not been difficult to engage with a good number of girls. We
consider it more fruitful to engage with the 11-14, rather
than the 15-19 bracket. Geographic coverage has been uneven
– many areas of the country are very rural with very
small schools
Evaluation: Short term evaluation has been achieved by
crude questionnaires before and after formal engagements.
While these are of questionable value, we can see a clear
shift in view of science and technology toward the positive.
Portability: Materials are routinely packaged for delivery
across all sites, and in due course more generally. Complete
154
‘bundles’ are posted as freely web-available resource.
Staffing: Very few university staff have the training or skills
that go with full-time teaching and from time to time this
has been an issue. Teachers will know classes well and will
defuse both the over-eager or disruptive, and will have techniques
to motivate the bored. Often, Technocamps staff
had to learn quickly how to ‘crowd manage’, for example by
having ready a repertoire of illustrative practical activities
or videos that provide relevant variety.
Welsh: Serious students of CS in most countries would benefit
from working at least partly in English, and pursuing it
in a minority language such as Welsh would be eccentric.
Nevertheless, many of the target group operate in Welsh as
a first language. It has not been easy to meet this need
from the project personnel, and a solution has to be found
by full engagement of Welsh speaking teachers who share
the project aims.
The project has been a success by all measurements: pupil
and teacher reaction is positive and the targets are being hit
with ease. It will continue to develop for the next 21 months,
improving what we do, particularly in devising techniques
for longitudinal monitoring. We aim to leave in place a suite
of web-readable materials most of which will be applicable
anywhere and would permit anyone so minded to bring ‘real
CS’ into schools of their choice. Generally, we are accruing
a range of first-hand experience of putting challenging CS in
front of children much younger than those who normally see
it, and doing so often in unusual and sometimes challenging
environments (and languages!). We are developing conclusions
on what represents best practice in these endeavours,
some of which we have presented here. At project end, we
will collate and publish these as a guide to efficient HE intervention
in high school Informatics.
5. REFERENCES
[1] Arduino. 2012. http://www.arduino.cc/.
[2] British Science Association. National Science and
Engineering Week, 2012. http:
//www.britishscienceassociation.org/web/nsew/.
[3] Computing at School. Computing For the Next
Generation . . . , 2012.
http://www.computingatschool.org.uk/.
[4] M. Gove MP. Speech given to BETT, January 11 th ,
2012. http://www.guardian.co.uk/education/2012/
jan/11/digital-literacy-michael-gove-speech.
[5] Royal Society. Shut down or restart?, January 2012.
Final report of the Computing in UK Schools group
http://royalsociety.org/education/policy/
computing-in-schools/report/.
[6] Technocamps. Creating the Next Generation of
Technologists, 2012. http://www.technocamps.com/.
[7] Wikipedia. History of the Welsh language, 2012.
http://en.wikipedia.org/wiki/History_of_the_
Welsh_language.

Abenteuer Informatik – Hands-on exhibits for learning
about computational thinking
Dr. Jens Gallenbacher
Didaktik der Informatik,
Technische Universität Darmstadt
Hochschulstr. 10
64289 Darmstadt
+49 6151 16-2573
jg@di.tu-darmstadt.de
Abstract
Computational thinking is one of the pillars of the ACM-CSTA
standards for teaching computer science from kindergarten to
college. Our approaches Abenteuer Informatik – Informatik
begreifen (adventures in informatics – hands on computer
science) and Abenteuer Technik are well established in the
german-speaking community as means to connect computer
science with other subjects and as means of clarifying some
prejudices against computer science, especially problematic for
establishing computer science as subject in schools.
Categories and Subject Descriptors
K.3.2 [Computer and Information Science Education]:
Curriculum
General Terms
Human Factors
Keywords
Computational thinking, curriculum, computer science education,
Abenteuer Informatik, computer science unplugged
1. Introduction
Since the invention of commercial electronic computers in the
middle of the last century the public perception of relevant
aptitudes for computer-users shifted continuously: From
scientificly trained expert over the well-informed programmer to
the "just"-user today. In the same manner the computer-science
education in schools changed.
While in the beginning it has been obligatory to learn basics of
computer-technology and programming, because standardapplications
were not available or to expensive for schools,
today's focus of cs-education often is computer literacy, like the
curriculum for european computer driving licence.
The stereotypic public and political opinion about computer
science also changed radically: On one hand the search for the
correct buttons in office, on the other hand a cryptic science, that
is everything about Devices and nothing about humans.
Since 2006 Jeannette M. Wing, Professor at CMU, gives
distinction to the term "computational thinking" and elaborated,
that computer science is not the science about computers. It
particularly produces concepts of (human) thinking, that amongst
others may be used to program machines, but mostly benefit other
155
sciences and dayly use. In short, Wing proved computer science
provides general education (in europe often stated by the german
term "allgemeinbildend"). She legitimized "true" computer
science as subject in general schools, so computational thinking is
positioned prominently in the standards of the CSTA and in the
proposal for a computer science curriculum in Great Britain,
which are likely to be implemented by many schools.
Adventures in informatics – hands on computer science is a very
practical approach with similar goals. The book [Gallenbacher
2012] is sold over 10.000 times and in the exhibition over
100.000 people played, puzzled to get their hands on computer
science at 31 different locations, amongst others ars electronica
center in Linz and Heinz Nixdorf Forum in Paderborn. It
fascinated kids as well as their grandparents and many politicians.
Reason enough to evaluate, how both approached could benefit
from each other.
2. Computational Thinking
Wing describes “Computational Thinking is the thought
processes involved in formulating problems and their solutions so
that the solutions are represented in a form that can be effectively
carried out by an information-processing agent Informally,
computational thinking describes the mental activity in
formulating a problem to admit a computational solution. The
solution can be carried out by a human or machine, or more
generally, by combinations of humans and machines.” [Wing
2010]
A position paper published by the german “Gesellschaft für
Informatik” [Biundo 2006] states the situation in a very similar
way.
3. Abenteuer Informatik, Abenteuer Technik
and Computational Thinking
Abenteuer Informatik as concept to teach computer science
without computers provides many approaches for teaching
computer science with focus on computational thinking. The full
exhibits referenced by the examples in this chapter may be
downloaded from www.abenteuer-informatik.de - most of them
are available in german, english and spanish language.

The Monkey Puzzle
Sometimes you know exactly how to solve a problem but fail
nonetheless, because it would cost too much time or is "not
scalable". Problems of that kind exist in many disciplines.
If you need a bridge over a ditch 1m wide, you only need to lay a
metal or stone plate across it, and it‘s done ! However, a bridge
across a gap of 10m cannot be made of ten such plates laid end to
end; we need first a precise static analysis of the construction,
with an outlay considerably greater than 10 times that of the first
example. If you want a bridge over a gap of 100m , the necessary
outlay will be many times greater. The longest bridge span
achieved to date is about 2000m. Each additional meter requires
an immeasurably greater planning effort and immense material
and building costs.
The peculiarity of the second set of monkey puzzles is not just
that the computer would take a very long time to find a solution;
it is fundamentally incapable of answering this problem! So
computational thinking provides means of recognizing, that there
are certain problems, which by their nature can never be
completely solved algorithmically.
Informatics Letter by Letter
The formal term "information" is very important to computer
science, as expressed by the also used "informatics", which is
derived from Informatik, build of information and automation.
Information is important to all disciplines in our modern world -
but what is information at all?
An approach by Shannon to use language for defining information
can easily be adapted for computational thinking. Please try to
decipher the following text. It is from a well known story, but is
missing all except the frequently used letters.
hen aune as tee eas o, the enhantess shut he in a toe, hih a in a
oest. e toe ha no oo, ut hih u as a ino. hen the enhantess ante to
o in, she ae hese eo the ino, an ie „aune, aune, et on ou hai.“ aune
ha aniient on hai, ine as sun o. hen she hea the oie o the enhantess
she oun he hai oun a hoo o the ino. e hai e tent as on, an the
enhantess ie u it.
Interesting, isn‘t it? Here we have retained about 70% of the
original letters, but the result is gibberish. Below is the same text,
with almost the same number of letters as in the previous
example, obtained by removing only the vowels from the original.
Can you read the story now?
Whn Rpnzl ws twlv yrs ld, th nchntrss sht hr n twr, whch ly n frst.
twr hd n dr, bt hgh p ws wndw. Whn th nchntrss wntd t g n, sh
plcd hrslf blw th wndw, nd crd „Rpnzl, Rpnzl, lt dwn yr hr.“
Rpnzl hd mgnfcnt lng hr, fn s spn gld. Whn sh hrd th vc f th
nchntrss sh wnd hr hr rnd hk f th wndw. hr l twnty yrds dwn,
nd th nchntrss clmbd p by t.
Apparently the amount of information has to do something with
the frequency: A text containing just common symbols contains
no information or at least less information than the text with
infrequent symbols.
In computer science, this knowledge is used to encode texts and
other information and save memory or bandwidth, e.g. by using a
huffman-tree or just zipping files. In other disciplines, this is used
as well - like in linguistics: In ancient sanskrit the symbol for T
followed by A, which is a very common combination, is written
as
156
If you want to use T without A, which is shorter, but less
common, you have to write with more strokes:
So computational thinking is not a new discipline, but known for
some millennia...
Abenteuer Technik
In our science lab for children we implicitly use computational
thinking too. We give out tasks like "building wings for an
efficient wind generator out of some cardboard, wood and plastic
film" or "building a water wheel generator out of milk carton".
The degrees of freedom are intentionally set very high, so the
children have to model the problem first and then make a
conscious decision to focus on optimizing one or two out of more
than 100 possible parameters.
4. Conclusion
Approaches to teach computer science and engineering without
using complex and abstract technology, like [Bell 2006],
[Gallenbacher 2008, 2012] are most suitable to implement
computational thinking in computer science education.
5. References
[1] [Bell 2006] Tim Bell, Ian H. Witten, Mike Fellows:
Computer Science Unplugged - Teachers edition,
downloaded from http.//www.csunplugged.org/
[2] [Biundo 2006] Susanne Biundo, Volker Claus, Heinrich C.
Mayr: Was ist Informatik, unser Positionspapier,
Gesellschaft für Informatik, 2006
[3] [CSTA 2010] Cameron Wilson, Leigh Ann Sudol, Chris
Stephenson, Mark Stehlik: Running On Empty: The Failure
to Teach K–12 Computer Science in the Digital Age, Report
of the ACM-CSTA, http://www.acm.org/Runningonempty/
[4] [CSTA 2011] Deborah Seehorn, Stephen Carey, Brian
Fuschetto, Irene Lee, Daniel Moix, Dianne O’Grady-
Cunniff, Barbara Boucher Owens, Chris Stephenson, Anita
Verno: CSTA K–12 Computer Science Standards Revised
2011, Report of the ACM-CSTA, ISBN 978-1-4503-0881-6
[5] [Gallenbacher 2008] Jens Gallenbacher: Abenteuer
Informatik, Exhibition, http://www.abenteuer-informatik.de
[6] [Gallenbacher 2012] Jens Gallenbacher: Abenteuer
Informatik, Springer-Spektrum, ISBN 978-3827429650
[7] [Wing 2006] Jeannette M. Wing: Computational thinking,
COMMUNICATIONS OF THE ACM March 2006/Vol. 49,
No. 3
[8] [Wing 2008] Jeannette M. Wing: Computational thinking
and thinking about computing, Philosophical Transactions of
the Royal Society A (2008) 366, 3717–3725
[9] [Wing 2010] Jeannette M. Wing: Computational Thinking:
What and why?, CMU, 17. November 2010

Gaming and Mathematics: A Cross Curricular Event
Sharon Jones Ed.D
Charlotte Mecklenburg Schools
1430 Alleghany St.
Charlotte, NC 28208
1-980-343-5992
sharont.jones@cms.k12.nc.us
ABSTRACT
Computer Science education can motivate students, through
students’ interests and experiences but, it can be a challenge to
create meaningful and engaging assignments that allow for both
creativity and learning while also using modern technology
practices. This poster will highlight the liaison between computer
science and math as students use Build Your Own Blocks
(BYOB) to build projects that will use traditional computation
math skills with computer gaming processes. The poster will
highlight an after school workshop success story.
1. INTRODUCTION/BACKGROUND
Computing or computer science as a scientific discipline predates
the invention of computers. The first decades of the twentieth
century saw development of fundamental concepts in this
discipline (Gal-Ezer, Beer, Harel, Yehudai, 1995). More recently,
fueled in part by the invention of computers and their widespread
use, the study of computing has bloomed, and CS is now
recognized as an autonomous scientific discipline (Gal-Ezer, et.
al, 1995). The disciplines concepts include the analysis of
algorithmic processes as well as the design and implementation of
computing systems and these concepts influence work in other
disciplines (Gal-Ezer, et. al, 1995). As the ideas of the discipline
and the importance of technology innovation become assimilated
in our everyday culture it becomes clear that computer science
high school curriculums should be implemented to reflect the
growing importance (Gal-Ezer, et. al, 1995). However, high
school computer science education in the United States has shown
significant declines in both the number of introductory CS courses
being taught (CSTA, 2010), and the intention of students
(especially those from underrepresented populations) declaring
computing as a major (HERI, 2009). However, the U.S.
Department of Labor projects that between 2008 and 2018, 1.4
million computing jobs will have opened in the U.S. If current
graduation rates continue, only 61% of these jobs could be filled
by U.S. computing degree-earners (NCWIT, 2011). When
including only computing bachelor’s degrees, this percentage
drops to 29% of projected job openings that could be filled.
American students need a 21st-century computing education if we
want a workforce that is innovative, competitive, and wellemployed
(NCWIT, 2011). To try and change the educational
landscape, the National Science Foundation (NSF) has put in
place an initiative, the CS10K Initiative, to have 10,000 teachers
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
Conference’10, Month 1–2, 2010, City, State, Country.
Copyright 2010 ACM 1-58113-000-0/00/0010 …$15.00.
Renada Poteat
Charlotte Mecklenburg Schools
11201 Old Statesville Road
Huntersville, NC 28078
1-980-343-3842
renada.poteat@cms.k12.nc.us
157
Beth Frierson
Charlotte Mecklenburg Schools
1430 Alleghany St.
Charlotte, NC 28208
1-980-343-5992
mary.frierson@cms.k12.nc.us
teaching computer science by 2015.
To support the NSF’s CS10K initiative and Computer Science
Principles efforts to prepare both high school teachers and
students to be creators in computing, the course Beauty and Joy of
Computing (BJC) was created. This course was developed to
encourage students to experience Computer Science in a new way.
The BJC course combines Moodle-based computer programming
labs, lectures ranging from artificial intelligence and parallelism
to the social implications of computing and technology, and small
discussion sections. In the non-programming part of the course
there is a balance of fundamental optimism about the future of
computer technology with an understanding of its limitations and
potential. The BJC class has been implemented in five major
universities with great success. Therefore, to reach the CS10K
goal, the BJC curriculum has begun to be adapted for the K-12
curriculum. As selected members of the development team to
bring BJC to the high schools, we have had the privilege of seeing
firsthand the impact the course has on students. The course offers
insight into all aspects of computer science and helps to dispel
misconceptions established about the computer science discipline
and show cross curricular elements. The students are lead by the
software used in the class to build their own projects that link
computer science and all other disciplines.
Common Core standards are being implemented across the United
States. Common Core standards define the knowledge and skills
students should have within their K-12 education careers so that
they will graduate high school able to succeed in entry-level,
credit-bearing academic college courses and in workforce training
programs (NCDPI, 2011). The standards stress not only
procedural skill but also conceptual understanding, to make sure
students are learning and absorbing the critical information they
need to succeed at higher levels - rather than the current practices
by which many students learn enough to get by on the next test,
but forget it shortly thereafter, only to review again the following
year (NCDPI, 2011).
Using the platform of common core and the computer science
concepts, student exposure on how to integrate and understand
how to apply subjects to multiple disciplines is significantly
increased.
2. WORKSHOP DETAILS
2.1 Workshop Overview
We created a cross departmental workshop, Get Your Game On,
to assist students that were struggling with Algebra concepts and
used computer gaming as a way differentiation strategy. The three
day workshop after school taught students fundamental
programming concepts (ie. variables, object, and looping) while
infusing Algebra I questions provided by the schools math
department. At the conclusion of the workshop students were able
to learn basic programming concepts while reinforcing algebraic

concepts that they previously learned in class. The intentions were
to introduce students to computer science concepts and how
computer science is a part of every curriculum in a non-
threatening environment and spark their interest to learn more
about the discipline
2.2 Workshop Data
There were 15 (n = 15) students that participated in the workshop
with 13 (86.7%) males and 2 (13.3%) females. The majority of
participants were in the 11 th grade (40%) and were 17 (33.3%).
We implemented a limit on participants to allow evaluation and
analysis. Prior to the workshop, students were asked to complete a
survey including the following questions:
1. I like to play games to learn educational material (PlayGames)
2. I know what BYOB is and how to use it (BYOBis)
3. I can see how math and gaming are connected (Connected)
4. I am sure I can learn programming (LearnProg)
5. I am interested in computer science (Interested)
Results from the survey are represented in Table 1.
Table 1. Workshop Pre Survey
Variable Yes No Sometimes
PlayGames 26.7 13.3 60.0
BYOBis 0 100.0
Connected 86.7 0 13.3
LearnProg 100.0 0
Interested 86.7 0 13.3
On the last day of the workshop, students were asked to take a
post survey to evaluate their experience and content. The
following questions were included in the survey:
1. Now that you know what BYOB is, would you use it in the
future? (Future)
2. Do you feel the game you created will help you learn a math
concept? (LearnMath)
3. After the workshop, I am more interested in Computer Science.
(InterestedCS)
4. Did you enjoy the workshop? (Enjoy)
5. Would you recommend this workshop to someone?
(Recommend)
6. Any suggestions for future workshops? (Suggestions)
Results from the survey are represented in Table 2.
Table 2. Workshop Post Survey
Variable Yes No Somewhat
Future 90.0 10.0 0
LearnMath 100.0 0 0
InterestedCS 90.0 10.0 0
Enjoy 100.0 0 0
Recommend 80.0 0 20.0
158
The survey included a section for students to post suggestions to
better the workshop. Below are a few responses:
I thought it was completely awesome and epic! it was
great as is.
Try Gamemaker now that you have it. Other than that
was an amazing course.
more time to make game
Tell more people :)
It was well put together, great job guys
From the data collected, it was concluded students enjoyed the
workshop.
3. MOTIVATIONS
Digital games, whether computer-, game console-, or handheldbased,
are “Purposeful, goal-oriented, rule-based activity that the
players perceive as fun” (Klopfer, 2008). Students are motivated
by competition, challenge, and the ability to tell a story through
interaction using the computer (Prenksy, 2001). The use of BYOB
allowed students the interaction of creating the games along with
the challenge of learning algebra in order to make the game
function properly.
The workshop provided students the opportunity to work with
educational gaming software that they had not been privy to
before. The actual learning of the software motivated them to
figure out the math problems in order to have their games function
properly. The use of the computer gaming software was a catalyst
to having students work on algebra problems that they would
previously shied away from.
4. SUMMARY
Due to the positive feedback from students and teachers, we will
use the format of this workshop as a template for a series of
computer science and math workshops. Through continuing to
offer opportunities for students to see the relevance of computer
science and other disciplines, the outcome would predict
increased interest and application.
5. REFERENCES
[1] (2005, January 1). Retrieved from Computer Science Teachers
Association website: http://csta.acm.org/
[2] (2012, January 1). Retrieved from National Center for Women
and Information Technology website: http://www.ncwit.org/
[3] Gal-Ezer, J., Beeri, C., Harel, D., & Yehudai, A. (1995). A
high-school program in computer science. Computer Science,
28(10), 73-80.
[4] Klopfer, E. (2008). Augmented learning: Research and design
of mobile educational games. Cambridge, MA: MIT Press.
[5] Prensky, M. 2001. Digital game-based learning. New York:
McGraw Hill.
[6] Yu, Y. T. and Lau, M. F. (2006). A comparison of MC/DC,
MUMCUT and several other coverage criteria for logical
decisions. J. Syst. Softw. 79, 5 (May. 2006), 577-590. DOI=
http://dx.doi.org/10.1016/j.jss.2005.05.030.

Turi: Chatbot software for schools in the Turing Centenary
ABSTRACT
Mathew Keegan
Computer Science
Aberystwyth University
Penglais
Aberystwyth SY23 3DB
Wales
mtk8@aber.ac.uk
We describe a workshop designed for 11-19 year-olds that
considers the nature of intelligence and introduces the Turing
test in various ways.
Chatbots as mimics of intelligence are considered at length.
Pupils are invited to use our system Turi in which they can
build and test their own chatbot.
The materials are free, open source and available for all
to download [1].
Categories and Subject Descriptors
K.3.1 [Computing Milieux]: COMPUTERS AND EDU-
CATION Computer Uses in Education Miscellaneous
General Terms
Computer science education, artificial intelligence
Keywords
Alan Turing, Computer science education, Artificial Intelligence,
Chatbots
1. INTRODUCTION
The Turing Test provides a readily understood means of
discussing the nature of intelligence; we have chosen the Turing
Centenary year of 2012 to join the various schools outreach
projects based on his work. We are doing this as part
of the Technocamps project, a large, multi-site European
funded project aimed at introducing technical informatics
into schools across the convergence area of Wales [5].
Students are introduced to the difficulty of defining intelligence;
then with the addition of the Turi software, they can
get to grips with the ideas of pattern matching and callresponse
conversations through practical interaction with
∗ Corresponding author.
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
WiPSCE 2012, Hamburg, Germany
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$15.00.
Roger D. Boyle
Computer Science
Aberystwyth University
Penglais
Aberystwyth SY23 3DB
Wales
rob21@aber.ac.uk
159
Hannah M. Dee ∗
Computer Science
Aberystwyth University
Penglais
Aberystwyth SY23 3DB
Wales
hmd1@aber.ac.uk
a chatbot’s internal representation of sentences, written in
AIML (Artificial Intelligence Markup Language).
The activity we describe runs as a 4 hour workshop (but
this would be easy to adjust). The software is free, open
source and available for all to download [1].
2. BACKGROUND AND MOTIVATION
AI chatbots have been widely used in teaching, but there
is little if any use of them in a school context. We believe,
however, that they are an excellent medium for introducing
concepts of AI and for experimenting with simple pattern
matching programs. Our software facilitates this by providing
a simple multi-user chatbot editor sitting on top of an
AIML-based chatbot engine. AIML is a live AI Markup Language;
it has the same visual appearance as many markup
languages, with tags in angle brackets delimiting tokens.
Thus students are exposed to a markup language, if only
a subset of one.
There are a number of excellent chatbot programs available
to use for free on the Internet. In particular, we have
used Cleverbot [3] and Jabberwacky [4] as examples of just
how good a chatbot can be. This both motivates students
and demonstrates what they should be doing with Turi.
3. SOFTWARE IN CONTEXT
Whilst it is possible to use the Turi software as a standalone
system, we have devised a workshop suited to mixedability
pupils in the age range 11-19. We have found that
with mixed-ability groups there is a real benefit in terms
of student attention to embed computer use within a varied
program of activities, including short video clips, discussions,
pen-and-paper, and physical activities.
The day proceeds initially by introducing students to the
idea of the Turing test, exploring concepts surrounding Artificial
Intelligence, and giving practical experience of two
real chatbots:
1. The Telephonic Turing Test: introducing the idea
of the Turing test without actually using computers or
referring to AI. A student and helper leave the room
with a mobile phone. Students remaining in the class
guess which of these two has the mobile phone, by
asking questions using text messaging.
2. Is it intelligent or not?: Printed pictures of various
creatures and items are given to the class, one per student.
These include some AIs from fiction and reality,
some animals, and some wildcards (e.g. a rock, a tree,

Sherlock Holmes). Students sort themselves in order of
intelligence, which can result in some illuminating and
amusing conversations (“Is a washing machine more or
less intelligent than a flea?”).
Once ordered, students indicate if they can do certain
things (e.g., be creative). This shows that certain characteristics
are more common at the intelligent end of
the line, triggering discussions on those features that
are necessary/sufficient to call an agent intelligent.
3. Chatbot comparison: We explore the idea of chatbots,
and what kinds of conversations might expose
the fact that they are programs rather than people.
Students write down three questions to ask a chatbot,
along with a sentence describing what the question is
testing (e.g. “Does it have a memory?”). The questions
are put to two chatbots and the answers recorded;
the chatbots are then compared in discussion.
4. TURI: SOFTWARE OVERVIEW
The Turi software provides an easy way for multiple instances
of the open source chatbot Program O [2] to be constructed
and allocated to individual students, with a simple
tabular interface for students to build their chatbot by
adding and editing AIML statements. It also enables students
to chat with their own bot and those created
Turi is web-based software, written using MySQL and
PHP within the CodeIgniter framework. Client side elements
of the software are written in HTML and JavaScript,
and have been tested on all common web browsers and platforms.
The server we currently use is a Virtual Private
Server with 1GB of RAM and a dual-core processor. We
have run side-by-side workshops with 30 children in each,
meaning that 60 separate chatbots were being edited at the
same time with no appreciable load visible in the server logs.
The student interface is deliberately simple. The fundamental
unit of the chatbot is the input-output pair, which
defines an input to the chatbot and the output the chatbot
will give upon receiving that specific input. There are
three views for the student chatbot creator: New Phrase,
Try Chatbot, and Chatbot Editor.
New Phrase permits chatbot input-output pairs to be added
or edited. Students type in input and the desired response –
examples of the four main kinds of call-response supported
are (see Table 1): simple call-response, *-response, *star
and call-random. This permits users to copy-andpaste
AIML templates for the common functions.
Try Chatbot allows the user to test out phrases that have
been entered, or pre-loaded. This provides a text box to
type a conversation into, and responses are shown above the
box scrolling upwards as the conversation continues.
Chatbot Editor is the view in which users get to see all of
the input-output sentence pairs entered thus far, with the
option to delete or to edit each one.
All three views are minimalist in layout, with clear links
and simple instructions. This simplicity is the result of
user testing early in the design phase, when we found that
younger users were happiest with clear tabular layout.
Administrative functions are hidden from the student view
completely. Student chatbots are created by setting a passphrase
for the session; students start Turi in a browser by
typing the day’s passphrase into the login screen. Each time
the passphrase is typed, Turi creates a new chatbot with a
160
I/O pair Description
call-response Exact match of call generates response
*-response Wildcard matches *, for a given response.
*-star is replaced by text which
matched * in the input.
call-random Given the sentence which matches the input
call, one of a random list is selected as
response.
combination It is possible to combine these.
Table 1: The types of input-output pair covered by
Turi in our sessions
unique ID, and the student is ready to start. The instructor
has no need to determine in advance exactly how many chatbots
are required, and there is no need to give each student
a login. At the end of the session the instructor can either
drop the Turi instances for that class or can allow students
to create a name and a password for their chatbot. In the
second case students are able to continue working at home,
or over multiple sessions.
Turi has been written in English but this is easy to amend.
Chatbots created by students can be in any language with a
European alphabet – we have used chatbots successfully in
classes where some students have been creating call-response
pairs in French, German and Welsh.
5. EVALUATION, EXPERIENCE, IMPROVE-
MENTS
We have run the workshop with several hundred pupils
in the age range 12-19 (the number grows monthly). We
re-iterate that the aim is to introduce them to ideas about
AI and conversation, rather than to enable them to create
conversational agents.
The largest chatbot created during early engagements had
27 input-output pairs, and the average size of the created
chatbots had just under 10. All response types were evident
with the majority being simple. Later, we pre-seeded each
chatbot with a set of phrases showing the kinds of things it
is possible to do. This encouraged much greater creativity.
Feedback from teachers has been very positive as the material
is seen as novel and stimulating. We have found the
module to be one of the more popular workshops we run,
and disproportionately popular with girls. We have also deployed
the material at open access events for the general
public and provoked lengthy, involved interactions.
We continue to develop: in particular an inbuilt spell
checker has been found to be essential, and we will consider
employing a speech synthesiser.
6. REFERENCES
[1] H. Dee. Technocamps AI module.
http://users.aber.ac.uk/hmd1/ai.zip, 2012.
[2] E. Perreau. Program-O, 2010.
http://sourceforge.net/projects/program-o/.
[3] Rollo Carpenter. Cleverbot, 2012.
http://cleverbot.com/.
[4] Rollo Carpenter. Jabberwacky, 2012.
http://jabberwacky.com/.
[5] Technocamps. Creating the Next Generation of
Technologists. University of Swansea, 2012.
http://www.technocamps.com/.

ABSTRACT
Learning Fields in Vocational IT Education –
Why Teachers Refrain From Taking an Opportunity
Simone Opel
Didactics of Informatics
University of Duisburg-Essen
Schützenbahn 70
45127 Essen, Germany
simone.opel@uni-due.de
Vocational education in Germany is characterised by a learning
venue cooperation between vocational schools and vocational
training companies. For a better combination of theory
and practice the curricula in the field of computer science
(CS) and information and communication technologies (IT)
are arranged in so-called “Lernfelder” (learning fields), which
gives the teachers the leeway to develop different learning
situations on their own. Unfortunately, it seems that teachers
often do not put this idea into practice. The question
is: Why would vocational IT school teachers willingly relinquish
these benefits? The elicitation study described in this
paper explores the IT teachers’ knowledge of and attitudes
towards the concept of learning fields in order to find an
answer to that question. It is part of a project which aims
to develop exemplary learning situations and helpful tools
for several learning fields, which will make it easier for the
teachers to create lessons in the context of learning fields.
Categories and Subject Descriptors
K.3.2 [Computers and Education]: Computer and Information
Science Education—Computer science education,
Curriculum, Information systems education
General Terms
Human Factors, Theory
Keywords
Vocational IT Education, Computer Science Education, Learning
Fields, Learning Situations, Teachers’ Attitudes, Empirical
Study, Elicitation Study
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
Copyright 20XX ACM X-XXXXX-XX-X/XX/XX ...$15.00.
161
Torsten Brinda
Didactics of Informatics
University of Duisburg-Essen
Schützenbahn 70
45127 Essen, Germany
torsten.brinda@uni-due.de
1. INTRODUCTION
1.1 The Vocational School System in Germany
The German school system and especially the ways to
professional life differ from the systems of most other countries.
Students can attend an upper secondary school, called
“Gymnasium”, where they get a general qualification for university
entrance. Another way is attending general or intermediate
secondary schools, where students get general education
and are prepared to take up training and education
at a company and part time vocational schools (so-called
“Duale Berufsausbildung” – dual vocational education and
training). During this time students are employees of their
companies. The students’ age is between 16 and 25 years
and their previous skills are quite heteregenous. At the end
of the training students receive a vocational certificate from
the chamber of industry and commerce or the chamber of
trade. The heterogenity of the students implies different
types of learners in each class, which need different types of
instruction to benefit from lessons [3]. For this reason the
learning content in the curriculum is arranged in so-called
“Lernfelder” (learning fields). Learning fields do not include
specific aims to be reached or skills to be acquired, but they
describe different competencies which students should gain.
By alternating multidisciplinary theoretical and practical
training, students should acquire the competencies to apply
their skills in new professional situations. The idea of cooperation
between vocational schools and training companies
and the integration of work and study spreads in vocationals
systems of different countries [4] [6]. So it can be worthful
to exchange the experience with these concepts in the field
of IT.
1.2 Significance of “Learning Fields” for
IT/CS Education
A learning field contains typical business and working areas,
which are reflected and reconstructed professional activities
[2]. A learning field includes theoretical and professional
skills from different subjects as well als social and
personal competencies. The concept was developed to better
meet the requirements of all partners of the dual vocational
education and training. The different learning fields
are designed openly; it is up to the teachers to implement
the learning fields into suitable learning situations. A learning
situation is one didactically prepared working process.
It contains theoretical knowledge, several working skills and
different competencies to solve complex problems and can in-

clude difficulties from different subjects. All activity-oriented
methods can be used. That concept may cost time in the
beginning, but one part of the concept is to teach in teams,
so teachers from different disciplines can share the workload
and support each other. Concepts of interdisciplinary collaborative
teaching and learning can be found in different scenarios
of vocational training and education, e. g. in mechatronics
[8] or in computer engineering education in Egypt
[7]. In the area of IT the definition of the learning fields has
not always succeeded. Most learning fields seem to be identical
to the former subject [5]; it is tempting for IT teachers
to work with the familiar subjects without thinking further
about the specifics of the learning field. It appears that especially
in IT-centered vocational schools the concept is hardly
implemented by the teachers; we tried to find the problems
teachers meet when translating the described competencies
from several learning fields into adequate learning situations.
The question was: why would vocational school teachers refrain
from using the leeway given to them by the curriculum?
Therefore we conducted an online survey with several vocational
IT teachers 1 . In this paper we present the first part
of the results of this elicitation study. It is part of a larger
project with the purpose to convince vocational IT teachers
to devise their classes according to the learning field concept
by developing exemplary learning situations or helpful tools
for several learning fields.
2. METHODOLOGY
We asked all vocational IT teachers in Bavaria (Federal
State in Germany) to participate in the online survey. 28
teachers answered (a response rate around 30 %). The online
questionnaire consisted of three sections. First, all participants
were asked for their age, sex, vocational discipline and
their years of teaching (as range). The second section contained
16 closed questions about the IT teachers’ knowledge
of and attitudes towards the concept of learning fields. We
used a 5-point scale answer format with options from “does
not apply at all” (1) to “fully applies” (5). The third part
consisted of open questions, which explored the teachers’
attitudes toward the concept of learning fields following the
theory of planned behaviour by Aizen and Fishbein [1] and
the situation at different schools.
3. RESULTS
We got replies from 21 men and seven women with the
average age of M = 47 years. The median of the years of
teaching is within the range of 11 to 15 years. 16 teachers
are basically educated for teaching IT. The answers of the
second part of the questionnaire were summarised following
the theory of planned behaviour by Aizen and Fishbein [1].
Only the first question about familiarity with the concept of
learning fields does not represent this theory, but shows an
individual estimate of the personal skills on this topic. The
participants reported a very high familiarity with the concept
(M = 4.21). The results of the items about the teachers’s
attitudes (M = 3.36), their self-effiacy (M = 3.21) and
their subjective norms (M = 3.25) indicate that the teachers
are principally open-minded to the concept of learning
fields. Only the factor control belief (M = 2.83) shows a
lower value, which could mean that the teachers see some
difficulties to implement the concept. These results were
1 IT teachers include also teachers for computer science
162
verified by a content analysis of the answers on the open
questions. The responses to the first four questions were
paraphrased and categorized [1]. We produced N = 151
statements. The statements about the teachers’ attitudes
(n = 47, 31.1 %) specified the information of the closed questions;
the teachers described the advantages and difficulties
they see in dealing with the concept. The statements about
the teachers’ control belief (n = 102; 67.6 %) dealt with the
school equipment, the strong heterogenity of the classes or
the teaching staff composition. Only two statements about
the subjective norm were found, which praise the colleagues’
and the headmaste’s support. The examples of organization
and learning situations are not analysed yet.
4. DISCUSSION AND CONCLUSION
The question of this study was: Why would vocational
school teachers refrain from putting the concept of learning
fields into practice? The results of this study show different
aspects of the problems teachers have. The answers show
that the teachers are motivated to implement learning fields
into suitable learning situations, but it also seems that they
need support to do that. In our opinion it seems to be
helpful for the IT teachers to support them by developing
guidelines with examples for learning situations and the related
teaching material. It is also important to train the
teachers on how to develop learning situations themselves,
also in difficult environments, and how to evaluate criteria
for appropriate learning situations.
5. REFERENCES
[1] M. Ajzen and I. Fishbein. Belief, attitude, intention,
and behavior: An introduction to theory and research.
Addison-Wesley, Reading, MA, 1975.
[2] R. Bader. Das Lernfeld-Konzept in den
Rahmenlehrplänen (in German). Die berufsbildende
Schule, 50(7–8):211–213, 1998.
[3] A. Kluge, S. Ritzmann, D. Burkolter, and J. Sauer. The
interaction of drill and practice and error training with
individual differences. Cognition, Technology and Work,
13(2):103–120, 2011.
[4] C. Liang. The development research on higher
vocational education curriculum based on the working
process, volume 155 LNEE of Lecture Notes in Electrical
Engineering. Springer, Berlin Heidelberg, 2012.
[5] S. Opel. Lernfelder in der Praxis des IT-Unterrichts (in
German). vlb-akzente Berufliche Bildung in Bayern,
19(07):19–20, 2010.
[6] Y. Ren and L. Zhao. Research and practice of ’teaching,
learning, practice integration teaching model’ in higher
vocational and technical education, volume 137 AISC of
Advances in Intelligent and Soft Computing. Springer,
Berlin Heidelberg, 2012.
[7] M. Salama and T. Thabet. A curricular reform
proposal for egyptian computer engineering education
ECEE. In 2010 IEEE Transforming Engineering
Education: Creating Interdisciplinary Skills for
Complex Global Environments, 2010.
[8] S. Shooter and M. Mcneill. Interdisciplinary
collaborative learning in mechatronics at bucknell
university. Journal of Engineering Education,
91(3):339–344, 2002.

ABSTRACT
This study examines the potential for rethinking dated educational
technologies. The mechanised Turtle Robot is taken as a test case
to examine whether dated educational technologies can be renewed
as a means of maximizing the tools and research of the
past towards the new wave of interest in computing education.
This paper will present research in progress and explore the Turtle
Robot and other educational tools in the context of Technocamps,
a Wales-based project aimed at inspiring young people 11-19
years in computing.
General Terms
Experimentation, Human Factors, Languages
Keywords
Turtle Robot, Logo, Papert, Learning Tools, Programming, Electronics,
Out-of-the-box, Teachers, Educators, Learners, Renew,
Reuse, Recycle, Technocamps
1. INTRODUCTION
Keen not to contribute to the computing wasteland, this paper
presents the findings of examining alternative approaches to
learning from past educational tools. Using the Turtle Robot as a
test case, the Turtle Robot is explored as a potential for contemporary
educational engagement.
A recent report reviewing the BBC’s Computer Literacy Programme
examines the climate that led to the development of the
BBC Micro [1] as a way of identifying the various influencing
factors which bring about radical change in computing education.
This study examines another seminal educational tool, the Turtle
Robot, and specifically focuses on the mechanism of the robot as
a potential contemporary teaching aid rather than the approach
behind the robot.
In exploring this outdated but very prevalent technology still
available in classrooms today, the study examines whether the
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists, requires
prior specific permission and/or a fee.
Conference’04, Month 1–2, 2004, City, State, Country.
Copyright 2004 ACM 1-58113-000-0/00/0004…$5.00.
Save Our Turtle Robots?
Emma Posey
Technocamps Computer Science
College of Science Swansea University
Singleton Park, Swansea
emma.posey@technocamps.com
163
robot mechanism has any relevancy, potentially through modification,
as a learning tool.
Comparisons are made with other robot simulators and their connection
to real robots, especially through Technocamps, a project
engaging young people in computing.
2. TURTLE ROBOTS
As part of the drive to encourage learning using computers, the
Turtle Robot was developed using Logo, a computer language
developed as a learning tool by Wally Feurzeig and Seymour
Papert. Papert was a keen early advocate of learning via computers
and he believed computers, and specifically Logo, could
help young people plan, problem-solve, as well as develop critical
thinking and logic.
Papert initiated the Turtle Robots, a 3D robot in the form of a
turtle. The turtle, Papert claimed, was useful as it was an object
and therefore could be understood in real terms. Critically, Logo
as a computer language, allows the user/programmer to introduce
new words to the program – effectively developing a personalised
vocabulary to define new procedures and commands. Papert
identified the benefits of computational thinking [3] [4], a literacy
which helps young people’s thinking to be ‘step-by-step, literal,
mechanical’. An agreed definition for computation thinking and
ways in which it can be taught continue to be a challenge [5].
The ‘Turtle Graphics’ project [6] used in the educational tool
Scratch [7] from MIT Media Lab and others such as RoboMind
[8] use a simulated robot to engage young people in programming,
one using a graphical drag and drop, the other syntax.
AberBots, modified by Technocamps Aberystwyth, is a robot
simulator whose program runs on real research robots, the Pioneer
and IDRIS. The AberBots program is scripted and as with the
Turtle Robot, there is a direct relationship between program, its
simulation and a real robot. The study compares these various
programs and other programmable toys and draws on findings to
analyse their ability to engage young people in programming
using simulated and real robots.
With the ability these days to build and program one’s own creative
machines using products such as LEGO/Logo and more recently
Lego Mindstorms, the ready-made object of the Turtle
Robot may seem too restrictive or prescriptive in today’s terms.

3. NEW WAVE EDUCATIONAL TOOLS
There is a wave of interest and political will within the UK for
‘real’ computing in education, (rather than the predominant ICT
culture) [9] [10].
The recent move to overhaul computer education has computational
thinking at its heart [12]. There is a danger that older tools,
resources and experience get swept aside for the new. Especially
in computing, older technologies and learning tools become
quickly disregarded and discarded. The computer wasteland is
vast with an increasing rate of obsolescence in the industry.
Yet there are Turtle Robots hibernating in schools, gathering dust
whilst schools with ‘Green Flag’ initiatives and ‘Eco Gangs’ aim
to apply the mantra of ‘Renew, Reuse, Recycle’ across all that
they do. There are numerous sensors and motors available in discarded
toys and computer hardware. In educational, environmental
and resource terms the exploration of existing technology
in schools makes real sense.
Further, a key obstacle or ‘grand challenge’ in getting computing
back in to schools is not the examining boards and curriculums,
but the skill base of ICT teachers, some with little or no training
in ICT let alone computing and many with little time or inclination.
However, there are some teachers and educators who have had
previous involvement with computing and they refer back to Logo
and the Turtle Robot (as well as the BBC Micro) as a starting
point in their own development. Those with some previous experience
of Logo may be confronted with new programs to learn –
revisiting a key reference point, the Turtle Robot, may be a useful
starting point in their resumed computing in education journey.
4. TECHNOCAMPS
Technocamps is a European project aimed at inspiring 11 to 19
year olds to engage with computer programming and electronics.
The three-year project is based in Wales
(www.technocamps.com). Technocamps has been exploring ways
in which young people can be encouraged to think imaginatively
about computing.
At Aberystwyth Technocamps we are particularly keen on outside-the-box
technology and most of the devised workshops
closely follow Papert’s aim of encouraging young people to control
and manipulate computers in the world rather than on the
screen.
164
5. SAVE OUR TURTLE ROBOTS?
The research project has explored the benefits of the ready-made
Turtle Robot machine and whether working with existing equipment
can save time and resources for schools. The study explores
various ways to modify the design of the Turtle Robot so that is
can be utilized as a relevant contemporary learning tool. It also
investigates other, more recent, programming languages, environments
and hardware with a view to unearthing any lasting
benefits to rethinking this outdated technology.
6. ACKNOWLEDGMENTS
My thanks to Technocamps for allowing me to utilise the research
and findings of the project through its modules and tracking with
specific thanks to Dominic Roberts, Jonathan Roscoe, and Barry
Thomas.
7. REFERENCES
1. Blyth. T., The Legacy of the BBC Micro, NESTA, May
2012
2. Papert. S., 1980. Mindstorms: Children, Computers,
and Powerful Ideas, The Harvester Press
3. Wing. J., Computational Thinking, Communication of
the ACM, Viewpoint, Vol. 49, No. 3, pp33-35, March
2006
4. Barr. V., and and Stephenson. C., Bridging Computational
Thinking to K-12: What is Involved and What is
the Role of the Computer Science Education Community?,
ACM Inroads, Vol. 2(1), 2011, 48-54.
5. diSessa. A., Changing Minds: computers, learning and
literacy, Cambridge, MA: MIT Press, 2000.
6. http://scratch.redware.com/project/turtle-graphics
7. http://scratch.mit.edu
8. http://www.technocamps.com/resources
9. The Royal Society, Shut down or restart?, The way
forward for computing in UK schools, January 2012
10. Livingston. I., and Hope. A., Next Gen: Transforming
the UK in to the World’s leading talent hub for the
video games and visual effects industries, NESTA, February
2011
11. Computing: A curriculum for schools, Computing at
Schools, Body of Knowledge Working Group, March
2011