(SpringerBriefs in Business Process Management) Learning Analytics Cookbook_ How to Support Learning Processes Through Data Analytics and Visualizatio
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
12 2 Learning Analytics Kitchen
sometimes the only digital data is a final grade. Teachers also tend to build their
appraisals of students in an intuitive and experience-based way instead of an
objective data-based, evidence-based way. More data would be the key to creating
more holistic profiles of the learners, which requires collecting information about
learners from various sources and storing, interpreting, and aggregating it.
For example, learning analytics uses datasets from institutions that are protected
from external access and use. However, an increasing number of open and linked
data sources from governments and organizations like the Organisation of Economic
Cooperation and Development (OECD) can be used to investigate target groups for
certain courses or programs (d’Aquin et al. 2014). Among the providers and users of
these closed and open datasets is a movement toward more standardized metadata
for learning analytics (i.e., new specifications for learning technology and frameworks
like xAPI 1 and IMS Caliper 2 ). The use of such metadata standards allows data
to be combined and results gained in various scientific disciplines and educational
scenarios to be compared (Berg et al. 2016). A comprehensive uptake of such data
could lead to a paradigm shift in educational science, a field that is currently more
accustomed to small-scale experimental studies than to big-data-driven ones like
those performed at Google and Facebook (Kramer et al. 2014). Apart from strengthening
the research side with standardized educational metadata and being able to
convert this data into one unified format, such standards could also create a market
space for educational tools and open educational resources and their analytics.
2.5 Instruments: Technologies, Algorithms, and Theories
That Carry Learning Analytics
Several technologies can be applied to the development of educational services and
applications that support educational stakeholders’ objectives (Drachsler et al.
2015). Learning analytics takes advantage of machine learning (where computers
can “learn” from patterns), analysis of social networks, and classical statistical
analysis and visualization techniques (Fazeli et al. 2014; Scheffel et al. 2017a).
Through these technologies, learning analytics can contribute tailored information to
stakeholders and report on demand. For instance, learning analytics can be applied to
develop a system that identifies students who are in danger of dropping out.
1 http://tincanapi.com/
2 http://imsglobal.org/caliper/index.html