Learning analytics can support teachers by providing statistics based on the data traces that students generate when interacting with online learning environments. MOOC platforms generate vast amounts of learner data, but it’s the teacher’s responsibility to obtain, prepare and analyse this data. ELAT is a browser-based, local processing tool that attempts to minimize the first two, so teachers and researchers can focus on analysis.
About the project
MOOCs log information with every click of every student, such as what they clicked, in which page, and when. In edX courses, this data is available for the course instructors but it’s just a large file with series of clicks, or events, for every student in every course running in their institution.
Processing this data to obtain useful information requires
a) technological infrastructure,
b) moderate programming skills,
c) domain knowledge to aggregate the events,
d) time
Not every MOOC instructor has all four to spare. While there are some solutions that can take care of domain knowledge and processing time, they still require IT-related skills such as setting up a database and/or a programming environment. ELAT can do all of this in the browser (no install required), and locally (data stays in the user’s computer), creating a dashboard and structured files for deeper analysis.
Expected outcome
ELAT attempts to fulfil a need for simple processing for Learning Analytics, in the study, we review existing solutions for this problem, and we found that most of the tools in published work are outdated and simply can not be used. Others require institution-wide efforts, both in IT-infrastructure and setup, this analysis is available in the publication for LAK20. In addition, the project is based on MOOCdb, which defines a session-based database schema to aggregate the list of isolated events into meaningful learner interactions with the educational material. The publication also includes a technological evaluation, where we push the processing limits of ELAT, as well as a user-study with potential end users. Future work with ELAT consists mainly of software sustainability and expansion to other platforms, as well as improvement on the dashboard and out-of-the-box analytics.
Partners
Delft Data Science, Faculty EEMCS - TU Delft
Extension School - TU Delft
