On Monday 8 April 2019, I attended an event at the teaching lab at TU Delft about learning analytics in higher education. A number of workshops, presentations and panels explored this topic.
Opening and keynote
Marcus Specht opened the event, followed by a keynote by Hendrik Drachsler.
Learning Analytics Workshops
It was a tough choice between four workshops:
- Learning Analytics Dashboards (Ioana Jivet and Tom Broos)
- Design Thinking for Learning Analytics (Christian Glahn)
- Multimodal Learning Analytics (Daniele di Mitri and Marcus Speech)
- Trusted Learning Analytics Infrastructure (Stefaan Termier and Maren Scheffel)
I chose the multimodal learning analytics as it sounded very unique, and it was. Daniele talked us through some really interesting projects that are capturing traces of the learning process with wearable sensors, Internet of Things devices and high-frequency data collection technologies (called multimodal data).
“Through signal processing, machine learning and experience sampling, it is possible to use multimodal data to tailor personalise multisensorial feedback. Such feedback could steer the learner to optimise learning goals attainment or skills acquisition.”Daniele di Mitri and Marcus Specht, Multimodal Learning Analytics
Two of the examples of application they showed us were learning calligraphy, and public speaking. The latter was particularly interesting for me, as I teach this to students.
The event was rounded off with a panel discussion about the future of learning analytics. The panel was comprised of:
- Claudia Hauff, Associate Professor, EEMCS TU Delft
- Erna Kotkamp, Project manger ICT, Directorate ICT-FM TU Delft
- Tinne de Laet, Associate Professor, Engineering Science KU LEuven
- Heather Hamilton, Brightspace
Much of the discussions during the panel and the day in general related keeping the student’s needs central and being transparent. GDPR is a big topic, as it brings both expectations and legal requirements into data minimisation and clarity of purpose. The impact of culture in education was also covered.
My perspective as a lecturer
As a lecturer for almost 5 years now at The Hague University of Applied Sciences, my own take on this subject is that I would really like to have more data (and learning analytics) to be able to better understand my students and tailor their education as much as possible. The current systems we use (Blackboard, Osiris) provide very limited information. I’ve actively used tools such as Socrative, Google forms (surveys) and simply talking to students to try to gather some kind of analytics, but this is currently very limited.
As I know what is possible for example, the great learning analytics provided via the Udemy course I have, it often feels like I’m currently “teaching blind”.
I do understand student concerns about sharing data but I feel the key is, as was discussed, in being transparent. In learning, as in our digitalisation in society in general, giving informed consent to “trade” some level of data can lead to a better education.
What are your thoughts on Learning Analytics?