The project explores how human-AI systems use data science insights in the context of teaching and learning in higher education. We create and evaluate educational situations where partners of the joint human-AI system have varying access to data insights about the student process of learning. 

To empirically examine how learners make use of the data insights, we compare learning outcomes that result from different types of how individuals make sense of their learning process presented to them as learning analytics. We complement learning analytics dashboards with interactivity features that include AI-powered agents prompted around various pedagogical tacticts and compare how these interactive pedagogically enhanced learning analytics affect learner judgement and behaviour in their studies. 

We use ARTEMIS, an authentic learning technology adopted in TUM's programming courses, which enables collection of real-time, process-level student data as well as affords A/B testing. ARTEMIS has been developed by the Professorship of Applied Software Engineering and is successfully adopted to teach large programming classes.

The project contributes a generic framework of how student decision making with learning data in higher education can be designed for using AI technologies.

The project has received competitive funding from the Munich Data Science Institute.

MDSI: https://www.mdsi.tum.de/en/mdsi/research/funding-support/seed-funds/hybridd3/

CIT Collaborator: https://ase.cit.tum.de

Artemis: https://artemis.cit.tum.de/courses

Contact Laura Graf for more details: l.graf@tum.de