News

Self-regulated learning (SRL) is central to academic success because it is associated with resilience, motivation, and essential skills such as time management and critical thinking. However, it is challenging to measure. While questionnaires are prone to bias and log data is superficial, interviews were considered too resource-intensive.

Testing in large courses

Niels Seidel and Slavisa Radovic are now testing a multi-agent system (MAS) in a funded project that automatically conducts structured SRL interviews. Based on established interview protocols, the multi-agent system then generates individual feedback with learning strategy recommendations. In two courses with over 200 students each, the reliability is being examined in comparison to established methods and SRL support.

In-depth measurement methods practical for the first time

Why is this so promising? Automated interviews using MAS make in-depth measurement methods practical for the first time. Learners can thus receive individual feedback and recommendations for improving their self-regulation.

In earlier work, Seidel and Radovic have already shown that their prototype works efficiently. The current project aims to help students better assess their self-regulation skills and develop them in a targeted manner. This project is relevant for use in university teaching in all departments and types of higher education institutions. It can contribute to greater equity in opportunities in higher education by providing better and more tailored support services.