The behavioural data analysis from Adaptive Learning Platforms (ALPs) is used to create Learning Analytical Dashboards (LADs), which can provide teachers with an overview and recommend pedagogical actions as feedback on students’ behaviour. Existing research focuses on demonstrating ALP value through retaining student engagement and prediction of performance, designing multi-modal data, applying algorithms, and improving software and learning systems. Still, it lacks methods for evaluating ALP’s LADs as they are applied with a pedagogical aim. Our study addresses the empirical research gap of pedagogically grounded ALP LAD prototypes. The pro-totypes analyse student behaviour, aiming at enhancing ALP functionality and assisting teachers in pedagogical reflection. We analysed activity logs from 397 nursing students of Denmark’s University College Absalon from Fall 2020 to Fall 2023 using R. Our study explains students’ behaviour, content difficulty, meta-cognition, and performance through two LAD prototypes and aligns insights with learning theories. Results show that while ‘Activity Labyrinths’ occurrences are infrequent, our first LAD prototype enables teacher awareness of autonomy-related motivational issues. In the second LAD prototype, teachers find the LAD difficult to comprehend. Additionally, the study addresses the black box issue which hinders the design of ALP LADs by utilising behavioural data along with qualitative data in prototyping pedagogically founded LAD prototypes.
Se artiklen her