Abstract

Past studies have shown that student engagement in Massive Open Online Courses (MOOCs) could be used to identify at-risk students (students with drop-out tendency). Some studies have further considered student diversity by looking into subgroup behavior. Yet, most of them lack consideration of students’ behavioral changes along the course. Towards bridging the gap, this study clusters students based on both their interaction with the system and their characteristics and explores how their cluster membership changes along the course. The result shows that students’ cluster membership changes significantly in the first half of the course and stabilized in the second half of the course. Our findings provide insight into how students may be engaged in learning on MOOC platforms and suggest the improvement of identifying at-risk students based on their temporal data.

Recommended Citation

Yang, B., Shi, L. & Toda, A. M. (2019). Demographical Changes of Student Subgroups in MOOCs: Towards Predicting At-Risk Students. In A. Siarheyeva, C. Barry, M. Lang, H. Linger, & C. Schneider (Eds.), Information Systems Development: Information Systems Beyond 2020 (ISD2019 Proceedings). Toulon, France: ISEN Yncréa Méditerranée.

Paper Type

Event

Share

COinS
 

Demographical Changes of Student Subgroups in MOOCs: Towards Predicting At-Risk Students

Past studies have shown that student engagement in Massive Open Online Courses (MOOCs) could be used to identify at-risk students (students with drop-out tendency). Some studies have further considered student diversity by looking into subgroup behavior. Yet, most of them lack consideration of students’ behavioral changes along the course. Towards bridging the gap, this study clusters students based on both their interaction with the system and their characteristics and explores how their cluster membership changes along the course. The result shows that students’ cluster membership changes significantly in the first half of the course and stabilized in the second half of the course. Our findings provide insight into how students may be engaged in learning on MOOC platforms and suggest the improvement of identifying at-risk students based on their temporal data.