A moderated mediation model on how interaction and early engagement enhance students’ learning in MOOC environment

MOOCs (Massive Open Online Courses) experienced exponential growth during the COVID-19 pandemic, and their relevance in education continues today. However, research on improving learning outcomes in MOOCs remains limited. This study addresses this gap by collecting data from 25,745 learners (with 11,904 learners that fit the analytical criteria) enrolled in a computer science course on Coursera. It examines learners’ clickstream behaviors, homework participation, and forum discussions to explore how these behaviors relate to final learning outcomes. The findings reveal that overall learning performance is predicted by learners’ engagement, with this relationship mediated by interaction with peers and instructor and moderated by early engagement with course materials. Interestingly, early-engaging solvers—those prioritizing assignments over video viewing—demonstrate a stronger positive moderation effect on this link. These results provide new insights into how engagement patterns and interactions influence learning outcomes, offering practical implications for improving pedagogy in MOOCs and other online learning platforms.