Predicting Student Performance by Modeling Participation in Online Discussions in Introductory Online Mathematics Courses

• In Event: Considerations in Distance Learning

Abstract

This study investigated both the nature of students’ participation in online discussions as well as the content of their posts in terms of impact on student performance. Using emerging data mining techniques, we analyzed online discussion data automatically collected by a Learning Management System from 2,869 students enrolled in 72 courses contributed over 20,000 posts. Results showed that online listening (non-posting) behaviors significantly predicted student course performance. Further, the content of posts which related to allocentric elaboration (taking other peers’ contributions) and application of new knowledge, showed the highest predictive value of student course performance. The findings contribute to our understanding of students’ participation in online discussions by considering the quantity, breadth, and content of their interactions.

Authors

• Ji-Eun Lee, Worcester Polytechnic Institute

• Mimi M. Recker, Utah State University