Discovery of study patterns that impacts students’ discussion performance in forum assignments

Abstract

Student-centered courses rely on the active participation of the students in forum assignments. In this work, we investigate a course where the forum assignment discusses a clinical case among professional students (N = 94). We propose a method to discover navigation patterns related to performance grades, using behavioral actions in an LMS platform. We selected a set of significant course actions and built per-user sequences along the course module. Then, we applied the GSP algorithm to identify ordered patterns from this navigational data. The identified patterns were then used as features for a linear regression model, to predict the assignments’ performance, graded manually by the teachers, and controlling for factors that may influence it. Results show some rules correlated to the students’ performances. These results can be used to better inform course designers on how to improve the courseware and instructors on how to better guide their students.