Student Performance Pattern Analysis Across Modalities in an Introduction to Computing Course

Interaction patterns between students in on-campus and MOOC-style online courses have been broadly studied for the last 11 years. However, there is very little analysis comparing the habits of students completing the same course offered in both a traditional on-campus and MOOC-style online formats. Analyzing the differences in student approach in a direct comparison across modalities will allow researchers to identify effective cross-modality interventions. This study looked at clickstream interaction patterns for students in an ’Introduction to Programming in Python’ edX course representing 7 traditional semester-cadence course instances offered to oncampus students and 4 large-scale MOOC-style course instances offered to all learners to determine what, if any, interaction and performance patterns exist between the two modalities. While significant research has been done to determine the difference between learners in separate modalities, some overlap in student interaction and performance patterns was found, with consistent interaction patterns across modalities and course instances for high-performing students. While students in the MOOC cohorts initially under-performed when compared to the on-campus cohorts, final scores were closer across modalities. Even though MOOC students are afforded the opportunity to finish the course in a much shorter time frame than on-campus students, those who complete the course put forth more effort, consume more course content, take more attempts on each problem set, and still perform worse overall. These differences in student interaction and performance patterns need additional research to determine which interventions would be transferable across modalities to address at-risk student performance.