EdX Insights Data Science Capstone Project
This project examined data from student clicks (including click type, clock time, and time in video) on videos (n=323) in three MOOCs (Massive Open Online Courses) to determine if there were distinct viewing patterns among groups of students. We sought to predict, based on the click data, whether or not a student was likely to finish the course. The analysis suggests that a prediction would depend on multiple factors, or features. To date, the best-performing feature set included total click rate, fast-forward rate, rewind-rate, and the average percentage of videos watched, [data removed]. These results are promising but inconclusive.