Using machine learning techniques to understand student variables that are linked to completion in MOOCs

Jacqueline Wong

Gregori, Zhang, Galván-Fernández, and de Asís Fernández-Navarro (2018) examined the learner-support strategies that assist students in completing formal, conventional, and professional Massive Open Online Courses. Using machine learning modelling, Semi-Supervised Extreme Learning Machine (SSELM), they found that the number of days students accessed the course was an important factor that related to completion of the MOOCs. In particular, students who passed the MOOCs were also the ones who accessed the MOOC platform the most number of times during the second quartile. The authors suggested that implementing an early warning system is crucial during the second quartile of the course. They did not find students’ interactions with videos (i.e., pause, play, and stop) to be a significant factor to students’ completion of the MOOCs. A possible reason for the non-significant contribution of video interactions o students’ completion was that the videos were not engaging and students were more ‘doers’ than ‘watchers’.

The results of the study showed that maintaining student engagement before the second quartile and instructors’ presence are two important aspects that relate to student retention and success. They suggested that the instructors’ presence can help to maintain student engagement, as such, instructors are encouraged to invest effort in maintain student engagement beyond the first quartile. Instructors can do by commenting at the beginning of each module to help students recall necessary information, encourage students to participate in forums, and respond to students’ feedback, comments and concerns. 

Gregori, E. B., Zhang, J., Galván-Fernández, C., & de Asís Fernández-Navarro, F. (2018). Learner support in MOOCs: Identifying variables linked to completion. Computers & Education122, 153-168.


More information…