Predicting course completion in massive open online courses
- Between 2012 and 2016, millions of learners have enrolled in massive open online courses (MOOCs). Survey data show that most learners intended to complete their courses, but analysis of the course data shows that only a small fraction of them complete the courses. Educators are trying to support learners before they drop out, but this requires early prediction of who will complete the course and who will not. This thesis presents three techniques for predicting course completion in MOOCs based on data from three different sources: learner demographics, learner engagement with course materials, and learner interactions in course videos. The thesis shows that engagement data provide the highest prediction performance among the three data types, and that prediction performance improves the later in the course the prediction is performed. The results show that the prediction models in this thesis are a step towards in-course support of learners at risk of non-completion.
|Type of resource
|electronic; electronic resource; remote
|1 online resource.
|Halawa, Sherif A
|Stanford University, Department of Electrical Engineering.
|Statement of responsibility
|Sherif A. Halawa.
|Submitted to the Department of Electrical Engineering.
|Thesis (Ph.D.)--Stanford University, 2016.
- © 2016 by Sherif Adel Mahmoud Halawa
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...