RT info:eu-repo/semantics/conferenceObject
T1 Supporting Group Formation in Ongoing MOOCs using Actionable Predictive Models
A1 Er, Erkan
A1 Gómez Sánchez, Eduardo
A1 Bote Lorenzo, Miguel Luis
A1 Asensio Pérez, Juan Ignacio
A1 Dimitriadis Damoulis, Ioannis
AB Although the massive and open nature ofMOOCs necessitates more technology support for instructors,existing prediction research has been barely capable of offeringreal-world solutions. One critical case where MOOCinstructors could benefit from real-time technological supportis the design of collaborative activities, in particular groupformation. In this regard, this research work investigated theuse of in-situ learning technique to produce useful andactionable information that could assist instructors in groupformation while the course continues. Focusing on a particularMOOC context, a predictive model was created to compute theprobability that students would participate in groupdiscussions or not. Using these probability scores, actual groupbehavior was also predicted for three cases: at least 2, 3 or 4different students would post in group discussions. Accordingto the results, the model was able to accurately predictindividual student behavior as well as group behavior beforethe actual collaborative activity had taken place, suggesting itspotential for real-time use. Future research involves theexploration of other approaches for creating actionablepredictions and the application of the predictions in practice inan ongoing MOOC.
YR 2018
FD 2018
LK http://uvadoc.uva.es/handle/10324/31407
UL http://uvadoc.uva.es/handle/10324/31407
LA eng
DS UVaDOC
RD 20-sep-2024