Eduardo Gómez-Sánchez, from GSIC-EMIC research group, attended the Learning @ Scale 2019 held in Chicago, IL, USA from the 24th to the 20th of June 2019. Eduardo participated in the poster session presenting the work entitled “Informing the Design of Collaborative Activities in MOOCs using Actionable Predictions“.
This paper introduces an approach to translate the student engagement data from two runs of a MOOC into actionable information to support instructors’ design decisions of collaborative activities. Predictions regarding student participation in group discussions are considered actionable information. Machine learning models were trained using two transfer learning techniques to generate the intended predictions before the activity so that they could be operationalized by the instructors beforehand, which is infeasible with post-hoc approaches common in the MOOC literature. The models accurately predicted learners’ engagement. Through several learning scenarios, it is illustrated how these predictions could assist instructors in creating designs that are feasible considering the characteristics of a particular context (e.g., choosing the optimal group size, setting re-quirements affordable by the groups).
Er, E., Gómez-Sánchez, E., Bote-Lorenzo, M.L., Asensio-Pérez, J.I., Dimitriadis, Y. Informing the Design of Collaborative Activities in MOOCs using Actionable Predictions Proceedings of the Sixth ACM Conference on Learning @ Scale, L@S 2019, Chicago, IL, USA, June 2019.