Title:Resource Usage Analysis from a Different Perspective on MOOC Dropout

Abstract:We present a novel learning analytics approach, for analyzing the usage of resources in MOOCs. Our target stakeholders are the course designers who aim to evaluate their learning materials. In order to gain insight into the way educational resources are used, we view dropout behaviour in an atypical manner: Instead of using it as an indicator of failure, we use it as a mean to compute other features. For this purpose, we developed a prototype, called RUAF, that can be applied to the data format provided by FutureLearn. As a proof of concept, we perform a study by applying this tool to the interaction data of learners from four MOOCs. We also study the quality of our computations, by comparing them to existing process mining approaches. We present results that highlight patterns showing how learners use resources. We also show examples of practical conclusions a course designer may benefit from.