Learner Paths and Trajectories in MOOCs

In economics and business, MOOCs are developing fast and this evolution raises many questions. The most publicised facts concern the dropout rate: only a small proportion of students finish courses. All courses experience the same phenomenon, which probably depends on factors like the duration and length of the course at least as much as on its content, but there may be more systemic factors. In order to delve into the question of course efficiency, we record all that individual learners do in a recent course hosted by FutureLearn: “manage your prices, an introduction to revenue management and pricing strategy.” We have a wealth of quantitative information on learners’ behaviour, including performance measurement, time spent on resources, steps completed, messages posted, etc. Qualitative information, like comments made by learners, connexions between them, is also rich in content. We focus here on the available information on learner paths. In what respect do they enable us to assess the course efficiency? We design indicators that can be computed with available data and analyse different types of learner paths. Are social learners (those who participate in discussions) more likely to complete the course? More generally, what can we discover from the individual learning paths? Do students need more than one session? What would we like to have in order to model the course efficiency better? This paper is an attempt towards a general reflection on the use of data for answering these questions.