Identifying Learner Problems Framed within MOOC Learning Designs

Detecting learners who face problems in MOOCs usually poses difficulties due to the high instructor-learners ratio, the diversity of the population, and the asynchronous participation mode. Existing solutions mainly draw on self-reported problems in discussion forums and on dashboards displaying learners’ activity traces. However, these approaches cannot scale up easily or do not consider the course learning design. This paper presents a conceptual framework aimed at guiding MOOC instructors in the identification of potential learners' problems and indicators of such problems, considering the learning design of the course (e.g., types of activities, difficulty, etc.). An instrumental qualitative case study served for the evaluation and refinement of the framework. The results showed that the framework positively helped instructors to reflect on potential learners’ problems they had not considered beforehand, and to associate such problems with a set of indicators related to their learning designs.