Big Data Characterization of Learner Behaviour in a Highly Technical MOOC Engineering Course

MOOCs attract a large number of learners with largely unknown diversity in terms of motivation, ability, and goals. To understand more about learners in highly technical engineering MOOCs, this study investigates patterns of learners’ (n = 337) behaviour and performance in the Nanophotonic Modelling MOOC, offered through nanoHUB-U. The authors explored clusters of learner clickstream patterns using the k-means++ algorithm and found five clusters of learner behaviour, labelled according to learners’ use of materials: Fully Engaged, Consistent Viewers, One-Week Engaged, Two-Week Engaged, and Sporadic learners. The Kruskal-Wallis nonparametric statistical test yielded a significant difference (p< 0.01) between learners’ access of course materials in each cluster. The researchers then examined the participation and mean scores on course quizzes and exams for each learner group. One-Week Engaged learners, on average, scored significantly lower on the first week’s assessment. Two-Week Engaged learners, on average, scored significantly lower on the second week’s assessments. Other differences found in learners’ participation and performance on quizzes and tests based on the five clusters are discussed. These findings suggest that some of the high dropout numbers in advanced MOOCs may be related to learners’ performance on course assessments. In addition, integration of learner access to course material with course assessment scores provides a much richer understanding of learners in a MOOC.