Data analytics to predict dropout in a MOOC course on energy sustainability

Abstract

Massive open online courses (MOOCs) offer multiple advantages and vast training possibilities in diverse topics for millions of people worldwide to continue their education. However, dropout rates are high; thus, it is important to continue investigating the reasons for dropout to implement new and better strategies to increase course completions. The present study aimed to analyze the data of a MOOC class on energy sustainability to know why students drop out, identify causes, and predict dropouts in future courses. The method used was Knowledge Discovery in Databases to analyze association rules in the data. Using the Mexico X platform, an initial, validated survey instrument was applied to 1506 students enrolled in the MOOC course "Conventional Clean Energy and its Technology." The results indicated that association rules allowed identifying participants' behavior according to the type of responses with a determined confidence level. Also, the association rules were appropriate for working with a large amount of data. In the present case, results of up to 86% confidence were obtained based on the rules. This research can be of value to decision-makers, teachers, researchers, designers, and those interested in large-scale training environments