Early Prediction of Dropout and Final Exam Performance in an Online Statistics Course

Abstract

Higher education students who either do not complete the courses they have enrolled on or interrupt their studies indefinitely remain a major concern for practitioners and researchers. Within each course, early prediction of student dropout helps teachers to intervene in time to reduce dropout rates. Early prediction of course achievement helps teachers suggest new learning materials aimed at preventing at-risk students from failing or not completing the course. Several machine learning techniques have been used to classify or predict at-risk students, including tree-based methods, which, though not the best performers, are easy to interpret. This study presents two procedures for identifying at-risk students (dropout-prone and non-achievers) early on in an online university statistics course. These enable us to understand how classifiers work. We found that student dropout and course performance prediction was only determined by their performance in the first half of the formative quizzes. Nevertheless, other elements of participation on the virtual campus were initially considered. The classifiers will serve as a reference for intervention, despite their moderate performance metrics.