Malaysia MOOC: Improving low student retention with predictive analytics

Abstract

Massive Open Online Courses MOOCs have become more acceptable as a learning program globally, including Malaysia. One main issue that has been discussed since the implementation of MOOCs is the issue of low student retention or high dropout rates from the course. Various factors have been found to play a role in this issue including the interaction factor. Previous studies have experimented with various strategies to monitor student retention and apply intervention programs to improve the situation. The strategies include the usage of machine learning and data mining techniques in analysing students' online interactions to predict student retention rates. The implementation of these strategies produced promising result. However, in Malaysia, these strategies are not really implemented yet. Therefore, this paper discusses the issue of student retention in MOOCs, explores possible intervention plans using data mining and its suitability with the current platforms used for MOOCs. The proposed method includes predictive analytics that involves classification analysis. This paper suggests that the method can be applied to the current platform and complement intervention programs for the issue of low retention or high dropouts with several improvements.