Analyzing students’ online presence in undergraduate courses using Clustering

Abstract

The Higher Education Institute (HEI) are experiencing a major paradigm shift due to recent global pandemic. A sudden shift from face-to-face (F2F) and blended modes of study to completely online mode of delivery has introduced hidden challenges to facilitators and students alike. Student’s online engagement has become even more important for their academic success as F2F component is not there in most cases. Therefore, there is a need to investigate the effects of the various indicators of students’ online presence towards their academic performance. This paper explores the effectiveness of online presence in HEI where Covid-19 has shifted the course deliveries to fully online mode. Previously, Online Measurable Presence Model (OMPM) was used to find students effectiveness in a blended learning environment where two indicators used were Frequency and Duration. The chosen indicator in this research is frequency, which will be adequately used to quantify the effectiveness of the online presence in two mathematics courses in the Pacific. Clustering technique is used to create clusters of Frequency and see their relation to OMPM model. Prediction is made using neural network to see the accuracy based on model. The clusters would allow to build predictive models to predict future outcomes or occurrences and student performances, with a major focus on mathematics courses.