A Case Study via Bayesian Network: Investigating Factors Influencing Student Academic Performance in Online Teaching and Learning During COVID-19 Pandemic

COVID-19 pandemic has an impact on numerous Malaysian industries, particularly the education sector. Due to the changes in class delivery during the pandemic, students’ academic performance has been impacted, for instance, by decrease in CPGA. Hence, we would like to determine the variables that could influence academic performance of the undergraduate students in Malaysia during COVID-19 pandemic, whereby educators may be able to help the students to adapt the changes. In this study, 27 items categorized into ten factors are investigated, and Bayesian network (BN) is employed to discover the determinants and forecast the students’ academic performances. The accuracy of prediction between BN and support vector machine (SVM) model is compared. The findings indicate a significant relationship between students’ academic achievement and the factors including the type of learning device, faculty, attendance, age, self-efficacy, and teaching technique. Additionally, the MAPE for BN is 5% less than that of SVR in comparison with the actual results. This study assists higher education institutions to effectively strengthen the significant variables to raise students’ academic performance, especially for OTL.