Factors Affecting Student Satisfaction Towards Online Teaching: A Machine Learning Approach

Abstract

During the outbreak of the Covid-19 pandemic, universities were forced to adopt technology and collaboration tools to reinforce online teaching and sustain their operations. This radical change pushes universities, researchers, educators, practitioners and decision makers to explore the perceptions of students and provide high quality online teaching operations. This study offers an understanding of the factors influencing students' satisfaction with online teaching. Using data from an institutional survey, a machine learning approach is developed along with feature importance analysis using Permutation Importance and SHAP. The two techniques yielded similar results, where quality, interaction, and comprehension were the most significant predictors of satisfaction while student class, gender and nationality were insignificant. Such results support previous research conducted on similar data but with different statistical techniques. Other factors might be significant in the online environment such as student support, academic experience, and assessment.