Predictive Analytics for Students’ Quantitative Class Performance and Its Impact on Classroom Operations
Abstract. The application of predictive modeling in the field of education has steadily gained interest and acceptance in recent years. In this study, students’ course grade is predicted with variables controllable by instructors in the class setting. The best predictive model found is a stepwise regression model, and it accounts for 57% of the variability of a student’s course grade based on a sample size of 182 students. The results showed that a student’s clarity in career direction is most significant in predicting the semester grade, followed by his or her interest in the course. In addition, the results suggest that predictive models hold great promise for improving classroom operations in a way that will enhance students’ learning experience and, ultimately, their performance.
Yi, John C. “Predictive Analytics for Students’ Quantitative Class Performance and Its Impact on Classroom Operations.” Journal of Supply Chain and Operations Management10.2 (2012: Sept.): 65-75.