Predictive Analytics Approach to Improve and Sustain College Students’ Non-Cognitive Skills and Their Educational Outcome

Abstract

The application of predictive analytics in higher education has increasingly gained acceptance and interest over the years. In this study, a predictive model is developed to map students’ non-cognitive skills against their class performance. Our predictive analytics model identified the non-cognitive skills that predicted new students’ class performance based on the dataset collected early in the semester. Based on the predictive analytics results, tailored teaching to improve students’ non-cognitive skills was offered in a required class designed for undergraduate business students. The improvement in the average final semester grade for students in the tailored-taught classes based on our predicted analytics approach was 9%, which was higher than that of the class grade taught without the approach. The study finding also demonstrates a long-term, sustainable positive effect to the students with the predictive analytics approach.