Estimating Academic Success in Higher Education Using Big Five Personality Traits, a Machine Learning Approach

The most popular way of predicting academic success in higher education is to use students' existing course grades. In this study we propose a novel approach to predict academic success in higher education with use of personality traits rather than existing course grades. Our main focus on this multidisciplinary study is to get the benefits of psychology and computer science to predict academic success of students in higher education, by using Machine Learning. We have used Big Five features as the personality traits of 2,575 higher education students and tested our proposed method on 20 different course categories. At the end of this study we conclude that Machine Learning can be used for predicting academic success while using all of the Big Five personality trait dimensions. With our proposed method, Big Five traits of prospective students can be used to predict higher education student success for the applied department. Our method can also be applied to different departments or course groups. This approach can be improved such that, the higher education institutes can even suggest departments to students, that they can be more successful.