Within the higher education context, various studies have been undertaken on the effects of personality that can potentially affect students' academic performance. In addition, earlier studies have shown that in the context of higher education, a better understanding of the relationships between these differentiating factors, such as personality, learning approaches, motivation, course perceptions, academic backgrounds such as prior academic performance, prior knowledge and prior education institution will enable higher education institutions to better understand the determinants of student’s academic success . However, there seem to be few studies that have explored the relationships between these various determinants holistically. Also, not many studies examined the effect of personality using a full model (i.e., inclusion of personality) and reduced model (i.e., exclusion of personality) approach.

This study thus applied data mining techniques, in particular, decision trees, to examine the effect of personality on academic performance through the inclusion and exclusion of personality traits in the model (i.e., full and reduced models). The study also explored the relationships of the above-mentioned determinants holistically in relation to academic performance. C5.0 was selected as the champion model and the resulting paired T-test of the predicted outcome of the full model and reduced model revealed that personality does have a significant effect on academic performance. Personality traits such as Conscientiousness, Extraversion, and Agreeableness are found to be important predictors of academic performance. In contrast to some studies, Extraversion has been found to have a positive association with academic performance whilst Agreeableness has been found to have a negative association instead. It is hoped that this study could provide a deeper layer of understanding of the effects of personality traits and their complex interactions with other factors on academic performance, which in turn will aid higher education institutions in tailoring students’ learning needs leading to better academic performance.

Keywords: Personality, higher education, academic performance, full vs reduced model, data mining.