Prediction of Students’ Performance in SPM English and Mathematics Using Data Mining Methods

Sijil Pelajaran Malaysia (SPM) or Malaysia Certificate of Education is a national examination required to be seated by all Form Five secondary school students in Malaysia. This examination is compulsory for them to pursue post-secondary education. However, students’ performance on this examination particularly in English and Mathematics subjects has been concerning. According to the literature, advances in educational technology can improve students’ performance. Therefore, this study attempts to predict Form Four students’ SPM performance in English and Mathematics subjects using data mining (DM) techniques. Three main attributes namely students’ past academic performance, demographics, and psychological attributes were scrutinised to identify their impact on the prediction. Moreover, the predictive performance of the DM techniques was evaluated to find the best technique for predicting students’ SPM performance. The relationship between students’ SPM performance in English and Mathematics subjects was also examined. This study found that by using Decision Tree (DT) rules, the characteristics of students with low, moderate, and high performance in English and Mathematics subjects could be identified. Next, DT and Nave Bayes (NB) had the best predictive performance among the DM approaches in predicting students’ SPM performance in English and Mathematics respectively. Moreover, there is a connection between students’ SPM English and Mathematics performance. The findings may provide stakeholders with new insight into how to improve students’ performance in these subjects. Educators can intervene early with students who are at risk of receiving low performance for these subjects in SPM