Analysis of JAMB Examination Results: Using Cluster and Canonical Correlation Techniques.

Abstract

This study investigates JAMB examination results for 2006. The data consist of scores of 225 students. Out of twenty-three subject areas which the students sat for, eleven subjects were used for the study. The data was subjected to cluster and canonical correlation analysis .The essence was to divide the subjects into smaller number of classes such that objects in the same class are similar to one another. The complete linkage method of cluster analysis was employed to this effect, while canonical correlation analysis was employed to investigate the relationship between these groups.  The results show that two groups were formed. Group A consists of Government, religion, mathematics, commerce, and literature-in-English, while Group B consists of   English, biology, chemistry, agriculture, economics, and physics. Five canonical roots were obtained and only one is statistically significant. However, group A is inversely related to group B indicating results from mathematics, commerce and religion show the strongest relationship and weighted heaviest among the subjects. While Chemistry, Economics and Agriculture appeared to be low with weak relationship and weighted lighter among the subjects.

Keywords canonical correlation, cluster analysis