Analysis of Student Academic Performance Using Cluster Analysis: A Study of Federal Government Colleges in North West Nigeria

Understanding patterns in student academic performance is essential for designing targeted educational interventions in Nigeria's secondary school system. This study applies k-means and hierarchical cluster analysis to classify students in four Federal Government Colleges across Sokoto, Kebbi, and Zamfara States in the North West geopolitical zone based on academic performance profiles. Data comprising final examination scores in eight core subjects across three consecutive academic sessions are obtained for 850 students with appropriate institutional approval. Variables are standardised before clustering to eliminate scale effects, and the optimal number of clusters is determined using the elbow method and silhouette coefficient. Four distinct performance clusters are identified: high achievers, average performers, struggling students, and subject-specific specialists. The high-achiever cluster constitutes 18 percent of the sample and is characterised by consistent scores above 70 percent across all subjects. The struggling cluster, comprising 29 percent of students, records average scores below 45 percent in mathematics and science subjects. Discriminant analysis is applied to identify the subject combinations most predictive of cluster membership. Results indicate that performance in mathematics and English language at Junior Secondary School level are the strongest predictors of senior secondary cluster assignment. The study recommends early diagnostic testing and targeted remediation for students exhibiting struggling-cluster profiles. Keywords: cluster analysis, academic performance, Federal Government Colleges, k-means, North West Nigeria.