Regression Analysis of Factors Affecting Examination Performance in Mathematics Among JSS Students in Kaduna State

This study applies multiple regression analysis to identify and quantify the factors associated with Junior Secondary School students' examination performance in mathematics in Kaduna State, North West Nigeria, providing an empirical basis for targeted remediation and instructional improvement efforts. Mathematics performance at the junior secondary level has long-term consequences for students' academic trajectories, particularly for access to science and technical programmes at the senior secondary and tertiary levels. A stratified random sample of 320 JSS3 students from sixteen public and private schools in Kaduna North, Kaduna South, and Chikun local government areas is drawn, and data are collected on examination scores, student-level variables including prior achievement, attendance rate, and home study hours, teacher-level variables including qualification and years of experience, and school-level variables including class size, library access, and instructional material availability. Ordinary least squares regression models with robust standard errors are estimated, and model selection is conducted using stepwise and information criteria approaches. Results identify prior achievement, home study time, and teacher qualification as the three strongest predictors of JSS examination performance. School-level variables show statistically significant effects only when student-level variables are excluded from the model. The study recommends merit-based teacher deployment and structured homework support programmes. Keywords: regression analysis, examination performance, junior secondary school, mathematics education, Kaduna