Modeling University Student Academic Performance in Nigeria: A Comparison of Seemingly Unrelated Regression Equations (Sure) and Multivariate Regression

This study investigates the factors influencing academic performance of students at Modibbo Adama University of Technology, Yola, Adamawa State, using the Seemingly Unrelated Regression Equations (SURE) model. Secondary data were collected from the Department of Statistics and Operations Research, encompassing variables including age at entry, gender, mode of entry, school type, parent occupation, course scores from 100 to 500 level, and CGPA for each session across three student cohorts (2016, 2017, and 2018). The SURE model was employed to analyze three dependent variables: first year CGPA, final year CGPA, and total credit units passed, while accounting for contemporaneous correlation among the error terms. Model selection was conducted using log-likelihood, Akaike Information Criterion (AIC), and Bayesian Information Criterion (BIC). The SURE model demonstrated superior performance over multivariate regression, with lower root mean square error values (0.6291, 0.5596, and 3.1884 for first year CGPA, final year CGPA, and total credit passed, respectively). The model explained 55.6%, 60.1%, and 96.0% of the variance in first year CGPA, final year CGPA, and total credit units passed, respectively. JAMB score emerged as the most significant predictor across all three dependent variables (p < 0.001), while program type significantly affected all performance measures at the 5% level. Age significantly influenced total credit units passed, and student set significantly affected first year CGPA. Conversely, gender, mode of entry, and number of O-level sittings showed no significant effects on any performance measure. The SURE model effectively captures the interdependencies among different academic performance indicators, with JAMB score serving as the strongest predictor of student success. The findings support the continued use of University Tertiary Matriculation Examination (UTME) scores as a reliable admission criterion for academic programs.