Modeling the determinants of internship-to-job retention among second-year students at Makerere University: a binary logistic regression approach

Abstract

The transition from academic internships to permanent employment represents a critical milestone in the school-to-work transition for statistics and economics students. This study aimed to model the determinants of internship-to-job retention among second-year students at the School of Statistics and Planning (SSP), Makerere University. Utilizing a cross-sectional research design, primary data was collected from a sample of 240 respondents using structured questionnaires. The study employed univariate, bivariate (Pearson’s Chi-square), and multivariate (Binary Logistic Regression) analysis techniques to identify significant predictors of retention. The results revealed a retention rate of 30.4% among the sampled students. Bivariate analysis indicated that while academic performance (CGPA) and supervision quality showed associations with retention, they were not statistically significant at the 5% level. However, the Binary Logistic Regression model, which demonstrated high predictive accuracy (AUC = 0.8126), identified several key significant determinants. Age was found to have a positive influence on retention (OR = 1.135, p = 0.032), suggesting that maturity is highly valued by employers. In contrast, the organizational sector played a critical role; interning within the NGO sector significantly reduced the odds of retention (OR = 0.244, p = 0.036) compared to the public sector. Furthermore, the source of the internship was paramount, as students who secured placements via direct applications were significantly less likely to be retained (OR = 0.140, p = 0.039) than those placed through university-mediated channels. These findings suggest that internship-to-job retention is driven more by institutional signaling and organizational characteristics than by academic metrics alone. The study concludes that strengthening university-industry partnerships is essential for enhancing graduate employability. It is recommended that the School of Statistics and Planning formalizes more placement agreements with high-retention sectors and that students prioritize internships in organizations with established hire-back policies to optimize their career entry trajectories. Keywords: Internship Retention, Binary Logistic Regression, Makerere University, School-to-Work Transition, Statistics Students.