Abstract
Purpose: This study examines the application of a predictive regression analysis model to identify and address underperformance among tertiary education students. The objective is to shift higher education institutions from reactive to proactive support systems by integrating data-driven early warning mechanisms with personalised learning interventions.
Design/methodology/approach: A pragmatic mixed-methods approach was adopted, combining quantitative regression analysis with qualitative thematic review. The quantitative component used a Kaggle dataset (N = 682 after cleaning) to assess socio-demographic and academic predictors of student performance through multiple linear regression. Qualitative analysis examined challenges and institutional readiness for implementing predictive analytics in higher education.
Findings: The regression results indicate that parental education, family income, and learner stress levels are significant predictors of academic performance, whereas attendance, study hours, and extracurricular participation are statistically insignificant. The model was statistically significant overall, though with a modest R² (0.046). Findings suggest that socio- demographic context plays a critical role in shaping learner outcomes, and predictive models must be supported by institutional frameworks for intervention.
Research limitations/implications:
The study is based on secondary, cross-sectional data that lacks a national context and relies partly on self-reported variables. Results may not be generalizable across institutions or regions.
Practical implications:
Universities can embed regression-based early warning systems into student management systems to identify at-risk learners and provide timely, personalized interventions, thereby improving retention and graduation rates.
Originality/value:
This research bridges the gap between predictive analytics and actionable personalised learning strategies in tertiary education, offering a methodological framework for enhancing equity and student success.