Predictive Modeling in Medical Education: Identifying Factors Influencing Academic Success

Abstract

Background: Academic success in medical education is influenced by multiple factors, including demographics, prior academic performance, and psychosocial attributes. Predictive modeling can enable early identification of at-risk students, fostering tailored interventions.
Objective: This study aimed to develop and validate a predictive model for identifying key factors influencing academic success among medical students at a College of Medicine in Al Ahsa, Saudi Arabia.
Methods: A cross-sectional design was employed, analyzing data from 350 students. Variables included demographic characteristics, high school grades, admission exam scores, and psychosocial factors such as motivation, time management, and academic self-efficacy. Predictive analyses were conducted using multiple regression and machine learning techniques.
Results: High school grades and admission scores emerged as the strongest predictors of GPA, while motivation also showed a significant positive association. The random forest model achieved the highest predictive accuracy (77%), outperforming logistic regression (73%) and decision tree methods (69%).
Conclusion: The findings highlight the need for a multifaceted approach, combining academic and psychosocial metrics, to predict and enhance academic performance. Implementing predictive models can support targeted interventions, improving educational outcomes in medical programs.