Predictive Value of Dental School Admissions Criteria for Preclinical Restorative and Prosthetic Dentistry Performance: A Multiyear Cohort Analysis

Background This study aimed to develop a predictive model of dental student success in preclinical manual dexterity development as a function of admissions criteria, including potential interaction effects among key predictors. Success measures comprised overall grade point average (GPA) of manual dexterity courses, the pass rate on restorative and prosthetic dentistry hands‐on preclinical practical competency examination, and the longitudinal preclinical performance trajectory of student grades over time. The ultimate goal was to better inform admissions decision‑making and academic advising. Methods Admissions and performance data from 367 predoctoral students representing the graduating classes of 2023–2025 were analyzed. A first‑order multiple regression model including more than 11 admission variables was developed, followed by refinement through selected second‑order interaction terms to evaluate nonlinear and combined effects. Results Dental Admission Test (DAT) Perceptual Ability Test (PAT) scores and enrollment in a quantitatively or biomedically intensive undergraduate major were significantly associated with manual dexterity based GPA ( p < 0.001). Several additional factors showed significant relationships with both competency pass rate and longitudinal performance. First‑order models predicted average GPA poorly ( R 2 _pred = 0.02), whereas inclusion of interaction effects modestly improved predictive performance ( R 2 _pred = 0.08). Longitudinal performance trajectory displayed stronger predictive capacity ( R 2 _pred = 0.28), suggesting that longitudinal trends in preclinical performance are more amenable to prediction than static outcomes. GPA, DAT/PAT, and certain quantitatively or biomedically intensive science or health science professional majors as well as universities appeared to have positive correlation with manual dexterity development in preclinical restorative and prosthetic courses. Conclusions Regression models incorporating second‑order interaction terms including undergraduate GPA, DAT/PAT, undergraduate majors and possibly universities enhanced prediction of dental student success, although overall explanatory power remained limited for manual dexterity‐based GPA and competency examination performance. Longitudinal performance trajectory outcomes demonstrated the greatest predictive reliability. Larger, multisite datasets and inclusion of psychological or behavioral variables together with undergraduate majors and education may further improve model generalizability and support evidence‑based admissions and advising practices.