Predictive validity of admission tests and educational attainment on preclinical academic performance - a multisite...

Area of Science:

• Medical Education
• Psychometrics
• Higher Education Admissions

Background:

• Educational attainment and admission tests are historically used for medical student selection.
• Their predictive validity for academic performance is frequently assessed.
• Simultaneous use may obscure overlapping predictive validity.

Purpose of the Study:

• Assess predictive and incremental validities of educational attainment (high-school GPA) and admission tests (TMS, HAM-Nat).
• Conduct subtest analyses to understand admission tests' predictive power more deeply.

Main Methods:

• Survey of German medical school admission test-takers (2022-2023).
• Matched self-reported preclinical performance with admission test scores and high-school GPA (n=2113).
• Hierarchical multiple regression and Pearson's correlations, with multivariate correction for range restriction.

Main Results:

• Both admission tests (TMS, HAM-Nat) and high-school GPA individually predict academic performance.
• Admission tests show substantial incremental validity over high-school GPA.
• Subtests combining crystallized and fluid intelligence have moderate predictive effect sizes.

Conclusions:

• Admission tests and high-school GPA are suitable medical school selection criteria.
• Admission tests offer stronger predictive power than high-school GPA, addressing concerns about GPA.
• Content-rich subtests assessing both crystallized and fluid intelligence best predict preclinical academic performance.