Predictive Validity and Differential Prediction of the MCAT Section Scores for Medical Students' Performance in Preclinical Courses.

Purpose: Limited research has been conducted on the differential predictive validity of the Medical College Admission Test (MCAT) section scores (e.g., Biological and Biochemical Foundations of Living Systems [BBLS], Chemical and Physical Foundations of Biological Systems [CPBS]) for basic science courses (Anatomy and Histology). Accordingly, the purpose was to assess the predictive validity and differential prediction of these section scores to predict Anatomy and Histology performance across sex (men vs. women) and race (White vs. non-White). Methods: The authors analyzed data from 520 undergraduate medical students (sex: 292 women [56.15%], 228 men [43.85%]) in Anatomy and Histology courses. The authors utilized multiple linear regression and t-tests to test for statistically significant differences in slopes associated with each section score across sex and race groups. Results: BBLS and CPBS section scores explained more variance in Histology compared to Anatomy, particularly for the non-White and men groups (25% and 29% vs. 24%). For the differential prediction, t-tests were not statistically significant for most analyses, which provided evidence for the comparable predictive validity across the groups or lack of differential prediction, a desired psychometric property for fair assessments. The t-test associated with the BBLS section score, however, was statistically significant between the women and men groups only in Anatomy. Conclusions: Current results contribute to the broader evidence supporting the validity of MCAT section scores. The general absence of differential prediction (i.e., comparable predictions across groups) supports the fairness of MCAT-based inferences, bolstering confidence in its use for medical school admissions and promoting equity.