Evaluating the internal structure validity of admissions measures: do they capture distinct attributes?

Veterinary admissions committees aim to select applicants who will successfully complete a Doctor of Veterinary Medicine (DVM) program and contribute effectively to the profession. Although traditional admission procedures emphasize academic indicators such as undergraduate grade point average (uGPA) and standardized test scores, increasing attention has been placed on assessing social-emotional skills and school-specific contextual attributes. Given the high-stakes nature of admissions and the claim that multiple attributes are assessed at the time of selection, internal structure validity evidence is required to support the use of scores from admissions measures in the selection process. This study examines whether commonly used admissions measures reflect distinct academic, social-emotional, and contextual factors.

Participants included 338 applicants (78.17% female) in a retrospective cohort study from the 2022-2024 admissions cycles at the Faculty of Veterinary Medicine, University of Calgary, Canada. Admissions measures included uGPA, Medical College Admissions Test (MCAT) total and subscores, Multiple Mini-Interview (MMI) scores, Computer-based Assessment for Sampling Personal Characteristics (Casper) situational judgement scores, standardized essay ratings, and structured file review scores. Bivariate correlation and Exploratory Factor Analysis (EFA) were used to explore the relationships and internal structure between the scores from these admissions measures.

Correlations among admissions variables showed small to moderate associations, with expected clustering among academic indicators. EFA produced a three-factor solution accounting for 53.36% of the variance which were labeled: (1) Academic Ability (uGPA; MCAT academic subscores), (2) social- emotional skills (essay, Casper, MCAT social emotional sub score; partial MMI loading), and (3) contextual factors (file review; partial MMI loading).

Findings provide internal structure validity evidence supporting the claim that admissions measures capture multiple attributes and these measures should be considered collectively as a system of assessment. Notably, the MMI contributed to both social-emotional and contextual factors, highlighting its multifaceted nature. These findings provide a foundation for longitudinal research examining the predictive validity of academic, social-emotional, and contextual factors for performance throughout veterinary training.