The Experience of Using CASPer at One College of Veterinary Medicine in the United States.

Abstract

CASPer (Computer-based Assessment for Sampling Personal characteristics), an online, open-response situational judgment test, screens applicants to educational programs early in the selection process for non-cognitive abilities. CASPer was applied in Purdue University's DVM admissions cycle in 2020-2021 in two ways: applicants met a threshold criterion of eligibility for further review by having a reported CASPer Z-score >-1.5; and applicants were awarded points for Z-score >0 during in-depth review. The requirement that applicants meet a threshold CASPer Z-score affected male, underrepresented minority (URM) and international students adversely: higher proportions of these individuals were eliminated early from admissions consideration. Among those with Z-score >-1.5, first-generation, male, URM, and international applicants scored lower, on average, than non-first-generation, female, and non-URM applicants. Additionally, Z-scores were correlated with undergraduate grade point average (uGPA). The correlation coefficient r of Z-scores with cumulative and core uGPA was 0.20 (p < .001) and 0.12 (p < .001), respectively, in all applicants. Hoping to select high-potential applicants without overemphasizing cognitive abilities, we expected CASPer Z-score to be independent from uGPA which is already weighted appropriately in our process. The nonrandom distribution of Z-scores systematically influenced early rejection and possibly biased later selection of applicants. Concerned that CASPer incorporation may have unintentionally narrowed access to DVM admissions for some applicant groups, and not seeing added value in a standardized test whose predictive validity remains understudied in veterinary medicine-a profession aiming to diversify its workforce-our admissions committee suspended CASPer as an admissions criterion until systematic, multi-source information becomes available within the profession.