Impact of CASPer Cut Score on Diversity in Physician Assistant Program Admissions

INTRODUCTION: Physician assistant (PA) programs increasingly use holistic admission strategies to evaluate noncognitive attributes not assessed by traditional academic metrics. Computer-Based Assessment for Sampling Personal Characteristics (CASPer), a situational judgment test, was developed to assess qualities such as professionalism, ethics, and decision making. Limited data exist regarding the impact of CASPer on applicant diversity in PA education. METHODS: This retrospective observational study examined applicants to a single public PA program across 5 application cycles (2021-2025). Computer-Based Assessment for Sampling Personal Characteristics scores were standardized within cohort, and a hypothetical cut score set at one standard deviation below the mean was applied. Applicants were categorized by scores greater than or less than the cut score. Gender, race/ethnicity, and underrepresented minority (URM) status were compared using chi-square tests, t tests, and logistic regression. RESULTS: Applying a CASPer cut score would have removed 16.5% of applicants. Modeling the interview pool demonstrated a 12.1% relative decrease in men and 12.5% relative decrease in URM applicants after replacement of those with less than the cut score. Computer-Based Assessment for Sampling Personal Characteristics scores did not differ significantly by race/ethnicity or gender. Logistic regression showed that male gender and lower prerequisite grade point averages were associated with higher odds of falling below the cut score. DISCUSSION: Although CASPer is intended to assess noncognitive attributes, applying a cut score was associated with reductions in male and URM applicants. These findings suggest that CASPer may influence applicant diversity through how scores are applied rather than through group differences in overall performance. Programs should consider using CASPer as one component of holistic review rather than as a rigid screening threshold.