Effect of Applicant Gender on Multiple Mini-Interview Admissions Score.

Abstract

Admissions criteria for physician assistant (PA) schools vary considerably, but they often involve a combination of academic measures, noncognitive "humanistic" qualities, and mission-related items. To address noncognitive attributes, some PA schools have implemented the multiple mini-interview (MMI) format developed at McMaster University in Canada. This research project looked at differences in interview scores by gender over 3 consecutive admissions cycles at one PA program. Three years of pre-existing de-identified data gathered as part of the routine admissions process were analyzed retrospectively using SPSS-v25. Data were available for all interviewees for 2015, 2016, and 2017 (N = 350 total). Between-group differences were not statistically significant by gender. Reliability (Cronbach's α) was 0.865 for academic scoring and 0.694 for MMI scoring. Analysis of 3 years of admissions data from a single program did not show evidence for gender bias in MMI scores. Although this result is reassuring, it requires continued monitoring and replication.