Validity Evidence of Multiple Mini Interviews for Holistic Admissions in Physical Therapy Education

Background. In an effort to improve diversity of the physical therapy (PT) workforce, the profession has put forth a series of comprehensive strategies to meet this goal. One strategy advocates for PT education to explore holistic admission processes. Purpose. Through several study purposes, this quantitative analysis of the Multiple Mini Interview (MMI) as a tool for holistic selection contributes to the field of PT education in the following ways: 1) describes the development of the MMI, 2) provides validity evidence for the tool use in physical therapy education selection processes. Methods. Through a rigorous quantitative analysis of validity using Messick’s Unified Validity Framework, this research discussed development of the tool (content validity), acceptability and feasibility of the tool by raters and candidates (response process), reliability of the tool (internal structure), its predictive ability (relationship to other variables), and the differential functioning and incremental impact of the MMI on summative outcome measures (consequences). Results. Noteworthy findings revealed no significant differences in inter- and intra-group variation and no significant correlation to candidate characteristics (gender, age, race/ethnicity, and ESL/ELL) nor to selection GPAs and MMI total score. This is important highlighting the non-bias nature of the tool and its ability to inform constructs other than academics in the selection. Predictive relationships to summative measures and MMI demonstrated significance to clinical education (CE) scores. This is meaningful as in PT education selection literature minimal relationship between selection variables has been established to CE performance. Lastly, this study analyzed the differential functioning of the MMI on outcome measures within each of the candidate characteristic subsets and its incremental impact in comparison with tools traditionally accepted to evaluate non-academic characteristics similar to the MMI (e.g. personal statements, letters of recommendation). Results demonstrate the MMI does indeed function differently across subsets influencing CE scores for gender, race/ethnicity, and age and licensure pass rates for ESL/ELL. Further, the MMI showed a positive incremental impact of incorporating into selection processes on the outcome measures of clinical education scoring. Conclusions. Taken together, these findings provide strong evidence for the validity and utility of the MMI in PT education selection processes.