Validity Evidence of a Multiple Mini Interview for Selection of Medical Students: Universiti Sains Malaysia Experience

Abstract

Introduction: Student selection is important for selecting the best candidates into medical courses. MMI is a valid selection tool for the task. The School of Medical Sciences (SMS), USM have employed MMI as a medical admission tool to select potential candidates into its medical program. Objective: This paper described the implementation of the MMI and reported the preliminary evaluation data on its validity evidence. Method: a 9-station MMI (with 5 manned and 4 rest stations) was employed for the 2015 student selection exercise. Interview data were analysed and questionnaire surveys were administered to both interviewers and candidates. Validity, reliability, feasibility and acceptability were determined. Results: Unidimensional construct on CFA and interviewer comments provided evidence of construct validity. Overall reliability was 0.94, good enough for high-stakes decisions. Interviewers and candidates’ comments on feasibility and acceptability were also generally positive. The consistency of the difficulty and discrimination indices of similar stations between sessions was identified as among the areas for improvement. Conclusions: We are cautiously optimistic regarding the utility of the MMI in the SMS; future improvements are planned but the present implementation seems sufficient for the stated purposes.