The Multiple Mini-Interview (MMI) [1] is an interview format that uses many short independent assessments, typically in a timed circuit, to obtain an aggregate score of each candidate’s soft skills. In 2001, the Michael G. DeGroote School of Medicine at McMaster University began developing the MMI system, to address two widely recognized problems. First, it has been shown that traditional interview formats or simulations of educational situations do not accurately predict performance in medical school. Secondly, when a licensing or regulatory body reviews the performance of a physician subsequent to patient complaints, the most frequent issues of concern are those of the non-cognitive skills, such as interpersonal skills, professionalism and ethical/moral judgment.

Introduction

Interviews have been used widely for different purposes, including assessment and recruitment. Candidate assessment is normally deemed successful when the scores generated by the measuring tool predict for future outcomes of interest, such as job performance or job retention. Meta-analysis of the human resource literature has demonstrated low to moderate ability of interviews to predict for future job performance [2]. How well a candidate scores on one interview is only somewhat correlated with how well that candidate scores on the next interview. Marked shifts in scores are buffered when collecting many scores on the same candidate, with a greater buffering effect provided by multiple interviews than by multiple interviewers acting as a panel for one interview [3]. The score assigned by an interviewer in the first few minutes of an interview is rarely changed significantly over the course of the rest of the interview, an effect known as the halo effect. Therefore, even very short interviews within an MMI format provide similar ability to differentiate reproducibly between candidates [4]. Ability to reproducibly differentiate between candidates, also know as overall test reliability, is markedly higher for the MMI than for other interview formats [1]. This has translated into higher predictive validity, correlating for future performance much more highly than standard interviews [5-8].


History

Aiming to enhance predictive correlations with future performance in medical school, post-graduate medical training, and future performance in practice, McMaster University began research and development of the MMI in 2001. The initial pilot was conducted on 18 graduate students volunteering as “medical school candidates”. High overall test reliability (0.81) led to a larger study conducted in 2002 on real medical school candidates, many of whom volunteered after their standard interview to stay for the MMI. Overall test reliability remained high [1], and subsequent follow-up through medical school and on to national licensure examination (Medical Council of Canada Qualifying Examination Parts I and II) revealed the MMI to be the best predictor for subsequent clinical performance [5, 7], professionalism [6], and ability to communicate with patients and successfully obtain national licensure [7,8]. Since its formal inception at the Michael G. DeGroote School of Medicine at McMaster University in 2004, the MMI subsequently spread as an admissions test across medical schools, and to other disciplines. By 2008, the MMI was being used as an admissions test by the majority of medical schools in Canada, Australia and Israel, as well as other medical schools in the United States and Brunei. This success lead to the development of a McMaster spin-off company, APT Inc., to commercialize the MMI system. The MMI was branded as ProFitHR and made available to both the academic and corporate sector [9]. By 2009, the list of other disciplines using the MMI included schools for dentistry, pharmacy, midwifery, physiotherapy and occupational therapy, veterinary medicine, ultrasound technology, nuclear medicine technology, X-ray technology, medical laboratory technology, chiropody, dental hygiene, and postgraduate training programs in dentistry and medicine.

Utility of the MMI

The MMI requires less expenditure of resources than standard interview formats [9]. Test security breaches tend not to unduly influence results [11]. Sex of candidate and candidate status as under-represented minority tends not to unduly influence results [1, 12]. Preparatory courses taken by the candidate tend not to unduly influence results [13]. The MMI has been validated and tested for over seven years and the product is now available “off the self.” [8]

[1] Eva KW, Reiter HI, Rosenfeld J, Norman GR. An admissions OSCE: the multiple mini-interview. Medical Education, 38:314-326 (2004).
[2] Barrick MR, Mount MK. The Big 5 personality dimensions and job performance: a meta-analysis. Personnel Psychology 1991, 44:1-26.
[3] Eva KW, Reiter HI, Rosenfeld J, Norman GR. The relationship between interviewer characteristics and ratings assigned during a Multiple Mini-Interview. Academic Medicine, 2004 Jun; 79(6):602.9.
[4] Dodson M, Crotty B, Prideaux D, Carne R, Ward A, de Leeuw E. The multiple mini-interview: how long is long enough? Med Educ. 2009 Feb;43(2):168-74.
[5] Eva KW, Reiter HI, Rosenfeld J, Norman GR. The ability of the Multiple Mini-Interview to predict pre-clerkship performance in medical school. Academic Medicine, 2004, Oct; 79(10 Suppl): S40-2.
[6] Reiter HI, Eva KW, Rosenfeld J, Norman GR. Multiple Mini-Interview Predicts for Clinical Clerkship Performance, National Licensure Examination Performance. Med Educ. 2007 Apr;41(4):378-84.
[7] Eva KW, Reiter HI, Trinh K, Wasi P, Rosenfeld J, Norman GR. Predictive validity of the multiple mini-interview for selecting medical trainees. Accepted for publication January 2009 in Medical Education
[8] Hofmeister M, Lockyer J, Crutcher R. The multiple mini-interview for selection of international medical graduates into family medicine residency education. Med Educ. 2009 Jun;43(6):573-9.
[9] www.ProFitHR.com
[10] Rosenfeld J, Eva KW, Reiter HI, Trinh K. A Cost-Efficiency Comparison between the Multiple Mini-Interview and Panel-based Admissions Interviews. Advanced Health Science Education Theory Pract. 2008 Mar;13(1):43-58
[11] Reiter HI, Salvatori P, Rosenfeld J, Trish K, Eva KW. The Impact of Measured Violations of Test Security on Multiple-Mini Interview (MMI). Medical Education, 2006; 40:36-42.
[12] Moreau K, Reiter HI, Eva KW. Comparison of Aboriginal and Non-Aboriginal Applicants for Admissions on the Multiple Mini-Interview using Aboriginal and Non-Aboriginal Interviewers. Teaching and Learning in Medicine, 2006; 18:58-61.
[13] Griffin B, Harding DW, Wilson IG, Yeomans ND. Does practice make perfect? The effect of coaching and retesting on selection tests used for admission to an Australian medical school. Med J Aust. 2008 Sep 1;189(5):270-3

How the MMI works:

The MMI system consists of several key components working in unison:

1. Interview stations – the domain(s) being assessed at any one station are variable, and normally reflects the objectives of the selecting institution. Examples of domains include the “soft skills” - ethics, professionalism, interpersonal relationships, ability to manage, communicate, collaborate, as well as perform a task. An MMI interview station takes considerable time and effort to produce; it is composed of several parts, including the stem question, probing questions for the interviewer, and a scoring sheet.
2. Circuit(s) of stations – to reduce costs of the MMI significantly below that of most interviews [10], the interview “stations” are kept short (eight minutes or less) and are conducted simultaneously in a circuit as a bell-ringer examination. The preferred number of stations depends to some extent on the characteristics of the candidate group being interviewed, though nine interviews per candidate represents a reasonable minimum [3]. The circuit of interview stations should be within sufficiently close quarters to allow candidates to move from interview room to interview room. Multiple parallel circuits can be run, each circuit with the same set of interview stations, depending upon physical plant limitations.
3. Interviewers – one interviewer per interview station is sufficient [3]. In a typical MMI, each interviewer stays in the same interview throughout, as candidates rotate through. The interviewer thus scores each candidate based upon the same interview scenario throughout the course of the test.
4. Candidates – each candidate rotates through the circuit of interviews. For example, if each interview station is eight minutes, and there are nine interview stations, it will take the nine candidates being assessed on that circuit 72 minutes to complete the MMI. Each of the candidates begins at a different interview station, rotating to the next interview station at the ringing of the bell.
5. Administrators – each circuit requires at least one administrator to ensure that the MMI is conducted fairly and on time.