The article brings forth some interesting information. However, the research did not include assessing traditional interviews (panel or one-on-one) to find out if that method also correlates strongly to extraversion. This is quite plausible as applicants in the study already had higher mean scores of extraversion and agreeableness (see page 8 of study under Limitations).
In an earlier paper by Griffin & Wilson, entitled “Associations between the big five personality factors and multiple mini-interviews,” Australian researchers discovered that “MMIs tapped the constructs of extraversion and conscientiousness…” indicating that “…those who perform well on an MMI are likely to enjoy being with people, are full of energy, enthusiastic, action-oriented individuals and are likely to strive hard to achieve excellence, be reliable and hard-working, and persist even when a task is difficult or unpleasant.” They further concluded that MMIs and personality questionnaires are not interchangeable so they would not recommend personality tests be used as a screening tool for MMIs.

Re: Paul’s comments – Jerant et al in their paper compared their findings with those of Griffin & Wilson. From the Jerant et al paper Discussion section: “An Australian study had findings closer to ours, with higher Extraversion, Agreeableness, and Conscientiousness associated with higher MMI scores.(ref 31) MMI participants in the Australian study were younger than those in our study, just completing high school, and such individuals typically have lower levels of Agreeableness and Conscientiousness than older persons.(ref 39) Also, screening for MMI participation in the Australian study hinged solely on cognitive testing results. By contrast, screening to determine MMI participation at our school included faculty ratings of personal statements, recommendation letters, and extra-curricular experiences. Such ratings likely pre-select for higher Agreeableness and Conscientiousness,(e.g. team orientation, humanistic qualities).(ref 11)”

Thus the findings of the two studies are generally consistent with each other, and the difference between the studies in the finding regarding Conscientiousness is likely related to the substantial differences in the Australian vs U.S. med school applicant pools and pre-MMI screening processes.

Also it should be emphasized that Jerant et al study did not explicitly propose or imply that the personality screen should replace the MMI or be used as an admissions screening tool – that is a very different issue than the one the study addressed.

Jerant et al also explicitly called for RCTs comparing the MMI and traditional interviews to determine if either is more or less likely to select for particular personality attributes. While this remains an interesting empirical question, the lack of a traditional interview arm in the study does not actually undermine the conclusions of the paper regarding the tendency of the MMI to select for certain personalities – the focus of the paper was simply on what personality attribute(s) the MMI selects for – an important question since the process is being adopted by an increasing number of schools.

The work of Eva et al, Academic Medicine (79) S40, 2004 addresses the predictive validity of the MMI in medical school using the OSCCEs as the outcome measure. The scores on the MMI correlate with that measure of acquisition of clinical skills and performance in a clinical situation. Reiter et al (Medical Education 2007; 41: 378–384) addresses the predictive validity of the MMI for performance on the Medical Council of Canada Qualifying Examination (MCCQE). This is the best predictor of performance in practice. In particular Eva et al focused on the “subsections relevant to medical specialties and relevant to broaderlegalandethical issues(PopulationHealthand the Considerations of the Legal, Ethical and Organisational Aspects of Medicine [CLEO⁄PHELO])”.The MMI scores correlated with the CLEO/PHELO scores which would be expected of a selection technique based on the non-cognitive skills. The second study also found that MMI predicted performance in clerkship where the interpersonal skills are paramount. Regardless of the relationships to the P5 the selection of the MMI produces graduates with the neccessary level of professiolism,

Finally, although it is N of 1 as a rater (and admittedly a biased one), I do tutor approaximately 10% of the class every year. To my eyes and ears, whatever the level of extraversion there may be, there is wide spread respect for other students, for their opportunity to enter discussions and for their ideas.

We need to know what students and graduates actually do not their scores on the P5.

To paraphrase Rosenfeld: “We need to know what students and graduates actually do not their scores on some qualifying exam.”

At least the link between personality and behavioral tendencies (in multiple aspects of life and functioning) is well-established, in literally 1000s of studies worldwide.

By contrast, the link between medical qualifying exam performance (Canadian or otherwise) and actual performance in clinical practice is so scant as to be almost non-existent.

We need to be honest with ourselves as a profession – we have no idea whether any applicant screening process, including the MMI, predicts future clinical performance as a doctor.

Also the Eva et al studies of the MMI lacked a comparison non-MMI selected group – so it may well be that people who do better on traditional interviews (or some other process) also have higher qualifying exam scores – though again that wouldn’t mean they are necessarily destined to be “better doctors.”

What we need are multi-school RCTs comparing various screening processes including the MMI – and following those accepted over time into their eventual clinical practices – to know whether the MMI adds more value than harm relative to alternative processes.

Until that happens, anecdotes, personal biases, and “the hot new thing” will continue to carry the day in medical education and admissions, just as they have for centuries.

As a provider of mental health care the above seems ridiculous. It is trying to fit round pegs into square holes. Doctors provide a service to ALL KINDS of PEOPLE and no one personality type fits all. Whereas one patient warms up to the outgoing, friendly doctor, another could care less and want his expertise and knowledge. One person wants a good listener, another wants a take-charge type to navigate them through a frightening journey. One patient desire a collaborative approach while another wants someone more directive. There should be as many personality types in medicine as their are people as long as they are qualified academically. Patients are always free to choose. Stop trying to create a type!