2012 Grant Recipient's Abstract: Stuart Lubarsky, MD, MPHE

October 27, 2011

Stuart Lubarsky, MD, MPHE

"Assessing the Multiple Mini Interview (MMI) for Use in Selecting Applicants to a Neurology Residency Program"

The goal of this study is to develop and implement a novel tool — the Multiple Mini Interview (MMI) — for selecting applicants to a residency training program in neurology. Unlike traditional one–to–one or panel–based interviews, MMI applies to the selection process a fundamental principle of assessment: that multiple observations of an individual's performance in different contexts (ideally by multiple observers) are essential for diminishing bias and reliably gauging his or her abilities. Based on this 'multiple samples' approach, MMIs feature numerous interview 'stations' designed to assess applicants' ability to demonstrate specific attributes and skills (eg. collegiality, moral reasoning, etc.) across varying situations. Evidence is growing in support of MMI as an instrument for selecting candidates to undergraduate medical training programs. To date, however, few studies have investigated the use of MMI at the postgraduate level, and none, to our knowledge, have explored its use for selecting candidates applying to a residency program in neurology. We propose to create MMI stations using a blueprint of attributes/skills judged to be particularly desirable for incoming neurology trainees (eg. decision-making under conditions of uncertainty), and administer the MMI to the 2012 cohort of applicants to our neurology program. We aim to investigate whether MMI is a reliable and valid tool for selecting suitable future neurology residents. We will also examine the acceptability of MMI to applicants and interviewers, and the feasibility of adopting MMI for residency selection, compared to a traditional interview approach. We anticipate that our findings will be of interest to neurologist-educators seeking more defensible, evidence-based methods of selection than those in present use.

 email to a friend  |   print