Elam, C.L.; Johnson, M.M.S. Prediction of Medical Students' Academic Performances: Does the Admission Interview Help? Academic Medicine, 67;S28-S30, 1992.

PURPOSE: The purpose of this study was to determine if a positive relationship existed between medical students' admission interview rating and their performance in medical school.

METHODS: The study was conducted over a seven-year period at a southeastern public medical school. The admission committee conducted two one-on-one unstructured interview sessions for invited candidates. Committee members had access to applicant academic files prior to and during the interview. Interviewers completed a ratings form immediately following each interview which covered applicant acceptability. Additional data were gathered from medical school applications and medical school academic files.

RESULTS: The results indicated that admission interview ratings were modestly predictive of medical school performances as measured by the GPAs in basic science (years one and two) and the GPAs in clinical science (years three and four). The authors maintained that the results of regression analyses indicated that the predictive value of the interview ratings for medical school GPA generally increased over time in medical school, while the average predictive value of previous academic performance as measured by undergraduate science GPA and total MCAT score diminished. In addition, their results illustrated that the predictive power of traditional admission selection measures dramatically decreased with each succeeding year in medical school. The authors found that ratings from particular groups of interviewers were significantly associated with academic performance measures (i.e. ratings assigned by female interviewers were more highly correlated with medical students' performances than were ratings assigned by males; and only ratings assigned by physicians were consistently significantly correlated with medical school performance).

CONCLUSIONS: The authors stated that if the interview is expected to be a measure of noncognitive qualities and abilities, structured interview formats should be created to assess specific personal dimensions so as to increase interviewer consistency and to act as better outcome measures to chart the predictive power of the interview.

Contact Us    © 1995-2009 AAMC    Terms and Conditions    Privacy Statement