Medical schools in the United States serve to train the next generation of physicians, admitting students who will continue to advance each school’s mission. Admission committees are tasked with identifying those candidates who will be successful academically and who promote the objectives of the school with respect to mission. The Quillen College of Medicine at East Tennessee State University in northeast Tennessee seeks to attract and retain physicians with an interest in rural and primary care medicine. A total of 630 students were included in this study representing classes from 2001 to 2011. This study examined admissions data including MCAT scores, undergraduate GPAs, admission interview scores, and admission committee rating scores along with USMLE Step 1 scores to determine if there is any correlation of these variables with graduates selecting a primary care career or a rural practice location. With respect to data available at admission, only MCAT scores were shown to have a significant correlation to specialty choice. None of the admission data significantly correlated with practice location. USMLE Step 1 scores had a weak negative relationship with specialty choice and a negligible relationship with practice location. This study provides the admission committee information that these variables are insufficient by themselves to predict whether a medical student applicant will select a primary care specialty or practice in a rural location. Other data, perhaps even subjective data, would need to be analyzed to predict how well the admissions committee is addressing the college’s mission with its selection of medical students.

The ranks of the medical profession have typically been filled by Anglo-Saxon males. A slow change in attitude appears to be leading toward a desire to increase diversity in the medical profession. Previous attempts to allow for increased representation of those underrepresented in medicine failed due to legal challenges. The Association of American Medical Colleges, an organization which oversees medical education, residency training, and research of both American and Canadian allopathic medical schools, has introduced Holistic Review as a method for creating a diverse population in a medical school class; this study investigates the way medical schools are incorporating these strategies into their medical school admissions policies and how effective these policies are in increasing diversity in medical school classes. The implication encourage admissions committees to modify the importance of certain selection criteria in an ongoing effort to increase diversity in their medical school classes.

Differential predictive validity in assessing academic performance in institutions of higher education has been assessed for a number of years. Historically, this body of research focused on gender and ethnicity. This study extends that research to geographic region (e.g., rural and non-rural populations). Specifically, this study predicted relationships between preadmission variables of incoming grade point average (GPA) and medical college admissions test (MCAT) and output variables of medical school GPA and comprehensive osteopathic medical licensing exam (COMLEX). Results indicate incoming GPA and MCAT are good variables to use to predict academic performance in medical school and score on the licensing board exam. Further, rural populations presented similar scores on preadmission variables and, thus, are not at a disadvantage in the admission process. A second goal of this study was to explore differential prediction of medical school GPA and COMLEX Level 1 score for the MCAT for rural and non-rural populations. Results provide some evidence of differential prediction of COMLEX score for the physical and biological sciences MCAT sub-tests such that rural populations' performance on the COMLEX Level 1 exam was underpredicted. Hence, when rural and non-rural populations present the same physical sciences and biological sciences MCAT sub-test score, the rural sub-group is predicted to obtain a lower COMLEX score and non-rural sub-group is predicted to obtain a higher COMLEX score. Further, when the two sub-groups present different MCAT scores for the physical and biological sciences sub-test, they are likely to obtain similar scores on the COMLEX. Implications and recommendations for future research are discussed. / Master of Science