A validity generalization perspective on the ability of undergraduate GPA and the medical college admission test to predict important outcomes

Abstract

Background: Research on the validity of using the Medical College Admissions Test (MCAT) and undergraduate grade point average (GPA) for selection to medical school has produced conflicting interpretations. There is debate regarding the degree to which coefficients diminish over the course of educational and professional outcomes and disagreement over whether these two measures can predict clinical performance. Purpose: To summarize and interpret the validity literature using validity generalization techniques that account for measurement error. Methods: Validity generalization techniques were used to summarize MCAT and undergraduate GPA validity research. A meta-analysis was performed to evaluate validity coefficients for two outcome domains across educational and professional attainment levels. Results: The ability to predict academic performance decreases slightly for written tests. For clinical performance assessments, existing research does not allow an assessment of change across training levels. However, relevant studies suggest that MCAT and undergraduate GPA have a positive predictive relationship with clinical skills. Conclusion: A validity generalization perspective of the literature supports the use of MCAT and undergraduate GPA for selection to medical school. Copyright © 2007 by Lawrence Erlbaum Associates, Inc.