The predictive validity of the MCAT exam in relation to academic performance through medical school: a national cohort study of 2001-2004 matriculants

Acad Med. 2013 May;88(5):666-71. doi: 10.1097/ACM.0b013e3182864299.


Purpose: Most research examining the predictive validity of the Medical College Admission Test (MCAT) has focused on the relationship between MCAT scores and scores on the United States Medical Licensing Examination Step exams. This study examined whether MCAT scores predict students' unimpeded progress toward graduation (UP), which the authors defined as not withdrawing or being dismissed for academic reasons, graduating within five years of matriculation, and passing the Step 1, Step 2 Clinical Knowledge, and Step 2 Clinical Skills exams on the first attempt.

Method: Students who matriculated during 2001-2004 at 119 U.S. medical schools were included in the analyses. Logistic regression analyses were used to estimate the relationships between UP and MCAT total scores alone, undergraduate grade point averages (UGPAs) alone, and UGPAs and MCAT total scores together. All analyses were conducted at the school level and were considered together to evaluate relationships across schools.

Results: The majority of matriculants experienced UP. Together, UGPAs and MCAT total scores predicted UP well. MCAT total scores alone were a better predictor than UGPAs alone. Relationships were similar across schools; however, there was more variability across schools in the relationship between UP and UGPAs than between UP and MCAT total scores.

Conclusions: The combination of UGPAs and MCAT total scores performs well as a predictor of UP. Both UGPAs and MCAT total scores are strong predictors of academic performance in medical school through graduation, not just the first two years. Further, these relationships generalize across medical schools.

Publication types

  • Evaluation Study

MeSH terms

  • Achievement*
  • Cohort Studies
  • College Admission Test*
  • Education, Medical, Undergraduate / statistics & numerical data*
  • Educational Measurement
  • Humans
  • Logistic Models
  • Reproducibility of Results
  • United States