The Limitations of the GRE in Predicting Success in Biomedical Graduate School

Area of Science:

• Biomedical Sciences
• Graduate Education

Background:

• Graduate admissions committees historically rely on GRE scores.
• The predictive validity of GRE scores on graduate student success is not well-established, especially in biomedical research.

Purpose of the Study:

• To investigate the predictive validity of GRE scores on various graduate student success metrics in biomedical Ph.D. programs.
• To determine if GRE scores can predict outcomes such as graduation, exam passage, time to defense, publications, and grant acquisition.

Main Methods:

• Analysis of data from Vanderbilt University Medical School's biomedical umbrella program.
• Statistical modeling to assess the predictive power of GRE scores while controlling for other admissions data.

Main Results:

• GRE scores did not predict Ph.D. graduation, qualifying exam success, time to defense, conference presentations, first-author publications, or grant/fellowship awards.
• GRE scores showed moderate predictive ability for first-semester grades and weak to moderate predictive ability for graduate GPA and some faculty evaluation components.

Conclusions:

• GRE scores have limited utility in predicting key indicators of success for biomedical Ph.D. students.
• Biomedical Ph.D. programs should reconsider their reliance on GRE scores for admissions and student progress evaluation.