January 19, 2017

GRE may be poor predictor of science success: study

Since it was created in 1949, the Graduate Record Exam (GRE) has been one of the most widely used requirements for admission to U.S. graduate schools. As a predictor of performance and success in graduate school, however, the exam is not without its critics.

Since it was created in 1949, the Graduate Record Exam (GRE) has been one of the most widely used requirements for admission to U.S. graduate schools. As a predictor of performance and success in graduate school, however, the exam is not without its critics.

Now, researchers at Vanderbilt University’s School of Medicine and Peabody College of education and human development report that the GRE is a poor predictor of success in Vanderbilt’s Interdisciplinary Graduate Program (IGP). Over-reliance on these scores could reduce student diversity, they warn.

“These findings suggest admissions committees of biomedical doctoral programs should consider minimizing their reliance on GRE scores to predict the important measures of progress in the program and student productivity,” they conclude last week in the on-line journal PLOS ONE.

“Instead of using a GRE cutoff, admissions committees are encouraged to use a more holistic approach with an emphasis on subjective criteria such as letters of recommendation,” says first author Liane Moneta-Koehler, Ph.D., a postdoctoral research fellow in the medical school’s Office of Biomedical Research Education and Training (BRET).

The Vanderbilt IGP is an umbrella admissions program that encompasses 11 graduate programs: Biochemistry, Biological Sciences, Cancer Biology, Cell and Developmental Biology, Cellular and Molecular Pathology, Chemical and Physical Biology, Human Genetics, Microbiology and Immunology, Molecular Physiology and Biophysics, Neurosciences and Pharmacology.

The researchers examined the records of 683 students admitted to the IGP from 2003 to 2011, and analyzed how well the three GRE subtests (quantitative, verbal and analytical writing) predicted progress through the program and research productivity compared to other objective admissions criteria such as students’ undergraduate grade point averages (GPA).

They found that none of the GRE subtests predicted the likelihood that the students would earn their Ph.D. degrees, but their undergraduate GPA did.

GRE scores also did not predict research productivity, as measured by the number of scientific publications, presentations, grants and fellowships, but they did moderately predict first-semester grades, and they were weak-to-moderate predictors of graduate-school GPA and some elements of faculty evaluations.

Importantly, underrepresented minorities tend to have lower GRE scores. Thus relying too heavily on the test will exclude many of these students from graduate school, the researchers concluded, and “may in fact run counter to the goal of diversifying the biomedical research workforce.”

Moneta-Koehler’s co-authors were Roger Chalkley, D.Phil., senior associate dean for BRET, Kimberly Petrie, Ph.D., director of the BRET Office of Career Development, Abigail Brown, Ph.D., director of Outcomes Research, and Brent Evans, Ph.D., assistant professor of Public Policy and Higher Education at Vanderbilt’s Peabody College.