Is there a future for the GRE?

Each year about half a million people around the world with dreams of becoming graduate students take the Graduate Record Examination (GRE) General Test. Many graduate schools require the standardized test, which assesses skills in critical thinking, analytical writing, verbal reasoning, and quantitative reasoning, as part of their application process. Some schools also require a subject GRE, which tests knowledge in a particular field. But in the past few years, a number of graduate programs, particularly in the biomedical sciences, have decided to drop the GRE as a criterion.

The trend, known on social media as #GRExit, is starting to catch on in chemistry graduate programs. Last year, at least six chemistry departments jettisoned the GRE requirement for students applying for the 2019–20 academic year. These departments point to the exam’s high cost, as well as results from recent studies that suggest GRE scores are poor predictors of grad school success, as reasons for abandoning the test. They also point out that women and underrepresented minorities on average don’t perform as well on the exam as white men do. The test, they say, puts these applicants at a disadvantage at a time when chemistry departments are making diversity a priority.

The goal of an admissions committee is to offer spots in its program to applicants who are likely to finish the program and earn a PhD, says Sandra Petersen, director of the Northeast Alliance for Graduate Education and the Professoriate. So these committees should base their decisions on factors that indicate which students are most likely to succeed. Applicants’ GRE scores, she argues, simply don’t do that.

Petersen, along with colleagues from three other universities, looked at the GRE scores of 1,805 students who either were from the US or were US permanent residents and started doctoral programs in science, technology, engineering, and mathematics (STEM) in the 2000–2005 period. They then looked to see whether there was any correlation between GRE scores and completion of the doctoral program. And they examined how long students took to earn their degrees. It was the first study to look broadly at graduate programs in STEM fields (PLOS One 2018, DOI: 10.1371/journal.pone.0206570).

The group found no difference between men’s and women’s scores on the verbal portion of the GRE. Men performed better than women on the quantitative portion of the GRE, but they were not more likely to complete a doctoral program. About 60% of both men and women who entered STEM doctoral programs graduated with a PhD. Men and women also took about the same time to finish—just under six years.

For women, there was no correlation between who scored well and who stayed in graduate school. For men, the team found one correlation with GRE scores: men who scored in the top 25% of the GRE’s quantitative section were more likely to leave graduate school without a degree than men who scored in the lowest 25%. When Petersen looked at the scores and completion rates of just engineering students, for whom one would expect quantitative skills to be a more likely indicator of success, the trend among men was still the same.

“We’re scientists and engineers. We should be data based,” Petersen says. “If the GRE doesn’t predict PhD completion, time to degree, or who will leave during the first year, what information does it give you?”

While no one has done a study focused solely on GREs and chemistry graduate programs, one has been done for physics. A team led by Casey W. Miller, associate dean for research and faculty affairs at the Rochester Institute of Technology, published a study last month examining GRE scores and PhD completion rates. In an analysis of 3,962 physics doctoral students who entered programs in the 2000–2010 period, the researchers found no correlation between program completion and the verbal portion of the GRE General Test or the subject GRE test for physics. Students with the highest GRE quantitative scores were more likely to complete their programs than those with the lowest scores, but the correlation was small (Sci. Adv. 2019, DOI: 10.1126/sciadv.aat7550). Miller thinks admission committees for physics PhD programs might be turning down applicants on the basis of their GRE scores, even though the scores don’t tell them which students will be successful.

David G. Payne, vice president of global education for the Educational Testing Service (ETS), which administers the GRE, says the test was designed to identify candidates who have the academic ability needed for graduate school. “The GRE test was never intended to predict PhD completion,” he says. Students typically leave graduate school for factors that have nothing to do with academics, such as changes in their families, financial concerns, or unhappiness with the program, according to a study by the National Center for Education Statistics. The GRE can’t predict those things, Payne says.

But many faculty members say the GRE is a distraction. “What I have found in my academic career is that some of my students may not have the best GRE scores, but they have lots of publications and graduate in 4½ years,” says Julia Chan, graduate recruiting chair for the Chemistry and Biochemistry Department at the University of Texas at Dallas. The department stopped requiring GRE scores from its applicants in the fall.

David Shultz, director of graduate studies for North Carolina State University’s Chemistry Department, says GRE scores “serve no purpose,” which is why the department decided to drop the requirement late last year. He notes that in the 26 years he’s been at the school, he’s never seen a case in which a student with a strong record of accomplishments, including research experience and publications, was denied a place in the program simply because of a low GRE score.

Shultz says the department is seeking students who know how to work in the laboratory, so applicants with a track record in research but low GRE scores are more attractive than applicants with high GRE scores and no research experience. “Those are the students who are going to quickly get results and make fast progress toward their PhD,” he says. “Research is expensive, so we really feel external pressure to select students that have research experience.”

A lot of data are building up that indicate what factors track strongly with success in graduate school, “and it certainly isn’t the GRE,” says Jason Gestwicki, director of the Chemistry and Chemical Biology Graduate Program at the University of California, San Francisco, which dropped its GRE requirement last year too. He notes that the department plans to track how getting rid of the GRE influences which students are accepted to the program.

When students take the GRE, they’re all taking the same exam, Gestwicki says. “What we’re doing by removing it is using a more holistic evaluation of a student, and that is prone to implicit biases. We shouldn’t point out that we think the GRE is disproportionately affecting disadvantaged students and then assume that by removing it we’re going to do better with those students. We should be much more statistically rigorous” and check that improvements are actually occurring. If dropping the GRE has a detrimental effect on diversity, he says, the program will reinstate the requirement.

Making admission decisions on the basis of letters of recommendation, research experience, or the selectivity of an applicant’s undergraduate institution usually works against underrepresented students, ETS’s Payne says. “There are many things that programs have to be mindful of if they’re not going to use the one standardized measure that can be used,” he says.

“One reason the GRE is particularly useful is that we get applications from many schools around the world with which we may not have personal familiarity,” says Robin L. Garrell, vice provost for graduate education and a chemistry professor at the University of California, Los Angeles. Garrell is also a member of the GRE board, an unpaid advisory position. The GRE “provides insights into students’ general readiness and basic skills,” she says. Also, Garrell points out, the GRE is not subject to differences in grading practices that vary across institutions. For example, the grade point averages at private schools are much higher than they are at public institutions, which giving an advantage to students who can afford to attend private colleges and universities.

Even so, the faculty at the chemistry departments dropping the GRE say the test is a small factor when they evaluate graduate school applications. “The information the GRE was providing us, we thought it was a data point,” says Jacob Ciszek, graduate program director in the Chemistry and Biochemistry Department at Loyola University Chicago. The department decided it could eliminate that data point and in doing so attract more applicants, particularly foreign students, who often must travel substantial distances to find GRE testing centers. “We think there are other data points we can use in its place,” he says, such as Skype interviews.

The GRE was simply becoming a box to tick, says Dana Baum, graduate program coordinator for the chemistry program at Saint Louis University. “If someone had a very low GRE score and we weren’t going to admit them, [the decision] was not just based on that. It was based on things we saw throughout the application,” she says. Looking to attract a broader range of students, she and her colleagues decided it didn’t make sense to require the GRE. But Baum adds that the decision to drop the exam isn’t set in stone. “We can assess in a couple years if it’s hurt us or it’s helped us.”

“We’re in the middle of an experiment,” says Susanna Widicus Weaver, the director of graduate studies for Emory University’s Chemistry Department, which also no longer requires GRE scores. She admits that she was resistant to the change initially. “It’s hard for me on a personal level because I’m a physical chemist, and that quantitative score is actually a fairly good indicator of students’ abilities in first-year classes. But beyond that it’s not a predictor at all of graduate school success.”

Another factor that influenced these chemistry departments’ decisions to drop the GRE is the exam’s price tag. It costs $205 for US-based applicants to take the GRE General Test. The same fee applies in other countries, with the exception of Australia, China, Nigeria, and Turkey, where the test costs $220–$255. Those who take the GRE subject test in chemistry, which only a few chemistry departments require but many recommend, pay an additional $150 for that exam. For the base fee, ETS will send test scores to four schools. Applicants who want their scores sent to more than four schools pay $27 for each additional report. Some graduate school applicants also take GRE preparation courses, which can cost more than $1,000.

“During our discussions, when we pulled up the cost of the GRE, a lot of people in the room were shocked,” Saint Louis University’s Baum says.

“I found that the cost of the GRE, especially the cost of preparation, is a burden, especially for students from underrepresented or disadvantaged groups,” UCSF’s Gestwicki says.

ETS provides free and relatively low-cost preparation materials on its website. And students who meet certain financial need requirements can get a 50% reduction in the cost of both the general test and the subject test.

Olivia Wilkins, a graduate student in chemistry at the California Institute of Technology, points out that graduate students don’t make much money. She says she often wishes she could recoup the few hundred dollars she spent taking the general and chemistry subject GREs. What’s more, she says, her GRE scores have made her less confident in her academic abilities. Although she did well on her general GRE, the first time she took the subject test, her score was disappointing. She didn’t think it was high enough to earn her a spot in a chemistry graduate program.

So the following year, when Wilkins was a Fulbright Scholar in Germany, she decided to retake the chemistry GRE. The exam wasn’t offered in Cologne, where she was studying, and Wilkins—who was pregnant at the time—had to make an overnight trip to a testing center in Brussels, which was four hours away by train. “I wasn’t feeling well and I ended up doing worse,” she recalls. Even though she’s doing well in graduate school, the low score gives Wilkins doubts about herself. “It still is a shadow that follows me around and makes me question whether I really know enough chemistry to get a PhD,” she says.

Wilkins teamed up with Ashley Walker, an undergraduate studying chemistry at Chicago State University, to create a list of chemistry graduate programs and their GRE requirements. They hope the list will be a guide for navigating the graduate school application process.

Walker says she’s seen the studies about how women and minorities fare on the GRE, and as a woman of color, she feels she’s at a disadvantage. “I’m actually quite nervous,” she says. Walker is a senior looking at graduate programs, but she says she’s been preparing for the exam since her sophomore year. She’s planning to apply to some schools that require GRE scores and some that don’t.

Both Wilkins and Walker say they’d like to see all schools do away with the GRE requirement. They say the exam tests you only on how much you remember and how quickly you can recall information during the marathon exam time of nearly four hours. “Realistically, that’s not how we do chemistry,” Wilkins says.