Degree In Sight

GRE

The Graduate Record Examination has been inducing anxiety attacks in aspiring graduate students since it was first administered in 1949. But even though GRE scores are often the first thing admissions committees consider when they evaluate a prospective student, a debate is raging over what the tests tell us about a student's true academic and professional potential.

Fortunately, the question of the GRE's validity has spawned its own subgenre of academic literature. Culled from the empirical data published over the last decade, here are a few things we know — and don't know — about how well this examination predicts the future.

  • There's no way to know whether a low GRE score translates into failure. Students with the lowest GRE scores aren't admitted into graduate psychology programs, so they never become psychologists. As a result, there's no way for researchers to know whether the very lowest-scoring students would have gone on to prove their predictors wrong. "It's certainly true that there's a restriction of range," says Robert Sternberg, PhD, a psychologist and provost at Oklahoma State University who's examined the GRE in his research. "If you had [greater] range, the predictive value of these studies would increase." This catch-22 makes some researchers wonder why the GRE looms so large in admissions decisions to begin with.

  • GRE scores do help reveal which students will do well in the classroom and which won't. Many studies have found that students with lower GRE scores are more likely to fail their preliminary examinations. Students with total scores higher than 1,167 usually end up with better grade-point averages than their classmates, more published papers and better ratings from faculty, according to a 2004 study by Dale Phillips, PhD, and Kristen McAuliffe in the School Psychologist Newsletter (Vol. 52, No. 2). "Based on the data that's out there, the GRE is consistently the strongest [predictor] we have of student success," says Nathan Kuncel, PhD, author of a 2001 GRE meta-analysis published in Psychological Bulletin (Vol. 127, No. 1).

  • The GRE's predictive powers diminish over time. In his 1997 study published in American Psychologist, (Vol. 52, No. 6) Sternberg found that GRE scores tell us most about how students will perform in the first year of grad school. That's because "you need the same kinds of skills in introductory courses as you do for the GRE," he says — namely, the basics, such as general reading and quantitative skills — but not necessarily imagination. As grad school grinds on, more abstract skills become increasingly important — for instance, intuiting which journal would be most likely to accept a particular kind of paper. "The GRE doesn't measure that," says Sternberg.

  • GRE scores are less reliable when it comes to predicting whether a student will eventually complete a psychology program. The exams may predict classroom performance fairly well, but grades aren't everything. Several researchers have found that the GRE tells us less about whether someone will finish school. Phillips and McAuliffe, for instance, found that GRE scores didn't differ much between students who eventually graduated and students who didn't. "Nothing predicts finishing very well," says Kuncel. In many cases, students drop out because of life circumstances — leaving to take care of an ailing parent, for example. Phillips's and McAuliffe's study support that claim: Only 9 percent of students who dropped out said it was because they couldn't hack the coursework.

  • The GRE's subject test in psychology tells us the most about a student's potential. Kuncel's meta-analysis found that the subject test outperformed the verbal, quantitative and analytical tests when it came to predicting students' grades and whether they'll eventually earn a degree. "That only makes sense," says Stephen J. Dollinger, PhD, a psychology professor at Southern Illinois University who's studied the validity of the GRE. "The student who enters graduate school knowing more psychology should have an easier time starting a thesis [and] passing prelims."

    But the subject test — usually 205 multiple-choice questions — measures more than just psychology knowledge, says Kuncel. A student who is especially passionate about psychology may outperform a fellow student who has been deemed brighter by the GRE's verbal and quantitative tests. Still, most master's programs and about half of doctoral programs in psychology don't insist that you take it. According to Kuncel, many admissions programs probably worry that they'd alienate prospective students by giving them another hoop to jump through.

    "That's the irony," he says. "The best single predictor is also not required at many programs." Still, Kuncel "highly recommend[s]" that prospective students take the test anyway, if only to convey their enthusiasm for the field.

In the end, your GRE score will certainly affect which program you get into, but it won't necessarily predict how well you do once you get there. Grad school success requires too many skills that can't be examined by the GRE, such as your ability to work well with lab mates, choose a research topic that gels with your professor's interests or pick up new therapy skills. And your test scores won't tell you much about how you'll actually perform as a professional researcher or practicing psychologist, either.

"Look at the great ones versus the not-so-great ones," says Sternberg. "The best of them come up with creative approaches. They're not necessarily good at taking tests."


Dave Jamieson is a writer in Washington, D.C.