For a few years now, there has been a debate raging in the science community: Should admissions at universities and colleges drop the requirement that students need to take the GRE for graduate schools? This is a conversation that has been steeped in inherent and implicit biases, data, and a gross misunderstanding of how standardized tests impact students. In this piece, I won’t go into details about how our own biases affect the decisions we make and opinions we form. Rather, the purpose of this post is to pull together the available data that show that the GRE is, in fact, an ineffective tool to predict the success of graduate students. To further this narrative, there are also personal stories from people who have taken the GRE included in this post.
Here is a list of US-based Geology/Geography programs that have dropped GRE.
First, a bit of background about the GRE test. The GRE, which stands for Graduate Record Examinations, is a standardized test that students take who are applying for graduate schools, including law and business school. The GRE itself is created by ETS, a company that touts itself as creators of ‘some of the most well-known and widely used educational assessments in the world’. There are different types of GRE tests, called Subject Tests, depending on what the student wants to focus on as their career path: Biology, Chemistry, Literature in English, Mathematics, Physics, and Psychology. There is no subject test currently for the geosciences, so students who wish to pursue a graduate degree in this field just take the General Test.
The General Test is broken down into 3 major categories: Verbal Reasoning, Quantitative Reasoning, and Analytical Writing. Each of these categories is supposed to test the student’s ability to draw conclusions from discourse and reasoning, summarize text, and distinguish major from minor points, measure the ability to understand, interpret, and analyze quantitative information, apply mathematical skills to problem-solve, and measure critical thinking and writing skills. These are just a few things the GRE is designed to measure, as stated on their website (https://www.ets.org/gre/revised_general/about/).
To take the GRE test, there is, of course, a fee involved. The current prices to take the test as of July 1, 2020 are as follows:
|All other parts of the world
(All data from https://www.ets.org/gre/revised_general/about/fees/)
Students cannot take a GRE test anywhere. There are specific testing centers that distribute the GRE, and they do so about 3 times a year in September, October, and April for the United States, but can be variable depending on other countries. Often, the testing centers are located in larger cities, away from more rural areas. The test itself takes about 3 to 3.5 hours, and students are not allowed to bring any snacks or drinks into the testing center with them. These restrictions are variable and can be more or less strict depending on the testing center.
After the student has taken the test, they receive part of their scores right away. The written portion of the exam is scored by a group of panelists who later give them the score on the written portion. The student then must immediately decide if they want to send those scores to the graduate schools they are applying to. The student doesn’t have to send their scores to any school if they don’t want to or feel like they need to retake the test. If you don’t send the scores at this time, it does cost extra to send them at a later time. Scores are reported on a scale from 130 to 170 in 1 point increments for the verbal reasoning and quantitative reasoning part of the exam, and from 0 to 6 in half point increments on the analytical writing portion.
Although the student can choose to send their test results to multiple schools, it does cost an additional $27 to send scores to additional schools. As stated on ETS’ website, these requests for additional scores are not refundable, cannot be canceled, nor can they be changed.
From the above section, it should be clear that the GRE test is serious business, as it costs quite a bit of money to take, takes a large amount of time, and is largely uncomfortable (seriously, no snacks?!?). The test is supposed to be an indicator of student success in graduate school, but there have been studies published that say otherwise. In addition, there are problems with the test, as it is not an equal predictor of success for men and women, nor among underrepresented groups. In this section, I’ll summarize some of those studies and their major findings.
The GRE as a Poor Predictor of Graduate Student Success
First, let’s start with the obvious assumption that the GRE test does predict success of upcoming graduate students. One of the earliest studies looking at the predictive power of the GRE was by Dunlap (1979). This study found that the best predictor a student success was the student’s performance on the basis of faculty interviews and undergraduate GPA. The GRE was a weak predictor of success. Dawes (1971) showed that GRE scores can be good predictors of grades and faculty evaluations, but only for first-year graduate students’ performance in psychology. Another study by Wood and Wong (1992) showed that the GRE, by itself, accounted for slightly less than 10% of the variation in the criteria of graduate performance against which the GRE was validated. This study also only looked at psychology students. A more recent assessment of student GRE scores as predictors of success in psychology departments was conducted by Sternberg and Williams (1997). These researchers at Yale University asked graduate student advisors to rate their graduate students on their analytical abilities, creative abilities, practical abilities, research abilities, and teaching abilities. The researchers also computed the students’ first-year, second year, and combined GPAs on a scale of 4 (high pass) to 0 (fail). Similar to the Dawes (1971) study, the data from Sternberg and Williams (1997) indicated that the GRE was a modest predictor of grades, but only for the students’ first year in graduate school. The GRE scores were not found to be useful in predicting other arenas of graduate performance, such as analytical, creative, practical, research, and teaching abilities.
Similarly, a study conducted by researchers looked at the predictive power of GRE scores in construction management programs in the United States. The study, conducted by Wao and others (2016) tried to correlate the success of graduate students (success meaning they finished their degree within two years and not drop out). They, too, found that the GRE scores were not correlated with graduate student success, and thus recommended that admissions committees should reassess requiring the GRE scores at all.
Personal Stories & Experiences
In high school, I was not a great student, and did terrible in math classes (my high school math teacher once told me to get away from his desk when I asked for help because I was ruining the signal on the TV while he watched basketball, DURING CLASS). When it was time for me to take the GRE, I was already working 20-30 hours per week to put myself through community college and help my family with expenses. I had to take a day off of work, paid $200 for the exam, and had to drive 45 minutes to the nearest testing center. I was so nervous about taking the test, I sweat the entire time even though the room was chilly. At the end of my exam, my scores didn’t meet the minimum most graduate programs required (300 points). I ran to my car, crying, and called my mom telling her I had failed. I couldn’t really afford to take the test again, but I did so twice more. I received about the same scores all three times, but in the end, I was out over $600 that I needed for school and my family. Today, I’m a Postdoctoral Fellow, and have personally grown and accomplished so much during my short academic career. No, the GRE was not a predictor of my success, but rather highlighted the fact that I came from an area with subpar high schools and from a family with lower socio-economic status.
–Adriane Lam, Postdoctoral Fellow, University of Binghamton
I an above average high school student, I didn’t significantly apply myself because I was easily distracted and often bored in class. But I did my work, just not to the best of my ability. I never was a strong test taker, and didn’t have excellent ACT scores (25). I went on to a local public institution and really did awful my first year. I was NOT able to coast through like I did in high school. This means I bombed my first year of my undergraduate studies. I spent the next 4 years (I spent 5 as an undergraduate) working to raise my GPA. I graduated with a 2.86 GPA. I have to reiterate that I am a terrible test taker, I get so anxious, my study habits were variable and awful. I struggled. I didn’t know that I could even ask for help until the last few years when I found supportive mentors. I dreaded the GRE, I got workbooks, flashcards, and even recorded myself reciting definitions of terms so I could listen to myself on my commute to work and school. The testing center wouldn’t let me bring in tissues or chapstick and said I could either take in my sweater or I had to take it off because I couldn’t take it off in the testing center. These are all comfort items for me and I felt naked and uncomfortable heading into the testing center. I took the exam and I got an okay score, around 1200 (I took the old version). One of my mentors said I needed to take it again because that wasn’t competitive. So, regardless of the expense, I scheduled another test. I did worse. This was a waste of my time and money, neither of which I had an abundance of. Today, I am a collection manager at an established museum and research institution – regardless of my GPA or GRE scores, I managed to achieve my goals.
–Jen Bauer, Ph.D., University of Michigan Museum of Paleontology
My story is from applying to PhD programs in 2013, both involving schools in the top 10 Earth Science PhD programs. At one I was told by the PI I was waitlisted because of my scores (which I can only assume meant my GRE scores, my GPA was a 3.8), only to then be admitted 1 week before the decision deadline after others had declined and I had accepted elsewhere (which was rather embarrassing). I was later told by the (junior) PI that he seriously regretted waitlisting me, which was quite humiliating. The second program was told when I visited informally before applications were due that there was a ‘formula’ involving GPA, quant and writing score that was unofficially used. I was below the threshold, told to retake the GRE before applying, all while trying to finish my MS degree. Knowing there was a threshold I needed was incredibly stressful.
I actually had a very positive experience with the GRE even if I’m not convinced of its utility as an application requirement overall. I’m also concerned about what standards are replacing it and how inequality is built in those too. Before applying to a geoscience grad program, I was nervous about “belonging” and being successful in the field. I didn’t do science undergrad, I was a little older, and black. Hitting the marks on my GRE helped reassure me that I was in fact qualified and belonged. I’m not sure how much my score actually factored in to my admission. Re the test itself, I bought an inexpensive test prep book (could have got it at library too), took it at a location walkable from my home, easy peasy. Test location was at an HBU so even that was encouraging. Largest drawback for me was the cost of the exam itself. That’s just one experience and if the data says the test is a useless barrier, then I trust the data. I always felt weaker in school on the quantitative side of things. The GRE wasn’t a class. It’s not math I’d apply in my field. It was a nut to crack, something to persevere and figure out. That’s something I do in my PhD, everyday. That’s how the GRE helped me feel ready.
I went to undergrad on a full needs-based scholarship, because my family had no money to send me to college. I worked part time jobs, sometimes more than one during a semester, through all 4 years of college. I wanted to apply to grad school, and spent nearly a month’s pay signing up for the GRE. Unfortunately, in the weeks leading up to the GRE, my grandfather died quite unexpectedly and I ended up needing to reschedule it because my mother needed help cleaning out his house out of state. I had to pay an extra $50 to reschedule (which wasn’t easy to scrape up). It took me over an hour and a half by bus to get to the exam because they didn’t offer the test where I went to undergrad (they still don’t)-they offered it in a city a short distance away, but without a car, I had to rely on public transportation. This added a lot to the stress of the day, seeing as the exam started at 8AM.
The amount of money it cost me to take this exam left me struggling to afford food, toiletries, and medical bills. I then had to figure out how to pay for the steep fees for the applications themselves. I didn’t know that you could send your scores for free *if* you did it the day you took the exam until I was already taking it. Because of that, I had to resend my scores after the exam for even more extra fees (I want to say $10 per school to send an email with my scores, but I can’t remember what the costs were back in 2010).
I got into grad school (M.S.) with a relatively low score in the quantitative section. Frankly, between taking 5-6 classes, extracurricular activities, and working (my senior year, I taught a lab section of a course for pay, tutored athletes, and worked as an office assistant in a music department all at the same time), I didn’t have any time to study for the GRE. I used an old copy of the GRE study book from the library when I had a free hour or two, but I didn’t study as much as I could have if I hadn’t been working.
I took it again when I applied to a Ph.D. program- I only did this because a mentor of mine told me I’d never get into a Ph.D. program with my quantitative scores (I scored in the 99% in the qualitative/literature section, but everyone said that didn’t matter in the STEM fields). I, again, had to spend just over two hundred dollars (my monthly take home pay as a student was around $850-$900/month, so this was tough to do, mind you). I had a car this time, but still had to drive an hour each way to take the test. This time, I was able to get a study book that I found from a thrift store and I was able to do a little bit better in quantitative, but not enough that it was worth the extreme stress I went through trying to figure out how to pay for the exam. Today, I’m an assistant professor at a university that has a significant number of low income students and the costs for the GRE have only gone up since I took it, while wages haven’t changed with those rising costs. I want to see a world where my students don’t have to forfeit time to study just so they can save up enough to take the test.
–Sarah Sheffield, Ph.D. The University of South Florida
As someone who grew up taking standardized tests every year starting in 4th grade, I have always been a good standardized test taker, (other kinds of exams not so much, my test anxiety really didn’t start until my undergrad when I felt that my scores would really determine my future) so the idea of the GRE didn’t bother me. At the time, it very much just felt like another annoying step to be able to apply to graduate programs. However, since the exam wasn’t offered in the same city that my college was located in, I had to drive 40 minutes to the exam after my classes ended that day to sit in a freezing cold cubicle with noise cancelling headphones to block out the sound of anxious typing from all the other students taking the exam with me.
One thing I did appreciate about the exam was knowing my quantitative and qualitative scores immediately after finishing the exam. I knew that my quantitative score would be my lowest (hello, math insecurities) but did hope that it would be a competitive score-an arbitrary number in my head that I had gotten from looking up “acceptable GRE scores for paleontology programs” on Google. My quantitative score was lower than what that number in my head was, but on my way out the door I made the decision that knowing how I test in math and how much better I would need to do to raise that score even a few points was not worth the stress and the $200 to retake the exam. I was 100% comfortable with this decision, but still nervous that my score might deter universities from accepting me for my Master’s. While the scores did not prevent me from being accepted into a Master’s program, I do think that they played a role in some of my rejections from PhD programs because they were schools with strict cut offs on GRE scores. However, I was accepted into a PhD program and have yet to see the need for anything I learned specifically to do well on the GRE.
–Maggie Limbeck, PhD student, University of Tennessee
Peer-reviewed Literature & References Mentioned
- Dawes, R. M., 1971. A case study of graduate admissions: Application of three principles of human decision making. American Psychologist 26, 180-188.
- Sternberg, R. J., and Wiliams, W. M., 1997. Does the Graduate Record Examination Predict Meaningful Success in the Graduate Training of Psychologists? American Psychologist 52(6), 630-641.
- Wao, J. O., Bivens, K. B., Eng, M., Hunt III, R., Ries, R., Carroll, J., and Ozbek, M., 2016. Predicting the performance and success of construction management graduate students using GRE scores. In Proceedings of the 52nd Associated Schools of Construction Annual International Conference, Brigham Young University, Provo, Utah.
- Wood, J. L., and Wong, A. C., 1992. GRE Scores and Graduate School Success. International News and Notes.
- Ledford, S. H., M. M. Monteagudo, A. N. Flores, J. B. Glass, and K. M. Cobb (2020), #GeoGRExit: Why geosciences programs are dropping the GRE, Eos, 101, https://doi.org/10.1029/2020EO145223. Published on 04 June 2020.
- Moneta-Koehler, L., Brown, A. M., Petrie, K. A., Evans, B. J., & Chalkley, R. (2017). The limitations of the GRE in predicting success in biomedical graduate school. PloS one, 12(1), e0166742.
- Petersen, S. L., Erenrich, E. S., Levine, D. L., Vigoreaux, J., & Gile, K. (2018). Multi-institutional study of GRE scores as predictors of STEM PhD degree completion: GRE gets a low mark. PloS one, 13(10), e0206570.