A large misconception that has not been addressed about the GRE

[The_Old_Wise_One]

Hello all,

First off, good luck to everyone out there applying for grad programs in psychology. It may be a lot of work, but it will be worth it when it pays off! 

Now, on to the point. I have been lurking around on GradCafe for some time now, but I just recently created this account in order to make some posts and such. After reading through the numerous postings, I have come to realize that a lot of people ask or comment on the GRE, both general and subject versions. One of the common discussions is that the general GRE is a bad predictor of graduate performance, it only measures one's ability to take the GRE, etc. Everyone is quick to demean the test. While these arguments may be valid, there is one thing that I have not seen mentioned––funding. Graduate programs, particularly clinical psychology Ph.D. programs (I say this because I am most familiar with them), oftentimes fund the entire education (up to 5 years) of those who are admitted to the program. Actually, if you check the admissions data given by most schools for clinical psych Ph.D. programs, it is rare to see data showing that a student's education was not 100% funded. Now, where does the funding come from? Well, it comes from a pool of money that the university keeps aside for funding graduate students (makes sense).

What most people may not know is that funding is: (1) an important factor in students deciding which programs to apply to, and (2) spread out across departments. While many people here likely understand #1, I doubt that many know about #2. This means that whether you are seeking admission and funding within a doctoral physics program or doctoral psychology program, the funding money comes from the same pool. That being said, what goes into deciding how much money is offered to each department? This is where things get interesting! Since most doctoral programs cost about the same (within the same university, of course), admissions committees need to come up with ways in which they can compare students across disciplines (it's starting to come together....). As many of you may know, GPA would be a terrible way to compare people from different fields. Is a 3.8 in undergraduate theoretical mathematics equivalent to a 3.8 in undergraduate psychology? Absolutely not! So, what are our options? Well, here is where the general GRE comes in!

The GRE is a baseline used to compare students across disciplines. It is simple as that. It is one of the only metrics that makes this possible. This is why some scholarships and outside funding options for masters programs require general GRE scores. This is why programs that do not fund their students are less likely to require general GRE scores. This is why the general GRE is an important factor when it comes to selecting doctoral candidates! Graduate programs have to compete with one another for a piece of the funding, thus we are all studying and banging our heads against keyboards trying to get a 320+ (and all the while wandering why this stupid test exists!). 

I hope that what I said above gives the community here a bit more insight into why programs require tests like the GRE. The test was created for the sake of making fair comparisons across disciplines, and regardless of how silly it may initially seem, I hope that people can realize how necessary it is. 

Please post comments below for good discussion! Also, post about any insights you may have that could help others understand the application process in more detail. Thanks for taking the time to read  

[Oshawott]

I think you may be overgeneralizing this. While GRE may be used in some schools to allocate funding, other schools do not. Various schools in Canada have a guaranteed minimum that the graduate school will offer, and whether the student gets more largely depends on the department's resources. For instance, at the University of Toronto, graduate students are guaranteed a minimum of $15k (might have increased to $17k after the TA strike). Some departments are poor and will only give that base funding, and tuition is removed from it. Other departments (e.g., the Psychology Department) will guarantee that base funding plus the cost of tuition. And then there are other departments that offer way more money (e.g., Chemistry).

How do these other departments offer more than others? GRE's can't possibly be the deciding factor--Psychology requires the GRE, yet Chemistry doesn't and they have more money to offer students. Departments get a lot of money from the number of students (grad and undergrad) enrolled in their program/courses (hence the generally good funding package in large Psychology departments across Canada) as well as the research overhead (likely the primary source of Chemistry's money, in combination with its large intro classes).

That being said, what goes into deciding how much money is offered to each department? This is where things get interesting! Since most doctoral programs cost about the same (within the same university, of course), admissions committees need to come up with ways in which they can compare students across disciplines (it's starting to come together....).

I think there's also a bit of a fallacy here. If GRE's are so important to get funding, then why don't all departments not accept students above a certain cut-off? How do you decide what department to fund more if all the graduates have a combined GRE of 320+?

Graduate programs have to compete with one another for a piece of the funding, thus we are all studying and banging our heads against keyboards trying to get a 320+ (and all the while wandering why this stupid test exists!). 

I would think that a university would have better data than a non-predictive score to decide what programs to invest in. A couple things off the top of my head:

1) Graduate Success--measured by placements in Academic or other highly sought after non-academic positions.

2) Research productivity--ostensibly measured by publication output. While some fields do lend themselves to faster publications, you can still normalize and look at how the particular department fairs in terms of research productivity compared to other departments in the same discipline.

3) Impact of research (perceived or real...)--maybe a department takes long to publish, but when they do, its generally a bigger impact in the field (citation count?) or society.

Yes, I am shooting myself in the foot here by suggesting other metrics that may not be useful, but at the very least, I would think that metrics like number of publications (relative to the field) and citations are more indicative of departmental success than the GRE score of the incoming cohort. The last two points would at least be a better explanation why STEM fields get more funding from the university than the social sciences.

I hope that what I said above gives the community here a bit more insight into why programs require tests like the GRE. The test was created for the sake of making fair comparisons across disciplines, and regardless of how silly it may initially seem, I hope that people can realize how necessary it is. 

Given that we're in the Psychology sub-forum, I'd think the obvious reason would be that the GRE's just serve as an easy filter for applications. Even if they don't predict anything, committees, especially at larger, more prestigious schools, can't go over the hundreds of applications submitted each year. What I've tended to notice is that programs that have less applicants (either because of the type of program or the prestige) are more likely to not require the GRE's or be more lenient in the cut-off (that isn't to say that's the sole reason for it).

Edited November 28, 2015 by Oshawott

[Eigen]

It's a good point, but I agree that you're overgeneralizing. 

I've helped our department write grants to secure funding (fellowships) for incoming grad students the last couple of years, and GRE scores are one of the smallest sections. 

Usually, departments are allocated funding based on their past success with picking graduate students. The department argues the success of the students it's funded in the past, and where those students have gone as a predictor that whatever algorithms they're using are working. 

GRE's are a good measure that equalizes applicants across a wide variety of backgrounds, sure. Generally, I see them used as a small piece of a much larger picture. They can help to counterbalance a low GPA, or can help to make a really high GPA look like perhaps it's just an easy school. 

I don't know anyone on admissions committees (in my field) that over-uses them- they are pretty honest with what the GRE scores, and are just another means of assessment. 

[The_Old_Wise_One]

Thanks for the reply! My only point here is to make the argument that the general GRE is used to compare students across disciplines for funding purposes. My mistake was leaving out the type of funding that I was referring to––I mean fellowships when I refer to funding in the above passage. Teaching and research assistantships have nothing to do with the funding money when it comes to comparison across disciplines, but for graduate fellowships the GRE plays an enormous role (along with undergraduate GPA) in deciding who gets funding from the university. I have never really considered TA or RA positions "funding" opportunities, as you are being paid for work outside of your schoolwork as opposed to being paid purely for being a part of the organization. 

Also it is rather misinformed to assume that the general GRE is a device meant "just to serve as a filter", as the psychology subject GRE is a better predictor of graduate GPA than the general GRE is. Another important note is that I did not mean to imply that departments themselves are given funding based on the GRE scores of their students, but that students (who I refer to above as the department) themselves are given funding by the university based largely on their GRE scores. When I refer to competition between programs, a better wording would have been "competition between students across programs".The time line looks like this:

1) University has pool of money that can be used for funding through graduate fellowships.

2) Students apply for program, and faculty members nominate certain outstanding candidates for fellowship awards and such. 

3) A committee decides who to fund, across disciplines.

Most PhD programs will waive tuition, give you a small stipend, and also give you opportunities to make extra money through TA or RAships. What I am referring to throughout all this talk is not those opportunities, I am speaking only of fellowship/university funding. Looking back at my post I can see where I should have been more detailed, and I am glad you brought all this up! Other people were likely misinterpreting me too. 

In the end, the university sees a graduate student as a graduate student, regardless of what they are studying. Funding provides some extra money for outstanding PhD students who the university assumes will provide a stimulating intellectual environment within their field. This money is very important, as graduate students will have difficulty affording daily living costs (depending on many factors of course, but in general) if not for the extra money provided by sources like university fellowships. In order to decide who gets access to that money, the university looks at a variety of factors, but one of the most important is the general GRE. 

If willing to read, there was an article published in 2001 that takes a meta-analytic approach to observing the general and subject GRE's and how they correlate with other factors such as (1) faculty ratings of grad students, (2) graduate GPA, and (3) research productivity. Link is below. 

Link: http://internal.psychology.illinois.edu/~nkuncel/gre%20meta.pdf

Though that article was written some time ago, I am sure that ETS has remained consistent over the years. If anything, their tests have become much better at classifying percentile rankings (as is clear when looking at the effect switching over from the old to new scoring system was). As for your main argument, I have to agree that the GRE does serve as a filter, but after that process, it comes into play once again when universities are trying to decide who to fund within the different programs. 

[The_Old_Wise_One]

Eigen,

I can see that your experiences have given you insight into the process, but it now seems as if you are trying to generalize your experience to the majority of schools. As mentioned before, the subject tests actually predict graduate success/research success better than the general GRE. If only being used for admissions criteria, the subject test would be a better fit! The link above contains the quantitative information. That being said, by the time grants are being written it is likely that most (if not all) students being discussed have good to great general GRE scores, so it makes sense to me that it would not be a huge topic of discussion at that point. However, if the university has enough money to give one more fellowship out to one of a few students in different programs, the general test is perhaps the only metric that can be validly used in deciding who gets the cash  

I just dislike seeing people on the forums talking down on the importance of the general GRE. It is a large portion of your application, and it should be treated as such. I don't want people to read that "it's just a cut-off" and then be disappointed that they did not try to achieve a higher score. It comes into play multiple times throughout the application process, not just in the beginning. 

 

[Oshawott]

@The_Old_Wise_One

Also it is rather misinformed to assume that the general GRE is a device meant "just to serve as a filter", as the psychology subject GRE is a better predictor of graduate GPA than the general GRE is.

I'm not entirely sure why you're bringing up the subject-test GRE, which I agree has utility. But how does the statement that the psychology GRE is a better predictor of graduate GPA (something that largely doesn't matter in most programs) at all relate to addressing the misinformation that the general GRE being used as a filter? If anything, it sounds like you're agreeing that the General GRE shouldn't be used since its the wrong test to base assessments from.

I'd like to get some evidence, anecdotal or otherwise, as to your assertions about the general GRE. There are multiple schools that specifically state a minimum GRE requirement (e.g., UI-Bloomington) and professors I have talked to outright state that in most schools, the GRE is used as a cut-off. At no point in later decision-making processes are GRE's exclusively used to make an informed decision on who to admit. While the GRE may be used in the final decision process, it is used as part of a larger picture (e.g., research experience, reference letters, the interview, whether the other graduate students like you). The only time I can see it making a difference at that point is if both applicants were equally as good in every other dimension, which I doubt ever happens.

Now the subject test--I can see why its important, especially since it normalizes scores across schools for general PSY knowledge. The reason I didn't address it when I called the general GRE a cut-off, however:

20 hours ago, The_Old_Wise_One said:

[...]

Well, here is where the general GRE comes in!

The GRE is a baseline used to compare students across disciplines. It is simple as that. It is one of the only metrics that makes this possible. This is why some scholarships and outside funding options for masters programs require general GRE scores. This is why programs that do not fund their students are less likely to require general GRE scores. This is why the general GRE is an important factor when it comes to selecting doctoral candidates! Graduate programs have to compete with one another for a piece of the funding, thus we are all studying and banging our heads against keyboards trying to get a 320+ (and all the while wandering why this stupid test exists!). 

I hope that what I said above gives the community here a bit more insight into why programs require tests like the GRE. The test was created for the sake of making fair comparisons across disciplines, and regardless of how silly it may initially seem, I hope that people can realize how necessary it is. 

Please post comments below for good discussion! Also, post about any insights you may have that could help others understand the application process in more detail. Thanks for taking the time to read  

Considering that the above quote was referencing a combined 320+ score, comparisons across disciplines (which you can't do with a subject test), you can see why my post did not take into consideration the Psych subject test in my response--as I had assumed that wasn't the topic at hand.

And in terms of comparing students across disciplines--Psychology is probably one of those disciplines where verbal and quantitative reasoning is important. But what about English? Why should that department be disadvantaged because its students score low on quantitative reasoning when they'd never have to use it? Do we now compare the Verbal score of English graduates with the Quantitative score of Math graduates? Or should we compare the scores they'd naturally be worse on (i.e., Quant for English, Verbal for Math) even though those scores would be even less indicative of success in their respective program? The common language argument only works if it makes sense to compare all disciplines verbal/quant scores, but it doesn't.

I also have to ask this: Why do schools care so much about Quantitative and Verbal reasoning? In research programs, especially, shouldn't Analytical Writing be more heavily emphasized given what it purports to measure? Related to this, some professors I've talked to who outright stated that the Verbal/Quant GRE's don't factor into their decision making process have actually stated that they look at Analytical Writing. Yet references to GRE's in graduate programs that advertise what they look for almost exclusively talk about Verbal and Quant. It seems counter-intuitive that the one score that could actually be predictive is largely being ignored in favor of the other two....unless they are being used as cut-offs.

1 hour ago, The_Old_Wise_One said:

 

I just dislike seeing people on the forums talking down on the importance of the general GRE. It is a large portion of your application, and it should be treated as such. I don't want people to read that "it's just a cut-off" and then be disappointed that they did not try to achieve a higher score. It comes into play multiple times throughout the application process, not just in the beginning. 

 

 

I don't think anyone on this forum "talks down" the importance of getting high GRE scores in the way you are thiking. Almost everyone will likely suggest that someone getting a 150V/150Q score should retake the test.

There is, however, a difference between acknowledging its role in the decision-making process and criticizing the fact that it plays a role in the first place (which is what almost everyone here does). We all know the GRE's are important and that aiming for a 320+ score will almost always guarantee that you are in the green zone with regards to that specific criteria, but that doesn't mean that we aren't critical of how it is being used.

Going back to my anecdotal examples with professors I have spoken with--while they do talk about the lack of predictability of the (general) GRE scores, they will always tell a student to get a high score because they understand the role it plays. It doesn't change the fact that they can criticize the test for what it is.

Edited November 28, 2015 by Oshawott

[Piagetsky]

I have also heard of the GRE being used as a cutoff and nothing more.  What I'd really like to know is how schools that don't require the GRE at all manage their application process without it. 

[Dedi]

I also have to ask this: Why do schools care so much about Quantitative and Verbal reasoning? In research programs, especially, shouldn't Analytical Writing be more heavily emphasized given what it purports to measure? Related to this, some professors I've talked to who outright stated that the Verbal/Quant GRE's don't factor into their decision making process have actually stated that they look at Analytical Writing. Yet references to GRE's in graduate programs that advertise what they look for almost exclusively talk about Verbal and Quant. It seems counter-intuitive that the one score that could actually be predictive is largely being ignored in favor of the other two....unless they are being used as cut-offs.

I've heard the opposite: the analytical writing score is largely ignored because, despite what it purports to measure, it only shows you how much word vomit you can make up in 30 minutes. Seriously, they don't fact check this stuff. You don't have to cite anything--and the longer you write (even if it is rambling, like many people do, and not concise scientific writing) you're bound to get a better score. I've been told I'm a very good writer because I'm straight to the point. I got a 3.0 on the writing portion, most likely because I didn't "write enough." Also, in real life you have time to check your work and revise it (and some people, such as myself, have bad first drafts). My guess is that there is no correlation between analytical writing scores and graduate school success, much like Q and V. Supposedly I have large deficits in my writing, according to the GRE, but I seem to make it around just fine.

[FacelessMage]

17 minutes ago, Piagetsky said:

I have also heard of the GRE being used as a cutoff and nothing more.  What I'd really like to know is how schools that don't require the GRE at all manage their application process without it. 

My old graduate program didn't require the GRE (neither did the other major university in town, but that's because it was a bilingual university and it wouldn't be valid for the French applicants). It based its admissions criteria solely on GPA, research experience, and fit with faculty. You basically got accepted into the program if a specific faculty member pulled for you and you had a decent GPA. 

[Oshawott]

2 hours ago, Dedi said:

I've heard the opposite: the analytical writing score is largely ignored because, despite what it purports to measure, it only shows you how much word vomit you can make up in 30 minutes. Seriously, they don't fact check this stuff. You don't have to cite anything--and the longer you write (even if it is rambling, like many people do, and not concise scientific writing) you're bound to get a better score. I've been told I'm a very good writer because I'm straight to the point. I got a 3.0 on the writing portion, most likely because I didn't "write enough." Also, in real life you have time to check your work and revise it (and some people, such as myself, have bad first drafts). My guess is that there is no correlation between analytical writing scores and graduate school success, much like Q and V. Supposedly I have large deficits in my writing, according to the GRE, but I seem to make it around just fine.

The issue pool is definitely something that would require fact-checking, but I'd think that the argument portion relies more on your ability to assess claims made (which is something that we do when we read research papers) and doesn't necessarily need prior research (in fact, one of the prompts asks you to give "types of evidence" that would be required to support the claim, not necessarily provide statement of fact).

With that said, do I think that my ability to reason why a restaurant shouldn't have changed from margarine to butter accurately reflects my ability to use my knowledge of Psychological research design to analyze a paper in my field? Definitely not, though I can see why the professor I talked to mentioned a preference towards it--not that having a preference towards Analytical Writing over V and Q necessarily means they even endorse it. Now that I think about it, why would one care about a numerical score on writing when schools can ask for SOP's and writing samples? And I've heard that the Analytical Writing Portion is now scored electronically, which means that it definitely is evaluated based on heuristics (i.e., long paper = good). Not sure if that's true but the ETS provides a ScoreItNow service...which automatically evaluates your writing.

It's interesting to get other perspectives on this because its really just boiling down to the fact that while everyone agrees that the GRE's are important for admissions (even if its to the extent that you need to meet a cut-off), how people use that information isn't even standardized raising even more questions about the validity of the test.

Edited November 28, 2015 by Oshawott

[Eigen]

4 hours ago, The_Old_Wise_One said:

Eigen,

I can see that your experiences have given you insight into the process, but it now seems as if you are trying to generalize your experience to the majority of schools. As mentioned before, the subject tests actually predict graduate success/research success better than the general GRE. If only being used for admissions criteria, the subject test would be a better fit! The link above contains the quantitative information. That being said, by the time grants are being written it is likely that most (if not all) students being discussed have good to great general GRE scores, so it makes sense to me that it would not be a huge topic of discussion at that point. However, if the university has enough money to give one more fellowship out to one of a few students in different programs, the general test is perhaps the only metric that can be validly used in deciding who gets the cash  

I just dislike seeing people on the forums talking down on the importance of the general GRE. It is a large portion of your application, and it should be treated as such. I don't want people to read that "it's just a cut-off" and then be disappointed that they did not try to achieve a higher score. It comes into play multiple times throughout the application process, not just in the beginning. 

 

Would you mind sharing your experiences in securing competitive funding for graduate fellowships for a department/program that differ from mine? I'm sure not everyone does it the same way, but I'd be shocked if metrics change enormously from state to state. 

The NSF grad fellowship (which is competitive across disciplines) doesn't even allow the submission of GRE scores, and hasn't for a while.

Your ideas are interesting, but I not seeing evidence to back them up, anecdotal or otherwise. For instance, you state that GRE scores are a "large portion of your application", but you are providing no support for why you think that is the case, in direct opposition to many, many other users on this board who have provided anecdotal evidence for why the GRE is primarily used as a cutoff. 

I can tell you not everyone who gets these University wide fellowships have high GRE scores. That's not considered as important as their evidence of research potential, as evidenced by research experience and letters of recommendation. 

GRE scores are used past application in fellowship decisions, sure, but I've never seen them used as a major component of those decisions either- they're usually, again, a cutoff (albeit a higher cutoff). 

And usually, from my experience, anyone up for a university wide fellowship would be funded by a TA or an RA regardless, the University is just putting them up for additional funding. That means that the GRE score isn't playing an additional role in applications (with respect to admission), but can have some minor role in securing additional funding. 

That doesn't, however, mean that the GRE comes into play "multiple times during the application process".

To take this off on a tangent, you make a statement in your original post that GPA cannot be compared across disciplines, and this seems curious to me. Why would you not say that a 3.8 in psychology is equivalent to a 3.8 in math? The whole point of the GPA system is that it does compare you to the other students in a major in such a way as to be relatively normalizable across disciplines (and even schools, to some degree). 

To make an argument that it's not, you'd have to be able to argue that grade inflation is more prevalent in some disciplines than other, or that a given discipline is objectively more or less difficult than another, which would be a really controversial point to make. 

[TenaciousBushLeaper]

6 hours ago, The_Old_Wise_One said:

Eigen,

I can see that your experiences have given you insight into the process, but it now seems as if you are trying to generalize your experience to the majority of schools. As mentioned before, the subject tests actually predict graduate success/research success better than the general GRE. If only being used for admissions criteria, the subject test would be a better fit! The link above contains the quantitative information. That being said, by the time grants are being written it is likely that most (if not all) students being discussed have good to great general GRE scores, so it makes sense to me that it would not be a huge topic of discussion at that point. However, if the university has enough money to give one more fellowship out to one of a few students in different programs, the general test is perhaps the only metric that can be validly used in deciding who gets the cash  

I just dislike seeing people on the forums talking down on the importance of the general GRE. It is a large portion of your application, and it should be treated as such. I don't want people to read that "it's just a cut-off" and then be disappointed that they did not try to achieve a higher score. It comes into play multiple times throughout the application process, not just in the beginning. 

 

You're making yourself look silly, stop it.