GRE - does raw score or percentile matter more?

[VMcJ]

One professor told me people of his department puts a lot of emphasis on GRE scores (quantitative, especially). He even said it was probably the second most important part of the application, with the first being the LoRs and the third the SoP. He said this was not right for him, as the SoP should have a bigger value. They look into the quantitative section as a proxy for predicting students' ability to perform well in empirical research.

If GRE verbal questions were all about interpreting texts, I would be totally fine with them. But to ask students to know ridiculous, never-used vocabulary is simply absurd. I would not change the quantitative part of the test. And I would make the criteria by which they evaluate the writing section more clear.

Full disclaimer: I am an international student (I probably made some grammar mistakes that gave it up, LOL) and my scores were 162V-168Q-4.5W. So I may be biased toward thinking GRE has a problem in the verbal section but not in the quantitative.

[wb3060]

Thanks everyone for all the input. I too have been under the impression that GRE mostly matters as a screening factor (and this is the advice I have received), particularly if you have a good writing sample, research experience, methodological training, etc. that provides more information on these areas. I'm sure though that it varies by program. I imagine GRE quant scores are given more consideration in very quant-heavy programs. 

[ultraultra]

12 hours ago, shane94 said:

Being deterred from applying to Ph.D. programs because your friends are told that GRE is important? Well, if they have the time to perform excellently in class to get fantastic letters, to do research diligently and write a publishable WS, and to have tons of research experiences, why can't they spend a month to periodically study for GRE and get a "threshold" score (around 330)? 

It occurs to me that the unhealthy part of grad school applications is not school's over-emphasis on GRE (and in fact schools don't emphasize on GRE, at least PSC depts), but people not realizing that GRE is a way to simply get your file reviewed, and nothing more. Study for a month and get a 330, and you are set for standardized test portion.

I don't know your situation, but... taking a month off work and school to study for an exam is not a feasible for option for a lot of people, particularly people who have to work to support themselves and/or who need to maintain certain grades or activities for financial aid. Not everyone has access to external sources of funding such as grants and loans (and parents), and many people (including myself) hold research assistant jobs just as much to support themselves financially as to build their portfolio. Not to mention the expensive cost of prep courses/materials, and the massive expense of the exam itself (which, as an international student, you are not able to waive), which is of course followed by the expense of the applications themselves (again, not waive-able). Not everyone is in a financial position to drop everything and study, nor is everyone in a financial position to gamble on applying with sub-perfect scores.  

Everyone knows that academia is generally inaccessible to most groups. My problem with the GRE is that it makes top programs (and, to some degree, eventual success) that much more inaccessible, all the while measuring things of only tangential importance to graduate success.

I don't disagree with your general point, that beating the threshold is something people just need to do. But I personally think the GRE causes more problems than it solves.

Edited January 30, 2016 by ultraultra
typo

[s1994]

16 minutes ago, ultraultra said:

I don't know your situation, but... taking a month off work and school to study for an exam is not a feasible for option for a lot of people, particularly people who have to work to support themselves and/or who need to maintain certain grades or activities for financial aid. Not everyone has access to external sources of funding such as grants and loans (and parents), and many people (including myself) hold research assistant jobs just as much to support themselves financially as to build their portfolio. Not to mention the expensive cost of prep courses/materials, and the massive expense of the exam itself (which, as an international student, you are not able to waive), which is of course followed by the expense of the applications themselves (again, not waive-able). Not everyone is in a financial position to drop everything and study, nor is everyone in a financial position to gamble on applying with sub-perfect scores.  

Everyone knows that academia is generally inaccessible to most groups. My problem with the GRE is that it makes top programs (and, to some degree, eventually success) that much more inaccessible, all the while measuring things of only tangential importance to graduate success.

I don't disagree with your general point, that beating the threshold is something people just need to do. But I personally think the GRE causes more problems than it solves.

You make GRE sound so formidable... 3 hours/day for 1 month should be sufficient for one to hit that 325/330 threshold (assuming one reads intensely and critically throughout her four years in college), without any prep courses. Hitting 330 is not that hard; it really isn't... What I get from this thread is that people are "devilizing" or under-emphasizing GRE a little bit, which can be misleading and problematic to prospective students when they read this thread next year. 

However, I do respect your opinions. Congrats on your NU offer and good luck with your other applications! 

Edited January 30, 2016 by shane94

[TakeruK]

1 hour ago, shane94 said:

You make GRE sound so formidable... 3 hours/day for 1 month should be sufficient for one to hit that 325/330 threshold (assuming one reads intensely and critically throughout her four years in college), without any prep courses. Hitting 330 is not that hard; it really isn't... What I get from this thread is that people are "devilizing" or under-emphasizing GRE a little bit, which can be misleading and problematic to prospective students when they read this thread next year. 

However, I do respect your opinions. Congrats on your NU offer and good luck with your other applications! 

I disagree with your definition of what is "easy". 90 hours of study is not that affordable to many many students. This is 90 hours on top of the regular courseload in a bachelor's degree---why should this be a necessary step in order to go to graduate school? Studies have shown (see this and references therein) that there are no correlations between GRE performance and other measures of graduate school success (whether it's grades, post-PhD placement, papers, etc.). Instead, there is strong correlation with gender and race. This is why I think we should "demonize" the GRE because by doing so and by putting pressure on our departments to do so, it is the way we can remove the GRE from the admissions process. 

I do agree with you that we do not want to be misleading though. Under-emphasizing the GRE could do more harm than good right now because you're right, many departments still use GRE scores as cutoffs despite ETS recommendations on their guide to interpreting their scores. So, we should make it clear the importance of the GRE. However, this doesn't mean we should refrain from demonizing it. And I definitely think we should not perpetuate the false statement that if you can't do well on the GRE, you can't do well in grad school. (or that there is something wrong with you if you can't achieve 330)

[s1994]

People are all so right. 

[PoliticalOrder]

14 hours ago, shane94 said:

You make GRE sound so formidable... 3 hours/day for 1 month should be sufficient for one to hit that 325/330 threshold (assuming one reads intensely and critically throughout her four years in college), without any prep courses. Hitting 330 is not that hard; it really isn't... What I get from this thread is that people are "devilizing" or under-emphasizing GRE a little bit, which can be misleading and problematic to prospective students when they read this thread next year. 

However, I do respect your opinions. Congrats on your NU offer and good luck with your other applications! 

 

In the past I studied significantly more than that and had a lot of trouble reaching 320, let alone 330. I mean, 330 is approximately 165/165, that means you are hitting the top 90 percentile in both categories (which means that for every 100 takers, you did as well or better than 90 of them in both math and verbal)...to say 'it's not that hard' is completely false. 

Honestly, on this board I see more of the other phenomena...people way overstate the importance of the GRE for Ph.D. applications. The simple fact is this:

YES, committees use GRE as an important cut-off measure. YES, you could potentially get axed against similar candidates because of a relatively poor GRE score. HOWEVER, the GRE has extremely diminishing returns for your acceptance...once you reach the later stages of the review process, the more important factors that give the reviewers more information such as writing samples, LOR, SOP, and research experience matter significantly more for if you get in or not.

You do not need stellar GRE scores to get into top programs (even top 5 programs). Don't believe me? Go through the results section...people get into places like Yale and Princeton with scores throughout the 150-160 range in math. And this isn't some random thing either, there are dozens people who get in with these scores (even as low as 148-153) every cycle.

------

Bottom line is get as good of a score as possible. Preferably 160/160+, but it you don't that doesn't mean you are doomed, especially if you have an otherwise very strong application. 

[s1994]

3 hours ago, PoliticalOrder said:

 

In the past I studied significantly more than that and had a lot of trouble reaching 320, let alone 330. I mean, 330 is approximately 165/165, that means you are hitting the top 90 percentile in both categories (which means that for every 100 takers, you did as well or better than 90 of them in both math and verbal)...to say 'it's not that hard' is completely false. 

Honestly, on this board I see more of the other phenomena...people way overstate the importance of the GRE for Ph.D. applications. The simple fact is this:

YES, committees use GRE as an important cut-off measure. YES, you could potentially get axed against similar candidates because of a relatively poor GRE score. HOWEVER, the GRE has extremely diminishing returns for your acceptance...once you reach the later stages of the review process, the more important factors that give the reviewers more information such as writing samples, LOR, SOP, and research experience matter significantly more for if you get in or not.

You do not need stellar GRE scores to get into top programs (even top 5 programs). Don't believe me? Go through the results section...people get into places like Yale and Princeton with scores throughout the 150-160 range in math. And this isn't some random thing either, there are dozens people who get in with these scores (even as low as 148-153) every cycle.

------

Bottom line is get as good of a score as possible. Preferably 160/160+, but it you don't that doesn't mean you are doomed, especially if you have an otherwise very strong application. 

170/161 (331) 2/19/15

167/170 (337) 2/19/15

169/154 (323) 2/19/15

162/159 (321) 2/19/15

167/168 (335) 2/19/15

162/160 (322) 2/19/15

160/169 (329) 2/18/15

This is the stats for admitted Yale students on Grad Cafe.

"people get into places like Yale and Princeton with scores throughout the 150-160 range in math"

Who is giving out the misleading information? Who is demonizing GRE? I think it is pretty clear. 

 

If people spend 100+ hours AND still cannot get 160+ in quant (=HS algebra, plus a few tricky though definitely doable questions), I do not know (1) how would that person be admitted into a top 10, and (2) how can that person survive the quant sequence in the first two years. 

Note THE world "and" above. One can definitely be extremely awesome in everything else although having a 15x quant score. However, if someone has spent 100+ hours/tons of money on courses AND still could not get 160+, that probably indicates that he or she should not consider applying to top 10s in the first place. 

Edited January 30, 2016 by shane94

[PoliticalOrder]

12 minutes ago, shane94 said:

170/161 (331) 2/19/15

167/170 (337) 2/19/15

169/154 (323) 2/19/15

162/159 (321) 2/19/15

167/168 (335) 2/19/15

162/160 (322) 2/19/15

160/169 (329) 2/18/15

This is the stats for admitted Yale students on Grad Cafe.

"people get into places like Yale and Princeton with scores throughout the 150-160 range in math"

Who is giving out the misleading information? Who is demonizing GRE? I think it is pretty clear. 

 

If people spend 100+ hours AND still cannot get 160+ in quant (=HS algebra, plus a few tricky though definitely doable questions), I do not know (1) how would that person be admitted into a top 10, and (2) how can that person survive the quant sequence in the first two years. 

Note THE world "and" above. One can definitely be extremely awesome in everything else although having a 15x quant score. However, if someone has spent 100+ hours/tons of money on courses AND still could not get 160+, that probably indicates that he or she should not consider applying to top 10s in the first place. 

With someone so arrogant about who and who shouldn't apply to top 10 programs you seem to lack some pretty significant reasoning abilities.

You basically proved my point...out of the people who got admitted to Yale last cycle 4/7 had quant scores in the 150-160 range (I counted the 161 because it's virtually the same thing)...what exactly are you arguing against here? Because that data definitely supports the statement "people get into places like Yale and Princeton with scores throughout the 150-160 range in math."

Secondly, you are conflating things here. 1) the GRE quant score does not represent one's ability to succeed in a quantitative methods sequence, nor 2) ability to succeed at a top 10 program. The evidence (or lack thereof) does not support either of these arguments. Try again.

 

 

[s1994]

1. PoliticalOrder, Maybe you should avoid personal attacks. 

2. Also, "the GRE quant score does not represent one's ability to succeed in a quantitative methods sequence".

Really? People who have spent 40 or 50 hours to study but still couldn't do HS algebra can master game theory, multilevel models, and Bayesian data analysis. I am convinced.

Edited January 30, 2016 by shane94

[VMcJ]

Well, three out of seven Yale admittances on Grad Cafe for 2015 were indeed in the 150-160 range. Median was 161 and mean was 163, but still. Verbal scores appear more important and I find that highly problematic, as I said before, especially if departments want to receive more applications from international students.

 

[PoliticalOrder]

9 minutes ago, shane94 said:

1. PoliticalOrder, Maybe you should avoid personal attacks. 

2. Also, "the GRE quant score does not represent one's ability to succeed in a quantitative methods sequence".

Really? People who have spent 40 or 50 hours to study but still couldn't do HS algebra can master game theory, multilevel models, and Bayesian data analysis. I am convinced.

What, first you claim I am giving out misleading information then provide data that actually supports my position in an arrogant fashion and you expect me not to call you out? 

Secondly, a test that is designed to trick you purposely, is timed heavily, and relies on math that doesn't necessarily have any applicability to probability or statistics is a good indicator of your ability to succeed in a quant methods sequence? Anyone with high school level math and a good work ethic can succeed in political science graduate quant sequences. 

Edited January 30, 2016 by PoliticalOrder

[s1994]

 

2 minutes ago, PoliticalOrder said:

Anyone with high school level math and a good work ethic can succeed in political science graduate quant sequences. 

I will keep this quote, and we will talk about it in two years at APSA (if we are both surviving the program at that time). We will both laugh at this quote. Good luck on your applications.

[PoliticalOrder]

5 minutes ago, shane94 said:

 

I will keep this quote, and we will talk about it in two years at APSA (if we are both surviving the program at that time). We will both laugh at this quote. Good luck on your applications.

You do realize you have been debating with someone who is already been through a quant sequence right?

[s1994]

6 minutes ago, PoliticalOrder said:

You do realize you have been debating with someone who is already been through a quant sequence right?

No. None of the people who are in a top 10 PSC program has told me that "with HS algebra, one can succeed in the quant sequence at a top 10." They all told me that it's hard, and you have to be smart and to work really hard to know what you are actually doing in these classes. And they all told me that lots of people drop out of the programs.

Moreover, I also do not see any grad student *in this thread* telling me that quant sequences at top 10s are so easy that with HS algebra and work ethic, everyone can succeed. I never do. 

Edited January 30, 2016 by shane94

[GradSchoolTruther]

The whole premise of the thread is asinine. It depends on who is interpreting scores. There is not one poster on here who has a definitive answer for the field.

[VMcJ]

Hey, people. No need to get angry at ourselves.

I do think it is possible people that are unable to get a 160+ score in quantitative even with training to succeed, especially if there are other indicators of mathematical ability whatsoever (e.g. a quantitative WS). But it is more likely that those people might fail quant sequences than people with the highest scores.

[ultraultra]

23 hours ago, shane94 said:

No. None of the people who are in a top 10 PSC program has told me that "with HS algebra, one can succeed in the quant sequence at a top 10." They all told me that it's hard, and you have to be smart and to work really hard to know what you are actually doing in these classes. And they all told me that lots of people drop out of the programs.

Moreover, I also do not see any grad student *in this thread* telling me that quant sequences at top 10s are so easy that with HS algebra and work ethic, everyone can succeed. I never do. 

PoliticalOrder was referring to the fact that they are currently attending a top program and have already completed a graduate quant sequence. For the record, I am an MA student in the middle of a PhD quant sequence. So far I'm acing it, and doing better than a lot of the more advanced students. I also have a heavily quantitative RAship which I excel at. This is all despite the fact that I got sub-160 on the GRE Q section. 

---

The sticking points for me of all you have said are as follows:

1) For reasons outlined above, I do not think that taking 90 hours to study for the GRE is possible for everyone, and in fact I think the GRE presents a massive undue accessibility barrier for many people (especially on the basis of class). 

2) I do not think that achieving 160+ on the GRE is possible for everyone. Not only because not everyone can study for 90 hours, and not because people aren't capable of understanding the concepts, but because the GRE isn't about conceputal understanding, its about memory recall under time pressure, which is not a skill everyone has nor one that you need to succeed in political science. Consider this: Kaplan and other top GRE prep companies do not focus on giving students a deep conceptual understanding of the math. They help students understand the math on a shallow level, but the majority of the courses/materials are about ways to shave seconds off your time through tricks like doubling and halving to multiply AND tricks for memorizing formulas (especially geometric). The companies know that success on the GRE is not about mathematical aptitude - it's about taking a timed test efficiently. That's what the GRE tests. Not to mention all the studies that show that all standardized tests test is the ability to complete standardized tests AND that things like knowing the interface ahead of time can increase your scores. Has nothing to do with innate intelligence or mathematical aptitude. 

3) Because of the skills it requires (as outlined above) I do not think that success on the GRE is any indicator of one's ability to succeed with quantitative methods. To succeed at quantitative methods you need a conceptual and instrumental understanding of certain branches of mathematics, which you then apply to the study of political phenomena. Most of the topics on the GRE are irrelevant to statistics, and (as above) acing the GRE is really about memory and timing, rather than conceptual understanding or mathematical aptitude. I really fail to see why you think a GRE Q score reflects one's ability to do quantitative political science. If I'm working on a quantitative project, I don't need to memorize anything, nor do I need to complete questions in 2-2.5 minutes a piece... I can take my time, use online resources to ensure I fully understand the model and its application, and receive/respond to feedback from mentors and peers. That's how social science is done... not in a room with 15 other test takers, staring in silence at a screen while a timer ticks away, and where I'm not allowed to take off my sweater without putting my hand up to request an attendant.

4) I understand that you think GRE scores are an indicator of work ethic (and by extension likelihood of success) but for all the reasons above, as well as (and especially) my comment earlier about accessibility, I think that's fallacious and misplaced.

Again, this is all a separate conversation from the practical reality which is that, yes the GRE is a requirement, and yes people should try to maximize their score to avoid being under mandatory cut-offs at programs they hope to attend. But overall I agree with @TakeruK (and @PoliticalOrder) about the ineptitude of the GRE as an indicator of scholarly potential, and would really encourage you to read the link they shared about how GRE scores have been shown within the Astronomy field to correlate with race and gender, but have little to no bearing on actual academic success.

Edited January 31, 2016 by ultraultra

[Determinedandnervous]

I think this conversation has been a little derailed. I agree that graduate schools should place less emphasis on the GRE, as it is not representative of the overall ability of graduate students to succeed in the program. However, we don't live in such a world where graduate programs are going to agree, and we shouldn't act according to how we wish the process operated, but rather how it actually does. As long as admissions committees are inundated with applications, the GRE will be used as a screening tool. Different schools will have different cutoffs, and we should adjust our expectations accordingly. Maybe when some of us are on these committees themselves, we can move the charge to de-emphasize GRE scores, but that's about all we can do.

[TakeruK]

4 minutes ago, Determinedandnervous said:

I think this conversation has been a little derailed. I agree that graduate schools should place less emphasis on the GRE, as it is not representative of the overall ability of graduate students to succeed in the program. However, we don't live in such a world where graduate programs are going to agree, and we shouldn't act according to how we wish the process operated, but rather how it actually does. As long as admissions committees are inundated with applications, the GRE will be used as a screening tool. Different schools will have different cutoffs, and we should adjust our expectations accordingly. Maybe when some of us are on these committees themselves, we can move the charge to de-emphasize GRE scores, but that's about all we can do.

I agree with most of this. In the application stage, we have to act based on how schools actually use GRE scores, not how we want them to be used. 

But that is not the same as writing things like "330 is not hard to achieve" etc. because it also gives applicants the wrong impression that if they don't have a 330, then they might as well not apply.

Finally, I don't think we need to wait until we are actually on these committees to move to de-emphasize GRE scores. As I shared in my link above, there is already a current movement to do so in Astronomy. The link was a letter from the American Astronomical Society (AAS) President and the link shows a draft letter---in January 2016, the AAS Council met and approved this resolution and they will be preparing materials for people to present at their own departments to convince them of this change.

So, yes, at the stage where change actually happens, it has to come from someone with a lot of power and respect. These are all tenured faculty members on the AAS Council. However, this letter and actual commitment to change is the result of several years of work by a lot of people, including junior scientists like grad students and postdocs. There has been a quieter campaign in the last year where some researchers (the Nature article from 2014 cited in the letter) have been going around to various schools and presenting the data showing the lack of correlation. These researchers got talk invites because people at these schools (students, postdocs, and faculty members) wanted to hear from the researchers and asked their department to invite them. And this is necessary because a change like this cannot just come as an "order" from the AAS (the AAS doesn't really have any legal power). We see this letter today because it's a result of a ton of influential universities already buying into the concept. And these universities bought into the concept thanks to work by a lot of people at all stages of their career. 

So, yeah, there's not much you can do while you are still applying, but you also don't have to wait until you are on an admissions committee to help start changing things! 

[Bubandis]

And ultraultra wins the thread! Seriously, though, all of those points were terrific, and effectively inarguable based on my experience.  The GRE, in its current form, ought to be tossed.

[ultraultra]

2 hours ago, Determinedandnervous said:

I think this conversation has been a little derailed. I agree that graduate schools should place less emphasis on the GRE, as it is not representative of the overall ability of graduate students to succeed in the program. However, we don't live in such a world where graduate programs are going to agree, and we shouldn't act according to how we wish the process operated, but rather how it actually does. As long as admissions committees are inundated with applications, the GRE will be used as a screening tool. Different schools will have different cutoffs, and we should adjust our expectations accordingly. Maybe when some of us are on these committees themselves, we can move the charge to de-emphasize GRE scores, but that's about all we can do.

I know you meant well but I find this response fairly frustrating, and on the basis of it I'm not sure that you've actually read the thread.  No one has said that applicants should eschew studying for the GRE. No one has recommended students "act according to how we wish the process operated." Everyone (including me) has emphasized that applicants should still try to get the highest score possible (whatever that may be given an individuals amount of free time and test-taking ability). Everyone has also agreed that ad comms take the GRE seriously, though there are disagreements about the cutoffs and the centrality of the GRE (to which the only answer is really, it varies). The main subject that people are debating is whether or not the GRE has merit as an indicator of academic potential, and thus whether or not it is right that committees take it so seriously. Of course this thread has no bearing on what committees will actually do... even so, why shouldn't we be allowed to have a discussion about this? If you don't think its fruitful, you're free to bow out, but I'm not sure why you're policing peoples' abilities to discuss a topic that is 1) interesting and 2) of great personal importance to many of us. Since when are we not allowed to discuss whether certain material realities are just or unjust?

Edited January 31, 2016 by ultraultra

[s1994]

People are here to know how to cope with the system so that they can get admitted into a top program, not to know why and how AdCom is doing something wrong.

GRE is awful, biased, and useless? Good to know. But you know what? No one cares. 

Good to hear that people are all acing your quant classes at your top notch programs currently. Congrats to you all. People like King and Imai will all be impressed by your quantitative excellence and hey, H and P will both take you with your 15x quant score. 

Edited January 31, 2016 by shane94

[s1994]

18 minutes ago, ultraultra said:

The main subject that people are debating is whether or not the GRE has merit as an indicator of academic potential, and thus whether or not it is right that committees take it so seriously. 

I see no one arguing that AdCom should take GRE seriously. I have been constantly saying GRE is good as a screen to weed out people so that the rest of the (stronger) applicants can get more attention with faculty's limited time. It's a win-win situation.

Is GRE so flawed that a top 10 department will miss tons of excellent students because of a cutoff at 320? Probably they will miss a few, but the top 10s will be doing just fine as always without these 160+155 students. 

Edited January 31, 2016 by shane94

[ultraultra]

Just now, shane94 said:

GRE is good as a screen to weed out people so that the rest of the (stronger) applicants can get more attention with faculty's limited time

In other words, you are saying the ad comms should take the GRE seriously...