| Quantitative | 800 | 94 percent scored lower (maybe they need to increase the difficulty) |
| Verbal | 800 | 99 percent scored lower |
| Analytical Essay | 4 | 37 percent scored lower (ouch) |
The essay is on a 0-6 scale. 0 is the score for people who can’t read or write english or generate original thoughts. 4 is the “there weren’t any glaring flaws with your essay, but it wasn’t particularly good either” I was hopeful I might get a 5. But, I didn’t expect to be in the 37th percentile. *shrug* [ETA: However, if you consider that there are only 3 groups remaining (4-6), even assuming the distribution between the three groups is even, group 4 would include the median GRE taker. Assuming it is lumpy in the middle, shallow on the edges, semi-bell-curvish distribution group 4 is even more expansive. I shrug and say, “whatevs.”]
Then again, how often do you have to write an essay with a hard 30 minute response deadline, and without any prior exposure to the subject? Same could be said of multiple choice tests, of course. Why do schools even care about these things?
I think my percentiles for analytic and quantitative are the reverse of yours. They don’t need to make quantitative harder as far as I’m concerned!
I can see grad schools wanting to see something standardized, since they don’t know how different programs genuinely compare. But I think they’d do a lot better with program-specific standards like some disciplines have, rather than something as generalized as the GRE.
I’m guessing the score distribution for “quantitative” is bimodal. Having one “general math” test that both technical and non-technical people take is a little strange.
By the way, they recently had to re-center the math subject GRE because 17% of test takers were getting the highest possible score.
The subject tests just make more sense! I’m a linguistics student. What little math I need is stats. I totally bombed the math and nobody cared. They only cared about the verbal and analytical. So why should I have to take the math if my program doesn’t care? Ditto for someone whose program doesn’t care about verbal. Etc.
The math part tests so little that it’s useless to everyone: either your field doesn’t need you to be able to learn five tricks well enough to apply them very quickly, or it does but your grades would already show you to be a hopeless case if you couldn’t. But the verbal has been shown to be a predictor of failure if you do poorly on it. Every field needs you to be able to communicate, and interested in reading more than absolutely necessary.
As a fellow linguist, I’d have to say that, while the specifics of math aren’t as necessary, logical, quantitative reasoning is key. Where I teach, my linguistics class is actually considered a “quantitative reasoning” class, that replaces stats or calculus for humanities majors. Of course, I’m sure the score on analytical was more important when I took the GRE’s, since that was straight-up logic.
It’s definitely an quantitative discipline, but the math tested on the GRE isn’t generally relevant. The old analytical section is much more relevant. The new version is less relevant to linguistics really. That change is frustrating. It doesn’t address the same skills.
I’m applying to PhD programs in linguistics this semester, and both of my advisers told me that the programs will look at the math more (though, to be sure, they want high scores across the board). I think the GRE is entirely ridiculous myself. I have a mediocre score on the Verbal, but good scores on everything else and am otherwise a strong candidate, but I still fear that this one little less-than-perfect Verbal score is going to put me out of contention for the schools I want to get into.
Where are you applying, if you don’t mind me asking?
I think I had a 670V when I took the GRE, which to me seemed too low, but it turned out to be ok. The math is much more important for formal, theoretical programs, so far as I know.
Yeah, I was told I needed 700s across the board. I ended up with a 580V, 760Q. Whatever. It’s higher than when I took it last month.
I’m hoping to get into Stanford, but I’m also applying to UT Austin, U-Pitt, and UCSB, all for sociolinguistics. Should none of that pan out, my back-up is to stay at San Francisco State (where I’m currently getting my BA) after graduation to get a Master’s, and then apply into programs again in a year or two.
What got me thinking was that alot of the verbal section was about reading an intricate passage for comprehension of particular points. Something important in all grad school programs. Granted, the art history one didn’t do much for me… =)
Some of it, at least. But they should be able to tell if you have those skills through other means, too.
The whole thing is so flawed, though. Standardized tests in general are. I don’t know why they still require them. I hear that undergrad programs aren’t all requiring SATs or ACTs anymore.
I think the argument for standardized testing for undergrad admissions is that there are lots of strange things that can happen in high schools to push someone’s grades one way or the other. For super-selective schools they can help remove the bias of “Well, you need a 4.0 from somewhere, but what it means that you got a 4.0 and this kid from another state got a 3.9 is kinda murky” Not that the standardized tests aren’t murky in their own ways…
I think a lot of the schools that are dropping the requirement are state universities, whose missions are explicitly to provide educational opportunities for children of the state’s taxpayers. For that goal, it makes sense to build the university bigger, be generous with the admissions and… …I guess the best answer is to move people down to the next tier of state university when they fail one, and let them transfer up when they do well.
The worst of both worlds (between standardized tests, and just going by high school grades) for undergraduate admissions would probably be the following:
Every county in the US cooks up a standardized test. The tests are given in each neighborhood. The test administrator can choose to weight which questions are important the day of the exam. The only requirements are that the scores for the county-wide exams range from 0 to 1600. Then the scores are all sent to colleges, which use them to decide which students are “appropriate for such-and-such university”
The more I think about it, the more I like the fact that we have community colleges and junior colleges.
They’re cheap.
They’re everywhere.
Adults can take classes at them for fun.
Night classes are offered.
You can start (or re-start) whenever you’re ready.
You can transfer to other universities after finishing.
Also : If you’re a foreign-born postdoc, your girlfriend can get a student visa for going to a community college near where you have your postdoc. I’m probably the only person who thinks this is an advantage. (Personally I think anyone with a valid visa should be allowed to sponsor one significant other, regardless of spousal status. If they break up, the s.o. loses the visa)
The percentage of each section that has certain features changes after the computer decides which “bracket” you belong in. Depending on where the computer places you and how well you do on previous questions determines if you see more analogies or antonyms or reading comp, etc.
I hate that stupid computer.
How much did you prepare for the GRE? I’m taking mine NOvember 4th, and my Math GRE on November 8th. I’m trying to focus on the Math GRE because its my understanding that the general GRE exam is a pushover…
The quant section of the general GRE is about as difficult and at the same level as the SAT math section. If you’re taking the Math subject GRE, you’re fine as long as you can remember back that far in your math education.
The verbal section generally produces much lower scores, as most people don’t take grammar courses in college and the vocabulary can be pretty arcane.
prep for the GRE, not alot, focused on the essays. I should have done more, but I wanted the scores quickly.
All I know is that I’m glad I took the GRE before they took out the old “analytical” section and put in the new “analytical essay”
P.S.
http://www.nytimes.com/2005/05/04/education/04education.html
that is comforting.
I heard once that they were planning to bifurcate the GRE at some point, with a version for math/science-intensive fields and a version for humanities/social sciences, or something like that. The main point being, of course, so that the distribution of quantitative scores might be less absurd.
Well, the GRE is so useful to begin with….
I thought the CS subject test was semi-useful.
The rest? Possibly relevant if you went to undergrad at some university that nobody has heard of.
I used to write essays with no prep on subjects I was not familiar with all the time. Of course I usually had a grounding in backing material on which to draw.
The problem was that the better I got at spitting out such essays in 30 minutes or less, the less thinking I felt was going into them. As this trajectory began to accelerate, I dropped out of school.
The essay is graded on a half point scale.
Well, I looked up the distribution. 54% score under 4.5, so I’m median, at least. Average is 4.1, so I’m slightly below that. Mode is 4.5.
http://www.ets.org/Media/Tests/GRE/pdf/994994.pdf