Page 1 of 1

Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 7:16 am
by omgmath
So, without disclosing the test content, what was the exam like for everybody?

I felt this was more difficult than the practice tests, for sure. It certainly felt longer.

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 7:27 am
by muneerusman
I felt the same. i.e. the questions were, on average, more difficult than those in the sample test. If everybody felt the same, that means the scores will be adjusted accordingly

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 10:15 am
by rahulshr
It wasn't very tough - marginally harder than practice tests though. A lot of calculus questions were familiar - in the sense very similar to what I had practiced.

However I screwed up for totally different reason - messed up sleep cycle over years - and had slept only 6-8 hours over last 3 days - and last night only 1-1.5 hours.

On the test... I was targeting to do 55 or so total, and I was well placed after first hour (even with heavy head, breezed through first 30 questions within 60 minutes, skipping 3-4 max).. But then fatigue settled in, and I could not do even 15 questions in remaining time (100+ minutes!!)...it was weird state of mind, I would start with formulating the problem well, but would redo 2-3 times ..getting different answers, or eventually skipping......that too some calculus stuff, which I was confident about otherwise (esp. during my normal waking hours)

With that, I do not think I will get good enough score to use in apps - most of which do not require subj-maths score - my primary reason to take subj-maths was to cover up for the long gap since college.
Wish it was conducted more frequently, with a time slot in afternoon :(.

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 1:28 pm
by arsenalmath
I thought it was easier than the October one, but there were certainly quite a few difficult problems

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 1:45 pm
by mathgirl28
I definitely felt the test was manageable. Can anybody venture a guess as to what the 80th percentile range might be in terms of raw score?

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 1:58 pm
by tundrawolf
I just took it. I thought it was extremely difficult: I left 6 blank, and probably 5 others I got wrong. Certainly I don't think I got any more than 20 wrong/omitted. I think I should be able to crack Berkeley's magic 80th percentile, no? Is 80th percentile considered the minimum possible to be competitive, or a real indicator of strength?

Why does nobody want to discuss the test questions?

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 2:03 pm
by mathgirl28
discussing test questions ever is against the rules and if ets finds out they will have your score revoked. 80th percentile is definitely the threshold that i think a lot of the top schools have for their programs (berkley stanford etc.)

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 2:08 pm
by strikeang
are the test questions universal across the globe? Im not sure if we have taken the same test if we were at different locations.

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 2:10 pm
by tundrawolf
mathgirl28 wrote:discussing test questions ever is against the rules and if ets finds out they will have your score revoked. 80th percentile is definitely the threshold that i think a lot of the top schools have for their programs (berkley stanford etc.)
Do you mean threshold to even be considered, or that would be sufficient for admission?

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 2:11 pm
by mathgirl28
considered, i believe. GRE wont rule you in but i think it can rule you out for some of the programs... but someone might know better than i do...

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 2:19 pm
by tundrawolf
how many finite abelian groups did you get? I got 10

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 2:53 pm
by exxx
Ugh. I was hoping to break 80th percentile and I don't think I made it. Hopefully I'm at least in the 70's.

It was much easier than October, but I still messed up pretty badly. I've always bombed on standardized tests. After leaving the exam there were so many problems I thought "shit I know how to do that."

I wish it were offered in the Summer.

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 2:57 pm
by exxx
mathgirl28 wrote:discussing test questions ever is against the rules and if ets finds out they will have your score revoked.
Heh. Every Chinese student I've ever spoken to has seen every question on every past math GRE, because the questions are shared on Chinese language websites. Even my Chinese ODE professor said when he took the GRE twenty years ago, before the internet, he had seen every question on the exam ahead of time. ETS is a joke, their exams have no integrity.

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 3:03 pm
by tundrawolf
For the question about ideals and N(A), the right answer was I and II only, right?

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 4:37 pm
by deckoff8
questions werent really hard they just took forever to do. still much harder than practice obviously.

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 5:07 pm
by goombayao
exxx wrote:
mathgirl28 wrote:discussing test questions ever is against the rules and if ets finds out they will have your score revoked.
Heh. Every Chinese student I've ever spoken to has seen every question on every past math GRE, because the questions are shared on Chinese language websites. Even my Chinese ODE professor said when he took the GRE twenty years ago, before the internet, he had seen every question on the exam ahead of time. ETS is a joke, their exams have no integrity.

Actually, that makes it sound like Chinese students have no integrity.

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 5:14 pm
by deckoff8
it seems like its less a big deal for US students since schools kinda know whats up with foreign GRE scores, but it seems most unfair to Chinese students who actually don't want to cheat but is kinda "forced" to.

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 5:29 pm
by talkloud
goombayao wrote: Actually, that makes it sound like Chinese students have no integrity.
Sounds like both. A huge testing organization like ETS should really have more security than "Promise not to tell? Cross your heart?". I didn't realize that they actually reuse questions. That's kind of messed up.

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 5:35 pm
by topology_shatal
Damn, wish I was I Chinese student.

I was actually shocked by that they allowed to get out to the restroom in time of exam and get back to the test after that. Indeed, some big integrity problems.
Level of test seemed to me smth like the one Princeton review book has in the end of it.

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 5:57 pm
by metalshredder
Hahaha you guys with the remarks about Chinese students are funny.

Anyway, I didn't feel bad about the test that much. I started to get a headache during the test so that was kind of bad. This guy in my room started talking to the proctor from the back. I guess he was asking a question...but I had to shush him.

In terms of the material, I really don't think that it was extremely difficult. It seemed to be a test that's on par with the other ones. I'm really annoyed about that algorithm problem that's slapped on the end. I saw it and thought "ooh an algorithm!" Then I read it and skipped away from it. I felt like my time would have been better used solving some other problem.

Re: Feedback - November 12, 2011

Posted: Sat Nov 12, 2011 11:59 pm
by muneerusman
Anyone has an idea as to approximately what raw score gives a percentile of 80?

Re: Feedback - November 12, 2011

Posted: Sun Nov 13, 2011 12:26 am
by exxx
I talked to my Chinese friends and he said he'd seen many of the questions on this exam online, with even the EXACT same answers down to the correct letter choice.

I don't blame him, it's a simple tragedy of the commons. ETS has to expect natural patterns of human behavior brought on by necessity. We do not, unfortunately, live in a world of Kantian deontology. Chinese students help each other because they care about one another and have a sense of common purpose. Personally I think that shows more integrity than the dog-eat-dog nature of American students.

It's ETS's job to deal with that, but they refuse to. If ETS refuses to, colleges should pay attention and stop using their exams. When the California universities threatened to stop using the SAT, ETS reformatted it. Departments should be more demanding of ETS, or they should find someone else, instead of supporting a bloated inefficient monopoly that has lost all integrity because, being the only dogs on the block, they simply don't need to maintain it.

There are two things ETS really has to do to respond to this: first, they have to stop reusing any questions at all. Second, they should put all past questions online for common viewing by ANYONE, not just those who can read Chinese. These two simple actions would completely restore the integrity of the test.
muneerusman wrote:Anyone has an idea as to approximately what raw score gives a percentile of 80?
I'm wondering that myself. Someone posted in the October scores thread that their scaled score of 800 turned into an 81%. That's a pretty high scaled score for just an 81%'ile, in the past you could get in the 700's and still be in the 80'th percentile. I'm hoping things change for November but if anything it seems it would be worse.

Re: Feedback - November 12, 2011

Posted: Sun Nov 13, 2011 12:57 am
by castlevania87
exxx wrote:I talked to my Chinese friends and he said he'd seen many of the questions on this exam online, with even the EXACT same answers down to the correct letter choice.

I don't blame him, it's a simple tragedy of the commons. ETS has to expect natural patterns of human behavior brought on by necessity. We do not, unfortunately, live in a world of Kantian deontology. Chinese students help each other because they care about one another and have a sense of common purpose. Personally I think that shows more integrity than the dog-eat-dog nature of American students.

It's ETS's job to deal with that, but they refuse to. If ETS refuses to, colleges should pay attention and stop using their exams. When the California universities threatened to stop using the SAT, ETS reformatted it. Departments should be more demanding of ETS, or they should find someone else, instead of supporting a bloated inefficient monopoly that has lost all integrity because, being the only dogs on the block, they simply don't need to maintain it.

There are two things ETS really has to do to respond to this: first, they have to stop reusing any questions at all. Second, they should put all past questions online for common viewing by ANYONE, not just those who can read Chinese. These two simple actions would completely restore the integrity of the test.
muneerusman wrote:Anyone has an idea as to approximately what raw score gives a percentile of 80?
I'm wondering that myself. Someone posted in the October scores thread that their scaled score of 800 turned into an 81%. That's a pretty high scaled score for just an 81%'ile, in the past you could get in the 700's and still be in the 80'th percentile. I'm hoping things change for November but if anything it seems it would be worse.
I agree with you. ETS is a horrible business corporation. The October test I had in Canada was EXACTLY the same test used in previous years in somewhere else (found out later, otherwise wouldn't just get 82%). My friend who took in the US also confirmed that he had seen the test before. So I think ETS may just have different versions of test and circulates around the world.

The GRE subject is definitely not a key indicator in your profile. Many schools are aware of that, though they still recommend students do well on it. I browsed through the results of acceptance on this forum, many admitted to top schools didn't even have stella performance on the GRE subject. It's their LORs, research experience that put them in the spot.

Not to scare your guys, Chinese students are aiming 95%+ in the math subject. They consider anything below that is not impressive. However, how many eventually get into those top programs each year? I once came across a Chinese guy who got 99% gre math but only ended up with NYU master (no funding).

Don't worry too much about these standardized test, your chances lie in your past research experience and strong LORs.

It's unlikely ETS will change the test unless a few big schools start complaining and threaten their profits.

Retrospectively, it's kinda stupid to sign at the beginning of the GRE test. ETS's real implication for that is: you need to vow not to disclose the test content to others, because we will use the same test again next time.

Re: Feedback - November 12, 2011

Posted: Sun Nov 13, 2011 3:27 am
by nikhil_may92
hiiii guys
how much percentile we can expect for raw score of 40+???
thanks in advance

Re: Feedback - November 12, 2011

Posted: Sun Nov 13, 2011 7:29 am
by PNT
Does anybody remember the logic question i think ~Q10? ugh that was not fun.

I feel i did pretty badly but really have no clue, i hate these types of tests, i make too many simple errors that wouldn't trip me up otherwise if i had time to double check.

Re: Feedback - November 12, 2011

Posted: Sun Nov 13, 2011 7:54 am
by rahulshr
PNT wrote:Does anybody remember the logic question i think ~Q10? ugh that was not fun.

I feel i did pretty badly but really have no clue, i hate these types of tests, i make too many simple errors that wouldn't trip me up otherwise if i had time to double check.
3 of the options were easy to rule out, and it boiled down to two (!s1 -> s2, !s2->s1) - which I could not resolve by simple logic manipulations. At that point I just quickly drew the truth tables - and the choice became clear.

For logic questions involving 2-3 statements, number of possible inputs/outputs is just 2x2 = 4 and 2x2x2 = 8.. So it should be faster and more accurate to actually draw the table (when stuck between choices.)... I did not use this in practice, and it occurred out of the blue in the actual test :)


Some of the calculus questions were picked as it is, or with slight variation (different answer choices though) from one of the books I used for practice - Anton's, PR, Schaum's.

Re: Feedback - November 12, 2011

Posted: Sun Nov 13, 2011 8:14 am
by PNT
tundrawolf wrote:For the question about ideals and N(A), the right answer was I and II only, right?
N(A)? i can only remember one question that was on ideals and i must of read the question wrong because none of the given answers were correct??

Re: Feedback - November 12, 2011

Posted: Sun Nov 13, 2011 8:18 am
by PNT
rahulshr wrote:
PNT wrote:Does anybody remember the logic question i think ~Q10? ugh that was not fun.

I feel i did pretty badly but really have no clue, i hate these types of tests, i make too many simple errors that wouldn't trip me up otherwise if i had time to double check.
3 of the options were easy to rule out, and it boiled down to two (!s1 -> s2, !s2->s1) - which I could not resolve by simple logic manipulations. At that point I just quickly drew the truth tables - and the choice became clear.

For logic questions involving 2-3 statements, number of possible inputs/outputs is just 2x2 = 4 and 2x2x2 = 8.. So it should be faster and more accurate to actually draw the table (when stuck between choices.)... I did not use this in practice, and it occurred out of the blue in the actual test :)


Some of the calculus questions were picked as it is, or with slight variation (different answer choices though) from one of the books I used for practice - Anton's, PR, Schaum's.
ugh truth tables so simple :( the worst part about this is i have been teaching first years truth tables this year. God i am an idiot.

Re: Feedback - November 12, 2011

Posted: Sun Nov 13, 2011 2:06 pm
by blahblah1
any guesses how much the qestion count of early forties wud fetch in terms of raw score? if the october test was much tufer and the raw score was much higher for a 81 %iler.. this test is gonna b much worse i feel

Re: Feedback - November 12, 2011

Posted: Mon Nov 14, 2011 12:26 pm
by person
;;;;;

Re: Feedback - November 12, 2011

Posted: Tue Nov 15, 2011 3:18 am
by path
tundrawolf wrote:how many finite abelian groups did you get? I got 10
I think I got 10 as well. It is a simple question, merely counting all the possibilities, but it is easy to miscount too!

Does anyone know why is Math GRE offered only three times per year?

I honestly think it should be offered more often and definitely have more time slots. Morning tests are terrible for some of us who are used to go to bed around 5 a.m. for the past 4 years. For this, and many other reasons (some of which were mentioned by others above) I feel that this test, for the most part, is not a very good representative of ones math skills/knowledge.

In your opinion, would strong letters of recommendation and two or three submitted papers make up for a relatively low GRE score, say 65-70%tile?

Another question, how exactly does the raw score and percentile score work together? In other words, is the scaled score adjucted each time, that is to say, a raw score of say 40 coud place you on the 78%tile on say October, but in November the same raw score could place you at the 80%tile or even 70%tile?

I was really hoping to be on the 80s, but I guess that's all over now.

Re: Feedback - November 12, 2011

Posted: Tue Nov 15, 2011 4:15 am
by castlevania87
path wrote:
tundrawolf wrote:how many finite abelian groups did you get? I got 10
I think I got 10 as well. It is a simple question, merely counting all the possibilities, but it is easy to miscount too!

Does anyone know why is Math GRE offered only three times per year?

I honestly think it should be offered more often and definitely have more time slots. Morning tests are terrible for some of us who are used to go to bed around 5 a.m. for the past 4 years. For this, and many other reasons (some of which were mentioned by others above) I feel that this test, for the most part, is not a very good representative of ones math skills/knowledge.

In your opinion, would strong letters of recommendation and two or three submitted papers make up for a relatively low GRE score, say 65-70%tile?

Another question, how exactly does the raw score and percentile score work together? In other words, is the scaled score adjucted each time, that is to say, a raw score of say 40 coud place you on the 78%tile on say October, but in November the same raw score could place you at the 80%tile or even 70%tile?

I was really hoping to be on the 80s, but I guess that's all over now.
The rescaled score depends on which test you took. Obviously a test with many hard questions will be scaled higher for the same raw score. Then the percentile is determined by your rescaled score: how many students score below that rescaled score.

But I have a feeling that for the same raw score the percentile should not differ too much

Re: Feedback - November 12, 2011

Posted: Tue Nov 15, 2011 9:52 am
by rahulshr
path wrote:
tundrawolf wrote:how many finite abelian groups did you get? I got 10
I think I got 10 as well. It is a simple question, merely counting all the possibilities, but it is easy to miscount too!

Does anyone know why is Math GRE offered only three times per year?

I honestly think it should be offered more often and definitely have more time slots. Morning tests are terrible for some of us who are used to go to bed around 5 a.m. for the past 4 years. For this, and many other reasons (some of which were mentioned by others above) I feel that this test, for the most part, is not a very good representative of ones math skills/knowledge.

In your opinion, would strong letters of recommendation and two or three submitted papers make up for a relatively low GRE score, say 65-70%tile?

Another question, how exactly does the raw score and percentile score work together? In other words, is the scaled score adjucted each time, that is to say, a raw score of say 40 coud place you on the 78%tile on say October, but in November the same raw score could place you at the 80%tile or even 70%tile?

I was really hoping to be on the 80s, but I guess that's all over now.

100% agree mate. I also have a habit of going to bed at about 5-6 am in the morning. And 1-2 months before the test I was fearing that it will affect my performance, so I tried various things to change my sleep cycle - like went w/o sleep for 2 days, worked out extra hard - just so that I can sleep out of fatigue.. but it just did not happen - one can say I did not try enough, as it is not something which can't be changed.
So, in two days before test, I slept only 4-5 hours, so that I can sleep on the day before the test - but even with fatigue, lying in the bed - I could not sleep more than 1 hour and had to wake up at 6..

And the result - after having done 30 questions in first hour, I could not even do 15 in remaining time - not even five in last hour, not even one in last 15 minutes - and those were the type of questions I would crack in practice, with minimal pencil work - in my normal waking hours.


But - as I understand GRE-Subject test is just one part of your CV - if you did well , show it off,... if you did not - then you have to emphasize on other parts of the apps- recommendations, publications, college grades.