Useless GRE math test

[AcidThinking]

Hello guys,

To what extent do you believe in this test? Praise and criticize. But of course explain.

[dexter]

The word around the block is that it is only meaningful if you do badly. I personally don't care much for it, and it's very easy to slip and do badly if you aren't well-prepared. Doesn't mean you aren't good enough, but it makes your case that much harder to argue because there's plenty of other applicants with solid scores.

The GRE is apparently good at what it does because I haven't been seeing grad admission directors complaining about it (at least not publically). So I think it is here to stay for a while, and people who want to get into good schools will have to score reasonably well to be considered.

[colldood]

it's crap and mostly meaningless. but schools care, to varying degrees

[jplusip]

It's absolutely meaningless. You could be fresh off of finishing Little and Big Rudin and be an expert in Sheaf Cohomology and fail the thing.


Your best bet for preparing for this test is to take the Physics "Math course" and the Computer Science "math course" the same semester you plan on taking this test,

[AcidThinking]

I totally agree. That's what I used to think: if you're an engineering major who has taken few extra maths courses then you would perform better.

The mentality required to get a high score on this test is not mathematical, but engineering mentality: you know, be down-to-earth and pragmatic, solve it quickly and accurately, see what tricks you can do to maximize your score.

[marco]

I agree, I think that the test is successful in assessing the level of knowledge you have in undergraduate math but definitely does not measure your ability to finish a PhD in Math.

[colldood]

I agree, I think that the test is successful in assessing the level of knowledge you have in undergraduate math but definitely does not measure your ability to finish a PhD in Math.

Since when do math majors spend all day doing these kind of absurd problems? I was busy proving things and trying to understand abstract ideas.

[marco]

My point is that if one is not comfortable with basic concepts in Math (regardless of major) then it is likely that one will not do well in this test. But this test does not measure how well one can grasp abstract ideas and prove results which is what one does in a Math PhD and what we did as undergraduates in Math. I personally don't care for the test and really hope that universities use this piece of information (GRE scores) of our apps with caution.
In a nutshell, I personally don't think that someone who scores in 80 percentile or higher is more likely to be a better mathematician than one who scores in the 60 percentile. It just means that he is a more pragmatic, swift and precise problem solver. However, if one scores percentile 30 or lower than it is pretty clear that this person does not really know his math (especially considering that 50% of the problems are of Calculus).

[vonLipwig]

The test is a very good assessment of your ability at undergraduate topics (particularly calculus, but also a mix of others). I would expect the results to be highly correlated with success. I do not claim that this correlation is perfect, and accept that this is a bit rough on people who specialised early in more abstract areas and haven't done many calculations recently. Keep in mind that the admissions committees see your GRE scores not in isolation, but with your transcripts, etc.