how to explain big GRE score jump

[RLB]

I'm planning on submitting applications to several of the top MFE programs this year, and was hoping someone might be able to provide insight or advice on how to handle an issue with my GRE score. Last year, I took the GREs, and for whatever reason had an absolutely horrible day, and out of some stupid misguided hope that the math section (out of 2) I'd bombed on was the "experimental" one, I accepted the score rather than canceling it. And it was very bad - like below 600 bad :wall.

I retook the test this year and scored an 800 on the quant section -a score much more consistent with my overall record (I did very well on on the SATs, got an 800 on the math SAT II, cum laude BA from an Ivy undergrad, got As in various advanced math courses, etc).

My question is, how do I explain this (or do I need to explain it at all?) in my applications. Obviously schools will see the bad score when they get the official report, and my concern is that the fact that it was sooo bad that it will scare off the adminissions people (I'm imagining someone sitting there thinking, "we can't let anyone into our program that could ever do this poorly on the GRE - even drunk!"). In reality, it was just a "bad day" and I don't have a great explanation for it (and I generally think this excuse is almost as much BS as "I'm just a bad test taker", but in this case, I don't have a better one. I could lie and say I studied a lot since last year, but that's not actually true...)

So what are eveyone's thoughts? Am I making a mountain out of a molehill? And if not, how would you address it?

 

[Andy Nguyen]

....eh, my evil twin took the last one?
The risk is that people may end up thinking that your evil twin took this one and you took the last one.
If I remember correctly, schools only see the score that you report, not your historical scores. That's how ETS make money, I suppose.

 

[samuel]

I think general GRE's are silly, especially the math section. Anyone (even Gauss) can mess up arithmetic a few times in any one 30min time interval. And whala, no perfect score, no 800. On the other hand, mediocre undergraduates with B's/C's in their math classes could get lucky a few extra times and score 700. Although a low score on any exam looks bad, I believe a valid case can be made for some test takers; that low GRE scores are uncharacteristic instead of being the norm. Being sick, hungry, nervous, or poor (these tests are expensive, $140 a pop), could definitely screw with ordinary long division routine.
Anyway, just my two pennies, anyone who gets high 90s on all exams in courses such as real analysis, PDEs, etc, shouldn't have to take the math section of the general GRE. In this case, the results of Gen GREs are irrevelent.
Plus, the subject math GRE test is more fun for math lovers; this should be a welcomed substitute for any math program.

I expect that admissions committies care much more about recommendation letters, writing samples, and the four years of straight A's, than one timed test.

 

[doug reich]

It depends where you're going. If you work on a trading desk, you have to very quick with the arithmetic, not so quick with the PDEs. To succeed in an MFE, it is the other way around.

There is no independent, intermediate test of mathematics between GRE General and GRE Math Subject, so the world has agreed on the former.

 

[mwes]

Anyone (even Gauss) can mess up arithmetic a few times in any one 30min time interval. And whala, no perfect score, no 800

Depending on the difficulty of the exam you sit, you can miss a couple questions and still make 800. I personally think too many people receive 800s in math (5% versus sub-1% for verbal). Instead of wacking people down from 800 if they make a simple arithmetic mistake, they should create a new category of difficulty. Then make it impossible to score over 700 unless you get some of those tougher questions correct.