Friday, June 06, 2008

GRE Prep

On July 12, I'm scheduled to take the GRE, which these days consists of a couple of essays you have to write, a verbal section, and a quantitative section. I am most interested in my quantitative score, though I plan to do well on the other parts as well, of course. I am generally good at standardized tests, and did well on the SAT back in the day.

One interesting thing is that the GRE uses adaptive testing, where the questions you get depend on how you answered the questions that came before. It's all on the computer, of course, and there is no going back to previous questions. In the beginning of a section, you'll get a question of medium difficulty, and then they move the difficulty up or down depending on whether you answer it correctly. So it's especially important to get the earlier questions correct so that you move quickly into the more difficult questions. My understanding is that once you reach a point where you are answering about half the questions wrong, that level of difficulty determines your score.

I took a sample test for the quantitative section and scored a 690 (out of 800 total), which is pretty bad. I feel like I need a 750 to be a credible candidate for a science-based graduate program. (I'm calling this a "feeling" because I don't know or can't remember where I may have picked up this magic number, if indeed I didn't invent it completely.) Granted, when I took the sample test I wasn't giving it my full and complete attention, but then I wasn't stressed as I might be under actual testing circumstances either.

I had a particularly hard time with the geometry problems, as always. And in general I think I have a hard time accessing my cleverness to solve the problems, which is what they require. (The math GRE questions are a pretty low level of math - basic algebra and geometry - but they often have a "trick" to them.)

My tendency is to employ brute-force methods, which is a terrible strategy when you don't have a calculator and you're under a time crunch. Take this question, for instance:

If the average of six consecutive numbers is 18.5, what is the average of the first five of those numbers?
This can be solved by brute force, as was my initial inclination. Letting x stand in for the first number, you know that
((x) + (x+1) + (x+2) + (x+3) + (x+4) + (x+5)) / 6 = 18.5
It's pretty easy (though arithmetically annoying) to solve this for x, and then take the average of just the first five in the series. But there is a very easy way to solve this problem with a little bit of thought instead, and that's the way I need to remember to do things on the GRE.

So, to that end, I bought a GRE Math Workbook, specifically the Kaplan one. The first part is large review sections on Arithmetic, Algebra, and Geometry. These are divided into sections, each of which ends with a test with easy, medium, and hard GRE-style problems. Each test has an answer key with descriptions of how to solve the problems, sometimes with several methods described. The second part is divided according to the three question types used on the GRE, with several practice tests for each question type.

I haven't been reading the actual review sections, because I basically do know math, of course, but I have been taking all of the tests, even the ones in the Arithmetic section. (It's not basic problems, but GRE-type tricky problems.) I feel like this process is making me more clever and definitely more ready to kick the GRE's ass.

A part of me rebels against the idea of preparing for a standardized test. What am I, some kind of goody-two-shoes who has to work hard because I'm not that bright? Fortunately, that part of me has been hastily stuffed into a sack and sat upon.


Sally said...

It seems that 750 Q is a reasonable goal for competitive PhD programs, but I'm not sure that masters programs would need to see something that high.

I too have been struck by the degree to which knowing how to actually set up and execute actual math is unnecessary to get the right answers. I am leery of my ability to figure out the trick/gimmick in such a time-pressured situation.

Where did you take your sample test? Was it on online test that approximated the computer-adaptive format or an old-school paper-based test? I continue to find the computer-adaptive element the most nerve-wracking part of this whole adventure.

Tam said...

I think I took my test at the Princeton Review web site. I'm not sure whether it was adaptive or not.

Jason Galbraith said...

This Saturday, huh? You don't need it, but best of luck!