Back in the days of the old GRE, there was only one book out on the market written by ETS (the creators of the test): the hoary 1991 tome Practicing to Take the GRE. At that point, there were few, if any questions, relating to statistics (probably a straightforward median and mode question).
Of course many students would come out of the test reporting how many questions they’d seen on standard deviations or weighted averages. There was a clear discrepancy between what the GRE led students to believe was on the test, and what was actually on the test.
Statistics on the GRE today
Surprisingly not that much has changed. Sure, the new Official Guide has an entire section on Statistics but only a few questions are actually scattered throughout the book, leaving students unsure how many statistics questions are on the GRE. Kaplan and the other usual suspects also give short shrift to this concept, so I still hear the refrain: there were lots of standard deviation questions.
Just how common is statistics on the test?
Don’t get me wrong – the GRE hasn’t become one big test on statistics. But if we were to take all the median/mode, averages/weighted averages, and standard deviation questions on a GRE, there could be as many as eight questions, or roughly 20% of the test. Imagine doing only a few practice problems, only to miss six out of those eight questions. You are already close to 160 out of 170.
How can I study for statistics?
My advice is to not only go through the section in the Official Guide to the GRE, but to also do as many statistics practice questions as possible. The thing is statistics looks deceptively simple when the books cover the usual mean, median, mode business. Questions on the GRE take this relatively straightforward knowledge and concoct these fiendishly difficult questions.
Don’t overdo it
At the same time, you should not be ferreting through your college textbook on statistics, trying to find the standard deviation to the nearest thousandth on a set of a hundred numbers.
The statistics on the GRE is more of a big picture, more conceptual than it is about crunching numbers. This fact will become evident once you start doing a few questions. To give you a foretaste: solving a standard deviation question is less about using the cumbersome formula and more about getting a sense of the standard deviation based on the numbers. Doing so will allow you to eliminate most, if not all, of the incorrect answer choices.