I had a student write me an e-mail earlier this year complaining about a lack of improvement on his last few practice tests. On three consecutive tests, he scored a 580, 560, & 580. Frustrating for him. But when I looked at the dates of those tests, I wasn’t surprised by his lack of improvement— he had taken the tests on three consecutive days. If students were able to improve by doing as many test questions as they could possibly find, then every student would do better on the verbal section than on the quant section because he or she has already spent two and a half hours “studying” before the verbal section begins.
Tests are not good learning tools. Tests are good for assessing what you would likely score on a real GMAT if you took the test with similar problem solving abilities and timing strategies.
This doesn’t mean that tests aren’t important for improving your GMAT score. I’d argue that they are the single most important tool that everyone (yes, everyone) can use to (eventually) improve his or her GMAT score. But think back to the last practice GMAT that you took: by the time that you finished a two and a half hour test that included 78 quant and verbal questions, how many questions did you remember? Probably not a lot.
When that student scored a 580 on his first practice test, he wasn’t satisfied with his score, but he didn’t do anything differently the next time he sat down and took another test. If insanity is doing the same thing over and expecting different results, then the GMAT was making this student insane. So I pointed out two areas for improvement that every student can look for after finishing a practice test: timing and weaknesses. Let’s take a look at one such practice test to see how you might be able to improve your overall GMAT score:
This is a student’s trajectory over the course of an entire quant section of a practice exam. When a student gets a questions correct, the next question becomes more difficult and when incorrect, the next question becomes easier.
Problem #1: For the most part, this student’s true ability level falls in the 60-80%ile. But his final quant percentile was below 60% because he got the last six questions wrong. As the GMAT tried to offer easier questions, this student was unable to answer the questions. If you’ve ever taken a practice test, you could probably guess that this student ran out of time and was forced to guess on the last several questions. Even though this student was capable of correctly answering ~75%ile level questions, his score did not reflect this.
Solution: The next time this student sits down to take a test, he needs to do a better job of keeping track of his timing during the test. Maybe he needs to write down benchmarks for how much time he should have left after finishing 10, 20, and 30 questions (55, 35, & 15 minutes, respectively). Maybe he decides to spend less time on questions near the beginning of the test or cap how long he spends on any one question. Maybe he will guess on three questions that he looks at and thinks will be too difficult to answer in just two minutes. And maybe he realizes after his test that this strategy doesn’t work, but at least he knows more about what doesn’t work and he can continue to refine his test-day strategy to search for something that will.
Problem #2: Take a look at question #17. While the GMAT considers this question to be a 45%ile level question, our student did not. And if this had been the very last question on his test, he would not have been able to answer this question correctly, even if he had two minutes to do so. Similarly, questions 15, 16, & 20 were beneath, what the GMAT would consider, his true ability level. And if those four questions were the last four questions on his test, he would have had a dip similar to the one that occurred due to timing.
Solution: Spend time reviewing questions. For every test that you take, you should be spending at least two hours reviewing test questions— studying questions that you got wrong or spent too long on and trying to get better at these questions. The easiest way to score a 700 on your exam is not to get better at 700+ level questions, but being 100% certain that you can correctly answer every question between 200-690 in less than two minutes.
Getting better at anything involves identifying weaknesses and working to improve on them. The same thing is true on the GMAT. Doing problems doesn’t help you improve; learning better ways to solve problems does. So the next time you sit down to take a practice test, understand that your goal isn’t to get a 700 on this test, but to work towards getting a 700 on the real one.