All tests have a 'standard error'; if you could answer an infinite number of GMAT questions, the test could measure you exactly, but because they can only ask a relatively small sample of questions, your test score can be a bit off your 'true level'. This partly comes down to luck - unless you're a perfect test taker, you'll likely be guessing on some questions, and if you're lucky, you'll get a few of these right, while if you're very unlucky, you'll get none of these right. That kind of luck tends to average out the more questions you see, and the length of the GMAT was set so the error due to luck would be quite small. In fact, the test was made computer adaptive so it could be made shorter, while still having a small error; the adaptivity allows the test to estimate your level more quickly and more reliably.
On the GMAT, the standard error is a bit less than 30 points, which means that about 70% of the time, your test score will be within about 30 points of your 'true ability'. About 30% of the time, your score will be 30 points or more away, but it will almost never be much further than about 50 points off. Of course, if you take two tests and get a similar score each time, then you have a pretty reliable indicator of your current 'true level'. The standard error should be smaller for very high ability test takers, though, since luck is not as important for the person who rarely needs to guess.
But with the exception of the inherent error in the test (which is of course present in any similar test), GMATPrep is a good predictor of your GMAT ability - and is much better than any other test you could try, besides the real thing. And on a side note, if you could consistently score 49 on the Verbal, you're so far above the mean that your score will be outstanding overall, just as long as your Quant is respectable. Good luck!
For online GMAT math tutoring, or to buy my higher-level Quant books and problem sets, contact me at ianstewartgmat at gmail.com
ianstewartgmat.com
