Hi wpro,

Percentile data can be deceptive since it measures your performance relative to OTHER Test Takers (not relative to the Test). For example, if you got half the questions correct, but everyone else got fewer questions correct, then you would have a much higher percentile (by comparison) even though your performance was not exceptional. This is one of the many reasons why the Scaled Scores are also included - THAT data gives you a way to assess your performance relative to the Test itself.

Then there are the experimental questions, which do not affect your score. However, if by chance you get more of the experimental questions correct and fewer of the questions "that count" correct, your score could fluctuate significantly.

Don't try to over think or "game" the exam. Instead, focus on these questions that you got wrong. WHY did you get them wrong? Were they too hard or did you make a silly/minor mistake? Keep the silly mistakes form happening and everything else will fall into place.

GMAT assassins aren't born, they're made,
Rich

Hi wpro,

The simple answer to what you're asking is "precision" - NO silly mistakes, EVER. You can still afford to get lots of questions wrong, but it has to be because the question was too hard (and not because of a little mistake, unlucky guess, short on time, etc.).

The more complicated answer is that since percentiles are based on the "pool" of people who took that particular test, the percentile data vs. the Scaled Score data can vary wildly. For example, the pool of people who take the 1st Kaplan CAT will likely be different from the pool of people who take the 1st MGMAT CAT. Hitting the 95th percentile on one exam likely would NOT require the exact same performance as hitting the 95th percentile on the other exam. And NONE of the CATs can truly match the Official GMAT in this regard (although many come close enough to be considered realistic).

Instead of basing your goal on the percentile, base it on the Scaled Scores. THOSE numbers provide data about how you performed vs. the actual Test, which is a far more helpful to your overall studies.

GMAT assassins aren't born, they're made,
Rich