Hey everyone,

I just received a good question about practice tests and wanted to share with the group:

QUESTION:

In a span of ten days, I have taken two tests one
from GMAC, GMATPrep Test1 & one from Veritas, GMAT
Simulator. Below is the summary of the tests.

Test Total Verbal Quantitative
GMATPrep Test1 700 39 47
Veritas Test1 660 38 44

Total Incorrect Verbal Quantitative
GMATPrep 14 11
Veritas 10 7

From the above data, it is clear that despite
getting more correct answers on Veritas test, my
overall score was lower than that on GMATPrep.
This has confused me. I don't understand how its
possible that you score less despite getting more
number of questions correct. Could you please help
me by explaining this?


MY REPLY:

Thank you for your inquiry - actually, yours is a situation that corresponds almost perfectly with a blog post that we recently wrote about the GMAT scoring system, so I'll link that here for your review: http://blog.veritasprep.com/2010/08/und ... rithm.html. The way that the GMAT is scored, the difficulty levels of the questions that you answer correctly/incorrectly questions has more bearing on your score than does the sheer number of correct/incorrect answers, so it's not at all unusual for someone to miss a greater number of questions and end up with a higher score, provided that the missed questions are all on the upper threshold of difficulty.

One other consideration, and one that works nicely in your favor - the official GMAT practice tests use the same scoring algorithm and logic as the official GMAT, but any other practice test that you take can only attempt to mimic that algorithm. For that reason, it's almost always the case that the practice test companies will build in a slight negative bias into the scoring so that they counteract the potential for a score that overestimates your score. It's much more beneficial for a test to tell you "study harder" than to tell you that it's okay to rest on your laurels, and since no corporate practice test has the true scoring algorithm or difficulty-level research per question that GMAC does, the tests are typically built to skew slightly low on scoring so that they don't give you a false sense of security. Think about it from a business perspective - no one would ever complain that "my practice tests said 660 but I ended up scoring 700", but someone whose practice tests were all 700+ who scores 650 might be miffed.

Overall, you'll learn a lot more from the experience of taking the tests than from focusing specifically on the score...after all, the only score that really matters is the score that you see on the official test on test day.