Manhattan/Williamsburg GRE & GMAT Math Tutor

The Quantitative Score range is 0 - 60, and it says that scores over 50 are extremely rare.  I consequently was shooting for a 60 when I took the test on June 6th 2012.  I was dismayed to receive this score report:



Although the pursuit of a perfect score may be a fools errand, I wanted a 60, and was very disappointing that I might have made a mistake on a question, so I took to the internet to find out how rare scores higher than 51 really are.  What I found was a lot of confusion and misinformation, but almost no one even claimed to score higher than a 51. Curious.

To get to the bottom of it, I spoke to an authoritative source within GMAC who was very helpful.  Let me set the record straight and say that for at least 14 years there has not been a score higher than 51 on the Quantitative GMAT.  In the past there were some scores higher than 51, but when they migrated to a computer-based test, 51 became the maximum. (it's also mentioned here.)

This lead me to believe that, unlike the old version of the GRE (which ran out of room in its scoring range), the GMAT could differentiate the top couple percent without having to goto a new scoring range.  They could use the scores 52-60 to differentiate the top two percentile.  In fact they were going to do this recently, but the schools said it would be too confusing if some 51s were maximum scores, and others weren't.  So that was scrapped, and they will probably only differentiate the top better when/if they change to an entirely new scoring range/system (like the ETS did with the revised GRE in 2011).

Speaking of the top of the Quantitative Scorers, 2% getting the max score seems a bit high [not nearly as bad as the old GRE (~6%!), but not as good as the new GRE (<1%)]... let's examine what's going on: (the following distribution graphs come from this great tool from GMAC.com)

Here's the Verbal Scores:


A nice normal curve, max and min's are rare, and the mean is also the mode.

Now for the Quantitative:



Which doesn't look nearly as normal, the most frequent score is not the maximum, but it's close.  And the max extreme of 51 is more frequent than each scores less than 23, this is not an ideal distribution.

Now lets look at Quantitative Scores from Americans (maybe we are just getting smarter over the years?):

Here we have a relatively normal distribution, the mean is the mode, and the extremes are rare... So it's not that we are getting much smarter.... Let's look at GMAT Quantitative scores from international test takers:


This is rather shocking, the Chinese and other Asian test takers score really well... And since the gmat is getting more popular internationally, this explains why the total curve is getting top-heavy today.

In 2010, 30,264 GMAT tests were taken by Chinese citizens (out of china's population of 1.3 billion), and 127,061 Americans took the test in that period (out of our population of 0.31 billion).  So it might be that only an elite class of students takes the test in china, and that is why they all score so well...  Or, maybe they are all just smarter in math than us Americans : )

Either way, this is why so many test-takers (2%) got a perfect 51 on the Quantitative section of the GMAT this year.

For more info on the GMAT, check out GMAT.com and GMAC.com