GMAT scoring inconsistency

[csnyder30]

Long time forum watcher first time poster. I recently took the GMAT (for the 5th and final time) and scored a 690 (Q47-78%; V39-87%) and was wondering why my score was a 690 instead of a 700. I have searched high and low on the internet and have not been able to find a 690 total score with that Q/V split. All of the Q47/V39's I see equate to a 700 combined score. I have surmised that one of three things is happening... 1) every post that I read describing the scaled score breakdowns on the various GMAT blogs (beatthegmat, GMATClub, MGMAT, etc,) is inaccurate and the people posting are lying about their splits and/or combined score...2) the GMAT calculated my total incorrectly, or 3) one can actually have different component verbal and quant scaled scores that aggregate to the same combined score. This third scenario is what I find most troubling because it implies that there is some "black box" or heretofore unknown factor in how the GMAC/GMAT combines the component scores to arrive at a combined score. If this were true, wouldn’t that argue that the test in not at all standardized? In other words you would expect the same component scaled scores to produce the same scaled combined score? I realize that the percentile scores have shifted over time in terms of which percentile corresponds to which scaled score.

Have contacted GMAC about this and it's working its way through some maze of a bureaucracy as Pearson is only the administrator, there is actually a third party who runs the GMAT program and a second third party who examines the test results and does the scoring. This all sounds like its set up to discourage and/or deter customers from contacting them. At any rate, would love someone’s feedback on whether it’s worth fighting or if there is a plausible explanation for my combined score.

[tomslawsky]

I agree 100% and I don't like an exam that has a secret formula for scoring when the consequences of the exam score are so high. I also think the GMAC should score in increments of 5 rather than 10. I have been n oticing that people are reporting a lot of the same verbal/quant scores and a scoring difference of 20 friggin points. The GMAC should have to answer for this and at least verify that score reporting isn't being botched and covered up internally. The million dollar question is how to generate enough pressure to force the GMAC to change policy.

[csnyder30]

Not sure if there are enough people out there with similar circumstances, but I am willing to start a petition [:-)]! They don't two bits about this, but why should they. GMAC/GMAT has a virtual monoploy. Guess there are some positive signs that GRE is putting more pressure on GMAC/GMAT, which I hope keeps everyone honest.

[StaceyKoprince]

Over time, the scores do change - the percentiles, the specific combinations that lead to a particular score, etc. If you heard of someone getting the same two-digit subscores the day before you did and getting a higher 3-digit score, then yes, you should wonder what's going on and ask. But if someone had the same two-digit subscores a year ago and had a higher 3-digit score, then that's just reflecting minor changes to the test and the scoring algorithm over time.

People also do post things incorrectly all the time - not on purpose (usually), but I've talked to plenty of students who give me score combos that are so off that I can actually notice that something's wrong, and I ask them to clarify / correct. So I'm sure that also happens just as frequently with numbers that are too close for me to know that there's a typo.

Also, this:

One can actually have different component verbal and quant scaled scores that aggregate to the same combined score

is true, if what you mean is that someone else could have, say, a slightly higher verbal than you but a slightly lower quant and the two of you would end up with the same combined score. This has to happen sometimes - there are more combos of two-digit scores than there are 3-digit scores available for every two-digit pair to have its own unique 3-digit score.

[csnyder30]

Thanks Stacey - I have seen posts as recent as October and November claiming a Q47/V39 split combining to a 700, so I have contacted GMAT/GMAC.

What I meant to say in this quote:

One can actually have different component verbal and quant scaled scores that aggregate to the same combined score

Was that two people with the same sub scores combining to two different combined scores. This is a situation that I would find troubling.

[StaceyKoprince]

I have seen posts as recent as October and November claiming a Q47/V39 split combining to a 700, so I have contacted GMAT/GMAC.

Then it's a good idea to contact them, just to make sure. (Though be prepared for the answer to be: the algorithm just changed slightly. The change has to come sometime, right? So, at some point, the score is going to be one thing one day and a different thing the next day - the change has to be applied at some point in time. Unlucky. But true.)