How Would Old GMAT Scores Correlate With New GMAT Scores?

Just wondering, is all. I took the GMAT back in 1984 and did OK on it, more than good enough to get into grad school then for a master’s degree.

Perusing some universities’ sites on their PhD programs today, I come across words to the effect of the usual successful applicant has a GMAT score of 680, with 700+ being fairly common.

Well, my score 26 years ago was not close to 680, let me tell you. But I know that some tests, after being extensively revamped, have had their scoring also change radically.

So it got me to wondering how I would score on today’s GMAT.

Is there a way to do a rough comparison between the two scores or is that simply not doable?

If you could assume that the test hasn’t changed significantly–that is, someone who took the test back then would score the same relative to the other takes if they took the test now–then you could compare the percentiles associated with each score. I don’t know whether that’s a reasonable assumption, but it doesn’t look too bad at first glance.

However, there are a couple other factors that are relevant here:
[ol]
[li]You were looking at master’s programs back then, and you’re comparing that to current PhD programs. You should not expect the admissions standards for those two types of programs to be the same.[/li][li]More people are applying to graduate business programs these days. As a result, schools can be more selective about who they admit and still get sufficiently many matriculants.[/li][/ol]
You should also consider where the school you went to falls on the rankings and compare that to the PhD programs you were looking at. Top schools are more selective.

I dont think its possible to compare. The GMAT today has a writing assessment which counts in the score. I think they had just started that when I took the test.

Plus test takers today have far more prep tools that you had in 1984 and even I had in 1994. So there is probably a fair amount of grade inflation over the years. I am not even sure that my score would even make the cut if I had been applying today.

In any case, they toss out scores over five years old so you would have to take it anyway.

Yeah, I knew I’d have to take it again some day, but was just wondering about my old score.

Neptunian Slug, I did not realize that there was a whole new component to the test; thanks for the info.

Ultrafilter also made some good points about there possibly being more competition nowadays so the schools may be mroe selective. Thanks!

I don’t know the exact formula, but GMAT scores are calibrated to percentiles. So, your score isn’t as good a comparison as your percentile. If I recall correctly, it would be like this: a 650 20 years ago would have been 90th percentile; a 650 now would be closer to 80th percentile. (Note: those are made up numbers, just explaining the trend.)

Hmmm…that would go a good ways toward explaining why universities won’t accept GMAT scores over 5 (I think) years old.

Thanks!

Not exactly. The writing assessments are scored separately, but the scores do not affect your “main”, 200 - 800 score.

I took the gmat in 1984 and my score of 720 was listed as the 99.9+%. It remains a mystery to me why the 1984 test was scaled so that only 1 in 1000 takers got scores of between 720 to 800. The score was a big factor I felt in getting in to Stanford where I went. I did a search on this today because looking at the current Stanford GSB website, it showed their median gmat as 730 which is currently the 94th percentile. I’ve perceived there was inflation in gmat scores through the years but I did not expect it to be this much. As I recall my roommate at the time got a 650 and that was about the 94th percentile then.

In 1994, the 99th percentile score began at 710. Now the 99th percentile begins at 760 and a 710 is 92nd percentile.

Hey matey, allow just a correction on my behalf (perhaps you did only a typo error or you don’t remember well): Most probably you mean that your GMAT score was at the 99th, not 99.9th percentile. I took the old, paper-and-pencil GMAT in 1991 and my score of 730 was equivalent to the 99th percentile, which started from 720.

If it is me who is mistaken, pls excuse me. Perhaps the 1984 year scores used a different scale than in 1991. If it is so, I would appreciate if you told me: I am always interested in learning new facts in the area of psychometrics and academic assessment tests.

The answer to the original query is that the GMAT has been completely reinstated in the fall of 1997, becoming as of that time a Computer Adaptive Test (CAT) Vs a paper-and-pencil test until then: http://social.mbastrategy.com/blog/20-about-the-gmat/169-do-you-know-gmat-history.html
The scoring scale apparently also changed, making the scaled scores pre- and post 1997 not directly comparable. Comparing percentiles is therefore more valid for such purposes, not totally valid though, as many more students are now taking the GMAT, i.e. about 250,000 p.a. Vs only a few thousands in the 50’s, when the test was launched. This means that the average"ability"of the test takers is now lower than in earlier years. Taking into account the Flynn effect, this translates to that today’s testees are not as much superior to the “average person” of today, as the earlier testees were, compared to the "average person"of their era.
However, having read a book on Item Response Theory, IRT, a theory that is applied by the test administrator, Educational Testing Service, ETS since decades now (at least since the 80’s, when I took another ETS-administered test, the GRE General Test and became familiar with what the scaled scores really mean) I suspect the CAT mechanism essentially facilitates the test takers to reach their full potential, therefore scoring higher scaled scores than when the GMAT was a paper-and-pencil test. What I mean is that, perhaps, a GMAT testee before 1997 having scored, let us say, 700 would score 730 or so if s/he took the CAT version. This is a story I can’t really make short, so I won’t expand on it. However, if my suspicion is correct, then pre- and post- 1997 GMAT scaled scores remain of course not directly comparable (even more so now), but there is an additional reason for the percentiles to become such, always taking into account the Flynn effect!