All GMAT Scores Over 7XX Are Treated The Same...

Well everyone agrees that AdComs use GMAT scores as an evaluation point below the 700 or so level. Lets stick to whether or not above that level higher scores affect the AdCom's decision making process.

So given that, it seems you are arguing that stronger applicants tend to have higher GMAT scores but higher GMAT scores do not imply stronger applicants (at least at that > 700 level). Further, you seem to be arguing that weaker candidates tend to have lower GMAT scores but lower GMAT scores do not imply weaker applicants (again at the > 700 level).

This is an interesting premise but again one that I have seen no data to support. Further, given the correlation between the two, and the fact that GMAT scores are the only fully objective evaluation point, doesn't it make more logical sense that AdComs would take advantage of that correlation (whether or not it is direct)?

To continue with that thought, given two candidates similar in all other aspects but one with a 730 GMAT and one with a 760 GMAT, it seems likely that the GMAT score would at least bias the AdCom's decision process subconsciously (due to past correlations) even if they try not to consciously give that preference.

I have often thought about this. Of the following admission criteria, only GMAT scores are objective
Essays: I know of many cases where the essays are drafted by admission consultants. Those essays are bound to be good. Even otherwise, most of the candidates spend weeks in writing the essays and get them reviewed by "experts". And frankly, everyone waxes eloquent, inflates achievements....
Recommendation: I once talked to a B-school professor, and he told me that they don't give much weight to the LORs cuz all candidates get good LORs. (My views: candidates select those recommenders that are likely to give good recommendation)
Extra curricular: These are not a decisive criteria - I believe you just need to show some ECs that shows you are a well-rounded individual.
Work-ex: This is objective as well. The brand matters a lot. A candidate with 3 years in one of the top IBs or McK, BCG has a far better chance than someone from a nondescript company - but hey, this is justified. The recruitment of these companies must have already done the candidtate-filtering that the B-schools can avoid. Again, views are personal, without enough data to back and I m open to contradictory viewpoints.

Any takers?

I agree completely both that brand of employer is very important and that it should be. AdComs know more or less how to compare titles and responsibilities at blue chip firms. However, that same criteria at an unknown start-up is easily inflated and is very difficult to compare objectively.

speaking of questionable information, i always disagree when people say that ECs are not an important part of an application.

i suspect that ECs do not matter much for someone in their late 20's or early 30's who is well established, but for young applicants (25 or under), i think they matter a lot. most of these people are still in the "crappy" job phase of their careers and need to add value any way they can. you may work at an impressive wall street job, but how much leadership experience do you have? you manned the copy machine for a while and put together some basic financial models? right, you have no meaningful leadership experience at work. so you'll make this up with grades and GMAT, right? wrong. everyone else has top grades and GMAT scores, too. you need to find a way to stand out. for some candidates, i think ECs matter a lot, but it has to be something meaningful and (preferrably) related.

in my own case, i manage a seven figure non-profit endowment (equity and fixed income) and decide which non-profit causes to support with the $45,000 budget the fund produces. additionally, i manage ~100 volunteers in connection with the non-profit. i suspect that my volunteer work is more impressive than a lot of candidates' actual work ex (mine included). sorry to thread-jack, but i have seen the notion that ECs are worthless repeated a lot lately and i think it is wrong.

anyway, regarding the GMAT, i doubt that there is any meaningful difference between 760 and 730. to start with, both scores are above the average at even the pickiest schools (Stanford). additionally, there is a 20-40 point standard deviation in the score anyway (i can't remember the exact standard deviation but it is around there). so i doubt that a 760 is very (if at all) statistically different than a 730. the difference may come down a luck of the draw between a few questions (e.g. if you get questions you click with on one particular test, you may get a 760, and if not, you may get a 730 -- it doesn't really matter). does the higher score subconciously influence the adcom's perceptions? probably, but the application is read by several people, so i think this affect is probably somewhat diluted by the time the final decision is made.

parsifal - there was a thread with very similar research posted here a year ago. The poster then came up with results and conclusions similar to yours. When it comes to gmat obviously the higher the better, however, I find it difficult to say that "two candidates are similar in every other way."

To me, the essays are a big distinguishing factor. No two applicants are ever going to write two identical essays; this means that it is not very easy to say that candidate <A> 3years@IBMinNYC/3.5GPA/26Yo/700 and candidate <B> 3years@IBMinNYC/3.5GPA/26Yo/790 identical in every way other than GMAT. If <A> had a gmat of 700 and <B> a GMAT of 790, I'd be initially more intrigued by <B> ; its human tendency to look at easily quantifiable aspects of comparison first.

However, then <A>'s essays could make laugh and keep me interested till the last word, whereas <B>'s essays could bore me to tears. In this scenario I, as an ad-com, wouldn't hesitate to ding <B> straight away. So your analysis can be narrowed to "what is the probability that two such similar candidates will have such different essays ?" - very difficult to tell in my opinion.

Well this all seems reasonable except it doesn't explain the correlation between GMAT and acceptance rate. Unless, of course, you are arguing that people who write better essays tend to have higher GMAT scores. I'm not sure that I am ready to buy that correlation.

Well just because two scores are within the standard deviation does not mean that there is no statistical difference. Just as there is the possibility of the test takers competencies being the same, the differential could just as easily be as large as 790 to 700. This certainly implies that on average the bias should be given to the 760 over the 730.

I also don't understand the argument that once the score is above the average at a school it is no longer an important data point. If the GMAT implies the degree in which a candidate possesses a valuable competency up to 7XX, then why would it not imply an even greater level of competency beyond that. This seems to me like someone arguing that once you are up to the "average" level of work experience, no credit is given to anything beyond that. Clearly, this is not the case.

I agree that essays do make a difference, but given your example, both <A> and <B> would spend weeks drafting their essays and have them reviewed by scores of friends/well wishers. In the end, the difference in the quality of essays wont be so high as to cause an outright rejection for <B>
Of course, the interview can do the further filtering to check if the candidate matches the quality of the submitted essays, but the adcoms will not invite candidate <A> for interiview - remember there would be too many such candidates with really good essays with low GMAT.
My point is that the adcom will invite <B> for interview to further probe him/her (assuming of course that he/she has written a reasonably good essay that have been reviewed by many).

I personally believe that not all GMAT scores are the same. That statement is probably more significant when the scores are further away from 800 and becomes slightly insignificant as scores approach the maximum(800). For instance, I believe there is a big difference between a 740 and 760. That difference is less significant when comparing 770 and 790. As for ad-com bias based on GMAT, I am sure that exists too. Just like bias exists based on brand name of UG, employer, country of origin, gender, perceived soft skills of certain ethnicities, etc...I dont buy the statement that GMAT scores above the schools average dont make ANY difference. So, I agree with livehard on this count.

Not everyone with a high GMAT has a good application. But there is enough number of high GMATs with well-rounded personality who put together great applications which helps them get in. So, I agree with dabots on this count. The % acceptances is probably indicative of a correlation rather than a cause and effect.

In the end, a lot of what happens depends on the subjective view point of the admissions committee. Perception matters. You can be a really great in your chosen line of work, but if the reviewer does not fully comprehend the level of difficulty associated, then you automatically become less favored. Multiply this randomness with the number of such admissions factors and the number of people involved, you get a huge level of unpredictability.

Net net, as an applicant our job is to do whatever we can to maximize every facet of our profile. That's why I dont pay heed to statements like "you are above the average GMAT for the school, dont waste time on your GMAT".

based on the information available, it is reasonable to conclude that once you are above average in one category, the adcom turns its attention to other aspects of the application as you have effectively "cleared the hurdle" in that category. your competency has been proved -- the relative degree of compentency above a certain point has a deminishing positive impact on your application. a 790 will not save you from deficiencies elsewhere. if you have a 790 and perform well in every other relevant category, bully for you -- you will probably get in, but not because you earned a 790. success in the process is not dictated solely by your GMAT score. the schools are adamant about this, and even if you do not trust the commentary that comes from the adcom, it seems obvious that the schools could have average GMATs closer to 750 if they wanted to. yet they do not, because GMAT is only one part of the equation.

additionally, i am not sure that a one school sample is enough to base a theory upon. kellogg is at the low end of the m7 (sorry guys, i know some of you love kellogg) and is, as business schools do, trying to climb the ranks. it's natural to think that kellogg would have a penchant for higher GMAT scores as it tries to improve its rank and relative prestige. alas, we are limited by the available data here. i suspect, for what it's worth, that you would see a different trend at harvard. you would probably also see a difference within applicant demographics.

in terms of being above average in other categories, i suspect the same rule of thumb applies. 9 years of work ex with a 3.0 and 600 GMAT score probably will not get a person into kellogg. that person may have a lot of experience, but they "fail" in other categories, and the additional benefit of the extra work ex is not enough to offset those failures.

i acknowledged that there may be a small statistical difference. certainly the score that is 30 points higher should be given additional weight despite the random elements involved. the data you have provided indicates that higher scores are priveleged, and this seems reasonable and intuitive. the disconnect is that you are basing your entire argument on the easily quantifiable data rather than on the subjective aspects of the applicant pool that cannot be measured. it's not possible for us to prove that someone with a 760 should be accepted over a similar applicant with a 730 every single time. is the higher score applicant accepted most of the time? perhaps, but we can't tell. this is business school, not a pure measure of academic merit (see law school). it's tempting to try to prove something with the limited data set available, but in the end we cannot prove anything. the process remains a black box.

I don't understand why people believe that 20 points is such a large difference. Can you explain it to me? Twenty points represents a handful of questions on a randomly generated exam. Your score will almost certainly fluctuate 20 points every time you take it.

Bsd-I dont quite buy your assumption. I believe that the content of your essays is more critical than the style. This is NOT to say that style does not matter, but style without content is useless. Essays are critical, no doubt, but your choice of stories and your "message" are the biggest takeaways for the adcom IMO.

It is a question of %iles. 740 is 98% and 760 is 99%. Simply speaking, once you get 760 all test takers fall under the same category because you are better than 99% of the people taking the exam. Again, this is something you can choose to disagree with. If you do disagree, then it does make my point that since there are so many opinions around it is always in the applicant's best interest to maximize GMAT scores.

I believe that GMAT is nothing more than a verification that x years out of school you can / can not successfully handle the quantitative rigor of the program. As also can you effectively communicate as a leader and as a member of student community in a lucid and correct language. I believe that once you score above 700, you establish that fact to a large extent.

As you can tell, I am having a real hard time getting myself to do any real work today...



I disagree. Most all candidates have some relative deficiency in an aspect of their profile and hence have to be relatively exceptional in another aspect to compensate. This can come from work experience, extracurricular activities, or academic dimensions. A GMAT of 790 certainly fits the bill as an exceptional academic dimension and I do believe it can compensate for weaknesses in other aspects. Just as exceptional quality of work experience (I was talking about quality not length of work experience earlier) can compensate for a low GMAT score.



I agree with all of this. We are certainly working with a dearth of quality data.



Again, I was talking about exceptional quality of work experience not length. Low GMAT scorers are regularly admitted based on exceptional work experience.



I agree that it is a black box and nothing can be proven with the available data. However, it still seems logical (and the limited data seems to support) the fact that higher GMAT scores imply better candidates. This in no way implies that it is the only, or anywhere near the most important, criteria used.

Or I could take the guy with 760 and take TWO interesting candidates with slightly lower GMAT.