GMAT Club had a chance to attend this year the GMAT Summit that took place last week, December 6th in Los Angeles.
This is a bi-annual event held since 2009. You can read our previous reports
Here is the report from this year's event as summarized, reported, and posted by MGMAT here (much better so than we could have done)
News from the GMAC Summit (from the makers of the GMAT)
Stacey Koprince — December 10, 2013
GMACLast Friday, I attended the biannual GMAC Summit, a special conference that the makers of the GMAT put on for test prep companies. I want to share various tidbits that you should know!
IR has existed long enough now that GMAC is starting to be able to draw some conclusions about the efficacy of the section. Dr. Lawrence Rudner, chief psychometrician of GMAC, is quite pleased with the section’s performance to date.
Though they still need to collect more data to be sure, early results indicate that IR is actually a little bit better of a predictor of grades in business school than are the quant and verbal scores. It will still be a while before they can collect as solid / extensive data as they have for quant and verbal, but perhaps it will be the case that, eventually, IR will become the most important section! (Don’t worry: if you are applying right now, nothing has changed. Even if you aren’t planning to apply until next year, it’s unlikely that the importance of IR will change extensively by then.)
There were no admissions officers in attendance, but we did hear from GMAC that they have heard that admissions consultants are starting to consider using IR as a tiebreak for borderline cases. For example, let’s say a school considers 680+ a strong score and 630 to 670 an average score. For the pool of 630 to 670 candidates (only a few of whom are likely to be admitted), one potential tiebreak is the IR score.
If IR is not your thing, don’t worry: it’s unlikely that any school is making this tiebreak decision based solely on the IR score. After all, many different variables go into an application; they might also decide to use number of years of work experience, under-represented industries, or some other factor. If you do tend to perform well on IR, though, then bonus: that’s an extra mark in the plus column for you.
Interestingly, US students are tending to do a bit better on IR than all other students. (This is also true for the Verbal section of the test, while non-US students tend to do better than US students on the quant section of the test.) A lot of people consider IR more of a quant section, but verbal is just as important. If quant is your strength, then you’ll feel that IR is testing verbal more, and vice versa.
Scoring and Timing
I have only one piece of info for you here, but it’s quite an important piece of data. As we were discussing the scoring algorithm, someone asked the age-old question: whether certain questions were “worth more” than others. Dr. Rudner indicated (as he always has in the past) that the earlier questions are not worth more than the later ones. He did, though, indicate something that we suspected but had never heard officially confirmed: “outlier” questions ultimately count less towards your score.
What’s an outlier? Briefly, an outlier is a question for which your performance was unexpected. Read on to understand what this means.
An outlier is always relative to your own performance. (Note: we’re also talking only about the questions that count towards your score; the experimentals don’t matter here.) Most of the questions you answer will be within a certain range of difficulty. As a general rule, you’ll answer more of the easier questions in your range correctly and you’ll answer more of the harder questions incorrectly. This is the expected behavior.
There are two types of outliers:
(1) easier questions that you get wrong
(2) harder questions that you get right
If you were mostly answering 70th percentile questions correctly but you answered a 50th percentile question incorrectly, then that question will ultimately count less towards your overall score.
The Takeaway: don’t stress over missing a question that you think is easier (but note that you can’t let this happen too many times or those questions will no longer be outliers!)
If, on the other hand, you were mostly answering 70th percentile questions correctly but you answered a 90th percentile question correctly, then that question will also ultimately count less towards your overall score.
The Takeaway: don’t waste a bunch of time and mental energy trying to get a really hard (for you) question right.
Now, some of you are thinking, “But what if I can get a lot of those really hard questions right? Then they won’t be outliers and they will help my score!”
That’s true. But can you spot the problem with that line of thinking?
If you were already capable of answering that level of question effectively and efficiently, you wouldn’t need to be spending extra time or thinking, “Oh, I really need to get these right!” You’d already know how to do them. J
Plus, spending extra time on a really hard (for you) question does not actually guarantee that you’re going to answer it correctly, so even if you do happen to get one right every once in a while, that question is going to end up being an outlier for you—and you’ll miss other, easier questions later in the section because you will have mismanaged your time.
It gets even worse. Those easier errors are no longer outliers, because there are too many, so your entire score is pulled down. (And the farther down the overall score is pulled, the more the couple of harder questions that you got right turn into outliers, so they can’t help pull your score up!)
What to do? What we’ve been telling you all along: maintain a steady performance throughout the test, neither rushing nor spending too much time, and letting go of a question when needed (everyone will need to do this!).
The Quant Scoring Scale
Some of you may have noticed that the top quant score of 51 is now only the 97th percentile. We were wondering whether GMAC would add to the scoring scale (maybe including scores of 52 or 53) in order to provide more categories of rankings at the top of the scoring scale.
They have no plans to do this. The top business schools have told them that any score of 48 (76th percentile) or higher is good enough for them; they don’t need to have additional levels of score discrimination at higher levels. Because they don’t need that info, they actually don’t even want GMAC to change those scoring scales. The schools already have their systems set up according to the current scales. In other words, if it ain’t broke, don’t fix it! (Have you heard that idiom before? Don’t worry—they won’t use it on the GMAT.)
After the last Summit, 2 years ago, we reported that the test has been “scrubbing” out American idioms and constructions that are used only in the Americanized version of English. Dr. Rudner reaffirmed this information. He also indicated that OG13 still contains examples of Americanized English that will no longer appear on the real test—so if you speak British English, Australian English, Indian English, or any other form, and you see a practice problem that seems wrong from your English’s point of view (such as collective nouns, which are singular in American English but plural in British English), then don’t worry about learning that construction. They’ve pulled such constructions from the real test.
Also, a note to American-centric students: they have started including certain details that might make you a little more uncomfortable (but your international counterparts have been dealing with such discomforts for years). For example, a math problem dealing with money might use pounds instead of dollars. Money is still money, but it’s always a bit easier when you’re reading about your own currency. A reading passage might talk about a historical figure who was very well known in Asia but not so well known in western countries; again, American-centric students have been benefiting from these kinds of subtle advantages for decades.
That’s all I’ve got for you for now; let us know if you have any questions!