Couple questions... please advise.

There are certainly more than 2. We have been trying to get the answer from GMAC for a long time

Having said that, here is something that we can say: the experimental questions would be around the middle of the test. Not at the beginning or the end. Using that logic - you can pretty much try to keep your educated guesses between 15-30q.

https://gmat.crackverbal.com/gmat-scorin ... ed-partly/


If you cancel your scores:

You will not be able to view them at a later time unless you choose to reinstate your score. Fees apply.
You will not be eligible for a refund of any test fees.
The score cancellation will not be reported or otherwise indicated on all future score reports.
You must wait 16 calendar days to retake the exam.

If you do NOT cancel your scores:

You will receive a printed copy of your unofficial report for Quantitative, Verbal, Integrated Reasoning and Total scores at the test center.
You will have 72 hours to cancel your score online via your MBA.com account for a $25 fee.
An Official Score Report, including the scores for the Analytical Writing Assessment, will be made available to you and your designated GMAT programs within 20 calendar days after you take the test.

From:
https://www.mba.com/india/the-gmat-exam/ ... cores.aspx

HTH,

Arun

Thanks a lot!! This is really an eye opener. When I gave my mock test I always tend to put random answers for last 4-5 questions, because of lack of time at the end of the test. Although I used to make only a few wrong answers, and those mainly toward at the end, it costed me very much, plunging my score to V31-34. Now I know, at least partly if not fully, where to speed up and where to wait and spend time to getting the answer right. So, my best bet is to make sure I should go through first 15 and last 10 questions with enough time and focus.

Be careful who you trust, AnotherGmater --there is a great deal of incorrect / outdated information in this response from CrackVerbal. First, we do in fact know the exact number of experimental questions per test (23 out of 90 questions total = 25.6% of all questions), and second, we know that they are evenly distributed among the test, not concentrated in the middle of each section. We know this not because the GMAC has told us directly, but because we have inferred these facts from analysis of the ESRs (Enhanced Score Reports). This has been common knowledge among most serious GMAT tutors for several months now.

IR has 9 questions that count, and 3 experimental. We can tell that the IR section has exactly 9 non-experimental questions because the "percent correct" always aligns with a denominator of 9.

Quant has 28 questions that count, and 9 experimental. We can tell that the Quant quarters are split into 7 non-experimental questions each because the "percent correct" always aligns with a denominator of 7. The first 3 quarters of the Quant section have 2 experimental questions each, and the final quarter has 3 experimental questions (2/2/2/3).

Verbal has 30 questions that count, and 11 experimental. We can tell that the Verbal quarters are split into 7 non-experimental questions (2nd and 3rd quarters) or 8 non-experimental questions (1st and 4th quarters) each because the "percent correct" always aligns with a denominator of 7 or 8. The first quarter of the Verbal section has 2 experimental questions, and the final 3 quarters have 3 experimental questions each (2/3/3/3).

Brilliant Analysis!! Awesome! Thanks a lot for the whole explanation.
So, I guess, the best way would be try to attempt all the questions with equal focus and attention irrespective of which quarter/part I'm in, since we don't know that experimental question might appear anywhere in that quarter.

Correct! Happy to help.

Hello McElroy,

Thanks for the inputs.

Just to clarify our aim has never been to misguide users. The numbers you have cited were common knowledge among test prep instructors for a long time - just that we did not have data to back it up

Point #1: Total Experimental Questions

In 2013 when I confronted the GMAC official Dr.Lawrence Rudner (the chief Psychomatrician at GMAC) he refuted the numbers. I have written about it here:
https://gmat.crackverbal.com/gmac-test-prep-summit-2013/
(Point #2 from a blog I wrote 3 years ago: I say "experimental questions may not be 1/4th as previously assumed".)

Also, I am not sure it was common knowledge with all "serious" GMAT tutors. Here is a blog from a few months ago from ManhattanPrep by Stacey Koprince:
"we don’t know exactly how many counted questions are in each quadrant; these reports strip out the nonoperational, or experimental, questions"
https://www.manhattanprep.com/gmat/blog ... re-report/

Point #2: Spread of the Experimental Questions

The belief was that the experimental questions are through out the test and there is no way to know when you'd encounter one. However, again it was told in a closed forum of invited guests that GMAT doesn't have enough data to throw experimental questions at the beginning of the test. And towards the end, the GMAT is trying to fine-tune the algorithm so giving experimental questions at that stage doesn't help. I merely took that piece of information as the truth as it was given to me by the Chief Psychomatrician at GMAC at that time. Also, in your analysis you cannot infer the distribution of experimental questions in each quarter either. All we can infer from the ESR is the total number of questions that are counted towards your actual score and hence, the total number of experimental questions. For all that we know, such questions might be concentrated in the 2nd and 3rd quarters of the test.

Having said that, at CrackVerbal we work closely with GMAC to ensure that there is no incorrect information. I hope you agree that as a community it is our responsibility to educate GMAT test-takers, and not to point fingers at each other

Arun

PS: I have reported your post to GMAC asking for a specific clarification. I can update this post once I get some information.

In fact, "the distribution of experimental questions in each quarter" is exactly what we can infer. Because the ESRs are broken down by quarter, and experimental questions are not included in the results, we can conclude the exact number of experimental sections in each quarter of the test. Thus, we can state definitively that the experimental questions are evenly distributed, whether the GMAC admits this fact or not.

Sadly, the GMAC is not always a reliable source of information on the GMAT, but the release of the Enhanced Score Reports has served as a useful window into the structure of the test, whether intentional or not. For the most part, asking the GMAC for honest advice on the GMAT is like asking the IRS for honest advice on how to save money on your taxes. Given the hundreds of millions of dollars made by the GMAC each year, the organization lacks the motivation to be overly forthcoming with test-takers (and tutors) about the large number of experimental questions on each test, as well as the severity of the adaptive scoring algorithm.

To reiterate:

IR has 9 questions that count, and 3 experimental. We can tell that the IR section has exactly 9 non-experimental questions because the "percent correct" always aligns with a denominator of 9.

Quant has 28 questions that count, and 9 experimental. We can tell that the Quant quarters are split into 7 non-experimental questions each because the "percent correct" always aligns with a denominator of 7. The first 3 quarters of the Quant section have 2 experimental questions each, and the final quarter has 3 experimental questions (2/2/2/3).

Verbal has 30 questions that count, and 11 experimental. We can tell that the Verbal quarters are split into 7 non-experimental questions (2nd and 3rd quarters) or 8 non-experimental questions (1st and 4th quarters) each because the "percent correct" always aligns with a denominator of 7 or 8. The first quarter of the Verbal section has 2 experimental questions, and the final 3 quarters have 3 experimental questions each (2/3/3/3).

These findings have been supported by dabral and others.

Thanks! The analysis does make sense.

However, going back to the original point - the idea is not to mislead anyone as you stated in your earlier post

Arun