GMAT Focus Algorithm - importance of the first 6 Questions is GONE

I don't have an exact timeline yet. Hoping to have them available by the time the new tests are live in November.

The current version of GMAT Club Tests will remain available as long as the Classic GMAT is available. I don't think it makes sense to shut them down before that and we will probably keep them live for 3-6 months after the end as well so people can check their progress, mistakes, etc.

Thank you for your response. I appreciate it.

Should I still study geometry to stay on the safe side then? Or can the gmat website help regarding the same?

Posted from my mobile device

Thanks for the analysis, bb. Based on this data you have derived from the official GMAT Focus practice exams, could it be that the GMAT Focus is / will be section adaptive like the GRE, instead of question-level adaptive like the current, "classic" GMAT?

In other words: perhaps the exam's difficulty adjustments are not made until after the 1st and 2nd sections—based on your overall performance on those sections, and that as a result, your performance only affects the difficulty of the next section(s)—which (the second part only) is something interesting and new that GMAC has in fact already revealed about the new exam?

Given the additional fact that you can now mark up to 3 questions per section for later review—and possibly even change your answers to those questions—it would seem impossible to maintain the exact same question-level adaptive nature of the test, anyway.

Of course, there is also a third possibility, which is that the new GMAT Focus uses a hybrid of the question adaptive and section-level adaptive scoring systems: it might only make adjustments after the first half of questions in each section (Q/V/DI), for example. In this case, however, there is still a chance that difficulty adjustments would be made based on answers that are later changed by the student—which is illogical and unfair.

If the scoring / difficulty algorithm has in fact changed, then this would of course delineate a marked and important change from the question-adaptive algorithm of the current GMAT, where adjustments to the test's difficulty levels are made after each and every response on the Quant and Verbal sections (except for the 9 "pretest" / unscored experimental questions: 3 on Quant and 6 on Verbal).

The IR section (3 pretest questions) on the current exam is not adaptive, but GMAC has indicated that the Focus's new Data Insights section (IR + DS) is in fact adaptive. The question remains, however: does this mean question-level adaptive, or section-level adaptive?

To state the obvious: the question-level adaptive scoring algorithm currently used on Verbal and Quant sections of the GMAT is also the most logical explanation for the current "front-loaded" nature of the Quant and (to a lesser extent) Verbal scoring systems—where questions at the beginning count more than those at the end. If the scoring algorithm has changed for the GMAT Focus, then so must our strategies to succeed on the GMAT, especially those related to pacing / time management.

Anyway, very interesting take, and I look forward to learning more.

The red part is something of SUPER important. This could open up the gate to an entire new world in terms of strategy, approach, time management.............🤔

I think it’s still too early to tell was just one data point of a single test and limited difficulty observations. I’m planning to take some verbal tests during this weekend, anyone else interested or board, feel free to join me 😉. Just send me a PM.

Posted from my mobile device

I have just finished a second quant test where I have answered the first 15 questions correctly and the last 6 questions incorrectly.

Result: Q80, 66th Percentile, or about Q48

There is basically little to no difference in scoring and placement of the questions which does bring a big question - is the adaptability built differently OR is it due to the very small question pool?
As a matter of fact, my score is lower after missing the last 6 than missing the first 6.
Very interesting (said nobody )

If the evidence from the official GMAT Focus practice exams continues to point us toward the conclusion that the GMAT Focus is in fact no longer question-level adaptive, but instead, section adaptive, then I think that there are indeed many of us—GMAT tutors and students alike—who will (and should!) find this news not only highly interesting, but highly relevant as well.

This would also suggest, for example, that test-takers will no longer be able to leverage the GMAT scoring algorithm, in particular on the GMAT Focus Quant section, by "front-loading" their performances: performing well on the first 2/3 - 3/4 of questions, and then simply surviving the last 1/3 - 1/4 of questions, while also making sure to confirm all their answers in time, to avoid the penalty for not finishing.

If that's the case, having 3 edits to change any answer from the section makes more sense now.

Also, bb when can we expect IR questions in the new Forum Quiz?

Exactly. How could the GMAT Focus be question-level adaptive anymore, when you can now go back and change up to 3 answers at the end?

For now at least, the logical conclusion is that it isn't, and that it's now section-level adaptive like the GRE: except unlike the GRE, your overall performance on the first and second sections of the GMAT Focus supposedly affects the initial difficulty level of subsequent sections (V/Q/DI).

make sense this ........but is too early to draw conclusions though!

The main reason why missing early questions on the classic GMAT is so devastating is that the early questions are medium-level questions, and by missing them, you drive the difficulty down to easy, thus hurting your score.

In contrast, in this case, your second, fourth, and sixth questions were hard questions, even though you were missing questions. Missing hard questions would not be expected to wreck your score.

What I'd like to see is what happens if someone misses 10 or more questions in a row on the Focus Edition. Does the difficulty decrease significantly? Seeing whether it does, we should get a clearer understanding of whether the Focus Edition is question adaptive.

Thanks Marty - I will play with this scenario. Missing 10 is a lot! but worth it for the experiment.

Here is another interesting find I discovered yesterday - I left 2 questions unanswered at the end of the test (I wanted to do all others correctly but somehow managed to miss one, #13 out of 21). Here is the result:



I am expecting and hoping nobody will be in a situation where they will be leaving questions unanswered but it seems the penalty has become way more severe for leaving questions unanswered. I mean 52nd percentile! with only 1 mistake and 2 unanswered

Wow. That's brutal.

You did better missing 6 in a row in your other examples.

I guess nobody will be leaving questions unanswered. Apparently, doing so is no longer a viable option.

Gmat Focus Official Mock 2,

Quant - All correct,
Verbal - 6 incorrect (4 CR and 2 RC),
Data Insights - 6 incorrect and 1 unattempted.

Would like to hear others points on this performance : bb , MartyMurray any thoughts?

Hi. THanks for sharing your mock test results! Can you tell me which question numbers you got wrong in Verbal?
At the same time, not sure I have a lot to share in terms of an opinion on your score besidees that it looks great! I think more on the DI would be nice. I do feel some things are still a bit unpolished in the test scores, something I need to research a bit more before sharing and potentially making a lot of stupid mistakes and misleading others