GMAT Strategy

800_or_bust wrote:The official GMAC prep exam questions are definitely tighter than most of the questions from non-GMAC sources. I went back and took a second look at that latest 800score exam. Here was question #5 (a medium difficulty SC question I had missed):

Bruce Springsteen and Meryl Streep are eligible for retirement benefits now, joining the ranks of other celebrities like Neil Diamond and Dustin Hoffman, both having turned 65 this year.

(A) joining the ranks of other celebrities like Neil Diamond and Dustin Hoffman, both having turned 65 this year.

(B) both having turned 65 this year, and join the ranks of other celebrities such as Neil Diamond and Dustin Hoffman.

(C) both turned 65 this year, joining the ranks of other celebrities like Neil Diamond and Dustin Hoffman.

(D) becoming 65 this year, joining the ranks of other celebrities such as Neil Diamond and Dustin Hoffman.

(E) and, because they have turned 65, they will have joined the ranks of other celebrities like Neil Diamond and Dustin Hoffman.

The credited answer is (B). I selected (A). My problem with (B) is the modifier appears to be misplaced - it looks like its modifying retirement benefits instead of the subject, and also the use of the conjunction "and" seems weird here. I would think a participial phrase would be more appropriate. (A) results in a weird sentence - with the "both having turned 65" modifying Neil Diamond and Dustin Hoffman. Grammatically, however, it seems rock solid except for the use of "like" instead of "such as." I guess I should have just keyed in on that minor detail.

Any insight on how the unscored experimental questions affect the scoring algorithm? Are the experimental questions that you receive based on your CAT performance? Or can you be given an unscored question of any difficulty level at any time in the exam? I assume for somebody scoring in the upper 40s to around 50, any experimental can be no more difficult than the actual questions.

DavidG@VeritasPrep wrote:

Any insight on how the unscored experimental questions affect the scoring algorithm? Are the experimental questions that you receive based on your CAT performance? Or can you be given an unscored question of any difficulty level at any time in the exam? I assume for somebody scoring in the upper 40s to around 50, any experimental can be no more difficult than the actual questions.

The experimental questions, to my knowledge, will not be impacted by how well you're doing. The whole point of including them is for GMAC to acquire data about what % of test-takers are likely to answer the question correctly. This way they can properly rate the difficulty level of the question once it becomes part of the exam.

800_or_bust wrote:

DavidG@VeritasPrep wrote:The experimental questions, to my knowledge, will not be impacted by how well you're doing. The whole point of including them is for GMAC to acquire data about what % of test-takers are likely to answer the question correctly. This way they can properly rate the difficulty level of the question once it becomes part of the exam.

Interesting. Might explain why some people - not all, but some - tend to perform better on the actual examination than the practice exams. Reading some debriefs, it seems this is fairly common for higher scorers, i.e. those around 700+. If the issue is timing, in theory, you might expect that with some easier experimental questions thrown into the mix, you'd have more time to devote to the harder problems. That is, if you take about 2 minutes or so per question on average for harder quant questions but only 1-1:30 on easier ones, the CATs that don't simulate experimental questions might tend to exacerbate timing issues.

Marty Murray wrote:

800_or_bust wrote:

DavidG@VeritasPrep wrote:The experimental questions, to my knowledge, will not be impacted by how well you're doing. The whole point of including them is for GMAC to acquire data about what % of test-takers are likely to answer the question correctly. This way they can properly rate the difficulty level of the question once it becomes part of the exam.

Interesting. Might explain why some people - not all, but some - tend to perform better on the actual examination than the practice exams. Reading some debriefs, it seems this is fairly common for higher scorers, i.e. those around 700+. If the issue is timing, in theory, you might expect that with some easier experimental questions thrown into the mix, you'd have more time to devote to the harder problems. That is, if you take about 2 minutes or so per question on average for harder quant questions but only 1-1:30 on easier ones, the CATs that don't simulate experimental questions might tend to exacerbate timing issues.

That's really interesting.

The first time I took the test, I think that one or two quant questions I got were probably experimental as they were weird and, and maybe I should not even have bothered working on them. That experience provides a good reason to guess now and then no matter how good at quant you feel you are, because an experimental question might be so weird that it won't even make it onto the actual test ever. I got one that used set notation in a way that I have not seen since high school or college, and never in any GMAT question before or since, and another used infinity or something related to infinity, and I felt that the answer was debatable. Maybe those questions were legit, but they seemed funny to me.

The second time I took the GMAT, the quant questions seemed so easy that I was actually thrown off a little. That perception was partly due to more preparation, but maybe also I just got some easy experimental ones too.

800_or_bust wrote:So I completed my official GMAT this morning. Finished with a 780 (51Q/47V/8IR). The proctor stated that tied the highest score she had ever seen on the GMAT.

But I was a little disappointed I wasn't able to get over the 800 hump, especially since my last practice exam was an 800 (51Q/51V/8IR).

It's worth noting the 51Q was only a 97th percentile. Has 51Q dropped from 98th percentile to 97th percentile? This is the first time I've ever seen a 51Q be anything other than a 98th percentile.

GMAC really should consider switching things up a bit to stem the steady rise in quant scores. Quant was too easy. I finished the section with 10-15 minutes left.

Marty Murray wrote:

800_or_bust wrote:So I completed my official GMAT this morning. Finished with a 780 (51Q/47V/8IR). The proctor stated that tied the highest score she had ever seen on the GMAT.

Nice work.

But I was a little disappointed I wasn't able to get over the 800 hump, especially since my last practice exam was an 800 (51Q/51V/8IR).

Man, I guess one has to be reeeaaally on and reeeaaally careful to score V51, partly because occasionally the right answers to the questions can be a little debatable.

Maybe if you were vegan, assuming that you are not, you would have had the edge to get a couple more verbal right answers and hit 800. Maybe I should have suggested that before ...

Anyway, at least something about the test was challenging for you.

It's worth noting the 51Q was only a 97th percentile. Has 51Q dropped from 98th percentile to 97th percentile? This is the first time I've ever seen a 51Q be anything other than a 98th percentile.

Q51 dropped to 97th percentile some years ago.

GMAC really should consider switching things up a bit to stem the steady rise in quant scores. Quant was too easy. I finished the section with 10-15 minutes left.

If I have it right, the distribution of quant scores is fairly normal in a lot of places, but in China people take the test and memorize the questions, and then provide the questions to test takers. So Chinese test takers who have already seen the questions before they take the test bump up the average quant score, and making the quant section more difficult would not do much to solve that.

Anyway, I guess you had some fun playing the GMAT game for a few months and scored high enough to get into any school you want to. It's been fun having you around and reading some of the answers you posted. All the best!

diegocml wrote:Man, you've nailed it! Congratulations! Wish I had your brain while taking the GMAT!

Marty, what's up with being vegan and scoring high on verbal? What's the link?!

Also, this Chinese thing, it seems like Suits to me. Nuts and messed up!

Congrats once again 800_or_bust!

diegocml wrote:Man, you've nailed it! Congratulations! Wish I had your brain while taking the GMAT!

Marty, what's up with being vegan and scoring high on verbal? What's the link?!

800_or_bust wrote:So I completed my official GMAT this morning. Finished with a 780 (51Q/47V/8IR). The proctor stated that tied the highest score she had ever seen on the GMAT. But I was a little disappointed I wasn't able to get over the 800 hump, especially since my last practice exam was an 800 (51Q/51V/8IR). It's worth noting the 51Q was only a 97th percentile. Has 51Q dropped from 98th percentile to 97th percentile? This is the first time I've ever seen a 51Q be anything other than a 98th percentile. GMAC really should consider switching things up a bit to stem the steady rise in quant scores. Quant was too easy. I finished the section with 10-15 minutes left.