GMAT Strategy

Im wondering if there are potentially strategies one can use to help increase their score payout.

One thing that I'm postulating around is... if i'm behind time on the quant section and in the middle of the section. I think if you need to.. I think its better to guess on a PS then on DS (if your goal is to regain your time). I think its much harder to make an educated guess on a PS question then on a DS question, so its better to guess blindly on the PS then on the DS.

Has anyone experienced this thought process ?

Also, these experimental questions.... I'm guessing that you dont want to devote too much time to any one question, since there is that possibilty that it wont be counted (and your just wasted more then 2.5mins). Its bad to think about placement of experimental questions, but if GMAT wanted to get a more accurate reading on those Experimental questions, I dont think they would put them towards the end, where ppl might have to randomly guess on those questions due to time restrictions, artificially lowering the question level for GMAT ppl to analyse. Instead I would think that they would be more dispersed thruout the initial-middle stages rather then the end.

What I'm getting at is... it might be better to make a blind guess in the middle of the section (to regain your time), rather then on the final questions. Anyone have thoughts on this ?

Thanks

1. It's better to make guesses on stuff you can't solve, be it PS or DS. If you spend more than a minute and can't really seem to come up with at least a starting point for your solution, then just guess (as "educately" as possible) and move on. I say do not underestimate DS - it's as fierce as PS at least! If I had to choose between DS and PS, I'd say guess on DS because most of us just aren't that familiar on "in love" with the format. With PS, you can sometimes plug in some answer choices and see what you get, so that's already a starting point to help you with your guess.

2. Don't worry too much about experimental questions. As far as I know, they're quite evenly spread across the whole test, simply because the test makers want to get as much realistic data about how people perform (whether it's at the beginning, middle or end).
However, note that even if you set your goal to guess at the middle and not at the end, that's not always realistically possible: people tend to guess at the end because they run out of time (I'd say that's the no. 1 cause for guessing), which always occurs at the end.

And one final tip: don't waste precious time trying to figure out if a question is experimental or not, as far as you're concerned each question is important

Hi Dana,

Thanks for the reply.

What I meant between guessing on hard PS and guessing on hard DS.

So the "easy plug in" type of question I was not really thinking of. The way I was thinking is that a hard PS question may have answers that deviate only slightly (and its hard to realistically eliminate answer choices)... whereas a hard DS question, you may be able to easier narrow it down to "A/D" or "B/C/E" even if you know you dont understand the question, so your educated guess goes further in that regard..

I ranged from 620-650 on last 5 Prep Test (1 GMAT prep and 4 Manhattan preps), taken within the last 2 weeks, and I've always had to rush the ending on the quant, tho I'm use to that.

Duke -

A few things:

First you are correct and Dana is also correct -- if you are beyond the 1 minute or 1:15 point and you don't have a strategy then you should make an educated guess and move on and that is for both Data Suff and Problem Solving.

However, Data Suff is different in that it is designed to not be a choice between 5 answer choices but rather between 2 or 3 choices. This is because you can usually evaluate at least one of the statements in a data suff problem. For example, you are looking for a specific value for x and statement 2 says "x is positive." This is clearly not sufficient, so you can eliminate choices B and D even if you do not understand the question stem or statement 1.

For this reason, if you are behind on time and trying to blindly guess at one or two questions to get back on pace (and yes, the late middle of the test is the place to do this, not at the end) then you would guess at problem solving questions (without reading them - if you are trying to make up time) and not at Data Sufficiency. This is not only because DS can usually be quickly paired down to 2 or 3 choices, but also because DS is supposed to be more like 1:30 whereas PS is more like 2:30.

Your speculation on the placement of experimental questions that do not count toward your score is right on. In order to get good data on their questions, and in order to penalize you for poor time management, the last 7 or so questions are logically less likely to be experimental. Also, you are exactly right about not spending more than three minutes on a question - it may not count!

If I can indulge in quoting myself, here is something that I wrote earlier in the week that I think will specifically help you:

"Many people base their "strategies" on the fear of missing questions. This leads them to the following approach - when they encounter a strange and difficult problem they become focused and alert and put maximum time and attention into this problem. Then when they see a problem that they can get right they try to not only get it right but do so in the shortest possible time so that they can make up the time they just devoted to the really difficult problem.

Here is the flaw in that scenario...between experimental questions that don't count and the chance of randomly guessing the correct answer, a test taker has about a 1 in 3 chance that a blind guess will not count against her. Even if she is able to narrow the problem down to just 2 choices - and assuming that one of those two is the correct choice - her odds are barely better than 50%. In other words, by continuing to devote her time and attention to a problem that continued to stump her, this test taker used three minutes or more and barely improved her chances of getting that question right. (And do keep in mind the very real chance that even if she does get it right that it will be an experimental question and not count in her favor).

Now that she has used too much time on the question that she struggled with she needs to make up that time somewhere - so she will make it up on the next few questions by hurrying so that she can save 20 seconds a question. This is the worst possible thing that she could do - other than just ending the test immediately. GMAT questions are deliberately tricky and are designed to specifically funnel test takers to the wrong choice - even when they fully understand the problem! The test writers want you to try to hurry so that you fall through their trapdoors into incorrect answers - for example, by answering the length of one side rather than the perimeter of the equilateral triangle. Of course the questions are usually set up so that you have to solve for the incorrect choice first and then go through that to the correct choice. If you are trying to "save 20 seconds" then you walk into these traps.

The better way to approach the quant section is this:

When you see a question that you can get right - the devote the attention and the time that it takes to get it right. There is a difference between struggling for 3 minutes to come up with a strategy for a question you find very difficult (this is usually a mistake) and spending three minutes doing what it takes to ensure that you get a particular question (that you know you can solve) correct. It is a matter of approach to the test and it can make a huge difference. Do not try to steal 20 seconds from questions that you can get right only to hand that time over to a question that you will likely miss anyway.

Here are two things to think about:

1) If you do not have a specific strategy for solving a quant problem in about 1 minute and 15 seconds then you will want to choose from the viable answer choices and move on. It is precisely this type of question - where you have no strategy for a long time - that can sap the time and energy right out of your quant section on the test.

2) Set a time standard for yourself - I like to say that after question 25 I should have 25 minutes left (for the last 12 questions). If I am 5 minutes behind, I will just skip (randomly guess) at the next two problem solving questions I see. This will bring me back to the right pace. This is a better strategy than the alternative of " I can catch back up." Be careful here, trying to catch up usually means saving 20 or 30 seconds per question and we already know this is a mistake. Better to sacrifice two or three questions for the good of the test.

Finally, focus on getting questions right rather than on fearing that you might get some wrong. You will get questions wrong. But if you get say, 30 of the 37 right you will be back up to that 44 and this strategy will be one that gives you a consistent score rather than being subject to the swing you saw recently."

Good luck!

Hi David, EXCELLENT analysis.... something everyone should read.

I'm also glad that I capitalized on what I thought smart minds who make the GMAT would do. That is tend to not have much experimental questions in the final 7 questions of each of the quant and verbal section. Logically it just makes sense for them to not have them there, since its more likely that incorrect answers would be generated due to time limitation, and the level of difficultly paired with them would be lower then actual. That would entail an easier GMAT for test takers over time. The only way to eliminate the "error due to time limitations" from the equation, would be to take it out completely, and throw those experimental questions in the initial-middle stages of the section, when time typically is not a strong factor....

On my Preps test... I guessed towards the end... however, on the Actual GMAT I took the above theory into hand, I concentrated my guesses "due to time limitations" in the middle of the test... (on PS for quants and on RC for verbal) .... and I am quite certain that I scored well on the final 7 questions on each section (since I had sufficient time to work out/reason the answer)

For example a HARD PS (in the middle) could have answers like (if its a detailed circle/triangle/radius/perimeter combinated question)

a. 16
b. 16 + 3(root2)
c. 32 + (root2)
d. 16 + (root2)
e. 16(root2)

You likely cannot narrow down these answer choices until you actually get an answer. Tho you accept that 1) it will take me too long to calculate this 2) Its too difficult for me to get .... Therefore take your guess (1/5) instantly and save your time. Don't throw away good time on a bad question (for you).

Now on a HARD DS (in the middle) question... you could be given a combination of question stem and statement choices that you dont understand. But you can have a better chance of recognizing that even tho I dont understand how to get the complete answer.. I can elminate 2 or 3 answer choices in about 1min ... essentially grouping it into "A/D" or "B/C/E" based on maybe understanding 1 statement or the question stem alone... hence devoting that 1min to this question increased your odds to (1/2) [A/D] or (1/3) [B/C/E] ... even if you accept that the question is too hard for me.

I actually took my actual GMAT yesterday and scored what I expected to with only 10 weeks of study only. That should be sufficient for me to get into the B-School I applied to for this coming September (the rest of my application they have & its reasonably strong comparitively). They advised me to email them my unofficial score report to speed up any acceptance.

At the end of the day, utilizing smart strategies can only help maximize your potential scores whatever range it is in.

Thanks Duke, David. Both of you have put forward few good strategies for the GMAT takers (certainly for me).