Say you are a math whiz and won't ever get a question wrong. But you aren't good enough to fly through the questions.
If you keep answering correctly, the questions keep getting harder/stay very hard. Which means, it takes longer to do every questions. Hence, you don't have time to do every question and are forced to guess at the end.
However, if you purposely answer the hard questions incorrectly at the beginning (through the first 10), then you make the CAT think that you aren't very smart. So now you can answer all the questions correctly.
In which of these scenarios will you end up losing more points? Will the guessing cost you more than the number of points you are gaining by answering the hard questions correctly,
or
will answering more mediumish questions correctly be better?
5-Day Free Trial
5-day free, full-access trial TTP
Available with Beat the GMAT members only code
MORE DETAILS
Is it good to purposely answer incorrectly at the beginning?
This topic has expert replies
-
Ashley@VeritasPrep
- GMAT Instructor
- Posts: 199
- Joined: Tue May 17, 2011 6:06 am
- Location: Cambridge, MA
- Thanked: 192 times
- Followed by:121 members
- GMAT Score:780
This is a clever question 
, but I DEFINITELY would not advise purposely answering incorrectly at the beginning. An overall point is that the test would be rather poorly designed indeed if it were wired to reward that strategy, whereas I think in reality the test is quite well designed to reward making good decisions. True, at times, to make a "good" decision is to make an educated guess on a problem and move on, rather than spending a disproportionate amount of time on it, but I think the test would be pretty whacked out if it were set up to reward people's deliberately sabotaging themselves early.
I also think that spending drastically more than 2 minutes on a question does not actually -- in most cases -- tremendously increase your chances of getting it right. Even the most difficult questions are written to be doable in 2 minutes on average by anyone who can wrap his head around them, so generally, spending much longer on a question may mean you're missing something.
My best advice would be to capitalize on the fact that the first ten questions likely WILL be more quickly answerable than the questions you will later get if you get all of those first ten right. If you can get those first ten questions done -- but correctly! -- in ten minutes (as opposed to twenty @ 2 min/question), for example, you'll have bought yourself about extra 22 seconds for every remaining problem, which is a pretty significant increase from the initial 2-minute timeframe.
After that, I would just keep checking your time as you go along, and adjust *as you go along* if the time elapsed begins to be more than 2x the number of questions you've answered.
Finally, guessing -- if you wind up doing so at the end -- at least gives you a 1 in 5 chance of getting a question right -- in the worst case scenario, that is... if you are left without time even to make educated guesses. Purposely answering incorrectly at the beginning, in contrast, means you have to spend time solving the questions in order to guarantee that you'll choose the incorrect answer -- and then absolutely not get any credit for it! Sounds terrible!
That's my two cents; others, naturally, feel free to weigh in!
, but I DEFINITELY would not advise purposely answering incorrectly at the beginning. An overall point is that the test would be rather poorly designed indeed if it were wired to reward that strategy, whereas I think in reality the test is quite well designed to reward making good decisions. True, at times, to make a "good" decision is to make an educated guess on a problem and move on, rather than spending a disproportionate amount of time on it, but I think the test would be pretty whacked out if it were set up to reward people's deliberately sabotaging themselves early. I also think that spending drastically more than 2 minutes on a question does not actually -- in most cases -- tremendously increase your chances of getting it right. Even the most difficult questions are written to be doable in 2 minutes on average by anyone who can wrap his head around them, so generally, spending much longer on a question may mean you're missing something.
My best advice would be to capitalize on the fact that the first ten questions likely WILL be more quickly answerable than the questions you will later get if you get all of those first ten right. If you can get those first ten questions done -- but correctly! -- in ten minutes (as opposed to twenty @ 2 min/question), for example, you'll have bought yourself about extra 22 seconds for every remaining problem, which is a pretty significant increase from the initial 2-minute timeframe.
After that, I would just keep checking your time as you go along, and adjust *as you go along* if the time elapsed begins to be more than 2x the number of questions you've answered.
Finally, guessing -- if you wind up doing so at the end -- at least gives you a 1 in 5 chance of getting a question right -- in the worst case scenario, that is... if you are left without time even to make educated guesses. Purposely answering incorrectly at the beginning, in contrast, means you have to spend time solving the questions in order to guarantee that you'll choose the incorrect answer -- and then absolutely not get any credit for it! Sounds terrible!
That's my two cents; others, naturally, feel free to weigh in!
Ashley Newman-Owens
GMAT Instructor
Veritas Prep
Post helpful? Mosey your cursor on over to that Thank button and click, please! I will bake you an imaginary cake.
GMAT Instructor
Veritas Prep
Post helpful? Mosey your cursor on over to that Thank button and click, please! I will bake you an imaginary cake.
-
David@VeritasPrep
- GMAT Instructor
- Posts: 2193
- Joined: Mon Feb 22, 2010 6:30 pm
- Location: Vermont and Boston, MA
- Thanked: 1186 times
- Followed by:512 members
- GMAT Score:770
I have to agree with Ashley.
I would not advise going to the extreme that some do which is to spend any amount of time to try to get the first 10 right, however, I cannot advise purposefully getting questions wrong.
If you think that a question is way too difficult and you would not get it right even with 3+ minutes then it is better to not even begin such a question.
I see your logic and I understand what you are saying about the questions getting harder, however you have to understand that missing harder questions will not hurt your score nearly as much as missing easier questions and the best way for this to work out for you is to do your best on every question that you can do and if a question completely stumps you do not, as Ashley mentioned, be afraid to guess and go.
Now doesn't it feel better to know that you won't have to intentionally answer incorrectly?
I would not advise going to the extreme that some do which is to spend any amount of time to try to get the first 10 right, however, I cannot advise purposefully getting questions wrong.
If you think that a question is way too difficult and you would not get it right even with 3+ minutes then it is better to not even begin such a question.
I see your logic and I understand what you are saying about the questions getting harder, however you have to understand that missing harder questions will not hurt your score nearly as much as missing easier questions and the best way for this to work out for you is to do your best on every question that you can do and if a question completely stumps you do not, as Ashley mentioned, be afraid to guess and go.
Now doesn't it feel better to know that you won't have to intentionally answer incorrectly?
-
nav!n
- Junior | Next Rank: 30 Posts
- Posts: 10
- Joined: Fri Apr 01, 2011 5:43 am
- Location: India
- Thanked: 2 times
- Followed by:2 members
This can be a critical reasoning question on Gmat 
IMO Depends what score are you aiming at quant , I think some one good in quant will not go with this strat
IMO Depends what score are you aiming at quant , I think some one good in quant will not go with this strat
Dear life , When I said "Can my day get any worse" It was a rhetorical question not a challenge!
Let's say you do answer the first 10 incorrectly. Now you're writing the quant out of fifty. Even if you get everything else right, you're limiting your score at the very outset.
Also, every other question will not be easy. Question 11 will be, then 12 will adapt to 11, and the last few questions will be set as a response to the 40-45 correct responses you just gave.
So no, I don't recommend it. If you're good at quant, focus and make the most of your advantage!
Also, every other question will not be easy. Question 11 will be, then 12 will adapt to 11, and the last few questions will be set as a response to the 40-45 correct responses you just gave.
So no, I don't recommend it. If you're good at quant, focus and make the most of your advantage!
GMAT/MBA Expert
-
Ian Stewart
- GMAT Instructor
- Posts: 2621
- Joined: Mon Jun 02, 2008 3:17 am
- Location: Montreal
- Thanked: 1090 times
- Followed by:355 members
- GMAT Score:780
If you're well above average in Quant, then the questions early in your test are much more likely to be easy/medium than the questions later in your test, assuming you proceed normally through the test. The key point about GMAT scoring, which David mentioned above, is this:
* it hurts you *a lot* to get a 200-level question wrong - then the algorithm starts to wonder if you're completely incompetent;
* it barely hurts you at all if you get an 800-level question wrong - then the algorithm only thinks you might not be an 800-level test-taker, but it can still think that you're a 770-level test taker.
If you answer the first 10 questions incorrectly, you'll be answering some 200-level questions wrong, and it will be pretty much impossible to get a respectable score. It's a terrible test-taking strategy. If you know you struggle with time, then you may (like almost all test takers) need to guess at some questions. You do not want to guess at randomly chosen questions -- that is, you don't want to guess at all the questions at the start of the test, or at all the questions at the end of the test. Instead you want to choose carefully which questions to guess at: ideally you guess at the questions which won't hurt you much if you guess wrong, and those are the really hard questions. You want to be sure you save time on those questions so you can use that time on questions you'll be able to answer.
And I wanted to reinforce Ashley's point above: hard GMAT questions are normally *not* more time-consuming than easy GMAT questions, at least once you see what the question is about. They don't make questions hard by asking you to multiply 8-digit numbers, or to complete a 40-step algebra problem. Instead, questions are hard because they are more abstract (letters instead of numbers), they include traps or exceptional cases that not all test takers will notice, or because they require you to apply concepts in unusual ways. If you see what to do on a real GMAT question, you'll normally be able to complete the problem within a minute or ninety seconds. If you can't see what to do, then all the time in the world might not help.
I think many test takers greatly overestimate the value of extra time (as do those prep books which suggest spending inordinate amounts of time on earlier questions - another terrible test-taking strategy). A study was done with the GRE Quant section, which is quite similar to the GMAT, where the same pool of test takers was given two GRE tests, one with normal time, and one with 50% extra time. With the extra time, scores rose by about 1% -- having extra time makes only a minuscule difference to your score. Having too little time, on the other hand, makes a huge difference - then you are forced to guess randomly. You want to avoid the situation where you have too little time at the end of the test, in which case you might be forced to guess at questions which are below your ability level, by saving time in the middle of the test by guessing at questions you can't see how to do reasonably quickly.
* it hurts you *a lot* to get a 200-level question wrong - then the algorithm starts to wonder if you're completely incompetent;
* it barely hurts you at all if you get an 800-level question wrong - then the algorithm only thinks you might not be an 800-level test-taker, but it can still think that you're a 770-level test taker.
If you answer the first 10 questions incorrectly, you'll be answering some 200-level questions wrong, and it will be pretty much impossible to get a respectable score. It's a terrible test-taking strategy. If you know you struggle with time, then you may (like almost all test takers) need to guess at some questions. You do not want to guess at randomly chosen questions -- that is, you don't want to guess at all the questions at the start of the test, or at all the questions at the end of the test. Instead you want to choose carefully which questions to guess at: ideally you guess at the questions which won't hurt you much if you guess wrong, and those are the really hard questions. You want to be sure you save time on those questions so you can use that time on questions you'll be able to answer.
And I wanted to reinforce Ashley's point above: hard GMAT questions are normally *not* more time-consuming than easy GMAT questions, at least once you see what the question is about. They don't make questions hard by asking you to multiply 8-digit numbers, or to complete a 40-step algebra problem. Instead, questions are hard because they are more abstract (letters instead of numbers), they include traps or exceptional cases that not all test takers will notice, or because they require you to apply concepts in unusual ways. If you see what to do on a real GMAT question, you'll normally be able to complete the problem within a minute or ninety seconds. If you can't see what to do, then all the time in the world might not help.
I think many test takers greatly overestimate the value of extra time (as do those prep books which suggest spending inordinate amounts of time on earlier questions - another terrible test-taking strategy). A study was done with the GRE Quant section, which is quite similar to the GMAT, where the same pool of test takers was given two GRE tests, one with normal time, and one with 50% extra time. With the extra time, scores rose by about 1% -- having extra time makes only a minuscule difference to your score. Having too little time, on the other hand, makes a huge difference - then you are forced to guess randomly. You want to avoid the situation where you have too little time at the end of the test, in which case you might be forced to guess at questions which are below your ability level, by saving time in the middle of the test by guessing at questions you can't see how to do reasonably quickly.
For online GMAT math tutoring, or to buy my higher-level Quant books and problem sets, contact me at ianstewartgmat at gmail.com
ianstewartgmat.com
ianstewartgmat.com
-
David@VeritasPrep
- GMAT Instructor
- Posts: 2193
- Joined: Mon Feb 22, 2010 6:30 pm
- Location: Vermont and Boston, MA
- Thanked: 1186 times
- Followed by:512 members
- GMAT Score:770
Thanks for making those points Ian and Ashely. This has turned into an informative discussion! I have been doing some research lately - not formal scientific yet, more anecdotal, but what seems more and more likely on the GMAT - let's say the Quant section for now, is this:
Imagine a graph where the x axis it the amount of time spent on a question and the y-axis the percentage of people who answer correctly. So we are looking at percent correct as a function of time spent on the question.
It seems more and more likely that this curve would have a bell shape, with only around 20% of people getting questions right in under 30 seconds since this is very likely to be pure guessing. from 30 seconds to 1 minute the percentage would increase as people who are very familiar with number properties get some questions right - particularly some Data Sufficiency.
from 1 minute to 1:30 the curve would rise the highest percent of correct answers coming in the period from say 1:30 to 2:30.
NOW is where this gets interesting. From my work with students I have noticed that as a student gets to more than 3 minutes on a question it becomes increasingly less likely that the student answers correctly! By the time you get to 5 minutes the odds of getting a question right are back down to just above the 20% mark that indicates pure guessing. This is because as Ian indicates GMAT questions are not difficult because they require time consuming calculations but because they have traps and require a proper strategy and 5 minutes is no more likely to yield a correct answer - one that uses a proper strategy and avoids the traps - than is 3 minutes or less.
To add what Ian has told us about the GRE seems to lead in this same direction with more scientific research.
It all adds up to a sort of approach where you take the time needed to get the questions right that you can do. (Saving 20 seconds only to miss a question that you SHOULD get right is a terrible error). And when you reach a question that is not yielding a strategy in a reasonable amount of time you really must move on. If you are guessing as to how to begin at 1 min or 1:30 you will still likely be guessing 4 minutes later.
Imagine a graph where the x axis it the amount of time spent on a question and the y-axis the percentage of people who answer correctly. So we are looking at percent correct as a function of time spent on the question.
It seems more and more likely that this curve would have a bell shape, with only around 20% of people getting questions right in under 30 seconds since this is very likely to be pure guessing. from 30 seconds to 1 minute the percentage would increase as people who are very familiar with number properties get some questions right - particularly some Data Sufficiency.
from 1 minute to 1:30 the curve would rise the highest percent of correct answers coming in the period from say 1:30 to 2:30.
NOW is where this gets interesting. From my work with students I have noticed that as a student gets to more than 3 minutes on a question it becomes increasingly less likely that the student answers correctly! By the time you get to 5 minutes the odds of getting a question right are back down to just above the 20% mark that indicates pure guessing. This is because as Ian indicates GMAT questions are not difficult because they require time consuming calculations but because they have traps and require a proper strategy and 5 minutes is no more likely to yield a correct answer - one that uses a proper strategy and avoids the traps - than is 3 minutes or less.
To add what Ian has told us about the GRE seems to lead in this same direction with more scientific research.
It all adds up to a sort of approach where you take the time needed to get the questions right that you can do. (Saving 20 seconds only to miss a question that you SHOULD get right is a terrible error). And when you reach a question that is not yielding a strategy in a reasonable amount of time you really must move on. If you are guessing as to how to begin at 1 min or 1:30 you will still likely be guessing 4 minutes later.
