<ol>
<li>He shouldn't guess because he doesn't know what he's doing.</li>
<li>This is the same scenario as the first case. The bottom line is that the person doesn't understand the concepts being tested in the question. I don't care how much time a person spends with the question, there are way too many traps on the math section for you to be randomly guessing.</li>
</ol>
<p>Here's a scenario I've cooked up</p>
<p>You're on the last section of a given subject (we'll say math here, because of its finality). You have guessed on two problems, but later while taking the test (or while talking about the test on break), you realize that you got them wrong. However, you are extremely confident that you got all of the rest right. You come up on #16, and can only turn it into a 1/3 guess. Therefore there is a 2/3 chance that you will get it wrong, and lose an extra point that you would have otherwise gotten by rounding. On the other hand, there is a 1/3 chance that you will get it right and get all of your points. Logically, you should protect your rounding point and omit.</p>
<p>That said, I have never omitted a question in all of my standardized testing days. I want my bloody 2400 or I'm coming back anyways.</p>
<p>I only omit math questions that I have no clue how to do anyway.</p>
<p>To neethu16 </p>
<p>Exactly! There was this absolutely difficult talent test that I took. There were 90 questions. I worked out 48 questions and I made educated guesses on the remaining 42. (There are no penalties for guessing). My score was 48!!! All the questions that I had guessed on were wrong.</p>
<hr>
<p>Actually, I think it’d be very likely that within that 48 questions, you got a few wrong because you actually guessed a few right in the ones that you completely guessed but coincidently that overall score was the same as the number you believed you answered correctly. Ie: You got 3 wrong within the 48 questions you BELIEVED you answered correctly, but actually got wrong, and you answered 3 correctly in the 42 that you completely guessed on.</p>
<p>From the previous page, I never ever said that the monkey will always get 8 right. A 700 on the CR is neither impressive nor valid support for a clear misinterpretation.</p>
<p>People need to move on from the notion that this is somehow about “luck” or being bad at “guessing”. It’s pure statistics. If you honestly think your guess is worse than the probability suggests, TI83’s have a nice random integer function. For 50/50’s you might as well flip a coin.</p>
<p>
</p>
<p>Yes you do. The most likely scenario is simply the probability multiplied by the total. The deviation from there follows a very predictable curve.</p>
<p>I don’t think any of you understand probability.</p>
<p>Just because you have a 1/5 chance of getting the question right doesn’t mean that you will get 8 correct out of every 40 questions. You could easily get 7 or 9 wrong and it is even possible to get every one wrong or every one right.</p>
<p>Think about it, if you flipped 20 coins you won’t get 10 heads and 10 tails every time even though the probability of getting either one is 1/2.</p>
<p>So by guessing you are gambling, you might break even, you might gain points, and you might lose points.</p>
<p>So guessing gives you no statistical advantage, but there is a larger risk and there is a potential for greater gains.</p>
<p>So I have no real advice on whether to guess or not, I always guess personally, but it gives you no real advantage.</p>
<p>@pckeller</p>
<p>What are you talking about???</p>
<p>Guessing is not affected by how long you spend on a problem, and if you can’t do a problem, then you can’t do it and it doesn’t matter how many of the other problems you could answer.</p>
<p>It sounds like your reasoning is that they spent so much time on that problem that they need to answer to make up for the time spent. That logic makes no sense because no matter what, they lost that time and guessing gives no real advantage.</p>
<p>If you flipped 20 coins, the most likely scenario would be 10 heads and 10 tails. It’s a probability curve. I’ve said this once before, and this is both mathematically and statistically correct. It’s even possible to figure out the mathematics behind the probability of ANY scenario, but such probability would be LESS than that of 10.</p>
<p>It gives the monkey no statistical advantage. What pckeller is suggesting is that we are not monkeys - merely reading the question will allow us, as sentient beings, to make a more educated guess than a monkey or a random generator. This in my mind is common sense.</p>
<p>Actually because of the collegeboard’s traps we are less likely to get the right answer than a monkey after reading it if we guess randomly.</p>
<p>And I understand how probability works. I realize that the most likely situation is 8 correct out of 40, but that does not mean that you will get 8 correct every time. That is why I said that guessing is like gambling. It gives no statistical advantage, but has greater risk and has a potential for gain.</p>
<h2>Actually because of the collegeboard’s traps we are less likely to get the right answer than a monkey after reading it if we guess randomly.</h2>
<p>Well, you wouldn’t be guessing randomly than would you? If your best guess is indeed WORSE than a monkey’s best guess, than I’m out of words to say… That is, if that is truly the case for any particular person than like I pointed out earlier, they shouldn’t be expecting any sorts of high marks. If your not wanting or expecting high marks, omitting might certainly be the right path to follow.</p>
<p>Let me try another way to explain why I believe that you should answer the questions you spend time on…I doubt I will convince you, but I’ll take a shot.</p>
<p>I’ll start this way: if you are moving at a pace that matches your ability and score goal and you are working carefully, you should not have to guess more that two or three times ON THE WHOLE TEST! So all this talk about how many you are likely to get right if you guess 8 times is completely misguided. If you are guessing that many times, you have gotten in too deep. You need to do more preparation and then follow a more reasonable strategy.</p>
<p>But suppose you are well prepared and you follow a strategy that matches your reasonable score goal (say 50 - 100 points better than your last score). And two or three times, you come to a problem where you feel you have to guess. It is silly to believe that a well-prepared student has a worse chance than a monkey. Stay calm, read, think and then take your best shot.</p>
<p>I also like the earlier post that pointed out the effects of rounding on raw scores. Your first two wrong answers cost you no more than blanks. Then, the third one costs you. But 4, 5 and 6 are no charge. And if you are getting more than 6 wrong, you are going to fast and answering too many. You have to have a reasonable strategy when you take this test. And not every student on this site gets 700+ in math.</p>
<p>4.0 said:</p>
<p>“Guessing is not affected by how long you spend on a problem, and if you can’t do a problem, then you can’t do it and it doesn’t matter how many of the other problems you could answer.”</p>
<p>And you also said that no one here understands probability!</p>
<p>But if we are trying to decide whether guessing is a good risk, we have to apply conditional probability. Here’s what I mean: consider two test-takers. One has moved through the first 15 questions in a 20 question section and has answered nearly every one correctly. The other has answered about 2/3 of them correctly. Now they both come to #16 and they decide to guess. Do you believe that they have the same chance of guessing right?</p>
<p>OK, that’s my last comment on this thread! It’s a controversial issue…</p>
<p>the SAT is a REASONING TEST. This is because it is not trying to test anyone’s knowledge here, but rather their ability to reason. Reason which answers can be eliminated, which answers are ridiculous, reason which answers are the right ones without knowing the actual subject matter. That’s why they made it the way it is, as a Reasoning Test. Your knowledge on subjects come into play for the SUBJECT tests. If you know your math take the math subjects, literature if you know your books.</p>
<h2>l the SAT is a REASONING TEST. This is because it is not trying to test anyone’s knowledge here, but rather their ability to reason. Reason which answers can be eliminated, which answers are ridiculous, reason which answers are the right ones without knowing the actual subject matter. That’s why they made it the way it is, as a Reasoning Test. Your knowledge on subjects come into play for the SUBJECT tests. If you know your math take the math subjects, literature if you know your books. </h2>
<p>True…but relevant?</p>
<p>I just omit if I have absolutely no idea.</p>
<p>Sure – I could probably go tick for tack and start eliminating answers. Narrow it down a few. But, still. It eats away too much time. Not worth it, in my book.</p>
<p>However, I am not saiyng omit everything that racks your brain. But on the hard ones – you’re doomed. They are meant to be multiple step and multiple thought processes. If you think you can just jump and be like ‘HEY you look like a cute answer’ – it’s not going to work. Omit and go back and make sure your answers are in tip top shape.</p>
<p>Use POE. I don’t omit much, because if you can eliminate some you have a higher chance.</p>
<p>The only questions I omit is one in the math (There is always a tricky one at the end that I have no clue about, and rather not loose .25). </p>
<p>I’d say use POE and don’t omit too much, though sometimes you should.</p>
<p>everyday, plenty of kids get 27 correct and 27 wrong in the Math section out of the total of 54 questions. these kids guessed and did not omit answers according to the advice you give above. Obviously: i would rather have the score for 27 correct and 27 omits than the score for joe guesser who got 27 correct and 27 wrong (6.75 points deducted meaning raw score of 20.25 which rounded down is 20 questions correct). do your own math - would you rather not omit any questions and get a score based on 20 right or answer what you know + what you can narrow down by process of elimination and get a score based on 27 right - omit the ones you don’t know. as for lucky guessing, i guess the kids who get 27 wrong were not quite as lucky at guessing as your statistics would indicate. it happens everyday; students go into every test and get TONS of questions wrong. think again about lucky guessing. erase all the wrongs from students exams and almost everyone of them has a much higher score. Duh…</p>
<p>Omitting becomes useful when you’re shooting for a near perfect score, and wan’t to avoid being rounded down by having a raw score ending in .25</p>
<p>For instance, if there’s three questions you need to guess on in the whole math section.
You can afford to make an educated guess on 2. But, on that third question, if you guess and are wrong, you end up losing a whole extra point due to rounding.
Statistically sure, if you can eliminate 2-3 choices, chances are you’ll get 1-2 of those 3 guesses correct.</p>
<p>Omitting serves no useful purpose for the average SAT taker, as long as that average SAT taker has enough reasoning skill to eliminate a few wrong answers.</p>
<p>That’s what I do ^.^
Generally I leave the second to last or last blank.</p>