Proper Guessing

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

thegenius

Senior Wharf Rat
Lifetime Donor
20+ Year Member
Joined
Jul 12, 2004
Messages
5,668
Reaction score
5,418
If you are forced to guess on a bunch of multiple choice questions, does it matter what stragedy you choose to guess for getting the highest possible score?

Take 10 questions you must guess on:

One method is to randomly guess on each question. If you have to guess on a question with 4 answers, there is (presumably) a 1/4 chance of getting that one right. Questions are independent events, so getting all 10 questions correct means you have a (1/4)**10 probability (0.000095%).

You could also guess all A's, B's, etc. I don't know enough about how to calculate these type of statistics, but to me the former method is an equal probability to the latter method.

Can anyone elaborate on this if they have good science behind it? (maybe provide a website link.)
 
Seems like it'd be a good idea to just consistantly guess one thing...I don't want to work out the math though...it would probably be wrong anyway 🙂 If you guess randomly, like you said, you could easily get them all wrong. If you guessed consistantly you could feel confident knowing that at least one of the ten questions you guessed on was C, you know?

HOWEVER, I think it is EXTREMELY rare that you wouldn't be able to eliminate at least one, if not two answers...thus making a strategy of guessing one letter consistantly not really viable 🙂
 
The Princeton Review people claim that you should always guess the same letter, if you have to guess on a bunch. I'd go with C.
 
if the correct answer is randomly placed into A,B,C,D then it doesnt matter what you guess


see a basic statistics book if you need proof
 
I've thought about this actually (sad) and here's what I think. If the answers were random, it wouldn't matter if you pick the same letter every time or not. But they aren't random because you almost never see 3 or 4 of the same letter in a row. So if you pick C 4 times in a row, you know you didn't get them all right. But you probably got 1 of them right. Now, if you guess a different letter every time, you have a better chance to get them all right, but you also have a better chance to get them all wrong. I remember when I was in high school I was taking a multiple choice test with 75 questions on it, but you only had to get 60 right to get 100%. So after #60 I figured I had gotten them all right and I just guessed randomly on the last 15. I got every one of them wrong and only got 59 correct. Ok, so that was a random guessing story.
 
GRRRRRRRRRRRRRRRRRRRRRRRR Stop guessing. The MCAT Program office hates your guts when you guess correctly. RARRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRR
 
medicalstudent9 said:
if the correct answer is randomly placed into A,B,C,D then it doesnt matter what you guess

see a basic statistics book if you need proof

I believe you - i just don't have a stats book in front of me.

It's puzzling why so many people think that guessing all B's is the right way to go. It doesn't make any difference at all.

The idea is, over a span of 10 questions, you are bound to get 1 or 2 C's in there. So you are guaranteed of getting 1 or 2 C's. But you are also guaranteed then, of getting the remaining 8 or 9 wrong. 😀 😱
 
thegenius said:
I believe you - i just don't have a stats book in front of me.

It's puzzling why so many people think that guessing all B's is the right way to go. It doesn't make any difference at all.

The idea is, over a span of 10 questions, you are bound to get 1 or 2 C's in there. So you are guaranteed of getting 1 or 2 C's. But you are also guaranteed then, of getting the remaining 8 or 9 wrong. 😀 😱

Yeah, the Bs thing was a joke
 
Top