taking multiple choice tests cold

[Jason Strasser (strassa2)]

Ive had some experience in this area.

The first step is to find questions that give hints about other questions. Even in college you see this all the time.

The second step is to find spots where you have groups of questions where one is the right answer. You can have two questions and you KNOW the answer to one of them is C, answer C for both. Dont try to be a guessing hero.

Third is to rule out answers that are obsurd and run well flipping coins! Whenever you can reduce variance and find spots where you can answer sets of questions where you are guaranteed to get an amt of questions right that would beat randomly guessing.

[Dan.]

Does Duke have a lot of multiple choice tests? I haven't had one at Northwestern, nor has anyone I know.

[CalledDownLight]

Where can I find these classes? I hear EOS11 has them, anything else?

[T50_Omaha8]

"Never" means never right, and "always" means always wrong.

[tbrown932]

The second step is to find spots where you have groups of questions where one is the right answer. You can have two questions and you KNOW the answer to one of them is C, answer C for both. Dont try to be a guessing hero.

WRONG.

[TurtlePiss]

If you're making a guess (this could be after you've eliminated some candidates), make sure to make it totally random, don't go by feel. Partition the clock into appropriate sections and use the seconds hand to make your answer.

[TheDivineRod]

I think i read somewhere that the longer answers are more likely to be the correct ones.

[goofyballer]

[ QUOTE ]
The second step is to find spots where you have groups of questions where one is the right answer. You can have two questions and you KNOW the answer to one of them is C, answer C for both. Dont try to be a guessing hero.

[/ QUOTE ]

Isn't this just reducing variance by taking 50% on the two questions? If you're trying to be a "guessing hero" then it's probably because you feel a little more than 50% sure about which one's correct, in which case you'll experience more variance but get more questions right in the long run.

[HOWMANY]

[ QUOTE ]
[ QUOTE ]
The second step is to find spots where you have groups of questions where one is the right answer. You can have two questions and you KNOW the answer to one of them is C, answer C for both. Dont try to be a guessing hero.

[/ QUOTE ]

Isn't this just reducing variance by taking 50% on the two questions? If you're trying to be a "guessing hero" then it's probably because you feel a little more than 50% sure about which one's correct, in which case you'll experience more variance but get more questions right in the long run.

[/ QUOTE ]

There could be 2 questions with the same set of 4 answers for both and you know that one of the choices is correct for one of the questions but you're completely clueless on the other 3 choices so the best decision here is to fill in the one you know is the correct answer for one of the questions instead of trying to be a hero and guess which question the guaranteed answer is for and then picking from the other 3 and guessing which one it answers as well.

[Misfire]

I know on tests like the GRE/GMAT, you can pretty safely avoid answers on math questions with "easy" numbers. For instance, a problem dealing with numbers like 100 and 50 and 25, the right answer will not likely be a multiple of 25. On reading comprehension, you can often avoid answers with lots of language lifted from the passage. Not that these questions are hard anyway, but if you're short on time, knowing what not to guess helps.