Balanced answers
- Thread starter SES
- Start date
Help Support Professional Engineer & PE Exam Forum:
snickerd3
Taking suggestions
false
Capt Worley PE
Run silent, run deep
Some people are full of it.
ptatohed
Licenced to Spell
Maybe I am missing something.
1.) Why is this relevant in the first place?
2.) But 2, if the answers (A, B, C, D) are random and each have equal opportunity of being the answer, why wouldn't the amount of A,B,C,D's be pretty close to balanced?
1.) Why is this relevant in the first place?
2.) But 2, if the answers (A, B, C, D) are random and each have equal opportunity of being the answer, why wouldn't the amount of A,B,C,D's be pretty close to balanced?
IlPadrino
Well-known member
I say "true" or very nearly so. Why not?
mrt406
Well-known member
- Joined
- Sep 29, 2009
- Messages
- 78
- Reaction score
- 6
If there were in infinite number of questions then this would be true. But I'm sure they don't go out of their way to make sure there's the exact same number of each answer.
Balanced? Probably. Perfectly balanced? Probably not.
Balanced? Probably. Perfectly balanced? Probably not.
civilized_naah
Well-known member
It is relevant because in your guessing period, if you do a count of those already answered (with a high degree of confidence, of course - otherwise this has no significance) and IF you observe that one of those choices (A,B,C,D) has a markedly lower frequency than the other three, then you are better off putting down that choice for the unanswered questions.
IlPadrino
Well-known member
mrt406
Well-known member
- Joined
- Sep 29, 2009
- Messages
- 78
- Reaction score
- 6
civilized_naah
Well-known member
IlPadrino
Well-known member
Now you're just talking out of your ***... 'cause I sure smell the ****! Ever heard of psychometrics?
mrt406
Well-known member
- Joined
- Sep 29, 2009
- Messages
- 78
- Reaction score
- 6
Ok, ok.... granted I'm not SURE of it. But I'm most nearly completely sure of it. I'm also not sure that a team of sealions wrote the test, or that the whole PE license/test/industry isn't a fabrication of my imagination.... but I'm pretty darn certain of it.
Yes, which is why I'm "sure" that they don't specifically make the exact same number of each answer choice. If that were the case, you could determine the answer to a question based upon the number of answer choices selected on the other questions, and not on the question itself. That would degrade the quality of the exam, which is the exact opposite of what psychometrics hopes to achieve.
That being said, my assumption is that the answers are randomized within the choices, with each having the same probability. As I said before, if you had an infinite number of questions then, yes, each choice would appear the same number of times. But there isn't. So you might get the same number of answer choices, you might not. I certainly wouldn't go into the exam expecting there to be.
Wouldn't this be true only if you were sure all of the other answers were correct? When you got to the end of the test and the answers were not balanced, how would you know which answers to change to balance it?
Last edited by a moderator:
Ble_PE
I need a beer.
Just pick C every time.
snickerd3
Taking suggestions
It was A when I took it. scared the crap out of me too. a run of 6-8 questions all with the same letter for the correct answer and I was certain of my selection too.
knight1fox3
Jedi MASTER & Friend of Capt. Solo
Capt Worley PE
Run silent, run deep
Choose E.
Don't be a slave to the machine.
Don't be a slave to the machine.
ptatohed
Licenced to Spell
So, let's say that among a cluster of questions that you felt really, really good about, the answer was B for each. Now, later in the exam when you are guessing..... do you choose B because B is the "common answer for this exam" or do you choose one of the other answers, because "all the B's have surely been used up"?
IlPadrino
Well-known member
But you don't know how they do it? There are certainly tests that use a balanced answer key - you could google this if you'd like to learn more.
