Page 1 of 1

basic math question

Posted: Sun May 30, 2010 12:02 am
by handlebutton
If top 15% gets A for Class 1
If top 15% gets A for Class 2
Then if student gets two A's, what % of the class is he in - for sure?
Is it 7.5% or something like that? Sorry, I majored in history and I'm not sure how this works.

Re: basic math question

Posted: Sun May 30, 2010 12:07 am
by Mrs. Jack Donaghy
is said student only taking those two classes?

Re: basic math question

Posted: Sun May 30, 2010 12:12 am
by handlebutton
I'm trying to answer this question for 4 law school classes, of course, but I'm reducing n to 2 so I can get an answer to the question.

Re: basic math question

Posted: Sun May 30, 2010 12:14 am
by Leeroy Jenkins
top 15%

Re: basic math question

Posted: Sun May 30, 2010 12:16 am
by handlebutton
Top 15% is the only "for sure" answer, I think. The person could be the top-ranked kid in the two classes if no other student got both A's, so he could be top 1% or whatever, but the worst he could be is top 15%

Re: basic math question

Posted: Sun May 30, 2010 12:32 am
by Mrs. Jack Donaghy
handlebutton wrote:Top 15% is the only "for sure" answer, I think. The person could be the top-ranked kid in the two classes if no other student got both A's, so he could be top 1% or whatever, but the worst he could be is top 15%
titcr.

Re: basic math question

Posted: Sun May 30, 2010 12:38 am
by thesealocust
edited / never mind

Re: basic math question

Posted: Tue Jun 01, 2010 6:54 am
by johnstuartmill
handlebutton wrote:I'm trying to answer this question for 4 law school classes, of course, but I'm reducing n to 2 so I can get an answer to the question.
The answer for four classes is also 15% (assuming that the 'A' cutoff in those classes the same). After all, you can't rule out the possibility that everyone who gets an A in one class also gets one in all the others, right? But that's pretty unlikely -- four A's would probably put a person much higher than that. We just can't say anything with any certainty except for that 15%.

Re: basic math question

Posted: Tue Jun 01, 2010 9:02 am
by tomwelling
johnstuartmill wrote:
handlebutton wrote:I'm trying to answer this question for 4 law school classes, of course, but I'm reducing n to 2 so I can get an answer to the question.
The answer for four classes is also 15% (assuming that the 'A' cutoff in those classes the same). After all, you can't rule out the possibility that everyone who gets an A in one class also gets one in all the others, right? But that's pretty unlikely -- four A's would probably put a person much higher than that. We just can't say anything with any certainty except for that 15%.
Top 15% is the only thing we can say with complete certainty, but it is not a likely scenario. The probability is much greater that student would be higher; I am too far removed from my last math class to do any adequate analysis on what the percentile with the highest probability would be.

Re: basic math question

Posted: Tue Jun 01, 2010 4:10 pm
by johnstuartmill
tomwelling wrote:
johnstuartmill wrote:
handlebutton wrote:I'm trying to answer this question for 4 law school classes, of course, but I'm reducing n to 2 so I can get an answer to the question.
The answer for four classes is also 15% (assuming that the 'A' cutoff in those classes the same). After all, you can't rule out the possibility that everyone who gets an A in one class also gets one in all the others, right? But that's pretty unlikely -- four A's would probably put a person much higher than that. We just can't say anything with any certainty except for that 15%.
Top 15% is the only thing we can say with complete certainty, but it is not a likely scenario. The probability is much greater that student would be higher; I am too far removed from my last math class to do any adequate analysis on what the percentile with the highest probability would be.
There's no way to answer that question without more data than is in the OP.

Re: basic math question

Posted: Mon Jun 07, 2010 1:39 am
by Tautology
If law students' grades in every class were completely independent (0 correlation), you would expect about 15% of 15%, or 2.25% (if my mental math doesn't suck) of students to have A's in both classes in the example given. However, grades are almost certainly not independent; some students tend to do well in a number of classes and some students tend to do poorly in most. As everyone else has said, it's always possible that all 15% from the first class get A's in the second and so you can never be sure that you're doing any better than 15% (correlation of 1). Without more information, there's really no way to know where you fall in between those two numbers (there is always the possibility that there is a negative correlation, and so fewer than 2.25% would have A's in both classes, but this seems quite unlikely).