# LR Question - Feb 1992 LSAT (#4)

#### dta

##### LR Question - Feb 1992 LSAT (#4)
« on: March 19, 2004, 10:55:26 AM »
I'm taking old LSAT's in preparation for the June LSAT. I have a question about the following LR question from LSAT #4 administered on Feb 1992. The question is #23 in Section #1.

I got the question right by POE but i'm still not entirely sure I see what the flaw in the argument is. Could someone please provide a hypothetical situation that would serve as a counterexample to the argument and thus demonstrate that it is flawed? That is, present a scenario in which a random group of people is tested for cocaine use and it is NOT the case that the vast majority of those testing positive for cocaine use do in fact use cocaine?

#### jgomez

##### Re: LR Question - Feb 1992 LSAT (#4)
« Reply #1 on: March 19, 2004, 10:57:09 AM »
can you post the question and answers here.

#### dta

##### Re: LR Question - Feb 1992 LSAT (#4)
« Reply #2 on: March 19, 2004, 11:16:43 AM »
OK. I was a little nervous about doing so b/c of possible copyright restrictions or whatever. But here goes:

===========================
When 100 people who have not used cocaine are tested for cocaine use, on average only 5 will test positive. By contrast, of every 100 people who have used cocaine 99 will test positive. Thus, when a randomly chosen group of people is tested for cocaine use, the vast majority of those who test positive will be people who have used cocaine.

A reasoning error in the argument is that the argument

(C) fails to take into account what proportion of the population have used cocaine
===================================

As I stated, I don't dispute that this C is the right answer choice - i got it right through POE. But I honestly can't see what's wrong with the argument. Given the premises (5/100 non-cocaine users test +; 99/100 cocaine users test +) I don't see how to construct a hypothetical scenario that meets these premises but not the conclusion of the argument.

Thanks for any help.

#### xrayspec

##### Re: LR Question - Feb 1992 LSAT (#4)
« Reply #3 on: March 19, 2004, 11:25:36 AM »
This topic comes up in epidemiology wrt the accuracy of disease tests (eg HIV)

Suppose 2% of the population uses cocaine.

In a random sample of 5000 people, you'd expect to find 100 (2% x 5000) cocaine users and 4900 non-users.

99% of the cocaine users will test positive. 99% x 100 = 99 positives.
5% of the non-users will test positive. 5% x 4900 = 245 positives.

So in this case, the majority of people who test positive have NOT used cocaine.

#### dta

##### Re: LR Question - Feb 1992 LSAT (#4)
« Reply #4 on: March 19, 2004, 11:38:57 AM »
doh! it's so obvious now! thanks xray!!!