I am just learning probability in my Discrete Structures class and am very lost. This is the example given in the book and I have no idea how to solve this problem.

Problem: Suppose one in 1000 people have a certain disease. Suppose medical testing is not perfect (as is the case in real life) and consequently, only 99% of the people with the disease are tested positive. Suppose 2% of the people who don’t have the disease also test positive. What is the probability of actually having the disease, given someone tests positive?

Things I do know:

  • People with disease = 1/1000
  • Tested positive with disease = 99/100
  • People tested positive who don't have disease = 2/100

I'm not sure where to got from here with the information that I know. What are the next steps?

  • $\begingroup$ You want to use Bayes' theorem: $P(A|B) = \frac{P(B|A)P(A)}{P(B)}$ $\endgroup$ – nomen Dec 6 '13 at 5:20

As nomen pointed out, you need Bayes' Theorem here but you will find it helpful the expand the denominator as follows

$P(A|B) = \frac{P(B|A)P(A)}{P(B|A)P(A) + P(B|\bar{A})P(\bar{A})}$

where the bars indicate complements. In particular, let A be having the disease and B be testing positive. Then

$P(B|A) = \frac{99}{100}$

$P(A) = \frac{1}{1000}$

$P(B|\bar{A}) = \frac{2}{100}$

$P(\bar{A}) = 1 - P(A)$

Plug and chug from here.

  • $\begingroup$ Thanks for the help. So I used what you gave me to get (0.99 * 0.001) / ((0.99*0.001) + (.999*.02)) = .04721 Sorry I don't know how to format the math problem. Now I am working on this problem: Redo the Medical Testing problem, as follows. Suppose Pr(Dis) = x, with everything else being the same. For what value of x is it the case that Pr(Dis | Pos) is 0.5? $\endgroup$ – ztoolson Dec 10 '13 at 3:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.