0
$\begingroup$

Question: One in two hundred people in a population have a particular disease. A diagnosis test gives a false positive $3$% of the time, and a false negative $2$% of the time. Ross takes the test and the report comes positive. Find the probability that Ross has the disease.

What I solved: Assume the probability that a person has disease is $\frac{1}{200}$. Then, we have the following probabilities:

  • person has disease and test is negative: $\frac{2}{100}$
  • person does not have disease and test is positive: $\frac{3}{100}$
  • person has disease and test is positive: $\frac{98}{100}$
  • test is positive: $\frac{98}{100}+\frac{3}{100}=\frac{101}{100}$

I am not sure about the above line the probability is $\frac{101}{100}$ I think this is not correct can any one tell me where I went wrong.

$\endgroup$
  • $\begingroup$ It is hard to follow your work, you string the sentences together so we can't see where one equation ends and the next begins. I could try to edit it for you, but I fear that I'd introduce errors. Can you clarify? Insert Line breaks in between each equation. $\endgroup$ – lulu Aug 8 '17 at 18:39
  • 1
    $\begingroup$ Probability never gets over 1, for anything - ever. $\endgroup$ – stretch Aug 8 '17 at 18:48
  • $\begingroup$ You must have missed the class where Baye's Theorem was discussed, and not done any reading. You can find how to approach problems like this by looking at the "Related" Q & A. $\endgroup$ – stretch Aug 8 '17 at 19:00
  • $\begingroup$ Related. perhaps a duplicate: math.stackexchange.com/questions/2279851/… $\endgroup$ – Ethan Bolker Aug 8 '17 at 19:13
0
$\begingroup$

The probability that the person has the disease and the test is negative is not $2/100$, but $1/200\cdot 2/100 = 1/10000$.

The likelihood that the person has the disease and the test is positive is $2/100\cdot 98/100 = 196/10000 = 0.0196$, not $98/100$.

The likelihood that the person does not have the disease and test is positive is $199/200\cdot 3/100 = 597/20000 = 0.02985$.

So the likelihood of having a positive test is then $196/10000 + 597/20000 = 9989/20000 = 0.04945$.

I guess Bayes' theorem comes in when computing the likelihood that a person with a positive test actually has the disease, although that was not stated.

$\endgroup$

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.