What is the difference between "probability of H given E" as calculated by Bayes theorem and the "positive predictive value" of E?


AFAIK, those two are pretty unrelated.

Probability of H given E is simply the probability that the event H will happen when we already know that the event E happened.

On the other hand, when we do a statistical inference (that is, draw conclusions about population using a sample of it), we often want to know if something like the mean of a variable is zero.

We might do a test and come to the conclusion that it's indeed zero. However, we are never 100% sure, and we might have told that it's zero, but it's not, whenever this happens we have a false alarm. 1-(proportion of false alarms) gives you the positive predictive value, that is the proportion of times we don't have a false alarm in our test.


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.