Let’s say you have a treatment T, and you apply it to a bunch os subjects S, and get some outcomes O.
The Null Hypothesis is that outcomes O were NOT caused by T, and would have happened to S regardless. In some medical experiments they actually do this with fake treatments which don’t do anything.
A Type-1 error rejects the null hypothesis when it is actually true, sometimes considered a false negative, that is you didn’t actually give the subject the treatment and their outcome was what you were looking for.
A Type-2 error accepts the null hypothesis when it is actually false, sometimes considered a false positive, that is you think the treatment didn’t work when you actually treated the subject.
Latest Answers