Timeline for One-sided Fisher's exact test and its complement
Current License: CC BY-SA 3.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 2, 2011 at 4:38 | comment | added | Charlie | Ah, based upon the original question, I assumed that A was the outcome under control, say, and B was the outcome under the treatment regime. The update to the question suggests that this is not what the OP had in mind. | |
| Oct 1, 2011 at 22:25 | comment | added | Karl | yes, Fisher's exact test is a type of permutation test, but what does "A superior to B" have to do with Fisher's exact test? | |
| Oct 1, 2011 at 20:56 | comment | added | Charlie | A permutation test is one way to implement Fisher's exact test and I think that it's the easiest way to think about just what hypothesis that it's testing (i.e., that all the observations come from the same bag and that assignment to groups is just an arbitrary label). You could also use formulas to calculate the p-values directly without needing to perform the permutations, but I don't think that is as clear. | |
| Oct 1, 2011 at 19:09 | comment | added | Karl | here you seem to be describing a permutation test, but the question concerns "Fisher's exact test", though in that case the "A superior to B" part is confusing. | |
| Oct 1, 2011 at 18:17 | history | answered | Charlie | CC BY-SA 3.0 |