4 ms·
I just realised the OP probably used a Fisher test or something like it. The Bayesian approach isn't necessarily better then.
by ogogmad 4y ago
I just realised the OP probably used a Fisher test or something like it. The Bayesian approach isn't necessarily better then.
- drc500free 4y agoMy stats are pretty rusty, I checked if the two binomial distributions were different. No idea if that reduces down to a Fisher test!
- kuhewa 4y agoBarnards, which isn't as conservative as Fisher's exact. My understanding is Fisher's would be necessary if the number of positive cases was constrained (not applicable to this kind of work really but e.g. 600 years ago someone had ten eggs and distributed amongst the laypeople and the friars).