[R-sig-eco] evaulating multiple responses using Chi-square

[Dixon, Philip M [STAT]]

Yes, the idea has been considered, by no less than RA Fisher.  If you want a name, it is sometimes known as Fisher's combined probability test.  Usually applied to one-sided tests of the same null hypothesis.   Yes, the p-values can come from different "sources".  Before using it in the community context, I would want to think carefully about what it tells me when applied to two-sided tests of different species.

Philip


	[[alternative HTML version deleted]]