Dear list: I am analyzing community data in which we have 6 samples per date across 6 dates per year at 4 sites over 3 years. Is it appropriate to permute across all rows in such an analysis? Or should I be restricting the permutation in some way (i.e. block by date)? I realize that this might be my call, but I was hoping to get some advice, philosophical or otherwise. I am new to mvabund and have done some research but there are still few examples out there. Yours, Jon.
--
Jon Benstead
Professor
Department of Biological Sciences
University of Alabama, Box 870344
2099 Bevill Building, 201 7th Ave.
Tuscaloosa, Alabama 35487-0344