[R-sig-eco] Pariwise tests in ANOSIM (and/or ADONIS)

[Pierre THIRIET]

Dear Jari,

from this perspective, is it correct to run multiple pairwise tests (1 
factor 2 levels) using adonis, and then adjust p-values. If yes, which 
adjustement method would you suggest, Bonferroni correction ?
Thanks in advance,
All the best,
Pierre

Le 05/06/2013 07:00, Jari Oksanen a écrit :
> On 05/06/2013, at 03:25 AM, Simon de Lestang wrote:
>
>> I have what I am sure is a very simple question relating to the package vegan, although I have not been able to solve it via Google.
>> Using the function anosim I am only able to produce a global result in the output.
>> Is it possible to produce all pairwise comparisons (corrected) in a single output or are these to be conducted individually? (individual pairwise tests).
>> I guess my question would also relate to the function adonis.
>
> Dear Simon,
>
> Pairwise tests are not possible in vegan. My understanding is that the non-R software with such tests makes separate pairwise tests using subsets of data with only two levels of a factor in one tests. We don't provide that in vegan and have no plans to provide this in the future.
>
> Cheers, Jari Oksanen