-------- Original Message --------
Subject: RE: Multiple group comparisons with EDMA
Date: Thu, 25 Aug 2011 22:11:35 -0400
From: F. James Rohlf <[email protected]>
Reply-To: [email protected]
Organization: Stony Brook University
To: [email protected]
One reason the Bonferroni correction is popular (despite being overly
conservative in applications such as when making all possible pairwise
tests) is that it is simple to apply. You need only multiply the raw
probability values by the number of tests you are making. Equivalently,
you could divide your desired type I error rate, e.g., 0.05, by the
number of tests you are making. It does not matter whether the tests are
based on EDMA or some more conventional statistical method.
----------------------
F. James Rohlf, John S. Toll Professor
Dept. Ecology and Evolution, Stony Brook University, NY 11794-5245
Please consider the environment before printing this email
-----Original Message-----
From: morphmet [mailto:[email protected]]
Sent: Thursday, August 25, 2011 3:40 PM
To: morphmet
Subject: Multiple group comparisons with EDMA
-------- Original Message --------
Subject: Multiple group comparisons with EDMA
Date: Thu, 25 Aug 2011 15:17:50 -0400
From: Murat Maga <[email protected]>
Organization: University of Washington
To: <[email protected]>
Hi all,
I have 4 groups of samples that I want to run EDMA with all possible
combinations. I do bootstrapping to obtain CI for pairwise distances. Is there
an equivalent of Bonferri correction to maintain the significance level?
Thanks,
M
