-------- Original Message --------
Subject: Re: Canonical variates from first PCs of GPA residuals
Date: Fri, 13 Feb 2009 12:57:11 -0800 (PST)
From: Dennis E. Slice <[email protected]>
To: [email protected]
References: <[email protected]>
morphmet wrote:
...
Thanks for your interesting points, Dennis.
Just a quick question on within-group PCs and a comment on Dennis' point on
resampling.
You can do something, perhaps, with PCs from a single-group PCA, or
use within-group PCs for dimension reduction and still examine between
group differences.
Interesting. The question then is: how will you do the comparison by using
"within-group PCs" so that specimens are in the same shape space?
See the most recent posting.
...
I meant something like a test for mean shape differences using Procrustes
distances and permutations. No data reduction needed there.
True enough. My earlier comments were about statistical testing in a
dimensionally reduced PC space constructed from the overall data set
containing group differences to be tested.
Though I have done most of the typing, I should point out that this
problem was pointed out to me in a note from Fred Bookstein. Errors made
in attempting to elucidate it are clearly mine.
-ds
Cheers
Andrea
--
Dennis E. Slice
Associate Professor
Dept. of Scientific Computing
Florida State University
Dirac Science Library
Tallahassee, FL 32306-4120
-
Guest Professor
Department of Anthropology
University of Vienna
========================================================
--
Replies will be sent to the list.
For more information visit http://www.morphometrics.org
