-------- Original Message --------
Subject: Re: Procrustes distances from DF
Date: Sat, 24 Mar 2012 11:05:46 -0400
From: Aki Watanabe <[email protected]>
To: [email protected]
Dear Sämi,
Since you have four data points (i.e., four distances between equivalent
growth stages), which is not statistically powerful, I would use a
non-parametric test to test how significant your observed data is
relative to permuted data. I don't have too strong of a background in
statistics, but I would randomize the captive/wild assignment of each
data and run a regression analysis after each iteration on shape and
growth stages (although I'm not sure if regression analysis is
appropriate here since distances between subsequent growth stages are
not equivalent). This would be used to test whether the regression
coefficient (slope) that you obtain from the observed data is
significant compared to the distribution of permuted data. Another way
of checking for significance would be to use MANCOVA on wild/captive and
growth stages and see if both the species and growth stage terms are
significant. This would be a parametric test, so it does make certain
assumptions: homoscedasticity, normally-distributed error.
Again, I'm not an expert in statistics, so others may likely suggest
better ways.
Cheers,
Aki
On Thu, Mar 22, 2012 at 3:06 PM, morphmet
<[email protected]
<mailto:[email protected]>> wrote:
-------- Original Message --------
Subject: Procrustes distances from DF
Date: Thu, 22 Mar 2012 04:55:04 -0400
From: Wechsler, Samuel <[email protected]
<mailto:[email protected]>>
To: [email protected] <mailto:[email protected]>
<[email protected] <mailto:[email protected]>>
Dear morphometricians
I am so free to ask you another question. I have shape differences
(between captive and wild fish) at 4 different developmental time
points. The Discriminant function in MorphoJ tells me the Procrustes
distances between the mean shapes for each time point. So I have four
Procrustes distances which indicate how different the mean shapes of
captive and wild are at each time point. Is it legitimate to say that
the differences in mean between the captive and wild fish become
smaller
if the Procrustes Distances become smaller with time?
Cheers
Sämi
--
Aki Watanabe
Department of Biological Science
Florida State University
King Life Science Building
319 Stadium Drive
Tallahassee, FL 32306-4295
University of Chicago - AB '09
Biological Sciences and Geophysical Sciences
Website: http://sites.google.com/site/akinopteryx/home
Weblog: http://akiopteryx.blogspot.com/
