-------- Original Message --------
From: - Fri Mar 6 08:42:53 2009
X-Account-Key: account4
X-UIDL: UID21014-1098736360
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00000000
X-Mozilla-Keys:
Return-Path: <[email protected]>
Received: from pro14.abac.com (localhost. [127.0.0.1]) by
pro14.abac.com (8.14.3/8.14.3) with ESMTP id n26CgZX3013922 for
<[email protected]>; Fri, 6 Mar 2009 04:42:36 -0800
(PST) (envelope-from [email protected])
Received: (from dsl...@localhost) by pro14.abac.com
(8.14.3/8.14.3/Submit) id n26CgZ7R013919 for
[email protected]; Fri, 6 Mar 2009 04:42:35 -0800
(PST) (envelope-from dslice)
Date: Fri, 6 Mar 2009 04:42:35 -0800 (PST)
X-From_: [email protected] Fri Mar 6 04:42:35 2009
Received: from mx3.abac.com (mx3.abac.com [216.55.191.209]) by
pro14.abac.com (8.14.3/8.14.3) with ESMTP id n26CgYpb013894 for
<[email protected]>; Fri, 6 Mar 2009 04:42:35 -0800 (PST)
(envelope-from [email protected])
Received: from nautilus.iim.csic.es (nautilus.iim.csic.es
[193.144.47.33]) by mx3.abac.com (8.14.1/8.14.1) with ESMTP id
n26CgNbj039866 for <[email protected]>; Fri, 6 Mar 2009
04:42:30 -0800 (PST) (envelope-from [email protected])
Received: from [192.168.41.37] (helo=rebeca) by nautilus.iim.csic.es
with esmtp (Exim 4.63) (envelope-from <[email protected]>) id
1LfZNQ-0005wl-Es for [email protected]; Fri, 06 Mar 2009
13:42:08 +0100
From: Rebeca <[email protected]>
To: <[email protected]>
Subject: allometry: size correction of multiple groups
Old-Date: Fri, 6 Mar 2009 13:42:21 +0100
Message-ID: <003a01c99e58$fe6b5f10$2529a...@rebeca>
MIME-Version: 1.0
Content-Type: multipart/alternative;
boundary="----=_NextPart_000_003B_01C99E61.602FC710"
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.3350
thread-index: AcmeWP3qFjtbQZYGRyWnXKl5e1hNpw==
X-SA-Exim-Connect-IP: 192.168.41.37
X-SA-Exim-Mail-From: [email protected]
X-SA-Exim-Scanned: No (on nautilus.iim.csic.es); SAEximRunCond expanded
to false
X-Spam-Score: 0.101 (HTML_MESSAGE)
X-Diagnostic: Not on the accept list
X-Envelope-To: morphmet
Dear Morphometricians:
I am comparing four fish populations of the same species. I would like
to know if there are differences among the groups due to the geographic
location of each fish population. However, the size composition of the
fish in the samples is different and I would like to correct for size
effects before comparing the populations.
So far, I have been using MorphoJ to analyze my samples. In this
software, there is an option in the Regression menu that performs a
multivariate regression of shape on size and gives you back the
regression residuals, which can then be used to perform a CVA (canonical
variate analysis). I read on the user manual that when one has multiple
groups, and want to consider this group structure, the size correction
should be based on pooled within-group regression. When I do the CVA
after this size- correction (using pooled within-group, the separation
among the groups improves considerably and the misclassification is very
low.
However, I have read that a basic assumption to correct for size
(generally) is that all groups share the same allometric trajectory
(same slopes) and I suspect that the slopes are not the same for my
groups. So, My question is then, does anyone knows if the pooled-within
group regression routine in MorphoJ is assuming that all groups have the
same slope? Is there a better option to correct for size?
I don’t have much experience in this subject and I would really
appreciate any advice on this!
Thanks again,
Rebeca
<>< <>< <><
Rebeca P. Rodriguez Mendoza
PHD student
Marine Research Institute (IIM - CSIC) – Fisheries Dept.
C/ Eduardo Cabello, 6
Tel. 986 23 19 30 ext. 254
Fax: 986 29 27 62
36 208 Vigo, Spain
--
Replies will be sent to the list.
For more information visit http://www.morphometrics.org
