-------- Original Message --------
Subject: MorphoJ bug warning: pooled within-group regression
Date: Wed, 24 Feb 2010 03:55:41 -0500
From: Chris Klingenberg <[email protected]>
Reply-To: [email protected]
Organization: University of Manchester
To: [email protected]

Dear morphometricians

I have just become aware of a programming bug in the computations for
the residuals from pooled within-group regression (Regression procedure,
with option for 'Pooled regression within subgroups' selected). The
mistake affects only the computation of residuals.

Depending on the directions of the regression vector and the differences
between groups, this bug tends to exaggerate group differences. This
effect is large in some of the examples I've tried.

This is especially important when the pooled within-group regression is
used for size correction in several groups simultaneously. The results
obtained in this way with the existing versions of MorphoJ (all 1.00 and
1.01 versions) should be considered as unreliable and possibly misleading.

I sincerely apologize to all those of you who are affected by this problem.


I am putting the finishing touches on a new version (1.02a), which will
include a correction of the bug. I will make another announcement to
Morphmet when I upload the new version to my lab web site.

Best wishes,
Chris


(P.S.: Andrea Cardini will


--
***************************************************************
Christian Peter Klingenberg
Faculty of Life Sciences
The University of Manchester
Michael Smith Building
Oxford Road
Manchester M13 9PT
United Kingdom

Telephone: +44 161 275 3899
Fax: +44 161 275 5082
E-mail: [email protected]
Web: http://www.flywings.org.uk
Skype: chris_klingenberg
***************************************************************



--
For now, new message AND replies should be sent to:

[email protected]

/* Replies will be sent to the list. */
For more information visit http://www.morphometrics.org

Reply via email to