-------- Original Message --------
Subject: RE: number of landmarks
Date: Wed, 28 May 2008 08:18:51 -0700 (PDT)
From: F. James Rohlf <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Organization: Stony Brook University
To: [EMAIL PROTECTED]
CC: [email protected]
References: <[EMAIL PROTECTED]>
There is no 'maximum possible number of specimens' that can be used in a
geometric morphometric study!
The minimum number depends on the type of analysis to be performed. For
a PCA there is no minimum other than to avoid the trivial cases. Need at
least two points to define PC1 and at least 3 to define two axes.
What is often pointed out as a sample size requirement is that if you
wish to perform some analysis that requires the inverse of a
within-group covariance matrix (such as CVA or discriminant functions)
then its degrees of freedom must be equal to or large than the number of
shape variables: 2p-4 for 2D data or 3p-7 for 3D data (where p is the
number of landmarks). Ideally, of course, the sample size should be much
larger.
For grouped data that you describe, the degrees of freedom for a
pooled-within groups covariance matrix would be the total sample size
minus the number of groups.
------------------------
F. James Rohlf, Distinguished Professor
Ecology & Evolution, Stony Brook University
www: http://life.bio.sunysb.edu/ee/rohlf
-----Original Message-----
From: morphmet [mailto:[EMAIL PROTECTED]
Sent: Wednesday, May 28, 2008 7:44 AM
To: morphmet
Subject: number of landmarks
-------- Original Message --------
Subject: number of landmarks
Date: Tue, 27 May 2008 15:30:10 -0700 (PDT)
From: Oliver Betz <[EMAIL PROTECTED]>
To: [email protected]
Dear list members:
I have a question regarding the maximum
possible number of specimens with respect to the
number of landmarks. We are acquiring 3D
landmarks on 20 different species, each
represented by 1-3 individuals. After GPA we
would like to examine the variation of the
species by PCA. In addition we would like to use
the method of thin plate splines as obtained for
3D datasets in the software morphologika to
visualize and interpret shape differences further.
I read that in the case of 3D landmarks the
number of specimens should be greater than at
least three times the number of landmarks. In our
case, having 20 species (each represented by 1-3
individulas, i.e. ca. 40 specimens in total), the
maximum number of landmarks allowed should amount
to 40 / 3 = 13. Is this correct?
On the other hand, I was wondering whether these
restrictions hold only to the application of
statistical tests following the PCA, so for pure
visualization puposes as descibed above even more landmarks might
bepossible.
I will be happy if you send your replies to
<oliver Betz> [EMAIL PROTECTED] I
promise to summarize the incoming replies in a sepertate e-mail to the
list.
My best wishes,
Oliver
Univ.-Prof. Dr. Oliver Betz
Zoologisches Institut
Abt. Evolutionsbiologie der Invertebraten
Auf der Morgenstelle 28E
D-72076 Tübingen
Germany
phone: 0049-(0)7071-2972995
E-mail: [EMAIL PROTECTED]
http:/www.uni-tuebingen.de/agbetz
--
Replies will be sent to the list.
For more information visit http://www.morphometrics.org
--
Replies will be sent to the list.
For more information visit http://www.morphometrics.org
