Marianna and others paying attention,

I would not advise forcing things into MorphoJ because that software is easy to 
use. As stated clearly in MorphoJ's help file, the Procrustes ANOVA implemented 
in that software is for measurement error and asymmetry designs. For other 
biological and statistical designs it states: 

"It is therefore not a replacement for general implementations of MANOVAs of 
various designs (e.g. crossed factorial designs, nested designs). " For the 
latter types of analyses, one really does require more flexible statistical 
software. 

Your original post did not state the nature of your design; it asked about 
permutations. But to reiterate, all manner of designs may be implanted in R, 
and geomorph provides relatively straightforward access to them for shape data 
analysis. It allows factorial, nested, and cross designs (as well as asymmetry 
and ME designs), allows type I, II or III sums of squares (the latter 2 of 
which are quite useful and even necessary in biology, as we have usually have 
imbalanced data and factors that do not have an inherent 'order' in the model 
design), and of course can do permutations: which was your original query to 
this group. And unlike many R-packages, the help files are extensive and have 
examples that one can use as a guide. 

Dean

Dr. Dean C. Adams
Distinguished Professor of Evolutionary Biology
Department of Ecology, Evolution, and Organismal Biology
Iowa State University
https://faculty.sites.iastate.edu/dcadams/
phone: 515-294-3834

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of 
andrea cardini
Sent: Friday, January 10, 2025 1:15 AM
To: Morphmet2 <[email protected]>
Subject: [MORPHMET2] quick one on design of MorphoJ Procr. ANOVA

Dear All,
this is something I forgot to mention when we discussed how to replicate 
MorphoJ's Procrustes ANOVA in a different software. It's mainly for Marianna 
but a few others can be interested.

When the analysis is replicated (for instance, to do the permutation tests 
instead of the parametric ones), I suggest to check first the appropriateness 
of the design in relation to the study questions. In MorphoJ, if I am correct, 
the analysis was originally tailored for the study of directional (DAS) and 
fluctuating (FAS) asymmetry, as described in detail in the ref. below and some 
of the others available in the help:
Klingenberg, C. P., M. Barluenga, and A. Meyer. 2002. Shape analysis of 
symmetric structures: quantifying variation among individuals and asymmetry. 
Evolution 56:1909–1920.
Of course, even when the analysis is specific to symmetry/asymmetry, there can 
be alternative approaches. My point is not to argue for one or the other. It's 
just a clarification for those who want to use the same precise design as in 
MorphoJ.

With that design, if I remember well, the main factor (group, if
present) is tested against individual variation; individual variation against 
FAS; DAS against FAS, and FAS against residual variation (meas. 
error estimated from individual replicates). Check carefully as I may be wrong. 
To check, one can copy/paste the ANOVA table in xls and manually compute the MS 
ratios: the results should be the same F ratios as in the table.
Whether this is the design one needs for her/his study it may depend on the 
specific questions, as I mentioned. Also, if one has more than one main factor, 
the order of the factors matters in MorphoJ (easy to verify by swapping them) 
and again depends on the study questions. Finally, the only interaction in the 
output is individual by side (FAS).
With R I am sure one can modify the design according to specific needs 
(different F ratios, interactions with main factors etc.) and run permutations.

All the best

Andrea

--
Andrea Cardini
E-mail address: [email protected], [email protected]
WEBPAGE: https://sites.google.com/view/alcardini/

--
You received this message because you are subscribed to the Google Groups 
"Morphmet" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion visit 
https://groups.google.com/d/msgid/morphmet2/13aa11cb-26a4-4dd5-9302-3df8958ec403%40gmail.com.

-- 
You received this message because you are subscribed to the Google Groups 
"Morphmet" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion visit 
https://groups.google.com/d/msgid/morphmet2/CO6PR04MB84274460FEE32B1997B88281A21C2%40CO6PR04MB8427.namprd04.prod.outlook.com.

Reply via email to