Hi Popko,
My sense is that if the polytomy is hard, it is never necessary to
subtract a degree of freedom. If the polytomy is "firm" (sensu Purvis &
Garland 1993) - in that the true tree is fully bifurcating but we cannot
resolve a multifurcating node due primarily to rapid divergence of the
descendant taxa - then subtracting a degree of freedom will probably be
excessively conservative. If the polytomy is truly soft (sensu Maddison
1989; Garland & Diaz-Uriarte 1999), then subtracting a degree of freedom
for each extra "furc" will probably also be conservative. [As Rohlf
2001 points out, this would imply zero degrees of freedom if the
phylogeny was completely unresolved (p. 2154), which seems excessive.]
Some of these issues are discussed in Garland & Diaz-Uriarte (1999;
Syst. Biol.) and elsewhere.
I'm sure that other subscribers to the list could add significant
insight to this issue beyond what I am able to offer.
- Liam
--
Liam J. Revell
University of Massachusetts Boston
web: http://www.bio.umb.edu/facstaff/faculty_Revell.html
(new) email: liam.rev...@umb.edu
(new) blog: http://phytools.blogspot.com
On 2/22/2011 5:31 AM, Popko Wiersma wrote:
Dear all,
Can somebody tell whether one should subtract degrees of freedom when
applying PGLS with a phylogeny containing soft polytomies? And does it
make a difference if polytomies originate from data from multiple
specimens of species?
cheers, Popko Wiersma
- I reposted this message because the first attempt did not show the
text directly in the message body. Apologies for any confusion -
_______________________________________________
R-sig-phylo mailing list
R-sig-phylo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
_______________________________________________
R-sig-phylo mailing list
R-sig-phylo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
