One option that I can think of is to code the binary trait at the genus level (it sounds like this should be fairly simple for most of them, then maybe use a cutoff for the rest) and then do a phylogenetic logistic regression (e.g. using phylolm)?
On Sat, Jul 20, 2019, 1:15 PM Elizabeth Christina Miller < ecmil...@email.arizona.edu> wrote: > Hello, > > I want to know if a binary character is related to species richness (not > diversification rates). I am using genera as units, and using PGLS to > relate genus richness to the proportion of species that have a particular > state. From the model diagnostic plots I am worried that my data are > violating assumptions of PGLS. Most genera either have no species with this > trait or all species having the trait (proportion = 0 or 1), with a few > genera in the middle. As such, the corrected residuals still have > structure. Is there a better way to go about testing this question? > > Thank you! > > [image: model diagnostics richness.png] > > -- > Elizabeth Miller > PhD Candidate > Ecology and Evolutionary Biology > University of Arizona > > > _______________________________________________ > R-sig-phylo mailing list - R-sig-phylo@r-project.org > https://stat.ethz.ch/mailman/listinfo/r-sig-phylo > Searchable archive at > http://www.mail-archive.com/r-sig-phylo@r-project.org/ >
_______________________________________________ R-sig-phylo mailing list - R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo Searchable archive at http://www.mail-archive.com/r-sig-phylo@r-project.org/