maybe after transitioning normal lin mod with likelihood to normal lin mod with bayes, we could go binomial to multinomial on categorical data that had been semi-arbitrarily split, for binomial purposes, into two categories (e.g. herbivore or not) when full data had more categories (e.g. herb, pred, detritivore, etc)
maybe after transitioning normal lin mod with likelihood to normal lin mod with bayes, we could go binomial to multinomial on categorical data that had been semi-arbitrarily split, for binomial purposes, into two categories (e.g. herbivore or not) when full data had more categories (e.g. herb, pred, detritivore, etc)