How to measure the "goodness of fit" in generalized linear models for repeated measures. To make clear: I'm searching for something which is based on the (deviance) residuals, something like an adj. R^2 for linear models. Something that could make me state about the efficacy of the model? Any thought would be appreciated!
Nakagawa and Schielzeth have some comments on R2 in GLMMs, which is more accurately what "GLMs for repeated measures" are - this might help clear up some confusion when you're looking for answers. Whether you're a biologist or not, I find Nakagawa generally produces very interpretable papers for people that are not professional statisticians. http://onlinelibrary.wiley.com/doi/10.1111/j.2041-210x.2012.00261.x/full