Oh jeepers. GLMMs are a bit of a nightmare for lots of reasons. I think that this book covers them: http://www.amazon.com/Generalized-Linear-Mixed-Models-Applications/dp/1439815127/ref=sr_1_1?ie=UTF8&qid=1404905964&sr=8-1&keywords=generalized+linear+mixed+models But generally, the issues are the same as for linear mixed models. And there, my rule of thumb is to use unstructured, unless it won't converge, and then I use variance components. Maybe use autoregressive (or that one that's a bit like autoregressive, but more extreme in the rate of descent, and I've forgotten its name) if I'm desperate. Jeremy On 8 July 2014 17:11, Jennifer Wathan <[log in to unmask]> wrote: > Hi all, > > Does anyone know of a good reference that explains the different types > of covariance structures that you can use (and which is applicable when) > when running Generalised Linear Mixed Models? > > Thanks, > Jen > > ----------------------------------- > Jen Wathan > > [log in to unmask] > +44 (0) 1273 876602 > > Mammal Vocal Communication and Cognition Research Group > School of Psychology > University of Sussex > Falmer > BN1 9HQ > > http://www.lifesci.sussex.ac.uk/cmvcr/Home.html > >