[R] GLM, LMER, GEE interpretation

Hi, my dependent variable is a proportion (prob.bind), and the independent variables are factors for group membership (group) and a covariate (capacity). I am interested in the effects of group, capacity, and their interaction. Each subject is observed on all (4) levels of capacity (I use capacity

Re: [R] GLM, LMER, GEE interpretation

Daniel Malter daniel at umd.edu writes: Hi, my dependent variable is a proportion (prob.bind), and the independent variables are factors for group membership (group) and a covariate (capacity). I am interested in the effects of group, capacity, and their interaction. Each subject is

Re: [R] GLM, LMER, GEE interpretation

I agree with Ben. Theoretically, Laplace (lmer) provides a better approximation. Chunhao Quoting Ben Bolker [EMAIL PROTECTED]: Daniel Malter daniel at umd.edu writes: Hi, my dependent variable is a proportion (prob.bind), and the independent variables are factors for group membership

Re: [R] GLM, LMER, GEE interpretation

: [R] GLM, LMER, GEE interpretation Daniel Malter daniel at umd.edu writes: Hi, my dependent variable is a proportion (prob.bind), and the independent variables are factors for group membership (group) and a covariate (capacity). I am interested in the effects of group, capacity

Re: [R] GLM, LMER, GEE interpretation

Daniel Malter daniel at umd.edu writes: Thanks for your answers. I appreciate your help. I tried the glmmML. However, it seems glmmML does not allow for a quasibinomial fit as I did with the models I used. I have large overdispersion which I account for using a quasibinomial with scaling