I am wondering if I can calculate robust standard errors and how I can do it (if possible) after multilevel quasipoisson using glmmpql.
An example would be the following code:
glmmPQL(nabs ~ gen + sco + pedu, random = ~ 1 | region/school, family = quasipoisson, data = data))
After this, how can I (or is it possible to) calculate robust standard errors?