I have claim frequency panel data for five years consisting of age, CC, make of car, age and gender of insured person and geographical area. I have run a Poisson regression to model the claim frequency in R. Now I want to test for a Poisson fixed and random effect. How can I proceed?
There's a nice presentation on fixed and random effect analysis in R. It's not for Poisson data, but you may still find it helpful: http://www.princeton.edu/~otorres/Panel101R.pdf
Maybe it is too late now, but you can consider pglm package from Yves Croisant.
This package does poisson and negbin with fixed and random effects.