ABSTRACT
This chapter considers regression analysis of count data, including simple count data and panel count data. The former can be regarded as a type of complete data, while the latter is a type of incomplete data. For their inference, we first discuss the maximum likelihood approach under the Poisson and negative binomial regression models for simple count data and then present two likelihood approaches for regression analysis of correlated count data under the framework of non-homogeneous Poisson or negative binomial processes. In addition, two estimating equation-based methods are also discussed that do not rely on any distribution assumption and an illustrative example is provided.
