ABSTRACT
This chapter explains that “count data” are non-negative integers that can be appropriately modeled with Poisson or negative binomial regression. It provides an explanation of the purpose and history of these types of statistics in general and in the planning field. The mechanics section goes over the model estimation and assessment of fit. It also explains the testing for overdispersion and how to handle zero-inflated models. It also looks at underdispersed data and introduces the idea of more complex models and the statistical packages that include them. It also includes a section on how to interpret the results. The “Step by Step” section instructs on how to use SPSS and R to estimate Poisson and negative binomial regression models. The chapter concludes with two planning case studies.
