Included in this course is an e–book and a set of slides. The course is divided into two parts. In the first part, students are introduced to the theory behind count models. The theory is explained in an intuitive way while keeping the math at a minimum. The course starts with an introduction to count tables, where students learn how to calculate the incidence–rate ratio. From there, the course moves on to Poisson regression where students learn how to include continuous, binary, and categorical variables. Students are then introduced to the concept of overdispersion and the use of negative binomial models to address this issue. Other count models such as truncated models and zero–inflated models are discussed.
In the second part of the course, students learn how to apply what they have learned using Stata. In this part, students will walk through a large project in order to fit Poisson, negative binomial, and zero–inflated models. The tools used to compare these models are also introduced.
Courses : 2
Specification: Modeling Count Data using Stata