On the hurdle negative binomial regression model with excess zeros for a right truncated count data

Abstract

A Poisson regression model is well-known for modeling the data with response variable in form of counts. However, one often encounters the situation with excess zeros occurred in the observed responses. Therefore, Poisson model is not suitable any more for this kind of data. Thus, we propose to use a hurdle negative binomial model. Furthermore, the response variable in such cases is truncated for some values. So, a truncated hurdle negative binomial model is introduced on count data with many zeros. In this model, we consider a response variable and one or more than one explanatory variables. The estimates using the maximum likelihood method are discussed and the goodness-of-fit for the regression model is examined. We study the effects of right truncation in terms of parameters estimation and their standard errors via an example and a simulation.