Journal of Statistics Applications & Probability


The over-dispersion and the excess zeros are two main problems that are related to the Poisson regression model. To handle these problems, different models like zero inflated negative binomial and zero inflated generalized Poisson were proposed. Lately Conway-Maxwell Poisson (COM-Poisson) had proposed to handle the over-dispersion for count data in cross-sectional case. However, there is no application of the COM-Poisson model in longitudinal case. In this paper, the zero inflated COM-Poisson model was proposed and developed to deal with longitudinal count data. Under two different working correlation structures, exchangeable and autoregressive of order 1, AR(1). The zero-inflated COM- Poisson (ZICP) regression model is considered as modification of the COM-Poisson regression model that allows for an excess of zero counts in the data and over-dispersion problem. this model was compared with the Zero Inflated Poisson regression model and zero inflated negative binomial model. The results show that the COM-Poisson model is very suitable to longitudinal count data, even in presence of dispersion. It gives the smallest AIC values. Also, it is insensitive to the choice of the working structure.

Digital Object Identifier (DOI)