Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/31006 
Year of Publication: 
2005
Series/Report no.: 
Discussion Paper No. 423
Publisher: 
Ludwig-Maximilians-Universität München, Sonderforschungsbereich 386 - Statistische Analyse diskreter Strukturen, München
Abstract: 
Poisson regression models for count variables have been utilized in many applications. However, in many problems overdispersion and zeroinflation occur. We study in this paper regression models based on the generalized Poisson distribution (Consul (1989)). These regression models which have been used for about 15 years do not belong to the class of generalized linear models considered by McCullagh and Nelder (1989) for which an established asymptotic theory is available. Therefore we prove consistency and asymptotic normality of a solution to the maximum likelihood equations for zero-inflated generalized Poisson regression models. Further the accuracy of the asymptotic normality approximation is investigated through a simulation study. This allows to construct asymptotic confidence intervals and likelihood ratio tests.
Subjects: 
central limit theorem
likelihood
maximum likelihood estimator
overdispersion
zero-inflated generalized Poisson regression
Persistent Identifier of the first edition: 
Document Type: 
Working Paper

Files in This Item:
File
Size
305.79 kB
598.52 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.