Generalized poisson regression model

P. C. Consul, Felix Famoye

Research output: Contribution to journalArticlepeer-review

195 Scopus citations

Abstract

The generalized Poisson distribution has been found useful in fitting over-dispersed as well as under-dispersed count data. Since a number of models and methods have been proposed for the regression analysis of count data either with under-dispersion or with over-dispersion, we define and study a generalized Poisson regression (GPR) model which is useful in predicting a response variable affected by one or more covariates. This regression model is suitable for both types of dispersions. The methods of maximum likelihood and moments are given for the estimation of parameters. Approximate tests for the adequacy of the model are considered. Asymptotic tests are given for the significance of regression parameters. The GPR model has been applied to four observed data sets to which other regression models were applied earlier.

Original languageEnglish
Pages (from-to)89-109
Number of pages21
JournalCommunications in Statistics - Theory and Methods
Volume21
Issue number1
DOIs
StatePublished - Jan 1 1992

Keywords

  • Count data
  • deviance
  • generalized Poisson distribution
  • hypothesis testing
  • maximum likelihood
  • over-dispersion
  • under- dispersion

Fingerprint

Dive into the research topics of 'Generalized poisson regression model'. Together they form a unique fingerprint.

Cite this