The Power of Statistical Tests for Overdispersed and Zero Inflated Count Data in Soe Discrete Regression Models

Abstract

quotCount data regression models have been widely used in statistics to model newlineresponse variables that are assumed to be observed without error. Poisson newlineregression model is basically used as a standard model for analyzing the count data. newlineThere are two strong assumptions for Poisson model to be checked: one is that newlineevents occur independently over time or exposure period, the other is that the newlineconditional mean and variance are equal. In practice, counts have greater variance newlinethan the mean are described as overdispersion. This indicates that Poisson newlineregression is not adequate. There are two common causes that can lead to newlineoverdispersion. The first one is additional variation to the mean or heterogeneity newlinewhich is a negative binomial model is often used.quot newline newline