|
Activity Number:
|
330
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 11, 2004 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
SSC
|
| Abstract - #300771 |
|
Title:
|
Generalized Linear Model, Zero-inflation, and Overdispersion
|
|
Author(s):
|
Dianliang Deng*+
|
|
Companies:
|
University of Regina
|
|
Address:
|
3737 Wascana Pkwy., Regina, SK, S4S 0A2, Canada
|
|
Keywords:
|
binomial models ; generalized linear model ; overdispersion ; Poisson model ; score test ; zero-inflation
|
|
Abstract:
|
Discrete data in the form of counts often exhibit extra variation that can not be explained by a simple model, such as the binomial or the Poisson model. Also, these data, sometimes, show more zero counts than what can be predicted by a simple model. Therefore, a discrete generalized linear model (Poisson or binomial) may fail to fit a set of discrete data either because of zero-inflation or because of overdispersion or because there is zero-inflation as well as overdispersion in the data. We deal with the class of zero-inflated overdispersed generalized linear models and propose procedures based on score tests for selecting a model that fits such data. We show that in certain cases and under certain conditions the score tests derived using the general overdispersion model and those developed under specific overdispersion models are identical. Empirical level and power properties of the tests are examined by a limited simulation study. Two illustrative example are given.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2004 program |
