A Poisson mixed model with nonnormal random effect distribution
Contribuinte(s) |
UNIVERSIDADE DE SÃO PAULO |
---|---|
Data(s) |
07/11/2013
07/11/2013
2012
|
Resumo |
In this paper, we propose a random intercept Poisson model in which the random effect is assumed to follow a generalized log-gamma (GLG) distribution. This random effect accommodates (or captures) the overdispersion in the counts and induces within-cluster correlation. We derive the first two moments for the marginal distribution as well as the intraclass correlation. Even though numerical integration methods are, in general, required for deriving the marginal models, we obtain the multivariate negative binomial model from a particular parameter setting of the hierarchical model. An iterative process is derived for obtaining the maximum likelihood estimates for the parameters in the multivariate negative binomial model. Residual analysis is proposed and two applications with real data are given for illustration. (C) 2011 Elsevier B.V. All rights reserved. CNPq FAPESP (Brazil) |
Identificador |
COMPUTATIONAL STATISTICS & DATA ANALYSIS, AMSTERDAM, v. 56, n. 6, pp. 1499-1510, JUN, 2012 0167-9473 http://www.producao.usp.br/handle/BDPI/43104 10.1016/j.csda.2011.12.002 |
Idioma(s) |
eng |
Publicador |
ELSEVIER SCIENCE BV AMSTERDAM |
Relação |
COMPUTATIONAL STATISTICS & DATA ANALYSIS |
Direitos |
closedAccess Copyright ELSEVIER SCIENCE BV |
Palavras-Chave | #COUNT DATA #GENERALIZED LOG-GAMMA DISTRIBUTION #MULTIVARIATE NEGATIVE BINOMIAL DISTRIBUTION #OVERDISPERSION #RANDOM-EFFECT MODELS #GENERALIZED GAMMA-DISTRIBUTION #APPROXIMATE INFERENCE #REGRESSION-MODELS #LINEAR-MODELS #DIAGNOSTICS #COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS #STATISTICS & PROBABILITY |
Tipo |
article original article publishedVersion |