Hierarchical models with normal and conjugate random effects: a review

被引：3

作者：

Molenberghs, Geert ^{[1
,2
]}

Verbeke, Geert ^{[1
,2
]}

Demetrio, Clarice G. B. ^{[3
]}

机构：

[1] Univ Hasselt, I BioStat, Martelarenlaan 42, B-3500 Hasselt, Belgium

[2] Katholieke Univ Leuven, I BioStat, B-3000 Leuven, Belgium

[3] Univ Sao Paulo, ESALQ, Piracicaba, Brazil

来源：

SORT-STATISTICS AND OPERATIONS RESEARCH TRANSACTIONS | 2017年 / 41卷 / 02期

关键词：

Conjugacy; frailty; jointmodelling; marginalized multilevel model; mixed model; overdispersion; underdispersion; variance component; zero-inflation; LOCAL INFLUENCE DIAGNOSTICS; LONGITUDINAL COUNT DATA; LINEAR MIXED MODELS; NEGATIVE VARIANCE-COMPONENTS; LOGISTIC-REGRESSION-MODELS; COMBINED GAMMA FRAILTY; RANDOM EFFECTS PROBIT; MARGINAL CORRELATION; BAYESIAN-APPROACH; COMBINED BETA;

D O I：

10.2436/20.8080.02.58

中图分类号：

C93 [管理学]; O22 [运筹学];

学科分类号：

070105 ; 12 ; 1201 ; 1202 ; 120202 ;

摘要：

Molenberghs, Verbeke, and Demetrio (2007) and Molenberghs et al. (2010) proposed a general framework to model hierarchical data subject to within-unit correlation and/or overdispersion. The framework extends classical overdispersion models as well as generalized linear mixed models. Subsequentwork has examined various aspects that lead to the formulation of several extensions. A unified treatment of the model framework and key extensions is provided. Particular extensions discussed are: explicit calculation of correlation and other moment-based functions, joint modelling of several hierarchical sequences, versions with direct marginally interpretable parameters, zero-inflation in the count case, and influence diagnostics. The basic models and several extensions are illustrated using a set of key examples, one per data type (count, binary, multinomial, ordinal, and time-to-event).

引用

页码：191 / 253

页数：63

共 109 条

[1] Parametric overdispersed frailty models for current status data [J].