Bayesian Model Selection for Longitudinal Count Data

被引:2
作者
Ariyo, Oludare [1 ,2 ,4 ]
Lesaffre, Emmanuel [1 ,4 ]
Verbeke, Geert [1 ,4 ]
Quintero, Adrian [3 ]
机构
[1] Katholieke Univ Leuven, Interuniv Inst Biostat & Stat Bioinformat I BioSt, Leuven, Belgium
[2] Fed Univ Agr, Dept Stat, Abeokua, Nigeria
[3] Icfes Colombian Inst Educ Evaluat, Stat Dept, Bogota 111071, Colombia
[4] Fed Univ Agr, Dept Stat, Abeokua, Nigeria
来源
SANKHYA-SERIES B-APPLIED AND INTERDISCIPLINARY STATISTICS | 2022年 / 84卷 / 02期
基金
比利时弗兰德研究基金会;
关键词
Replication sampling; Marginal likelihood; Bayesian model selection; DEVIANCE INFORMATION CRITERION; OVERDISPERSION; PARAMETERS; REGRESSION; INFERENCE;
D O I
10.1007/s13571-021-00268-9
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
We explore the performance of three popular model-selection criteria for generalised linear mixed-effects models (GLMMs) for longitudinal count data (LCD). We focus on evaluating the conditional criteria (given the random effects) versus the marginal criteria (averaging over the random effects) in selecting the appropriate data-generating model. We advocate the use of marginal criteria, since Bayesian statisticians often use the conditional criteria despite previous warnings. We discuss how to compute the marginal criteria for LCD by a replication method and importance sampling algorithm. Besides, we show via simulations to what extent we err when using the conditional criteria instead of the marginal criteria. To promote the usage of the marginal criteria, we developed an R function that computes the marginal criteria for longitudinal models based on samples from the posterior distribution. Finally, we illustrate the advantages of the marginal criteria on a well-known data set of patients who have epilepsy.
引用
收藏
页码:516 / 547
页数:32
相关论文
共 52 条
[1]   Bayesian model selection techniques as decision support for shaping a statistical analysis plan of a clinical trial: An example from a vertigo phase III study with longitudinal count data as primary endpoint [J].
Adrion, Christine ;
Mansmann, Ulrich .
BMC MEDICAL RESEARCH METHODOLOGY, 2012, 12
[2]   Comparison of Additive and Multiplicative Bayesian Models for Longitudinal Count Data with Overdispersion Parameters: A Simulation Study [J].
Aregay, Mehreteab ;
Shkedy, Ziv ;
Molenberghs, Geert .
COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION, 2015, 44 (02) :454-473
[3]   A hierarchical Bayesian approach for the analysis of longitudinal count data with overdispersion: A simulation study [J].
Aregay, Mehreteab ;
Shkedy, Ziv ;
Molenberghs, Geert .
COMPUTATIONAL STATISTICS & DATA ANALYSIS, 2013, 57 (01) :233-245
[4]   Model selection for Bayesian linear mixed models with longitudinal data: Sensitivity to the choice of priors [J].
Ariyo, Oludare ;
Lesaffre, Emmanuel ;
Verbeke, Geert ;
Quintero, Adrian .
COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION, 2022, 51 (04) :1591-1615
[5]   Bayesian model selection in linear mixed models for longitudinal data [J].
Ariyo, Oludare ;
Quintero, Adrian ;
Munoz, Johanna ;
Verbeke, Geert ;
Lesaffre, Emmanuel .
JOURNAL OF APPLIED STATISTICS, 2020, 47 (05) :890-913
[6]   Simultaneous Bayesian modelling of skew-normal longitudinal measurements with non-ignorable dropout [J].
Ariyo, Oludare Samuel ;
Adeleke, Matthew Adekunle .
COMPUTATIONAL STATISTICS, 2022, 37 (01) :303-325
[7]   Negative binomial loglinear mixed models [J].
Booth, JG ;
Casella, G ;
Friedl, H ;
Hobert, JP .
STATISTICAL MODELLING, 2003, 3 (03) :179-191
[8]  
BRESLOW NE, 1984, J R STAT SOC C-APPL, V33, P38
[9]   APPROXIMATE INFERENCE IN GENERALIZED LINEAR MIXED MODELS [J].
BRESLOW, NE ;
CLAYTON, DG .
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1993, 88 (421) :9-25
[10]  
Celeux G, 2006, BAYESIAN ANAL, V1, P651, DOI 10.1214/06-BA122