Posterior bimodality in the balanced one-way random-effects model

被引:17
作者
Liu, JN
Hodges, JS
机构
[1] Minneapolis Med Fdn, Nephrol Analyt Serv, Minneapolis, MN 55404 USA
[2] Univ Minnesota, Minneapolis, MN USA
关键词
Bayesian analysis; bimodality; hierarchical model; prior distribution; random effects; variance components;
D O I
10.1111/1467-9868.00384
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 [统计学]; 070103 [概率论与数理统计]; 0714 [统计学];
摘要
Although some researchers have examined posterior multimodality for specific richly parameterized models, multimodality is not well characterized for any such model. The paper characterizes bimodality of the joint and marginal posteriors for a conjugate analysis of the balanced one-way random-effects model with a flat prior on the mean. This apparently simple model has surprisingly complex and even bizarre mode behaviour. Bimodality usually arises when the data indicate a much larger between-groups variance than does the prior. We examine an example in detail, present a graphical display for describing bimodality and use real data sets from a statistical practice to shed light on the practical relevance of bimodality for these models.
引用
收藏
页码:247 / 255
页数:9
相关论文
共 13 条
[1]
[Anonymous], BUGS EXAMPLES VERSIO
[2]
Atkinson AC, 1998, J ROY STAT SOC B, V60, P521
[3]
Investigation of bimodality in likelihoods and posterior densities [J].
Firat, MZ ;
Thompson, R .
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION, 1996, 54 (04) :379-386
[4]
The posterior distribution of the fixed and random effects in a mixed-effects linear model [J].
Harville, DA ;
Zimmermann, AG .
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION, 1996, 54 (1-3) :211-229
[5]
HOESCHELE I, 1988, J STAT COMPUT SIM, V33, P149
[6]
LIU J, 2001, THESIS U MINNESOTA M
[7]
LIU J, 2001, 2001025 U MINN DIV B
[8]
LUCAS TW, 1994, BIOMETRIKA, V81, P385
[9]
WHEN IS CONFLICT NORMAL [J].
LUCAS, TW .
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1993, 88 (424) :1433-1437
[10]
Montgomery D. C., 1991, DESIGN ANAL EXPT