A New Bayesian Analysis of a Random Effects Model
针对单向方差分析中的典型随机效应模型，提出一种新贝叶斯分析方法，定义可交换先验分布，将问题视为估计而非F检验，并得到衡量变异程度的参数函数的后验分布。
SUMMARY For a typical random effects model in the one way analysis of variance a new Bayesian analysis is offered. An exchangeable prior distribution is defined over the parameters. It is suggested that rather than use the usual “F” test one should consider this as an estimation problem. The posterior distribution of a function of the parameters which measures the amount of variation present is found.