方差分量的广义p-值检验  

Generalized p-value test for variance components

作　　者：赵静[1] 王磊[1] 

机构地区：[1]滨州学院数学与信息科学系,山东滨州256603

出　　处：《安徽大学学报（自然科学版）》2010年第2期34-37,共4页Journal of Anhui University(Natural Science Edition)

基　　金：滨州学院科研基金资助项目(BZXYLG200707)

摘　　要：利用广义p-值检验法给出了含两个方差分量的线性混合模型中一种新的方差分量的检验方法,对单个随机效应和两个模型随机效应方差分量间的比较分别作出了检验,这些检验都是精确检验,且计算简单,便于使用.A new procedure on the basis of generalized p-value test for testing variance components in linear mixed model with two variance components was developed. The authors developed tests for variance component of a single random effect and similar procedures for obtaining the test for comparison of variance components of random effects in two models. The procedures were exact tests and easy to compute and use.

关 键 词：线性混合模型 广义P-值 广义置信区间 精确检验 

分 类 号：O212.1[理学—概率论与数理统计]