On the robustness of empirical likelihood ratio confidence intervals for location

被引:6
|
作者
Tsao, M [1 ]
Zhou, J [1 ]
机构
[1] Univ Victoria, Dept Math & Stat, Victoria, BC V8W 3P4, Canada
关键词
breakdown point; confidence interval; empirical likelihood; Huber's location M-estimate; robustness;
D O I
10.2307/3316056
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
The authors examine the robustness of empirical likelihood ratio (ELR) confidence intervals for the mean and M-estimate of location. They show that the ELR interval for the mean has an asymptotic breakdown point of zero. They also give a formula for computing the breakdown point of the ELR interval for M-estimate. Through a numerical study, they further examine the relative advantages of the ELR interval to the commonly used confidence intervals based on the asymptotic distribution of the M-estimate.
引用
收藏
页码:129 / 140
页数:12
相关论文
共 50 条