ON COMBINING INDEPENDENT TESTS IN LINEAR-MODELS

被引：3
|
作者：
JORDAN, SM [1 ]
KRISHNAMOORTHY, K [1 ]
机构：
[1] UNIV SW LOUISIANA,DEPT STAT,LAFAYETTE,LA 70504
来源：
STATISTICS & PROBABILITY LETTERS | 1995年 / 23卷 / 02期
关键词：
COMBINED TEST; FISHERS TEST; RECOVERY OF INTERBLOCK INFORMATION;
D O I：
10.1016/0167-7152(94)00102-E
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 problem of testing the common mean of several normal linear models when the variances are unknown and unequal is considered. A minor modification to the combined tests given in Zhou and Mathew (1993a) is proposed. The powers of the modified tests are numerically compared with their original versions and a traditional test. Comparison studies indicate that the modified tests have better power properties than their original versions. Practical situations where the modified tests are preferable to the traditional test are pointed out.
引用
收藏
页码：117 / 122
页数：6
相关论文
共 50 条
12345