On a test of normality based on the empirical moment generating function

被引:8
|
作者
Henze, Norbert [1 ]
Koch, Stefan [2 ]
机构
[1] Karlsruhe Inst Technol, Inst Stochast, Englerstr 2, D-76131 Karlsruhe, Germany
[2] Univ Mannheim, Inst Math, A5 6, D-68159 Mannheim, Germany
关键词
Test of normality; Empirical moment generating function; Weighted L-2-statistic; Consistency; Contiguous alternatives; STATISTICS; FIT; POWER;
D O I
10.1007/s00362-017-0923-7
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
We provide the lacking theory for a test of normality based on a weighted L-2-statistic that employs the empirical moment generating function. The test statistic has a non-degenerate asymptotic null distribution, and the test is consistent against general alternatives. As a parameter associated with the weight function tends to infinity, an affine transformation of the test statistic approaches squared sample skewness.
引用
收藏
页码:17 / 29
页数:13
相关论文
共 50 条