A comment on locally most powerful tests in the presence of nuisance parameters

被引：5

作者：

Marohn, F ^{[1
]}

机构：

[1] Katholische Univ Eichstatt, Math Geog Fak, D-85071 Eichstatt, Germany

来源：

COMMUNICATIONS IN STATISTICS-THEORY AND METHODS | 2002年 / 31卷 / 03期

关键词：

score test; local asymptotic normality (LAN); optimal test sequence; efficient estimator sequence;

D O I：

10.1081/STA-120002852

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：

Let P-kappa,P-lambda be a k-parametric family of distributions, where kappa is an element of R is a one-dimensional parameter of interest and lambda is an element of Rk-1 is a nuisance parameter. Assuming L-2-differentiability of the underlying model, we give conditions for the asymptotic optimality of locally most powerful (LMP) tests with estimated nuisance parameter. Various examples as scale-location families are discussed.

引用

页码：337 / 349

页数：13

共 13 条

[1]

[Anonymous], THEORY POINT ESTIMAT

[2]

Bickel Peter J, 1993, Efficient and adaptive estimation for semiparametric models, V4

[3]

Choi SS, 1996, ANN STAT, V24, P841

[4]

GALAMBOS J, 1987, ASYMPTOTIC THEORY OR

[5] ON LARGE-SAMPLE ESTIMATION AND TESTING IN PARAMETRIC MODELS [J].

HALL, WJ ;

MATHIASON, DJ .

INTERNATIONAL STATISTICAL REVIEW, 1990, 58 (01) :77-97

[6]

Ibragimov IA, 1981, Statistical Estimation

[7]

MAROHN F., 2000, Extremes, V3, P363

[8]

MULLER UU, 2001, IN PRESS SEL P S INF

[9]

Pfanzagl Johann., 1994, PARAMETRIC STAT THEO

[10]

Serfling R.J., 1980, APPROXIMATION THEORE

← 1 2 →