A FUNCTIONAL-APPROACH TO ESTIMATING FINITE POPULATION-DISTRIBUTION FUNCTIONS

被引：13

作者：

KUK, AYC

MAK, TK

机构：

[1] UNIV NEW S WALES,DEPT STAT,KENSINGTON,NSW 2033,AUSTRALIA

[2] CONCORDIA UNIV,DEPT DECIS SCI & MANAGEMENT INFORMAT SYST,MONTREAL H3G 1M8,QUEBEC,CANADA

来源：

COMMUNICATIONS IN STATISTICS-THEORY AND METHODS | 1994年 / 23卷 / 03期

关键词：

AUXILIARY INFORMATION; DESIGN-BASED ESTIMATION; DISTRIBUTION FUNCTIONS; FUNCTION COMPOSITIONS; RATIO ESTIMATOR;

D O I：

10.1080/03610929408831293

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：

Existing estimators of a finite population distribution function that utilize auxiliary information are often constructed by a pointwise argument. As a result, these estimators are not always monotone. We adopt a functional approach to the problem and propose two estimators based on compositions of functions. Asymptotic variance formulae are derived for the proposed estimators. Comparisons are made with existing estimators in a simulation study using three natural populations.

引用

页码：883 / 896

页数：14