A class of estimators in two-phase sampling with subsampling the non-respondents

被引：8

作者：

Diana, Giancarlo ^{[1
]}

Perri, Pier Francesco ^{[2
]}

机构：

[1] Univ Padua, Dept Stat Sci, I-35121 Padua, Italy

[2] Univ Calabria, Dept Econ Stat & Finance, I-87036 Arcavacata Di Rende, Italy

来源：

APPLIED MATHEMATICS AND COMPUTATION | 2013年 / 219卷 / 19期

关键词：

Asymptotic variance; Auxiliary information; Efficiency; Regression type-estimator; MISSING DATA; NONRESPONSE; REGRESSION;

D O I：

10.1016/j.amc.2013.03.106

中图分类号：

O29 [应用数学];

学科分类号：

070104 ;

摘要：

Motivated by Singh and Kumar [14], we introduce in this paper a general class of estimators for the population mean of a study variable when two auxiliary variables are used in the presence of non-response. The minimum asymptotic variance bound of the estimators belonging to the class is determined and the optimality of Singh-Kumar estimators discussed. The best estimator in the class is analytically found in accordance with the auxiliary information used and the efficiency gain that can be achieved upon competitive estimators is shown by an empirical study. (C) 2013 Elsevier Inc. All rights reserved.

引用

页码：10033 / 10043

页数：11

共 17 条

[1]

Cochran W.G., 2007, Sampling techniques

[2]

DIANA G, 2007, METRON, V65, P99

[3] An improved class of estimators for the population mean [J].