Cost-effective Estimation of the Population Mean Using Prediction Estimators

Fujii, Tomoki ; van der Weide, Roy

World Bank, Washington, DC

关键词: prediction;    double sampling;    maximum likelihood;    generalized method of moment;    regression estimator;   

DOI  :  10.1596/1813-9450-6509
RP-ID  :  WPS6509

学科分类：社会科学、人文和艺术（综合）

来源: World Bank Open Knowledge Repository

【 摘 要 】

This paper considers the predictionestimator as an efficient estimator for the population mean.The study may be viewed as an earlier study that proved thatthe prediction estimator based on the iteratively weightedleast squares estimator outperforms the sample mean. Theanalysis finds that a certain moment condition must hold ingeneral for the prediction estimator based on aGeneralized-Method-of-Moment estimator to be at least asefficient as the sample mean. In an application tocost-effective double sampling, the authors show howprediction estimators may be adopted to maximize statisticalprecision (minimize financial costs) under a budgetconstraint (statistical precision constraint). This approachis particularly useful when the outcome variable of interestis expensive to observe relative to observing its covariates.