AN EFFICIENT CLASS OF CHAIN-TYPE EXPONENTIAL ESTIMATORS FOR POPULATION MEAN UNDER TWOPHASE SAMPLING SCHEME

Abstract

This paper intends to develop an efficient estimation strategy to estimate mean of a finite population using information on two auxiliary variates, which are associated with study variable y under two-phase (double) sampling setup. The properties of the proposed estimation procedure have been examined up to first order of approximations and empirical studies are performed on seven data sets to show the dominance over some contemporary estimators. A separate numerical computation is also performed to illustrate the predominance of the suggested estimation procedures for various combinations of correlation coefficients. Recommendations are also made for survey statisticians.