A class of estimators of the population mean in stratified random sampling

Abstract

In stratified random sampling from a finite population, when information on an auxiliary variable is available, a well known estimator for the population mean is the "combined ratio estimator" (Hansen, Hurwitz and Gurney, 1946). In this paper a class of estimators for the population mean which incorporate infor-mation on an auxiliary variable is considered. The starting point of this research are results in Ceccon, Diana and Salvan (1991) and Diana (1991). The class under study includes the "combined ratio estimator" and some other estimators proposed in the literature; see in particular Kaur (1985). Some improved estimators belonging to the class are singled out and their behaviour is discussed, with the help of an empirical illustration, with reference both to bias and to mean square error.