A Class of Estimators for Population Variance in Two Phase Sampling

In this paper we have suggested a general procedure for estimating the population variance in two-phase sampling through defining a class of chain-type estimators. A large number of estimators are identified as members of the suggested family. Asymptotic expression for mean squared error of the proposed class of estimators is derived. Optimum conditions are obtained under which the proposed families of estimators have the minimum mean squared error. The proposed class of estimators has been compared with some contemporary estimators of population variance. The performances of the proposed class of estimators have been supported with suitable numerical illustrations and suitable recommendations are made.