Generalized Family of Efficient Estimators of Population Median Using Two-Phase Sampling Design

For estimating the population median of variable under study, a generalized family of efficient estimators has been proposed by using prior information of population parameters based upon auxiliary variables under two-phase simple random sampling design. The comparison of proposed family of estimators has been made with the existing ones with respect to their mean square errors and biases. It has been shown that efficient estimators can be obtained from the family under the given practical situations which will have smaller mean square error than the linear regression type estimator, Singh et al estimator (2006), Gupta et al (2008) estimator and Jhajj et al estimator (2014). Effort has been made to illustrate the results numerically as well as graphically by taking some empirical populations considered in the literature which also show that bias of efficient estimators obtained from the proposed family of estimators is smaller than other considered estimators.