Objectives: To estimate the finite populations mean using two auxiliary variables in double sampling as well as the efficiency of the proposed class of estimators. Methods: The mixing of estimators became more popular in developing more efficient estimators for estimating finite population parameters but while mixing two or more estimators, we should consider the basic purpose and the conditions under which the individual estimators are developed and are efficient. This paper deals with a class of mixed estimators of population mean by mixing ratio estimator and dual to product estimator in two phase sampling scheme using SRSWOR scheme to select the sample units at both the cases, i.e. Case-I: When − X is unknown but − Z is known and Case-II: When both − X and − Z are unknown. The purpose of mixing these two estimators is that both these estimators are designed to be used effectively for population mean when the population correlation coefficient between the study variable and the auxiliary variable is highly positive. Results: We observe that the proposed class of estimators is more efficient than the existing estimators which are available in literature and the empirical study indicates that the proposed class of estimators t 01 and t 02 performs better than the other existing estimators of t ′

This paper deals with a class of estimators of finite population mean using a combination of two mixed classes of estimators by exploring the information on two auxiliary variables. We have assumed that the study variable y is highly correlated with both the auxiliary variables x and z. The optimum properties of the proposed class of estimators is studied both theoretically and empirically. The minimum variance bound(MVB) estimator of this class is also derived and compared with several other competing estimators in terms of its bias and percent relative efficiency.