Regression-type estimation of a finite population mean in two-phase sampling using auxiliary variable and attribute

In this paper, by making regression adjustment, a class of estimators of the finite population mean under two-phase sampling is suggested which incorporates auxiliary information on quantitative and qualitative variables. Making approximation up to first order, bias and mean squared error (MSE) are obtained. A few particular cases of the estimators are discussed. The numerical and empirical comparisons of these estimators with ordinary ratio and regression estimators are carried out using a Monte Carlo simulation.