Estimation of mean for a finite population using sub-sampling of non-respondents

In this study a generalized estimator for the estimation of population mean assuming the presence of non-response has been advised for two different situations. Expressions of the bias and mean squared errors of the general class have been obtained. The optimum conditions under which the proposed class has minimum mean squared errors are provided. A real life example is also provided to reveal the effectiveness of the proposed general class of estimators. The study shows that the proposed class of estimators performs better than the existing estimators.