A class of hybrid type estimators for variance of a finite population in simple random sampling.

In this paper, a class of hybrid type estimators is proposed in estimating the finite population variance using single auxiliary variable in simple random sampling. Expressions for the bias and the mean square error (MSE) are derived up to the first order of approximation. Theoretically comparisons are provided to show that the proposed estimators perform more efficiently than various existing estimators. Empirical study and simulation results are carried out to confirm numerically that the proposed estimators are more efficient than the usual unbiased sample variance estimator, ratio and linear regression estimators by Isaki, Singh et al., and Shabbir and Gupta estimators. [ABSTRACT FROM AUTHOR]