Optimality of novel estimators for variance estimation under successive sampling

This article presents a class of estimators allied with Searls (1964) using a novel approach to estimate population variance on two occasion under successive sampling. Isaki (1983) encourages us to address the problem of estimating finite population variance in survey sampling, its application to the case of successive sampling is highly captivating, and the methodology proposed here will be useful to those involved in similar studies in the future. To the best of our knowledge, this is any authors' first foray into this field to provide more efficient estimators than the regression type estimator. An empirical study based on real population and artificial population proves the utility of the proposed methods.