On estimating the variance of the ratio estimator

Abstract

An alternative general class of estimators of the variance of the ratio estimator is considered. The class includes the classical variance estimator v₀ . Asymptotic expansions for the biases and variances of the proposed estimators are obtained and a comparison of the relative merits of the estimators v₀ v_s1 and v_s2 is made. Conditions under which each of these estimators is more efficient than the other two are derived. Among the estimators v₀ v₁ v₂ v_s1 v_s2 the results of the empirical study seem to favour the use of v_s1

Keywords:

• Asymptotic expansion
• bias
• empirical study
• optimal variance estimator
• ratio estimator
• variance