Two New Estimators for a Normal Coefficient of Variation based on Ranked Set Sampling

Abstract EN

This article addresses the estimation for the coefficient of variation (CV) of the normal distribution considering ranked set sampling.

Using EM algorithm and linear interpolation technique, two new estimators for the CV are proposed.

The finite sample size properties of the proposed estimators are examined by simulation studies in terms of their relative efficiency, Pitman measures of closeness and bias criteria.

It turns out that the proposed estimators are substantially more efficient and less sensitive to the perfectness assumption than those recently suggested in the literature.

Also it was verified the superiority of the proposed estimators using empirical data set.