Key points are not available for this paper at this time.
Abstract An estimator, γ, of μ is developed which utilizes a priori information concerning the value of the population coefficient of variation. γ′ is shown to have a smaller mean squared error than the sample mean, γ. A table of relative efficiencies for various coefficients of variation and sample sizes is presented.
Donald T. Searls (Tue,) studied this question.