On Selection Procedures for Normal Populations with Common known Coefficient of Variation with an Application to Multivariate Normal Populations.

摘要

There are instances in the biological and physical sciences in which a linear relationship between the sample means and the sample standard deviations seems to exist. It is not unreasonable in such cases to assume that the population standard deviation sigma is a multiple of the population mean Theta, i.e. sigma = (sq.rt.b) Theta, b > 0, and Theta also is assumed to be positive. The purpose of this paper is to investigate some problems of selection when the k underlying populations are normal with unknown positive means Theta sub 1,..., Theta sub k and a common known coefficient of variation sq.rt.b. Three selection rules are investigated, and some comparisons are made among these rules. (Author)