Asymptotic Efficiency of OneR-Factor Experiment Relative toROne-Factor Experiments for Selecting the Best Normal Population

Abstract

In this article, we consider two single-stage procedures for ranking means of normal populations with a common known variance, for problems involving an r-way (r≧2) classification without interaction. For the problem of selecting the best level of each of the r factors, the asymptotic (p*→7) efficiency of one r-factor experiment relative to r one-factor experiments is determined. Applications to some related problems are noted.