A note on "Planning an experiment in the company of measurement error" by Levin and Subkoviak

Abstract

Levin and Subkoviak (1977) contend that experimenters should consider the reliability of their measures when comparing the power of fixed-effects completely randomized designs with that of randomized block designs. The two major purposes of this note are (1) to show that under the conditions specified by Levin and Subkoviak, it is not necessary to consider the reliabilities when comparing these two designs and (2) to identify certain errors in the illustrative example used by Levin and Subkoviak.