Abstract
Blocking of experimental units is frequently used to improve the precision of an experiment, yet many discussions of optimal plot size do not include a factor for the effect of blocking. Since there is a relationship between the choice of optimal plot size and number of plots per block, such an omission may make recommendations inappropriate for blocked experiments. An adjusting formula of H. F. Smith allowing for blocking is expressed more intuitively in terms of the correlation among plots within blocks.

This publication has 0 references indexed in Scilit: