Working rules for determining the plot size and numbers of plots per block in field experiments

Abstract

SummaryWorking rules for determining plot size and number of plots within a block in field experiments are proposed based on the value of the intrablock correlation (ρ), which can be obtained from the analysis of variance of a randomized-block experiment. The method uses Binns' (1982) equation to relate this correlation to Smith's (1938) empirical law. The rules are: (1) if ρ is greater than 0·5, use an incomplete block design or reduce the plot size in order to increase the number of replications; (2) if ρ is less than 0·1, an increase in plot size is effective; (3) if ρ is between 0·1 and 0·5, an increase in plot size and a decrease in the number of plots per block may be helpful in combination.