The value of a uniformity trial in field experimentation with rubber

Abstract

1. Sanders’ method of utilising previous crop records to correct experimental results by means of a linear regression is briefly described. 2. The method is applied to yield figures from rubber trees in Sumatra, and the precision of a dummy experiment is thereby increased nearly fourfold when the “preliminary” and “experimental” years are consecutive. When the “experimental” year is three years later than the “preliminary” year, the error variance of the adjusted yields is reduced to about a half. 3. It is concluded that not only has the method of correction been of value in the particular instance investigated, but that a uniformity trial utilised in this way should be of practical value in any major field experiments with rubber.