Abstract

                                Research workers who have to deal with field trials are often faced with the problem of deciding upon the number of replications that should be adopted for a particular set of treatments. In most cases the number of replications is decided arbitrarily and this leads to a result where even large differences between treatment means fail to attain statistical significane. Any method by which the workers will be enabled to determine the number of replications necessary to render a predetermined difference, of say 10% between the treatments, to be statistically significant, would be very useful. An attempt is made in this paper to evolve such a method, utilising the data that were available from the Paddy Breeding Station, Coimbatore.

Key words :