DESIGN AND ANALYSIS OF SERIES OF INCOMPLETE BLOCK EXPERIMENTS

Abstract

A methodology to design and analyze series of incomplete block experiments is proposed in this paper. For a simple combined analysis, a
particular basic scheme should be chosen as a starting point. Then, replicates of the experiment by independent randomizations of the basic scheme, are done on each location. The analysis requires the construction and solution of a reduced normal equations system for treatments eliminating blocks, as a simple extension of the corresponding basic design, which renders the sum of squares for treatments eliminating blocks, SC(TEB). The sum of squares due to treatments x locations interaction (SC Trat x L) is calculated by solving the reduced normal equations system for treatments eliminating blocks, for each location, so that the sum of squares for treatments on the k-th location, SCT_k; is obtained, and SC(Trat x L) corresponds to the difference between the sum of the SCT_k over all locations and SC(TEB ). The SAS programs to perform these analyses are included.