Extended Complete Block Designs Generated by BIBD

Abstract

In many biological and industrial experiments the number of experimental units and their groupings into blocks are such that the experimenter cannot influence them without sacrifice of experimental purpose or material. This paper is directed to those experimental situations in which block size is constant and exceeds the number of treatments; the cost of experimental units is high and this cost is principally associated with blocks as a whole; and a balanced design is requlired. Tests of pertinent hypotheses are developed, easily applied expressions for estimates of treatment effects, and their variances are found. The selection of the most effective designs for particular experimental situations is considered and tables of optimal designs for these are presented.