Abstract

Assuming the presence of first order residual effects, besides periods, experimental units and direct effects of treatments, a class of minimal strongly balanced changeover designs has been proposed. For v treatments, the designs require 2v experimental units for v/2 periods. It is seen that in terms of variances of estimated elementary contrasts in treatment effects, these designs are partially balanced based on the circular association scheme. The efficiency factors for the estimation of various treatment effects of the proposed designs relative to the orthogonal designs have been tabulated for v £ 24.

Key words: Minimal strongly balanced designs, changeover design, crossover designs, repeated measurements designs, residual effects, partially balanced changeover designs.