Designs involving sequences of treatments with residuals proportional to direct effects

Designs involving sequences of treatments are advantageously used when observations are taken repeatedly from same experimental unit period after period as a new treatment is applied in each period. Hence, inclusion of residual effects into the model becomes the most important feature of such designs. Six popular classes of designs involving sequences of treatments are studied under a model with carryover effects proportional to the direct effects where the unknown proportionality parameter assumed values from -1 to +1. The objective of this empirical study was to determine the value of proportionality parameter for which maximum information can be drawn from the design.