ABSTRACT
I n this chapter an experiment is considered where the basic treatmentstructure is three-way and there are no independent replications of these three-way treatment combinations. In order to use the methods of Chapters 1 to 3 on three-way and higher-order treatment structures, one must look at the experiment as though it is two-way. For the three-way this can be done by letting all possible combinations for two of the factors in the three-way represent one of the factors in the two-way and letting the levels of the third factor in the three-way represent the second factor in the two-way. That is, if one had a three-way treatment structure with factors A, B, and C, one could consider the experiment as a two-way with factors A*B and C, A*C and B, or A and B*C.
