ABSTRACT
I n this chapter it is assumed that the multiplicative interaction modelwith a single interaction component has been determined to adequately describe a given two-way treatment structure data set. Complete analyses of data sets described by such models are discussed in this chapter. The estimation of a2 is discussed in Section 3.1. Tests of hypotheses on row and column main effects are discussed in Section 3.2, and methods which show how to locate subsets of the data table which are free of interaction are given in Section 3.3. The knowledge gained in these three subsections is utilized to provide an improved estimate of cr2 in Section 3.4.
