ABSTRACT
In previous chapters, where the principles of designing experiments and several different designs were introduced, the focus was on understanding and specifying the structure of the experimental units, i.e. the structural component of the model. In all of these situations, the explanatory component of the model, consisting of two or more treatment groups, has been represented using a single factor. Testing the null hypothesis of ANOVA allows us to answer the broad question of whether the treatments differ from one another, but usually we are interested in more structured comparisons between treatments, and investigating these is the subject of this chapter.
