The Choice Of A Model

Abstract

Abstract Optimum experimental designs depend upon the model or models to be fitted to the data, although not usually, for linear models, on the values of the parameters of the models. This chapter is intended to give some advice on the choice of an appropriate form for a model. Whether or not the choice is correct can, of course, only be determined by analysis of the experimental results. But, as we shall see, designs that are optimum for one model are often almost optimum for a wide class of models.