TABLE OF CONTENTS In Chapter 5 we considered factorial experiments and gave a general method for their analysis. Among the special situations that are of considerable interest in future designs is the case of f factors in which each factor is at just two levels. These levels might be two extremes of temperature, two extremes of pressure, two time values, two machines, and so on. Although the case may seem rather trivial, since only two levels are involved, it is, nevertheless, very useful for at least two reasons: to introduce notation and concepts that will apply to more involved designs, and to illustrate the main effects and interactions really present in this simple case. It is also true that in practice many experiments are run at just two levels of each factor. For example, the ceramic tool-cutting study of Chapters 1 and 5 is a 2 2 2, or 2 3 factorial, with four observations per cell. Throughout this chapter the two levels will be considered to be fixed levels. This is quite reasonable because the two levels are chosen at points near the extremes, rather than at random.
In the simplest case, two factors are of interest and each factor is set at just two levels. This is a 2 2 = 2 2 factorial, and the design will be considered to be completely randomized. Our investigation of current flow in an integrated circuit (Section 5.1) is of this...
