Blocking [Spu(2)] Factorial Treatment Structures

The primary advantage of conducting a 23 experiment according to the plan in Table 8.1 when there is no real three-factor interaction is that one is still able to measure all of the main effects and all two-factor interactions. Furthermore, one is usually able to measure these effects more accurately since the experimental error variance is usually reduced by blocking the experimental units into subgroups of more homogeneous units.