Chapter 7 One-Way and Higher Designs for Independent Groups 371 bivariate data in columns 7 and 8 of plasma. Then the commands con=con2way(2,3) linconMpb(z,con=con$conAB) would test all hypotheses based on the linear contrast coef cients typically used when dealing with interactions. I 7.11 Nested Designs Briefly, a two-way nested design refers to a situation where there is a hierarchy among the levels of two factors under study. This is in contrast to a completely crossed design as considered in Section 7.2. For example, a goal might be to compare the efficacy of two medical procedures. The first method is used in K randomly sampled hospitals, with n participants used within each hospital, and the same is done for another K randomly sampled hospitals for the second method. For various reasons, the efficacy of a method might depend on the hospital where it is used. Here, the factor hospital is nested within the two levels