11.8 Consider a 26 factorial experiment and suppose that the experimenter has only enough resources to handle just a fraction of all possible treatment combinations. Let this fraction be chosen by the identify relationship I=+ABCD +CDEF = +ABEF (i) Give the treatment combinations that make up this fraction. (ii) Assuming that all interactions involving 3 or more factors and all 2-factor interactions not involving factor B can be estimated from this fractional factorial. (iii) Suppose we have 2 replications (that is, 2 EUs) for each treatment com- bination in a CRD; outline the ANOVA table (giving sources of variation, d.f. and E(MS)) based on the assumptions given in (ii). (iv) Suppose we need to use blocks of size 8 and we have 4 blocks available; under the assumptions in (ii) give a suitable arrangement without sacrific- ing information about the main effects and 2-factor interactions involving factor B. (v) For the design in (iv) outline the ANOVA table, giving sources of variation and d.f. (vi) Describe how you would obtain the ANOVA table in (v) with SAS.