Kirstine Smith developed forms of these things that minimized the largest value in the hat matrix's diagonal, satisfying the criterion of G-optimality. The Paley construction for Hadamard matrices was used to generate the set of orthogonal Plackett-Burman designs for these things. The analysis of split-plot examples of these things requires two error terms. The resolution of these things indicates the degree of aliasing between main effects and interactions. Ronald Fisher promoted the use of Latin squares in designing these things when there are two blocking factors. These things are full factorial when they cross all possible factor combinations, (-5[1])and they are balanced when all combinations have equal numbers (-5[1])of replicates. (-5[1])For 10 points, name these procedures that use randomization to isolate the causal (10[1])effects of treatments. (10[2]0[1])■END■ (10[2]0[4])