Two-level compromise designs for estimating main effects and detecting interactions

This report describes a procedure for constructing experimental designs for use in estimating main effects, and detecting the presence of interactions in factorial models. D-optimality and tr(L)-optimality are the two design selection criteria used in constructing these compromise designs. A catalogue of compromise designs for k = 4 through 9 and n = k + 2 through 2k + 1 is given in Appendix II, where k is the number of experimental factors and n is the number of experimental runs.