In this talk, guests will introduce a new approach to building supersaturated designs (SSDs). This method will divide the screening factors into several groups, and ensure that factors in different groups are orthogonal to each other, and factors in the same group are weakly correlated. At the same time, the first group is designated as a pseudo factor group, so as to obtain an unbiased estimate of the error variance, and Developing effective, design-based model selection procedures. Simulation results show that these designs, used together with our model selection procedure, can identify more active main effects than previous supersaturated designs. These designs and their automated analysis are new in JMP 15. This lecture will provide a demonstration of the GO SSD method. Using 12 experiments and a 4-block orthogonal supersaturated design method, the five most significant factors can be obtained.
