Selection of Active Factors by Stepwise Regression in the Data Analysis of Supersaturated Design

Supersaturated design is a type of fractional factorial design in which the number of columns is greater than the number of rows. Several articles have considered methods for constructing supersaturated designs to assure a low level of nonorthogonality for all paired columns. Stepwise selection is recommended for the selection of active factors in data analysis for supersaturated design, under an assumption of the effect sparsity. This article considers selection errors of stepwise regression in the analysis of supersaturated designs. Specifically, type II error in the selection of stepwise regression is examined using simulation. Based on the results of simulation, some guidelines for data analysis of supersaturated design are discussed.