Tutorial: Using DOE to Identify Main Effects and Interactions
Abstract: Design of experiments (DOE) is a powerful statistical tool that allows maximum learning with minimum experimentation. The traditional approach to experimental design—varying one factor at a time and observing the results—can produce a large number of experimental runs while failing to reveal important interactions. As this discussion illustrates, interactions are best found by varying more than one input at a time in an experiment. A companion template is available at www.asq.org/sixsigma/tools-exchange/docs/3-factor-doe-template.xls.
Keywords: Interactions; Main Effects; Balanced Design; Quality Tools;