Design & Factor Selection
Design Types & Categories
Subsidiary Factors (Noise, Outer Array) are factors that may be contributors to the response, but are not controlled in normal operations; however, they must be controlled for the experiment. When not included in the design, they contribute to error.
Subsidiary factors may not interact with main factors, but may interact with other subsidiary factors. If interaction with any other factor is to be estimated, then the interacting factors should be included as main factors. When subsidiary factors are incorporated in a design, they are usually considered as random effects during the analysis, or as noise factors.
Taguchi uses the labels Inner Array for main factors and Outer Array for subsidiary factors. The Analyst should be alert to the possibility that the effect of a subsidiary factor may be so strong as to warrant its promotion to the status of a main factor.
Subsidiary designs cause the Main Factor design to be replicated. Each subsidiary replicate is at known different levels of the Subsidiary factor(s). Ordinary replicates incorporate the effect of all factors not in the design at whatever levels they may have during the experiment.
Learn more about the DOE tools for designed experiments in Six Sigma Demystified (2011, McGraw-Hill) by Paul Keller, in his online Intro. to DOE short course (only $99) or online Advanced Topics in DOE short course (only $139), or his online Black Belt certification training course ($875).