The operating window method is a novel tool in quality engineering. It has been successfully practiced in some industrial sectors but has received scant attention in academic research. If a critical factor for competing
failure modes can be identified, the probability of failures can be reduced by widening the operating window of this factor, thus improving the system's quality and reliability. In this research a comprehensive and rigorous foundation is given for the operating window method. A general performance measure is derived which include the operating window SN ratio as a special case. New strategies for the design and analysis of experiments are proposed, which improves over the existing practice. Extending the ideas of operating window we develop a more general
framework called "Failure Amplification Method" for categorical response optimization. Illustration is given with the analysis of a paper feeder experiment at Fuji Xerox.
