A response-based method for analyzing data from Taguchi experiments

Abstract

The engineering ideas of the Taguchi method have been widely recognized but the signal-to-noise ratio (S/N ratio) has drawn much criticism. Response surface methods do not use S/N ratios but it is not easy to build an adequate response surface model. In this paper, we propose a relatively simple method for analyzing the data from Taguchi experiments. The proposed approach directly uses the average and standard deviation of responses as performance measures. We combine the two measures into the square deviation from target and the optimal factor level is identified as the one that has the smallest deviation from target. We also introduce the concept of relative target, with which the proposed approach is applicable for all three kinds of quality characteristics without a need to execute data transformation. Two real-world examples are included to illustrate the appropriateness of the proposed approach.