Design of Experiments Using the Taguchi Approach: Sixteen Steps to Product and Process Improvement

摘要

There are numerous books and articles that discuss the Taguchi approach to experimental design. The better resources provide balanced presentations, adequately addressing the advantages and disadvantages of using the Taguchi approach. Design of Experiments Using the Taguchi Approach is a text that provides, however, a one-sided view of the Taguchi methods. The book is "designed to provide you with a comprehensive background in the design of experiments (DOE)/Taguchi technique," and the author presents the material in a very understand-able and interesting manner. As stated in the Preface, the book is written for anyone with or without a college degree, and "a formal course or background in statistics is not required." Although the book is well written, the author does not address any of the criticisms of the Taguchi approach that are well-documented in the literature. For example, see Box (1988), Box, Bisgaard, and Fung (1988). Montgomery (1999, 2001), Nair (1992), or Pignatiello and Ramberg (1992). To illustrate. the author thoroughly presents the use of and analysis of signal to noise (S/N) ratios. Four advantages/disadvantages of analyses of S/N ratios are listed. According to the author, "the only disadvantage of S/N analysis is that some numbers (which would be positive or negative) may look strange" (p. 373). The author does not address the well-documented problem that optimizing the S/N ratio (say for larger is better) does not guarantee minimizing variability nor does it always lead to the optimal settings of the variables under investigation.