×

Quality by experimental design. 2nd ed., rev. a. exp. (English) Zbl 0927.62115

Quality and Reliability. 43. New York, NY: Marcel Dekker, Inc. 504 p. (1999).
This book is supposedly aimed at non-statisticians and is written in a chatty style. It is organized in five parts: the philosophy of experimentation, experimental design, sorting the signal from the noise, empirical model-fitting, and using empirical equations. Most of the BASIC programs that were in the first edition have been removed, and more space has been devoted to exposition and to additional topics such as split-plot designs.
The author’s enthusiasm for the subject is evident and, in places, one finds some really good beginnings at explaining principles and techniques. Much of it is correct; but far too often, the subtleties of designs and statistical models are apparently either of no consequence to the author, or beyond his reach. The non-statisticians to whom the book is directed are typically engineers or managers in manufacturing companies who could benefit greatly from using experimental methods in their work. The author is very good at explaining the benefits; but for implementing the techniques, that audience needs clear explanations and well-defined steps. Instead, we have stream-of-consciousness blabber and abundant use of mathematical notation (much of it poorly chosen and with lousy typography) that will confuse more than enlighten these readers.
In summary, I would say that this book far too technical to be of use to non-statisticians, and far too lacking in authority to be of use to statisticians. I cannot imagine using this book as a textbook, but I can imagine using the examples and case studies therein to supplement a course based on a more competently written text. Some of the examples are excellent; others are good illustrations of what not to do.

MSC:

62P30 Applications of statistics in engineering and industry; control charts
62K99 Design of statistical experiments
62-01 Introductory exposition (textbooks, tutorial papers, etc.) pertaining to statistics
PDFBibTeX XMLCite