Applications of Statistics to Industrial Experimentation (Wiley Series in Probability and Statistics)

Rating:
Summary: great book by great statistical consultant
Review: Cuthbert Daniel had over thirty years of experience consulting on experimentation in a variety of industries in 1976 when the book was written. He learned statistics from many of the great statisticians who were his colleagues, including John Tukey, Stu Hunter, Fred Mosteller, Bill Cochran and Henry Scheffe. As he says, the book is concerned only with confirmatory industrial experimentation where statistical methods can play an important role. He provides sage advice on the steps in planning and choosing a design for a particular experimental situation. Topics included are the factorial designs, fractional factorials, incomplete factorials (including the popular Plackett-Burman designs), trend robust plans and nested designs including split-plot designs. There is no coverage of response surface designs or other exploratory designs.

This is a terrific book that is fun to read. The only drawback is that it predates many of the recent developments such as the robust parameter designs motivated by the work of Taguchi. For a thorough up-to-date treatment see Hamada and Wu (2000) "Experiments" and my Amazon review of it.