Statistics For Experimenters

Box, George E. P., J. Stuart Hunter, and W. G. Hunter (1978) Statistics for Experimenters: An Introduction to Design, Data Analysis and Model Building (ISBN 0-471-09315-7)

Statistics for Experimenters is a very readable book on statistical methods and experimentation - much more so than many DOE books. Practical advice and gems of wisdom are woven throughout the pages of the book. The analysis and modeling methodologies are all placed within the context of statistical ("critical") thinking and the iterative learning process of induction and deduction -- something that does not occur anywhere else that I know of. As stated in the Preface, "Actual problems are never so straightforward that they can safely be solved mechanically. Therefore, we will emphasize the necessity for thinking about the real nature of the scientific problem itself ... as well as for understanding potentially useful statistical principles and their practical consequences." Box, Hunter and Hunter provide the foundation for our focus on critical thinking (induction and deduction) and the use of Thought Maps (as opposed to roadmaps) to guide the many phases of project work and learning. Statistics for Experimenters is a timeless book that should be used as a guide prior to experimentation, a reference for DOE analysis, and a source of insights and ideas once DOE results are had. If one were limited to owning only one book on design of experiments, Statistics for Experimenters should be the one. Reviewed by Cheryl Hild