To resolve issues related to the limitations imposed by response surface methodology, it is prudent to add the robustness. This valuable book with the title ‘Robust Response Surfaces, Regression, and Positive Data Analyses’ by Rabindra Nath Das has compiled so far known concepts and tools about adding robustness in the data analysis in using robust response surface applications. A basic knowledge of regression methodologies will ease the comprehension of the contents in the book. The book emphasizes the design issues, rotatability types and optimality, estimating model parameters and replicated measures.
In essence, there are 10 well written chapters covering topics such as motivation for robustness ideas, basics of response surfaces, brief review of literature about the research direction of response surfaces, first-order correlated model, rotatable and optimum designs, inter-class structures, compound symmetry structures, tri-diagonal structure, second order designs, intra-class versus inter-class structures, robust regression designs for non-normal data, weakly robust rotatable designs, slope rotatability, balanced designs, maximum directional variance, correlated errors, positive data analysis for log-normal or gamma models, Taguchi approach and resistivity. The appendix displays the datasets in full.
The examples in illustrations are real life oriented such as infant mortality, hydrology and human blood biochemicals. The bibliography is exhaustive and up-to-date. The book is quite useful to statisticians, quality engineers, medical professionals, economists and researchers in response surfaces. The book is suitable to be a text book in graduate level course to teach response surface methodologies. I highly recommend this book.