Statistical Distributions in Engineering

The second revision of this popular book [originally published in 1978, with Ž rst revision in 1989 reviewed in Technometrics by Walker (1989)] updates the previous edition in the following ways: (1) an additional author, (2) the deletion of chapters on discriminant analysis, factor analysis, and categorical data analysis, (3) reorganization of material on methods requiring maximum likelihood estimation, (4) the addition of a chapter on the analysis of repeated measures, and (5) the inclusion of SAS computer code and output. Electronic data for many problems, in ASCII, StataQuest, Minitab, and SAS formats, comes in a 3.5"  oppy disk. The title word multivariable accurately re ects that the text is devoted to univariate regression analyses, in contrast to multivariate regression analyses (i.e., more than response). The level of the text has been preserved from earlier editions and is somewhat akin to the applied linear statistics and regression works by Kutner, Nachtschiem, Wasserman, and Neter (1996). Methods are presented and developed without the usually requisite matrix algebra, and only a meager amount of calculus experience is needed. Most examples and problems arise from the health and social sciences, but a smattering involving manufacturing, Ž nance, and marketing are included. The book would be most appropriate for upper-level undergraduates in applied statistics or as a survey course for graduate-level students in the medical, social, or biological sciences. Indeed, there is so much material covering adequate levels of breadth and depth in applied regression, that there’s something for nearly everyone. Most chapters are preceded by brief previews and concluded by numerous interesting exercises (odds answered in the back) and a brief (albeit a bit arcane and dated) reference list. As for organization, the chapters are arranged thusly (this reviewer’s opinions added):