Synopses & Reviews
Synopsis
With Regression Diagnostics, researchers now have an accessible explanation of the techniques needed for exploring problems that compromise a regression analysis and for determining whether certain assumptions appear reasonable. The book covers such topics as the problem of collinearity in multiple regression, dealing with outlying and influential data, non-normality of errors, non-constant error variance and the problems and opportunities presented by discrete data. In addition, sophisticated diagnostics based on maximum-likelihood methods, scores tests, and constructed variables are introduced.
Synopsis
Its principal themes, sometimes treated independently, include problem-flagging statistics, variable transformations, analytical graphics, and the spirit of Tukey's exploratory data analysis. Regression Diagnostics. . . combines these themes nicely. . . . The volume is . . . an accurate and detailed portrayal, resulting in a valuable contribution. . . . All in all, this volume is highly recommended not only for systems theorists but also for those sociologists and others desiring an accurate portrayal of feedback concepts. The book is careful and comprehensive . . . and generally brings the reader up to date on the feedback literature.
--Contemporary Sociology
This excellent, concise, and practical handling of diagnostic methods suffers in no way from its use of social-statistics illustrations. The 80 pages are as good as anything I have seen in promoting, explaining, and illustrating the diagnostic tools for regression.
--Technometrics
Linear least-squares regression analysis makes very strong assumptions about the structure of data--and, when these assumptions fail to characterize accurately the data at hand, the results of a regression analysis can be seriously misleading. With Regression Diagnostics, researchers now have an accessible explanation of the techniques needed for exploring problems that comprise a regression analysis, and for determining whether certain assumptions appear reasonable. Beginning in Chapter 2 with a review of least-squares linear regression, the book covers such topics as the problem of collinearity in multiple regression, dealing with outlying and influential data, non-normality of errors, non-constant error variance, and the problemsand opportunities presented by discrete data. In addition, sophisticated diagnostics based on maximum-likelihood methods, score tests, and constructed variables are introduced. The book concludes with suggestions on how regression diagnostic techniques can be effectively applied in research, and offers advice on implementing these suggestions through the use of standard statistical computer packages.