Introduction to linear regression analysis

This fourth edition of a regression textbook covers both standard and less common applications of linear regression, linking the methods to current mathematical and scientific research. It mixes theory with practical application so readers can learn the core principles behind building regression models across many fields. Readers should already know basic statistics, including hypothesis tests, confidence intervals, and the normal, t, chi-square, and F distributions.

The opening chapters introduce regression modeling and typical uses, then move through basic inference, checking model adequacy, and polynomial models with their variants. Later material explains how transformations and weighted least squares address model inadequacy, how to handle influential observations, and how indicator variables relate regression to analysis of variance. Further chapters cover variable selection and model-building strategies, the sources, effects, and remedies for multicollinearity, robust regression such as M-estimators, the basics of nonlinear regression, generalized linear models, and regression with SAS.

The examples draw on MINITAB, SAS, and S-PLUS, the major procedures in those packages. The text is aimed at practitioners in engineering, the physical and chemical sciences, economics, management, the life and biological sciences, and the social sciences. A companion FTP site provides data sets, detailed problem solutions, software tips, and PowerPoint slides, and the topics are arranged in increasing order of complexity.

The book is designed for upper-undergraduate and beginning graduate courses in mathematics, engineering, and the natural sciences. Scientists and engineers can also use it as a reference or for independent study.