The material in this module is aimed at students or researchers who have already had some exposure to simple linear regression and inference about the slope and intercept. It describes how transformations of the explanatory and response variables can be used to model nonlinear relationships. Topics covered include
The module works through an analysis of a particular data set that exhibits a nonlinear relationship. A series of models for the data is considered and each model is explained (and criticised!) with the help of interactive diagrams. The main exposition also includes several links to pages with 'refresher' material that allow readers to revise some basic concepts.
These pages are part of the CAST package, a complete computer-based introductory statistics textbook with over 300 interactive diagrams. Readers who have never met simple linear regression or least squares are advised to study the CAST pages about bivariate data before attempting this module.