Researchers from many fields can benefit from applied knowledge of general linear models. This class of models includes the t-test (paired and two sample), regression, ANOVA, and ANCOVA. Like all statistical methods, certain assumptions must hold in order for these models to be successfully implemented. What happens when one or more of these assumptions are violated by the data? In this short course, we discuss the primary assumptions required for general linear models and present methods for assessing the validity of each one. We will discuss methods that address violations in these assumptions (e.g. Box-Cox transformations, weighted least squares). Alternative modeling strategies, such as semiparametric and nonparametric methods, may eliminate the need for some assumptions and will be briefly introduced.
Prerequisite knowledge for this course is a familiarity with general linear models, including ANOVA and ANCOVA. A variety of datasets will be used to illustrate the testing of assumptions as well as alternative procedures. The data will be analyzed using R.