Introducing the generalized linear model: general linear model

This chapter reviews the generalized linear model (GLZM), which is an extremely useful and increasingly popular framework approach to analysing data. Since it relies on making assumptions about the distribution of data, it is parametric. In particular, the chapter looks at the general linear model (GLM), a sub-framework of the generalized linear model which uses the normal distribution. GLM can be used to conduct analyses that parallel all the most commonly used parametric null hypothesis significance tests (NHST), such as t-tests, Anova, and linear regression. In other words, GLM can be used to test for differences between two or more samples and for relationships between two samples. GLM can also be used to do both at the same time, and it can handle more than two explanatory variables. The chapter then outlines the multiple model approach to analysis before introducing alternatives to a GLM within the GLZM framework.