ABSTRACT
In Chapter 20 the point was made that several statistical tests, such as ANOVA and regression, are all examples of the general linear model (GLM). However, they have certain assumptions which make their use inappropriate when data do not conform to those assumptions: for example, when the outcome measure or dependent variable is on a categorical scale, such as pass or fail on an aptitude test. To cope with a wider range of data than is covered by the GLM, statisticians have devised an extension of it, which is slightly confusingly called the generalised linear model.
Logistic regression is in the family of tests which conform to the generalised linear model and can be used to analyse data with a categorical outcome variable.
