ABSTRACT
This chapter introduces multiple regression, a way of constructing descriptive models for how the mean of a measurement is associated with more than one predictor variable. It aims to deal with simple confounds, using multiple regression. The chapter introduces graphical causal models as a way to design and interpret regression models. Causal inference always depends upon unverifiable assumptions. Linear regression models do all of the simultaneous measurement with a very specific additive model of how the variables relate to one another. But predictor variables can be related to one another in non-additive ways. A common question for statistical methods is to what extent an outcome changes as a result of presence or absence of a category.
