I've read that adding an interaction to a regression model is accomplished by creating a new variable, which is the product of two existing variables. However, I'm curious whether continuous and categorical variables matter.
Suppose I have four variables, con1, con2, cat1, and cat2 (con=continuous, cat=categorical.) How would I formulate interactions between the following variable combinations, (con1, con2), (con1, cat1), and (cat1, cat2)?
