Alright, done with day 3! Today we talked about hierarchical regression, logistic regression, moderation, and polynomial regression. We discussed the conceptual background for these concepts and then demonstrated how to do these analyses in R. Please see the attached file for screenshots of the whiteboard content we reviewed today.

Some key points: 

• Hierarchical regression is about looking for a change in the amount of variance accounted for by each subsequent model (model n+1 should explain more variance than model n)
• Polynomial regression and moderation analysis are based in hierarchical regression
• Logistic regression is for when our outcome is dichotomous and uses odds ratios (i.e., when X increases, the odds of Y decrease, etc.)
• When doing hierarchical regression, there is no set rule on when to stop, and it is open to interpretation (consider statistical significance vs practical significance)

Some resources: 

• On multicollinearity
• More on multicollinearity