Lecture42 (Data2Decision) Multiple Regression

Intro to multiple regression, interactions, multicollinearity, model building.

[FRANCESCO-wjrs]

If I have three groups, how do I proceed with a post-hoc test to check slope differences? 

Can I use the result from lm in R, for example? And then switch the references to match all pairwise comparisons.

For example, if I have three groups: I can make the first lm (y ~ a*b). And then see if group 2 and 3 differ from 1, directly from the significance level of the interactions term.

Then, I change the reference for level 2. And then observe if it differs significantly from 1 and 3, etc.