1. How do the concepts R^2 and correlation relate?
2. In the absence of information about X, what is the best model we could use to estimate reality?
3. What does the error term represnt conceptually? What does a model equation look like with an error term included (both linear and quadratic)? Do we want error term to be large or small? Why?
4. Imagine we were studying the relationship between coffee consumption and happiness. In simple English, what does a positive linear relationship between the variables mean? A positive quadratic relationship? A positive j-shaped relationship (containing both a linear and quadratic term)? Which model would you guess describes the real relationship between variables, just based on intuition?
5. What does error variance tell us? What doesn't it tell us?
6. Why do we calculate R^2? What does R^2 of 0.88 tell us? Would you say that a model with that R^2 is a good representation of reality?