The coefficient of determination, \(R^2\), measures how much of the variation in the response is explained by the model and is defined as

\[ R^2 = 1 - \frac{\texttt{RSS}}{\texttt{SYY}} \]

and \(0 \le R^2 \le 1\).

Q: What are some possibilities for \(\frac{\texttt{RSS}}{\texttt{SYY}}\)?

Residual plots show the residuals versus the fitted values. We can see this by looking at

\[ \texttt{my.lm <- lm( a $\sim$ b)} \]

and then plotting it with

\[ \texttt{plot(my.lm)} \]

to see if any of our assumptions do not hold.