Topics 1/25 Lecture

1. Review of simple linear regression
2. The "regression effect" (Section 7.5.1, Sleuth). Example of father and son heights.
3. Example: Relationship between elemental carbon (Y) and mass (X) of diesel exhaust. Fitted regression line, interpretation of slope, linear correlation coefficient, R2, hypothesis test for nonzero slope of regression line.
4. Features of residual plots. Horn-shaped patterns in residuals suggest heteroskedasticity and possibly the need for transformation.
5. Confidence intervals for the mean response
6. Prediction intervals for a future observation