Multiple regression models based on coded and un-coded data. Multiple regression models for output variables measured in different units. Lack of fit. Canonical path.

Determine when regression analysis is the appropriate technique in analyzing an experiment. Check the important assumptions underlying regression analysis.
Read and interpret the results of the R output. Study the graphical presentation of the surface associated to the model.

Matrix equation equivalent to a second-order regression model. Canonical analysis (Linear Algebra comes to the rescue). Surface associated to the second-order regression model on the coordinate system of eigenvectors.