=+51. The article “The Analysis and Selection of Variables in Linear Regression” (Biometrics, 1976: 1–49) considered a data set of 32 observations on the following variables: y 5 fuel efficiency, x1 5 engine type

(straight or V), x2 5 number of cylinders, x3 5 transmission type (manual or automatic), x4 5 number of transmission speeds, x5 5 engine size, x6 5 horsepower, x7 5 number of carburetor barrels, x8 5 final drive ratio, x9 5 weight, and x10 5 quarter-mile time.

Use the summary information (top of page 571) on the best model of each given size to select a model, and explain the rationale for your choice.