7.9.16. On Page 204, Bowerman et al. (2005) present a dataset concerning sales prices of houses in a city in Ohio. The variables are: y is the sale price in $10,000; x1 is the total square footage; x2 is the number of rooms; x3 is the number of bedrooms; and x4 is the age of the house at the time data were collected. The sample size is n = 63. For the reader’s convenien ce the data are in the dataset homesales. Consider the linear model y = α + 4i=1 xiβi + e.
- (a) Use the Hogg-type adaptive scheme discussed in Section 7.6 on these data; i.e., use the function adaptor. Which score function did it select?
- (b) Comment on the estimated regression coefficients as to significance and what they mean in terms of the problem.
- (c) Fit the model using the selected score function in Part (a).
- (d) Using the Studentized residuals from Part (b), perform a residual analysis which includes at least a residual plot and a q−q plot. Identify all outliers.
- (e) Based on your analysis in (d), what, if any, other models would you fit to these data?
