The marketing manager at Turfco, a lawn mower company, believes that monthly sales across all outlets (stores, online, etc.) are influenced by three key variables: (1) outdoor temperature, (2) advertising expenditures, and (3) promotional discounts. A portion of the monthly sales data for the past two years is shown in the accompanying table.

Sales (units)Temperature (°F)Advertising (in $1,000s)Discount (in percent)

17,235.................33............................15...............................5.0

19,854.................42............................25..............................5.0

⋮..........................⋮..............................⋮.................................⋮

22,571.................44............................21..............................5.0

a. Estimate the model Sales = β0 + β1Temperature + β2Advertising + β3Discount + ε and test for the joint and individual significance of the explanatory variables at the 5% level.

b. Examine the data for evidence of multicollinearity. Provide two reasons why it might be best to do nothing about multicollinearity in this application.

c. Examine the residual plots for evidence of heteroskedasticity.