The Polk Utility Corporation is developing a multiple regression model that it plans to use to predict customers’ utility usage. The analyst currently has three quantitative variables (x1, x2, and x3) in the model, but she is dissatisfied with the R -squared and the estimate of the standard deviation of the model’s error. Two variables she thinks might be useful are whether the house has a gas water heater or an electric water heater and whether the house was constructed after the 1974 energy crisis or before.
Provide the model she should use to predict customers’ utility usage. Specify the dummy variables to be used, the values these variables could assume, and what each value will represent.