Question #2: A large land developer had just completed the division of a parcel of land into 500 lots. From his experience, he established selling prices for 20 lots and asked his daughter Jane to set prices for the remaining lots. Jane, a recent M.B.A. from NYIT, knew that size, view, slope, and elevation are the only four variables, which could influence the price of a lot. Both Jane and her father tried to quantify the impact of these variables on the price of a lot, without any success. Jane decided to use Multiple Regression, a technique she had learnt from her favorite professor in the M.B.A. program. She collected data (Area in thousands of square feet, Elevation in feet, Slope in degrees, View scale 1 for poor up to 9 for excellent, and price in thousands of dollars) for the 20 lots priced by her father and did four regression runs as shown in the following page. She picked the best regression equation and priced the remaining lots. Her father looked at her prices and agreed that she had done an excellent job.

(a) What is wrong with the regression runs she did not use? Jane used four regression runs, but only one of them was selected as the best equation. To determine the equation used by her, we would need to see the details of the four regression runs. If you can provide the regression output or the equations used in each run, I can help you identify the equation selected by Jane.

(b) Specify the equation used by her? Why?

(c) Which variables are significant in determining the price of a lot? Which is most significant? (d) If she were to set a price with 99% confidence on a lot with excellent view, 10o slope, 20,000 square feet area and 200 feet elevation, what will it be?

(e) Can she use this equation to set prices of lots in other developments?