Evaluating multiple regression models, not-for-profit (continuation ofProblem 10-35).

(Chapter Appendix)

Required 1. Given your findings in Problem 10-35, should Hanks use multiple regression analysis to better understand the cost drivers of overhead costs? Explain your answer.

2. Hanks decides that the simple regression analysis in Problem 10-35 should be extended to a multiple regression analysis. $he finds the following result:

Regression 3. Overhead costs = a + (bl X number of academic programs) 4- (b2 X numher of enrolled students)
Variable Coefficient Standard Error t-Value Constant $3,335.54 $4,344.06 0.77 Independent variable 1:
number of academic programs $ 214.04 $ 61.84 3.46 Independent variable 2:
number of enrolled students $ 2.24 $ 1.11 2.02 r2 = 0.81; Durbin-Watson statistic —
The adjusted R2 = 0.766 1.91.
The coefficient of correlation between number of academic programs and number of students is 0.60. Use the format in Exhibit 10-22 to evaluate the multiple regression model.

(Assume linearity, and constant variance and normality of residuals.) Should Hanks choose the multiple regression model over the two simple regression models of Problem 10-35?

3. How might the president of Eastern University use these regression results to manage overhead costs?