Refer to the Business Week magazine data in Problem 13 of Chapter 8. The relationships among the yield (Y), 1989 rank (X2), and P-E ratio (X3) were studied in that problem using a random sample of data from the magazine's compilation of information on the top 1,000 companies. The results of the regression of yield on 1989 rank, P-E ratio, and 1990 rank (X1) are presented next. Use this output and the output from Problem 13 in Chapter 8 to answer the following questions.
a. Test whether 1989 rank, P-E ratio, and 1990 rank, taken together, significantly help to predict the yield.
b. Test whether 1990 rank is associated with the yield after taking into account the contribution of 1989 rank and P-E ratio.
c. Should 1990 rank be included in the model to control for confounding? Explain your answer, including discussion of any assumptions you made in reaching your answer.

Transcribed Image Text:

Regression of Y on X1, X2, and X3 DF Sum of Squares Mean Square F Value Pr> F Source Model Error Corrected Total 29.97725832 18.10303668 19 48.08029500 9.99241944 8.83 0.0011 16 .13143979 RSquare Coeff Var Root MSE Y Mean 0.623483 42.79588 0636922.485500 Source DF Type I SS X1 X2 X3 Mean SquaeVal 182894871 0.00159198 28.14671763 1.82894871 0.00159198 28.14671763 162 0:2218 0.00 09705 24.88 0,0001 1 Source DF Type III SS X1 X2 X3 Mean Square F Value PrS F 3.39453697 1.35441258 1 28.14671763 3.39453697 1.35441258 28.14671763 3.00 0.1025 120 0.2901 24.88 0.0001 Estimate 7.421187723 -0.013351684 Standard Error t ValuePr> I Parameter Intercept X1 X2 X3 0.98839344 7.51 ,0001 -0 01335184 00770835 173 1025 109 0.2901 0007620880 0.00696539 0.261846279 005249866 4.99 0.0001