Refer to Problem 7 of Chapter 9 to answer the following questions about the relationship of homicide rate (Y) to city population size (X1), percentage of families with yearly incomes less than $5,000 (X2), and unemployment rate (X3).
a. Determine the squared partial correlations r2YX1|X3 and r2YX2|X3 using the computer output here. Check the computation of r2YX2|X3 by means of an alternative formula, using the information that rYX2 = 0.8398, rYX3 = 0.8648, and rX2X3 = 0.8154.
b. Based on the results you obtained in part (a), which variable (if any) should next be considered for entry into the model given that X3 is already in the model?
c. Test H0: pYX2|X3 = 0 using the t test described in Problem 2.
d. Determine the squared partial correlation r2YX1|X1,X3 from the output here and/or in Problem 7 of Chapter 9, and then test H0: pYX1|X2,X3 = 0.
e. Determine the squared multiple partial correlation r2Y(X1,X2)|X3, and test H0: pY(X1,X2)|X3 = 0.
f. Based on the results you obtained in parts (a) through (e), which variables would you include in your final regression model? Use α = .05.

Transcribed Image Text:

Regression of Y on X3, X2, and X1 ANALYSIS OF VARIANCE Sum of Mean DFI Squares Square F Value Pr>F Source Model Error Corrected Total 3 16 19 1518.14494 506.04831 24.02 0001 337.05706 21.06607 1855.20200 Root MSE Dependent Mean Coeff Var 4.58978 20.57000 22.31297 R-Square 0.8183 Adj R-Sq 0.7843 PARAMETER ESTIMATES Squared Partial Squared Partial Standard Error Parameter DF Estimate t Value Pr>t Corr Type Corr Type Il Variable Intercept X3 X2 X1 -36.76493 4.71982 1.19217 0.00076294 7.01093 1.53048 0.56165 0.00063630 524 <0001 308 0.0071 2.12 0.0497 1.20 0 2480 0.74795 0.21441 0.08244 037280 0 21972 0.08244