2. Refer to Table 9.1, excluding the observation for DC Let predictors in that table (excluding violent crime rate), the bivariate model has P-value below 05 = muider rate. For the five test of independence for the

a) Fit the multiple regression model, using all five predictors. Are the P-values for the partial tests all significant? Explain why results of these tests may differ from those in the separate tests of independence

b) Use backward climination to select a model. deleting a variable if it does not have P- value below 10 Interpret.

c) Use forward selection to select a model. adding a vanable it it has P-value below 10. Interpret.

d) Use stepwise regression with the .10 level for significance. Interpret.

e) Compare results of the thee selection procedures How is it possible that a vanable (percent with a high school education) can be the first variable dropped in

(b) yet the sec- oid added in (e)