solution

 

. In the previous two exercises, you analyzed the males and females using separate multiple regressions with the high school grades as explanatory variables. Here we will run the analyses together. Recode the variable Sex into a new variable Gender that is an indicator variable. (You can do this by setting Gender equal to Sex minus 1.) Construct three interaction

variables that model the interaction between gender and each of the high school grade variables. Then run a multiple regression

using seven explanatory variables: the three high school grade variables, gender, and the three interaction variables.

(a) Report the fitted equation and the results of the significance tests for the regression coefficients.

(b) Substitute the value 0 for Gender and simplify the fitted model. This is the model for males. Verify that this is the same fitted

model that you obtained in Exercise 11.117.

(c) Repeat part (b) to obtain the results for females and verify that this is the same fitted model that you obtained in Exercise 11.118.

(d) Use software or the method for comparing the coefficients of a collection of q explanatory variables using the values of R2

(page 600) to test the null hypothesis that the three interaction terms are all zero.