Suppose you have income and education data for a random sample of 1000 couples. You have three education variables, the years of education of the husband (educ_husband), the years of education of the wife (educ_wife), and the total years of education in the family (educ_family), which is defined as the sum of the husband's and wife's education:
educ_family = educ_husband + educ_wife
Using this data, you have estimated the following three regression models:
Model 1:
family_income = $\beta_0 + \beta_1$educ_family + $\beta_2$educ_wife + u
Model 2:
family_income = $\beta_0 + \beta_1$educ_family + u
Model 3:
family_income = $\beta_0 + \beta_1$educ_wife + u
Now you wish to test whether education of the husband and education of the wife have the same impact on family income. Which of the following are valid test statistics for implementing this test?
(Check all that apply.)
An F-test comparing the SSR from Model 1 with the SSR from Model 3
A t-test of the statistical significance of the wife's education in Model 1
An F-test comparing the SSR from Model 1 with the SSR from Model 2
A t-test of the statistical significance of family education in Model 1