Question 8 Using data from Miami University surveys of students graduating between 2013 and 2017, I estimated salary and log-salary regressions. Salary is measured in $1000. The control variables included in the regression are defined as follows: Female: dummy variable that equals one if the graduate is female; 0 otherwise. GPA: the student's grade point average at graduation. ACT: the student's ACT score when admitted to Miami. Some models also include dummy variables indicating the year of graduation which ranges between 2013 and 2017 in the sample. The dummy variable for 2013 is omitted from the regression. The total and residual sum of squares is provided at the bottom of the table for each specification estimated. Below each coefficient, the standard error of the estimate is presented in parentheses. The results are in Table 2 below. IV. Vi From specification (1), what is the t-statistics for the null hypothesis that male and female graduates have the same salary holding GPA and ACT constant? Can you reject the null at 1%? In specification (2), an interaction term between female and GPA is added to the model. If a student's GPA increases from 2.0 to 4.0, what is the predicted change in salary (in dollars, not thousand dollars) for a. Men b. Women Specification (3) updates (2) by adding dummy variables for the year of the student's graduation. In this sample, the graduation year ranged between 2013 and 2017. Using the information provided, calculate an F-statistic for the null hypothesis that salaries are equal across the 5 possible years of graduation. Based on your answer, can you reject the null that salaries are equal across the different years of graduation? Show the basis for your calculation. Use specification (3) to estimate the salary for a male who graduated in 2013 with 3.0 GPA and 25 on the ACT. Test the null hypothesis that the effect of grades on salary is identical for men and women. Conclude your hypothesis test at 10%. In specification (5), the square of a person's graduating GPA is added to the regression. Given that GPA can range from 0 to 4. What GPA maximizes salary for a A male graduate? b. A female graduate?