Practice work of Econometrics 60%
Problem 1: 30%
Use the data in GPAforEXAM.dta to answer these questions. It is a sample of Michigan State
University undergraduates from the mid-1990s, and includes current college GPA (colGPA), high
school GPA(hsGPA), ACT test score and a binary variable indicating whether the student owned a
personal computer (PC).
(i) How many students are in the sample (3%)? How many students owned their own PC (3%)?
(ii) Find the sample average (3%) and sample deviation (3%) of college GPAs.
(iii) Estimate the simple regression equation ，and report your estimates for
and (3%). Interpret these estimates of (3%).
(iv) Run the regression colGPA on PC, hsGPA, and ACT and report it in appropriate form (3%). Are
PC hsGPA and ACT jointly statistically significant at the 5% level, why(3%)?
(v) What is the R-squared from the regression (iv) (3%)? What do you make of its magnitude(3%)?
Problem 2: 30%
Use the data in CEOSALforEXAM to answer this question.
(i) Draw a scatter plot of lsalary against lsales, and the fitted regression line on the same graph. (6%)
(ii) Estimate the model: . Report the results
in the usual form with the usual OLS and heteroskedasticity-robust standard errors (3%).
Compute RESET test to see whether polynomials are missed in this OLS result (3%).
(iii) Apply the White test for heteroskedasticity. What is the test result (3%)? What do you conclude
(3%)?
(iv) Estimate the equation in part (i) by LAD(3%). Explain the estimate of b1 compared with the OLS
estimate (3%)?
(v) Is there any evidence that the more advanced degree the CEO gets the more Salary he earns?
Construct your hypothesis, run the test and explain (6%). (Hint: consider the indicators of
college and grad.)