STAT Question

Mid-Term

Take-Home Problem

In this take-home problem we will be focusing on expanding on Exercise 6, page 177 – 178 in your Real Econometrics, edition 2 textbook. You are to use STATA for this exercise. The data set you will be using is globaled.dta. The dependent variable here is ypcgr with the explanatory variables being listed in Table 5.14 of your textbook. (We disregard the variables “name” and “code” as they are no way beneficial in modeling the variation in ypcgr.) Each of the following parts of this Take-Home problem are worth 10 points for a total of 60 points.

a) Estimate a standardized regression model with ypcgr as the dependent variable and explanatory variables edavg, testavg, and ypc60. Don’t forget to suppress the intercept (constant) in your regression. Report your regression here. Interpret the coefficients of your regression model. Which has a greater effect on growth? Years in School or Test Scores?

b) Use the “test” statement in STATA to determine if the effect of Years in School is equal to the effect of Test Scores on growth. Report your results here. State your null and alternative hypotheses of this test and your conclusion.

c) Here we are going to replicate the results of part b) by using the restricted regression and unrestricted regression F-statistic based on R-square measures. First, report the restricted regression that is implied by the null hypothesis of part b). (Remember you need to suppress the constant term in your restricted regression.) You have already run the unrestricted regression in part a). Explicitly show me how you would calculate the F-statistic reported in part b).

d) Now add the standardized versions of the variables proprts and open to your regression of part a) and run it being sure to suppress the intercept. Report your regression equation here. Which is more important, property rights or the openness of the economy in promoting growth? Explain your answer.

e) Is the adjusted R-square of the model in part d) greater or less than the adjusted R-square of the model of part a)? Which model do you prefer between the model of part a) or the model of part d)? Explain your answer.

f) Report the STATA code you used to complete this problem.