Use the data in JTRAIN3.RAW for this question.
(i) Estimate the simple regression model re78 = (0 + (1 Strain + u, and report the results in the usual form. Based on this regression, does it appear that job training, which took place in 1976 and 1977, had a positive effect on real labor earnings in 1978?
(ii) Now use the change in real labor earnings, cre = re78 - re75, as the dependent variable. (We need not difference train because we assume there was no job training prior to 1975. That is, if we define drain = train78 - train75 then ctrain = train78 because train75 - 0.) Now what is the estimated effect of training? Discuss how it compares with the estimate in part (i).
(iii) Find the 95% confidence interval for the training effect using the usual OLS standard error and the heteroskedasticity-robust standard error, and describe your findings.