a) Do there appear to be important regional differences? Use an appropriate hypothesis test to explain your answer.
b) Explain why the regressor West is omitted from the regression?
c) Juanita is a 28-year-old female college graduate from the South. Molly is a 28-year-old female college graduate from the West. Jennifer is a 28-year-old female college graduate from the Midwest.
(i) Construct a 95% confidence interval for the difference in expected earnings
between Juanita and Molly.
(ii) Explain how you would construct a 95% confidence interval for the difference in expected earnings between Juanita and Jennifer. (Hint: What would happen if you included West and excluded Midwest from the regression?)
Extracted text: Question 5 The data set (2012 Current Population Survey, U.S.) used for this question consists of information on 7440 observations of full-time, full-year workers. The highest educational achievement for each worker was either a high school diploma or a bachelor's degree. The workers' ages ranged from 25 to 34 years. The data set also contains information on the region of the country where the person lived. The following variables are included in the data set: AHE = average hourly earnings (in 2012 US dollars, S) = binary variable (1 if college. O high school) = binary variable (1 if female, O if male) = age (in years) College Female Age Northeast - binary variable (1 Region = Northeast, 0 otherwise) Midwest = binary variable (1 if Region = Midwest, 0 otherwise) = binary variable (1 i Region = South, 0 otherwise) = binary variable (1 i Region = West, O otherwise) South West The following regression results are obtained based on the above data set: Regressor College (X,) Dependent variable: average hourly earnings (AHE) (1) 8.31 (2) (3) 8.32 8.34
Extracted text: The following regression results are obtained based on the above data set: Dependent variable: average hourly earnings (AHE) Regressor College (X,) (1) (2) (3) 8.31 (0.23) 8.34 (0.22) 8.32 (0.22) Female (X,) -3.85 -3.81 (0.22) 0.51 (0.04) -3.80 (0.23) (0.22) Age (X,) 0.51 (0.04) Northeast (X,) 0.18 (0.36) Midwest (X,) -1.23 (0.31) South (X,) -0.43 (0.30) 2.05 Constant 17.02 1.87 (0.17) (1.18) (1.18) Summary Statistics and Joint Tests F-statistic for regional effect = 0 Standard error of regression 7.38 9.79 9.68 9.67 0.162 0.180 0.182 N 7440 7440 7440 Note: Standard errors in parentheses.