Answer Happy

15. Treatment effects without randomization Suppose you want to evaluate the effectiveness of a job training program using wage = Bo+B₁program+u as a model. You take 400 employees and estimate the simple regression as: wage = 22.6-8.6 program where wage = hourly wage in dollars program = 1 if the employee participated in the training course, 0 otherwise Suppose that in addition to performing the simple regression, you perform a multiple regression with three controls that gives the following result: wage = 19.5+ 7.5program +0.9prevearn +0.85educ - 0.65 age where wage = hourly wage in dollars program = 1 if the employee participated in the training course, 0 otherwise prevearn = hourly wage in dollars prior to the training course becoming available educ = years of education age = age in years per hour. (Hint: Calculate The difference between wage and wage of an employee who participated in the training program is S wage - wage holding prevearn, educ, and age constant when computing wage.) per hour. (Hint: The difference between wage and wage of an employee who did not participate in the training program is $ Calculate wage - wage holding prevearn, educ, and age constant when computing wage.) True or False: The difference between regression results from wage = 22.6-8.6 program and those from wage = 19.5+7.5program +0.9prevearn+0.85educ-0.65 age is likely a result of nonrandom selection into the training program. True False