Размещено 2 года назад по предмету
Экономика
от lila_ch
Evaluating a Job Training Program: The data on male workers were used to evaluate a job training
program, where the variable we would like to explain, ???? = ????????????????98 is labor market earnings (in
thousands of dollars) in 1998, the year following the job training program. The variable ????????????????n is the
binary participation indicator, where ???? = 1 means that the male worker participated the program,
and ???? = 0 means he did not. Other variables are ????????????????96, the earnings in 1996 (the year prior to
the program); ????????????????, years of schooling; ????????????; and ????????????????????????????, marital status (married). Like the
????????????????n, ???????????????????????????? is a binary variable, where ???????????????????????????? = 1 means the man is married.
The simple regression estimates are
????????????????̂98 = 10.61 − 2.05???????????????????? (I)
n = 1,130, ????
2 = 0.016
We add the four controls and perform a multiple regression:
????????????????̂98 = 4.67 + 2.41????????????????????+. 37????????????????96+. 36???????????????? − .18???????????? + 2.48 ???????????????????????????? (II)
n = 1,130, ????
2 = 0.405
Questions: (15 points)
(1) Which model is better, (I) or (II)? Why? (5 points)
(2) Interpret the estimates 2.41 on ????????????????????. (5 points)
(3) Holding all other factors fixed, what is the change in ????????????????̂98 if a male worker increases one
more year of education and gets one year older? (5 points