Related questions
Problem 1-
Experiments/Randomized Control Trial
Suppose you are interested in studying the effect of being a part of the labor union on an
individual’s hourly wage.
You collect data on 1000 people and run the following regression.
where is a dummy variable which is equal t to 1 for people who are a part of labor union and
0 for others.
a. Suppose you estimate . Interpret
b. Do you think is biased or unbiased? Explain.
Now suppose you conduct a randomized control trial to answer the same question. You
randomly assign some individuals to be a part of the labor union and others to not be a part
of the labor union. The first step you take is to ensure that the randomization was done
correctly. Then you estimate the following equation:
=1 if the student is assigned to be a part of Labor Union
=0 if the student is assigned to not be a part of Labor Union
c. Why is it important to ensure that the randomization is done correctly?
d. Name any two variables that you can use to test if Treatment and Control groups are
balanced.
e. Suppose you test for balance using 20 relevant variables and find that the treatment
and control groups are balanced. Do you think produced from the randomized
control trial will be biased or unbiased? Explain.
to generate a solution
a solution
- Text book imageManagerial Economics: Applications, Strategies an...EconomicsISBN:9781305506381Author:James R. McGuigan, R. Charles Moyer, Frederick H.deB. HarrisPublisher:Cengage LearningText book imageText book image
- Text book imageText book imageText book imageEconomics (MindTap Course List)EconomicsISBN:9781337617383Author:Roger A. ArnoldPublisher:Cengage Learning