1. Would you use a survey, data analysis or an experiment to answer the following question: How much of the gender gap in earnings is due to men and women choosing different college majors?

Group of answer choices

Survey

Data analysis

Experiment

2.

Would you use a survey, data analysis or an experiment to answer the following question: Do college applicants understand the relationship between the college major they are choosing and longer-term earnings?

Group of answer choices

Survey

Data analysis

Experiment

3. You run an experiment among college students to test the effect of incentives to walk vs. no incentives to walk on average steps per day. Students who receive incentives to walk (D=1) encourage their friends who don't receive incentives (D=0) to walk with them. Does this likely make your observed treatment effect larger, smaller or no different than the population ATE?

Group of answer choices

The observed treatment effect is smaller than the ATE

The observed treatment effect is the same as the ATE

The observed treatment effect is larger than the ATE

4.

REUSE THE PLEASE E TOWELS HELP SAVE THE ENVIRONMENT Towel reuse sign You are running an experiment to test the effect of two different hotel door hanger messages on hotel towel reusage: Message A "Save the environment" vs. Message B "75% of hotel guests reuse their towel." One possible level of randomization for this experiment is the hotel room level: you randomize each room to receive either Message A or Message B. Another possible level of randomization is the city level: all the company's hotels in city A get Message A and all the hotels in city B get Message B. What is another possible level of randomization