Multiple Choice
The following estimation methods should not be used to test for randomization when is binary:
A) linear probability model (OLS) with homoskedasticity-only standard errors.
B) probit.
C) logit.
D) linear probability model (OLS) with heteroskedasticity-robust standard errors.
Correct Answer:

Verified
Correct Answer:
Verified
Q9: The New Jersey-Pennsylvania study on the
Q10: Causal effects that depend on the
Q12: Let the vertical axis of a
Q12: Experimental effects, such as the Hawthorne effect,<br>A)generally
Q14: In the context of a controlled
Q16: A repeated cross-sectional data set a.
Q18: All of the following are reasons for
Q20: Heterogeneous population<br>A)implies that heteroskedasticity-robust standard errors must
Q23: Quasi-experiments<br>A)provide a bridge between the econometric analysis
Q46: (Requires Appendix material)Discuss how the differences-in-differences estimator