Problems - Ch12

  1. Create a simulated data set with 4 types at 1, 2, 3, 4. Re-estimate. How accurate are they now? Explain the difference.

  2. Using the twins data, replicate the IV estimates of measurement error.

  3. Using a DAG explain how the IV approach solves the measurement error problem.

  4. Create a bootstrap version of the mixture model estimator. Can this be used to estimate the identified set?

  5. Estimate the difference in difference model for each of the types of restaurants in the Card and Krueger data.