Exam 15: Quantile Regression, Count Data, Sample Selection Bias, and Quasi-Experimental Methods

What is non-negative count data? Why does it present a concern for OLS? How might you control for non-negative count data in the estimation process? Explain.

Suppose you observe that several different communities in a large metropolitan area raised their sales tax by one percentage point in 2012 while several others did not.In an effort to determine how the increase affected car sales in the affected communities you estimate a difference-in-difference estimator for 2011 and 2013 car sales and get = 987.25+ 17.08201+ 1.47- 8.32(2013* Treatment (311.24) (5.33) (0.56) (2.46) a)Why would you want to estimate a difference in difference model? b)Draw a graph with the 4 means on it and explain where the difference in difference estimator is on the graph.

Non-negative count data presents a challenge because OLS estimates

Difference-in-difference estimators attempt to

Quantile regression is different than OLS in that it

Non-negative count data occur when the dependent variable takes on

You can choose between the Poisson and the Negative Binomial models by performing a

In which of the following cases would you want to use difference-in-difference estimation?

The second-stage in the Heckman selection correction is including the _____ in the second-stage regression to control for the potential sample-selection bias.