Are child seats are no safer than seat belts? The paper "Successful Scientific Replication and Extension of Levitt (2008): Child Seats Are Still No Safer than Seat Belts" by Jones and Ziebarth is a replication of earlier results by Levitt. Levitt found that there is no empirical evidence for the claim that child seats in cars are safer than seat belts. A copy of the replication paper by Jones and Ziebarth can be found at http://ftp.iza.org/dp8590.pdf Jones and Ziebarth's paper is appeared in Journal of Applied Econometrics. This journal has a data archive and the data for the paper can be found at http://qed.econ.queensu.ca/jae/2016-v31.5/jones-ziebarth/

Since the conclusion that child seats are no better than seat belts is interesting and controversial, I want you to assess in this project how sensitive the conclusions of the paper are to change the model specification. For this project, I would like you to carefully read the papers by Levitt and by Jones and Ziebarth, and then empirically assess how sensitive the conclusions of Levitt are to changes in the specification. In your conclusions section, clearly state your conclusions about the robustness of the empirical results and motivate them.