YOU MUST ALSO SUBMIT A PEER EVALUATION FORM (BELOW) TO RECEIVE CREDIT FOR THE PROJECT.
There are two deliverables due on April 29 that each group will be submitting together: your final report and slides for your presentation. Each individual member must complete an anonymous evaluation of your partners, to be turned in separately. You will also be presenting your project to the class on April 29. Any one person from your group may present the project, or a combination of members. Presentations should be 10-12 minutes long and should follow the same general structure as your paper. You will all be graded the same for the proposals, report, slides and presentation. The only heterogeneity in grades will come from the anonymous evaluations.
For the report:
Your paper should be 8-12 pages long including figures and tables, double-spaced, 12 point font, no more than 1 inch margins. Make sure to use proper citations for any references to other papers. All reports should begin with a title page and a short, approximately 100 word abstract. Sections of the report should include (at minimum) an intro, methodology, data, results, and conclusion section. Follow a structure similar to the proposals. Intro is your opportunity to state your research question and present the motivation for answering this question. You should present descriptive statistics/summary stats and possibly visualizations of the data in the data section. Make sure you describe empirical challenges, why X might be endogenous, what the ideal experiment would be, etc. in your methodology section. Are there still endogeneity issues even after implementing your method? Describe the weaknesses of your approach, why you still might not be identified, what you would have to be concerned about. You might just be mitigating these problems with your approach.
The results section should discuss your main findings. Interpret your estimates for the reader and discuss statistical significance. You should include results from multiple specifications. One of the specifications should be your “base specification,” the others should be included mostly for robustness checks and falsifications. Discuss whether your coefficients change significantly across specifications. If they do, why?
Figures and tables:
You may include figures and tables at the end or embedded within the document. No figure or table should take up more than half a page. No need to report all coefficient values if including multiple controls, just make note of it in the way that I discussed in class (e.g., a row labeled “State Fixed Effects”: no, yes, yes…). Report your main coefficient values of interest first. Make sure to include standard errors, and stars indicating p-values (***p<.01, **p<.05, *p<.1). It is possible that you will not have any figures, but it would be nice to show some sort of graph illustring the source of variation in X you are exploiting (e.g., diff-in-diff graph, time-series variation in X, histogram of the cross-sectional variation in X, etc.).
