check assignment sheet
1. The aim of this 5-7 page paper is to present the data you have obtained, explain the econometric methods you intend to use in testing your hypothesis and present the preliminary results
2. Even though, the introduction and literature review will not be regarded, include it in your paper, so as to have a good understanding of you topic.
3. Some advice regarding the data and empirics: i. Include descriptive statistics of your variables: means, std. deviations, min/max, number of observations Interpret the numbers- describe your data in the body of your paper. Make sure units are clear to the reader. Talk about an average observation as well as about outliers. ii. Make sure that the units of observation are clear to the reader and the sample selection is well explained. iii. Provide as much evidence from the data as possible – even if it doesn’t support your hypothesis. iv. For clarity, put your tables at the end of the paper. v. Discuss differences of means. I.e. if your theory is that FDI into countries with capital export controls is lower, tell us the average FDI (scaled appropriately) into countries with and without such controls. If your theory is that repeat customers tip better, tell us the average tip for repeat customers versus first-time visitors. vi. Explain why the simple differences-of-means, or simple OLS regressions, might be biased, i.e. why another empirical approach is more suitable in your case. Describe your empirical model. State the equation(s) clearly and list all the variables you are planning to include. If you are planning on testing different models, describe it clearly. Motivate your planned empirical approach using the theory as much as possible.
4. Some advice regarding the data and empirics: i. Tables should be self-contained to the extent possible and make sure names of the variables and units of measurement are clear to the reader. ii. Present the basics first (differences of means, OLS, etc.). Then explain why the simple correlations might be biased before you try anything fancier (fixed effects, 2SLS, FGLS etc.). iii. Discuss your results in the context of the theory and other empirical papers on the same topic. REMEMBER: Before you get involved in the actual regression analysis, first present evidence from the data itself that support your claim/hypothesis.
5.If you are planning on using panel data methods, at this point ignore it and run an ols regression ignoring panel structure of your data. Same if you are using probit/logit model on binary data or Tobit model on censored data. If you are planning on using multivariate probit or logit model, you don’t have to present any preliminary results at this point.
