Some suggestions about how to write an empirical paper
General Issues There are four basic questions that the paper must explicitly address. The introduction of the paper is a good place answer these: 1-What is the question you are trying to answer? 2-Why is this question important? 3-What is you contribution? How are you answering the question and in what way is your answer to the question different from previous work? 4-What are your main findings? What are the implications of the findings? You should always include an abstract in which you succinctly answer these questions.
Comments on the literature review section -it should be short -it should include only papers that are directly relevant to your paper, not every single paper written on the topic. -In this section you should explain in more detail how your paper is different from what was done before, and what exactly your contribution is. For example: Is it new data? New/better methods? Do you have a new model/theory? Are you answering a new question? Are you able to overcome some problem that was prevalent in previous literature? Specific comparisons are useful. For example: “this paper improves on work by X (19..), Y(19..) and Z(19..) because we do K”
Comments on modeling -Start with the simplest model you can think of. Later on you can make things more complicated. -Think about whether you need to use continuous or discrete time. -Do you need an infinite time horizon? Will a two period model be sufficient? -Which functional forms for utility and productions functions will you use? Are there particularly simple functions to work with? -Is a model necessary? Do we gain any insights from the model or could you have described the process with words? -Provide intuition for the results -It is useful to be specific about which assumptions drive the results of interest.
Comments on the data description section -Always describe the name and source of the data, and the period it covers. -Do you have a panel, a cross section? A time series? -Describe the unit of observation. Is it a person? A household? A country? How many observations do you have? -What kind of variables are available? Is there some important variable that is not available? -Descriptive statistics of the data should always be included and commented upon. Graphs are nice. -Describe the sample you are using: why do you use this sample? -This section should alert the reader about potential problems with the data that might affect you estimations. For example: how was the data collected? Are there measurement error issues? How are the variables coded? -It is useful to think about what the ideal data set to answer the question would be. Is the data that you have different from what you would like? How will that affect your results? -If you are using more than one data set. Why? What are the relative advantages and disadvantages of each?
Comments about the econometric model -Why are you using a particular specification? Is the specification suggested by theory? By previous empirical work? -Describe in words where the identification comes from. -Are there any issues related to the econometric specification? -If you write down an equation, the paragraph immediately below should explain exactly what all the symbols in the equation mean. -It is useful to state what you expect to find prior to the estimation -if you do use different specifications or estimate different models, explain why. -The variables included must be justified (briefly).
Comments on the results section -Always interpret the coefficients of your estimations. For example “we find that b=0.003, so that increasing X by one unit increases y by 0.003. The implied elasticity is...”. This is particularly important if you are not estimating a simple OLS regression. And even with OLS it is useful, especially when the magnitudes of the variables are not immediately apparent, for example when x is in logs. -Are the effects that you find statistically significant? Economically important? -Are they what you expected (in terms of the sign and the magnitude)? If not, can you think of reasons why you are finding an unexpected result? -What are the potential problems with your estimates? Can you do anything about them? -Are you findings similar or different from what others have found? What is the source of the difference?
Comments about tables -The tables should be self explanatory. They must have a descriptive title, include the name of the dependent variable, the method of estimation, the sample used and any other relevant information that the reader needs to understand the tables. -Always use labels rather than symbols or programming names. -It is preferable to report standard errors than t-statistics. -Do not include 8 decimal places! -For some reason in economics no vertical lines are used in tables. Look at a journal.
Comments on the Conclusion -This is a good place to summarize your findings. -It is a good place to point out limitations of the results and possible extensions. -It is a good place to speculate in a more casual manner about the implications of your results. -In general it should not contain any new results.
Other General Comments -Substantiate general claims that are not proved in the paper either by providing relevant data or citing studies. -Make an effort to write concisely and to the point. Try to make the paper as short as possible. Digressions and marginal comments belong in footnotes and not in the main text. -Long proofs, details on data collection and other secondary documents belong in the appendices of the paper. -You should try to keep the text under 25 pages.
京公网安备11010202008513号 京ICP证101109号 京ICP备12012021号
发表于 2012-6-22 21:17:00
收藏
收藏
发表于 2012-6-22 22:33:20
楼主
