Effect of Smoking on Baby During Pregnancy

Effect of Smoking on Baby During Pregnancy

Your group will use the data(R) set BWGHT2. You are encouraged to go out and get other data (from the real world) if you think it will help your results (you might update these data files and look for other interesting factors).
The Model
In this section you want to discuss the basic behavioral, informational and institutional assumptions you are making. It is important to be explicit about which assumptions may be driving the results. You should discuss how the results are sensitive to changes in parameters.
1) Not every paper will require the development of a full theoretical model. If it is possible to explain the assumptions, the mechanisms at work and the results clearly, then a model could be redundant. Every paper, however, should include the empirical model that will tested (see below).
2) If the question you wish to address requires (or even benefits from) a formal model, you should start with the simplest model you can think of.
You can complicate things later if you need to (i.e. if you find that you cannot make
any of the predictions you would like to test). Think carefully about functional form and continuous vs. discrete time (again, go for simplicity but be open about the assumptions and how they might drive the results). There are two important things to keep in mind here:
a. The model should be simple and help illustrate the question you will be testing but should not be ad hoc. That is, don’t use a two period model because it is simple and gives you the result you want, if adding one more period to the model would negate these results. The form of the model should be appropriate to the problem. The easiest way to achieve this is to build on previous models that have been used to test this or related topics.
b. In the end, you need an empirical model, so the theoretical model you develop must lead somehow to what you are testing. You will need to address how, say, the regression you are running is a linear form of an equation from your optimization problem. Be explicit about how the empirical model differs from the theoretical model, e.g. if you are unable to estimate certain parameters or if you need to assume a particular functional form. The writer just write a part of the project.