It is highly recommended that you pick a topic of personal interest using original data. Stuff out of textbooks rarely turns out to be interesting. Think small – like can I do something which will get me promoted on the job – rather than large – can I estimate ISLM curves. Really good projects in the paper have involved estimating what kind of personals ads get a strong response, forecasting church attendence, modelling occupancy rates in UW residence halls, forecasting the demand for lift tickets at Snoqualmie, studying cable deregulation, forecasting box office revenues, and estimating the density of Douglas fir trees.
· Good papers usually do more than just describe how an outcome is different between two categories. There is no hard and fast rule about this, but in general think multiple regression.
· One way to frame your thinking is as follows: “I’m going to estimate the following parameters. If the parameters look like this, I will conclude A. In contrast, if the parameters look like that, I will conclude B.” In this way, the framework of your paper is set out before you begin. (You just have to do all the hard work to fill in the details.)
The paper must begin with a one-page Executive Summary, which explains everything one will learn from the paper. The paper proper can be no more than five pages. (The paper should be double-spaced and in a reasonable font size.) You may include technical appendices if you think it necessary – but you cannot assume that I will read past the five pages. The paper proper must be self-contained. You must include an appendix with computer output for all results presented in the paper so I can look at it if I’m curious. This doesn’t count against the five page limit. You must have an EViews file containing your data available to give me, although I don’t generally request them.
A word about the Executive Summary: an Executive Summary is just that – a summary. It is not an introduction. The summary should state what question you asked and tell the answer you found, briefly describing your method. (So it should be redundant to someone who reads the whole paper.) Think of the summary as one page going to your boss. Your boss trusts your work and wants a summary of what she needs to know and no more. Here are two rules about Executive Summaries:
1. It should be possible to understand the paper by reading the Executive Summary and never reading anything else in the paper.
2. Nothing should be lost if the reader skips the Executive Summary and reads the paper instead.
The proposal paragraph you turn in must identify the question you plan to work on AND identifiy specifically the data source you are going to use.
It is okay with me if this paper is part of a project for another class as well, provided you let me know (in advance and by email explaining the extent of overlap). Obviously, you have to clear it with the instructor of the other class as well.
