Final Project Assignment

17.871

Spring 2009

Assignment summary. The purpose of this project is to conduct a research project in political science with quantitative methods. Your main goals are to address an interesting question, further develop and practice the quantitative skills you have learned in this class, and demonstrate your learning to us. You will make one 5-10 minute and one 15 minute oral presentation, and turn in a final research paper, 15–20 pages long.

Presentations General considerations. The best presentations are practiced beforehand. You will be graded on the quality of your presentations.  To improve presentation quality and to save time, distribute handouts. Take care of the technical aspects of the presentation before the class time begins. When you are not presenting, you are participating by listening and giving feedback. You must be present for all the presentation sessions. Time passes faster than you imagine, so practice first to make sure you're within the time limit. So that everyone will have an opportunity to present, I will have to cut presentations off after the allotted time.

First presentation. 5-10 minutes to a small group on April 15. The first presentation is to get feedback from your instructors and other students. You should inform us of your subject, how you intend to pursue it, the data you plan to use, any special problems you think you will encounter, and the bivariate relationship between your 

dependent variable in your key explanatory variable. Please bring copies of your slides for everyone in class and your instructors. By the first presentation, you must have gathered your dependent variable and your key explanatory variable.

• Inform us of your subject: What is the substantive issue you intend to pursue? What is the key puzzle or question you wish to address? To whom would the issue be of interest? Have others researched this area before? If so, what have they found?

• How do you intend to pursue the subject: What are your major dependent variables (focus on one or two)? What are your major independent variables? What “unit of analysis” will you be studying? (That is, will you be studying individuals, counties, countries, etc.?) Will this be a time series analysis? Cross-sectional? A combination?

• Your data: What would the ideal data be to do this project? If you can’t get your ideal data, what data can you get your hands on? Tell us actual sources. It would also be good at this point to report descriptive statistics of your data set, graphs that show simple, basic relationships, etc.

• Special problems. Perhaps there’s a crucial variable that will be difficult or impossible to measure directly, or some regression assumption will be violated. Quantitative analysis is rarely straightforward, so tell us any wrinkles you’ve encountered thus far, or anticipate encountering over the next month.

• Bivariate finding. You absolutely must present the bivariate relationship (scatter plots, cross tab, regression, etc.). You must interpret the bivariate coefficient and show us the labeled data points where possible.

Second presentation. 15 minutes. The second presentation should summarize the basics of your first presentation and then report your findings. The summary of the first presentation should be very brief, focusing on reminding us of the subject you are pursuing, your basic approach, and the source(s) of your data. Your findings should be reported in much the way we learned data analysis in this course. First, present your data. Identify the variables in your analysis: how are they measured, what are their means and dispersions? Second, present the most important bivariate relationships. Third, present your multivariate analysis. Draw to our attention the core results and whether they confirm or disconfirm your original conjectures. Discuss any cases that appear to be poorly described by your analysis and what might be done to correct this. Fourth, tell us what you conclude about your subject from your analysis.

Research paper. The logic of exposition in your research paper should roughly parallel the second presentation. The one section you should add at the beginning of the paper that you won’t want to talk too much about in the second presentation is a discussion of previous research. While you don’t have to do a comprehensive search of the literature, you need to do a little library work to see if anyone has written about your topic before and (if they did) what they found. Don’t worry if you’ve found that someone has previously done something identical to your project. Replication of others’ results is an important part of normal science. Journal articles in political science (which is the model your write-up should follow) often follow this outline:

I. Introduction

II. Literature review

III. Data

IV. Model specification (i.e., what are the variables and how are they measured?)

      V. Results 

           A. Preliminary results (simple relationships and a first cut at multivariate analysis)   

                 Make sure you interpret your key coefficients!

           B. Sensitivity analysis and other searches for anomalies 

           C. Re-estimation; further analysis

VI. Discussion of results from a substantive perspective and address alternative 

      explanations. Make sure you address internal validity and external validity.

VII. Conclusion

The first page (after the title page) must have an abstract of no more than 250 words that summarizes your project and findings.

The abstract page must contain your name and the title of your paper.

The paper should be about 20 pages long, including text, graphs, figures, appendices, and bibliography. There is no hard and fast page limit for the final page. Write succinctly. Edit. Hone. Eschew verbiage. Make the text flow. If there’s a particularly long, involved methodological problem that you attend to, consider moving it to an appendix and only make reference to it in the body of the paper.

Read Kate Turabian’s style book (A Manual for Writers of Term Papers, Theses, and Dissertations) before you write your final draft. Follow this style book for most things. In addition, the American Political Science Review’s style sheet is the authority for special cases that pertain to political science. The following URL extracts the information concerning citations: http://depts.washington.edu/pswrite/cite.html. Look at recent editions of the American Political Science Review or American Journal of Political Science to see how tables and figures should appear.

Finally, there are some mechanical issues that you must attend to in writing this paper. (Indeed, you should always attend to these issues when writing for the social sciences.) If you do not follow these strictures, your paper will be returned to you for rewriting, possibly resulting in your receiving an I for the subject. When you turn in your final paper, you will need to include a check-list indicating that you have made sure it meets these criteria.

• Double-space everything, except footnotes and tables.

• Use footnotes, not endnotes.

• Use the author-date form of citation.

• Properly number the pages (page 1 is the first text page).

• Include figures and tables at the end of your paper, consecutively numbered. (Include all the figures first, then the tables.)

• Make explicit reference to each table and figure in the body of the text.

• Don’t use fancy fonts. Make sure the font size is no smaller than 11pt and no larger than 12pt. Never use bold face. Use underscoring or italics for emphasis, book titles, etc.

• Use a constant 1" margin on each page.

• Tables and figures should be self-explanatory.

• Staple the paper in the upper left-hand corner. Don’t use a report cover.

What you will hand in

According to the regulations of the faculty, all final papers are due on the last day of this class. That means by 5pm. Any extensions must be approved beforehand.

To receive a grade in this class you must produce the following:

• Two copies of your paper, handed in at my office by 5pm on the last day of this class. We will grade one of the papers and return it to you, if you wish. We will keep the second copy, as an example to people who take the class in the future. Make sure you have another copy for yourself.

• Completed checklist form. If you want to see your paper alive again, include a self-addressed envelope capable of holding our paper or stop by my office. (If the envelope is addressed to go off campus, I will happily supply the postage.) 

• An extra copy of your abstract, which I will keep. Your data and the STATA code capable of reproducing the results you report. See below for further instructions about this.

If any of these elements is missing, you will receive an I.

How to turn in your data

For the benefit of future students in 17.871, to assist us in giving you feedback, and in the interest of scientific replication, you are required to turn in the data you use to produce the results in your paper, plus the code used to generate the results. Here’s what you have to produce.

Data. If you gather the data yourself, you must submit the computer file of that data. You must also make available a “codebook” that identifies the variables: (1) variable name, (2) data source, (3) column location of the variable in the file, and (4) any coding rules 5 you used to create the variable. If you used a data set produced by someone else, such as an ICPSR data set, you must tell us the full citation of the data set. If it is not an ICPSR data set, and it is not available through the MIT-Harvard Data Center, we will treat it like you gathered the data yourself.

STATA code. You must save the STATA code from the “.do files” that produced the results you report in your paper. Give the code a file name that corresponds with the results reported in your paper (e.g., table1.do, figure2.do). You will then e-mail the data to us, as an attachment. If the data is too large to e-mail, let us know and we’ll work out something.