The final project requires you to conduct an independent research project. For the project you will perform an exploratory analysis of data using R. You will then submit a written report that explains your research questions, provides necessary background information, describes your data and methods, and summarizes your findings.

Components of the project

To help keep you on track, I have broken the project down into several discrete tasks, which you will complete over the last several weeks of the semester:

  1. Receive topic approval: You will first send me an email about your plans.
  2. Load the data: You will then send me an email demonstrating that you can access and load your data.
  3. Submit a rough draft: You will send me a rough draft. I will provide feedback and circulate the draft to two other students who will also provide feedback.
  4. Present your project: you will present the project to the class via Zoom.
  5. Submit your final report: after receiving feedback from the class, you will have about a week to revise and submit your final report.

Instructions and due dates for each of these tasks are listed in the Due Dates section of the syllabus.

Final Report Guidelines

This section provides suggested formatting and organization guidelines. I am open to alternative approaches (e.g., a policy brief) if they better suit your interests. But: you must first clear any formatting changes with me before deviating from these guidelines.


By default, I expect the following format:

  • 12-point, serif font (e.g., Times New Roman) for all body text
  • Sans serif (e.g., Arial or Helvetica) font is encouraged, but not required for all headings
  • 1" margins
  • Begin with a title page containing the following information:
    • Title
    • Your name
    • Date
    • Abstract: A single-spaced, 150-word (maximum) summary of the problem you address, the data you use, the most important findings, and the conclusions you draw


The main body of your report should include the following headings:

  • Introduction: The introduction has three parts:
    • Context: the opening paragraph should explain the broad context needed to understand your research problem
    • Problem: what problem will your research address? If you are trying to answer a specific research question, persuade the reader that the absence of an answer to the question presents a problem. The context paragraph above should be written to provide only enough background information as necessary to help readers see why the problem you address is really a problem. see the Topic Ideas section suggestions about problems to study
    • Solution: In no more than one paragraph, describe how your analysis addresses the problem. Specifically, explain what data you use, what analyses you perform, what results you find, and what these results suggest about the answer to the problem.
  • Background: provide whatever relevant background knowledge is necessary for readers to understand the scope of the problem and to justify your choice of data and methods for addressing the problem.
  • Research design: Explain and justify the data you will use and the methods you will use to analyze the data. How will you clean the data? What are the key independent and dependent variables in your analysis? What values can each of these variables take on? What are the most likely confounding variables and how will you account for them in your analysis?
  • Empirical analysis: Report your results, providing well-formatted and easy-to-understand plots for your most important findings. Your analysis must include at least one linear regression model. Explain your results in language that would be understandable to an interested observer who has not taken this class (like a prospective employer). The reader must be able to understand why you are including every figure and table and how to interpret each one. Embed all of the main tables and figures in the body of the text rather than at the end of the document. Feel free to include an appendix with more technical analyses such as regression diagnostics or supporting information that a non-expert may not understand or care about.
  • Empirical limitations: no study is perfect, so it is important to help the readers understand the problems with your study. These limitations usually include (A) alternative explanations for the patterns you find in your analysis and (B) concerns about whether the results you find in your study will generalize to other relevant populations, settings, or times.
  • Conclusion: briefly summarize your main results and explain what readers should conclude based on those findings. Limit this section to three or fewer paragraphs.
  • Statistical appendix/appendices:
    • Provide a table with appropriate summary statistics for all variables you include in your analysis.
    • Unlike the midterm, you do not need to include your R in an appendix. Instead, you will submit it separately (see details in the Due Dates section of the syllabus)


The paper will be graded on three criteria:

  1. Clarity of writing: the report is well written, free of spelling and grammatical errors, and fully explains everything required in the organization section. Particular emphasis will be placed on the design and analysis sections. Ensure that an interested non-expert would be able to understand all analyses you report. For help with grammar, prose, and copy editing, consult with the FSU Writing Center
  2. Appropriate use of methods: the analysis relies on the best available methods, given the research question. You are required to include at least one regression model.
  3. Scope of analysis: on the midterm, I saw huge variation in how much analysis people provided to address their research problems. There is rarely a single plot or regression that will provide all of the relevant information. This part of the grade will reward people for examining their research problem from several different angles. Keep in mind, however, that you must explain each plot or table you create in writing, so you will not be rewarded for providing analysis that is redundant, irrelevant, or not thoroughly explained and justified.

Topic Ideas

Since this is the first time this course has been offered, I lack real examples of successful projects (Note that your project may be used as an example for future classes. If you would prefer that I do not use your project as an example in future classes, please let me know via email.). But I have many ideas for appropriate research questions. Here are a bunch of ideas, organized around three broad themes. Keep in mind, I want students to tailor their research toward their career interests, so feel no need to rely on these ideas. I prefer creativity and originality!

  1. The central purpose of data in campaigns is to enumerate citizens to contact. So a simple approach for this project is to choose a constituency and then provide exploratory data analysis to answer one or more of the following questions:
    • which citizens will support which parties?
    • which citizens will donate to campaigns?
    • which citizens will turn out to vote in general elections?
    • which citizens will turnout out to vote in primary elections?
    • which neighborhoods have the most supporters of a given party?
    • which neighborhoods participate at the greatest rates?
  2. Many staffers or policy advocates may also be interested in the opinions of different constituents. Therefore, you may wish to examine questions such as:
    • which people have the strongest support for a specific policy?
    • which regions have the strongest support for a specific policy?
    • how is opinion on a policy changing over time? Is it changing the same ways in all regions? Is it changing the same ways in all demographic groups?
  3. More broadly, many representatives need a basic understanding of their constituencies. Projects may use data to answer questions like:
    • Where do constituents of different types live? e.g., where do supporters live? where do opponents live? Do they live in rural or urban areas? Do they tend to live in apartments, townhouses, or single-family dwellings?
    • What are the major economic concerns of constituents? How many are employed? How has the economic landscape changed over time?
    • Has the demographic makeup of the district changed?

Data Sources

Here is a list data sources you might find useful: