Project

Project Options

Option 1: Data analysis

Overview: Perform a causal analysis on a dataset of your choice.

Collaboration: Groups of up to 3. Individual work is fine.

Resources for finding data:

It’s fine with me if you use a dataset associated with an existing project (like an honors project) or a project for another class (as long as it’s ok with the other instructor).



Option 2: Learn an advanced topic

Overview: Dig deeper into existing course topics or learn a new topic. Examples could include:

  • Methods for transportability (generalizability) of effects
  • Interference
  • Specialized considerations for particular study designs
  • Machine learning in causal inference
  • Individual causal effect estimation

Collaboration: Groups of up to 3. Individual work is fine.



Option 3: Other

If none of these options piques your interest, I’m happy to discuss alternatives with you. Some ideas:

  • Design a Shiny app to illustrate causal concepts to students
  • Write (part of) an R package for making it easier to work with causal graphs
  • Critique of applied research
    • This is a good option for those who would like to do a data analysis but cannot find adequate data to pursue their question.
    • Find and read papers that study a question of interest to you. Critique these papers from a causal inference lens. This will involve constructing your own causal graph to guide a critique of the authors’ data collection and analysis methods.
    • Discuss what remains uncertain in this line of research, and propose an analysis plan for a new causal study to rectify the limitations of prior research.





Deliverables

Each group will turn in a paper and give a presentation.

Important dates:

  • Final paper due Wednesday, December 18 at 5pm
  • Final presentations will span the last 2 class days of class (Monday 12/9 and Wednesday 12/11) and our final exam period (Tuesday 12/17 from 10:30-12:30)

Paper

  • Format (data analysis option): Your paper should resemble an academic journal article and have the following sections:
    • Introduction:
      • Conduct a literature review using at least 3 sources. This literature review should provide necessary background for your investigation and explain how your work provides new understanding.
    • Data and Methods:
      • Describe the context behind and contents of your data. (Examine who, what, when, where, why, and how questions)
      • Describe any data cleaning and/or processing that you performed.
      • Show the causal graph that you use to identify what variables need to be included in the analysis. Discuss key decision points in your causal graph construction.
      • Describe the statistical analyses performed. This should include modeling to estimate treatment effects as well as placebo tests and/or sensitivity analyses.
    • Results:
      • Interpret causal effect estimates in context, being careful to note the particular estimand that was targeted.
      • Interpret measures of uncertainty (confidence intervals and/or p-values).
      • Discuss results from placebo tests and/or sensitivity analyses.
    • Discussion:
      • Summarize key results and the extent to which they were what you expected or were surprising.
      • Discuss limitations of the analysis in terms of data quality, scope of the analysis, and generalizability.
      • Discuss concrete future directions for the project.
  • Format (advanced topic option): Your paper should resemble a textbook chapter and have the following general structure:
    • Introduction:
      • Give an overview of the method/topic and its importance in causal inference.
      • How does this method/topic relate to topics that we have covered in class?
    • Subsequent sections: These sections should correspond to key concepts for this topic. Each section should explain a concept and show examples.
    • Conclusion:
      • Summarize when researchers would want to use this method/topic in their own work.
      • Discuss the limitations of this tutorial paper in terms of scope. e.g., “We did discuss concepts X, Y, and Z in this tutorial, but important subareas within this topic include A, B, and C (include citations to relevant articles where readers can learn more).”
  • Bibliography: In-text citations and references should follow the APA format.

Presentation

  • 10-12 minute presentation with 2-5 minutes for questions
    • Schedule posted here
  • Target audience is our class (students who have taken causal inference)
  • Share your slides with the instructor the day before your presentation. We will use the instructor’s laptop for presentations to streamline and avoid technical difficulties.





Timeline

Milestone 1

Due date: Friday, November 8 at 5PM

As part of this milestone, you will pick a project option, pick a specific topic, and form a group (if not working alone).

You will submit a short write-up that varies by project option. (No page length requirements—just convey the information that you need to.)

  • Data analysis option: You will write a draft of the Introduction, Data, and Methods sections of your paper.
    • Introduction: Give background for your topic, including relevant theory, domain knowledge, and prior research. Incorporate findings from at least 3 journal articles.
    • Data: Describe the data that you have and how it was collected. Include any limitations or cautions that are important to keep in mind about the data. Think about the who, what, when, where, why, and how behind your data to inform your discussion in this section.
    • Methods: Write a detailed plan for the analyses that you will conduct. This part can be written in bullet points.
  • Advanced topic option: Your final paper will be structured as a tutorial, and for this milestone, you’ll make progress towards writing this tutorial.
    • Introduce the big idea behind this topic. Make connections between your topic and the topics we covered in class so that others can have a better sense of how this topic fits into the field of causal inference.
    • Explain 1 idea that you learned from your research so far. Include a citation to at least 1 journal article.
    • Write a plan for your remaining research. Describe the ideas are you planning to learn in the remaining time. Describe a simulation study that you will perform to explore some of the theory behind your topic.

Milestone 2

Due date: Friday, November 22 at 5PM

For both the data analysis and advanced topic options:

  • Conduct analyses according to your Milestone 1 plan as well as updates to your plan based on instructor feedback
  • Submit a rendered HTML of your analyses and interpretations.