Group Replication Project

EC031-S26

Aleksandr Michuda

As described in the syllabus, you will be working in groups to replicate a paper. This project will be due at the end of the semester and you will present your work to the class. The project will be worth 15% of your final grade.

The replication crisis in social science has led to changes in peer-reviewed journal policy that increases open access, reproducibility checks for code and transparent data. This means that many reputable peer-reviewed journals provide the code and often the data to replicate a paper. For instance, if you go the website of the American Economic Review (https://www.aeaweb.org/journals/aer) and click on any of the papers in a current or relatively newer edition, you will see a link on the bottom of the page to the replication package.

Replication packages come in all different sizes. Despite open data and replication packages, replication is still hard. This is because the code is often written in a way that is difficult to understand, the data is not clean, or the output is not provided.

You will break up into groups of 3-4 people and choose a published, peer-reviewed paper (from the list in the link below) that you are interested in and try to replicate the results in the paper.

You will submit a video with a presentation or some other way of describing your findings. Your project will be graded on the following criteria:

  1. How well you replicate the results (as a function of how easily replicable the paper was)
    1. If you choose a paper that was easy to replicate, then you need to make sure that everything replicates perfectly. But oftentimes, this is not the case. So if the replication package was difficult to implement for whatever reason (difficult to understand code, missing data, missing output), you should highlight this in your presentation, and explain why replication wasn’t possible.
  2. A breakdown of issues you faced, how you overcame them or what you did to get as close as possible to replication
  3. Extensions to the original paper
    1. Replicating is hard enough, but the true test of a researcher is to push beyond that and think of an interesting extension. This need not be a completely new and innovative idea. It might just be using the data and exploring the data more (running a regression with a new interaction that the original paper didn’t)
  4. Topics and thoughts about future research
    1. Did you end up still being interested in this paper?
    2. Would you want to continue on this line of research?
    3. What are some questions you would want to explore in future research?

The papers you choose for the replication can be chosen from this list: Replication Paper List. Look through the papers and choose one. You cannot choose the same paper if another group has already chosen it.