Semester project overview

Further details about the semester project will be presented in Week 1 of the semester, at which point this document will be updated.

Instructions for graduate students

The goal for this project is for you to implement a complete open–source reproduction of an important research paper from your dissertation field. The general steps to completing this project are as follows:

  1. Identify a published paper whose key results you would like to reproduce, and whose reproduction seems feasible.1.

  2. Secure the dataset generated in the original study, either directly from the paper or by contacting the authors.

  3. Reproduce, to the best of your abilities, the analyses and figures presented in the original publication. Your work on this reproduction analysis should be tracked with version control, and the final report should be in the form of a qmd document.

  4. Write a report that describes (i) the motivation, methods, and results of the original publication; (ii) the steps you took to reproduce these results, including any problems that got in the way of your reproduction; (iii) what you learned from the reproduction exercise.

1 We will discuss characteristics of “good” paper to reproduce in class. Generally, this means that the data are available (or made available by the authors), but the code is not easily available.

Instructions for undergraduate students

Undergraduate students will choose an openly available dataset that you find interesting, from a list provided by Gaurav in Week 2 of the course. Your task is to create two non-trivial visualizations2 As above, undergraduate students will write a report that describes (i) the motivation and methods that went into collecting the original dataset; (ii) the steps you took to generate your figures, including any problems that got in the way; (iii) what you learned from the exercise. Undergraduate students will choose an openly available dataset that you find interesting, from a list provided by Gaurav in Week 2 of the course. Your task is to create two non-trivial visualizations3 As above, undergraduate students will write a report that describes (i) the motivation and methods that went into collecting the original dataset; (ii) the steps you took to generate your figures, including any problems that got in the way; (iii) what you learned from the exercise.

2 Here, I take “non-trivial” to mean something more than just plotting data from two existing columns against one another.

3 Here, I take “non-trivial” to mean something more than just plotting data from two existing columns against one another.

Grading

Your final grade will be determined as follows:

  • Proposal presentation (~10 minutes each; Week 10 of semester (October 28/30)): 10 points
  • Final project presentation (15 minutes each with 5 for questions; Week 15 (Dec 2/4)): 20 points
  • Final writeup (Dec 10): 50 points
  • Participation and engagement: 20 points.

Further details for each will be provided as the semester progresses.