Skip to content

This GitHub repository stores all the files and instructions necessary to reconstruct the data and analyses for my PhD dissertation in Political Science at the University of Michigan.

Notifications You must be signed in to change notification settings

yivash/shaun-mcgirr-dissertation

 
 

Repository files navigation

Shaun McGirr's PhD Dissertation: "Andrei Drives Bentley, Boris Drives Lada: Why Corrupt States Have Clean Agencies"

This GitHub repository stores all the files and instructions necessary to reconstruct the data and analyses for my PhD dissertation in Political Science at the University of Michigan, defended in 2016 and available here: http://hdl.handle.net/2027.42/135764

The main aim of storing the files and data on GitHub is to maximize the reproducibility of my findings. As the necessary steps cross a few different software tools, it makes sense to store them in a place built for that use case.

Disclaimer: I am by no means a software developer and the code in this repository will almost never be the best way of doing something. As any PhD candidate can attest, completion is more important than perfection.

The contents of this repository are organised in to folders as follows:

  1. Preliminaries: overview of the software tools you will need

  2. Obtain: downloading the Russian government procurement data from zakupki.gov.ru

  3. Unpack: unzipping and parsing the downloaded files in to a usable format

  4. Construct: calculating the quantities of interest (ie measures) necessary to test my theory

  5. Test: statistical analyses testing the hypothesised relationships between these quantities of interest

  6. Present: collation and tidying of results in to presentable objects (tables, graphs)

The final product was compiled using Overleaf's online LaTeX tool: www.overleaf.com

About

This GitHub repository stores all the files and instructions necessary to reconstruct the data and analyses for my PhD dissertation in Political Science at the University of Michigan.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • R 91.4%
  • HTML 8.1%
  • TeX 0.5%