gevhaz / reputation_and_regime_changes

Reproducible statistical analysis for my Master's Thesis on the efficacy of Human Rights Treaties

Requirements

The R programming environment is required, along with the following packages:

• gdata
• tibble
• lubridate
• ggplot2

The code depends on having the right datasets available. These are:

• Polity IV Annual Time-Series, 1800-2018 from Center for Systemic Peace

• Country and Territory Ratings and Statuses, 1973-2019 (Excel) from Freedom House

• Ratification data for the International Covenant on Civil and Political rights and its Optional Protocol from the Status of Ratification Interactive Dashboard from the United Nations Office of the High Commissioner.

How to use

Run the main project file in R:

R project.R

After that, the results of the thesis can be reproduced by calling one of the functions run_pure_regime_change_test() or run_main_test(). For the pure regime change test, only one argument is needed: the test number. It can be 1 or 2. The main test also takes the test number as the first argument but additionally needs to know which treaty to use to divide the regimes into populations of ratifiers and non-ratifiers. This argument should be "CCPR" or "PROTOCOL". There is also a third, optional argument that determines what type of plot will be produced. "BARS" for a bar plot and "HISTOGRAM" for a histogram. An example test would be:

run_main_test(1, "CCPR", "BARS")

which would then show some statistics about the selected sample and produce a stacked bar plot with two groups – ratifiers and non-ratifiers – and bar heights depending on the fraction of the sample that saw improvements, decline or no change in their ranking.