Unofficial Listing of GA Stat Effects Data
As indicated by the title, this thread lists data vis-a-vis the stat effects of the various GA proposal categories; more specifically, the non-relative difference categories effect on the stats. This record works by gathering data regarding stat changes from the most recent resolution(s) in each category, for every nation which (1) has not answered any issue and (2) was in the World Assembly in time to have stat changes effected from the resolution. This data is collected using a set of Python and R scripts, which may be found here; feel free to run your own analyses using these scripts! How to use the scripts is in the Readme file of the linked repository. Please also feel free to reply to this thread!
Notes,
Last updated: 8 August 2023 / GA #678.
Notes,
- This listing is entirely unofficial. It is not affiliated with, endorsed by, or informed by the Issues Editing team, the GA Secretariat, or any site staff. Obviously I am not (and have never been) a member of any site staff group. This also means that I make no guarantee that the facts given here accurately represent the stat effects of listed categories.
- Some categories have not had any resolution passed in sufficient recency for it to be possible to gather meaningful data regarding their statistical effects. Some categories may therefore be missing from here. If you want to write a resolution in that category to fix that, good for you.
- This is updated manually whenever I want to. However you can run your own analyses if you want to see new data before I get around to updating this thread.
- Some stats are highlighted green. This is because the observed effect is very strong; ie the mean change is over 30 and stat changes the other way are de minimis (less than 0.05 * the mean effect). Thus, stats highlighted in green are likely to be the most affected by that category.
- Linear relationship data between the original value of a stat and the change to the stat is included if the p-value of said relationship is under 0.05. There is no check to ensure that a linear relationship model is accurate. Run this check yourself if you are really interested.
Last updated: 8 August 2023 / GA #678.
