Statistics in the Animal Kingdom

A particularly misleading headline after the release of the UK Census inspired this little exercise. To better understand questionable practises in data journalism we'll play the 'bad guys'. For obvious reasons it's not a great idea to misrepresent actual data, especially if this is done at a university. Instead we use a fictional dataset that bears little resemblance to the real world.

Hence, we play in the Animal Kingdom which is extremely loosely inspired by the CGP Grey's Politics in the Animal Kingdom. We recently had a census of the four major groups: lions, monkeys, elephants, and gazelles. The census measured age, education level, income, and habitat. Their politics is dominated by two issues: deforestation and taxes. Players in this game take the role of spin doctors, trying to misrepresent the census data combined with surveys as crassly as possible.

Note: most of the creative writing and a sizeable fraction of the world building is due to ChatGPT. The logs of the two writing sessions can be found here. Any real world references in these logs (such as to Donald Trump, David Attenborough, and the TV show Yes Minister) are meant to solicit a specific response from ChatGPT and are not to be interpreted deeper nor are they authorised. The artwork was generated using Stable Diffusion.

The code is licenced under GPLv3, the data under CC-BY-4.0.

Playing instructions

Warning Please don't look at any file in the git repository until you have finished the game to avoid spoilers.

Disclaimer You may post results you have obtained with this dataset anywhere you like but please indicate that this is fiction. This repository is provided for educational purposes; I do not endorse misrepresentation of real data!

  1. Make sure you have access to a python installation with numpy, pandas, and matplotlib. If you don't want to install these locally, you can play in Binder
  2. Read the introduction here and pick the camp for which you want to play. You may choose to play for the Office of National Statistics (ONS) in which case you must represent the data honestly.
  3. Read the introduction to your camp: Monkey, Lion, Elephant, Gazelle, ONS.
  4. Download the python module survey.py here if you're playing for a camp or ons.py here if you're playing for the ONS.
  5. Look at the example notebook here (download) to understand the survey module.
  6. Misrepresent the data you can access to advance the cause of your camp.
Go to Introduction