Chimpanzees engage in competitive altruism in a triadic Ultimatum Game

This repository contains all the raw data, the statistical analysis code and the plotting code used in authoring the paper "Chimpanzees engage in competitive altruism in a triadic Ultimatum Game" by Sánchez-Amaro et al. By cloning the repository you can reproduce all the figures and essential statistical numbers found in the paper.

The data and code are both licensed under the Creative Commons Attribution copyright license, like the paper itself. See the LICENSE.md file for details.

Replication quickstart guide:

After cloning the repository, the various .R scripts in the analysis/ folder should be run in the following order. Note that all scripts expect to be run with the analysis/ folder as the working directory.

1. Run data_reformat.R. This will produce a number of intermediary data files in the data/ folder, all derived from the single raw data file data/competitive_altruism_dataset.csv.
2. Run each of the fit_*.R scripts. They can be run in any order. Each will fit one or more brms models and save the fitted models as .rds files. All further scripts require that these .rds to exist and will fail with errors if they do not.
3. Run the knit_summary.R script. This will load all the saved model files from the previous step and produce a matching Markdown file (model_summaries.knit.md) and PDF file (model_summaries.pdf). The model summaries included in the paper's Electronic Supplementary Material are precisely this PDF file. The Markdown file produced from the model fits used for the published paper was committed to the repository prior to initial submission of the manuscript. This means that if you re-fit the models, you can use git diff to easily confirm that your new numbers do not differ substantially from those seen in the paper.
4. Run each of the plot_*.R scripts. They can be run in any order. Each will save one or more figures in the plots/ folder. All figures used in the paper will be reproduced, plus some extras. Minor differences may be visible between your reproduced figures and those in the paper, due to variations in MCMC model fits.