Raw data is found in the data directory. It consists of experimental data
(data/raw_data_anonymized.csv) and interview data (split across four files
data/Interview_Data*.csv).
The anonymised experimental data is derived from a non-shared raw data file,
using the script data/anonymize.R.
A logistic model is fit to the experimental data, implemented in Stan. The Stan
model definitions are in analysis/stan_code*.c and the R script which fits
them is in analysis/fit_model.R.
Two Bayesian GLMMs are fit to the interview data, implemented in brms. This
code is in analysis/Interviews.R.
The analysis/results.Rmd RMarkdown document loads the fitted stan models and
presents tables of results as reported in the manuscript. The helper script
analysis/knit.sh knits this document to a PDF file.