singmann/roulette_timing

Data and Code: A speed-of-play limit reduces gambling expenditure in an online roulette game

This repository consists of data and code for the manuscript "A speed-of-play limit reduces gambling expenditure in an online roulette game" by Newall, Weiss-Cohen, Singmann, Boyce, Walasek, and Rockloff.

The code to reproduce the analysis is contained in the RMarkdown file bayesian_analysis.Rmd. Running the code requires that the data folder containing the data is in the same folder as bayesian_analysis.Rmd.

Compiling (or "knitting") this RMarkdown document produces the output file bayesian_analysis.html which contains numerical results and figures. Compiling also creates the results figures as files in folder figures as well as saves the Bayesian MCMC chains as a binary R file in folder model_fits. Both of these folders are created during compilation in case they do not yet exist.

To ensure computational reproducibility, the code uses the checkpoint package. This makes sure that R package versions from a specific checkpoint date (2021-10-20) are used. Sometimes running checkpoint from within RMarkdown can fail. In this case, running the following R code once before compiling the RMarkdown document will install the required packages. Please note that the working directory of the R session needs to be set to the directory (see getwd() and setwd()) in which the bayesian_analysis.Rmd file is (and in which ideally no further R files should be as checkpoint will install all packages found in the folder in which it is run.)

install.packages("checkpoint")
library("checkpoint")
create_checkpoint("2021-10-20")