paolocrosetto/temptation_experiment_data_and_analysis

Data and analysis for the temptation experiment

Data and Analysis for "Hard vs. soft commitments: Experimental evidence from a sample of French gamblers"

In this repository you will find the original data and the analysis carried out for the paper "Hard vs. soft commitments: Experimental evidence from a sample of French gamblers", by Paul Bettega, Paolo Crosetto, Dimitri Dubois and Rustam Romaniuc.

The paper

The paper is available as a GAEL (Grenoble Applied Economics Laboratory) Working Paper here TODO ADD LINK.

The data

Data are stored in the /Data folder.

• temptation_data contains the clean data issued from our oTree app, and is used for all analyses but the MCMC estimation
• MCMC contains the results of the MCMC estimation

The analysis

Analyses are fully run using R. They depend on the following packages

• tidyverse: R dialect used here

• broom: to clean test results and coerce them into a data frame

• magrittr: for extra specialized pipes like %$%

• gtools: to set up the list of combinations needed to perform pairwise tests over pairs of treatments

• esvis: to compute cohen's d effect sizes

• hrbrthemes: plot theme

• kableExtra: formatting and exporting nice tables

To reproduce the analyses, simply run the FDJ_analysis.R script. It calls all other analysis scripts and saves figures as high-resolution .png in the /Figures folder and tables both as .tex source and rendered .pdf files in the /Tables folder.