Welcome to the Liquid Democracy Replication Package! This repository provides the code, data, and materials needed to reproduce the analyses and results presented in the study Tracking Truth with Liquid Democracy. Below, you'll find a detailed structure and description of the repository, along with information on each file and dataset.
The code folder contains Jupyter notebooks and an R Markdown file used in the empirical analysis and graph generation.
| File | Description | Paper Material |
|---|---|---|
EmpiricalAnalysis.ipynb |
Main notebook for performing empirical analyses of the experiments conducted and presented in the paper. | Figures 1, 2, 5, 6, 7, 8, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24 & Table 2, 3, 8, 13, 14 |
R_notebook_LiquidDemocracy.Rmd |
R Markdown file with performs statistical analyses specified in the equatiosn described in the paper. | Figures, 9, 10, & Tables 5, 6, 7, 17 |
The data folder includes various datasets used throughout the analyses. Below is a table describing each dataset and its columns.
| File | Description |
|---|---|
anova.csv |
Contains the data to perform an ANOVA analyses for gender correlations. |
data.csv |
Raw data collected on voting preferences and participant choices (anonymized). |
LDvDD_means.csv |
Aggregated means data comparing Liquid Democracy (LD) and Direct Democracy (DD) results. |
LDvDD.csv |
Detailed dataset comparing LD and DD by question. |
phi_w_gmm.csv |
Estimated |
R_paper_pre_study.csv |
Data from the pre-study stage, used in R analyses. |
R_paper.csv |
Final dataset used for the main study analyses in R. |
(See the data/README.md for column information about the data.)
The figures folder stores all figures generated by the code files. These figures include visualizations and graphs that illustrate key points from the analyses.
The material folder contains supplementary material used for the surveys conducted during the study. This includes images of place (participants had to guess where the image was from), sounds (participants had to guess what soundtrack this was), etc. Some of the material was borrowed from Stanford's wisdom of crowd experiment.
This directory maps every single Figure and Table in the paper to the piece of code that produed it.
- Clone the repository.
git clone https://github.com/ManRev/liquiddemocracy.git
- You may need a R interpreter to run
R_notebook_LiquidDemocracy.Rmd. Example: download R Studio here. The packages you will need are listed in the first cell of the notebook. - To use
EmpiricalAnalysis.ipynb, you will need to install packages fromcode/requirements.txtPython 3.7.6 was used. - Navigate to the
codefolder and open the main analysis notebooks to begin reproducing the results. - Reach out to mrevel@mit.edu shall you have any question!