Replication Data for: How Much Disagreement is Good for Democratic Deliberation?

This distribution includes the data to reproduce the main analyses and balance tests published in the paper. Importantly, the data are organized in a directory structure, so it is best to view the dataverse as a "tree" rather than as a "table." The analyses can be run as-is from the subdiretories in the "main" folder using OpenBUGS. To create the data and inits files for the analysis, load the dataset_goodsites.RData workspace in R and implement disagreement_datasetup.R. To create all figures use disagreement_plots.R. Note that the "goodsites" are the field sites where there was nearly full compliance with assignment so the main text only analyzes these. All data and analyses regarding the "badsites" with less then full compliance are used as a replication in the paper's appendix.