This project studies the interactions on X (formerly Twitter) of members of the US Congress. The aim is to identify polarization movements at work in Twitter interactions, and particularly in the practice of retweeting, whether based on political party, age or gender, which could shed light on the underpinnings of American politics. The data was compiled from Twitter API and posted by Christian G. Fink, Gonzaga University and relies on 530 Twitter user IDs for each member of the 117th United States Congress. The author retains congress members that issued at least 100 tweets over the timeframe, leaving 475 available for analysis. 117,974 tweets were retrieved overall from February 9, 2022 to June 9, 2022. C.G. Fink provides two initial files.
Congress members are identified only by names. In order to deepen the analysis, we added manually the sex, age (in 2022), political party and chamber (Senate or House) for each individuals. Table \ref{descriptive_table} describes the overall distribution of Congress members based on the additional data.
A general exploration and visualization of the network is done in Graph visualization and exploration.ipynb
Graphs for descriptive statistics are drawn in Attribute and centrality visualization.ipynb
Analysis of polarization and other statistics is done in Centrality, Polarisation and Blau index.ipynb