Our Analysis Methodology: Technical Details

We’ve added a new Analysis Methodology page. This page provides details on our ideology, leadership, and prognosis analyses so that they can be understood better by our users and so they can be replicated by researchers in other domains.

GovTrack pioneered three large-scale (i.e. BigData) statistical analyses of legislative information. You can find the results of these analyses on different pages throughout the site, such as on the pages for Members of Congress and for bills. The new methodology page outlines how those analyses are performed, and where sensible it evaluates whether the analysis was successful.

The Ideology Analysis compares the sponsorship and cosponsorship patterns of Members of Congress to put them on a scale roughly from liberal to conservative. We first began publishing this “principal components analysis” in 2004, then calling it a political spectrum. Follow the link for charts, Python source code, and references.

The Leadership Analysis looks at who is cosponsoring whose bills to see who the legislative leaders are. It’s a little like if you scratch my back will I scratch yours? The analysis is based on Google PageRank, the algorithm Google uses to order search results. We first began publishing leadership scores in 2010. As far as we know, this analysis is unique to GovTrack. Follow the link for charts, Python source code, and references.

The Prognosis Analysis computes a probability that a bill will be enacted based on a logistic regression model. The analysis helps us identify which bills are important, and it helps explain the legislative process in practice by showing the factors that contribute to a bill’s success or failure. We first began publishing leadership scores in 2012. Click the link for methodology, references, the list of factors considered, accuracy, and precision-recall charts.