To report bugs, errors, or to suggest features, please contact us at bwalenz-cs-duke-edu.

Congressional Voting Records

We retrieve bulk raw data of Congressional voting records (2009-01-06 to 2016-09-22) from GovTrack.

Images of legislators are hotlinked to GovTrack.

Presidential Positions

Data on the President's positions on votes in the Congress (2009-01-01 to 2014-12-31) are from the Congressional Quarterly, last retrieved on 2016-09-24.

Key Votes

We use key votes identified by various research, lobbying, and political organizations:

Contested Elections

Information on contested races in the 2016 Elections is from the Cook Political Report, last retrieved on 2016-09-16.

Choice of Legislators to Visualize

When a large group of legislators is being compared, it becomes impossible for iCheck to show everybody in the group in one visualization. In this case, iCheck shows a subset of legislators that may be more easily "recognizable." Specifically, iCheck uses a combination of three measures: leadership roles in the Congress; appearances on Sunday news shows (tracked by Congressional Quarterly's Face Time, last retrieved on 2016-09-16); and online social influence (as measured by Klout; last retrieved on 2016-08-23). Note that one can always see every legislator in the group by moving your mouse over the bars in the rank view.

Minimum Number of Votes Required for Calculating Percentage Agreement

When calculating percentage agreement between a legislator and a target, iCheck requires that they cast a minimum number of 10 votes together (because a small sample would not give a meaningful assessment of agreement). Consequently, for example, iCheck will exclude, from a comparison group, any legislator who cast fewer than the required number of votes in the selected period. As another example, in the trend view, there may be gaps in a legislator's timeline, corresponding to times when this legislator cast fewer than the required number of votes in a 90-day window.