https://senate-similarity.netlify.com/
I found an article from Alberto Cairo's blog, the functional art, lauding the virtues of a US Senate voting similarities visualization from Gramener. He made suggestions for improvements that I wanted to take on. His ideas turned out to be too ambitious for this homework. So I have scaled back my expectations. Andrew Gelman made an argument that a simpler version of the graphic would be more useful and Alberto followed up with a new version. Cairo's new version with Gelman's suggestions is simpler and follows many of the guidelines we learned from the class. The new version uses more natural to interpret features like cartesian coordinates and leans on the political left-right relationship to make it more intuitive.
I think the polar coordinate system is more attractive and engaging, but Gelman makes a good argument against its usefulness. I have given the arc position meaning by placing the circles representing senators based upon their ideology score from GovTrack.us. Cairo addresses this issue in the same way and makes for a good comparison when determining the better visual.
I have shared my version with a few friends, and about half of them are confused with the selected senator at the center as pointed out by Gelman. Interestingly, when I showed them Cairo's version of the same kind, they dismissed it quickly understanding it, but because my more "complex" version engaged them, they clicked around and explored.
I like this visualization for the way to shows a complicated relationship between party affiliation and vote support. The interactivity provides a method for exploring some exciting edge cases.