I spent some time this morning analyzing the Twitter data again. This time I tried to analyze it using LightSide, but I ran into a few problems. Either way, I’m interested to see exactly what happens when I get the software to work. Today I kept getting error messages. On top of that, my rig at home is a bit outmoded for this sort of analytic text processing.

Twitter References


I’m in the process of analyzing a sample of Twitter references to incumbent candidates from the 2012 Congressional Election. There’s a lot of data to sort through, but one of the things that struck me is just how much attention that a small group of candidates received. The Top 25 candidates received over two-thirds of all tweet mentions.

On top of this, almost all of the most tweeted candidates were members of their party’s leadership or members of the U.S. Senate, which represents more people per capita than the House of Representatives. As best I can tell, the only exceptions were Allen West (R-FL) and Michelle Bachmann (R-MN) who both had a propensity to say crazy things that drew national media attention. Anyway, I guess this just goes to show how American political discourse tends to be: (1) dominated by radicals who play to their base or (2) largely captured by poll-tested party narratives. So much for deliberative democracy. So much for thoughtful candidates trying to run on their beliefs.