About the Author:  Bill leads the FortiusOne data analysis team. You can follow him on twitter @billfgreer. Read more from this author

The Elections are over and Barack Obama won. Aside from historic nature of electing our first African-American President, this election was also historic based on voter turn out and the technology that was used to help make the election possible and entertaining. This election was surrounded by new technologies and innovation, including Holograms from CNN, Get out to Vote Drives on Facebook, Pandora (they had a find your polling place widget), and other social media platforms. By far the coolest Election application was the Twitter Vote Report. This service allowed users to tweet information on the go from their polling locations, giving information about wait time, ranking the polling location, and other quality indicators, all of which was updated to a live map. FortiusOne was joining in on the fun by making as much election data public and available as possible. Here are some examples of our political/election datasets and maps, including the full twitter results:

Twitter Voter report (end of day), USA, Nov. 4 2008
Polling locations, Maine, 2008
2008 Polls vs. 2004 Election Results with Socio-Economic Indicators, USA, 2008
Voting Districts, New York, 2000
FEC, Individual donations to Obama campaign during August, 2008, USA, August 2008

The most interesting data was the Twitter Vote Report Data, so we thought we would try to run a little analysis on where the Tweets were coming from and where the wait times were longer.



The first map shows counts of Twitters – Seems to correlate strongly with high tech corridors in the region.

The second is pretty interesting – really shows a spatial divide in the region in terms of average wait time. Lower wait times in the Western portion of the region, higher wait times in the Eastern portion. Orange represents high values and blue low values.

Seeing that the wait times were longer on the eastern part of DC we decided to run some statistical analysis to see if the longer wait times were correlated with race and ethnicity.

the numbers in the tables are correlation coefficients – a correlation coefficient has a possible range of -1 to 1. Anything closer to 1 or -1 indicate stronger associations between the variables. Also, negative coefficients indicate negative association and positive coefficients indicate positive associations. So, the percent black field a correlation coefficient around 0.15, which means that counties that have higher percentages of blacks had higher average wait times. This only shows a slight correlation, but its there.

Here are the correlations between Twitter users and race

Keep an eye out, we’ll be updating post-election data with full results, voter turn out, and other interesting tid-bits. Feel free to let us know what you would be interested in seeing in Finder!

Popularity: 12% [?]