Twitter is no longer only a social media platform used to interact with peers, coworkers and favorite celebrities. Now it is even being relied upon to predict winners of competitive elections.
In a study conducted at Indiana University, new research accurately exhibited public opinion extracted from social media platforms to be correlated with how the public will vote. Using an archive of billions of randomly sampled tweets from 2010 prior to the 2012 presidential election, combined with a specific method, researchers were able to predict the winner in 404 out of 406 competitive races. This is the first time a working model for predictive Twitter analysis has seen such success.
This study also demonstrated the significant correlations between how many times an incumbent for the House of Representatives was mentioned in a tweet and his or her success of winning. Even when the public discussed a candidate in a negative context, it was still an indicator of whether he or she was on the cusp of victory. Ultimately, the more a candidate is mention on Twitter, the more success he or she will see.
Here at PMBC Group, a Los Angeles Technology PR Firm, we are intrigued by the study conducted at the University of Indiana, and its data that can predict the winner of competitive races utilizing Twitter. We look forward to witnessing how this will influence the way campaigns are run in the future.
Have you discussed politics on Twitter?
PMBC group is a multi-level PR firm based in Los Angeles, specializing in Corporate PR and Tech PR. Visit us at http://www.pmbcgroup.com.