Study: Twitter Activity Nearly Flawless In Predicting Election Winners
Bloomington, Ind. (CBSDC) — The more Twitter activity there is about a political candidate, the more likely they are to win their election, a new study finds.
According to researchers from Indiana University, campaigns can use Twitter to successfully predict the winner of most political races. A candidate’s “tweet share” percentage – or how often the candidate is mentioned on the popular social network in comparison to their opponent– correlates to their likelihood of winning.
“We plotted it and thought, ‘Holy moly, it was a very strong correlation,’ ” Fabio Rojas, a sociology professor at Indiana and one of the study’s co-authors, told the National Journal. He added that preliminary analysis of last year’s congressional elections show similar results. “Think of this as a measurement of buzz.”
The study’s analysis of tweets from the 2010 midterm elections found that the Twitter data correctly predicted the winner in 404 of the 406 House races across the country.
Rojas and the researchers found that not only does the raw, total number of tweets about a candidate not matter, but whether or not the comments were positive or negative did not matter either.
Initially, the researchers simply looked at the hundreds of thousands of 2010 race tweets only to find that popular candidates such as Rep. Michele Bachmann, R-Minn., and candidates from wealthier or bigger districts dominated the social network. But the findings did not correlate with the winner.
Instead, the researchers looked at the “horse race” factor to directly compare the number of tweets a candidate earns vis-à-vis their opponent. The candidate who won 404 of 406 times was the one who had the higher percentage of tweets about them versus their rival.
And when it comes to positive or negative, the study found that any publicity is good publicity.
“Are you going to talk about the guy who loses or the guy who wins?” Rojas told the National Journal. “You’re going to talk about the winner, even if you hate the winner.”
Rojas said that the study should revolutionize campaign research, and possibly save millions of dollars in campaign financing because politicians can now just check their Twitter statistics.
“The point is, it’s cheap,” he said. “Once you start up software for collecting tweets, it’s very cheap. It took one of my Ph.D. students a couple of weeks to set it up.”
Follow WNEW on Twitter.