Validate a Social Media Movement

Rochelle Sharpe, Adrienne Debigare, and David Larochelle

       It didn’t take long for conspiracy theories about the Boston marathon bombers to take hold on social media.

       On the same day Dzhokhar Tsarnaev was captured in Watertown, Ma, the Twitter feed “FreeJahar got created – and it’s been growing ever since.

       “Look at his hair,” someone tweeted earlier today. “Would a terrorist have nice hair?”

       While hundreds of Twitter followers and more than 6,000 fans of a Facebook page asking whether Jahar has been set up, a bigger question emerges about social media. How can the public possibly know whether the conspiracy theorists are a growing movement or merely a tiny band of people who’ve figured out how to hijack social media to amplify its message?

      Certainly, the Jahar conspiracy theorists now have the nation’s attention, with large newspapers, like the New York Post, picking up the story. “Smitten teen girls stir up #FreeJahar mania for Boston marathon bombing suspect,” the Post’s headline blared last Sunday.

      We propose a tool to assess the legitimacy of opinions spread on social media. It would essentially apply the age-old basic journalism questions, the who, what, when, where, why, to assess the validity of social media posts.

      How many supporters are needed before a movement is significant? The White House has struggled with this question with their We the People petitions. They initially required 5,000 signatures within 30 days to get a response. After being inundated with more petitions than they could handle, they raised the number to 25,000 and then to 100,000.

      When covering an emerging social movement online, it is important to know who actually supports it. We developed a prototype tool to analyze the membership of the main Jamar Facebook group. Although the New York Post portrayed supporters as smitten teenage girls, we found that the Facebook group was overwhelmingly male. We also found that the overwhelming majority of members were likely to be US based. (The facebook group went invite only before we could perform additional analysis.)

      Encouraged by the success of our prototype, we propose the creation of a more powerful suite of tools for analyzing social media movements. These tools would allow journalists to quickly analyze a social movement quantitatively rather than just qualitatively.

      The proposed tool suite would be able to show the following:

  • Word clouds based on the contents of tweets containing a hashtag or posts in a facebook group.

  • What social media services the conversation was taking place on.

  • How many users took the extra step of signing an online petition.

  • How users of a twitter hashtag are connected

  • What types of sources are being cited — research papers, mainstream media articles, blogs, memes, etc.

  • Where do links in the conversation go?