October 19, 2021

I, Science

The science magazine of Imperial College

Twitter spreads misinformation and adds little to political debate, so how useful is it? ...

Captain Twitter

An American voter sits alone in her living room watching a much-anticipated Obama vs Romney debate on TV. It’s the 2012 elections and she’s following the debate on her phone’s Twitter app at the same time. She reads another sarcastic tweet from a well-known political commentator and for the fifth time in two minutes, hits the retweet button.

“During the debates we found that people tended to reply to one another less and retweet more,” said Drew Margolin, Assistant Professor of Communication at Cornell University: “We call it the rising tide of retweets.” Margolin was disappointed by the results of his recent research, which he hoped would find that Twitter enabled a number of diverse and original voices to contribute to the TV debates during the run up to the last election. In fact, he found that the bulk of the attention was given to a few key commentators who conversed amongst themselves and were slavishly retweeted: “What you have is a recreation of what we already had pre-social media, which is people just echoing what the TV talking head personality is already saying,” he said.

Twitter started as a social media platform in 2006 and has grown in popularity at an astronomical rate since then with over 58 million tweets now posted every day. The company has 650 million users worldwide and was recently valued at $18 billion. But does the software add much to social interaction among its users? Drew Margolin identified that the TV debates mainly served media elites such as star political commentators Bill Maher and Sean Hannity, who have an exceptionally good awareness of how social media works and exploited their knowledge to dominate the Twitter discourse and reap the rewards of exceptionally high public attention for themselves.

Counting the limitations

Manipulation of Twitter in this way brings into question how useful it is as a tool for public discourse. Dr Miles Osborne of Edinburgh University’s School of Informatics in the UK has studied a number of social media platforms and has tried to uncover their value and their failings. He’s found that Twitter performs worse than other social media such as Facebook and blogs when it comes to discussion because of the way it’s set up. The other platforms allow users to see the history of a conversation before they add their comment: “When you can see the posts all at once, it’s much easier to have a back and forth discussion. In Twitter, it all goes into the ether. This is a limitation of the technology.”

Perhaps not the only limitation. Supporters of the technology say a central function of Twitter is that it enables immediate reaction, but the value of this has been questioned in a recent study by Kate Starbird, Assistant Professor in the Department of Human Centered Design and Engineering at the University of Washington. Her work examined the spread of misinformation on Twitter in reaction to the 2013 Boston Marathon bombing. She found that a lot of misinformation was retweeted but, when the information was discovered to be wrong, the corrections weren’t retweeted to nearly the same degree: “In one case there were 90,000 tweets that spread the rumour and about a thousand that corrected the rumour,” she said.

Rumour spreading is a coping mechanism during disaster events and a natural human response but the Twitter technology enabled rumours to spread further and faster, which increased the number of people misinformed. When the Twitter crowd misidentified the bombing suspects, the percentage of crowd correction tweets was relatively high – as many as 20% of the initial tweeters posted a correction tweet, but this is still only a fifth of the initial tweeters and Starbird feels it isn’t enough. She’s working on an algorithm to detect crowd correction and then amplify that signal: “When somebody somewhere in the crowd is attacking the information, the algorithm would make that more visible to people who are seeing the information as it appears in Twitter.”

Instructions for use

Starbird’s crowd correction tool may be two years away and in the meantime Twitter is still susceptible to spreading misinformation while seemingly adding little to public political debate. Is there any use for Twitter at all? Osborne thinks Twitter can be helpful as a news aggregator because it brings stories from several different sources to one place, which saves the user from chasing news on multiple websites. It also offers gossip, jokes, and sentiment, but this seems like an underuse of a technology that has so much potential and so much reach.

More can be done according to Margolin who feels the key is to create Twitter instructions or a code of practice. He sees Wikipedia as the gold standard in this area that works well because it has both good technology and an accepted method of use. In contrast, he says Twitter: “doesn’t have rules, it just has the technology.”