Social media’s potent ability to rapidly diffuse information and disseminate ideas has made it an important part of today’s news distribution. The massive rise of communities like Reddit, the aggregation and discussion site, is a testament not only to people’s desire for rapid, real-time information, but also their belief that crowd-sourced news is somehow better; with faster updates and more cross-checking by diverse participants making it essentially more accurate.

But false memes, or social ideas, can spread wildly, with sometimes dangerous results. The Boston Marathon bomber manhunt in Watertown is an example of just how fast misinformation can spread. When the FBI released pictures of the bombing suspect on the internet after the attack in April 2013, thousands of social media users acting as sleuths began frantically posting updates on internet forums. Unsubstantiated and inaccurate information was circulated and a missing student wrongly identified as the culprit. The student’s name was reported on Reddit discussions, Twitter feeds and internet sites to the pain and distress of his family and friends. What’s worse, social media wanted to believe that its reporting techniques were superior to old school methods, such as traditional journalism and reporting.

For a better understanding of how speculation and false information can be picked out of a sea of competing ideas and quickly become “fact”, we studied the event as it played out on Reddit. We examined five live thread discussions which took place before the identity of the bombers was made public, assessing 42,500 comments over a period of nine hours. These comments included the emergence, dissemination and survival of false rumours, which were quashed just as quickly as they emerged once the true culprits were identified. But in the meantime, they gained rapid popularity and legitimacy. We wanted to understand what drives people’s beliefs (for or against emerging memes) in social media, and in this case particularly in news-source media.

What we found was that, while social media provides for an engaging and rapid dissemination of ideas, its unique features raise warning signs for users and society, especially in the context of collective sense-making.

The messenger does not matter

Most surprising, and perhaps of greatest concern, was the lack of consideration given to the origins of an idea or the credibility of the messenger whose opinions were being taken as fact. While traditionally readers and news services will take into account the reliability, expertise and quality of a source before quoting that person, the “messenger” seems to carry little weight on social media. Reddit scores its contributors on multiple counts: a set of historical scores representing their experience and popularity on the site with regards to starting discussions or sharing links, and comments made; and a popularity score  which is peer-voted, giving readers the chance to up-vote or down-vote an individual’s comment. Taken together, these measures provide an indicator of the credibility of the messengers/contributors. In our study, however, it had no impact on whether or not an individual’s post influenced the position taken by subsequent contributors.

The influence of noise and momentum

While the quality of the messenger didn’t seem to matter, the effort contributors put into conveying their message did. The more space (or words) individuals took to explain their theories or ideas, no matter how misguided, had an impact. Moreover, momentum played a significant role. The more individuals were bombarded with information – multiple messages from different sources in quick succession – the more likely they were to be persuaded into that line of thinking and to attempt to convert others. That is, online, people were susceptible to simply the timing and succession of information – fast-following comments give the appearance of momentum and, in turn, credibility.

We also found that simply more “noise” around an idea led to adoption – comments that were neither pro nor con but just showed “activity” drew people to commit to positions. Relatedly, we found that polarisation in the debate made it very hard for people to remain neutral or judicially cautious as they entered the fray. Unfortunately, when faced with having to make a choice for or against an idea, many will swing in the wrong direction.

The free-wheeling nature of social media democratising news and information – all things to applaud – creates an environment where mistakes are easily made and false ideas spread rapidly. While these mistakes can be quickly rectified as new information comes to hand, the damage may be serious. Traditional media, aware of the cost of getting things wrong, persists with fact-checking. In the case of the Watertown manhunt, mainstream news outlets were chastised for taking too long to cross-check facts with police reports, delaying filing stories until they were verified. Meanwhile on social media, where individual responsibility is non-existent and there is very little to lose, getting “facts” out fast remained the priority. But it does come with a cost. Particularly worrying is the susceptibility to momentum effects and other dynamics. We need to be careful not to overestimate the quality, or underestimate the challenges, of social media.

Social media is a completely new paradigm of communication and one we are only beginning to understand. The power, reach and speed of discussion forums like Reddit are phenomenal. While there are many positive examples of crowd-sourcing as a catalyst for idea collaboration and technological innovation for the greater good, the potential for the spread of misinformation cannot be ignored. This continued area of work has implications for organisations managing product rumours and for governments combating the proselytisation of extreme ideologies.

Charles Galunic is a professor of Organisational Behaviour  and The Aviva Chaired Professor of Leadership and Responsibility at INSEAD.

Chaitanya Kaligotla was awarded a PhD in Decision Sciences from INSEAD and is a Postdoctoral Appointee at Argonne National Laboratory.

Follow INSEAD Knowledge on Twitter and Facebook.