Monday 28 April 2014, 10:32
User-generated content is a horrible term, yet no-one has successfully come up with an alternative or even a clear definition of what they mean when they use the ugly acronym UGC.
But as we all try to come up with some sort of agreement on this, myself and two colleagues - Sam Dubberley and Pete Brown - embarked on an eight-month research project, funded by the Tow Center for Digital Journalism, to study the way UGC is integrated by 24-hour news channels.
It’s important to note that for the purpose of the study we defined UGC as “videos or photos created by people unrelated to the newsroom who are not professional journalists”. It was hard enough doing that. If we’d included online comments and tweeted tips and suggestions it would have been impossible.
The research project has two phases, and the first report was published last week. In this first phase we studied the television and online output from eight 24-hour news channels over three weeks. The eight broadcasters were: al-Jazeera Arabic, al-Jazeera English, BBC World, CNN International, Euronews, France 24, NHK English and Telesur.
When I say studied I mean we sat down and watched 1,164 hours of television news and read/watched 2,254 web pages. Not a simple task. As we couldn’t watch all of the channels live, we had to record all of the TV output and capture all the web pages using a variety of techniques - some straightforward; some involving pressing ‘record’ in the middle of the night.
Here are the headlines. UGC was used on air and online every day. It was used when journalists couldn’t get there, or there was just nothing else. Yet very little of it is credited to its original source or labelled as content that differs from ‘professional’ material created by the news organisations themselves, or that from agencies or freelancers.
On average there were 11 pieces of UGC used per day on television and 19 online. There was a steady use of UGC every day, but the numbers increased significantly during breaking news events.
During our three-week coding period, the Glasgow helicopter crash occurred, which prompted a significant use of UGC as newsrooms scrambled to use Twitter pictures before their own cameras or the agencies arrived at the scene.
Many people equate the use of UGC with the Syrian conflict. Our figures demonstrated that in fact 40% of the total UGC used was from Syria. But there were many other stories that used UGC: from the Ukraine and Bangkok protests, police brutality in Egypt, riots in Singapore and Black Friday shoppers, to a man rescued after surviving a shipwreck. What we didn’t find were many dancing elephants or singing cats. There was almost no video that you would be described as ‘viral’ - the type that you might expect as an ‘and finally’ piece.
The key result for us is the lack of crediting of UGC. When you compare how many times broadcasters add on-screen credit, the numbers vary tremendously. The average was 13%, but that hides big differences between broadcasters. Fifty-one per cent of the UGC used by CNN International was credited, compared to 13% on Euronews, 9% by BBC World and 1% by al-Jazeera English.
Perhaps unsurprisingly, more content was credited online (80% by CNN International, 49% by BBC World), which is clearly because there is a culture of giving credit online; and also the fact that social content can be embedded directly from social networks, which makes crediting easier too.
The issue of labelling, as opposed to crediting, was also fascinating. Crediting is when a broadcaster explains who shot the picture or video - for example, ‘Courtesy: Claire Wardle.’ Labelling is when a broadcaster explains, either as an on-screen description or in the script, that the footage has been created by someone unrelated to the newsroom. This might take the form of: ‘amateur footage’, ‘activist video’ or just ‘from:youtube.com’, or a script saying ‘these pictures have come into us and we’re unable to independently verify them.’
We expected to see a great deal of this, as we thought broadcasters would want to distance themselves from the content, either because of the quality or as an insurance policy if it turned out to be fake or inaccurate. But on average, across networks, 72% of the UGC broadcast on television had no label or description.
One element that was integral to us when designing this project was to move away from the dominant focus on Anglo-American news output. We think we’ve managed to do this - and that’s important because it helps to give a more holistic, global view of how UGC is being used in different contexts.
This is the first phase of the research, which included analysis of one US (CNN) and one UK (BBC World) channel amongst the eight that we looked at. We've also just finished transcribing 61 interviews with journalists from countries as diverse as Iceland, Indonesia, Japan, Kenya, Australia, Iraq, and many more, for the final report which will be published at the end of May.
The reason these interviews are so crucial is that they’ve enabled us to understand the reasons behind the patterns we observed. Some of it is ignorance; some of it fear of ‘screen clutter’; some of it caused by a reliance on agencies for UGC (when some agencies don’t give any information about the uploader). But the pressure broadcasters face in rolling news environments causes a lot of these issues. Do you put the pictures out quickly, or do you stop to double-check whether the person wants to be credited and add an on-screen graphic? Most broadcasters thought it was more important to get the pictures up.
All of this will be explored in the next phase of the research, so, as they say in this business, stay tuned. It will be out at the end of May.
Other College of Journalism blogs by Claire Wardle
Thursday 24 April 2014, 15:36
Wednesday 30 April 2014, 10:22