In an age of misinformation, sustainable newsroom collaborations should become a priority

The benefits for collaborative verification projects demonstrate that funders should be supporting newsrooms and fact-checkers to work together on a long-term basis

By Claire Wardle (Versão em português brasileiro)

It’s not often that when you hear seasoned journalists admit something has rekindled their love for the craft. This year has been very difficult for journalists globally, and most of what we hear relates to journalists under physical, verbal and online attacks from politicians and trolls.

So earlier this month when a stream of WhatsApp messages from Brazilian journalists popped up on my screen, full of fist pump emojis and love hearts, it was clear something was going on. The messages were being shared in a WhatsApp group used by the 62 journalists who powered our collaborative election project, Comprova. On a daily basis, the group shared hundreds of messages as they decided what pieces of misinformation to debunk, how to verify the rumors, doctored photos and misleading memes, and how to word headlines responsibly. But as voting ended and the project concluded, the journalists started sharing with each other what the project had meant to them personally. And the messages were astonishing.

These “debunking” projects are often judged by whether or not they slow down or stop misinformation. Understanding this impact is almost impossible to do without data from the platforms, but we are in the process of measuring and evaluating the impact of the project using data we collected (more about this in a subsequent post).

But as well as the impact on the audience, we also need to acknowledge what these projects do to a news industry. This type of “CrossCheck” project was piloted in France during the election in Spring 2017. Our post-project evaluation showed that the project had significantly improved verification skills in newsrooms large and small, but perhaps more importantly, built significant trust between journalists in normally competitive newsrooms. Comprova has shown similar results.

You can read more about the details of Comprova here, but the basic premise is that journalists from 24 Brazilian newsrooms worked collaboratively every day to find, debunk and report on different rumors, false and misleading content circulating ahead of the Brazilian elections. Slack hasn’t taken off in Brazil so the reporters collaborated in a WhatsApp group they called the Comprova Verificadores (the nickname they gave themselves). The group was so busy that on some days there were more than 1,000 messages shared. One journalist would write up the report, and the other journalists were asked to then “crosscheck” the reporting. If they agreed with the quality of the work, they added their newsroom’s logo, and once three newsrooms had done this, the report was published on the central site.

The magic of the CrossCheck model, is that while Comprova did build up somewhat of a following, over a three month period we could not (and should not) compete with the newsrooms who were part of the coalition. So along with the report being seen on the Comprova website, partners that were happy with the “crosscheck” published their own versions on their sites, or broadcast stories on their own radio or television channels. For example, the central Comprova website had about one million unique page views. Just one of our 24 partners reports traffic of almost 7 million uniques.The amplification machine was powerful.

Jornal do Commercio, based in Recife, ran full pages of Comprova debunks in its weekend editions several times during the project.

In total 147 reports were published on the main Comprova site and many many more were published via the partners (we’re in the process of collecting the final number). Jornal do Commercio, based in Recife in north eastern Brazil, published a collection of the debunks in a weekend edition several times during the project replicating the site for the print audience.

One of the elements that the journalists are most proud of is that not one correction was made during the project. As we saw in our previous project in France (which similarly made no errors in 10 weeks), forcing newsrooms to work together increases standards. Being held to account by your editor is one thing but being held accountable by your peers in competing newsrooms is another.

Having these cross-newsroom communication channels also led to in-depth discussions about the ethics of covering a particular piece of content, or how to frame the headline. We saw the same conversations play out in France. Doing journalism in an age of misinformation is increasingly challenging. Newsrooms are having to figure this out as they go.

In the U.S., newsrooms make poor judgments daily in terms of giving unnecessary oxygen to conspiracies or fringe personalities, or wording a headline or tweet in a way that reinforces a falsehood. These often result in a Twitter eruption of indignation but little changes. These types of supportive inter-newsroom channels allow for thoughtful consideration of these complex issues and new challenges.

The overwhelmingly positive experience for those journalists who took part in Comprova means that most want the project to continue. We will be traveling to São Paulo this week for a well-earned “wrap party” and to make plans on how to establish Comprova 2.0 for an ongoing collaborative effort to fight misinformation in Brazil. Unsurprisingly, misinformation didn’t stop on election day.

This project was ambitious and at times we were close to giving up. There were many sleepless nights and five overnight flights to Brazil. But it has all been worth it. Comprova has created a sustainable response to misinformation, created a team of journalists who could now train other journalists on how to discover, verify and responsibly debunk misinformation, and for some it has rekindled their love for journalism.

And for the academics in the room, we now have the most incredible misinformation datasets, which has always been the point of these projects. We will ask in our research how can we do good work in the field, but work that ultimately helps the academic community learn more about the information ecosystem across platforms?

This election project is the sixth one we’ve created; we have tried different methodologies and have learned a great deal. Our key takeaway after №6 is: While elections are the time when funders and newsrooms get excited about what is possible, we need to build ongoing sustainable collaborations in preparation for when the next elections come. Misinformation is not just a problem during election seasons. We need to provide newsrooms with training on social discovery, verification, and responsible reporting techniques when it comes to misinformation. We then need to establish ongoing collaborative projects to work so those newly found skills don’t die, and we need to build networks of “ambassadors” willing to spread quality content in WhatsApp groups. We have a lot of work to do, but journalists in France and Brazil in particular have shown what is possible.

Even I have had my enthusiasm for this topic rekindled. And it’s been a tough year!

(Versão em português brasileiro)

Comprova website
Previous articles about Comprova:
- Comprova wraps in Brazil
- What WhatsApp “API Access” Meant for Comprova