Take South Africa. When the pollster asked people how many immigrants they thought made up the country’s population, the average guess was 29%. The actual figure is 5%, it says.
How many of South Africa’s prisoners were born in a foreign country? The average guess was 37%, against the real figure of 6.3%. And nearly 80% thought South Africa’s murder rate had risen since 2000. It actually fell 29%, Bobby Duffy, who manages the firm’s social research institute, said.
These and other insights were shared at Global Fact V, the International Fact-Checking Network’s fifth gathering of global fact-checkers, held in Rome from 20 to 22 June 2018.
(Note: We cannot speak to the accuracy of the Ipsos MORI findings about South Africa but only use them to illustrate the nature of the wider debate.)
‘Why we are wrong about nearly everything’
Just as startling was people’s response when they were told the facts. They still insisted they were right. “People come into the country illegally so they aren’t counted” or “it is what I see in my local area” were typical pushbacks.
In the UK, even generously estimating for illegal immigration wouldn’t add one percentage point to the 13% share of immigrants in Britain, Duffy says in his forthcoming book on misperceptions. Britons guessed 25%.
The book – The Perils of Perception: Why We’re Wrong About Nearly Everything -and its premise was perhaps scant comfort to the more than 200 fact-checkers who gathered in Rome to take stock of the battle for accuracy in public debate.
As Duffy emphasised to the audience, in the face of identity, ideology and partisanship, facts alone won’t change misperception. It is equally important to understand the role of emotion and story narratives in shaping beliefs.
The conference has grown rapidly, from three dozen fact-checkers in 2014 to more than 150 from about 50 countries in 2018. The 230-strong gathering (more than 1,000 applied) included academics and, significantly, representatives from Google and Facebook (including Whatsapp).
Some notable discussions included:
- Audience scepticism about the mainstream media spilling over into fact-checking. Is selection bias of what is checked a concern? To answer this, US-based Politifact went to pro-Donald Trump states and did local fact-checking. Their overall take was that fact-checking could be better explained to audiences, especially local ones. “We don’t do a good enough job, probably, of explaining how we do what we do,” said Aaron Sharockman, the executive director of Politifact.
- Some fact-checkers have overhauled their entire model. Checknews.Fr sees itself as “the human search engine”. In response to a perceived mistrust of the media, they no longer listen out for claims in the media. Instead, they respond only to research queries from their readers.
- Is the (fake) videogeddon upon us? While manipulated videos and deepfakes are a talking point, experts felt that the technology is still rudimentary and pretty hard to do for normal people. The tools to detect them are great, but classical fact-checking is still key. However, it could get harder for fact-checkers in future.
- Clara Jimenez Cruz of Maldito Bulo, a Spanish fact-checking platform, said it was important to share fact-checks on the same platforms used to spread misinformation.
- What are the social media giants doing to help counter misinformation? Facebook’s Grace Jackson said the firm had found that on its newsfeed, structured headlines that included a rating and a synopsis of a checked claim were more effective in holding the attention of fast-scrolling readers.
Tessa Lyons, a product manager at Facebook, said the technology giant was working with fact-checkers to downgrade false articles, using machine learning to find and flag duplicates of already debunked articles and looking for ways to get claims rated faster.
- And does fact-checking have a “women problem”, given that audience demographics seem to lean towards men? Interestingly, women are well represented in most fact-checking organisations.
- What is the future of automated fact-checking? Spoiler alert – fact-checkers are not about to be replaced by robots!
This was the type of information (and misinformation) amphitheatre fact-checkers found themselves in, with Rome’s Colosseum fittingly just a few steps away.
We will see you at the next the annual conference – in Cape Town in 2019!
PS: Africa Check picked up the inaugural “Best correction obtained” award for the fact check that led the World Bank to correct a public brief about depression in Nigeria.
© Copyright Africa Check 2017. You may reproduce this piece or content from it for the purpose of reporting and/or discussing news and current events. This is subject to: Crediting Africa Check in the byline, keeping all hyperlinks to the sources used and adding this sentence at the end of your publication: “This report was written by Africa Check, a non-partisan fact-checking organisation. View the original piece on their website", with a link back to this page.