Facebook fake-news labeling doesn’t work, study says

Labeling fake news on Facebook isn’t working very well, a new Yale study shows.

Nearly a year after the U.S. presidential election, the fallout from fake news keeps piling up — and so is the scrutiny of Facebook’s efforts to fight it.

The study published this week by Yale University researchers found that tagging news as “disputed by third-party fact-checkers” made participants just 3.7 percentage points more likely to correctly determine headlines were false.

Get tech news in your inbox weekday mornings. Sign up for the free Good Morning Silicon Valley newsletter.

The study, first reported by Politico, also found a “backfire effect” — “particularly among Trump supporters and those under 26 years of age” — in which untagged fake news stories were then seen as more accurate than real ones. The researchers said that’s because it’s impossible for Facebook’s new system, which the social network rolled out in March, to label all fake news as such.

The findings are concerning because young people disproportionately get their news from social media, the researchers said. For example, Pew Research published a study last week that showed 78 percent of Americans under 50 years old use social media as a news source.

The study, which the authors will submit for peer review, involved more than 7,500 participants. They were split into groups and shown news articles that were posted on Facebook in 2016 and 2017. The control group got no headlines labeled as disputed, while other groups were shown some headlines that were flagged as disputed. They were then asked to judge whether the headlines were real or fake, and whether they would be willing to share the stories online.

“The main potential benefit of the tag is that it (slightly) increased belief in real news headlines,” the researchers said. “This seems insufficient, however, to stem the tide of false and misleading information circulating on social media.”

Some specifics from the findings, which resulted from studies conducted in July and August:

  • There was “a large backfire” among people ages 18 to 25 but “little backfire” among people 26 years old and up.
  • There was “a large backfire” among people ages 26 and older who are Donald Trump supporters. There was “a positive spillover” —  less belief in untagged fake news — among the same age group who are Hillary Clinton supporters.

The researchers noted that they also found that Trump supporters had less confidence in third-party fact-checkers, local and news organizations and social media. They had more confidence in friends and family as sources of information.

Since labeling fake news doesn’t seem to be working well enough, what should Facebook do?

“We don’t have good answers to this yet, in large part because Facebook is extremely reticent to give researchers access to data,” one of the researchers said in an emailed response to SiliconBeat Wednesday. David Rand, an associate professor in psychology, economics and management at Yale, added that he is doing further study.

Along the same lines, some of Facebook’s partner fact-checkers complained last week that they don’t know whether their work is making a difference because the company won’t share internal data.

When reached for comment, a Facebook spokesman said Wednesday: “Disputed flags is just one product that exists inside a larger multi-pronged strategy” of fighting fake news, and that the flagged items are “powering other systems that limit the spread of news hoaxes and misinformation.”

Of the Yale study, he said, “this is an opt-in study of people being paid to respond to survey questions. It is not real data from people using Facebook.”

The Yale study also found, from surveys conducted in July and September, that Facebook’s new practice of putting publishers’ logos next to news stories isn’t effective.

“Slapping down a big logo also doesn’t do anything,” Rand told Politico. “People don’t find mainstream media outlets particularly credible.”

Wednesday, Facebook released new guidelines that detailed, among other things, how purveyors of fake news could become ineligible to make money from Facebook ads.

 

Photo: Facebook CEO Mark Zuckerberg introduces the redesigned Facebook News Feed at Facebook’s headquarters in Menlo Park on March 7, 2013. (John Green/Bay Area News Group)

 

Tags: facebook, fact checking, fake news, news

Source

The post Facebook fake-news labeling doesn’t work, study says appeared first on Tech News Blogs.

Powered by WPeMatico