A Blog by Jonathan Low

 

May 2, 2018

Explainers Are Tedious. Fact Checks Feel Partisan. Can Truth Be Delivered Another Way?

Research is demonstrating, repeatedly, that for a variety of bio-chemical and psychological reasons, fake news is proving to be more engaging and believable than its real alternatives.

The question for society and the business it nurtures is whether the problem posed by disinformation is not just political but a threat to the integrity of - and trust in - products and services of an apolitical nature as well. And if so, what can be done about it? JL


Laura Owen reports in Nieman Lab:

It's more difficult than it has been in the past to assume that legitimate news will win over false news. The European Commission proposed new regulations that would require web platforms to provide transparency about search rankings. The rules aim to prevent web platforms from offering unfair terms to small businesses that use their services. The commission also urged platforms to do more to halt the spread of “fake news,” on their sites, or face regulation. It asked web platforms to draw up a code of conduct by July
“Contextual fact-checks can be remarkably successful in correcting misperceptions."

That’s one of the findings of a new whitepaper, written for the Knight Foundation by Emily Thorson, assistant professor at Syracuse University.
“Not all gaps in public understanding reflect a lack of interest,” writes Thorson. “They can also occur when information is presented in a format that people are not capable of understanding.” When Thorson ran a couple of experiments around how Americans understand the national debt, she found that “when news coverage fails to provide the information necessary for understanding an important issue, people fill in the blanks by drawing on the fragmented information available as well their own experiences, which can in turn lead to substantial misperceptions.”
But the ways that newspapers often address this problem, Thorson argues, are problematic. Explainers can be “tedious in practice…people do not come to the news looking for a civics lesson.” Fact-checking initiatives, meanwhile, can also be problematic: “When a news organization fact-checks someone of a reader’s own party, he is likely to rate that news organization as more biased. This paradox illustrates the challenge of fact-checking: Readers profess to want it, but dislike it when the fact-check paints their own side in a negative light.” Here’s what Thorson suggests instead:
This white paper makes the case for a different approach: contextual fact-checking. While the goal of traditional fact-checking is to correct misleading or false statements made by elites, the goal of contextual fact-checking is to correct areas of confusion and misperception among members of the public. By moving the focus from misinformation (false information) to misperceptions (false beliefs), news organizations can simultaneously correct misperceptions among the public and potentially increase readers’ ability to meaningfully engage with the news. At the same time, by moving away from highly politicized “fact-checks,” they minimize the potential for partisan backlash.
In an experiment, Thorson had 391 participants read one of three sample USA Today articles, one of which included a box with contextual information:

Afterward, the participants were asked to answer two questions about the national debt. While the study was small, Thorson found that the contextual fact-check was “extremely successful at reducing misperceptions about the national debt,” nor did it seem to engender any kind of partisan backlash toward the newspaper from Democrats, Republicans, or independents.
Thorson also offers some ideas on how these fact-checks could be included in articles.
The example in the experiment included a highlighted box at the end of the article. Similarly, a contextual fact-check could be included in a sidebar or mouse-over text. An alternative strategy is to include the factcheck in the main content. This approach has precedent: For example, when most media outlets refer to a member of Congress, they provide her district and party affiliation. The same strategy could be adopted for mentions of issues. For example, an article mentioning Donald Trump’s campaign statement, ‘I am going to protect and save your Social Security’ might also include the statement, ‘Social Security is a federal program that provides benefits to retired and disabled Americans. It is funded through taxes on people who are currently working.’
Do we need to shift how we think about the First Amendment? That’s what Duke professor Philip Napoli writes in a new paper, “What If More Speech Is No Longer the Solution? First Amendment Theory Meets Fake News and the Filter Bubble,” in which he writes that the way news is produced has shifted so much that it’s
more difficult than it has been in the past to assume that legitimate news will systematically win out over false news. Thus, just as it has been asked whether the assumptions underlying the Second Amendment right to bear arms (written in the era of muskets and flintlocks) are transferrable to today’s technological environment of high-powered, automatic assault weapons, it may be time to ask whether this fundamental aspect of First Amendment theory, crafted in an era when news circulated primarily via interpersonal contact and print media, and in which electronic media were just beginning to develop, is effectively transferrable to today’s radically different media environment.
“First Amendment theory has seldom grappled with the issue of truth versus falsity,” Napoli writes, adding, “As a result, the First Amendment has essentially facilitated the type of speech that, ironically, undermines the very democratic process that the First Amendment is intended to serve and strengthen.”
Who’s checking the fact-checkers? Poynter’s Daniel Funke profiles a couple recent projects. Real Clear Politics’ Fact Check Review seems the most legit, though its methodology has also been criticized by some of the fact-checkers it checks. And here’s some interesting upcoming news from PolitiFact:
In March, PolitiFact received the results of a months-long linguistic analysis of its fact checks, conducted by researchers at Carnegie Mellon University and the University of Washington. The report, funded by a grant the Knight Foundation awarded PolitiFact in June, used natural language processing to analyze about 10,000 articles between 2007 and 2018 to determine whether or not there was a Republican or Democratic bias in the wording.
PolitiFact executive director Aaron Sharockman told Poynter in a message that the outlet will publish the results in the coming weeks.
When one party is in power, newspapers aligned with the other party do better. A new paper by University of Richmond’s Allison M.N. Archer (h/t Matt Grossmann), “Political Advantage, Disadvantage, and the Demand for Partisan News” (to be published in an upcoming issue of The Journal of Politics) looks at how partisan newspaper circulation changes when the president’s party changes. “While many scholars have considered the effects of media consumption on political outcomes like voter turnout or candidate choice, I examine whether political outcomes affect the consumption of partisan news…” she writes. She finds that “when parties are electorally advantaged in presidential contests, demand for their affiliated newspapers decreases relative to demand for papers affiliated with disadvantaged parties.” Archer looked at two local Florida papers’ circulations between 1932 and 2014, plus “the aggregate circulations of local, daily partisan papers in the United States from 1932 to 2004.” When a Democrat was president, the Republican-leaning papers’ circulations increased; the opposite was true when a Republican was president.

Archer also looked briefly at partisan cable news, which, though newer, appears to follow the same patterns: “Even in the months right after the election of Republican Donald Trump, prime-time viewership for MSNBC increased 55 percent from one year prior — a growth rate that is larger than its rivals…In sum, these patterns suggest the results based on local partisan newspapers may generalize to another, more contemporary form of partisan media.”Further research on these other forms of media would be good, especially considering how rare it is these days for any region to have two local papers competing with each other. In the case of the Florida papers that Archer looked at, for instance — The left-leaning Tampa Bay Times and right-leaning Tampa Tribune — the Times acquired the Tribune in 2016, and there’s now just one newspaper for Tampa Bay. (One of the questions in the acquisition FAQ for readers: “I’ve preferred the Tribune because I want conservative viewpoints on the editorial pages. Does the Times publish those?”)
The EU tells big tech companies that the clock’s ticking. The European Commission rolled out new guidelines for online platforms and search engines in Europe. From The Wall Street Journal:
The European Commission — the bloc’s executive body— proposed new EU-wide regulations that would require web platforms, including search engines like Google, to provide more transparency about their search rankings. The rules also aim to prevent web platforms from offering unfair terms to small businesses that use their services to sell or promote products.
Ahead of European Parliament elections in 2019, the commission also urged platforms to do more to halt the spread of disinformation or “fake news,” on their sites, or face possible regulation.
It asked web platforms to draw up a voluntary code of conduct by July with the aim of seeing changes by October. It said the EU was first pursuing industry self-regulation, but if that fails it would explore other options “including regulatory ones targeted at a few platforms.”
Some of the other measures the EC outlined include increased media literacy programming and a network of European fact-checkers; those are here. Reuters’ Rasmus Kleis Nielsen argues that a lot of these are awfully vague (thread).

0 comments:

Post a Comment