Tagging fake news on Facebook doesn’t work

Jason Schwartz for Politico:

Facebook touts its partnership with outside fact-checkers as a key prong in its fight against fake news, but a major new Yale University study finds that fact-checking and then tagging inaccurate news stories on social media doesn’t work.

The study, reported for the first time by POLITICO, found that tagging false news stories as “disputed by third party fact-checkers” has only a small impact on whether readers perceive their headlines as true. Overall, the existence of “disputed” tags made participants just 3.7 percentage points more likely to correctly judge headlines as false, the study said.

This is particularly disappointing:

The researchers also found that, for some groups—particularly, Trump supporters and adults under 26—flagging bogus stories could actually end up increasing the likelihood that users will believe fake news.

You are the product

John Lanchester:

What this means is that even more than it is in the advertising business, Facebook is in the surveillance business. Facebook, in fact, is the biggest surveillance-based enterprise in the history of mankind. It knows far, far more about you than the most intrusive government has ever known about its citizens. It’s amazing that people haven’t really understood this about the company. I’ve spent time thinking about Facebook, and the thing I keep coming back to is that its users don’t realise what it is the company does. What Facebook does is watch you, and then use what it knows about you and your behaviour to sell ads. I’m not sure there has ever been a more complete disconnect between what a company says it does – ‘connect’, ‘build communities’ – and the commercial reality. Note that the company’s knowledge about its users isn’t used merely to target ads but to shape the flow of news to them. Since there is so much content posted on the site, the algorithms used to filter and direct that content are the thing that determines what you see: people think their news feed is largely to do with their friends and interests, and it sort of is, with the crucial proviso that it is their friends and interests as mediated by the commercial interests of Facebook. Your eyes are directed towards the place where they are most valuable for Facebook.

I finally got round to reading this—I currently, and temporarily, have a lot of free time on my hands, so I’m reading everything—and it’s fantastic. Recommended reading for anyone interested in the nascent subject of web platforms (in fact this piece is reminiscent at times of John Herrman, who is currently the writer of the most interesting and relevant articles on the topic).

Designing headlines to make them more useful

Nieman Lab is running their annual predictions for journalism. Melody Kramer’s piece about designing headlines caught my eye:

In other words, how can we encode as much useful information as possible in a headline? Colors, fonts, shading, size, position, pictures, interactivity, history, metadata — basically all the design elements of information encoding across multiple dimensions. Which of those are most helpful to enhancing the headline? How can we test them?

For example, could we think of a headline as something that one can hover over, and immediately see source material? Or how many times the headline has changed? Or how other publications have written the same headline? (How does that help readers? How could that help publications?)

Let’s go broader. Why are headlines text? Could they be something else? What is the most important element at the top of a page? Is it five to fourteen words or is it something else entirely?

The whole piece is interesting and (typically for Mel) full of good ideas. Later she discusses the role of text:

Do we only think of mainly-text-based solutions because of the current nature of the platforms we share on? What if that changes? How could that change? A lot of current restrictions around headlines come from social and search restrictions and it would be interesting to think about that impact and how publications might bypass them with headline-like constructs (like Mic’s multimedia notifications or BuzzFeed’s emoji notifications.) They’re take the headline space and reworking it using images. What could we use besides images? In addition to images?

This is key. We use text because, well, text. It’s demanded by the channels we use to disseminate content. As readers we can react in non-textual ways: Facebook, Buzzfeed and others allow us to offer what might be very nuanced reactions using (barely?) representative icons and emoji. But as publishers, our platforms—both those that we own and third-party sites in our extended IA—generally haven’t evolved to a point where we can implement much of what Mel imagines.

This is a shame, as there’s plenty wrong with text and how it is used. Alan Jacobs wrote a short post in November, disagreeing with another post that championed text over other forms of communication:

Much of the damage done to truth and charity done in this past election was done with text. (It’s worth noting that Donald Trump rarely uses images in his tweets.) And of all the major social media, the platform with the lowest levels of abuse, cruelty, and misinformation is clearly Instagram.

No: it’s not the predominance of image over text that’s hurting us. It’s the use of platforms whose code architecture promotes novelty, instantaneous response, and the quick dissemination of lies.

This is problematic, and brings me back once again to Mike Caulfield’s excellent take on the layout and purpose of Facebook’s news distribution:

The way you get your stories is this:

  • You read a small card with a headline and a description of the story on it.
  • You are then prompted to rate the card, by liking it or sharing them or commenting on it.
  • This then is pushed out to your friends, who can in turn complete the same process.

This might be a decent scheme for a headline rating system. It’s pretty lousy for news though.

[…]

So we get this weird (and think about it a minute, because it is weird) model where you get the headline and a comment box and if you want to read the story you click it and it opens up in another tab, except you won’t click it, because Facebook has designed the interface to encourage you to skip going off-site altogether and just skip to the comments on the thing you haven’t read.

No conclusions this end, but plenty of interrelated issues to ponder:

  1. How do we (re-)engineer headlines to be more useful by revealing more information than is currently available in a few short words?
  2. How do we maintain the curiosity gap without ever-increasing reliance on clickbait?
  3. How do we continue the battle against fake news and propaganda masquerading as unbiased thought?
  4. How do we reconcile this with third-party distribution platforms that can only (barely) cope with text, and that treat content as a title and comments box only?

There’s something else in here about headlines and metadata and their role in content discovery and dissemination, and how users decide what to read and when. I was talking about this today with Richard Holden from The Economist and it’s sparked a few assorted thoughts that are yet to coalesce into anything new or meaningful. Perhaps in time.