Tag Archive for: clickbait

,

The Source Test Card

You’ve read a headline, felt sure of it, and found out later there was nothing behind it. Nobody lied to you, exactly. You just took the claim at face value because checking felt like too much work.

This card makes checking quick. It starts with one move that catches most bait before you click: say in one plain sentence what you’re being told, what happened and who did it. If you can’t, you haven’t been told anything yet.

Then three questions we use on this channel before trusting any claim. Stake: who gains if you believe it? Source: how do they know? Steelman: what’s the strongest honest case on the other side? Each one comes with what to look for and the warning signs that mean keep checking.

The full walk-through, with the research behind it, is in the video.

/by

Tag Archive for: clickbait

Why Headlines Lie Without Lying: 8 Tricks Explained

You know the feeling. You tap a headline, read three paragraphs, and realize the article never said anything the headline didn’t already hint at. Nothing in it was false, exactly. You just got moved from one page to another.

I spent 30 years in marketing writing lines whose whole job was to get the click, so I know how these are built. Below are the eight moves that show up most, with a quick check for each one. At the end there’s a single habit that catches all eight.

1. The missing noun

“Doctors want you to stop eating this common food.” Which food? The headline won’t say. There’s a hole in the sentence exactly where the important word should be, and the only way to fill it is to click.

In 2016, Facebook built a filter to catch headlines like this and published real examples of what it was targeting. One of them, word for word: “He Put Garlic In His Shoes Before Going To Bed And What Happens Next Is Hard To Believe.” That’s the missing noun in its purest form.

The check: if you can’t say what the story is about from the headline alone, the headline was built for the click.

2. The villain frame

Somebody “slams” something. Somebody “destroys” somebody. Here’s an invented example so I’m not picking on anyone real: “Celebrity chef slams the way you cook your eggs.” The actual story is a chef mentioning in an interview that he likes his eggs a little softer. The verb turned a preference into a fight, because a fight gets more clicks than a preference.

The check: swap the verb for “said.” If the story falls apart when you change one word, that word was doing all the work.

3. The question mark

“Is your morning coffee slowly killing you?” A question lets a writer put an idea in your head without having to stand behind it, and it works better than you’d think.

In a 1981 experiment, researchers showed people headlines about made-up candidates for public office. Some headlines simply asked whether a candidate was linked to wrongdoing, without claiming he was. People who saw only the question still rated that candidate worse than people who saw a neutral headline about him. A preregistered repeat of the study in 2024, with 506 people, found the same thing.

There’s an old newsroom joke that any headline ending in a question mark can be answered “no.” When people have actually counted, it isn’t that tidy, so here is the honest version.

The check: read the question mark as “we couldn’t prove yes.” If they could have, it would have been a statement.

4. The number with no base

“Cases up 200%.” That sounds like a wave. It could also mean two cases last year and six this year.

Here’s a real case of what that costs. In 1995, British regulators warned that some newer birth control pills doubled the risk of a dangerous blood clot. “Doubled” was accurate. The actual risk went from about one woman in 7,000 to about two in 7,000. Many women stopped taking the pill after the warning, and researchers later estimated roughly 13,000 additional abortions in England and Wales the following year.

The check: ask “up from what?” A percentage without its starting number hasn’t told you anything yet.

5. Nobody did it

“Mistakes were made.” “Customer data was exposed.” Something happened in both sentences, and no one did it.

That missing name changes how you judge what happened. In experiments at Stanford, people read about the same famous Super Bowl wardrobe malfunction. Some read that a performer ripped the costume. Others read that the costume ripped. The people who read the version with a person in it blamed him more, and they wanted him fined 30 to 50 percent more.

The check: ask who did it. If neither the headline nor the story will say, that silence is information.

6. “Experts say”

Which experts, and what were they asked? My favorite example is an ad. In 2007, Colgate ran posters in Britain claiming that more than 80 percent of dentists recommended Colgate. Britain’s advertising regulator looked at the survey behind the claim and found the dentists had been allowed to name several brands, so other brands could have scored just as high. The regulator ruled the poster misleading.

The check: look for names and the actual question asked. No names usually means you’re reading someone’s opinion in a lab coat.

7. The quote that isn’t

A headline says someone “blasted” or “warned,” then gives you two or three words in quotation marks. Those words are real, but the sentence they came from might say something different.

Movie ads have done this for decades. In 1997, a critic panned a gangster film called Hoodlum but described its lead actor as “fierce, magnetic, irresistible.” The ad ran one word: “Irresistible.” The word was real, and the meaning had flipped.

The check: find the full sentence. If the article won’t give it to you, that tells you why.

8. The local story that goes national

One thing happens in one town, and it gets written up so it feels like it’s happening everywhere. Every October, for example, parents are warned to check Halloween candy for poison from strangers.

The sociologist Joel Best has tracked those reports back to 1958 and hasn’t found a single confirmed case of a child killed or seriously hurt by a stranger’s poisoned treat. The well-known 1974 death in Texas was a father who poisoned his own son. The real danger on Halloween night is cars. A CDC analysis found that children were about four times more likely to die as pedestrians on Halloween evening than on other evenings.

The check: ask how many times this has actually happened. One story is one story.

One habit that catches all eight

Every trick above works by leaving something out. So the first move is to fill in the blank yourself: say in one plain sentence what actually happened and who did it. If you can’t, you haven’t been told yet.

Then run the three questions we use on this channel before trusting any claim.

Stake: who gains if you click? The site gets paid whether you learn anything or not.

Source: how do they know? Look for names and full sentences.

Steelman: what’s the honest reason this headline might be fair? Sometimes there is one. More often, the most boring true version of the story is the real one.

Sources

  • Peysakhovich & Hendrix, Facebook Newsroom (Aug. 4, 2016). Further Reducing Clickbait in Feed.
  • Wegner, Wenzlaff, Kerker & Beattie (1981). Incrimination through innuendo. Journal of Personality and Social Psychology.
  • Social Psychology (2024). Preregistered replication, N = 506. DOI 10.1027/1864-9335/a000540.
  • Gigerenzer et al. (2007). Helping Doctors and Patients Make Sense of Health Statistics. Psychological Science in the Public Interest.
  • Furedi, A. (1999). The public health implications of the 1995 “pill scare.” Human Reproduction Update.
  • Fausey & Boroditsky (2010). Subtle linguistic cues influence perceived blame and financial liability. Psychonomic Bulletin & Review.
  • Advertising Standards Authority, UK (2007). Ruling on Colgate-Palmolive.
  • Slate (Nov. 25, 2009). How Do Movie Blurbs Work?
  • Best & Horiuchi (1985). The Razor Blade in the Apple. Social Problems; updated at joelbest.net.
  • Centers for Disease Control and Prevention (1997). Childhood Pedestrian Deaths During Halloween, United States, 1975–1996. Morbidity and Mortality Weekly Report 46(42).
/by

Why You Clicked That Headline Even Though You Knew Better

You know the one. You’re scrolling and a headline says, “He Put Garlic In His Shoes Before Going To Bed And What Happens Next Is Hard To Believe.” You know it’s going to be nothing. You click anyway. It’s nothing. You close the tab feeling a little dumb, and forty seconds later you do it again.

For 30 years I wrote headlines and email subject lines for a living, and the job was scored on one number: how many people clicked. I’ve built lines like the garlic one on purpose and tested them against plainer versions. So here’s how they’re put together, and a quick check that tells you whether a headline is informing you or holding something back.

Why it works: the curiosity gap

The trick is older than the internet. In 1994, a Carnegie Mellon psychologist named George Loewenstein reviewed decades of research on curiosity and landed on a simple explanation. Curiosity shows up when you notice a gap between what you know and what you want to know. The important part is how that gap feels. It’s less like interest and more like an itch, a small sense of something missing that you want to close.

Loewenstein called it an information gap. I call it the curiosity gap.

Look at the garlic headline again. It tells you almost everything: a man, garlic, shoes, bedtime. The only thing missing is what happened. That missing piece is the itch, and clicking is how you scratch it.

The three moves

Once you know these, you’ll see them every time you read a bait headline.

Move 1: Hide the who. The headline won’t tell you who the story is about. You get “he,” “this celebrity,” or “a common household item.” You get most of the story, minus the one word that would tell you whether you care.

Move 2: Point at the answer. The headline doesn’t say what happened. It tells you the answer is one click away: “what happens next,” “here’s why,” “the reason will surprise you.” In 2015, two Danish researchers, Jonas Blom and Kenneth Reinecke Hansen, studied 100,000 headlines and gave this move a name: forward reference. The headline points ahead at the article instead of telling you what’s in it. They found it most on sites that depend on clicks to make money.

Move 3: Promise a feeling. “Hard to believe.” “Priceless.” “Left everyone speechless.” The headline tells you how to feel about something it won’t show you.

The garlic headline uses all three. Who is “he”? That’s move one. “What happens next” is move two. “Hard to believe” is move three.

And here’s the twist. Facebook itself published that exact headline in 2016, as an example of what it was trying to push down in people’s feeds. A company that makes money on attention went after the most reliable click on the internet. Its stated reason was simple: people had told Facebook they wanted to see fewer stories with clickbait headlines.

Running the Source Test on a headline

Every bait headline makes you a promise: there’s something here worth your click. On this channel we run the Source Test on promises like that.

Stake: who gains if you believe it? The publisher. They get paid when you click, not when you’re glad you clicked. The article can be nothing and they still got paid, so there’s no reason for it to be good.

Source: how do we know this is done on purpose and not just lazy writing? Two pieces of proof. First, when Facebook built its 2016 filter, a team read thousands of headlines by hand and sorted out the clickbait using two questions: does the headline hold back what the story is about, and does it oversell it? Then Facebook built a filter to catch those phrases, the way your email catches spam. Nobody builds a spam filter for an accident.

Second, Upworthy, the site that made this style famous, tested its headlines like a lab. Researchers later published the archive: more than 32,000 experiments and about 150,000 headline versions shown to readers hundreds of millions of times. The winners became the headlines you remember from 2013.

Steelman: what’s the best case for the other side? Opening a gap isn’t a crime. Every mystery novel does it. Every good teacher does it. I did it at the start of the video. A headline that told you everything would be a summary, and nobody clicks a summary. So the gap by itself isn’t the trick.

The fill-in-the-blank check

Here’s the whole trick in two sentences. A gap the story fills is a good story. A gap the story never fills is bait.

You can tell which one you’ve got before you click. Ask one question: can I say what this story is about, just from the headline?

“When she looked under her couch cushions and saw this, I was shocked.” Saw what? You can’t say. Bait.

“The dog barked at the delivery man and his reaction was priceless.” What reaction? You can’t say. Bait.

The garlic one? You still can’t say what happened. Only the writer knows, and the writer left it out on purpose.

The honest part

You’d think this trick would cost a news site your trust. I looked for proof of that, and it isn’t clean. In 2016, researchers at the University of Texas showed headlines to about 2,000 American adults. Some were plain, some were questions, and some pointed at the answer like the ones above. The question headlines got slightly worse reactions. The ones that pointed at the answer didn’t: people rated them about the same as the plain ones.

So using this trick doesn’t reliably hurt the site, and that’s a big part of why it’s everywhere. The site doesn’t pay for it. You do, forty seconds at a time.

I’ll own my part. I open a gap at the start of every video I make, on purpose. The difference is that I tell you what you’ll get before I ask for anything, and then I give it to you. The gap gets filled. That’s the line between a story and bait, and I’d hold everyone to it, including me.

Sources

  • Loewenstein, G. (1994). The Psychology of Curiosity: A Review and Reinterpretation. Psychological Bulletin 116(1), 75–98.
  • Blom, J. N., & Hansen, K. R. (2015). Click bait: Forward-reference as lure in online news headlines. Journal of Pragmatics 76, 87–100.
  • Peysakhovich, A., & Hendrix, K. Facebook Newsroom (Aug. 4, 2016). Further Reducing Clickbait in Feed.
  • Matias, J. N., Munger, K., Aubin Le Quere, M., & Ebersole, C. (2021). The Upworthy Research Archive. Scientific Data 8, 195.
  • Scacco, J. M., & Muddiman, A. (2016). Investigating the Influence of “Clickbait” News Headlines. Center for Media Engagement, University of Texas at Austin.
/by