Stop Before You Share · Method · DoingResearch.net

Purpose

The first move is the shortest: stop. Before you believe a claim, react to it or pass it on, take a few seconds to ask what it is and where it came from. Most of the damage a false claim does happens in the moment between seeing it and sharing it, and that moment is the one you control.

Stopping isn’t a test of intelligence. In a 2021 study in Nature, Gordon Pennycook, David Rand and colleagues found that whether a headline was true had a large effect on whether people judged it accurate, but little effect on whether they said they’d share it. Most people said they cared about sharing only accurate news. The researchers’ explanation: when scrolling, attention drifts to other things (is this funny, surprising, on my side?), and accuracy never gets checked. In a field test on Twitter, simply asking users to rate the accuracy of one unrelated headline improved the quality of the news they shared afterward.

Stopping also tells you how hard to work. Mike Caulfield, who wrote the SIFT method these four moves come from, treats a quick check and a deep investigation as equally valid. The point of the pause is to pick the right one.

When you need it

A post makes you angry, scared, smug or delighted, and you want to share it right now.
The claim is big news (a death, an attack, a scandal) and you’ve only seen it in one place.
You don’t recognize the account, site or outlet it came from.
It confirms exactly what you already thought about a person or group.
You’re about to repeat it in a conversation, a class or a piece of writing.
Halfway through checking, you’ve opened twenty tabs and forgotten what you were trying to find out.

How to do it

1Notice your reaction

Name the feeling: outrage, fear, glee, vindication. Strong emotion is the cue to check, not a reason to skip it. Caulfield’s advice to student fact-checkers is to make that feeling a trigger: when something you want to share gives you a jolt, spend thirty seconds on it first. There’s evidence behind this. In experiments by Cameron Martel, Pennycook and Rand (2020), people asked to rely on their feelings rated fake headlines as more accurate than people asked to rely on reason, while their ratings of real headlines didn’t change.

2Ask what you know about the source

Do you know who made this? Not who shared it, but who first published it. Is it a news outlet, a company, an advocacy group, a stranger, a parody account? If you can’t answer, that’s your first job, and the next move, Investigate the Source, covers it.

3Ask what you’d need to know

Write the claim in one sentence, then ask what would have to be true for it to be right. A named study? A date and place? An official record? Would other outlets be reporting it? This turns a vague feeling (“seems off”) into specific questions you can answer.

4Decide how much checking it deserves

Match the effort to the stakes. A funny photo you’ll scroll past needs nothing. A claim you’ll share needs a quick check. A claim you’ll rely on, teach or publish needs the full trail. If it isn’t worth even thirty seconds, it isn’t worth sharing either.

5Stop again when you get lost

Caulfield’s second meaning of “stop”: partway through checking, when you’ve drifted down a rabbit hole, pause and remember what you set out to learn. Then go back to the question you wrote in step 3.

Try it

Your task

On April 23, 2013, the real Associated Press account on Twitter, @AP, posted that there had been two explosions at the White House and that President Obama was injured. Before looking anything up, run the stop questions on it: what’s your reaction, what do you know about the source, what would you need to know, and how much checking does it deserve? Then find out what actually happened, and how long it took the stock market to notice and recover.

Show a hint

Search for the date with “AP Twitter hacked”. Look for coverage from more than one outlet.

Show the answer and the path

The account was hacked. The tweet was false: the Associated Press said its account had been compromised and suspended it, and White House press secretary Jay Carney told reporters the president was fine. A group calling itself the Syrian Electronic Army claimed responsibility. According to NPR, the Dow Jones industrial average dropped about 130 points (about 1 percent) just after 1 p.m. Eastern and was back in positive territory by 1:45 p.m.

What the stop questions show:

The source was real and normally trustworthy. That’s why the tweet worked. “Who said it?” can’t be the only question.
What you’d need to know: an attack on the White House would be on every outlet within minutes. One tweet, and nothing else, is the tell.
How much checking: a claim this big deserves the minute it takes to see whether anyone else is reporting it. The market’s dip and recovery took less than an hour; anyone who waited for a second source saw the correction arrive.

Common mistakes

Skipping the pause because the source looks familiar. A trusted name can be hacked, impersonated or quoted out of context.
Only stopping for claims you dislike. The posts that fit what you already believe are the ones you’re least likely to question, and the ones most worth a second look.
Treating “I’m not sure” as permission to share with a disclaimer. “Not sure if this is true, but…” still spreads it.
Doing all-or-nothing checking. Thinking that checking means an hour of research leads people to skip it entirely. Thirty seconds is often enough to catch the worst.
Mistaking a gut sense of “something’s off” for a finding. The pause tells you what to check. It doesn’t tell you the answer.

Limits

Stopping finds nothing on its own. It only buys time for the other three moves. A pause followed by a glance at the same post doesn’t make it any truer.

The research on accuracy prompts shows real but modest effects, measured mostly in online experiments and on one platform at one point in time. A pause helps most against careless sharing. It does less against a well-made false claim from a source that looks reliable, or a claim you have strong reasons to want to believe.

Your feelings can also mislead in the other direction. A claim that seems dull or reasonable can still be false, and an outrageous one can be true. Emotion is a cue to look harder, not a verdict. And deciding what deserves checking is itself a judgment that your biases can bend: see Confirmation Bias on ThinkingCritically.net.