Most people who spread false information don't do it deliberately. They see something plausible that fits their view of the world and pass it on. That's normal behaviour, and it's the main mechanism by which misinformation propagates.
A few checks help surprisingly well. They take very little time and can become habit.
Who's actually saying this
The first and most important step: with something arriving via social media, the original source is usually unclear.
A screenshot of a headline is not a source. A post citing a study without linking to it is not a source. An account with a serious-sounding name is not a newsroom.
The practical step is to look for the origin. If a story is true and significant, more than one established outlet generally covers it. If a single channel has something of enormous importance, caution is warranted.
When was this
One of the most common forms of misinformation isn't false at all — it's old.
A real photograph of a real event gets shared in a new context. An article from five years ago circulates as though current. A statement from a different situation resurfaces.
Checking the date is the simplest and most effective single step. On many platforms the publication date isn't visible when a link is shared — you have to actually open the page.
The context of the image
Images are particularly vulnerable, because they persuade immediately and carry no provenance.
A reverse image search takes seconds and usually shows where a picture has appeared before. For photographs from conflict zones or disasters this is almost always instructive — a great many come from other years or other countries.
With generated images it's become harder, because the obvious tells are disappearing. Which makes the question of source more important, not less: who took this, when, and is there independent evidence the event occurred?
Headline against text
A systematic effect with little to do with intent. Headlines get sharpened because they compete for attention. The text is often more careful.
There's research on how many shared links were actually opened first, and the figures are low. A substantial share of propagation happens on the basis of the headline alone.
Anyone who reads the article regularly finds it says something different. Particularly common with science topics, where "a possible association in a small study" quickly becomes "researchers prove".
For studies: three questions
When a story cites research, three questions help enormously.
How many participants? A study with twenty people can be interesting and cannot support general conclusions.
Humans, animals, or cell culture? A great many spectacular health stories rest on experiments in mice or on cells in a dish. The distance from there to a statement about people is long and frequently isn't travelled.
Association or cause? Observational studies show two things occurring together. That's entirely different from demonstrating one causes the other. Phrases like "linked to" signal exactly this distinction and routinely become "causes" in retelling.
The feeling test
The most useful indicator of all, and it requires no research.
If a story provokes a strong emotional reaction — outrage, fear, validation — that's the moment to pause. Content that spreads fast is disproportionately content that produces strong emotion. That isn't accidental; it follows from how the platforms work.
Your own agreement is the most dangerous signal. We scrutinise information contradicting our worldview far more carefully than information confirming it. That's well established and affects every political direction equally.
What I've made a habit of
One more habit that turns out to matter, and it concerns corrections rather than checking.
When a story turns out to be wrong, the correction almost never travels as far as the original. This is well documented and the asymmetry is large. A widely shared claim reaches an audience; the correction reaches a fraction of it.
Which means the practical value of checking before sharing is far higher than the value of correcting afterwards. Thirty seconds spent before pressing the button does more than an hour spent afterwards trying to unwind it.
It also means being sceptical of your own memory. Most of us have absorbed claims that were later retracted and have no idea which ones, because we saw the original and never encountered the correction. That is not a failure of attention; it is a structural property of how information circulates. The only defence is a general habit of holding confidently remembered facts a little more loosely than feels comfortable.
Not forwarding anything I haven't opened and at least skimmed. Sounds obvious and evidently isn't.
With anything that makes me angry, waiting a night. If it's still true tomorrow, I can share it then.
And if I have spread something false, correcting it in the same channel. That's uncomfortable and it's the only way a correction reaches the people who saw the error.