Before you share something: a five-minute check
This is the same sequence used to build every case on this site, compressed to what you can do on your phone before you hit share.
Find the name behind "researchers say"
Search the specific claim plus a word like "study" or "study name." If you can't find a named author, institution, or publication within a minute of searching, treat the claim as unverified — not false, unverified — and don't pass it along as settled fact.
Read the summary, not just the headline
Most journal publishers put a plain-language abstract at the top of the actual paper, free to read even when the full paper isn't. That paragraph alone catches a large share of misrepresented studies — check whether it says what the post claims it says.
Ask what the study measured, specifically
"Linked to," "associated with," and "caused" are different claims. A study that found a correlation is not the same evidence as one that found a cause — and a lot of misleading content depends on you not noticing the swap.
Notice your own reaction
Content built for speed over accuracy is usually built to trigger surprise or anger first, because those emotions travel fastest online. If you feel a strong urge to share something within seconds of seeing it, that urge is itself a signal to slow down, not a reason to hurry.
Check who benefits from you believing it
Not as an accusation — as a routine question. Content that confirms your existing view spreads faster than content that challenges it, on every side of every issue. That's a reason to check it more carefully, not less.
Rebuilding the attention that makes step one possible
The checklist above only works if you can actually hold a claim in mind long enough to check it. That capacity is trainable — plasticity in the brain's attention and memory systems runs in both directions, and the same mechanism that lets it shrink under neglect lets it rebuild under practice.
Finish the interruption before switching back
Fully returning to a task after a single interruption takes over twenty minutes on average. Stacking small checks between tasks costs more focus than it feels like in the moment — closing one loop before opening the next is the actual fix.
One long-form read a day, notifications off
Even twenty to thirty minutes of sustained, sequential reading rebuilds the capacity to hold a claim and its context in mind at the same time — the exact skill a five-minute verification check depends on.
Try to recall before you look it up
People reliably stop encoding facts they expect a search engine to hold for them. Pausing to recall first — even if you're wrong, even if you then look it up anyway — is the part that keeps the memory system in practice.
Let unstructured time stay unstructured
A wait, a commute, a quiet stretch with nothing in your hand. Reaching for a phone the instant nothing is happening removes the one condition most reliably linked to unprompted, undirected thought.
None of this is about willpower as a character trait, and struggling with focus or memory day to day doesn't mean something is wrong with you — attention is shaped by environment for everyone, and the environment most people are in now is genuinely built to fragment it. These are practice habits, not a verdict on anyone's intelligence.
If you want to go further
- Read the method page — the same six questions used to build this site, written out in full.
- Walk through a real example on Case 001, where a real study gets stretched past what it found.
- Bookmark the standing reference studies so the next viral claim citing one of them is a thirty-second check instead of a search from scratch.