Sharing a story takes two seconds. Reading it with any care takes two minutes. That gap — between the reflex and the reading — is where most bad information travels. Not through elaborate forgeries, usually, but through real reporting stripped of context, dressed in a headline and dropped into a group chat.

There is no shortage of people telling you which outlets to trust. Trusting a logo is a shortcut, and it fails in both directions: good papers have bad days, and outlets you dislike occasionally get it right. A routine works better — twelve questions, applied to everything, especially the stories you are delighted to see.

Sourcing: who actually told them this?

Naming matters more than most readers realise. Anonymity has legitimate uses: a whistleblower who would lose their job, a witness to a crime, a source in a country where speaking openly is dangerous. But a story resting entirely on unnamed sources is a claim to hold at arm's length. Ask how many there are, whether they corroborate each other or simply repeat the same briefing, and whether the outlet explains why they cannot be named.

If a report is cited, find the report. Most are published, and the summary in paragraph two is often narrower than the finding. "A study suggests" can cover anything from a large trial to a small online poll, and the difference matters if you are about to repeat it as fact.

Ask too whether you are reading journalism or a rewrite. A great deal of local, trade and business news begins life as a press release. That is not automatically dishonest, but the framing is someone's communications strategy, complete with the quotes they wanted you to see. Court reporting works the same way: the ruling is usually available, and it frequently says something more careful than the coverage does.

Framing: what are the words doing?

Two headlines can describe one event and leave you with opposite feelings. The same council vote becomes "Council backs £40m regeneration plan" or "Council spends £40m while libraries close". Both sentences can be accurate. Only one of them is trying to make you angry.

Watch for passive constructions that remove the actor: "three people were killed" tells you less than "police shot three people" where the second is accurate. Watch for nouns carrying a verdict — clashes, chaos, burden, crackdown, u-turn. None is neutral, and none is necessarily wrong, but you should notice it.

Check who gets to speak. A story can quote six critics and describe supporters only in the third person, or the reverse. One piece proves little; a pattern across several is worth naming. Question headlines deserve suspicion too. "Is Britain broken?" asks nothing and plants everything, as does "concerns have been raised" or "critics claim" — vague attribution is how an opinion gets printed without anyone owning it.

Context: what has been left out?

The most common problem with a shared story is not that it is false. It is that it is incomplete.

• Percentages of what? "One in five" means little without the base, and small samples swing wildly.
• Compared with when? A "record high" is only meaningful against a stated time frame.
• Who is missing? Stories about a policy often quote the people running it and the people opposing it, and skip the people living under it.

Ask whether the same outlet covered the same subject a month ago with a different emphasis. Papers are entitled to report new evidence that contradicts an earlier line, but they should say so. When yesterday's certainty becomes today's silence, remember it.

Slant: whose interests does this angle serve?

Every outlet has a shape. Some are owned by a proprietor with commercial interests, some by a trust, some funded by members, some by a licence fee. None of that automatically corrupts the reporting, but all of it is information worth holding while you read.

Think tanks and campaign groups are often described as though they were simply researchers. Many do real research; most also have a position. Look at who funds them and what they exist to argue for — that applies to organisations whose conclusions you welcome as much as to those you don't.

Be sceptical of balance theatre: two loud opposites, no attempt to establish what is actually known. And notice when one side's claims are reported as facts while the other side's are reported as claims.

Numbers and pictures: the two easiest things to get wrong

Statistics are the simplest part of journalism to check and the easiest to mangle. A few traps:

• Percentages versus percentage points. A rise from 4 per cent to 6 per cent is a two-point rise and a fifty per cent increase. Both are true; only one usually makes the headline.
• Cash versus real terms, and totals versus per-person figures. Spending, crime and council budgets all change shape depending on which you use.
• Averages versus medians. An average wage is dragged upwards by a handful of very large incomes.
• Charts with cut axes, polls with leading questions, and any sentence containing "up to", "could" or "may".

Pictures deserve the same treatment. Library images, archive footage and stock photographs appear under current headlines constantly, and a tight crop can tell the opposite story to the wide shot. Look for a date and a caption credit.

One caution that sits outside media literacy: if a story touches on your health, your money or a legal problem, treat it as a prompt to ask a qualified professional rather than as advice. Reporting summarises; it cannot know your circumstances.

The twelve questions, and how to use them

1. Who is the source, and are they named?
2. How many independent sources back the central claim?
3. Can I see the original document, data or ruling?
4. Is this a press release or briefing with a rewrite on top?
5. What does the headline claim, and does the text support it?
6. Which words are doing the emotional work?
7. Who is quoted, and who is only described?
8. Percentages of what, and measured against when?
9. Cash or real terms, total or per person, average or median?
10. What did this outlet say about the same subject last year?
11. Who owns the outlet, and who funds the think tank it cites?
12. If I agreed with this story, would I still be asking these questions?

Run them in order and you will usually stop at two or three. Most weak stories fail early: no named source, no link to the underlying document, a headline the text does not support.

Two habits make the list stick. Apply it first to something you agree with, where your guard is lowest and the exercise teaches you most. And when you cannot verify something, say so — "I haven't checked this" is a perfectly respectable caption for a link, and better than the alternative.

Then check the date. Old stories come round again every few months, usually with the original year quietly missing. None of this will make you right every time. It will make you slower to share, which is most of the battle.

Photo: Kübra Şeref / Pexels