How to evaluate information: seven questions

Published 2026-08-15.

A few dozen numbers have probably already passed before your eyes today — in the news, on social networks, in an ad, in a politician's speech. How many did you check? None? That's normal: nobody checks everything, there is no time for that.

But there is a difference between "I don't check because I trust" and "I don't check because I wouldn't know where to start". This text is about the second case. Seven questions you can ask of any claim. Usually a few seconds is enough — and the claim either firms up or falls apart on its own.

1. Who is saying this — and how do they know?

"Scientists have found", "experts say", "it is estimated that" — who, exactly? The primary source is whoever collected the data or made the decision: the statistics agency, the court, the institution, the study's authors. Everyone else is retelling. The more retellings between the fact and you, the more room for distortion. Nobody is even lying — it's just that everyone rounds a little as they retell.

An example from this site: when we wrote about fuel-station price rules, we cited not articles about the law but the law itself and the official Seimas voting record — with links, so you could open them yourself.

2. What does this number actually count?

It sounds pedantic, but this is where most wrong conclusions are born. "Immigrants from the United Kingdom" in our migration table means people who lived in the UK before arriving. A large share of them are returning Lithuanian citizens, not Britons. And "permits issued" counts documents, not people: one person can receive both a permit and its replacement in the same year.

So before drawing a conclusion, it is worth asking: what is the unit here? A person, a document, an event, a decision?

3. Compared with what?

"It doubled!" sounds alarming — until you see it went from two to four. Percentages without a reference point let the same fact sound like growth or decline: it all depends on which year was picked for the comparison.

And two correct numbers can disagree. On this site, the number of Ukrainian citizens holding permits is roughly 50 thousand higher than Eurostat's. Neither is wrong — Eurostat simply does not count temporary-protection holders among permits. Two sources, two definitions, two correct numbers.

4. Who benefits if you believe it?

This is not an invitation to see conspiracies everywhere. It is hygiene — like washing your hands. Every message has a sender, and the sender has an interest: to sell, to persuade, to mobilise, sometimes simply to get attention. An interest does not make a claim false. But it explains why, out of ten true facts, precisely these three were chosen.

5. What is missing here?

A percentage without the absolute number. An average without the distribution. One year without the history. The missing part is often exactly the one that would change the conclusion. That is why our absolute-poverty card shows both the percentage and the number of people: 7.0% sounds abstract. Roughly 202 thousand people does not.

6. Is the data current — and does it say when it was collected?

A number without a date could be from this year — or collected a decade ago; the text won't show it. In our tables the period always stands next to the number. Not out of politeness: without it, the number simply means nothing.

7. Can you check it yourself?

If a claim carries no link to a source you can open, it is not yet a fact — it is a story about a fact. Every table on this site links to its primary source. That works against us too: do not take even our word for it — check.

These questions do not tell you what to believe. They only slow you down — and wrong conclusions feel most at home precisely in a hurry. Healthy scepticism is not distrusting everyone. It is the habit of separating what you know from what you were told.