TruthLens tools
Ten short scenarios built from the patterns that actually spread — misleading headlines, recycled images, missing citations, cherry-picked statistics. Every answer is explained, so a wrong guess still teaches you something.
Bring the thing you actually saw. The ten scenarios below teach you the patterns to look for in it — this tool does not analyse your paste automatically.
Nothing you type leaves this browser tab. 2000 character limit.
HEADLINE: “SHOCKING: City confirms what everyone suspected about its water supply” The article body contains a routine statement that the annual water quality report was published, with a link to the report.
What makes this headline misleading rather than the story being false?
A photograph of a flooded street appears with the caption: “Yesterday's flooding in the coastal town.” The photo is genuine and from three years earlier.
What should you check before treating this as breaking news?
A post reads: “Everyone in this thread is saying it, so it must be true.” The account has no profile information and the claim is repeated by other accounts using identical wording.
What does the repetition across accounts tell you?
“A new study proves that the recommended sleep is unnecessary for adults.” No study is named, no journal, no link — just the word “proves”.
What is the strongest reason to be sceptical here?
“Crime is up 400%.” The figure compares this year's crimes in one category against a near-zero baseline from a category that was abolished three years ago.
What is the problem with this statistic?
A graphic attributes a sentence to a well-known figure: “We never considered that outcome.” The quote is presented with the figure's name and title.
What is the fastest way to check this quote?
An industry-funded report concludes the industry's own product is safe. The funding is disclosed in small print at the bottom of the page.
Does the disclosure make the conclusion reliable?
Two outlets report different figures for the same survey. One cites the published paper; the other cites “sources familiar with the matter”.
Which is the stronger evidence, and why?
“New research” reveals a statistic first published nine years ago by a different team, on a smaller sample, in a different population.
What has gone wrong, and does it make the statistic false?
A chart is shared showing a sharp trend. It has no title, no source line, no units on the vertical axis, and no date range.
What is the most important thing missing here?
This scores pattern recognition, not truth. All scenarios are written for this test — no fabricated statement is attributed to any real person. A good score means you will recognise these structures more often; it does not mean you can identify whether a specific real claim is true.
It measures whether you recognise the shapes misinformation takes: the gap between headline and body, a real image with a wrong date, a percentage with a rigged baseline, a quote nobody can find. Those shapes are learnable, and they are much more recognisable once you have seen them side by side.
It does not test whether you would catch a genuinely new technique, and it does not verify anything real. A person can score well here and still share something false — the most effective material is usually the kind built from true facts arranged misleadingly, which no quiz prepares you for automatically.
There is no time limit and answers are explained as you go, so the result measures your reasoning rather than your recall. Most people score higher on a second attempt; that gap between attempts is the real measure of how much you learned.
No. Every scenario was written for this test. No fabricated statement is attributed to any real person, outlet or organisation, which is the same standard this tool asks you to apply to things you see online.
Seven or more out of ten means the patterns feel familiar. Below six is normal the first time — the explanations are written to be read. A low score is not a sign that you are gullible; these patterns are designed to work on attentive readers.
No. This tests familiarity with known patterns. The most effective misinformation is usually built entirely from true facts arranged misleadingly, which is not something a quiz can inoculate you against. Keep using the SIFT method on individual claims.
The scenarios describe what you would see and ask you what you would check. Testing whether people can describe the verification step is more useful than testing whether they recognise a specific file, which they would simply learn to recognise instead.
Yes. The explanations and the habits list are written to be discussed, and the print layout is clean enough to hand out. The scenarios avoid real names and events so there is nothing to misrepresent.
The fallacy quiz covers reasoning errors — the shape of an argument. This one covers publishing and sharing patterns: headlines, statistics, images, quotes, citations. The two overlap less than you might expect, and both are worth doing.