HowHint

How to Verify an Online Claim Using Primary Sources

A repeatable verification method for screenshots, statistics, study findings, and widely repeated claims online.

Original editorial guide illustration

Original editorial illustration; not a location photograph or product screenshot.

A search result that repeats your question is not evidence that its answer is correct. Online claims travel quickly from one post to another, and ten repetitions may trace back to the same misunderstanding. The solution is a compact verification workflow that identifies the original source and the precise statement it supports.

Step one: turn a headline into a testable proposition

A headline such as 'new study proves phones damage memory' mixes a source description, causal claim, population, and outcome. Rewrite it: 'In which participants, did which measured exposure cause which result, using what study design?' If the exact proposition cannot be stated, the claim is too vague to verify confidently.

Step two: trace the attribution backward

Open the linked article and follow its references. If it cites a press release, inspect the underlying study or report. If it cites an unnamed 'expert', search for the actual statement and its date. Prefer the responsible institution's original dataset, text of a law, official technical documentation, or peer-reviewed paper as appropriate. An official source can also contain error, so original does not mean infallible.

Step three: inspect scope and definitions

Check geography, time period, sample, age group, and unit. 'Doubled' might mean growth from one case to two; 'most' might refer only to survey respondents. For a screenshot, look for missing axes, cropped disclaimers, or a date that belongs to a previous version of the story. For policy statements, distinguish a proposal from an enacted requirement.

Step four: compare independent material

Find a second source that conducted its own analysis or offers a different type of evidence. Two outlets copying one press release do not count as independent corroboration. For scientific results, reviews and replication may matter more than a second article about the same paper. For fast-changing products, primary documentation and a current changelog are often more useful than evergreen lists.

Step five: record what remains uncertain

A useful note records the exact claim, source link, publication or observation date, strongest supporting evidence, strongest contradiction, and status: supported, unsupported, mixed, or not yet established. Uncertainty is an outcome, not a failure. Resist promoting preliminary findings into universal rules.

Worked example: a free software claim

Imagine a page claims that an editor is 'free forever, for everyone'. The company’s pricing page may show a no-cost tier but also mention limits on exports, commercial use, storage, or countries. A screenshot from last year might be accurate historically but false for today's offer. Your revised statement should preserve conditions: 'a free plan exists; features and eligibility should be checked before use'.

Publication checklist

Before sharing a claim, include the source URL, a defensible paraphrase, the date relevant to the claim, and any material limitation. Never fake testing, interviews, or credentials to make a story sound authoritative.

FAQ: Is Wikipedia unusable? No. It can be a starting map of references, but the precise claim should be checked against the cited primary or specialist source.

A claim-to-evidence log

Use this original planning worksheet with your own information. Its example decisions illustrate a method; they are not results from a product test or a visit.

CheckWhat to recordHow to use it
Exact claimPopulation, date, unit and wordingRewrite a vague headline as a checkable statement.
Original evidenceSource URL and supporting passageCheck whether the passage actually supports the claim.
OutcomeSupported, contradicted or unresolvedKeep unresolved details visible in the final explanation.

Download this planning worksheet (CSV)

Sources & verification

Source check: 2026-10-28. This is desk research and practical guidance; no firsthand product test or place visit is claimed. Provider terms can change.

  1. Library of Congress — primary-source analysis guides

Keep exploring

A useful next step from the published collection.