Research Guides
Three Articles Can Still Be One Source
Finding the same statistic on three websites can feel like confirmation. But if all three repeat one announcement, you have found three accounts of one source, not three independent measurements. Count the evidence behind the claim before counting the pages that mention it.
This is an editorial research method, not an automated Colquhoun.shop feature or a claim about any named publisher.
1. Write down the exact claim being checked
Separate the number from its population, period, unit and qualification. “Adoption increased” is not the same claim as “survey respondents reported higher planned spending.” Small wording changes can make repeated coverage appear to support a conclusion the original evidence never made.
Save a short working note identifying the passage and source URL. Avoid copying whole articles into a shared record when a citation and a concise description are enough.
2. Follow the attribution upstream
The University of Cincinnati Libraries recommends tracing claims to their original source and recognizing when an article is reporting on another article. Its guide frames this as a way to assess the trustworthiness and context of the underlying information. University of Cincinnati: Trace Claims to the Original Source
Open the cited report, study or announcement when available. Record the route: article A cites article B, which cites report C. If two later stories both point to C, label them as sharing that evidence source. If no attribution is given, mark the origin unresolved rather than inventing a connection from similar wording alone.
3. Distinguish reporting from measurement
Imagine an invented vendor survey reporting that 40 of 100 respondents plan to change software. A trade article summarizes the survey, and two blogs repeat the trade article. This is still one survey. The three later pages may help explain or criticize it, but they do not turn the sample into 300 respondents or establish three separate findings.
Read what each page actually adds. An interview, a new dataset or a methodological critique can contribute different evidence. A rewritten headline alone does not. Keep those contributions separate instead of applying one blanket label such as “confirmed by multiple sources.”
4. Check the original on its own merits
Locating the origin is not the end of evaluation. Review the method, sample, date, definitions and disclosed interests relevant to your claim. A primary source can still be incomplete, mistaken or promotional. A careful secondary analysis can be valuable when it explains limits the original presentation obscures.
If the original is unavailable, be explicit about which secondary account you inspected. Do not present a quotation as checked against the original when you only saw it repeated elsewhere. Narrow or omit a consequential claim when its evidence cannot support the intended wording.
5. Keep a claim-to-origin ledger
- Claim: exact proposition, unit, population and period.
- Coverage: pages where the claim was found.
- Origin: underlying document or observation, if traced.
- Relationship: repeated account, separate analysis, independent evidence or unresolved.
- Contribution: what each source adds beyond repetition.
- Publication wording: supported conclusion and remaining qualification.
When preparing a comparison, pair this with the comparison-table evidence checklist. The aim is not to reject every repeated story. It is to avoid making a claim sound better supported merely because the same underlying evidence traveled through more websites.