A claim appears on several websites, so it starts to feel established. But the pages may all repeat one announcement, study or witness account. Their agreement then shows repetition, not necessarily independent corroboration. Before counting sources, trace what each one contributes and where its evidence comes from. This does not mean secondary reporting is worthless. It can clarify, challenge or add context. The task is to distinguish those roles instead of treating every separate URL as another observation of the underlying event. A small evidence map often makes the relationship clearer than a long list of links.
| Record | Underlying evidence | Contribution |
|---|---|---|
| Original announcement | The organisation’s stated information | Primary statement of its position |
| Summary linking the announcement | The same announcement | Explanation or context; not another measurement |
| Separate relevant record | Its own documented source | Assess which particular detail it independently supports |
Define the claim and the unit of evidence
Write the specific proposition you want to support. A broad topic can contain several claims with different evidence trails. One article may independently confirm a date while repeating a figure from an announcement. Another may add an interview without checking that figure. Evaluate contribution at the claim level. Do not assign an entire page the label independent merely because some part of its reporting is original. The relationship between evidence and sentence is what matters.
Distinguish documents from the events, studies or observations they describe. Cochrane’s handbook makes this explicit within systematic-review methodology: multiple reports can concern one study and need to be linked rather than counted as separate studies. The general research lesson is about the unit being counted. It does not mean that an ordinary web article is a systematic review. It means that the number of documents and the number of underlying observations are different questions.
Trace the stated source behind each page
Look for links, attribution and descriptions of how information was obtained. Phrases such as according to a report or a named organisation said can identify a dependency. Follow the link and check what the underlying source actually supports. Do not assume that a page performed an independent investigation because it uses a different headline or introduction. Equally, do not assume copying solely from similar wording. Record the explicit relationship and preserve uncertainty where the source is not stated.
Keep a compact table with page, claim, underlying evidence and contribution. A row might show that a summary repeats the published figure but adds an explanation from another interview. That detail can be valuable without becoming a second independent measurement of the figure. The table should make this distinction visible. It also helps identify where you still need evidence instead of continuing to collect pages that lead back to the same origin.
Compare the underlying observations, not the domain names
Separate ownership and editorial presentation from evidential independence. Different domains can publish the same supplied material, while one publication can contain several genuinely distinct sources. The relevant question is how the particular claim was established. Did the reporter observe the event, obtain a separate record, interview another relevant person or merely repeat the original assertion? These possibilities need to be evaluated from the content and source trail, not inferred from the number of logos on your bookmark list.
Independence also does not guarantee correctness. Two accounts can be independently mistaken, and one well-documented primary record can be highly informative. Avoid turning the exercise into a simple score where more independent rows automatically settle the issue. Assess quality, relevance and agreement as separate dimensions. A clear evidence map helps you reason about them, but it does not replace judgement about whether the observations actually support the conclusion.
Keep secondary reports for the roles they genuinely serve
A secondary report may contain useful context, criticism or access to material not explained in the original source. Do not discard it just because it is not another independent observation of the main figure. Read what it adds. It may also identify a limitation you missed. The correct response is to label its contribution accurately, not to divide every source into wholly useful or wholly redundant. Several related reports can together clarify one underlying source without multiplying its evidential weight.
Watch for changes introduced during repetition. A cautious statement can become more definite in a summary, or a condition can disappear from a headline. Compare the wording with the original record before adopting the strongest version. Repetition may explain why a claim feels familiar, but familiarity is not an additional basis for certainty. If the later account makes a new claim, identify what supports that extension rather than assuming the original source covers it.
Resolve conflicts at the relevant layer
When two pages disagree, determine whether they refer to the same underlying source, different versions or genuinely different evidence. A revised report and an older summary can differ without representing two competing observations. A secondary writer may also have rounded or reinterpreted a value. Trace the discrepancy before choosing the more recent or more polished page automatically. The explanation should identify what changed and which source supports the statement you ultimately make.
If the underlying evidence conflicts, keep the disagreement visible and investigate proportionately. Do not average incompatible claims merely to produce a single answer. Consider definitions, timing and scope before deciding what the conflict means. A bounded conclusion can explain that several reports trace to one source while another relevant record differs. That is more informative than saying the internet is divided, which hides the structure and may exaggerate how many independent positions actually exist.
Write the source relationship into the final account
Use wording that describes the evidence accurately. You might say that several publications reported a named organisation’s announcement, rather than claiming several independent sources verified its contents. If a separate record corroborates a particular detail, identify that contribution. Keep the strength of the verb aligned with what was checked. Reported, stated, observed and verified are not interchangeable decorations; they communicate different relationships between your sentence and the available evidence.
Save the map with the research notes and update it if new material changes the relationships. A newly discovered source may add genuine independent evidence or reveal that two earlier records were related. Do not treat the first classification as permanent. The purpose is a transparent basis for the argument, not a certificate attached to a folder of links. Readers benefit when you can explain both how many documents you consulted and what those documents actually establish.
Count evidence at the level relevant to the claim, not by the number of pages carrying it. Trace dependencies, retain useful secondary contributions and distinguish repetition from independent observation. Then describe the relationship honestly in your prose. A shorter source list with a clear evidence map can support a stronger account than many links whose agreement comes from one unexamined origin.
