Research Guides
Reconcile Article Authorship Before Polishing the Markup
An article's visible byline names two authors, but its structured data names only the publisher. A validator may accept the code while the page still gives conflicting accounts of who wrote the work. Authorship review needs evidence from the page, not just a syntax check.
This worksheet is for an authorized editorial or metadata review. It does not infer private identities, award expertise scores or describe a Colquhoun.shop author-verification feature.
1. Separate the roles before comparing names
Google's Article guidance distinguishes authors from publishers, recommends including all displayed authors in markup and representing multiple authors separately. It recommends an identifying author URL and using the appropriate Person or Organization type. Google: Article structured data and author best practices
Start a role list from the explicit information available. “Written by,” “Reviewed by” and a site's brand name do not automatically mean the same thing. If a role is unclear, record the ambiguity rather than promoting everyone mentioned on the page into an author.
2. Compare three surfaces
Inspect the visible byline, the article's structured data and the destination of any author-profile link. Keep the page address and observation date with the review. A copied template can make one surface lag behind the others.
Follow only relevant public or authorized profile information. Matching a common name to an unrelated account elsewhere is not sufficient to establish identity. The task is to reconcile the publisher's stated attribution, not to conduct a broad personal investigation.
3. Resolve a fictional template mismatch
Imagine a report visibly credited to two fictional contributors, Mara Chen and Elias Reed. Its metadata contains a single author name copied from a previous article, while both profile links lead to the same generic staff page.
The reviewer records three separate issues: the incorrect metadata name, the missing second author and the unclear profile destination. They ask the responsible editor to confirm the intended attribution before changing it. A convenient technical repair should not decide who deserves authorship.
After confirmation, the implementer updates only the affected attribution fields and links. The reviewer then checks another article using the same template to determine whether the defect is isolated or shared.
4. Preserve honest uncertainty
If a page is intentionally attributed to an organization, do not invent an individual author to make it look more personal. If the responsible editor cannot confirm a profile destination, leave that question unresolved instead of attaching a plausible social account.
Likewise, do not manufacture credentials or transform a reviewer into a specialist author. Accurate role descriptions are more useful than elaborate metadata that cannot be supported by the publication record.
5. Copy this attribution consistency card
- Article: public address and review date.
- Visible attribution: names and explicitly stated roles.
- Metadata: author entries, types and publisher entry.
- Profile route: linked destination and what it identifies.
- Mismatch: exact conflicting or missing information.
- Editorial decision: confirmed attribution and approving owner.
- Retest: corrected surfaces and related templates checked.
Keep technical observations separate from the editorial decision. A broken link can be verified directly; assigning authorship requires the appropriate publication record or responsible editor's confirmation.
6. Verify the public result
Reopen the published article and inspect its live metadata, not only the editing interface. Follow the corrected profile links and confirm they identify the intended person or organization. Check that unrelated publication dates, content and attribution were not changed.
The result is a consistent, supportable account of authorship. It does not prove the article's claims are correct, establish the author's expertise or guarantee a search appearance. Those are separate questions requiring their own evidence.