Preprints, Corrections and Retractions: Understanding the Publication Record

Treat publication as a record that can develop

Research communication does not always move directly from an unpublished draft to a permanent final article. A manuscript may appear as a preprint, undergo revision, be published by a journal and later receive a correction or other notice. Readers need to understand which version they are using.

Before relying on a finding, inspect the record for dates, version labels, links and notices. A downloaded PDF may no longer reflect the current status of the article. The publisher or repository page is often the best place to check the relationship between versions and any updates that affect interpretation.

Understand what a preprint establishes

A preprint makes a manuscript publicly available before acceptance through a journal's usual publication process. Server screening should not be confused with journal peer review. ASAPbio's resources explain preprints, versioning and the importance of understanding their status. ^[1]

A preprint can support timely discussion and feedback, but readers should assess methods and claims with appropriate care. Authors should label the work clearly and avoid presenting it as peer-reviewed evidence when it is not. Check the target journal's policies, co-author agreement, rights and any risks associated with early disclosure before posting.

Follow changes between versions

When an updated preprint or journal article appears, do not assume it is identical to the earlier version. Questions, analyses, figures and conclusions may have changed. Read the relevant differences before replacing a citation or relying on a previously quoted result.

Authors should maintain clear links where the platform permits and explain substantive revisions. A version history helps readers understand the development of the work. It also prevents several versions of one study from being mistakenly counted as independent evidence in a review or used to create the appearance of repeated confirmation.

Distinguish correction from retraction

A correction addresses an error or omission in the published record according to the journal's process. A retraction alerts readers that the publication should not be relied on in the way originally presented. The reason and scope matter; neither label should be interpreted without reading the notice.

COPE's retraction guidance addresses the correction of the scholarly record and the need for clear, linked notices. ^[2] Retraction can arise from different circumstances, including unreliable findings or serious process concerns. It is not appropriate to infer an individual's intent solely from the existence of a notice.

Illustrative example: a changed outcome estimate

Suppose an institution's briefing cites a preprint reporting a favourable result for a teaching intervention. The later journal version uses a corrected dataset and reaches a more cautious conclusion. The institution should update the briefing and explain the change if it materially affects the advice being given.

If the original article is later corrected, the author should inspect whether the correction concerns a minor label or the central analysis. If it is retracted, the briefing should not continue presenting the finding as established support. The action should follow the substance of the update rather than treating every publication notice as equivalent.

Respond responsibly when your own work contains an error

Document the issue and assess which outputs are affected. Discuss it with relevant co-authors and contact the journal or repository through its correction process. Preserve the original records needed to understand the mistake. Do not silently replace a published file or minimise a problem because the result has already been cited.

Consider related materials such as presentations, policy briefs, websites and shared datasets. They may also require correction. Where responsibility or reliability is disputed, follow the appropriate institutional process while addressing urgent concerns about the record. Accurate communication should remain the priority throughout the review.

Use a publication-status checklist

• The cited version and its review status are clear.
• Repository and publisher pages have been checked for updates.
• Related versions are not counted as separate studies.
• Corrections and retraction notices are read in full.
• Claims are revised when the underlying evidence changes.
• Relevant communications link to the current record.
• Conclusions about responsibility are based on evidence and fair process.

A changing publication record is not automatically a sign that scholarship has failed. Transparent correction can demonstrate that errors are being addressed. The professional task is to keep interpretations aligned with the best available version and to make important changes visible to those who relied on the earlier account.