Retraction & Correction Policy

The principle

An article that has been published has been read, cited and relied on. Correcting it is therefore not a matter of editing the page: it means publishing a visible notice that tells anyone who encounters the article, now or in ten years, what changed and why.

Three commitments follow, and the journal applies them regardless of who the authors are or how long ago the article appeared.

• Nothing is changed silently. Every substantive change after publication is accompanied by a citable, permanently linked notice.
• Nothing is deleted. A retracted article remains available, marked as retracted. Removing it would break the citations of everyone who relied on it and hide the fact that it existed.
• The reason is stated. Notices say what was wrong, not merely that something was.

The four instruments

The journal uses four instruments, following COPE’s retraction guidelines. Which applies depends on how far the problem goes.

Corrections

A correction is published where the article’s conclusions remain sound but something in it is wrong in a way that affects how a reader understands or uses it. Typical grounds:

• An error in data, a table, a figure or a calculation;
• A mistake in the method description that would mislead someone repeating the study;
• An omitted or incorrect declaration — funding, competing interests, ethics approval;
• A change to the author list agreed under the authorship policy;
• Missing attribution for a passage or figure, where the problem is confined to specific items;
• A production error introduced by the journal — in which case the notice says so, so that no fault attaches to the authors.

Trivial errors that cannot mislead anyone — a misspelled word, a formatting fault — are fixed without a notice, and the article records the date it was updated. Anything touching data, meaning, attribution or a declaration gets a notice.

Expressions of concern

An expression of concern is published while a serious question is being investigated, so readers are not misled in the meantime. It is used where:

• There is credible evidence of a problem, but the investigation is not complete;
• An institutional investigation is under way and will take time;
• The authors have not responded, or their response has not resolved the concern;
• The journal is satisfied a real question exists but cannot itself establish the facts.

It is explicitly not a finding against the authors. When the matter concludes, the expression of concern is replaced by a retraction or withdrawn — and if withdrawn, the notice records that the concern was not upheld, so the authors are not left under a permanent shadow.

Retractions

An article is retracted where its findings can no longer be relied on, whether the cause was misconduct or honest error. Grounds include:

• Unreliable findings — from fabrication, falsification or a major error in method or analysis;
• Substantial plagiarism, under the plagiarism policy;
• Redundant publication — the work has been published elsewhere without disclosure;
• Unethical research — conducted without required approval or consent, under the research ethics policy;
• Compromised review — the peer review process was manipulated;
• Copyright infringement or another legal defect going to the substance of the article;
• An undisclosed competing interest serious enough that readers would have judged the work differently.

Retraction is not a punishment. It is a statement about the reliability of the findings, and it applies equally to an honest error the authors themselves discovered. Where authors identify a fatal problem in their own work and ask for a retraction, the notice records that they initiated it — which is to their credit, and the journal says so.

Where only part of an article is affected and the remainder stands independently, the journal issues a partial retraction identifying precisely what is withdrawn.

What a retraction is not

The journal does not retract an article because:

• Its conclusions are contested, or later work reaches a different finding — that is how research proceeds, and the answer is publication and debate, not retraction;
• Its findings are inconvenient to a government, institution, funder or donor — see editorial independence;
• An author has since changed employer, fallen out with a co-author, or would prefer it were not on their record;
• Someone has complained about it without evidence of a defect.

Removal

Removing an article defeats the purpose of the published record, and the journal treats it as a last resort. It is used only where:

• A court has ordered it, or the content is defamatory or otherwise unlawful;
• The article infringes a legal right in a way that cannot be cured by correction;
• Leaving it available would present a real risk to someone’s safety, or would breach the privacy of a research participant.

Even then, the article’s metadata — title, authors, DOI, date — and a notice stating that it was removed and on what basis remain in place, so the citation continues to resolve and the removal is visible rather than silent. Where a court order forbids explaining the reason, the notice says that a legal restriction applies.

How the process works

1. A concern is raised — by a reader, a reviewer, an editor, an institution, or the authors themselves.
2. Assessment. The editorial office establishes whether the concern is specific and credible enough to act on.
3. The authors are told what has been raised, shown the evidence, and given a fair opportunity to respond — normally within 21 days. Every author is copied, not only the corresponding author.
4. Investigation by an editor with no involvement in the original decision, with independent expert advice where needed, under the publication ethics policy. Where the question concerns research conduct rather than publication conduct, it is referred to the authors’ institution.
5. Decision by the Editor-in-Chief, taken on the reliability of the findings rather than on the authors’ cooperation.
6. Publication of the notice, and notification of the authors, their institution where relevant, and the indexing services.

Where the authors disagree with the outcome, they may appeal under the complaints and appeals policy. An article is not left uncorrected while an appeal proceeds; where a retraction is later overturned, the journal publishes a further notice reinstating the article.

A notice is published even where the authors do not agree with it, or cannot be reached. In that case it records that the authors did not agree, or could not be contacted, and which authors did agree.

What a notice contains

Every notice is a citable item with its own DOI, freely available, and linked in both directions with the article it concerns. It states:

• The title, authors and DOI of the article concerned;
• Who initiated the action — authors, editors, or the institution;
• The reason, stated specifically enough that a reader can judge what it means for the findings;
• For a correction, exactly what changed, so the original and corrected text can be compared;
• For a retraction, whether the cause was misconduct or honest error;
• Which authors agree with the notice, where they do not all agree;
• The date it was published.

Notices avoid language that alleges more than has been established. Where an institutional investigation reached a finding, the notice reports the finding rather than the journal’s own inference from it.

What happens to the article

The original article always remains available. What changes is how it is presented:

• A watermark and a banner on the article page and the PDF identify it as corrected, subject to an expression of concern, or retracted;
• The notice is linked from the article, and the article from the notice;
• The article’s Crossref metadata is updated so that reference managers, indexes and other services show its status;
• For a correction, the corrected version becomes the version of record, and the original remains accessible so the change can be verified;
• A retracted article remains in the issue in which it appeared. It is not removed from the archive or renumbered.

Requesting a correction

Authors who find an error in their own published work should write to the editorial office at ceoborjss@gmail.com as soon as they become aware of it, giving the article title and DOI, the exact location of the error, what it should say, and whether the conclusions are affected.

Readers may report a suspected error the same way, with the specific evidence. Reporting an error in one’s own work in good faith is treated as the responsible act it is, and never as an admission of misconduct. The obligation to report such errors is set out in the research integrity policy.