About & policies · Journal of Reproducible Statistics

Corrections, retractions & versioning

last updated 2026-07-30

A journal that reruns every paper takes on more post-publication duty, not less. A published result can stop reproducing. Maintaining the published record is therefore a core responsibility, and publication marks the start of scrutiny, not its end.

Versioning

Published articles are versioned. A substantive change publishes a new version with its own version note, the prior version stays resolvable, and the record shows what changed and when. Typo-level fixes that touch no result are applied silently. Anything touching results is a recorded amendment.

Corrections, retractions and expressions of concern

We operate the COPE-aligned set of post-publication actions, each as its own linked notice on the article.

  • Correction: an error that does not invalidate the work.
  • Expression of concern: a public notice on the article warning readers that a serious concern about its validity or integrity is under investigation and not yet resolved.
  • Retraction: findings that can no longer be relied on, whether through error or misconduct. Retraction notices state the reason and are permanently linked to the article, which remains visible and clearly marked.

Correction and retraction metadata will be deposited machine-readably with the DOI record, so a paper's current standing follows it into reference managers and indexes, and a reader will always see whether the version they cite still stands.

Reproducibility notes

When a later rerun, or a reader report, shows that a published result no longer reproduces (an environment rotted, a dependency changed, a data source moved), that is handled as a lightweight reproducibility note on the article. It is short of a retraction, used where the science still stands but the artifact broke, and it states exactly what no longer runs. Where the authors repair the package, the note records the fix.

Post-publication discussion

We will open accountable, attributable post-publication discussion. Comments will carry a real identity and are never anonymous, authors will always have a right of reply, and the exchange will be preserved alongside the article as part of the record. Until then, a substantive challenge (a real error, a result that no longer reruns) goes to contact@fairpressjournals.com and follows the corrections process below. Once verified, it routes into the corrections machinery above and can trigger a re-execution of the article's package.

How to report a problem

To report an error, a result that does not rerun, or an integrity concern, write to contact@fairpressjournals.com with the article and the specifics. Integrity concerns follow the procedure in publication ethics & malpractice. Disagreements with editorial decisions follow complaints & appeals.