Corrections and retractions

Publication isn't the end. Sometimes errors are found later, by you or by others. How they're handled matters for science, and for your reputation.

The options

• Correction (erratum or corrigendum): a mistake that doesn't change the main conclusions is fixed with a published notice linked to the paper.
• Expression of concern: the journal warns readers while an issue is investigated.
• Retraction: the paper's findings can't be relied on. The paper stays online, clearly marked as retracted, with a notice explaining why.

COPE's retraction guidelines set out when each is appropriate.

Why papers are retracted

Honest error is one reason. But a 2012 study by Fang, Steen and Casadevall of more than 2,000 retracted biomedical papers found that about two thirds were retracted because of misconduct: fraud or suspected fraud, duplicate publication or plagiarism. Only about a fifth were due to error.

If you find an error in your own paper

1. Check carefully, and tell your co-authors.
2. Contact the journal promptly, explaining the error and its impact.
3. Request a correction, or a retraction if the conclusions don't hold.

Researchers who correct their own errors openly are generally respected for it. Some journals call these "self-retractions" and treat them differently from misconduct.

Retracted papers keep getting cited

Studies show many retracted papers continue to be cited, often without mentioning the retraction. Always check whether a paper you cite has been retracted (Retraction Watch database, or the publisher's page).

RememberErrors happen. Fix them openly: corrections for small ones, retraction if the conclusions fall. And check what you cite.