Lucent Explainers

What a Retraction Means and How to Check a Paper

21 November 2026

A retraction is a notice from a journal saying a published paper should no longer be treated as reliable, usually because of an error, misconduct, or data that cannot be trusted. It does not erase the paper, it flags it. Retractions are a small fraction of all papers published, but common enough that checking a paper's status before you cite it is worth the half a minute it takes.

What a Retraction Actually Means

When a journal retracts a paper, it does not delete it. The paper usually stays online, often with the word RETRACTED stamped across every page as a watermark, and the publisher adds a short notice explaining what went wrong and when.

This is different from a correction. A correction fixes a specific error, such as a wrong figure or a mislabelled axis, while the paper's main conclusions still stand. A retraction is a bigger step. It means the journal no longer stands behind the paper's findings as part of the reliable record.

Because the paper stays searchable, it can still turn up in a database search or a reference list years later. That is exactly why checking its status matters. Nothing on the surface stops it looking like an ordinary paper unless you look for the notice.

Why Papers Get Retracted

The most common reason is an honest error. Someone, often the original authors, finds a mistake after publication, such as mixed-up samples, a flawed calculation, or software that produced the wrong numbers, and reports it themselves. This kind of retraction is not a scandal, it is the system working as it should.

Other retractions follow genuine misconduct: fabricated data, duplicated or doctored images, or plagiarism. These are often spotted by image-screening software, plagiarism checkers, or readers on post-publication forums who flag odd patterns in figures or statistics.

A growing number come from so-called paper mills, businesses that sell fake or templated manuscripts to researchers under pressure to publish. Once a journal spots one paper from a mill, it often finds a cluster of similar ones and retracts them together, which can make retraction counts jump for reasons that have nothing to do with any single author's honesty.

How Common Retractions Are

Retraction remains rare set against the total number of papers published each year. The overwhelming majority of papers are never retracted, corrected quietly if at all, and simply sit in the literature unchallenged.

The number of retractions has risen over the past couple of decades. This is partly because detection has improved, with better image-duplication software, plagiarism tools, and a more active community of volunteer sleuths checking published work, not necessarily because fraud itself is becoming more common.

Rates vary a great deal by field and by journal. Some fields rely more heavily on images or complex datasets that are easier to tamper with, and some lower-tier journals have weaker screening, so a handful of outlets account for a disproportionate share of retractions overall.

How to Check a Paper's Status

The fastest method is to search the paper's title or its DOI, a unique identifier every paper has, in the free Retraction Watch database. It is built specifically for this and covers a huge range of journals and publishers.

You can also check the publisher's own page for the paper. Look for a red banner, a watermark across the PDF, or a linked notice near the title that says the paper has been retracted or is under an expression of concern. Most major publishers now flag this clearly if you know where to look.

Tools such as PubPeer let you see whether other researchers have raised concerns about a paper's data or images, even before any formal retraction happens. Some reference managers and browser extensions will also flag a retracted paper automatically when you add it to your library, which is worth setting up if you read a lot of literature.

What a Retraction Doesn't Tell You

A retraction does not automatically mean fraud. Many are the result of an honest mistake that the authors themselves caught and reported, and treating every retraction as a scandal misreads what actually happened most of the time.

It is also worth distinguishing a retraction from an expression of concern, which is a lighter, often temporary notice a journal adds while it investigates a paper. The paper is not withdrawn at that stage, it is simply flagged as under review, and the outcome could still be that nothing is wrong.

A retracted paper's claims can keep circulating even after the notice goes up, because other papers already cited it and those citations do not update themselves. This is why checking the source paper directly, rather than trusting a citation secondhand, protects your own work from repeating a claim that the record has since withdrawn.

Common questions

Does a retraction mean the authors committed fraud?

No. Many retractions follow honest errors, such as a mislabelled sample or a coding mistake that the authors found and reported themselves. Only a minority involve proven misconduct such as fabricated data or plagiarism.

What's the difference between a retraction and an expression of concern?

An expression of concern is a temporary flag a journal adds while it investigates a paper, and the paper is not withdrawn at that stage. A retraction is the final step, issued once an investigation confirms a serious problem with the paper's findings.

Can a retracted paper still be cited?

It can be cited to discuss the retraction itself or as a historical note, but not as evidence for its original findings. Citing it without mentioning the retraction risks passing on a claim the record no longer supports.

Do retracted papers keep getting cited by other researchers?

Yes, often by accident. Citation managers do not always flag a retraction automatically, and some people cite a paper secondhand without reading it, so the claim keeps circulating in later papers, reviews, and sometimes even guidelines.

Does one retraction mean I should distrust everything else an author has published?

Not automatically. A single honest-error retraction says little about an author's wider body of work, though a pattern of several retractions, especially for misconduct, is a reasonable reason to look more closely at their other papers.

Lucent, once a week

We read the new research and write up what actually holds, in the same plain language as this page. No jargon, no hype, three papers a week.

Read the latest issue

Read next