REVIEW
Experts read it first
Peer review is a quality control system in which experts in the field judge a piece of work and whether it is suitable for publication.
Check a paperHOW SCIENCE GETS CHECKED
Checked,
but not guaranteed.
Most research news starts with a published paper.
ASK FIRST has it been reviewed?
1665 Year Philosophical Transactions began
- Experts check the work
- Preprints come first
- Mistakes get corrected
- How a finding reaches print
Before many papers appear in a journal, other experts read them and judge whether they hold up. That check, peer review, is useful, but it is not the last word, and not every paper you read about has had it.
THE SHORT ANSWER
Peer review is
a filter, not a stamp.
Three stages a research finding can pass through.
Knowing which one you are reading helps.
HISTORY
Since 1665
The Royal Society's Philosophical Transactions, launched in March 1665, calls itself the world's first and longest-running scientific journal.
See the numbers
PREPRINTS
Shared before review
Preprint servers let researchers post papers early. arXiv, founded in 1991, does not peer review what it hosts: content is presented as is.
Read news wiselyA QUICK PAPER CHECK
What stage is
this research at?
Pick the kind of paper behind a news story to see what checks it has had. The bar shows how much weight the publishing system itself puts behind it.
Based on guidance from Sheffield Hallam University Library, arXiv and the Committee on Publication Ethics (COPE). General information only.
FOUR QUESTIONS FOR A SCIENCE STORY
Read research news
like an editor.
None of these needs specialist knowledge. They simply help you place a finding in the publishing process before you share it.
Find the source
Look for the journal or server named in the story. A link to the paper itself is a good sign.
Ask if it is a preprint
Preprints are valuable for speed, but they have not yet been through peer review.
Check for notices
Look on the paper's page for a correction, expression of concern or retraction notice.
Wait for more studies
Peer review checks one piece of work. Confidence grows when other studies point the same way.
TWO NUMBERS WORTH KNOWING
An old system,
still being fixed.
One figure from the start of scientific publishing,
one on how often the record now gets corrected.
1665
Year Philosophical Transactions began
It appeared monthly and cost one shilling. By the 1830s, facing competition from other journals, the Royal Society introduced more rigorous and systematic expert peer review.
FIRST JOURNAL · ROYAL SOCIETY
10,000+
Research articles retracted in 2023
A Nature news analysis reported that retractions issued for research articles in 2023 passed 10,000, a new annual record.
RETRACTIONS · NATURE, 2023
COPE's guidelines say batch retractions may be needed where the publication process has been systematically manipulated, for example by paper mills.
WHEN THINGS GO WRONG
Correcting
the record.
Retraction is how journals tell readers that a paper's findings cannot be relied upon. COPE, the Committee on Publication Ethics, lists reasons such as major errors, data or image irregularities, fraud, plagiarism, unethical research, compromised peer review and undisclosed conflicts of interest.
A retraction notice should be published promptly, be freely available and link to the original article. Where doubts exist but the evidence is not yet clear, an editor may publish an expression of concern instead.
Questions
Questions people ask
What is peer review?
A quality control system for published research. Experts in the same field assess the strength of a piece of work and whether it is suitable for publication. Practice varies between journals.
What is the difference between single-blind and double-blind review?
In single-blind review the reviewers' names are not revealed to the authors. In double-blind review neither the authors' nor the reviewers' names are revealed. Some journals use open review.
What is a preprint?
A paper shared publicly before peer review. arXiv, founded in 1991 and hosting more than three million articles, moderates submissions but does not peer review them.
How many papers are retracted?
Nature reported that retractions of research articles in 2023 passed 10,000, a new annual record.
Does a retraction mean someone cheated?
Not necessarily. COPE says unreliable findings can come from honest error as well as misconduct, and that retraction is meant to correct the record, not to punish authors.
Where the facts come from
Updated