Journals publish a disproportionate number of articles by their editors.

An analysis of over 1,000 scientific journals published over 38 years shows that 12% of journal editors publish a fifth of their own research.

More than 1 in 10 researchers who are also scientific journal editors publish a fifth of their own papers in their journals, and 1 in 20 publishes a third of their own papers. This raises the question of whether editors’ submissions are viewed more favorably.

For more than a decade, there has been concern that a growing number of research papers contain errors. This is sometimes referred to as a replication crisis in science, as flaws can come to light if other research groups fail to replicate the results.

Part of the problem is that scientists are forced to publish as many papers as possible because it helps them get promoted and gain access to research funds.

Decisions about which articles to accept are made by the journal’s editorial board, which usually includes practicing research scientists. While editors seek advice on submitted papers from other scholars who are experts in this topic, known as peer review, they still have a lot of influence over the process.

To gauge the extent of the problem, Bedor Al-Shebli of New York University Abu Dhabi, United Arab Emirates, and her colleagues analyzed a database of more than 1,000 journals published between 1980 and 2018 by Elsevier, the company behind the fifth part of world magazines. science articles.

While there were large differences in self-publishing rates, 12 percent of the editors of these journals published more than one-fifth of their articles in their journals, and 6 percent published more than a third in their journals.

The team used the software to match each of these editors with a similar researcher, for example in the same scientific field. The results show that these comparison researchers tended to have only a small percentage of their papers accepted by the journal in question.

This makes it more likely that papers submitted to the journal by its editor will be treated more favorably, “which could be seen as an abuse of the scientific publishing system,” according to the AlShebli group.

“Publication in a journal should be a signal that the journal believes it is good science,” says Stuart Buck, who runs the Good Science Project, a nonprofit American organization that aims to increase scientific rigor. “At least, [self-publishing] looks like a conflict of interest.”

Dorothy Bishop of the University of Oxford says some editors may be trying to publish high-quality scientific data in their journals to improve their profile rather than advance their careers. In such cases, scientists should waive the editorial board’s decision about whether to accept the paper and state in the paper that it did, Bishop says.

This is recommended in the guidelines of the Committee on Publication Ethics, an international advisory body for journals. The editors whose research was included in the latest study may have been following this process, as Al-Shebli’s team did not appreciate how often such claims appear in self-published articles.

Elsevier has reached out for comment.