Are retraction notices becoming clearer?

Retraction notices issued by scholarly journals have, to an extent, become clearer over time, a new analysis suggests.

The authors of the study examined 768 retraction notices issued by journals run by the publishing giants Springer Nature and Wiley in 2010, 2015, and 2020 (Account. Res. 2023, DOI: 10.1080/08989621.2024.2366281). They obtained retraction records from the Retraction Watch Database, which contains more than 49,000 retractions to date.

The researchers used retraction guidelines issued by the Committee on Publication Ethics and recommendations from the research integrity blog Retraction Watch to rate retraction notices. Two authors of the study are affiliated with Retraction Watch.

The authors judged retraction notices using criteria including whether notices were freely available and easily accessible, whether they mentioned investigations carried out by journals or institutions, and whether they were candid and transparent about the reasons for retraction. The study authors also checked if notices specified whether the journal or the authors pulled the study and if the journal and all the authors agreed to the retraction.

They found that retraction notices released by Springer Nature became clearer between 2010 and 2020, while those issued by Wiley did not. For some of the criteria examined, the Wiley notices became worse over the time, the study found.

“Maybe we’ve set the bar relatively high because what are the incentives for publishers to do this?” asks study coauthor Misha Angrist, a social scientist at Duke University, referring to the clear communication of retraction notices. “If we truly want science to be self-correcting, it behooves us to be transparent about the ways in which authors get things wrong.”

“Retraction is a necessary but not sufficient step,” Angrist says, noting that many retracted studies continue to be cited after being pulled from the literature. “That has a trickle-down effect that diminishes the science that follows.”

Jodi Schneider, an information scientist at the University of Illinois Urbana-Champaign whose work has found that 96% of citations to retracted studies don’t acknowledge their status (Scientometrics 2020, DOI: 10.1007/s11192-020-03631-1), says the criteria the researchers used to analyze the retraction notices are “very sensible” but could be a little clearer.

Schneider points out that retraction notices sometimes appear before the conclusion of institutional investigations into the cause of the error, making it difficult for publishers to include the most up-to-date and accurate details on retractions.

Chris Graf, research integrity director at Springer Nature, says it’s impractical to expect retraction notices to always differentiate between misconduct and honest error.

“Publishers and editors in most cases will not have sufficient information to be able to make the differentiation, and the imbalance of power between authors may also mean that information that is provided to us may not be reliable, to the detriment of the more junior parties,” Graf says.

“Wiley supports the call for increased standardization of the retraction process, in addition to the consistent use of existing frameworks,” adds Mike Streeter, Wiley’s director of research integrity strategy and publishing ethics. He notes that Wiley has been involved in crafting the National Information Standards Organization’s recent recommendations on the standard communication of retractions, corrections, and expressions of concern.