A journal received a letter raising concerns about a study that had been published a year earlier. An independent reviewer commented that the study’s conclusion did not agree with the study design and results. When asked to clarify, the corresponding author replied that objections about articles should not be allowed beyond three months after their publication, which was this journal’s stated time limit for receiving letters to the editor. The editor felt the article still needed correcting.
Question(s) for the COPE Forum
- Is there a time limit for considering critiques on published research?
- Is there a time interval for retracting an article or issuing an expression of concern?
Forum advice and follow-up
The Forum drew the distinction between (1) setting time limits for receiving and publishing post-publication critiques and author replies, and (2) investigating and addressing concerns received at any time about the integrity of published research. If editors receive readers’ concerns about an article’s integrity, they are ethically bound to act. The editor needs to decide if the paper needs to be retracted, referring to the COPE Retraction Guidelines, and could ask the author’s institution to investigate.
This COPE Forum case is categorised under two COPE Core Practices:
- Allegations of misconduct: “Journals should have a clearly described process for handling allegations, however they are brought to the journal's or publisher’s attention. Journals must take seriously allegations of misconduct pre-publication and post-publication. Policies should include how to handle allegations from whistleblowers.”
- Post-publication Discussions and Corrections: “Journals must allow debate post publication either on their site, through letters to the editor, or on an external moderated site, such as PubPeer. They must have mechanisms for correcting, revising or retracting articles after publication.”
The topic of this case highlights variations on the theme of post-publication correspondence, which is a form of post-publication peer review that contributes to “self-correction” of the scholarly literature, in a general sense.
In the long-term process of self-correction, authors and peers may publish new papers that expand, elaborate, modify, or even overturn earlier work in the light of new technology, evidence, or insight. In the short term, many journals publish formal critiques as readers’ letters or commentaries with the author’s response, to encourage academic debate of alternative opinions, interpretations, and explanations. Some journals allow online comments in a journal website or external website that the authors may or may not read or respond to.
Critiques versus notifications
As illustrated by this case, journals need to have clear and transparent policies on how to submit critiques as part of academic debate. To make direct formal critiques, readers require details of the submission and publication process, such as length restrictions, peer review, and author replies. Time restrictions (eg, weeks to months) may be imposed for the sake of recency. In turn, journals should establish office procedures for separating genuine correspondence from nuisance or defamatory correspondence, and for separating “regular” article critiques from notifications of concerns about an article’s integrity, such as possible mistakes or misconduct.
Notifications of readers’ concerns need an initial assessment and, if there is a valid claim, a follow-up investigation that may involve the institution (eg, according to COPE flowcharts). If judged necessary by the editor, perhaps with the help of independent post-publication peer reviewers, journal action could include “self-correction” by the journal, in a direct and specific sense, by way of publishing a correction, a retraction, or an expression of concern (eg, according to the COPE Retraction Guidelines). Occasionally, as an alternative to a formal correction or expression of concern, journals may add a post-publication editorial note or an addendum to an article to make a certain point of clarification, if appropriate.
Journals may receive notifications of concerns about an article’s integrity at any time, directly or indirectly. They can follow the respective COPE flowcharts titled “Responding to whistleblowers when concerns are raised directly” and “Responding to whistleblowers when concerns are raised via social media”. Although not strictly social media, the PubPeer platform has a paid dashboard service to notify journal administrators of journal mentions. Whether and how journals track, use, or respond to possible ethical concerns that have been raised in external websites may vary widely. This is an evolving area for which COPE will be developing relevant guidance.
COPE flowcharts on critiques
In this case, the critique was in fact a reader’s notification of possible integrity problems with an article. As shown in the flowchart, this type of correspondence would not be processed and published in the same way as a scholarly critique. Post-publication notifications of concerns about an article’s integrity should also not be subject to the same time limit as a letter to the editor. The difference could be clarified in the journal’s instructions to authors and should also be clearly explained to the author, with a repeated request to respond to the new peer reviewer’s comments about the conclusions not fitting the study design and results.
As noted in the full case and Forum advice, the necessary article amendment depends on the seriousness of the error. If the author is unresponsive, however, the institution can be contacted and an expression of concern added in the interim. If the author remains uncooperative, the editor is bound to retract the paper for the sake of not misleading readers, researchers, practitioners, policy makers, and other stakeholders. The retraction notice should transparently explain what post-publication review procedures were followed and what happened.
As an additional follow-up step, the editor could retrieve the original peer reviewers’ comments to check their academic quality and whether the reviewers had been correctly matched to the paper.
Finally, although not relevant in this case, it is possible that a reader’s correspondence contains both a valid concern about an article’s academic integrity or research/publishing ethics and also an element of scholarly debate within the allowed time limit. For example, the reader may make a valid academic argument about the remainder of the paper or the general topic or approach. As shown in the flowchart, these two components can be identified and treated separately, and some leeway in timing could be allowed for the publication of the critique.
Trevor Lane on behalf of the COPE Education Subcommittee