An author of a coauthored article published in our journal ten years ago contacted the outgoing editors with a request to have their name removed. The author in question is Dr A of University 1 and they are the paper’s first author. Their stated reason for doing so was that they had recently discovered errors in a table in the paper. The second author on the paper, Dr B, provided the original data and conducted the analysis. The third and fourth authors (Drs C of University 1 and D of University 2) aided in analysis and write-up. After requesting they be removed as an author, Dr A followed up 5 days later with a second email indicating they wished to withdraw the request.
The wider context is that Dr B has been the subject of much attention in the discipline and the media for having a number of papers retracted, all of which feature the same dataset and similar errors. Those papers were voluntarily retracted by all of the coauthors involved. Dr B has also been subject to several investigations at their most recent employer (University 1) and their previous university. Dr B’s employment at University 1 was terminated in the same week that Dr A asked for their name to be removed from the paper in our journal. Dr B is in the process of taking legal action against their former employer.
As a result of the termination of Dr B’s employment University 1 -- in response to an open records request -- released all documents related to the termination to a media source that then published them online as part of a story about Dr B’s papers. This included the results of the investigations related to the retracted papers, and addressed similar errors to those identified by Dr A in the paper published in our journal. However, they had been found to be immaterial to the ultimate conclusions of the paper and were likely due to translational errors or to idiosyncratic processes employed by different statistical packages used at the time of the paper being written. The original data were, and remain, unavailable due to a hard drive failure. Dr A stated that he was aware of the content of these papers.
We were disinclined to acquiesce with the original request to remove Dr A’s name from the author list because first authors are ultimately responsible for ensuring that a paper meets the standards of academic integrity. We are also disinclined to accommodate the follow-up request to ‘drop the issue’ because we have now been made aware of a potential error and are duty-bound to investigate it regardless of the authors’ preferences. Two of Dr B’s retractions were in our journal, one of which generated controversy between the journal’s home association and the owner of the journal, and were featured in several media articles. There is therefore particular jeopardy for our journal in this matter.
Subsequent to the requests from Dr A, we sought advice from our publisher. They suggested that the editors send one further piece of correspondence with Drs A and B asking for a full explanation of the concerns which had been raised and then dropped, informing them that if they are unresponsive or provide an unsatisfactory response that the institution’s research ethics committee will be asked to investigate and give guidance on whether or not the research is reliable or requires retraction. We have chosen to seek advice from the COPE Forum prior to doing this.
We need guidance on how best to proceed in a manner that ensures scientific integrity of the journal and treats all authors fairly. As we see it, there are four courses of action.
- Retract the paper (with the acquiescence of the authors or not). This would require us to request an investigation by University 1 and potentially University 2.
- Because the data for the paper are no longer available, attach an amendment to the paper that indicates a “Concern of the Editors” regarding potentially faulty analyses.
- Allow for the publication of a Corrigendum/Errata (this may not be possible since the original data are no longer available).
- Allow the authors to publish a note to be attached to the paper that indicates the paper has minor errors but that these do not alter the findings, and that this has been verified by a disinterested third party (chosen by us, the editors of the journal).
Questions for the COPE Forum
- Which of the above courses of action would you recommend?
-
Are there any other actions we should take?
The Forum agreed that once journals are aware of a potential problem they must investigate, regardless of whether the concern has been retracted. An important issue in this case is that their ability to investigate is compromised because of the non-availability of the original data. It is therefore not possible to ask for re-analysis; all that is known is that there is an error in a table. Even if this is not important for the paper’s results, the error should still be acknowledged in some way, for example, via an Expression of Concern or an editorial note. This can note the issues raised with the data and the fact that it is impossible to determine its extent or impact. An Expression of Concern may be a permanent outcome if that best follows the journal’s guidelines. In other cases it may be followed by a Retraction depending on the outcome of any further investigation.
The editors could also consider removing the problematic table if it is not integral to the results (since it cannot be reliably corrected), although they may wish to get an expert’s advice here. Such an outcome could be flagged as a Correction or a partial Retraction depending on the impact on the paper.
The journal should also notify the relevant institution/s of the problem with the article as it may be relevant to their own investigations. Any correspondence with the authors, for example, over errata, corrections or Expressions of Concern, should go direct to them.
Regarding the author’s original request, the relevant section of the COPE Retraction Guidelines states that:
'Authors sometimes request that articles are retracted when authorship is disputed after publication. If there is no reason to doubt the validity of the findings or the reliability of the data, it is not appropriate to retract a publication solely on the grounds of an authorship dispute. In such cases, the editor should inform those involved in the dispute that they cannot adjudicate in such cases but will be willing to publish a correction to the author/contributor list if the authors/contributors (or their institutions) provide appropriate proof that such a change is justified.'
Read the Retraction Guidelines to see the latest version of this guidance.
More detailed advice on handling Expressions of Concern will be published by COPE in due course.