A publisher received communication alleging that a published article in one of their journals contained large portions of text taken without attribution from another article. Upon review of these allegations, it was determined that a table in the article had been reproduced. In the table the authors did not put direct quoted material within quote marks or otherwise identify the quoted material as such. The table values in a column are identical to that in another paper.
The authors responded that a mistake at the proofing stage had happened and unfortunately they were unaware of this problem. They apologised for the mistake. They said that in the proofing stage a column in the tables was inadvertently missed. They provided the corrected tables. They also said that the tables were needed to compare the results with those reported in the other published paper. Therefore, they compared the results using direct quotation on the column.
After the editors reviewed the submitted article it was determined that the "missed" data were not in their original submission. The authors responded that as they mentioned before, only a column was missed in the manuscript and that the corrected tables had now been provided.
The editors insist on retracting this paper because they do not believe the mistake was an error but an intentional reproduction of previously published work.
Question for COPE Council
- How should the publisher proceed with this case? The authors are asking for a corrigendum and the editors are insisting upon a retraction.
Advice on this case is from a small number of COPE Council Members. Most cases on the COPE website are presented to the COPE Forum where advice is offered by a wider group of COPE Members and COPE Council Members. Advice on individual cases is not formal COPE guidance.
This case involves a published manuscript that has been questioned. Upon investigation, the editors do not believe the authors innocently forgot to include a column in a table that was crucial to the findings. When the original files were checked, the column in question had not been submitted and the methods, and results sections were not clearly written to explain that comparison data were used. For this reason, the editors recommend a retraction, but the authors want a corrigendum. This is our understanding of the case and forms the basis for COPE’s comments; however, our advice still relies on editorial judgement and possible further investigation if the editors are not satisfied with the information currently in their possession.
The advice in such cases is to follow the COPE flowchart for suspected plagiarism in a published article. It would seem that the editors find the response from the authors insufficient and unsatisfactory, and it is clear from the flowchart what the next recommended steps should be.
If the editor believes that the omission was not intentional, the authors could be allowed to fix the oversight. The editor could let the authors revise the article and then the editor could make a decision or send it back to the reviewers to check.
However, it is unclear why the authors claim the results of others should have been included in their own results section, alongside their own (but omitted) data. Without an explanation in the methods, or an indication this was a replication study, or even mention of old and new data being compared in the discussion section, it is confusing for readers and later authors doing a systematic review. If the published version did not mention that a comparison was being carried out, the reader would naturally assume the conclusions stem from the results shown (which includes other researchers' results in this case).
In general, if the author's reply is unsatisfactory, the institution needs to be asked to investigate and yield the raw anonymised data, so the editors can decide if the conclusions remain the same. However, if the original conclusion cannot hold to begin with, because the study was really a comparison (ie, the conclusion needs to take into account the results of the comparison), yet a comparison was never mentioned anywhere, then retraction would be the right decision, as the revised version would need new methods, results, discussion and conclusion.