The editor received an allegation of image falsification from a whistleblower relating to two papers published more than ten years previously (under the previous editor and publisher). A senior editor reviewed the allegations according to COPE guidelines, and decided there was evidence of image duplication. The allegations were then put to the author who was unable to supply the original data and asked for the opportunity to repeat the experiments. The response was again reviewed by the senior editor and discussed at a meeting of the editorial board, which again confirmed the view that there was a problem with one particular image.
It was therefore decided not to accede to the request and the author was notified that a retraction for one paper would be issued (earlier it had also been agreed that a Corrigendum would be issued for the other article). This decision was challenged by the author who raised a formal complaint in relation to the handling of the case. As a result, an internal enquiry was conducted, which found that the matter had been investigated correctly, that all communications with the author had been handled appropriately but recommended that an Expression of Concern be raised instead of a retraction. The Editor agreed to this proposal.
Question for COPE Council
- Could the author have been permitted to repeat the experiments and, if so, would a Corrigendum have been more appropriate than the proposed Expression of Concern?
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.
According to the COPE flowchart on fabricated data in a published article, replying that the raw data cannot be found but being willing to repeat the test would be an unsatisfactory response. It is very problematic to consider "reproducing " the experiment as personnel and equipment may not be the same—basically creating a new research project. If the data are reproducible then there should be alternate data supporting the same conclusion. Also, most laboratories should have clear and consistent data preservation policies. Hence, according to the flowchart, the next step would be to inform the author that the institution will be asked to investigate. Where data are concerned, the institution should likely be involved to sort out if manipulation/falsification occurred, and if a retraction is warranted. An Expression of Concern would be appropriate while the institution investigates.
The journal should therefore ask the institution to investigate what happened and advise on the next steps. If the institution finds against the author, the journal could then proceed to full retraction. If the institution finds that an honest mistake occurred, then a correction could be done instead. If the institution refuses to investigate, then the editors must make an assessment of the importance of the figure to the conclusions of the paper and act accordingly. The key issue is that any inferences that have been drawn from that figure are unsafe. If the figure in question is tangential to the central argument of the paper, perhaps an Expression of Concern may be reasonable, but retraction may be the most likely course.