Investigation into violation of research integrity is concluded

In June 2017, PSI was made aware of allegations that members of its staff had submitted an article containing aspects of scientific misconduct to a scientific journal.

A preliminary review by experts showed that the allegations raised were solid. Therefore, in August 2017, the PSI director led a procedure to look into the allegations in accordance with Article 4 Section 1 of the Code of Procedure for Suspected Violation of Scientific Integrity in Research at PSI. A committee composed of external and internal experts was charged with the review.

Results of the investigation

The expert committee inspected an extensive set of documents, conducted four interviews, and held six meetings. In the process the committee also listened to those accused of misconduct. As a result, five flaws with regard to content and methodology were identified in the submitted, but still unpublished, paper. The flaws detected were judged and evaluated under the Guidelines for Integrity in Research at PSI. According to these guidelines, the following scientific misconduct was ascertained:

1. Insufficient documentation of analyses and calculations according to the applicable rules
2. Impermissible manipulation in the representation of different experiments
3. Non-reproducible experimental set-up
4. Running the risk of grey zones, namely with regard to the reproducibility of experiments
5. Insufficient reference to uncertainties in the discussion section of a publication

The investigative committee assumes that the authors misled the editor as well as the referees of Nature Communications and likewise accepted that the potential readers would be deceived. The deception arose through the omission of an adequate presentation of the methods used, merging different experiments and presenting them as directly belonging together, and the misleading correspondence with the referees of the journal. The investigative committee does not deny that the manuscript presents interesting scientific results.

In summary, the investigative committee concludes that, with their manuscript, the two authors did violate the rules of good scientific practice as stated in the guidelines of PSI.

Measures taken

The management of the Paul Scherrer Institute takes the incident very seriously and firmly holds that such conduct in the course of publication is not acceptable. For this reason, the following measures have been taken:

1. The editor of Nature Communications was informed about the findings of the investigation.
2. After that, the two staff members were obligated to take steps to withdraw the flawed, but still unpublished, manuscript. (Since other authors collaborated on the manuscript, withdrawal must occur with the consent of all authors.)
3. In all future publications in which they are acting as PSI authors or co-authors, the two staff members must also obtain prior written permission through the responsible laboratory head.
4. SNF, the Swiss National Science Fund, was immediately informed of the misconduct ascertained within the framework of the present investigation, since the authors of the manuscript used research funding from SNF for their work.
5. The PSI representative for research integrity, together with the PSI committee for integrity in research, is charged with subjecting the Code of Procedure and the guidelines Integrity in Research at PSI to a review. The goal is to prevent a co-author from waiting patiently until a manuscript has been submitted before expressing criticism towards third parties, instead of making use of the established procedures of PSI beforehand and first having recourse to the internal means for lodging a complaint.
6. Monitoring of participation in PSI-obligatory training and continuing education in the subject of Integrity in Research at PSI is being expanded.
7. The research group in which the two staff members worked has been dissolved.
8. Additional sanctions against the two staff members are being decided in a separate procedure, in accordance with federal personnel law.

The two staff members have accepted the director's ruling and apologised for their misconduct.

Significance of research integrity at the Paul Scherrer Institute

PSI has committed itself to excellence in research. To achieve this goal, various criteria must be fulfilled, including the commitment of highly motivated staff members and the development of a first-rate research infrastructure. Equally important, too, is an environment in which responsible conduct is fostered through cognisance of clear and documented rules derived from generally accepted norms and values. Clear rules form the foundation for credible work in research:

• First and directly, they help to disseminate robust knowledge by minimising the occurrence of errors.
• Second: Research is increasingly conducted in large groups whose active participants come either from different institutions or from diverse disciplines. The promotion of important values such as mutual respect, rules concerning authorship, confidential information, or handing over primary data is essential for a successful and sustained collaboration.
• Third: Transparent processes in a research institution such as PSI are important for preserving the trust of the public as well as that of our industrial partners and funding organisations.

Trust in research also depends essentially on the responsible conduct of the researchers. To sustainably ensure a high quality of research, every staff member is called upon to act responsibly and abide by these rules. A violation will not be accepted by PSI.