Research integrity
Research misconduct and detrimental research practices
Scientists should be familiar with the definitions of research misconduct and the procedures for dealing with it, regardless of whether they will ever be accused of misconduct or otherwise involved in a misconduct case.
Research misconduct (RM) is a breach of OSR standards and of those expected by OSR funders and sponsors, a betrayal of the trust placed in OSR by the public, and the failure to comply with the high expectations of the scholarly community for research integrity (RI) and accurate and experimentally-supported communication. OSR commits to vigorously investigating, and if warranted, taking action on any credible allegation of research misconduct (RM), pursuant to this document11.
Of note, due to increased media, regulatory, and public scrutiny, research institute and investigator reputations and funding depend as much on perceptions of integrity as on integrity itself.
Any activity that seriously violates the core principles of RI is considered to be RM. Such activities essentially amount to cheating, lying or stealing. More in detail they include, but are not limited to:
- Fabrication of data and/or results and their publication or inclusion in grant applications;
- Falsification (data manipulation, suppression, or modifications to experimental conditions that are not reported or disclosed e.g. to editors and peer-reviewers);
- Plagiarism including serious self-plagiarism (duplicate publications)
- Conscription of authorship without consent
- Wrongful denial of authorship to deserving persons
- Wrongful and deliberate obstruction of other scientists' research activity or attempts to diminish another person's scientific reputation;
- Deliberate damage of research work (including damaging, destroying or manipulating experimental designs, equipment, documents, hardware, software, chemicals, or other materials needed by another person to carry out their research);
- Destruction of primary (source) data and deliberate violation of good RDM practices5
- Gross neglect of supervisory duties including, but not limited to inducing, directing, encouraging, or knowingly allowing others to engage in fabrication, falsification, or plagiarism.
- Non-compliance with scholarly journal policies (both journal-specific and general), including, but not limited to: multiple simultaneous submissions to different journals, re-publication of one’s own prior work (duplicate publications) (see also item 3 above), failure to rectify/correct one’s own published record when errors or inaccuracies are found or reported by self or third parties.
- Not acting (by reporting, see below) on directly acquired knowledge of others’ potential RM.
- Forceful imposition of undeserved key authorship of scholarly communications such as manuscripts, conference abstracts, etc., by power harassment (i.e. via the abuse of hierarchical superiority), whereby refusal to comply might be perceived as being professionally very disadvantageous.
- Failure to disclose substantial financial or other conflicts of interest to internal or external monitoring bodies, organisations or publishers upon preparation, submission or publication of a manuscript or a grant application, or while acting as a reviewer for scholarly journals, funding agencies, or as a member of internal and external career advancement and recruitment committees or other professional duties.
11 The current OSR procedures (PSQ 017) for dealing with allegations od RM is available at http://www.hsr.it/wp-content/uploads/2018/01/017-1CattivacondottainRicerca2017.10.31.pdf
Seeking guidance on RCR, RM and DRP
Any individual who has questions/doubts on RCR practices and/or what might entail a possible RM or DRP is advised to contact the Head of the RIO (HRIO) for counsel and advice. This contact is confidential and need not be reported to other officers. Should the HRIO learn of a potential/alleged case of RM, s/he will ensure the prompt activation of the relevant internal procedure.
The HRIO provides an annual report on the RIO activities to the Chief Operating Officer for Research, the Scientific Director, the committee of Research Directors, the OSR Supervisory Body (Organismo di Vigilanza; OdV) and the Director of Human Resources.
Other breaches of RI, not so serious as to amount to RM, are referred to as questionable or detrimental research practices (QRP or DRP) and are nevertheless firmly discouraged and may also be sanctioned. Examples of DRP of varying seriousness include, but are not limited to:
- Misrepresenting research achievements.
- Knowingly supporting, or publishing in journals that undermine the accepted quality control standards of scholarly publishing (COPE, 2017; Kolata, 2017; Shamseer et al., 2017, NIH, 2017). The RIO is available for advice and assistance on establishing whether journals are of a questionable nature.
- Unauthorised removal of research samples or test materials, log books, and primary data from OSR premises.
- Acceptance or presumption of underserved/unsubstantiated authorship of a scholarly publication of scientific authorship or co-authorship.
- Ignoring putative RM by others or being party to inappropriate responses to RM (depending on the gravity of the case/allegation, this may imply RM, see point 12 above)
- Selective citation of papers that support one’s own findings or to please certain parties (reviewers, editors, colleagues).
- Exaggerating the importance and applicability of one’s own findings
- Self-plagiarism, excluding duplicate publications (which qualify as RM, see above) and limited re-use one’s own previously published text.
- Failure to fully disclose experimental conditions and materials.
- Failure to report negative (i.e. not respondent to one’s expectations) data.
- Directly connected to point 8 above, selective reporting, a.k.a. p-hacking or unduly influencing the data collection or inclusion process or statistical analyses performed to produce a statistically significant result. In general, the practice of reanalysing data in many different ways without a clear scientific rationale to yield a target result, or deciding to collect more data after failing to see the expected result within properly sampled groups are also instances of DRP.
- Failure to disclose potential financial or other conflicts of interest to internal or external monitoring bodies, organisations or publishers upon preparation, submission or publication of a manuscript or while acting as a reviewer for scholarly journals, funding agencies, or as a member of internal and external career advancement and recruitment committees or other professional duties.