Specific rules and principles
The chapters in this section deal more specifically with certain aspects of scientific integrity in general, but also more specifically with research with health implications.
Courtesy is essential in all circumstances, especially in the event of conflict between professionals.
Honesty, transparency, reliability, benevolence and respect are the key elements of good conduct between professionals.
In all but the most extreme situations, such as suspected data falsification or harassment, the parties concerned discuss their conflicts or disagreements among themselves and seek conciliation before appealing to the institutional bodies responsible for deontology, scientific integrity and ethics, or to the university ombudsman.
Resources allocated to a project are used honestly and transparently between the professionals involved.
Hierarchical relationships are not used to discriminate against those at lower levels.
Professional partners agree, ideally in writing, on the individual contributions and responsibilities attached to research during its design, implementation, analysis, publication and valorization phases.
A provisional list of signatories for the article(s) derived from the research is drawn up a priori when the projects are written, in particular for the key positions of the first 3 and last 2 authors. This list is respected if the commitments of the individuals on it have been fulfilled. If not, it must be modified. It complies with the recommendations of the International Committee of Medical Journal Editors (ICMJE).
The search for funding is at the heart of every researcher's concerns. This funding can be public and/or private.
The following principles must be respected with regard to private funding:
- The autonomy and independence of the institution to which the researcher is attached must be preserved.
- Freedom of teaching and research, as well as freedom of methodology and publication, must continue to be guaranteed.
- Financial contributions are governed by a written contract with the donor, setting out any requirements or conditions in detail.
- The funder must not interfere with research results. Decisions on staffing and acquisitions remain the sole responsibility of the institution that received the funding.
- The origin of the funds must be known and disclosed as part of the publication.
- Acceptance of funds must not generate conflicts of interest with the funding organization.
Health products are quoted under international non-proprietary names without relaying marketing messages, whatever the medium (written, oral, online).
Regulations concerning "gifts" and agreements with manufacturers are respected.
Public funding is managed in accordance with the rules of the public institution.
Plagiarism exists in the following cases:
- use of works (including those from unpublished sources), ideas (including structures) or formulations of third parties (persons outside the group in question) without correct indication of the source;
- use of third-party works with slight modifications or translations, without correct indication of the source ;
- reuse of substantial parts of one's own work from publications in scientific journals and research projects or from unpublished sources without proper indication of the source, especially if third parties have contributed to these works and projects ("self-plagiarism");
- reuse of co-authored publications without proper acknowledgement of the source.
The indication of the source must be explicit and, if appropriate, must allow access to it (at least before a password).
The verbatim reproduction of a source must be enclosed in quotation marks.
Duplication of an article (sending the same document - or a slightly different version - to two different journals) is contrary to scientific integrity, unless duly authorized by the publishers. In all cases, the second article must cite the first.
The faculty has access to plagiarism detection software, and a procedure for detecting plagiarism, particularly in medical theses and other dissertations validating degrees awarded by Sorbonne Université, has been set up.
All Sorbonne Université Health Sciences lecturers have access to this software.
Although there is no consensus definition of what constitutes a predatory journal (see this article), the following criteria are most often used to identify this type of journal, whose aim is to increase profits at the expense of authors, science and its integrity (see also this link from the Interacademy Partnership):
- Manuscript review procedures suspect of complacency and lack of rigor (e.g.: no information on the decision taken by the journal and on the comments of all reviewers after submission of their review). Poorly designed website.
- Name close to a non-predatory reference magazine.
- Abusive submission, manuscript review and/or publication fees.
- Lack of clarity in the description of the manuscript review process and in the application of various fees.
- Verification of ethical and regulatory compliance of research ignored or neglected.
- Aggressive and indiscriminate article solicitation policy. Non-specific contact e-mail address (e.g. @gmail.com).
- Falsification of their impact factor or indexing in PubMed or DOAJ.
For legal reasons, it is unwise to list "predatory" journals, and Sorbonne Université Health Sciences has chosen to list presumed non-predatory journals in the health fields that are recommended as preferred for submitting manuscripts and citing publications. See https://sante.sorbonne-universite.fr/recherche-0/liste-des-revues-non-predatrices.
It is advisable to refuse to review articles submitted to a journal suspected of being predatory or not "honourable".
If you have any doubts about the publisher or the journal, we recommend that you consult those around you and the reference sites available online (https://thinkchecksubmit.org/, https://doaj.org/ and https://oaspa.org/).
Sorbonne Université's Faculty of Health Sciences advocates the deposit of preprints, open access manuscripts and manuscripts (not final PDFs) accepted for publication in open archives managed (https://hal.sorbonne-universite.fr/) by our university library, in accordance with the Chart for Open Access to Publications adopted by Sorbonne University in 2019 which lists good publishing practices.
Research results must be published in such a way as to be disseminated as widely and openly as possible.
A provisional list of signatories for the article(s) derived from the research is drawn up a priori when the projects are written, notably for the key positions of the first 3 and last 2 authors. This list is respected if the commitments of the individuals on it have been fulfilled. Otherwise, it must be modified.
The author criteria proposed by the International Committee of Medical Journal Editors (ICMJE) are appropriate. They recommend that an author fulfill 4 conditions:
- Substantial contribution to study design, data acquisition, analysis or interpretation AND ...
- The preparation or critical revision of the substantial content of the manuscript AND ...
- Approval of the final version of the manuscript AND ...
- Agreement to be accountable for all aspects of the work, ensuring that any issues relating to the integrity and validity of the work are and will be resolved.
The other autorat criteria indicated in the ICMJE recommendations, notably the author in charge of correspondence and authors of multiple groups, are relevant.
Hierarchical links and the fact of being responsible for a structure associated with the work, or of having contributed to the funding of the work, do not legitimize either systematic autorat, or an author's place in the list of signatories to a publication.
Although pseudo-objective methods have been proposed (see article 1 and article 2) to determine the order of authors according to their contribution, there is no consensus recommendation on how to determine the place of authors.
The first page of a manuscript may list "collaborators" who do not fulfill the 4 conditions for authorship. These collaborators will appear on PubMed (MeSH code [IR]).
Persons acknowledged by name on a manuscript must give their consent.
Authors' affiliations are those that existed at the time the work was carried out.
It is appropriate to publish negative results if they are perceived as useful.
Withdrawal of an article or individual authorship is a mark of scientific integrity.
Conflicts between authors should try to be settled between themselves before calling on the ombudsman or the referent for scientific integrity.
Sorbonne Université's Faculty of Health Sciences endorses Inserm's recommendations concerning the signing of scientific publications.
What is a preprint? A preprint is an author's manuscript that has not been corrected by peer review. It is also known as an author manuscript, concept paper, draft or working paper.
Do I have the right to make a preprint public? If the preprint is intended for publication and the author assigns part of his rights to the publisher, the latter may impose conditions on the sharing of preprints. To find out the publisher's policy, consult the Sherpa/Romeo database. In the case of an open-access publication, with an open license such as CC-BY, the preprint can be made public without any problem.
Where can I share a preprint? It can be made public on a preprint server such as BioArXiv (to find a server, see the Directory of Open Access Preprint Repositories (DOAPR) or Asapbio), HAL, social networks or a personal website. Please note, however, that the last two distribution channels do not allow for the permanent archiving of scientific production, unlike some preprint servers.
Why make a preprint public by depositing it on a server?
- Increased search visibility and speed of publication: preprints are immediately referenced in search engines and can be integrated into your author profiles;
- Recognition: this "publication" allows you to gather comments from your peers;
- Easy preprint citation: some servers allow you to attach a perennial identifier (DOI) to the preprint, making it easier for third parties to cite it;
- Free deposit for the author and free access for readers;
- Perpetual archiving of scientific production (depending on the server chosen) ;
- Possibility of indicating preprints in response to a call for projects (e.g. ERC).
Some publication platforms allow you to submit preprints for open peer review. E.g.: F1000Research, Open Research Europe (ORE) for European H2020 and Horizon Europe projects, Wellcome Open Research for Wellcome Trust projects.
Focus on Peer Community In (PCI), a preprint evaluation service: communities of researchers offer a free peer review service once the preprint has been deposited on a server. Following this review, the manuscript can be submitted to a journal (PCI journal, PCI-friendly journals, other journals, etc.). List of existing thematic PCIs.
For further information: Ouvrir la Science practical guide and CoopIST practical guide.
Do you have a question or need advice? The Publications & Open Access department can help. Contact: publications@sorbonne-universite.fr
Good practice in research data management, while already part of laboratory routine, has become essential to formalize since the publication of the December 2021 decree on scientific integrity. Public establishments must now define a policy for the conservation, communication and reuse of the raw results of scientific work carried out within them. The data management plan is now mandatory, as proof that this policy has been implemented.
A data management plan is a summary document that helps to organize and anticipate all the stages in the data life cycle. It specifies how the data produced during a research project will be managed, from its creation or collection to its opening, if possible, and archiving. The data management plan is a guide to adopting consistent practices within a team, and documents the fate of the data produced.
The data management plan is also a deliverable for funded projects. The project data management plan must be submitted to the funder at three points during the project: six months after the start of the project; mid-project (except for ANR projects of less than 30 months for this second stage); at the end of the project.
There are several types of data management plan, depending on the funder. However, two main types stand out: a model organized according to the chronology of the project (what are the practices for collecting, analyzing, sharing and archiving) and a model organized around the acronym FAIR (are the data Easy to find, Accessible, Interoperable, Reusable?).
- Data management plan templates provided by funders are recommended, but not mandatory: ANR, Horizon Europe, ERC, etc.
- There are also models created by individual establishments, which may be better adapted to the discipline and research practices involved, e.g. Inserm, Inca, Institut Pasteur.
As the December 2021 decree on scientific integrity reminds us, proper management of research data is an essential prerequisite for science with integrity and trust. If a project's data is open, it must be precisely documented so that it can be reproduced and reused.
The data management plan, a deliverable testifying to the good practices adopted by the team, includes a number of points on ethics and the integrity of practices, on the protection and security of personal data and health data, themes to which good data management is intrinsically linked.
Support can be provided by the BSU's Research Data and Digital Humanities unit data-bsu@sorbonne-universite.fr (data management plan), the Data Protection Officer (personal data) or the Research Ethics Committee.
- The vast majority of data produced as part of research funded at least half by public funds is considered to be public data.
- The Valter Law (2015) and then the Law for a Digital Republic (2016) introduced a principle of openness by default for a large proportion of them.
- Sorbonne Université adopted an open research data policy in June 2021, encouraging all staff to commit to this approach.
- Research funders are also demanding that the data produced for the projects they support be opened up. This is the case for funding under the Horizon Europe program (HE projects and ERC projects), where the opening of data and other research results, such as codes and software, is required.
- There are, however, a number of exceptions to openness for data protected by a specific right or regulation, for example for the following categories of data:
- Data belonging to third parties and covered by business secrecy or industrial and commercial secrecy;
- Certain data relating to the protection of the nation's scientific and technical potential (in particular laboratories located in restricted zones);
- Data corresponding to personal information, except in the case of anonymization, personal consent or legal or regulatory authorization;
- Data from materials covered by Article 6 of the Nagoya Protocol (genetic resources, in particular);
- Data covered by a contractual obligation of confidentiality arising from a partnership with a private player.
- The question of data openness should be anticipated as far as possible, ideally as early as the response to the call for projects. Responses to calls for projects for Horizon Europe now include criteria relating to open science (often in part B2 of the forms), which require anticipating the extent to which data can be opened up, and establishing a provisional budget for their proper management.
- Support can be provided by the BSU's Research Data and Digital Humanities unit data-bsu@sorbonne-universite.fr (data management plan), the Data Protection Officer (personal data) or the Research Ethics Committee.
When a teacher-researcher or researcher is asked to speak to the public as an expert, it is advisable BEFORE intervening to :
- Clarify the scope of the intervention, reserving your answer to determine whether you have the right level of expertise.
- Find out about the media, yourself or from colleagues or your institution, to determine whether it is presumed to be "honorable".
- Indicate any links of interest.
- Specify any conditions to be met (proofreading, revision of documents or interviews, etc.).
When a teacher-researcher or researcher is asked to give a talk to the public as an expert, it is advisable DURING the talk to :
- Stay within your field of expertise - don't take a stand outside your field.
- Identify your institutional affiliation and any links of interest.
- Respect one's duty of reserve, bearing in mind that academics and teacher-researchers are the only civil servants who are not bound by an absolute duty of reserve (as long as a certain form is respected): article L952-2 of the Education Code affirms their freedom of expression.
When a teacher-researcher is asked to speak to the public as an expert, it's a good idea to exercise your right of reply AFTERWARDS, if necessary.
Focus your speech, get to the point, be clear and precise, using non-technical terms that can be understood by everyone.
Communication on social networks requires :
- Bear in mind that, voluntarily or involuntarily, we always represent to some degree the institution or institutions to which we belong.
- Target the audience you wish to reach and the objective of your communication.
- Never remain anonymous.
- Respect professional secrecy. Do not communicate any content that you would not share with a journalist, a competing team or an industrial company.
- Set your initial privacy settings.
- Check your sources of information to avoid misinformation and rumors.
- Hide or block accounts that generate insults or untruths.