Scientific integrity home page
- Details
The Observatory's position on Scientific Integrity
The Observatory's positioning is consistent with the advice of committees and actions implemented by Université Côte d'Azur. The Scientific Integrity Committee, which advises the OCA's Scientific Integrity Officer, aims to address scientific integrity issues by taking into account the specificities of research practices in Earth and Universe Sciences.HRS4R Label
Scientific Integrity agenda
Scientific Integrity news
How should scientific integrity misconduct related to a publication deposited in HAL be addressed and dealt with? Although reports are rare, they unfortunately do occur. The CCSD is implementing a system that will allow it to rely on the scientific advisors of its three supervising institutions.
> OFIS: Artificial intelligence: new regulations to promote transparency
Generative AI is changing the way we do research. The scientific integrity issues associated with these developments are leading institutions around the world to review their codes of good practice, and transparency is the watchword.
Alexei Grinbaum, a specialist in digital ethics, sheds light on these issues for Ofis. He explains why the regulatory requirement to mark AI-generated content is a step forward for scientific integrity.
AI enters the European code of conduct.
Regarding the use of generative AI, the guidelines are quite similar to the recommendations that abound elsewhere in the world. Any content or results using generative AI must be clearly identified as such, the methods used must be explained, and the use must be appropriate and compliant with regulations. Some uses are banned such as «direct use» of generative AI in applications for funding and approval of research work.
The lack of an English version on the Chinese Ministry's website, the one Ofis gives you to read was carried out using the multilingual translation model mbart-large-50-many-to-many-mmt (611M parameters), freely available on HuggingFace.
> OFIS : Retraction of published articles: a good thing ?
Paper mills, predatory journals, fraud or simple publication errors: correcting scientific literature is more necessary than ever. Not only for the scientific community, but also for society as a whole, which needs expertise and reliable results..
Retraction of published articles is at the heart of this process. For more than 12 years, Retraction Watch has been monitoring the evolution of this post-publication correction and commenting on topics related to scientific integrity. Today, it has 40,000 retractions in its database. Ivan Oransky, one of the site's co-founders, analyses the main trends.
> OFIS : It's not so easy to report scientific fraud ?
While scientific integrity has been declared a matter of vital national interest in the United States, little is known about how it is perceived by the country's most successful researchers. With this in mind, two researchers at Virginia Tech University conducted a survey of young recipients of funding from the National Science Foundation. They targeted two disciplines, “Civil and Environmental Engineering” and “Computer science and Engineering”, which, according to these researchers, are particularly exposed to strong competition.
According to the results published in Scientific reports, only 30.7% of respondents said they would report a breach of scientific integrity they witnessed - 8.6% said they would not report such a situation. The authors put forward two possible explanations: the pressure exerted on whistle-blowers and the lack of incentives to report breaches. Only 12.7% of respondents consider that such reporting would have a major impact on their field of research.
When asked about the sanctions to be implemented in the event of proven breaches that distort scientific results, the first measure recommended by the respondents was the retraction and correction of articles, followed by the dismissal of the authors of the breach or their listing in a public register.
> OFIS : A PaperMills detector for publishers
One year after the announcement of the project, the STM Integrity Hub - makes a first version of its Paper Mills article detector available to publishers. The Integrity Hub of the international association Scientific, Technical and Medical Publishers (STM) brings together numerous publishers in a collaborative approach to develop cutting-edge tools to help this community deal with breaches of scientific integrity. Paper mills, which produce and sell fraudulent articles, are a major target.
The tool, in the form of an application, enables publishers to automatically screen the papers submitted to them. Based on 70 key indicators, the detector pinpoints any problems and signals whether the article is or may be from a paper mill. It is then up to the editor to investigate further. The application must ensure confidentiality, data encryption, etc. and, to remain effective, does not reveal its criteria. Its vocation is to evolve constantly, so as to thwart the tactical adaptations that paper mills are bound to implement.
> OFIS : If in doubt about a publication, go directly to Science
Correcting scientific publications, a leitmotiv for all. In Strengthening the scientific record, three of the heads of Science's affiliated journals, headed by Holden Thorp, describe the new guidelines for speeding up post-publication criticism and discussion of their articles: e-letters submitted online now replace old-fashioned technical comments or technical responses.
In another notable development, they also invite anyone with concerns about the integrity of an article published in one of their journals to contact them directly at : Science_data@aaas.org. The e-mail will be examined in accordance with the code of the publication's ethics committee.
« We are increasingly concerned that some actors may attempt to exploit or unfairly distort the open research environment and misappropriate research results for economic, strategic, geopolitical or military purposes. This undermines the principles and values underpinning open, transparent, reciprocal and accountable international research cooperation, as well as the integrity of research, and may present security risks.».
The ministers will therefore continue to support the G7 working group on challenges to the integrity and security of the global research ecosystem. They also pledge their support for the Open Science Task Force, with encouragement for “research on research”.
> OFIS : The doctoral oath
A new provision in the French system to promote scientific integrity, the doctoral oath has been mandatory since 2023.> OFIS : ChatGPT
The arrival of the conversational robot, including in the world of research, raises new questions of scientific integrity. In particular on the status of author of a publication.
> ANR strengthens its policy on ethics, scientific integrity and professional conduct