Catégorie : EN

Catégories
EN

(De)generative AI and research integrity

Author : Jurij Selan

The phenomenon of artificial intelligence (AI) is inherently paradoxical. On one hand, it is generative. This generative quality has benefited contemporary research by enabling researchers to generate ideas and enhance research opportunities while saving time and costs.

On the other hand, the generative nature of AI appears inevitably to lead to antagonism, resulting in entropy through model collapse and becoming degenerative. In this article, we explore the extent to which the implicitly degenerative nature of AI could be regarded as the main long-term threat to research integrity (RI), as many other problems associated with the impact of AI on RI may be seen as its effects.

In the first part, we provide an overview of the impact of AI on RI, including AI ethics, the use of AI in education (AIED), AI as a “grey area” or questionable research practice (QRP), the implementation of principles for AI use in codes of conduct, and the attitudes of academic publishers and universities towards AI. In the second part, we examine how the collapse of generative AI into degenerative AI poses a critical threat to RI in the future.

We emphasise that the only way to prevent the harmful effects of the degenerative nature of AI on RI is to retain the original human-generated datasets as the basis for AI systems and continually add new human-generated datasets.

One of the key principles regarding the impact of AI on RI is therefore the responsibility to ensure that AI remains grounded in human-created reality. This, however, leads us to the sociotechnical perspective on degenerative AI, which we address in the third part, where we evaluate the broader social and moral impact of degenerative AI.

We stress a fundamental shift in human trust requirements towards society and make a plea for more inclusive anticipatory risk management of AI with respect to RI.

DOI : https://doi.org/10.1057/s41599-026-08248-y
Catégories
EN

Modernizing Legal Scholarship: Toward Open Access Compliance

Authors : Ana Rogers-Butterworth, Melissa Moreau

Introduction and Literature Review: Legal research often operates outside conventional academic scholarship standards, characterized by a proliferation of student-edited journals and a notable absence of rigorous peer review. While some law journals have sought to align with established academic standards, many have struggled to keep pace with emerging open access (OA) requirements, such as those outlined by Plan S. As funding agencies increasingly mandate immediate OA for research outputs, the field of legal scholarly communications faces urgent needs for adaptation and modernization.

Methods: This study analyzed the OA policies of 384 journals that included articles and reviews authored by Canadian law faculty members. Data were extracted from Web of Science and Open Alex, focusing on six law faculties across Canada known for their high research output. Quantitative methods were used to assess publishing policies concerning OA principles.

Results: The findings reveal a strong preference for hybrid OA journals, particularly those with an international focus, often produced by interdisciplinary publishers. Diamond OA journals, primarily centered in North America, ranked second. Notably, a significant number of diamond OA journals fail to meet established OA standards, alongside a considerable presence of closed-access law journals.

Discussion and Conclusion: A consistent theme among law-specific publications, whether from academic faculty or corporate law publishers, is a pervasive lack of compliance with OA standards and a limited understanding of their implications. This underscores the imperative for further education and policy reform within the legal publishing ecosystem to enhance access and uphold the principles of open scholarship.

URL : Modernizing Legal Scholarship: Toward Open Access Compliance

DOI : https://doi.org/%2010.31274/jlsc.20259

Catégories
EN

Data Availability Statements in Mega Journals: A Comparative Analysis of Global and Korea-Affiliated Publications in Health and Medical Research

Authors : Sanghee Oh, Yunseo Park, Seyun Sim

Data Availability Statements (DAS) have become a standard mechanism for promoting transparency and reproducibility in open-access mega journals, yet questions remain about how effectively they support meaningful data sharing in practice and how these practices vary across national research contexts.

This study examines data-sharing practices in the medical and health sciences through a comparative analysis of global publications and Korea-affiliated articles in three mega journals indexed in PubMed Central: PLOS ONE, Scientific Reports and BMJ Open (2020–2024). DAS from 176,145 articles were collected from PubMed Central using an automated pipeline with manual validation and classified into a seven-category typology reflecting levels of data accessibility and reuse.

Results indicate that although DAS inclusion increased over time, repository-based data sharing remains limited, while ‘data available upon request’ continues to be prevalent. Clear differences are observed across journals: PLOS ONE shows greater use of repository-based and in-article sharing, whereas Scientific Reports and BMJ Open rely more heavily on ‘data available upon request’. Korea-affiliated articles largely follow global trends, with slightly greater reliance on national public data repositories.

Repository use is concentrated among a small number of international multidisciplinary platforms and selected national biomedical databases. The findings reveal a persistent gap between formal DAS compliance and effective data accessibility, indicating the need for clearer, more actionable data-sharing guidance.

URL : Data Availability Statements in Mega Journals: A Comparative Analysis of Global and Korea-Affiliated Publications in Health and Medical Research

DOI : https://doi.org/10.1002/leap.2083

Catégories
EN

The regulatory ethos in science

Authors : Nicole C. Nelson, Lara Keuck

This paper introduces the concept of the regulatory ethos to describe some common values and ideals that underscore the close connection between validation and regulation that this issue of BJHS Themes explores. We identify the primary motivation for this ethos as making knowledge production processes traceable, and define the regulatory ethos as valuing plans over situated actions, uniformity over heterogeneity, auditing over communication, and validation over validity.

Standard operating procedures, reporting checklists, preregistrations, compliance rules and monitoring are key means through which this ethos is enacted. While regulators have been instrumental in promulgating this ethos, it is not confined to the regulatory sphere.

We argue that reforms aimed at enhancing rigour and reproducibility are an example of how the practices and values associated with regulatory science have diffused out into academic science. Identifying this ethos as regulatory in origin – rather than wholly new or as part of a broader process of modernization – allows us to see that alternatives are not unscientific per se, and better identify the strengths and weaknesses of the regulatory ethos of science, such as the risk that data produced through these procedures will be replicable, statistically rigorous, and transparent, but not meaningful.

URL : The regulatory ethos in science

DOI : https://doi.org/10.1017/bjt.2026.10041

Catégories
EN

The Role of Preprints in Neuroscience Scholarly Communication: A Citation Analysis

Authors : Behrooz Rasuli, Fatemeh Seyfzadeh, Aurore Thibaut, Olivia Gosseries

Preprints, scientific manuscripts publicly shared prior to peer-review, are now part of scholarly 25 communication as emerging information resources. While neuroscience researchers have increasingly 26 published preprints, the impact of preprints in this field remains unclear.

Through a bibliometric 27 approach, this case study explored preprint citation patterns. Results yielded over 33,000 citations to 28 preprints within Scopus-indexed neuroscience documents (1993-2022). Trends of citations and 29 citation motivations were investigated. Findings indicated that 1.62% of neuroscience publications 30 cited at least one preprint, with citations peaking at 6% in 2021.

Review and journal articles cited 31 preprints more frequently, compared to books, notes, and conference papers (X² = 1909.015, p < 32 0.001). The most commonly cited servers were bioRxiv, arXiv, medRxiv, and PsyArXiv. Regarding 33 journals, a moderate positive correlation (rs = 0.353, p < 0.01) was found between publications citing 34 preprints and journals’ CiteScores.

Using Scite.ai, 93% of citations were classified as ‘mentioning,’ 35 with considerably fewer being supporting or contrasting. Most preprint citations appeared in 36 Introduction and Discussion, highlighting their role in framing research questions and contextualizing 37 results.

The global overview of these results may help contextualize citation behavior in relation to 38 structural and cultural factors, such as disciplinary norms, policy frameworks, researchers’ attitudes, 39 and health emergencies.

URL : The Role of Preprints in Neuroscience Scholarly Communication: A Citation Analysis

DOI : https://doi.org/10.1162/QSS.a.490

Catégories
EN

Responsible research evaluation: integrating quality, leadership, and integrity in national systems. The case of Peru

Authors : Pablo Alejandro Millones-Gomez, Jean Paul Simon Castillo-Nunez, Kenyie Jossuet Paucca-Calla, Carlos Alberto Minchon-Medina, Vanessa Lizet Castro-Delgado, Samantha Sotelo-Llancari, Cecilia Ignacio-Punin, David Yeret Rodríguez-Salazar, Víctor Hugo Urrutia-Baca

Introduction:

National research evaluation systems often rely on publication-based metrics that equate productivity with performance while overlooking scientific leadership and research integrity. This study examines the Peruvian National Registry of Science, Technology, and Innovation (RENACYT) to inform a more multidimensional framework for research evaluation.

Materials and methods:

An observational, non-experimental, and analytical study was conducted using data from RENACYT, Scopus, and SciVal for 9,651 researchers during 2019–2024. Four dimensions were assessed across hierarchical levels: scientific production, journal-based impact (Q1–Q4), corresponding authorship as a proxy of leadership, and retractions as indicators of research integrity. Descriptive statistics, ANOVA, repeated-measures tests, and count regression models (Poisson, negative binomial, and zero-inflated specifications) were applied.

Results:

A total of 92,284 publications were identified. Productivity increased across RENACYT levels (3.9 publications in Level VII vs. 62.5 in Distinguished; F = 1,162.572, p < 0.001), although with substantial within-level dispersion and differentiated temporal trajectories (Time × Level: F = 44.662, p < 0.001). Higher levels concentrated Q1 output (30.8 vs. 0.7 articles per author; F = 1,090.183, p < 0.001), while differences became less pronounced in Q3–Q4 journals. Corresponding authorship increased with level (β = 1.624 for Level I, p < 0.001) but remained heterogeneous even among top categories. Retractions were positively associated with productivity (coef. = 0.013, p < 0.001) reflecting differential exposure to integrity-related risks rather than uniform patterns across levels.

Conclusion:

RENACYT captures gradients in productivity and quality but insufficiently differentiates leadership and integrity. These findings support the proposal of a hybrid evaluation framework integrating productivity with explicit recognition of intellectual leadership and research integrity.

URL : Responsible research evaluation: integrating quality, leadership, and integrity in national systems. The case of Peru

Original location : https://www.frontiersin.org/journals/research-metrics-and-analytics/articles/10.3389/frma.2026.1842222/full

Catégories
EN

Faculty Perspectives on Open Access: A Multidimensional Analysis of Open-Access Usage and Library Relationships at an R2 Institution

Authors : James E. Van Loon, Julia E. Rodriguez

Introduction: The focus of this study is the broad impact open-access (OA) publishing has on all aspects of faculty members’ work (including publishing habits, research activities, and teaching) and on faculty relationships with the library.

Methods: An online survey addressing the impact of OA publishing on faculty and on their relationship with the library was administered to full-time tenure-track faculty at an R2 institution. Responses were analyzed by disciplinary unit, and themes were developed from faculty qualitative responses.

Results: A majority of faculty respondents had experience with OA publishing, with payment of article processing charges (APCs) to OA publishers, and with the use of OA materials in research and teaching. Many faculty respondents reported a substantial impact of OA on their research, teaching, and authorship, but expressed concern with the funding of OA publishing fees, with the quality and reputation of OA journals, and with the ethics of the OA business model.

Discussion: Faculty respondents demonstrate differing levels of familiarity with OA materials in all three roles. Overall, they are more receptive to OA materials as readers and researchers and more demanding about the quality, impact, and cost of OA publishing as authors.

Conclusion: Despite recognition of the benefits of OA materials, faculty respondents express concern over the cost, quality, and ethics of OA publishing. Faculty respondents also indicate a desire for the institution to provide funding for OA publications and rely on the library for expertise in locating, evaluating, and utilizing OA publications.

URL : Faculty Perspectives on Open Access: A Multidimensional Analysis of Open-Access Usage and Library Relationships at an R2 Institution

DOI : https://doi.org/10.31274/jlsc.20328