Workflow for detecting biomedical articles with underlying open and restricted-access datasets DOI Creative Commons
Anastasiia Iarkaeva, Vladislav Nachev, Evgeny Bobrov

и другие.

PLoS ONE, Год журнала: 2024, Номер 19(5), С. e0302787 - e0302787

Опубликована: Май 8, 2024

To monitor the sharing of research data through repositories is increasingly interest to institutions and funders, as well from a meta-research perspective. Automated screening tools exist, but they are based on either narrow or vague definitions open data. Where manual validation has been performed, it was small article sample. At our biomedical institution, we developed detailed criteria for such screening, workflow which combines an automated step, considers both fully restricted-access We use results internal incentivization scheme, monitoring in dashboard. Here, describe detail procedure its validation, 11035 articles, 1381 articles with potential were subsequently screened manually. The highly reliable, witnessed by inter-rater reliability values ≥0.8 (Krippendorff’s alpha) two different samples. also report institution independent sample study. In largest three samples, 2021 institutional sample, underlying had openly shared 7.8% articles. For additional 1.0% shared, resulting 8.3% overall having and/or extraction then discussed regard applicability contexts, limitations, possible variations, future developments. summary, present comprehensive, validated, semi-automated detection publications.

Язык: Английский

It takes two flints to start a fire: A focus group study into PhD supervision for responsible research DOI Creative Commons
Tamarinde Haven

Accountability in Research, Год журнала: 2025, Номер unknown, С. 1 - 24

Опубликована: Янв. 28, 2025

Background Supervision is one important means of promoting responsible research. However, what a supervisor should do and how to foster supervisory climate unclear.

Язык: Английский

Процитировано

1

Why does cardiology have many extreme publishing authors? DOI Creative Commons
John P. A. Ioannidis

Hellenic Journal of Cardiology, Год журнала: 2025, Номер unknown

Опубликована: Янв. 1, 2025

Under diverse contributing factors in different scientific micro-environments, the number of authors who publish extreme numbers full articles a single year has increased. Cardiology is subfield that largest share with publishing behavior than any other science (outside physics). Between 2000 and 2022, 137 Cardiovascular System (CVS, Science-Metrix classification) have published over 60 at least one calendar are also highly-cited. The majority (70/137) from Europe. All 7 countries highest prevalence CVS per million population European countries. Issues massive authorship papers by administrative leaders discussed, including arguments favor sustaining this practice - refutation these arguments. Other major contributors to phenomenon publications clinical trials epidemiological studies highly-cited guidelines. Micro-environments instrumental creating both developed less Listing contributions does not solve problem since gamed; metrics probe gaming nevertheless available. Eventually, carries credit accountability. Number metric can be heavily gamed. Emphasis should given what makes impact on human lives.

Язык: Английский

Процитировано

0

Leading by example: how to empower supervisors as role models DOI Creative Commons

Miriam van Loon,

Joeri K. Tijdink, Natalie Evans

и другие.

Frontiers in Research Metrics and Analytics, Год журнала: 2025, Номер 10

Опубликована: Апрель 4, 2025

Supervisors are considered to play a pivotal role in stimulating responsible conduct of research (RCR). Their position as supervisors PhD candidates offers the opportunity be good models and show young researchers how properly. In this contribution, we delineate what it means “lead by example.” We inquire concept modeling is currently applied context supervision general, RCR specifically, present perspective empowerment fruitful approach help determine should focus on when aiming foster positive culture.

Язык: Английский

Процитировано

0

Open science interventions to improve reproducibility and replicability of research: a scoping review DOI Creative Commons
Leonie Dudda, Eva Kormann, Magdalena Kozula

и другие.

Royal Society Open Science, Год журнала: 2025, Номер 12(4)

Опубликована: Апрель 1, 2025

Various open science practices have been proposed to improve the reproducibility and replicability of scientific research, but not for all practices, there may be evidence they are indeed effective. Therefore, we conducted a scoping review literature on interventions reproducibility. We systematically searched Medline , Embase Web Science PsycINFO Scopus Eric 18 August 2023. Any study empirically evaluating effectiveness aimed at improving or methods findings was included. summarized retrieved narratively in gap maps. Of 105 distinct studies included, 15 directly measured effect an intervention replicability, while remainder addressed proxy outcome that might expected increase such as data sharing, transparency pre-registration. Thirty were non-comparative 27 comparative cross-sectional observational designs, precluding any causal inference. Despite investigating range addressing various outcomes, our indicate general base which research remains remarkably limited many respects.

Язык: Английский

Процитировано

0

Open Science in Three Acts: Foundations, Practice, and Implementation - First Act DOI Creative Commons
Ricardo Limongi, Pablo Rogers

BAR - Brazilian Administration Review, Год журнала: 2025, Номер 22(1)

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0

Workflow for detecting biomedical articles with underlying open and restricted-access datasets DOI Open Access
Anastasiia Iarkaeva, Vladislav Nachev, Evgeny Bobrov

и другие.

Опубликована: Июль 27, 2023

To monitor the sharing of research data through repositories is increasingly interest to institutions and funders, as well from a meta-research perspective. Automated screening tools exist, but they are based on either narrow or vague definitions open data. Where manual validation has been performed, it was small article sample. At our biomedical institution, we developed detailed criteria for such screening, workflow which combines an automated step, considers both fully restricted-access We use results internal incentivization scheme, monitoring in dashboard. Here, describe detail procedure its validation, 11035 articles, 1381 articles with potential were subsequently screened manually. The highly reliable, witnessed by inter-rater reliability values ≥0.8 (Krippendorff’s alpha) two different samples. also report institution independent sample study. In largest three samples, 2021 institutional sample, underlying had openly shared 7.8% articles. For additional 1.0% shared, resulting 8.3% overall having and/or extraction then discussed regard applicability contexts, limitations, possible variations, future developments. summary, present comprehensive, validated, semi-automated detection publications.

Язык: Английский

Процитировано

3

Workflow for detecting biomedical articles with underlying open and restricted-access datasets DOI Creative Commons
Anastasiia Iarkaeva, Vladislav Nachev, Evgeny Bobrov

и другие.

PLoS ONE, Год журнала: 2024, Номер 19(5), С. e0302787 - e0302787

Опубликована: Май 8, 2024

To monitor the sharing of research data through repositories is increasingly interest to institutions and funders, as well from a meta-research perspective. Automated screening tools exist, but they are based on either narrow or vague definitions open data. Where manual validation has been performed, it was small article sample. At our biomedical institution, we developed detailed criteria for such screening, workflow which combines an automated step, considers both fully restricted-access We use results internal incentivization scheme, monitoring in dashboard. Here, describe detail procedure its validation, 11035 articles, 1381 articles with potential were subsequently screened manually. The highly reliable, witnessed by inter-rater reliability values ≥0.8 (Krippendorff’s alpha) two different samples. also report institution independent sample study. In largest three samples, 2021 institutional sample, underlying had openly shared 7.8% articles. For additional 1.0% shared, resulting 8.3% overall having and/or extraction then discussed regard applicability contexts, limitations, possible variations, future developments. summary, present comprehensive, validated, semi-automated detection publications.

Язык: Английский

Процитировано

0