What a difference a research field makes! 📢 new paper on high rates of #OpenData sharing in #Biology at #EdinburghUni, led by Haya Deeb from our BioRDM team with 5 student researchers 👩🎓.
Previously on her #Metascience2025 poster. 🧵 1/3
https://doi.org/10.1371/journal.pone.0328065
@toothFAIRy @fresseng
Biology is the most data sharing friendly discipline.
The figures are note the same, but I guess it is to be expected since the methodology is different.
We found the sensitivity of ODDPub to Open data was around 50% compared to manual scoring [modulo many differences]. If DataStet is similar, then 30% sharing in French 'Fundamental Biology' articles in 2023 might correspond to ~60% by our manual approach.
I would prefer an institutional/regional/national resource! I know UKRN is evaluating several options.
📈 Bioscience researchers shared data in 92% of articles that we manually evaluated from 2023. In the chart 👇 orange shading shows 45% of articles shared ALL the relevant data, up from 7% in 2014👏. Sharing varied by data type as expected, 🧬 vs. 🔬, among several other factors.
Thanks to BIH QUEST @ChariteBerlin for ODDPub, which gave a parallel, programmatic evaluation.
In contrast, testing an international sample of circadian, neuroscience and mental health articles by the same manual method … 2/3
Edited for Alt-text.
… found only 8% of articles in 2023 shared ANY Open data, consistent with previous reports for psychiatry and psychology research.
In our bioscience sample that year, only 8% of articles shared no Open data. Many factors underlying this difference, of course.
https://doi.org/10.1136/bmjment-2024-301333
👇 CMHN is our analysis of articles from researchers in the MRC-funded "Circadian and Mental Health Network". 3/3
