A new study from IOP Publishing (IOPP) reveals significant differences in the adoption of open data sharing practices among physical science research communities, with barriers varying across disciplines. The study, which analyzed over 30,000 research articles, was detailed in the white paper "Bringing Researchers on Board: Navigating the Barriers to Sharing Data Publicly."
The study shows that while many researchers are open to sharing their data, adherence to the Findability, Accessibility, Interoperability, and Reusability (FAIR) principles remains low, with only 10% of physical science researchers meeting these standards alongside their published articles.
Key findings from the study include:
Environmental scientists lead in open data sharing, with over 80% of researchers openly sharing their data. However, legal constraints, often related to third-party ownership, hinder their ability to fully comply with the FAIR principles, with just under 60% adhering to them.
Physicists are similarly open to sharing data (over 70%), but only 18% of them follow FAIR principles. The primary barrier is the accessibility and understanding of the data formats used, making even available data difficult to use.
Engineering scientists face greater obstacles, with only 55% sharing their data and less than 8% complying with FAIR principles. Confidentiality concerns and the lack of known repositories to submit their data are significant challenges.
Materials scientists also encounter barriers, with over 70% sharing their data but only 5% adhering to FAIR principles, largely due to the confidential or sensitive nature of their data.
IOPP’s white paper emphasizes that open data sharing enables replication of research and strengthens trust in scientific results. The FAIR principles, introduced in 2016, aim to standardize metadata and make research data more discoverable, accessible, and reusable with proper attribution.
Since 2022, IOPP has required all authors to include a data availability statement in their publications, outlining whether and how their research data can be accessed. In 2023, this policy was extended to require authors who cannot share data publicly to provide reasons.
In addition to these policies, IOPP is introducing new content types to support open data practices, such as its open access Machine Learning journal series, which includes dataset, benchmark, and challenge articles. These efforts aim to promote transparency and advance the adoption of open data sharing across the scientific community.
Click here to read the original press release.
