Pվ��Ƶ

Author: Saskia Hoving

Editor-in-Chief

Open science is necessary to advance solutions to the world’s most pressing challenges, and data sharing is both a crucial element and a driving force to achieving the objectives of open science. But in general, current data sharing practices leave something to be desired. What are the challenges to data sharing and how can they be overcome? A panel discussion at the Charleston Conference delved into these questions and offered insights from diverse perspectives.  

In the evolving landscape of scientific research, data sharing and open science are becoming increasingly central. Data sharing is a fundamental part of open science but despite the considerable benefits for both researchers and research, substantial barriers to this practice persist – a topic that is explored in the longitudinal survey on researchers’ attitudes towards and experiences of open data, The State of Open Data. 

Researchers need support to enable them to openly share outputs from their work beyond traditional publications: preprints, data, code, protocols, methods, and more. However, this support alone won't suffice to encourage data sharing unless the evaluation system for researchers adapts to recognise and reward their sharing efforts.  

A recent panel discussion explored the practice of data sharing, the barriers researchers face, and the need for a fundamental change in research assessment to further support it. 

How to encourage and support data sharing: A panel discussion

The virtual panel session titled ‘Strategies for driving the transition to a culture of open science: What works, what doesn’t?’ was held at the 2024 Charleston Conference. Within the broader context of revising the parameters and culture of assessment criteria, the panel covered challenges to data sharing such as the need for relevant training and support, the requirements for technical solutions, and the importance of an incentive structure based on proper accreditation.

Panel moderator Laura Patton, Head of Government Affairs, US at Pվ��Ƶ, opened her remarks with results from the annual global survey The State of Open Data by Pվ��Ƶ, Digital Science, and Figshare. This survey, now its ninth year, provides insights into the attitudes of researchers towards open data. The main challenge to sharing data, the survey shows, is authors’ feeling that they are still receiving too little credit: 73% of respondents said that the work that goes into sharing data is not worth the credit they get for it. In a survey on research integrity (Pվ��Ƶ white paper forthcoming), Patton continued, researchers’ desire for training related to data sharing is clearly expressed, including use of repositories, storage, and data management strategies.

Jonathan Petters, Associate Director, Data Management and Curation Services at Virginia Tech University Libraries, stated that a change in culture is required to incentivise data sharing. For faculty members, there is little value in sharing data, software, or pre-registration of research plans. Against the traditional activities that count for tenure, these research products are fairly new, and the technical infrastructure to demonstrate their impact does not exist yet. Researchers themselves, Petters claims, do not consider these to be “first class research products” but this must change. 

Making it happen: Culture, recognition, and support for data sharing

The technical challenges associated with data sharing, according to Daniel Hook, CEO of Digital Science, are not the real barriers. The technology to share data, to give it a unique identifier and to make it available for others to use already exists, but the main challenges, Hook says, are entirely cultural.   

The curation of data records or demonstration of appropriate data handling are not factors in the evaluation system in a standard academic track. Recognising and valuing researchers' expertise and skills in data management are essential for fostering the cultural change needed to make data sharing a common practice.

In a fascinating insight into a unique success story, Erika Pastrana, Vice President, Nature Research Journals Portfolio at Pվ��Ƶ, illustrated the importance of a supportive, encouraging setting to achieve robust data and code sharing results. The journal Nature Computational Science set clear policies and structures in place to encourage and strongly advocate for code sharing and citing, and the team of editors individually addressed every paper to champion and support this. As a result, the journal reached an astonishing 100% code sharing compliance, with all published articles sharing their code publicly via a permanent identifier. While this is not easily scalable, Pastrana admitted, it does show that it is possible to get researchers to share their data, and perhaps this can inform and inspire behavioural change.   

Strategies for promoting data sharing and advancing open science

What would it take to support and encourage data sharing? From the panel discussion, several strategies emerged:

• Research assessment criteria must be reevaluated to include non-traditional research outputs such as contributions to data sharing or software development. This means promoting diverse research outputs beyond journal articles, which is in accordance with the recommendations of the Declaration on Research Assessment (DORA) (of which Pվ��Ƶ is signatory). Encouraging and supporting the publication of outputs such as data, software, protocols, or preprints means that the broad spectrum of research activities is given value and credit, and researchers are incentivised to share them as a valuable research product.
• Enhancing data linkages between publications, and other research outputs such as data, facilitates access to data and promotes its transparency and reuse. The use of persistent identifiers like DOI for data records and proper citation of them in related publications is an example of such linkage.
• Fostering collaboration and transparency to encourage researchers to share their data, methods, and findings. This requires a holistic approach that includes data sharing mandates, available platforms for data sharing that facilitate collaboration, and a cultural shift to advocating openness and collective research advancement.

From open access to data sharing: The road to open science

Similar to the transition to open access (OA), data sharing is driven by the alignment and collaboration of various actors in the scientific ecosystem, like publishers, funders, and libraries. In their support for the entire research lifecycle, academic libraries are key partners in the advancement of OA and open science. The role of libraries in facilitating research sharing and helping researchers to make their research and their data available is essential to advancing open science.

The necessary shifts in culture and priorities as well as the availability of technological solutions and training that were discussed in the panel are vital for advancing data sharing en route to open science. Raising awareness, allocating resources to funding and training, and creating incentives are also measures to promote and encourage data sharing. Above all, research assessment criteria need to change for data sharing to become a standard practice. Continued efforts from the various stakeholders are essential, and as in any such transition, collaboration and alignment are key. 

Here’s everything you need to know about open research and the library to be part of the transition to open science, from OA agreements to open data. 

Don't miss the latest news & blogs, subscribe to The Link Alerts!

Related content

• Making policy a reality – practical steps to open access workflows
• Facilitating the transition towards open access (OA) publishing
• Librarians in the age of open access: An evolving role
• Open access and the mission of the academic library
• Driving open access transformation: Insights from the Charleston Conference 2023
• Empowering librarians to promote open access
• Transitioning to open access: Reflections from three US organizations
• Open access agreements: Empowering librarians to be OA champions