Knowledge Exchange: researchers say what they would like to share - and how

While most researchers appreciate the benefits of sharing research data, they may be reluctant or unsure of how to share their own data. A study investigating the issues, based on interviews with 22 researchers in five research teams that have established data sharing cultures, has been published by the Knowledge Exchange, a body that exists to promote open sharing of data.

The five case studies in Denmark, Finland, Germany, the Netherlands and the UK, span arts and humanities, social sciences, biomedicine, chemistry and biology.

The study shows that when researchers talk about ‘data sharing’ they are referring to a variety of different ways in which research data are exchanged with other researchers. These are: private management sharing; collaborative sharing; peer exchange; sharing for transparent governance; community sharing; and public sharing.

The motivation for researchers to share research data are:

1. when data sharing is an essential part of the research process;
2. direct career benefits derived from sharing through greater visibility of one’s work, reciprocal data exchanges, and the reassurance of having one’s data recognised as valuable by others;
3. the norms that researchers are exposed to within their research circle or discipline;
4. (4) a framework of funder and publisher expectations, policies, infrastructure and data services as drivers.

Incentives vary across disciplines and over time

While the norms in specific disciplines are shown to be important, equally striking was the variation found within disciplines.

Incentives to share also vary across a researcher’s career. The strong influence of data sharing norms implies a key role for early training on research data sharing, as an integral part of research methods training.

Researchers’ experiences, data sharing practices and motivations were shown to be heterogeneous across the research groups and disciplines studied.

This points to the need to create a level playing field for all researchers to share data and to change the collective attitude towards sharing. Embedding data sharing training in research methods training is crucial for data sharing to become standard research practice.

There is also a vital role to be played by formal data policies - as long as they do not interfere with informal sharing and are sensitive to variations across disciplines.

Policies work at two levels, providing a collective voice and helping to clarify and change the norms of the research community. Instituting data sharing policies also avoids a mismatch of incentives: they create a positive motivation to share that benefits science in general, even in those cases where the direct benefits to individual researchers are weaker.

The report, ‘Sowing the seed: Incentives and motivations for sharing research data: a researcher’s perspective’ makes recommendations aimed at incentivising increased data sharing by different groups.

Amongst the recommendations, the study says research funders:

• Should adopt a data sharing policy that clearly indicates expectations for data accessibility, to provide a level playing field for all funded researchers. Policies may cover measures such as requirements for data management planning and clear guidance much of a grant budget can be allocated to data sharing;
• Provide funding and support services to researchers, for example, for data documentation, annotation and data deposit. This should be similar to the funding available for publishing in journals. The study notes that different disciplines have different needs in this respect, as some types of data require more preparation to make them available for reuse. In addition,  the type of data being shared influences what is needed, for example, sharing raw versus processed data; sharing data that supplements articles versus sharing in repositories;
• The focus on data sharing funding should be directed to two key time points: when research is being planned and upon completion of a research project, to prepare data and documentation for curation;
• Should continue to invest in data infrastructure that also provides rich context, detailed metadata and an account of the data creation. The kind of infrastructure researchers find most useful is where research data, papers and other outputs or resources are jointly available within a single data resource;
• Data sharing training should be embedded into research methods training for students and doctoral researchers, to help establish data sharing as standard research methodology and practice;
• Promote re-use of existing data resources via specific funding streams for secondary analysis and by setting expectations for research grant applicants to justify the need to create new data in research, that is, to demonstrate that existing data cannot address their research questions;
• Engage with publishers and commercial partners on IP and copyright of data that may limit data sharing by creating a working group to find ways to protect IP and share data, especially when research is intended for non-commercial use;
• Provide guidance to peer reviewers to evaluate data sharing plans and strategies in research proposals

Knowledge Exchange promotes the use and development of ICT infrastructure for higher education and research with the aim of making a layer of scholarly and scientific content openly available on the Internet. The partners in the Exchange are CSC - IT Centre for Science in Finland, Denmark’s Electronic Research Library, the German Research Foundation (DFG), Jisc in the UK and SURF in the Netherlands.

Access the report here: http://www.knowledge-exchange.info/Default.aspx?ID=733