To share or not to share: how researchers handle data

1. To share or not to share: how researchers handle data Michael Jubb RIN Fourth Bloomsbury Conference: Valued Resources 24 June 2010

2. Some definitions • Data • ‘evidence supporting research and scholarship’ • DCC Charter and Statement of Principles • ‘have no intrinsic meaning until converted into information through some process of analysis, interpretation and description: typically, the process by which experimental data become a research paper’ • Patterns of information use and exchange:case studies of researchers in the life sciences, RIN 2009

3. researchers’ perspectives • skills and training • funder and institutional perspectives and implications

4. 1. Researchers’ perspectives

5. data and information in the research process: some verbs • gather • evaluate • create • analyse • manage • transform • present • communicate • disseminate

6. But………. • gathering, creating, evaluating data not usually the primary object of research • few career rewards from sharing data • how it’s done in different disciplines varies greatly • big science not the norm

7. the research process: animal genetics

8. research process: transgenesis & embryology

9. research process: epidemiology

10. research process: neuroscience

11. why share? • completeness of the scholarly record • validation of results • re-use and integration • exploit what’s already been done • avoid duplication of effort • ask new research questions • researchers may have different interests as creators and users

12. evidence of benefits citation esteem and successful evaluations funder requirements altruism/reciprocity cultural/peer pressures enhanced visibility opportunities for collaboration, co-authorship easy-to-do no clear benefits/incentives competition; resistance to sharing intellectual capital desire for/fear of commercial exploitation access restrictions desired or imposed legal, ethical problems lack of time, funds, expertise practical and technical difficulties creators: motivations and constraints?

13. So do they do it?

14. ownership, protection and trust • responsibility, protectiveness and desire for control over data • concerns about inappropriate use • climate change data……….. • preference for co-operative arrangements and direct contact with potential users • decisions on when and how to share • commercial, ethical, legal issues • lack of trust in other researchers’ data • “I don’t know if they have done it to the same standards I would have done it” • lack of standardisation • intricacies of experimental design and processes

15. some conclusions…… • data are primary in the research process, but secondary as research outputs • data management, curation and sharing not yet embedded or the norm • genomics, bio-informatics, astronomy etc the exceptions • few career rewards from sharing data • resistance to open sharing of intellectual capital • real differences between researchers working in different disciplines and contexts • impact of funder policies? • impact of FoI?

16. 2. Training and skills “This has been identified in every study as a major problem, both training researchers to be e-researchers, and training the people running the systems to deal with researchers, and to understand the technology”

17. Researchers • “…a lot of scientists don’t get information and structures at all. It’s not what they’re trained to think about.” • “……..the idea of quality, provenance, and metadata about data is woefully inadequate in most science training.” • engagement between researchers and data management professionals • “…….now we manage our data, whereas before we didn’t” • issues of scalability in training and support to meet diverse needs of wide range of research groups

18. concern about low numbers of people with specialist expertise people from two kinds of backgrounds library/information professionals researchers need for co-ordination of effort and funding for capacity-building much depends at present on short-term project funding lack of career structure “Who’s training these people? We need training at the professional level for people who are actually going to run these data centres” “….the career structure for those people with expertise is miserable, because the number of places they can work is not large, and the universities don’t treat them as key staff” “So there’s a real danger of losing people to the private sector” Curators

19. some conclusions…… • how to promote cultural change • building capacity and capability among both researchers and information specialists • career paths and rewards • assessment of national requirement for skills in data curation and support

20. 3. Funder and institutional perspectives and implications

21. Policy drivers • increasing the efficiency of the research process • avoid duplication • promoting scholarly rigour and enhancing research quality • review and testing of data • enhance scope and quality of the scholarly record • enabling researchers to ask new questions • re-use of data • development of ‘data-intensive science’ • enhancing visibility of research • opening opportunities for engagement • “increasing the return on public investments in research” • OECD Principles and Guidelines for Access to Research Data 2007

22. Research Councils….. • BBSRC….. expects research data generated as a result of BBSRC support to be made available with as few restrictions as possible in a timely and responsible manner to the scientific community for subsequent research • recognises that different fields of study will require different approaches • expects that timely release would generally be no later than the release through publication of the main findings • supports the view that those enabling sharing should receive full and appropriate recognition by funders, their academic institutions and new users for promoting secondary research June 2010

23. Research Councils……. • NERC….requires that due consideration be given to the 'post project' stewardship of data prior to approval being given for a project • requires that recipients of NERC grants offer to deposit with NERC a copy of datasets resulting from the research supported, for research or other public good purposes, but without prejudice to the intellectual property rights • ensures that individual scientists, principal investigator teams and participants in programmes will be permitted a reasonable period to work exclusively on, and publish the results of, the data they have collected Updated Feb 2010

24. Wellcome Trust • …..considers that the benefits gained from research data will be maximised when they are made widely available to the research community as soon as feasible, so that they can be verified, built upon and used to advance knowledge. • ….expects the researchers that it funds to maximise the availability of research data with as few restrictions as possible • …..believes that data sharing for the benefit of the research community as a whole will only proceed if those using the data also adopt good research practice.…… [and] expects all users of data to acknowledge the sources of their data and abide by the terms and conditions under which they accessed [them]. 2007

25. But do the policies work? • funding and infrastructure • compliance and engagement with researchers

26. co-ordination between different funding bodies Research Councils, Higher Education funding bodies, JISC, universities clarity about roles and responsibilities piecemeal initiatives with limited take-up and impact infrastructure ‘driven by the science’? need for careful management of relationships between specialists and researchers disciplinary and institutional dimensions of scale and complexity dangers of “solutions looking for problems” “….we need a more co-ordinated strategy and real leadership to take things forward.” “…..it’s very easy in the current framework to pass the buck and do nothing.” “Things are funded in silos. So I just don’t think there is really a national strategy.” “…….you have to work pretty hard to demonstrate there’s a business case for reuse of data………there’s no point in paying to curate and store data if nobody ever does use it again.” building an infrastructure: leadership and co-ordination?

27. top-down and/or bottom-up: a real tension • bottom-up • develop policies and local services in response to what researchers themselves want • develop tools and environments within universities to equip the research community with appropriate processes and skills • top-down • national policy frameworks • national body/programme to catalyse change required for sustainable and ubiquitous service

28. Some policy and service implications……. • policies and services need to be informed (but not determined) by an understanding of the views and practices of researchers • different communities and contexts • single, one-size-fits-all approach won’t work • engagement with researchers • to identify and address constraints • to preserve exercise of informed choice • pragmatic and experimental policies • build on informal sharing already taking place • recognition of mutual needs • practicalities of sharing • what makes data intelligible and usable? • when is sharing useful enough to warrant the labour necessary to achieve it? • address barriers as well as drivers for change • incentives, self-interests and goals of researchers • sustaining of intellectual capital • professional recognition and reward structures

29. Some policy and service implications….. • the funding and sustainability challenge • sharing is not cost-free • co-operation needed between researchers, funders, institutions • complexities of the dual support system • benefits and evidence of value • are the benefits realised in practice? • does making it available mean that it’s used? • scope for publishers to promote sharing? • carrots as well as sticks?

30. Questions? Michael Jubb www.rin.ac.uk