Open, FAIR data and RDM
Download as PPTX, PDF6 likes1,889 views
This document summarizes key findings from surveys about researchers' data sharing practices and attitudes. It finds that while most researchers agree data should be shared, only a small percentage actually make their data openly available. Researchers typically share data through email, cloud services, or external drives rather than repositories. The document also discusses increasing emphasis on open and FAIR data in research funder policies, but notes researchers face barriers to compliance like unclear terminology, lack of skills and incentives, and confusion between open data and managed/FAIR data. It argues for engagement programs to help researchers better understand and participate in open scholarship.
Open, FAIR data and RDM
- 1. Open data, FAIR data and RDM: the ugly duckling Sarah Jones Digital Curation Centre, Glasgow sarah.jones@glasgow.ac.uk Twitter: @sjDCC Open Science Conference, Berlin, 13-14 March 2018 #osc2018 unless otherwise stated Slides available at: https://doi.org/10.5281/zenodo.1196631
- 2. ‘Questions’ CC-BY by Derek Bridges www.flickr.com/photos/derek_b/3046770021
- 3. How many researchers make data open? 79% of researchers have made data openly available The State of Open Data 2017 Digital Science 2300 respondents worldwide only 1 in 10 provides their research data as open data for the public Researchers and their data (2015) eInfrastructures Austria 3026 Austrian respondents 68% of researchers already share data or expect to do so in future Jisc DAF studies (2016) 1185 UK respondents 64% agree that they are willing to share their data Open Data: the researcher perspective (2017), Elsevier 1162 respondents worldwide 29% 32% 18% 21%
- 4. How do researchers share data? Less than 15% publish data in a repository. Elsevier: Open Data - the researcher perspective Over half only allow access on request. 54% share data by using external storage devices or email. eInfra Austria: Researchers and their data Of 13 methods stated, top 4 options for currently sharing data were: 1. Emailing data files (65%) 2. Cloud service e.g. Dropbox, Googledrive (59%) 3. Portable storage (35%) 4. Supplementary data (20%) Formal repository (public / institutional) c.12% Jisc DAF studies “When asked where they have published data, most commonly respondents had done so as an appendix to an article (just over 30%) with a data repository close behind (just under 30%) and 20% having published in a data journal.” Digital Science: The State of Open Data
- 5. Why do researchers share data? Digital Science: The State of Open Data Jisc DAF studies “For more than half of the researchers, the most attractive incentives for sharing their data were increased visibility and impact, new cooperation opportunities, recognition in professional circles, as well as their contributions being regarded as scientific output.” eInfra Austria: researchers and their data
- 6. Data storage and loss 17% of respondents had lost data 36% had experienced loss and 83% of this was due to physical storage media Digital Science: The State of Open Data More than one-third had experienced data loss. Strong preference to store on business/private computer, external hard drive & usb eInfra Austria: researchers and their data Jisc DAF studies
- 7. Wellcome OA compliance rates
- 8. Sharing of microarray data Increase from c.5-35% in under a decade Best-practice guidelines for sharing microarray data are fairly mature Two centralized databases have emerged Unusually strong data sharing requirements in some journals Piwowar, H. (2011) Who Shares? Who Doesn't? Factors Associated with Openly Archiving Raw Research Data. PLOS One https://doi.org/10.1371/journal.pone.0018657
- 9. ‘Confusion’ CC-BY-NC-ND by Allan https://www.flickr.com/photos/trekker308/5208629587
- 10. Data policy changes Emphasis on data sharing more than RDM Increasingly ‘open’ and ‘FAIR’ rhetoric 2002 (handbook) • General issues relating to data • Management responsibilities for data within NERC • Planning for the management of data • Access to, and charges for, NERC’s data • The implications for scientists holding data 2010 • Data acquisition • Data management • Access and use • Charging for Access to NERC's Data 2016 • Access to data • NERC’s environmental data centres • Data collection • Open access to data underpinning research publications www.nerc.ac.uk/research/sites/data/policy
- 11. Forerunners to FAIR OECD Principles and Guidelines for Access to Research Data from Public Funding (2007) A. Openness B. Flexibility C. Transparency D. Legal conformity E. Protection of IP F. Formal responsibility G. Professionalism H. Interoperability I. Quality J. Security K. Efficiency L. Accountability M. Sustainability Science as an Open Enterprise (2012) notion of ‘intelligent openness’ where data are accessible, intelligible, assessable and useable “Open scientific research data should be easily discoverable, accessible, assessable, intelligible, useable, and wherever possible interoperable to specific quality standards.” G8 Science Ministers Statement (2013)
- 12. Good understanding of FAIR, but… “We understand the basic principle of FAIR, but the terminology is often difficult to grasp immediately. Things could be explained better in plain language” “The term interoperable is quite confusing sometimes and mixed with re-use.” “I could do with help understanding the section on Making data interoperable as I don't understand a number of the terms and concepts.” Table from Q4, comments from Q5 To what extent do the following statements represent your experience of using the H2020 template? Agree Neither agree nor disagree Disagree I don’t understand what FAIR means 10% 17 16% 28 74% 125 Grootveld et al. (2018). OpenAIRE and FAIR Data Expert Group survey about Horizon 2020 template for Data Management Plans http://doi.org/10.5281/zenodo.1120245
- 13. Language is a barrier Respondents mentioned 40 terms which were unclear to them “Researchers are not familiar with the following terms/phrases : Metadata, standards for metadata/data, ontologies, mapping with ontologies, interoperability, ... . All the ICT jargon” “With the help from Swedish National Data Service we could clarify many questions. Without this help we would not be able to finish the DMP.” Grootveld et al. (2018). OpenAIRE and FAIR Data Expert Group survey about Horizon 2020 template for Data Management Plans http://doi.org/10.5281/zenodo.1120245
- 14. Conflation of FAIR and open Making data FAIR ensures it can be found, understood and reused Data can be shared under restrictions & still be FAIR Open data is a subset of all the data shared "As open as possible, as closed as necessary" Image CC-BY-SA by SangyaPundir
- 15. Confusion in DMPs Overly broad definitions of data, including publications, presentations, meeting minutes, dissemination materials, digital photos, project website… Talk of making data available by gold or green open access Blurring of methods to store and share data within consortium versus long-term preservation e.g. backup to googledocs, use Dropbox to give public access…
- 16. Be careful what you say… My data are sensitive We’ve signed a non- disclosure agreement with our commercial partners This doesn’t apply to me… I want to patent
- 17. CC-BY by SSG Robert Stewart https://www.flickr.com/photos/familymwr/4930276154
- 18. How do Open, FAIR & RDM intersect? Open FAIR data Managed data Internal Self-interest External Community benefit
- 19. Open data and FAIR data Managed data FAIR data Open data
- 20. Open data, FAIR data & RDM Managed data FAIR data Open data
- 21. All research data Managed data FAIR data Open data the wild
- 22. Increasing that which is FAIR & open Managed data FAIR data Open data the wild
- 23. OS advocacy Better science Greater impact Mandates € RDM issues Too big to email… Dropbox? Deadlines paperwork PRESSURE Not enough storage How does this even help me or my career?
- 24. Data engagement programmes Data champions at Cambridge & data stewards at TU Delft Local support & help Researcher-focused & led Explicitly recognise value and role of curation https://osc.cam.ac.uk/engaging -researchers-good-data- management Data conversations at Lancaster University Provide a forum for researchers to speak about their data Engage the non-converted Use peers to spread RDM / OS message www.lancaster.ac.uk/library/ rdm/data-conversations
- 25. Awareness of OS & initiatives European Commission (OSPP) Open Science Policy Platform. (2017) Providing researchers with the skills and competencies they need to practise Open Science. Report of the Working Group on Education and Skills under Open Science, doi: 10.2777/121253
- 26. ORD pilot & FAIR data ORD Pilot • Introduction of an Open Research Data pilot in 2014 • Expansion from 7 to 9 work areas in 2016 • ‘Open data by default’ since 2017. Need to actively opt out. FAIR data • New ‘FAIR Data Management guidelines’ in July 2016 • Increasing emphasis on data management as well as sharing • Mantra of “As open as possible, as closed as necessary” ? • Formal policy in FP9… • Even greater emphasis on research data management… • Mandatory DMP, even in cases of opt out… http://ec.europa.eu/research/participants/data/ref/h2020/grants_manual/ hi/oa_pilot/h2020-hi-oa-data-mgt_en.pdf
- 27. Manage data so it can flourish
- 28. Thanks for listening DCC resources on Data Management www.dcc.ac.uk/resources Follow us on twitter: @digitalcuration and #ukdcc Icons on slides 6, 16, 24 & 27 are copyrighted and used under licence
Editor's Notes
- #3: Do you publish open access Do you make your data open? Are your data FAIR?
- #11: Handbook covers issues like data ownership, data loss, minimum standards for stewardship and charging for data. By 2016, openness is a key principle. Data considered a public good and made freely available except in a few cases.
- #12: OECD – 13 principles e.g. openness, flexible, transparent, legal, interoperable, quality, secure, accountable, efficient… OECD preconditions: ‘data must be accessible and readily located; they must be intelligible to those who wish to scrutinise them; data must be assessable so that judgments can be made about their reliability and the competence of those who created them; and they must be usable by others.’ G8 statement adopted verbatim in the European Commission’s first data guidelines for the Horizon 2020 framework programme later the same year.