Transcript Data sharing and the ethics of consent

Openness, respect,
and participation
Consent in a time of transitions
Hallvard Fossheim
University of Bergen
Nordic Open Science and Research Forum 2016
Helsinki, 22 November
1
Good tidings
• Data sharing:
• Rational
• Reasonable (fair)
• Resource efficient
• Data sharing: not only an option, but an obligation
2
Good for research, good for us
• CUDOS
•
•
•
•
•
Communalism
Universalism
Disinterestedness
Organized
Skepticism
• Constructively critical openness
• Data sharing/re-use (vs needless, costly duplication and vs FFP)
• Validating existing research
• Doing non-/complementary research
• All the stuff in between
•
•
•
Nature’s «Scientific Data» initiative (http://www.nature.com/sdata/)
EU Guidelines to Open Access to Scientific Publications and Research Data in Horizon 2020 and Article
29.3 of the Model Grant Agreement
www.re3data.org registry over data repositories
3
Basic ethical principles
• Good consequences
• Respect for persons
• Justice
4
A challenge
• Social science: data about people
• anonymized/de-identified/personal // sensitive
• Big Data
• Size
• Approach
• Technology
• Data sharing: data life cycle vs project life cycle
• Sharing/re-use/handling
• Collecting, curating, mining, combining (Ohm)
• vs minimality? vs specificity?
• Relevant differences
• In data (surveys, public records, corporate consumer data)
• In agents (public offices, research institutions, NSOs, corporations)
• But for each, the whole data economy must be considered
5
Good consequences
• Data with a past and a future: define responsibilities & periodically evaluate risks
• Risks to privacy (Nissenbaum: private/public agency; spheres of privacy; private information)
• The ‘five safes’ in data handling (UK Data Service)
• Safe people
• vetted for ability to work appropriately with data
• Safe projects
• evaluation within an ethical framework
• Safe data
• minimized risk of disclosure
• Safe environment
• access in securely assured locations
• Safe outputs
• checked for disclosure potential
6
Respect for persons
• Consent
•
•
Informed
Voluntary
•
Unnecessary? Wrong answer—respect not reducible to safety
•
Impractical? Wrong answer—importance and quality of research first
•
Unworkable? Wrong answer—reflects mainly on researchers’ shortcomings
•
•
•
Methods; risks
Not for safeguarding researchers or institutions (1 month/year)
Impossible? Wrong answer—technology already there
•
•
Naive? Wrong answer—higher level
•
•
•
•
•
Never been easier
Cf data life cycle
Cf initiatives like Sandy Pentland’s New deal on Data
Cf rulings like EU Court, Grand Chamber, 14 May, 2014 (the right to be forgotten)
Analogy: Open Access (another prong of Open Science); requires something from
• research institutions
• funders
• journals
• governments/policy makers
Matters to persons, research, society, and political legitimacy
7
Trust
• A partial fix
• Between consent and mere safety
• Not unquestioning, but reasoned
• OECD Expert Group for International Collaboration on Microdata
Access 2014 (trust; method of access; legal relationship)
8
Summing up
• Data sharing (including re-use): YES
• Challenge: re-identification
• Good consequences  Five safes or similar frameworks
• Challenge: consent deficit = respect deficit
• No degree of data security can replace respect for persons
• Risk only shows more dramatically what is always true
• We should not let opportunity lead to lessening of respect
9
Thanks for your attention
[email protected]
10