Social scientists may be able to connect different sets of administrative information held about people technically, but should there be any rules about accessing and using the data? Such a concern extends to those areas of social research where there is open access to the data that people globally post on social media about themselves.
The accessibility of personal digital information, and the possibilities for linking together anonymised or pseudo-anonymised records of two or more large administrative or survey data sets, provide social scientists with unprecedented opportunities to transform our understandings of a range of social issues. Innovative technologies create new opportunities for knowledge but they also invade privacy and provide new channels for discrimination. The field is fraught with a range of ethical and social concerns. To stimulate and inform the important debate on these issues among social scientists, research funders and policy-makers, the International Journal of Social Research Methodology and ESRC National Centre for Research Methods held an innovation forum event, hosted by the Warwick Business School at The Shard, London.
The day began with Mandy Chessell, Master Inventor and Distinguished Engineer at IBM, delivering the NCRM Annual Lecture exploring the process of ethical decision-making about data linking in the business sector. Peter Elias and Rob Procter (both at Warwick) followed with presentations on the strengths and drawbacks of administrative data linkage and sharing social media information. Ian Dunt, editor of politics.co.uk, provided a political perspective on the policy-making process and civil liberties, while a panel consisting of Neil Serougi (independent researcher), Emma Uprichard (Warwick), David Martin (Southampton), Peter Smith (Southampton) and Rob Procter, identified opportunities and caveats in discussion with the audience.
The main deliberations revolved around:
- The benefits of data linkage:
We can know about who people are, what they do, how and why they do it, for the betterment of society. There are also potential cost savings as data from many sources can be processed automatically and linked cheaply.
- Purpose and power:
There is a business interest in the use of public data, both in terms of providing IT solutions for managing linkage, and in using the data for commercial purposes. The public, however, may feel that it is acceptable to link up administrative data only for the primary purpose for which was collected, such as health and social care provision. In particular, there is a concern that commercialisation of data and using data for non-intended purposes breaches rights to privacy. What is technically and legally possible is not necessarily what is ethically and socially right.
- Ethical governance
In the British context there is no legal requirement for consent to administrative data linkage (albeit European legislation may impose this) and the conventional, institutional and professional research ethics that social scientists work with may not be sustainable in an age of big data and data science. The old practices are founded on principles of anonymity and informed consent, both of which are disrupted where big data, the World Wide Web and the internet of things have built-in linkage and are resistant to anonymisation. New ethical principles and practices need to be developed.
Possible ways forward for ethical data linkage practices and governance could include a new socio-technical architecture for data management involving a personal data store, raising understanding and building trust through public debate, and setting up an independent ‘watchdog’, data ombudsman, ‘data parliament’ or citizen’s panel to ensure transparency, openness, a form of consent and appropriate use.
To watch the recorded presentations and listen to the audio podcasts go to www.ncrm.ac.uk - resources - podcasts or videos.