Posted on 21 March 2018 by Matt Barnard, Head of Communities, Work & Income
The scandal over Cambridge Analytica’s use of data harvested from Facebook users is yet another sign that we are now in an age when the collection, storage and use of information has moved from the realm of ‘geeks’ to the status of potential political bombshell.
The revelations about the use of data to try and influence the American election are so significant that billionaire Facebook founder Mark Zuckerberg has been called to give evidence to a parliamentary committee. Meanwhile, the UK’s information commissioner Elizabeth Denham is applying for a warrant to search the offices of Cambridge Analytica. The innocuous-seeming 1s and 0s that make up most data are coming back to bite their so-called masters.
In social research, assurances about confidentiality are a key plank of ensuring informed consent, one of the basic pillars of conducting ethical research. As the Social Research Association’s (SRA) ethics guidelines indicate, it’s the task of researchers to maintain the principles of confidentiality in order to protect the interests of research.
And in a groundbreaking study NatCen conducted, exploring the views and experiences of research participants of research and ethics, confidentiality was one of the research participants’ fundamental assumptions about what taking part in research meant.
So, the Cambridge Analytica / Facebook revelations have the potential to damage not just those organisations’ reputations, but the image of social research more widely. This is particularly problematic in the context of declining response rates to major surveys, which Keith Bolling of Kantar Public UK and Pattern Smith of Ipsos MORI recently argued may be down to a loss of trust.
However, the damage may go beyond the collection of primary data; it may make it even more difficult to link datasets, a research approach that has the potential to bring enormous public benefit.
Back in 2015 (which seems like eons ago), the Wellcome Trust published a report setting out not only the benefits of data linkage but some of the key barriers to doing it and argued that the tone of the debate needs to change. It contended that the assumption that no data can be released unless it is explicitly allowed should be changed to an assumption that all data can be used unless it is shown be unlawful or unethical.
This is the equivalent of changing organ donation from an ‘opt in’ to an ‘opt out’; both changes, to organ donation and data usage, could have significant public benefits, but both also have profound ethical implications.
In part to understand the implication of this kind of change, NatCen recently conducted three deliberative discussion workshops for University College London exploring public attitudes to health data linkage.
One thing the workshops demonstrated was that the issues are complex and difficult to grasp for some members of the public. However, it was also clear there was support from some for the status quo of on-going consent for each use of data and deeply felt concerns about whether data could be completely anonymised and whether the transfer of data was always secure.
The implication, as is often the case in social research, is that the issue of trust, confidentiality, and the collection and use of data is complex.
It also implies that there is a need for more dialogue to enable people to understand the issues and help the public to gain confidence that their data can safely be used to make society better and fairer.
What’s also clear is that the Cambridge Analytica / Facebook scandal is going to make that significantly harder.