Data governance sounds like an obvious thing to do, especially for statistics producers. We all have data so what could be more obvious than good governance that safeguards data uses and reassures people – research participants, public and data users alike – that their interests are ours as well? Legal compliance with appropriate procedures for privacy and security is an important part of the governance of data. But there is more to it than that – governance is about people, and making accountable decisions about data which can sometimes require you to resolve different interests. This blog illustrates the approach that has been developed to support data governance decisions for a particularly complex set of studies.
Managing Ethico-social, Technical and Administrative issues in Data ACcess (METADAC)
METADAC is one example of data governance for the large biosocial data resources of UK longitudinal studies (funded by the ESRC, MRC and Wellcome Trust). Data from these studies are highly specialised, involving longitudinal social, economic and health data spanning decades, and biological samples collected from participants. They therefore require a highly interdisciplinary data access committee including ethicists, social scientists, lawyers, clinicians, geneticists and survey statisticians – more details are published here.
METADAC pays particular attention to engaging with the study participants and members of the public. This matters because it is good practice, but as METADAC governs access to data from long-term studies, maintaining the confidence of subjects is also critical. Or conversely, losing their confidence would be calamitous. Indeed, one of the key questions the committee asks is “Will this upset or alienate participants?”
Public engagement to build public confidence
Public engagement in METADAC is twofold. First, study participant members sit on the committee providing insight into how participants may view data uses. Second, METADAC applicants must provide high-quality plain language summaries of their studies to communicate how data are or will be used. These summaries are published on the METADAC website.
Ultimately, if governance is to be effective it must be accountable, trustworthy and fair. This means responding to stakeholders’ concerns , and seeking to strike a balance between the views of different groups. Specifically this may mean championing the rights of less powerful groups, and taking particular steps to seek out their views and respond to .
Data linked from administrative sources will need sophisticated governance to sustain public trust. METADAC oversees research use in a very specific domain of large studies, but the principles underlying its approach – in particular the attention paid to public engagement – are transferable to other settings. We don’t really own all of this data, bt we are responsible for looking after it, keeping in mind how important it is to everyone involved.
The Royal Statistical Society’s new Special Interest Group on Data Ethics will be organising a meeting on Data Governance to look at these issues across the landscape to be held in Newcastle in May 2019.
Thomas King is Secretary of the RSS group on Data Ethics
Madeleine Murtagh is chair of METADAC