UK STATISTICS AUTHORITY
RESEARCH ACCREDITATION PANEL
Monday 13 December 2018.
Present Committee Members Professor Paul Boyle (Chair) Siobhan Carey (Northern Ireland Statistics and Research Agency) Tricia Dodd (Independent member) Rebecca Endean (UK Research and Innovation) Roger Halliday (Scottish Government) Sarah Henry (ONS) Glyn Jones (Welsh Government) Sarah Mathieson (Independent member) Advisors Peter Stokes (ONS) Andy Wall (ONS) Secretariat Simon Whitworth (UK Statistics Authority) Ross Young (UK Statistics Authority) Apologies Chris Dibben (Independent member) Kevin Fletcher (HMRC) Andrew Garrett (Independent member) Neil McIvor (Department for Education)
Professor Paul Boyle (Chair)
Siobhan Carey (Northern Ireland Statistics and Research Agency) Tricia Dodd (Independent member)
Rebecca Endean (UK Research and Innovation) Roger Halliday (Scottish Government)
Sarah Henry (ONS)
Glyn Jones (Welsh Government)
Sarah Mathieson (Independent member)
Peter Stokes (ONS) Andy Wall (ONS)
Simon Whitworth (UK Statistics Authority) Ross Young (UK Statistics Authority)
Chris Dibben (Independent member) Kevin Fletcher (HMRC)
Andrew Garrett (Independent member) Neil McIvor (Department for Education)
1.1 The Chair welcomed members to the first meeting of the Research Accreditation Panel.
1.2 Members introduced themselves to the other members of the Panel.
1.3 Members were informed that the secretariat would work with them to produce a short biography for the Research Accreditation Panel section of the UK Statistics Authority’s website.
2. Overview of Digital Economy Act (DEA)
2.1 Ross Young, Head of Data Governance, Legislation and Policy in the UK Statistics Authority, provided an overview of the The meeting heard that the DEA (Part 5, Chapter 5) includes an important new statutory framework to support the UK research community, both within government and beyond, that permits public authorities to share de-identified information with accredited researchers for the purposes of public good research.
2.2 Members were also briefed that under the DEA, the UK Statistics Authority is the statutory body that will oversee the accreditation of researchers, projects, processors and secure access The Authority has published a set of criteria, set out in the Research Code of Practice and Accreditation Criteria, that individuals, organisations and research projects must meet before being accredited for any of the functions set out in Chapter 5, Part 5 of the Act.
2.3 Members were briefed that data processors will be able to retain data where there is a clear research rationale for doing so subject to the consent of the relevant data-holding public authority.
3. Accreditation of Processors
3.1 Andy Wall, ONS’s Chief Security Officer, presented on the security accreditation process for data processors under the It was reported that an accreditation process and security control set had been developed that was linked to the breadth of the security control set required for the internal security standard of ISO 27001. This standard is accepted by the UK Government as a baseline for security. The intent of the process is that data processors complete a compliance spreadsheet which is then assessed by the ONS Security team.
3.2 Members were informed that ONS had engaged with a range of data partners who had indicated a general acceptance for the overall approach.
3.3 The Panel suggested that the approach outlined was sensible and approved the It was agreed that there should be a review of how the accreditation of processors was working after six months.
ACTION: The Panel requested that Pete Stokes present the plans to accredit the staff, training and capability of processors at the next meeting.
4. Accreditation of Researchers and Research Projects
4.1 Pete Stokes presented how researchers and research projects would be accredited under the DEA. It was reported that the process for accrediting and training researchers would be the same as the process used for the UKSA Approved Researcher scheme, which is conducted by ONS, as this fully meets the requirements of the DEA Research powers and related Code of Practice and Accreditation Criteria. Therefore, to be accredited a researcher would need:
i. An undergraduate degree (or higher) including a significant proportion of maths or statistics, or be able to demonstrate at least three years quantitative research experience;
ii. Have completed the ONS Safe Researcher Training (or recognised equivalent), and passed the related ; and
iii. Allow their details to be included in a public record of all accredited researchers.
4.2 It was reported that researcher accreditation lasts for a period of five years, subject to compliance with all terms and conditions of access to data, and researchers may complete multiple projects within that time (subject to each project being separately approved).
4.3 The Panel heard that the DEA states that research projects can only be approved if departments approve the use of their data for the projects and the projects are judged to “serve the public good”. The Panel has been set up to make this independent assessment that projects are in the public The public good is defined in the following way, which is consistent between the Approved Researcher scheme and the DEA:
i. To provide an evidence base for public policy decision-making
ii. To provide an evidence base for public service delivery
iii. To provide an evidence base for decisions which are likely to significantly benefit the UK economy, society or quality of life of people in the UK
iv. To replicate, validate or challenge Official Statistics
v. To replicate, validate or challenge existing research
vi. To significantly extend understanding of social or economic trends or events by improving knowledge or challenging widely accepted analyses
vii. To improve the quality, coverage or presentation of existing statistical information
4.4 The Panel agreed that the DEA requires that all project proposals require formal ethical The Research Accreditation Panel will be asked to ensure this has been completed and will have authority to request additional scrutiny by the National Statistician’s Data Ethics Advisory Committee (NSDEC), if considered useful. The relationship between the Panel and NSDEC was seen to be important.
ACTION: The secretariat was asked to present what will be considered by NSDEC and what will be considered by the Panel at a future meeting.
4.5 The following points were made in the discussion that followed:
i. Ministerial approval should not be required for every The Research Accreditation Panel want to quickly get to a position where Departments trust the Panel to make the decision about whether a project is for the public good.
ii. The public good definition set out in the DEA was designed to be broad and exploratory and methodological research will be possible under this public good It was suggested that some of this research will be useful as policy relevant research.
iii. Commercial researchers can access data under the All researchers accessing data under the DEA will have to publish the results of their research so no commercial advantage can be gained from accessing data under the legislation.
iv. The Panel suggested that they wanted to hear about how the accreditation of researchers and research will be communicated across the research community at a future It was suggested that examples of successful and unsuccessful research applications should be published to help researchers.
ACTION: Pete Stokes was asked to present the portal the researchers will use to interact with the accreditation process at the next meeting.
v. It was suggested that it was important that the Panel had clear oversight of how the processes were working so that any bottlenecks could be identified and The Panel asked for regular reports detailing key metrics, describing how the processes were working, to be presented at future meetings.
ACTION: Pete Stokes was asked to present an example of how this report may look at the next meeting including the metrics that will be presented.
4.6 Both the processes to accredit researchers and research projects were It was agreed that there should be a review of how these processes were working after six months.
5. Terms of Reference
5.1 Professor Paul Boyle presented the Terms of Reference for the Panel. These were approved by the Panel.
6. Open discussion about opportunities, challenges and risks
6.1 All members discussed the opportunities, challenges and risks presented by Research Strand of the The following points were made in the discussion:
i. All the processes that have been discussed sound sensible and members representing the Devolved Administrations felt that they had been well engaged in the process.
ii. The Research Accreditation Panel will monitor how these processes work in If problems occur in accessing data from some departments then the Panel would escalate these issues quickly at the highest level within the Departments concerned.
iii. The communication of the opportunities presented by the Research Strand of the DEA to the research community will be important in stimulating researcher demand to use the legislation to produce research for the public good.
iv. It was noted that other approvals panels exist for research that was not taking place under the DEA. Approval processes should be aligned so researchers don’t have to go through lots of different approvals.
ACTION: To help inform this, the secretariat was asked to map the approvals panels landscape across the UK and present this at a future meeting.
6.2 John Pullinger, the National Statistician, joined the meeting to thank members for their involvement in this work and offer his support to making a success of the Research Strand of the DEA.
7. Any other business
7.1 Members were informed that the secretariat would be in touch to arrange the meeting dates for the next six months.