Twenty-fifth meeting of the National Statistician’s Data Ethics Advisory Committee


Meeting agenda:

  1. Minutes and matters arising from the previous meeting
  2. Integrated Data Programme (IDP)
  3. County Lines Project
  4. Qualitative Research for a UK Child Abuse Prevalence Survey Feasibility Study
  5. Welsh Government’s post implementation review of the Children (Abolition of the Defence of Reasonable Punishment) (Wales) Act 2020
  6. CADE Guidance Update
  7. Any other business
    1. Ethics compliance review update

Present Members
Dame Moira Gibb (Chair)
Stephen Balchin
Rob Bumpstead
Vanessa Cuthill
Colin Godbold
Isabel Nisbet
Marion Oswald

UK Statistics Authority
Lily O’Flynn
Daniel Towler
Simon Whitworth
Ross Young
Emma Walker (for item 6)

Office for National Statistics
Peter Stokes (for item 2)
Tim Gibbs (for item 3)
Becky Vials (for item 3)
Meghan Elkin (for item 4)
Sophie Sanders (for item 4)

Other
Ellie Roberts, NatCen (for item 4)
Hayley Collicott, Welsh Government (for item 5)
Ryan Nicholls, Welsh Government (for item 5)

Apologies
Emma Uprichard

  1. Minutes and matters arising from the previous meeting

1.1       The Chair welcomed members to the 25th meeting of the National Statistician’s Data Ethics Advisory Committee (NSDEC).

1.2       Members approved the minutes from the previous meeting.

1.3        Daniel Towler updated the Committee with progress on actions from the April 2021 meeting.

  1. Integrated Data Programme (IDP)

2.1       Peter Stokes (Director of the Integrated Data Programme, ONS) presented this initiative which seeks to deliver the Integrated Data Service (IDS) by bringing together ready-to-use data to enable faster and wider collaborative analysis for the public good.  Peter requested advice from the Committee on the ethical considerations to be mindful of while developing the service, as this is to be done in a way that champions data security.

2.2        The Committee recognised the value in improving the way in which government data can be harnessed to support public good research. The NSDEC encouraged the IDP to take an ethics first approach and encouraging those they worked with to do the same.  The Committee recognised that the UKSA’s ethical framework has been successfully implemented and has enabled a broad spectrum of policy relevant research from across government and the wider research community.

2.3       The following points were raised by the NSDEC during discussion:

  • The different research governance processes that government analysts and non-government analysts would be subject to when accessing data via the IDS, including ethical consideration of both the collection and use of data were unclear to the Committee;
  • NSDEC stressed the importance of researchers giving ethical consideration to their uses of data in the IDS in all cases. The UK Statistics Authority’s ethics self-assessment tool and accompanying user support framework was identified as an efficient and convenient way for researchers to consider the ethics of their uses of data in the IDS. The Committee agreed that the ethics self-assessment process empowers researchers to quickly consider the ethics of their research, which is already widely accepted and used across government and the wider research community;
  • It was agreed that ethical consideration should not only be applied by analysts accessing and using data via the IDS. Data ethics must also be considered during the collection, linkage and assembly of linked data assets that the IDP will be providing access to. Subsequently, ethical consideration should be given to the IDP as an overall concept, but also at an application/research project level;
  • The Committee emphasised that the informed consent of data subjects should permit the use and linkage of data in the way that the IDP proposes to use it;
  • The Committee stressed the importance of the Integrated Data Programme having an accountability structure in place to ensure that appropriate ethical consideration is given to the collection, linkage and use of data within the programme;
  • It was agreed that there should be careful thought given to research projects that are considered, especially ethically sensitive projects, and what the implications of the IDS are on this. The NSDEC was identified as a suitable Committee that could provide ad hoc independent ethics advice to such projects via correspondence, where required, and;
  • The Committee suggested that ethical consideration should be included within the presentational materials that Pete and IDP members are regularly using to keep stakeholders informed of updates to the Programme.

2.4       Pete thanked the Committee for their comments and agreed that ethical consideration, whether that is the ethics self-assessment, or consideration from NSDEC for more ethically challenging projects, would be included in further interactions of the IDP and its associated policies and services.

2.5       Action – The secretariat to work with Pete and the IDP team to help integrate ethical considerations into the programme. The secretariat to also assist Pete in providing the Committee an update on the progress of the IDP at a suitable time.

  1. County Lines

3.1       Becky Vials and Tim Gibbs, from the Public Policy Analysis division in ONS, presented plans for a project, which is being commissioned by the Home Office,  and seeks to answer what the indicators of County Lines drugs activity are, and identify the indicators of a local area which is vulnerable to County Lines operations becoming  established. ONS plans to produce outputs to help answer these research questions, including indicators of vulnerability, frequency tables for police force areas, and a machine learning dashboard to show vulnerabilities for each area.

3.2        The following points were covered in the discussion:

  • The researchers confirmed that this was the only form of ethical review that the project is currently subject to;
  • The Committee understood that the outputs of this research are at the police force area level and advised the researchers to be mindful that police force areas may cover regions with differing levels of vulnerabilities to the establishment of County Lines activity. The researchers should be aware of this when communicating outputs, which may feed into decisions around County Lines interventions across police force areas;
  • The Committee suggested that the research team consider whether police forces have carried out similar projects to learn from any limitations of these and understand any findings and recommendations;
  • The Committee agreed that it would be useful to understand how the police would use this information, and consequently, how they will be controlled and limited in their use of them. The Committee therefore requested reassurance that the police forces benefitting from this information would only use it to support decision-making processes where permitted and appropriate;
  • The NSDEC requested to have more detail on what police data were being used for this project;
  • The Committee agreed that it would be most beneficial if the study could decipher causation rather than correlation, however stressed that if the former is not possible, that this is clearly communicated;
  • The committee emphasised the importance of ONS being transparent about their involvement with the project. Given the potentially experimental nature of this project, the Committee agreed that any outcome from this project would be legitimate and interesting, regardless of whether the project fulfils its initial aims in their entirety.

3.3       Action – The secretariat to work with the research team to provide the Committee with a full project application for consideration, as and when it is confirmed that this project will be feasible and valuable. 

  1. Qualitative Research for a UK Child Abuse Prevalence Survey Feasibility Study. NSDEC(21)10.

4.1       Sophie Sanders, Meghan Elkin (Centre for Crime and Justice, ONS), and Ellie Roberts (NatCen) presented this item. This project is an extension of the item previously considered by NSDEC in February 2019, which was a piece of qualitative research to inform a feasibility study of whether a child abuse prevalence survey could be effective. It had been identified that several areas need further investigation before ONS decides to proceed with a pilot survey in this area.

4.2       The NSDEC acknowledged the substantial public good that would come from this work, and appreciated the preparation involved to best realise the benefits. The Committee thanked the researchers for the project application submitted detailing developments on this project, as ethical risks had been thoroughly considered.

4.3       The following points were raised by the NSDEC during discussion:

  • The Committee recommended that a Data Protection Impact Assessment be completed for this project, if not already done so, to understand any privacy impacts associated with the processing of this sensitive information;
  • The NSDEC queried the suitability of conducting focus groups for adults on potentially challenging and sensitive issues, and suggested that 1 to 1 interviews may be more suitable;
  • The NSDEC requested reassurance that safeguards are in place to ensure participants are alone when engaging in the study remotely, so as to ensure that individuals are not being coerced to answer in a certain way if, due to Covid-19-related restrictions, interviews are conducted via video conference. The Committee emphasised their support for face-to-face participation wherever possible;
  • The Committee agreed that it would be vital to ensure the location of interviews is comfortable for the participant and does not have any relation to negative or distressing experiences;
  • The Committee emphasised the importance of being transparent with participants as to why they have been chosen to participate in the study so that participants are aware of the expectation to engage with potentially sensitive memories and experiences;
  • The Committee highlighted that while participants may indicate that they will be willing to answer questions on these sensitive issues if a child abuse prevalence survey was to be operationalised, this is only a hypothesis of willingness to answer which may differ during the survey collection, and;
  • The NSDEC emphasised the importance of ensuring the representativeness of the sample, and the need to be transparent in any outputs about the limitation of the sample used.

4.4       The Committee were appreciative of the thoroughness of the application, and acknowledged the significance and importance of the work

4.5       Action – The secretariat to work with the research team to understand what other ethics committees have considered in relation to this work and the subsequent outcomes.

4.6       Action – The research team to provide an update on the findings on the feasibility study, and involve the NSDEC should this study develop further.

  1. Welsh Government’s post implementation review of the Children (Abolition of the Defence of Reasonable Punishment) (Wales) Act 2020. NSDEC(21)11.

5.1       Ryan Nicholls and Hayley Collicott presented this item. The Children (Abolition of the Defence of Reasonable Punishment) (Wales) Act 2020 will remove the defence of reasonable punishment for parents in Wales from March 2022. This project will use Natural Language Processing (NLP) techniques to identify police records relating to parents/guardians that are reported to the police for physically punishing their children. This is to allow Welsh Government to monitor the impact that the Act has had on public services in Wales.

5.2       The Committee understood the requirement for this work, but advised that the quality of data, and methods used must be assured to ensure useful and valid outcomes that can ultimately realise the public benefit.

5.3       The following points were raised by the NSDEC during discussion:

  • The Committee advised that the researchers confirm whether the NLP techniques and the data being used enables the potential to realise the perceived public good of the project. The Committee suggested engaging with research teams who have previously used these techniques for similar purposes, and seeking further feasibility support from NLP specialists;
  • The Committee suggested that the researchers considered that the baseline that changes are being measured against, which is being calculated prior to the implementation of the act, could have already been impacted as a result of the passing of the act, thereby being a non-authentic baseline, and;
  • The NSDEC queried what the implications of the use of Welsh and English in police reports would be on the validity of results.

5.4       Action – Researchers to keep the Committee updated on the progression of the project and requested assurance that the methods will produce successful results.

  1. The Centre for Applied Data Ethics (CADE) Guidance Update. NSDEC(21)12.

6.1       Emma Walker (Centre for Applied Data Ethics, UKSA) presented this item which put forward two pieces of guidance due for publication by the Centre for Applied Data Ethics. Members of NSDEC are invited to provide comments on the structure, approach and content of these guidance pieces, which focus on the use of geolocation data for research and statistical purposes, and the maximisation and articulation of the public benefits of a research project.

6.2       The NSDEC agreed that the guidance helpfully supported researchers in considering the ethical implications of these issues.

6.3       The following points were raised by the NSDEC during discussion:

  • The Committee supported the inclusion of practical resources in these guidance pieces, such as checklists, and advised that the Centre should be clear on how researchers can best use these resources alongside existing frameworks, such as the ethics self-assessment tool, to support the thorough consideration of such issues. It was suggested that an index that assists researchers in finding the most relevant piece of guidance for their need would be a useful output in itself;
  • Regarding the guidance relating to the maximisation of public good research, the NSDEC suggested there be further guidance on the consideration of legal proportionality, as this has different implications to the consideration of benefits and risks, already included in the public good draft, and;
  • The Committee agreed that it is also vital for the probability for the public benefit to be realised is also considered, not just the potential theoretical benefit.
  1. Any other business

7.1       The quarterly Data Ethics Compliance Review, NSDEC 21(13) noted that the Data Protection Compliance team at the UKSA has found that some projects selected for review have had to be paused due to re-prioritisation during the COVID-19 pandemic. Due to this, two papers were presented to the Committee:

  • Annex A presents a revised plan for ongoing data protection and ethics compliance reviews.
  • Annex B presented a compliance review of project NSDEC(20)03: Investigating the risk factors and inequalities associated with infant mortality. The review found that the project had paused due to COVID-related reprioritisation. The project recommenced in planning stages in May 2021. A further data protection and ethics compliance review will be undertaken when phase 1 of the project completes

7.2       Simon Whitworth thanked the Committee for all their help in supporting the establishment of the Centre for Applied Data Ethics.

7.3       The Chair informed the Committee about the public launch of the Inclusive Data Task Force in September 2021.

7.4       The secretariat suggested that the next meeting be rescheduled for a later date than originally planned date 29th September. The next meeting date is to be confirmed.