| We uphold trustworthiness quality and value of statistics and data used as evidence | |
| Challenge official statistics producers to answer society’s questions, focusing on appropriate timeliness, granularity and data linkage | |
| Use rapid reviews to support producers as they develop new outputs and approaches in response to the pandemic | Fully delivered: We published 14 Rapid Reviews in 2020/21 at the request of producers. We expect the benefits gained from this approach to inform improvements to our regulatory toolkit beyond the pandemic. |
| Regulatory programme looking at outputs which draw on multiple sources or can offer detailed breakdowns | Fully delivered: We consistently challenged producers to make more granular analyses available, and endorsed good practice – for example, Business Demography statistics and Scottish Prison Population statistics. This will be a continuing key theme in 2021/22. |
| Systemic Reviews with a focus on society’s most important issues which cut across policy | Partially delivered: We published reviews on Post-16 Education and Skills, Mental Health, and Adult Social Care. We have reviews ongoing on Children and Young People, and Climate Change. At the invitation of HMRC, we also carried out a quality review across their statistics. |
| Continued advocacy for data linkage and use of multiple sources to answer society’s most important questions | Partially delivered: We have embedded the advocacy of data linkage into our regulatory work in 2020/21. We have been building our networks with partners with common goals such as ADR UK and NAO, running joint events and planning more guest blogs. We have published Reproducible Analytical Pipelines: Overcoming Barriers to Adoption; and Unlocking the Value of Data through Onward Sharing. An update on our implementation of our data linkage recommendations is planned for 2021/22. |
| Challenge official statistics producers to offer insight, clarity of communication and coherence | |
| Regulatory work focuses on clarity of communication | Fully delivered: We challenged producers to provide clarity of coherence and insight and endorsed good practice across a range of statistics – for example, Rough Sleeping statistics. |
| Systemic Reviews looking at whether statistics across broader topics are coherent and communicated effectively | Fully delivered: Our reviews of Defra user engagement, Post-16 Education and Skills, and Reproducible Analytical Pipelines all had strong focus in this area. Our review of Poverty statistics launched in November 2020 considers the extent to which the statistical system is providing coherent and comprehensive information required to support decision making. |
| 2020 Insight programme: using intelligence to identify key themes in our work that we can learn from and address across the statistical system | Partially delivered: Our discovery work has delivered a number of recommendations which are being developed into a plan for implementation. Some are quick wins, such as making improvements in the project closure report and other regulatory documents. Others are more innovative and complex and are being scheduled. |
| Increase voluntary application of the Code beyond official statistics, particularly for government analytical outputs | |
| Continue to promote voluntary application of the Code through support of the established Community of Practice, workshops and published case studies | Fully delivered: 26 statements of voluntary compliance with the Code have been published so far in total by a range of organisations, from producers of official statistics to those outside the public sector. In 2020/21 six statements were published, including for Department for Health and Social Care’s NHS Test and Trace statistics. We have strong interest from other organisations to adopt the Code. |
| Support government to publish analytical outputs, such as management information and research outputs, in line with the Code | Fully delivered: We have published a range of guidance and case studies in support of the Code – for example, we published a statement on Why trust and transparency are vital in a pandemic in November 2020. Our on-line Code course is now available for the government statistical system on the Learning Hub. We are working with partners in the GSS to develop scenario-based self-learning modules that will illustrate applying the pillars of the Code, and to roll out the Code across the Analysis Function. |
| Sponsor the Voluntary Application Award, in partnership with the Royal Statistical Society (RSS), rewarding excellence in the voluntary application of the Trustworthiness, Quality and Value pillars of the Code | Fully delivered: The first awards have been handed out and we have launched the 2021 awards. We had some excellent nominations in 2020 and we published Case Studies by our winners and those who were highly commended to share their experiences more widely. |
| We protect the role of statistics in public debate | |
| Focus on addressing data gaps and stepping in when statistics are being misrepresented | |
| Identify and highlight gaps in statistics and data | Fully delivered: We have challenged producers to address gaps in statistics and data across our regulatory programme. We’ve used what we have learned to offer insights. We published two blogs: Closing data gaps: understanding the impact of COVID-19 on income; and a blog from our Director General to start the process of demystifying data gaps more widely. The pandemic has highlighted the importance of looking to the questions that need to be answered with data, both now and in the future, and this will remain a key theme for us in 2021/22. |
| Intervene where we identify significant or persistent issues with how statistics are being used or misrepresented | Fully delivered: We have opened 323 cases in 2020/21, a threefold increase on previous years. We published an updated Interventions Policy and FAQs to help the public to engage with us and know what to expect. We anticipate our volumes of casework to remain higher than in the past, and we are continuing to improve processes to allow us to effectively deliver. Our review on Statistical Leadership, published in February 2020, highlights the importance of Governments across the UK showing leadership in the use of numbers and data. |
| Systemic Reviews to encourage producers to address gaps in statistics | Fully delivered: Our review programme has had a strong focus this year on increasing the public value of statistics and understanding where users’ needs are not being met/ their questions are not being answered by the available statistics. Examples of published Reviews which have addressed this area: Mental Health (England), Post-16 Education and Skill and Adult Social Care. |
| Develop automated tools and dashboards to provide intelligence on how statistics are used in social and mainstream media | Partially met: Our Twitter dashboard is now fully cloud based and able to run automatically. This has set the infrastructure for future projects which should now be much quicker to deploy. We are scoping projects on Google Trends (and how this might inform data gaps) and scraping Hansard to provide analysis of what is being talked about in UK Parliament. |
| Increase the profile of OSR to support impact of our interventions | |
| Greater use of regulatory statements on website, so people can see our position on key issues | Partially met: We have published a number of position statements during 2020/21 – for example: domain updates on crime, trade and the RPI; our expectations of producers in respect collecting and reporting data about sex; the impact of COVID-19 on the Labour Force Survey; and welcoming changes to pre-release access in Scotland. We need to make these statements more accessible in the coming year. |
| More regular contributions to Parliamentary inquiries and raising issues with relevant select committees | Fully met: Within the context of the pandemic, our Director General for Regulation has been called to offer evidence to PACAC and select committees. We have also contributed evidence to parliamentary reviews. We will seek to continue to seek to improve how we inform UK and devolved Parliaments in 2021/22. |
| Develop our website | Fully met: We launched the OSR website in 2020 and it has become an important vehicle for communicating our work. As well as our formal outputs we have published a series of blogs on a wide range of subjects by guests and our own team. In September, we delivered enhancements to comply with the new UK Government Accessibility Regulations. |
| Strengthen our regulatory presence on Social Media e.g. through our new Twitter account launched in 2019 | Fully met: We now have over 2000 followers on Twitter and have used social media in increasingly creative ways to engage with our audience. On our one-year anniversary of @StatsRegulation we posted our most engaged-with tweet ever. Our intervention that day had 105,000 impressions; 73,000 views of the video and 15,500 engagements. We increased by 150 followers in four hours. |
| We develop a better understanding of the public good of statistics | |
| Clarify our role in regulation of data and AI | |
| Publish a statement on our regulatory role in the data and AI landscape | Fully met: We published our statement and supporting blog in July 2020, and the subsequent report on Ensuring Statistical Models Command Public Confidence consolidated our messaging. |
| Work with organisations working on data and AI to further our knowledge, including their regulatory role and any gaps in regulation | Fully met: We have been working with ONS’s Data Science Campus and others like NAO and ADR UK and developing events, blogs and plans for workshops to build our team’s capability. |
| Build our evidence base on understanding of the public good | |
| Review the National Statistics designation | Fully met: This review will continue into 2021/22 but we have made strong progress this year, publishing a consultation document and engaging with Heads of Profession and the wider analytical community to inform our thinking. We published findings and next steps in June 2021. |
| Further develop a framework for judging misleadingness | Fully met: We published an initial think piece in May 2020 and a programme of engagement was carried out to inform a follow-up piece one year on. |
| Utilise our research and data and automation capabilities to develop our understanding of what is meant by public good of statistics and existing evidence | Fully met: We published our report: The Public Good of Statistics: What we know so far in December 2020 and work is progressing well on other elements of our research. We shared our latest findings and plans for further work with our external expert Steering Group in April 2021. |
