Summary of changes between the principles in Code 2.1 and the Standards for Official Statistics (OS) in Code 3.0 for OS producers
The following sections start with the principles from Code 2.1 and explain where they are included in the Standards for Official Statistics in Code 3.0. When we refer to T1 to T6, Q1 to Q3 and V1 to V5 we are referring to Code 2.1. When we mention a standard, e.g. standard (std) 1 we are referring to Code 3.0.
Trustworthiness principles:
T1 Honesty and integrity has similar practices to standard (std) 1 Show integrity. Plus, one practice was moved to std 4 Manage data responsibly; one practice was removed as it is covered by the Standards for Public Use, and a practice on challenging misuse was moved from T2.
T2 Independent decision making and leadership is very similar to std 2 Lead responsibly, emphasising that analytical leadership is all analysts’ responsibility. It also specifies where decisions are under the authority of the Head of Profession/ Chief Statistician. Quality management was moved to std 5 Prioritise quality. A practice on statement of compliance was added.
T3 Orderly release is very similar to std 3 Be transparent; additionally, a published revisions and corrections policy and release practice policy are required. The need to make the status of stats clear (OS, AOS, OSD) was added. Statistics leads no longer need to be named, but straightforward access to the team does need to be provided.
T4 Transparent processes and management has been split across other standards aligned to these areas of practice, e.g., business management to std 2, quality management to std 5.
T5 Professional capability has been split between std 1 and std 2.
T6 Data governance is very similar to std 4 Manage data responsibly – plus, a practice related to social acceptability was added; a data management approach is required; and a confidentiality policy needs to be published (new guide coming soon on public involvement and engagement from OSR).
Quality principles:
New: std 5 Prioritise quality has some new practices emphasising quality culture and a supportive learning environment. It includes some practices on innovation and improvement and a new requirement for a quality management approach to be published.
Q1 Suitable data sources is very similar to std 6 Be rigorous as this standard covers the area of practice for producing stats; the need for considering coherence and comparability is emphasised, particularly when considering new statistics across the UK.
Q2 Sound methods is split between std 6 Be rigorous and std 7 Be open about quality. Std 7 Be open about quality emphasises prominently communicating about the quality of the stats and summarising uncertainty.
Q3 Assured quality is mainly in std 6 Be rigorous – the Code now emphasises both input data and stats output quality dimensions; stats should be verified to be representative and statistical error quantified where possible.
Value principles:
V1 Relevance to users is similar to std 8 Be relevant – there is a strong emphasis on being accountable to users by involving users to inform stats planning. In addition, there is a requirement to publish an annual statistical work programme and a public involvement and engagement strategy. The new accountability guide draws together the messages across the Code to help steer practice.
V2 Accessibility is very similar to std 10 Be accessible – it draws out requirements around prominently explaining quality issues.
V3 Clarity and insight is very similar to std 9 Be clear – plus, it now includes a practice to explain how the stats add value and serve the public good and one about highlighting potential misinterpretations and clarifying meaning when stats are used inappropriately.
V4 Innovation and improvement is spread across relevant principles according to the area of practice std 6 Be rigorous and the Value principles.
V5 Efficiency and proportionality is split between std 6 Be rigorous and std 10 Be accessible.
Takeaway messages for producer teams:
- Know your organisational policies for release practices, quality management and data management and be clear where your local practices diverge from the organisational policies (see the guide listing required policies).
- Consider publishing a statement of compliance for your business area (practice 2.5) (in addition to the required organisational statement) and look at how you contribute to the stats annual work programme and involvement and engagement approach.
- Consider how you are monitoring input and output quality, as well as how you summarise quality issues and measure and describe uncertainty.
- Be a Code leader and encourage others to model good practice (practices 2.1/2.2) – check out the new Standards for the Public Use of Statistics, Data and Wider Analysis and see how you can support your colleagues.
General guide for next steps
- Check out the detail of the Code by using the short guides, which contain useful questions to consider and provide links to useful resources.
- Review your approach to TQV and see how well your team’s practices align with the Code and how you can improve.
- For more detail on the changes, refer to our Code mapping spreadsheet, which shows how the practices and principles from Code 2.1 correspond to the new Standards for Official Statistics / Code of Practice for Statistics edition 3.0, and vice versa.