With the Census in England and Wales taking place on 21 March 2021, we created a programme of webinars to showcase our plans for design and quality assurance. The series, which was carried out through November and December 2020, included a high-level introductory overview as well as 'In Focus' sessions that outlined specific aspects in more detail. These webinars gave attendees the opportunity to ask questions and provide feedback.

1. 1
Quality Assuring the
Census 2021 results
Welcome
Please turn off your camera and mute your microphone
Questions?
Visit www.sli.do and enter code 71547 to ask your questions
Technology problems?
Email sdr.stakeholder.engagement@ons.gov.uk
Starting at 10:15am
#Census2021

2. 2
Speaker
Pete Large
Joint Head of Quality Assurance of Census
Data
Census Statistical Design
#Census2021 www.sli.do 71547

3. 3
Aims for this session:
• Explain how the quality assurance of data contributes to the overall quality
of the census results
• Describe how we will assure both the census processes and the final
estimates
• Share our plans for comparing the census results to other sources of data
• Signpost where more information is available on our plans
• Invite suggestions of further checks or information on quality that should be
published
#Census2021 www.sli.do 71547

4. 4
Why are we running these webinars?
• To explain how the census works in collecting information
and producing great statistics
• To sign-post where more information is available
• To follow on from the material we published in October
• COVID response
• Statistical Design
• Findings from 2019 rehearsal
#Census2021 www.sli.do 71547

5. 5 #Census2021 www.sli.do 71547

6. 6
Census 2021 Quality Targets
High quality, flexible,
timely, accessible census
statistics for users
94% overall
response
At least 80% in
every local authority
75% Online
Response
Minimise variability
in response
Support
completion
#Census2021 www.sli.do 71547

7. 7
The Census 2021 results will
only be of value when they are
fit for purpose and when they
are trusted by users as a basis
for decisions
#Census2021 www.sli.do 71547

8. 8
Three strands of Quality Assurance
• Quality Assurance of Census processes
• Validation of the Census results
• Providing information on quality
#Census2021 www.sli.do 71547

9. 9
Quality Assurance of Census processes
Even with well designed processes, error can be
introduced
 sending out questionnaires
 completing questionnaires
 removing multiple responses
 coding write in answers
 correcting for non-response
#Census2021 www.sli.do 71547

10. 10
Validation of the Census results
• Population Estimates
- National and local authority
- Comparator sources
- Demographic analysis
• Topic analysis
#Census2021 www.sli.do 71547

11. 11
Population estimates: Comparator sources
Include:
• 2011 Census
• Mid-year population estimates
• Admin-based population estimates
• NHS Personal Demographic Service
• Benefits/Pensions data
• School census
• Higher Education Statistics Agency
• Council Tax data on households
#Census2021 www.sli.do 71547

12. 12
A demonstration of
standard population
estimates checks
#Census2021 www.sli.do 71547

13. 13
Standard checks
cover
• Total population
• Sex and age
• Households
• Communal
establishments
• Selected topics
• Migration
• Communities
• Local knowledge
(Fictional data)
#Census2021 www.sli.do 71547

14. 14
Total population
(Fictional data)
#Census2021 www.sli.do 71547

15. 15
Mortality Ratios
(Fictional data)
#Census2021 www.sli.do 71547

16. 16
Lower Super Output Area populations
(Fictional data)
#Census2021 www.sli.do 71547

17. 17
Sex ratios
(Fictional data)
#Census2021 www.sli.do 71547

18. 18
Population pyramids
(Fictional data)
#Census2021 www.sli.do 71547

19. 19
Household size
(Fictional data)
#Census2021 www.sli.do 71547

20. 20
Large Communal Establishments
(Fictional data)
#Census2021 www.sli.do 71547

21. 21
Communities
(Fictional data)
#Census2021 www.sli.do 71547

22. 22
Local information
(Fictional data)
#Census2021 www.sli.do 71547

23. 23
Quality assuring topic results
• Univariate distributions (national/LA)
• Spikes in small areas
• Age-sex profile/standardised measures
• Key multivariate relationships
• Write-in answers
• Expert Quality Assurance
#Census2021 www.sli.do 71547

24. 24
Signing off the data
• Quality Assurance in parallel with processing
• Independent of any contingency methods
• Evidence taken to Quality Assurance Panels
• National population estimates
• Population estimates for each local authority
• Results for each topic
• Final sign-off by National Statistician
#Census2021 www.sli.do 71547

25. 25
How we developed our plans
• Lessons from previous censuses
• Other census organisations
• Proposed approach published in January 2020
• Local Authority Quality Assurance Working Group
• Local Authority Partnership Plan
• … always open to new ideas …
#Census2021 www.sli.do 71547

26. 26
Publishing information on quality
• Response rates and confidence intervals
• Comparisons with other sources
• ‘QA packs’ for local authorities
• Census Quality Survey report
#Census2021 www.sli.do 71547

27. 27
Other aspects of the quality assurance
• Data science outlier identification
• Spurious responses
• Census Quality Survey
#Census2021 www.sli.do 71547

28. 28
More information
• Census Statistical Design
• Quality Assurance of the Census data
#Census2021 www.sli.do 71547

29. 29
Summary
• Building on successful approaches from previous Censuses
• Checking processes, population estimates and topic results
• Standard checks completed for all local authorities
• Using local and topic expertise outside the Quality Assurance
team
• Published outputs to help you understand the quality of the
results
#Census2021 www.sli.do 71547

30. 30
Quality Assuring the
Census 2021 results
Any questions?
Visit www.sli.do and enter code 71547
Please complete our evaluation survey and let us know how we did.
If you have any questions email us on SDR.stakeholder.engagement@ons.gov.uk.
#Census2021