National identity, ethnic group, language and religion question development for Census 2021

In December 2018, the government presented to Parliament a White Paper Help Shape our Future: The 2021 Census of Population and Housing in England and Wales (PDF, 967KB). This outlined the Office for National Statistics’s (ONS) recommendations to collect information on the topics of national identity, ethnic group, religion and language. These data are used for service planning and meeting duties under the Equality Act 2010.

We have reviewed the 2011 Census questions for Census 2021 and will collect this data in a way that is consistent with previous censuses. The questions used to collect data on these topics are:

• national identity
• ethnic group
• religion
• main language
• English language proficiency (ability to speak English)
• Welsh language skills

Census 2021 will be an online-first census, with a target of 75% online returns. We have reviewed and tested each question online and have recommended additional design and functionality amendments specific to the electronic questionnaire. We have also recommended additional changes to the question designs following our stakeholder engagement, evaluation and testing programmes.

The main recommended changes to question designs are as follows:

• on the national identity question, the response option “British” has been moved to the top of the response options on the questionnaire in England only
• the ethnic group question is presented in multiple stages on the electronic questionnaire
• the ethnic group question includes a new tick-box for “Roma”
• the ethnic group question includes a write-in response for “other African background”
• on the ethnic group question, the examples “Black Welsh” and “Asian Welsh” have been added to the high level ethnic group category descriptions on the Welsh questionnaire only
• the questions on language have been moved to after the question on religion

Since the publication of the White Paper, the Office for National Statistics (ONS) has concluded the final phase of testing on the topics of national identity, ethnic group, religion and language.

This report provides links to previously published research and the findings of additional testing that led to the final recommended questions for Census 2021 in England and Wales. The questions and response options for Census 2021 have now been finalised through the census secondary legislation: the Census (England and Wales) Order 2020, and Census Regulations for England and for Wales.

The evidence base for the recommendations made in the White Paper is discussed in the section Research that led to the 2018 White Paper recommendations. The evidence base for other changes is discussed in the section Research that led to the recommended Census 2021 question design.

In June 2015, the Office for National Statistics (ONS) held a formal, 12-week consultation process asking census users for their views on the topics that were required in the questionnaire in England and Wales. The aim of the consultation was to promote discussion and encourage the development of strong cases for topics to be included in Census 2021.

In May 2016, we published our response to the 2021 Census topic consultation. This set out our updated view on the topics to be included in Census 2021, including:

• a summary of proposals for new topics
• next steps
• an overview of our plans

The 2021 Census topic consultation revealed a clear requirement for information on the topics of national identity, ethnic group, religion and language for service provision and to meet duties under the Equality Act 2010.

A detailed summary of the consultation responses relating to these topics can be found in the following topic-specific consultation reports:

• Ethnicity and national identity (PDF, 850.5KB)
• Religion (PDF, 799KB)
• Language (PDF, 687.1KB)

In these reports, the ONS committed to collecting information on these topics. We also committed to reviewing the need for additional response options, the online presentation, and the guidance. We have provided an update on how we met these commitments in Annex 1.

In 2016, we began a comprehensive programme of research and development. A full list of the tests used in the development of the questions on national identity, ethnic group, religion and language is shown in Annex 2. Further details are provided in the summary of testing for Census 2021.

The tests used a range of qualitative and quantitative research methods. A short description of the different research methods and sampling techniques is given in the question and questionnaire development overview for Census 2021. Testing included participants from a wide range of backgrounds, purposely chosen to allow the research aims to be met. For example, testing of the ethnicity questions included people from a range of ethnic groups and religions, and testing of the main language question included people whose main language was not English, people who are bilingual, and Welsh language speakers.

Between November 2016 and January 2017, the ONS ran the “Ethnic group stakeholder follow-up survey” to gain a deeper understanding of user need for ethnic group information. We published the summary of responses in March 2019.

In December 2017, we published a further census topic research update, which explained how we evaluated 55 requests for additional ethnic group tick-boxes. Four ethnic groups that were taken forward for further stakeholder review and testing.

The 2018 White Paper outlined our proposals for the topic of national identity, ethnic group, religion and language. These were to:

• ask the same questions as used in the 2011 Census
• add a “Roma” tick-box to the ethnic group question
• improve the information on Somali and other African population by use of a write-in box
• develop search-as-you-type functionality for the ethnic group, national identity and religion questions

Alongside the White Paper we published a further census topic research update. This provided additional details of the research that supported the topic recommendations announced in the White Paper.

In June 2019, the ONS published The ethnic group prioritisation tool: 2021 Census in England and Wales (PDF, 220KB), which explained the evaluation of requests for additional ethnic group tick-boxes for Census 2021. The decisions made as a result of this evaluation, and an analysis of the prioritisation tool, have been published in the information paper: The ethnic group tick-box prioritisation: Census 2021 in England and Wales.

The questions for Census 2021 are now finalised. We have evaluated the question for its potential impact on data quality, public acceptability, respondent burden, financial concerns and questionnaire mode. We present details of this evaluation in Annex 3.

Current ongoing work around this topic focuses on finalising the search-as-you-type functionality used on the majority of these questions.

The starting point for reviewing the questions was the 2011 Census question designs. You can see these questions on the 2011 Census Household Questionnaire (PDF, 1.8 MB).

As set out in the White Paper Help Shape our Future: The 2021 Census of Population and Housing in England and Wales (PDF, 967KB), with further details provided in the December 2017 and December 2018 census topic research updates, we reviewed the response options for these questions and concluded that we should:

• add a Roma response option within the “White” category of the ethnic group question
• collect more detailed information on Somali and other African populations within the response option “African”

We also committed to using a search-as-you-type functionality in the electronic questionnaire to make it easier for respondents to self-identify their ethnic group, national identity and religion as they wish.

The work leading to these recommendations is summarised in Annex 1, this section focuses on how these decisions were then implemented within the questions.

References to tests are provided in the form (Year: Test number). For example, the fifth test conducted in 2017 is referenced as (2017:5).

National identity and ethnic group

A question on national identity was introduced in the 2011 Census in England and Wales. The question was developed to improve public acceptability of the ethnic group question by allowing respondents to express where they feel they belong, such as the country or countries they think of as home, irrespective of ethnic group, citizenship or country of birth.

A question on ethnic group has been asked since the 1991 Census. The Equality Act 2010 defines race as including colour, nationality, or national or ethnic origins. We collect information on ethnic group in the Census of England and Wales, rather than race. A person may also self-define their ethnic group to include their country of birth, language spoken at home or religion.

These data are used:

• for resource allocation and service delivery by central and local government
• to inform policy development and equality monitoring
• to help organisations meet their statutory obligations under the Equality Act 2010

The national identity question increases the public acceptability of the ethnic group question by allowing respondents to express their identity as British, English, Welsh , Scottish, Northern Irish or Other (write in). In the 2011 Census, the two questions were asked on a single screen on the electronic questionnaire and in a single column on the paper questionnaire to collect different aspects of a person’s identity.

2011 Census national identity question

How would you describe your national identity?
Tick all that apply
[ ] English
[ ] Welsh
[ ] Scottish
[ ] Northern Irish
[ ] British
[ ] Other, write in
(Write in national identity)
On the Welsh questionnaire, the response option “Welsh” came before “English”.

2011 Census ethnic group question

What is your ethnic group?
Choose one section from A to E, then tick one box to best describe your ethnic group or background
A White
[ ] English/Welsh/Scottish/Northern Irish/British
[ ] Irish
[ ] Gypsy or Irish Traveller
[ ] Any other White background, write in
(Write in ethnic group)

B Mixed/multiple ethnic groups
[ ] White and Black Caribbean
[ ] White and Black African
[ ] White and Asian
[ ] Any other Mixed/multiple ethnic background, write in
(Write in ethnic group)

C Asian/Asian British
[ ] Indian
[ ] Pakistani
[ ] Bangladeshi
[ ] Chinese
[ ] Any other Asian background, write in
(Write in ethnic group)

D Black/African/Caribbean/Black British
[ ] African
[ ] Caribbean
[ ] Any other Black/African/Caribbean background, write in
(Write in ethnic group)

E Other ethnic group
[ ] Arab
[ ] Any other ethnic group, write in
(Write in ethnic group)
On the Welsh questionnaire, Welsh was shown before English under Section A.

Displaying the national identity and ethnic group questions separately

Following changes to Government Digital Standards, the online form shows one question at a time, the national identity question followed by the ethnic group question. Because of space constraints on the paper questionnaire, the two questions appear in separate columns on the same page.

To aid understanding of the national identity question, when seen separately to the ethnic group question, we included a definition of the term “National identity” within a guidance accordion in testing up until the 2019 Rehearsal (2019:15).

In testing (2016:7), respondents expressed that acceptability of the ethnic group question could be improved with the addition of guidance on what the data are being used for. A guidance accordion was therefore added to explain this.

Developing the online version of the ethnic group question

The ethnic group question is the longest individual question on the census questionnaire. We investigated whether the question could be answered easily and accurately using a range of mobile devices. We conducted user experience (UX) testing to cover three main approaches.

Approach 1: Response options presented in a list

We tested (2016:1) a one-stage question design with all response options presented in a list. This was a similar format to the 2011 Census question. Participants found the amount of scrolling needed to view all of the response options frustrating.

Approach 2: Open write-in response

We developed an open write-in response with search-as-you-type functionality on the online questionnaire mode. This would provide suggested responses as respondents typed in their answer. Three versions of the question were tested (2016:7) with participants from different ethnic groups.

The test questions were:

• the 2011 Census design
• a one-stage design with an open question and no tick-boxes
• a two-stage design; the first stage with tick-box response options, and then a second stage question with a write-in box underneath for respondents to self-identify

Participants who were sure of their own ethnic group liked the open write-in response and considered the search-as-you-type functionality to be helpful on both designs. Participants who did not fully understand the question’s intent or the concept of ethnic group found the one-stage question difficult to answer on the electronic questionnaire as there were no visible response options.

Some participants responded using a paper version of the question with an open write-in response. The lack of search-as-you-type functionality on the paper questionnaire led to a high number of input errors, such as spelling errors, in the open write-in response.

Quantitative testing of the two-stage design (2017:3) found that the use of an open write-in response led to a significantly lower response to the second stage of the question. This would lead to lower quality statistics and so we stopped development of this question design.

Approach 3: Two-stage question with tick-box response options

We also tested (2016:3) a two-stage question design. The first question was “What is your ethnic group?” Respondents would choose their high-level ethnic group from a list of response options at the first stage, then a second question “What is your ethnic group?” would be shown underneath, with the more focused ethnic group response options from the 2011 Census. This design performed better than the one-stage design, but participants felt they were being asked the same question twice.

This was quantitatively tested against the two-stage open write-in approach (2017:3). The two-stage question design with tick-boxes on the first and second stages was widely considered to be the easiest to answer and had the lowest respondent burden. We decided to continue developing the question using this approach, and to investigate how to make the difference between the two questions clearer.

Acceptability of colour terminology

The use of colour terminology is believed to be essential to capture information on visible minority populations, and for collecting the data required to meet user needs. In the 2011 Census, colour terminology was used in the ethnic group question high-level categories and some response options. However, the terms people use to describe their ethnic group can change over time. The ONS aims to use terminology that is clear, understandable, acceptable and relevant.

We conducted a public acceptability test (2017:9) to understand public opinion on the acceptability of the terms used in the ethnic group question. The majority of participants were comfortable with the colour terminology used for the high-level categories, but a small proportion of participants felt uncomfortable with some of the terminology, including “White”, “Black” and “Mixed”.

We commissioned focus group research (2018:17) to:

• establish if colour terminology was still acceptable for Census 2021
• uncover any specific ethnic group terminology preferences
• assess the implications for data quality and comparability if changes to colour terminology were made

The focus groups were sampled according to self-identified ethnic group, and included participants who identified as "Black, Black British, African or Caribbean”, “White” or “Mixed or multiple ethnic group”. We tested five variants of the ethnic group question using a variety of colour and non-colour terminology. When non-colour terminology was used, the research found that:

• respondents could not easily locate their ethnic group under the high-level ethnic group categories, resulting in confusion, errors and an increase in multiple responses
• many Black and Black British participants identified using colour terminology, and found its removal unacceptable, viewing it as denying them an aspect of their identity
• while some participants identified as “European”, the term was viewed with suspicion as being “a mask for Whiteness”

Some variants also split African and Caribbean into separate high-level categories, which aligns with the approach taken by National Records of Scotland (NRS). Research found this negatively impacted on the clarity of the question for Black British participants, causing some to enter their ethnic group in the “Other” category.

This testing supported retaining the current high-level categories and continuing the use of the existing colour terminology for Census 2021.

Participants in the focus groups also commented that it was important to be able to re-assert their national identity within the ethnic group question by being able to identify as “Black British”. In the 2011 Census, “Black British” was included within the higher-level category description only. We have recommended including “Black British” in the “Any other” statement for Census 2021 as a prompt for respondents.

Using commas or slashes to separate examples

The 2011 ethnic group question used slashes to separate examples in the response options, for example, “Black/African/Caribbean/Black British”. These made the questions shorter and allowed the Welsh language response options to fit in the same space on the paper questionnaire. However, slashes are not accessible for screen readers and other assistive technologies.

We commissioned an external research agency to conduct cognitive interviews (2017:16) with participants from a range of ethnic groups to assess if it was appropriate to use slashes for Census 2021. We tested two versions of the question: one with slashes and one with commas.

Some participants found the slashes helpful, while others said they felt “divisive”. Most participants either expressed no preference or preferred commas to slashes as they were easier to read. Therefore, we have recommended using commas rather than slashes to separate the ethnic group examples, as they are better suited for screen readers and other assistive technologies.

Improving the “Mixed and Multiple ethnic group” responses on the paper questionnaire

When we reviewed the responses to the 2011 Census, we found that respondents who identified as Mixed ethnicity were more likely to respond across multiple categories and tick-boxes. We also looked at the average length of write-in response to the 2011 Census online question and it was longer than the 17 characters provided by a single line on the paper questionnaire write-in response. This meant respondents were unable to write their response in full.

We conducted focus groups (2018:17), which allowed us to investigate if providing two write-in lines on the paper questionnaire for the “Mixed and Multiple ethnic group” category could reduce instances of multi-ticking.

Participants initially stated a preference to remove the tick-box responses for the “Mixed and Multiple ethnic group” category entirely. However, when the tick-boxes were removed in practice, participants were confused about what information should be included in the write-in response.

Positioning of the “Roma” tick-box

Work on reviewing the ethnic group response options concluded that we should add a “Roma” tick-box. We conducted informal interviews and focus groups (2018:17) with members of the Roma community to test the positioning of a “Roma” tick-box response under different high-level ethnic group categories on the paper questionnaire. The findings from this research recommended that the “Roma” tick-box was placed under the “White” category, where it was more visible for respondents with low literacy levels, and close to the “Gypsy and Irish Traveller” response.

We conducted qualitative research (2018:25) to consider if a list of examples at the first stage of the electronic question could be used to direct Roma respondents to the “Roma” tick-box under the “White” category.

Four question variants were tested, with different examples for the “White” category:

• variant 1: “Includes any White background”
• variant 2: “Includes British, European or any other White background”
• variant 3: “Includes British, Roma or any other White background”
• variant 4: “Includes British, Northern Irish, Irish, Gypsy, Romany Gypsy, Irish Traveller, Roma or any other White background”

Respondents were most comfortable with variant 3 or 4, where Roma was explicitly listed under the “White” category, and found it to be the easiest to answer.

Following this testing, we have recommended an improved version of variant 4, by adding the “Roma” tick-box within the “White” category, and including “Roma” in the list of examples on the first stage of the electronic questionnaire, removing “Romany Gypsy”.

Improving the data collected on the “African” ethnic group

Work on reviewing the ethnic group response options concluded that we should not add a “Somali” tick-box, but that we should collect more detailed information on the “African” ethnic group.

We conducted cognitive interviews (2018:30) with participants who identified as Caribbean, Somali and Black African to see if adding a write-in response would improve the ethnic group data. While some participants would write in “Black African”, rather than provide more detailed information (for example, “Somali”), the new question design captured more detail on African identities than the 2011 Census. Therefore, we have recommended an open write-in response for the “African” tick-box response.

On the electronic questionnaire, respondents selecting “African” ethnic group are provided with a separate write-in box.

Because of space constraints, on the paper questionnaire respondents who select the African response option are instructed to use the same write-in box as those selecting the Other Black, Black British or Black Caribbean tick-box. To make this clearer to respondents we have reordered the tick-boxes so that “African” comes after “Caribbean”.

Religion

A question on religion was first asked in the 2001 Census. The census has historically measured religious affiliation. This is a measure of how a person connects or identifies with a religion, whether or not they actively practice it.

“Religion or belief” is a protected characteristic in the Equality Act 2010. The census does not collect information on the belief aspect of the protected characteristic.

The data collected from this question provide users with population data to inform equality monitoring and meet the requirements of the Equality Act 2010. This question has been made voluntary through the Census (Amendment) Act 2000.

In our response to the topic consultation, we stated that we intended to use the same question as in 2001 and 2011 to collect information on respondent’s religious affiliation. This was to allow comparability between censuses.

2011 Census question

What is your religion?
This question is voluntary
[ ] No religion
[ ] Christian (including Church of England, Catholic, Protestant and all other Christian denominations)
[ ] Buddhist
[ ] Hindu
[ ] Jewish
[ ] Muslim
[ ] Sikh
[ ] Any other religion, write in
(Write in religion)

On the questionnaire in Wales, the response option for “Christian” is followed by the example text “(all denominations)”.

Testing grammatically correct response options

In the 2011 Census, the response options to the religion question were affiliate oriented, for example “Christian” and “Muslim”, rather than “Christianity” and “Islam”. Testing of the 2011 Census question showed that participants expected to see these affiliate labels, even though they are grammatically incorrect responses to the question “What is your religion?”

As part of the focus testing for the ethnic group question (2017:21), we tested to see if this was still true.

As in the 2011 Census testing, participants preferred the grammatically incorrect version, as it reflects how people self-identify. Jewish and Sikh participants identified that changing from affiliate-oriented to religion-oriented responses suggests that respondents should only tick if they are actively engaging and practicing the religion.

The religion question is intended to capture affiliation. Any change in the implied concept of the response options could affect comparability with data from previous censuses, as fewer respondents would select the same response options. Therefore, we have recommended continuing to use the grammatically incorrect, affiliate-oriented response options.

We have recommended that the religion question remains the same as the 2011 Census question design for Census 2021.

Welsh language skills

A question on Welsh language skills has been included in every census in Wales since 1891. Language knowledge and skill covers whether a person can understand, speak, read or write a language, without reference to how well they can do it.

The need for data to monitor and promote the use of the Welsh language has been strengthened because of obligations under the Welsh Language Act 1993, which places a duty on public organisations in Wales to treat the Welsh and English languages equally, and the Welsh Language Standards (No.1) Regulations 2015, which requires Welsh Ministers and public authorities to take account of the Welsh language when developing policies and services.

2011 Census question

Can you understand, speak, read or write Welsh?
Tick all that apply.
[ ] Understand spoken Welsh
[ ] Speak Welsh
[ ] Read Welsh
[ ] Write Welsh
[ ] None of the above

This question was only asked on the 2011 Census questionnaire in Wales.

Asking the Welsh language skills question on the English questionnaire

In the 2015 topic consultation, we received a request from data users to collect Welsh language skills data in England as well as Wales. We conducted cognitive testing (2017:21) to establish the feasibility of asking the 2011 Census question for Wales on the questionnaire for England. Some participants were surprised to be asked the question, but most found it easy to answer.

We quantitatively tested (2017:22) the Welsh language skills question in England to discover the proportion of the English population to which a Welsh language skills question would apply. Around 1% of participants reported some level of Welsh language skill.

Asking a Welsh language skills question in England increases burden for a majority of respondents, by asking an additional question to meet a localised user need on a national questionnaire. In a later review of consultation responses, the user need for Welsh language skills in England was evidenced as not sufficient. Therefore, we have recommended that the Welsh language skills question will continue to be asked in Wales only.

Asking four separate questions on the electronic questionnaire

We tested (2019:6) a four-stage question design for the electronic questionnaire. Participants were asked about each individual Welsh language skill, with “Yes” or “No” response options after each question.

This question design indicated no impact on respondent burden. However, the responses were inconsistent with the 2011 Census question design. This could be indicative that respondents with lower skill levels were more likely to tick “Yes” for this question design than they would on the 2011 Census question. This approach would impact data comparability with previous censuses. Therefore, we have recommended keeping the 2011 Census question design of a single, multi-tick question.

Mutually exclusive responses

We tested (2019:6) the 2011 Census Welsh language skills question design with mutually exclusive functionality to see if this had any impact on data quality. If the respondent ticks any response option and then ticks “None of the above”, any previously selected tick-boxes are automatically deselected.

Testing showed that including mutually exclusive functionality had no impact on data quality. Therefore, we have recommended that this functionality be used. We also added an “or” to the paper questionnaire to emphasise that the “None of the above” option is mutually exclusive from the other response options.

Main language and English proficiency

The questions on main language and English proficiency were introduced in the 2011 Census. A person’s main language is the language they use most naturally. For example, this could be the language they use at home.

Data from these questions are used to identify people whose first language is not English. This helps councils and other organisations to plan language support, translation and study programmes at a local level, to promote integration and eliminate discrimination.

2011 Census language questions

What is your main language?
[ ] English
[ ] Other, write in (including British Sign Language)
(Write in main language)

On the Welsh questionnaire, the first response option was “English or Welsh”.

How well can you speak English?
[ ] Very well
[ ] Well
[ ] Not well
[ ] Not at all

In the 2011 Census paper questionnaire, the response options for the English proficiency question were displayed horizontally because of space constraints. We did not identify a strong user need for changing the question on English proficiency. Therefore, we will use the 2011 Census question design in Census 2021.

Defining “main language”

As testing had showed that bilingual respondents found it difficult to answer the question, we added guidance in an expandable accordion (stated previously) to the electronic questionnaire for Census 2021. Respondents can click the accordion text “What do we mean by “main language”?” to open a definition of “main language”.

This section provides the details on how the questions on national identity, ethnic group, religion and language continued to evolve following the publication of the 2021 Census White Paper as a result of respondent feedback from the 2019 Rehearsal.

Ethnic group and national identity

After the 2019 Rehearsal (2019:40) we reviewed the national identity guidance accordion defining “national identity” and refined the ethnic group question wording in Wales.

We tested the re-designed questions in qualitative testing (2020:1) across locations in England and Wales. The aims were:

• to assess whether making the instruction under the question stem, defining “national identity” visible at all times helps respondents to answer the national identity question
• to understand if respondents could easily identify with an amended ethnic group design that incorporates “Welsh” into the Asian or Asian British and Black, Black British, African and Caribbean high-level categories

Testing the national identity definition in a guidance accordion

We tested two versions of the national identity question: one with the guidance in an accordion, and one with the guidance visible as an instruction.

Participants did not click on the guidance accordion to read it when it was not immediately visible but did find the instruction helpful when it was visible.

We have therefore recommended moving the instruction out of a guidance accordion and into the question description. We also used participants’ views on the precise wording used to further increase the clarity of the instruction.

Ethnic group high-level category descriptions

We tested four versions of the two high-level category descriptions at the first stage of the ethnic group question:

• version 1: “Asian or Asian British” and “Black, Black British, Caribbean or African”
• version 2: “Asian, Asian Welsh or Asian British” and “Black, Black Welsh, Black British, Caribbean or African”
• version 3: “Asian or Asian British, including Asian Welsh” and “Black, Black British, including Black Welsh, Caribbean or African”
• version 4: “Asian or Asian Welsh, English, Scottish, Northern Irish or British” and “Black, Black Welsh, English, Scottish, Northern Irish or British, Caribbean or African” (in locations in England “English” was listed before “Welsh”)

When additional high-level identities were included in the response categories, respondents were more likely to record an identity that was inconsistent with how they described themselves. This was more likely depending on how many words were included in the response categories. The combination of ethnic minority and UK identity, such as “Black Welsh”, was noticed in England, but was rarely noticed by ethnic minority respondents.

Following this testing, we have recommended using the version 2 phrasing, with the high-level categories “Asian, Asian Welsh or Asian British” and “Black, Black Welsh, Black British, Caribbean or African”. These will appear on the questionnaire in Wales only.

National identity response option ordering

On the English questionnaire in England, the response option “British” has moved to the top of the national identity question to make it easier for people to respond to. On the English and Welsh questionnaires in Wales, “British” will not be moved, and the first response option will be “Welsh”.

Clear selection option on the religion question

Respondents who answer the religion question should only select one response. On the electronic questionnaire, this is controlled using radio buttons, rather than tick-boxes. Tick-boxes allow multiple selections to be made, while radio buttons only allow a single response option to be selected.

Because the religion question is voluntary, it must be possible for respondents to de-select their response if they wish to continue without providing an answer. We have developed the option to clear the radio button selection by including an additional “Clear selection” button. This will clear previously submitted answers.

Search-as-you-type functionality

The questions on ethnic group, national identity, main language and religion all feature open write-in responses.

These are presented on a separate page to the main question and are still under development. This is because including the functionality on a page with tick-boxes and write-in boxes reduces its accessibility to respondents using assistive technology such as screen readers. Also, we tested (2019:13) a write-in box that is not immediately visible and found it had lowered acceptability. We concluded that to have a write-in box on the second screen, users have to be clearly directed to this screen from the “Other” response options.

We are continuing to develop search-as-you-type functionality for these questions on the electronic questionnaire. Respondents can start typing a response, and a list of coded responses will be shown. The respondent can either click one of these coded responses, or continue typing their own response.

We are continuing to develop the search-as-you-type functionality for Census 2021.

Question design for online

In addition to the testing described earlier, the national identity, ethnic group, religion and language questions have undergone significant user experience (UX) testing (2017:2, 2018:2, 2019:1, 2020:2). UX testing focuses on understanding user behaviours as people interact with online services. Through observation techniques, task analysis and other feedback methodologies, it aims to develop a deep knowledge of these interactions and what it means for the design of a service.

UX research has taken place on a rolling basis since 2017. Before the Census Rehearsal in October 2019, 458 interviews have been conducted at 99 events. The UX testing programme will continue through 2020. All participants are purposively selected to include a wide range of ages and digital abilities.

UX testing included various iterations of the national identity, ethnic group, religion and language questions described elsewhere in this report. Feedback from this research informed decisions made on the design of these questions. For more information on UX testing, see the question and questionnaire development overview for Census 2021.

Welsh language question development

Between 2017 and 2018, an external agency with Welsh speaking researchers was commissioned to undertake focus groups (2017:17) and a series of cognitive interviews (2017:18). In 2018, further cognitive interviews (2018:41) were undertaken by the same agency. The qualitative research tested public acceptability and comprehension of amended and newly designed census questions in Welsh. The questions were tested with people across Wales with varying dialects and Welsh language proficiencies.

To ensure questions adhere to Cymraeg Clir guidelines, some changes to the text or questions across the census questionnaires were translated by our contracted specialist Welsh language translation service provider. These changes were quality assured by the Welsh Language Census Question Assurance Group (the Assurance Group). This group includes Welsh language and policy experts from the Welsh Language Commissioner and the Welsh Government and was convened to give advice on the accuracy, clarity and acceptability of the language as well as other policy issues pertaining to the Welsh language and bilingual design.

Following testing and reviews by the Assurance Group, the Welsh versions of the religion and language questions will remain the same as in the 2011 Census.

National identity

In the 2011 Census, the national identity question in Wales used gendered terms for national identities (for example “Cymro/Cymraes”). We tested (2017:17) the use of non-gendered terms (for example “Cymreig”) in focus groups. Testing showed that respondents preferred the non-gendered terms.

However, in correspondence with the Assurance Group, the decision was taken that the response options should remain the same as in the 2011 Census. This meant that the response options were grammatically correct, consistent with previous censuses and make a clear distinction between “Cymro/Cymraes” (Welshman/Welshwoman) as a national identity and “Cymreig” (Welsh) as an ethnic group.

Ethnic group

We focus group tested (2017:17) the two-stage electronic question design and the paper question design in Welsh. The recommended question designs tested well in further cognitive interviews (2018:41) and these were reviewed by WLCQAG. We have recommended the following changes to the question designs in Wales:

• to add “orau” (best) in “best describes” in the second-stage question stem and re-order the question stem
• to add “Cymraeg” (Welsh) to the list of “White” examples on the first stage of the Welsh electronic questionnaire
• to keep the English alphabet for each section, as changing the alphabet would have a wider impact on the questionnaire
• to add “Roma” (Roma) in Welsh as an additional tick-box
• to add the phrases “Asiadd Cymreig” (Asian Welsh) and “Du Cymreig” to the relevant high-level categories

The design of the questions recommended for Census 2021 was informed by the research and testing detailed in this report. The questions and response options for Census 2021 have now been finalised through the census secondary legislation: the Census (England and Wales) Order 2020, and Census Regulations for England and for Wales.

The paper questionnaires used in Census 2021 are available.

Guidance text and instructions are not part of the legislation, but we consider these to be finalised as well. However, it is possible that guidance text or instructions may change if there is enough evidence to support doing so.

The questions on national identity, ethnic group, language and religion all contain an option to write in an answer. These write-in screens are still under development and so are not shown in this report.

National identity

Figure 1: Online national identity question in English and Welsh

Download this image Figure 1: Online national identity question in English and Welsh

.png (238.1 kB)

The instruction “This relates to where you feel you belong, such as the country or countries you think of as home. This could be different from your citizenship or ethnic group” only appears on the electronic questionnaire. It does not appear on the paper questionnaire because of space constraints.

Ethnic group

Figure 2: Online ethnic group question in English and Welsh

Download this image Figure 2: Online ethnic group question in English and Welsh

.png (290.8 kB)

Routing

Online, the question is asked in stages. The answer to the first question informs which of the subsequent second stage options is displayed (Figures 3 to 7). Respondents can self-identify their ethnic group in stage 2, where they are routed to a third stage that contains a write-in box with search-as-you-type functionality. The third stage presents an open write-in box with search-as-you-type functionality. This final stage screen is under development and is not presented in this report.

Figure 3: Online White ethnic group or background question in English and Welsh

Download this image Figure 3: Online White ethnic group or background question in English and Welsh

.png (183.8 kB)

In Wales only, in the first tick-box, “Welsh” is first, “Welsh, English, Scottish, Northern Irish or British”, while in England, “English” is first.

Figure 4: Online Mixed or Multiple ethnic group or background question in English and Welsh

Download this image Figure 4: Online Mixed or Multiple ethnic group or background question in English and Welsh

.png (195.0 kB)

Figure 5: Online Asian or Asian British ethnic group or background question in English and Welsh

Download this image Figure 5: Online Asian or Asian British ethnic group or background question in English and Welsh

.png (180.7 kB)

Figure 6: Online Black, Black British, Caribbean or African ethnic group or background question in English and Welsh

Download this image Figure 6: Online Black, Black British, Caribbean or African ethnic group or background question in English and Welsh

.png (211.9 kB)

Figure 7: Online other ethnic group or background question in English and Welsh

Download this image Figure 7: Online other ethnic group or background question in English and Welsh

.png (125.6 kB)

Religion

Figure 8: Online question on religion in English and Welsh

Download this image Figure 8: Online question on religion in English and Welsh

.png (141.1 kB)

In Wales, the included example underneath “Christian” will be “including all denominations” as per the 2011 Census religion question.

Routing

On the paper questionnaire, the open write-in response appears beneath the “Any other religion” tick-box. On the electronic questionnaire, the open write-in response is displayed on a separate page and features search-as-you-type functionality. The second stage is under development, and as such is not presented in this report.

Language

Figure 9: Online Welsh language skills question in Wales, in English and Welsh

Download this image Figure 9: Online Welsh language skills question in Wales, in English and Welsh

.png (129.4 kB)

Figure 10: Online main language question in English and Welsh

Download this image Figure 10: Online main language question in English and Welsh

.png (101.9 kB)

In Wales, the English language version changes the first tick-box from “English” to “English or Welsh” because of user need for both the official languages in Wales to be included. In the Welsh language questionnaire, the first tick-box has “Welsh” first “Cymraeg neu Saesneg”.

Figure 11: Online English proficiency question in English and Welsh

Download this image Figure 11: Online English proficiency question in English and Welsh

.png (70.7 kB)

The English proficiency question response options are presented horizontally on the paper questionnaire because of space constraints and vertically on the electronic questionnaire.

On the electronic questionnaire, the Welsh language skills question includes mutually exclusive functionality to prevent respondents ticking multiple response options and “None of these apply”.

Routing

Respondents aged under three years old do not have to answer any of the language questions. On the paper questionnaire, this guidance is provided in the separate instruction booklet, because of space constraints. On the electronic questionnaire automatic routing, based on their answer to the date of birth question, means they will not be presented with the questions.

Only respondents in Wales will be asked the Welsh language skills question. On the paper questionnaire in England, this question is not shown, and the space where it would appear reads “This question is intentionally left blank” with routing to the next question.

Only respondents who have answered “Other, including British Sign Language” to the main language question will be asked the English proficiency question.

The “Other, including British Sign Language” on the electronic questionnaire routes to a second stage where respondents are able to write their “Other” language with search-as-you-type functionality. The second stage screen is under development, as such is not reported in this report.

Question order

In the 2011 Census, the question order was:

• national identity
• ethnic group
• Welsh language skills (in Wales only)
• main language
• English proficiency
• religion

This ordering maintained these questions as a set, and fulfilled the Government Statistical Service recommended requirement for the national identity question to be asked before the ethnic group question.

We tested (2019:6) moving the language questions in two different locations. One was positioned after the questions on national identity and ethnicity, and the other was positioned before the questions on national identity and ethnicity. The testing found that respondents were more likely to answer the Welsh language skills question if it was positioned after the question on national identity.

In the 2019 Rehearsal (2019:15), we positioned the suite of language questions before the ethnic group question. The results confirmed the findings of the earlier test, that respondents reporting having no Welsh language skills was lower than in the 2011 Census.

To improve the response to the Welsh language skills question, and keep the identity questions together, we have recommended moving the language suite of questions after the questions on national identity, ethnic group and religion. We therefore recommended that the topics should be ordered:

• national identity
• ethnic group
• religion
• Welsh language skills
• main language
• English proficiency

Question evaluation

The question designs put forward in this report are based on extensive research and assessment using evaluation criteria that were set out in The 2021 Census – Assessment of initial user requirements on content for England and Wales: Response to consultation (PDF, 796KB).

The evaluation considered the potential impact that including a topic on the census would have on data quality, public acceptability, respondent burden, financial concerns, and questionnaire mode. The evaluations were used in conjunction with the user requirements criteria to steer the development of the census questions and questionnaire.

A topic that has been assessed as having a “high” potential for impact is closer to the threshold for exclusion from the census than a topic that has been assessed as having a “low” potential for impact.

After completing the research and development phase, we evaluated the recommended questions against the same criteria using an updated tool that considers the type of evidence we have available and the Census 2021 context. A description of this updated evaluation tool is provided in the question and questionnaire development overview for Census 2021.

All questions meet our thresholds to ensure reliable information will be collected in Census 2021.

Table 2 provides the updated evaluation scores for the questions on national identity, ethnic group, religion and language.

For most questions, the measures have remained the same, or the potential for impacts has decreased. The exceptions are national identity and ethnic group; these questions have a heightened potential for impact for some measures because of the increasing complexity of the questions, and of the population’s interaction with these concepts. However, both questions still meet our criteria for inclusion in Census 2021.

Further details of the evaluation are provided in Annex 3.

Question harmonisation

As in previous censuses, there will be separate censuses in Scotland and Northern Ireland. The questions for England and Wales have been developed through close collaboration with National Records of Scotland (NRS) and the Northern Ireland Statistics and Research Agency (NISRA), who are responsible for conducting the censuses in Scotland and Northern Ireland, respectively.

Details of the differences between the questionnaires for England and Wales, Scotland, and Northern Ireland can be found in Annex 4.

We recognise that each country has its own user and respondent needs. However, we aim for harmonisation of census questions and topics where possible to produce UK-wide statistics that are consistent and comparable.

The national identity, ethnic group, religion and language questions were developed for use in the context of Census 2021 in England and Wales, a mandatory household form. Therefore, it is possible that in different contexts, such as social surveys, a different approach may be more suitable.

Recommended questions to use when collecting data on ethnic group, national identity, religion and language in social surveys can be found in the current harmonised principle.

The Office for National Statistics (ONS) will continue to develop the functionality for the electronic questionnaire, including search-as-you-type functionality for the national identity, ethnic group, religion and main language open write-in responses.

The Office for National Statistics (ONS) will engage with minority groups who have requested response options on the census to improve the data quality through the open write-in responses, and work with census data users to produce analytical products following the census that will meet their data needs.

In our response to the 2021 Census topic consultation, we made clear commitments to the public. We committed to continuing to ask questions on national identity, ethnic group, religion and language. We also made three further commitments:

• investigate how to present the ethnic group question on a range of devices
• review, test and evaluate additional response options
• review and improve the online guidance

More detail on each is provided in this section.

Investigate how to present the ethnic group question on a range of devices

The Office for National Statistics (ONS) investigated the best approach to presenting the ethnic group question online. Details of our user experience testing and online question development can be found in this report.

Review, test and evaluate additional response options

During the topic consultation, we received requests for additional tick-boxes for national identity, ethnic group, religion and main language. We evaluated each request as described below.

To help ensure that everyone can identify as they wish we have also added a search-as-you-type function to the electronic questionnaire.

Ethnic group response options

Following the topic consultation, we conducted an ethnic group stakeholder follow-up survey. The results were published in the summary of responses. The ONS reviewed and updated its 2011 prioritisation tool to evaluate requests for additional ethnic group tick-boxes. Details of how this was achieved are included in the information paper The ethnic group prioritisation tool: 2021 Census in England and Wales (PDF, 198KB).

In the December 2018 census topic research update we published our recommendation that all ethnic group tick-boxes from the 2011 Census are kept for Census 2021, with the addition of “Roma”. We would also add a write-in response for the “African” tick-box to allow for more detailed reporting of a respondent’s African ethnic group (for example, “Somali”).

We have published an information paper that provides the details of how the prioritisation tool was used to make these decisions: [LINK: Deciding whether to add tick-boxes to the Census 2021 ethnic group question in England and Wales].

National identity response options

We received a request to add “Cornish” to the national identity question for Census 2021. We evaluated the evidence provided, but there was not a strong user need for inclusion.

In parallel with the evaluation, we conducted cognitive interviews (2017:16) to evaluate the acceptability of adding regional identity tick-boxes, such as Cornish, to the national identity question. Participants considered the question to be asking about national identity, rather than regional identity, and reported that it would be unrepresentative to add particular regional response options, such as Cornish.

Based on the findings from the evaluation and testing, we have recommended that the national identity response options remain the same as in the 2011 Census. Respondents who wish to identify as Cornish or any other identities are able to do so using the open write-in response.

Religion response options

We received requests to include additional tick-box response options on the religion question, including one for “Jain”. Where evidence of a user need was provided we evaluated the requests on the strength of user need, including disadvantage and unmet needs within the community. We concluded that there was not a strong enough user need for a separate tick-box for any of the requested religions.

In Census 2021, people of all religions, for example Jains, will be able to identify their religious affiliation using the write-in response.

Main language response options

Following the topic consultation, the ONS worked with stakeholder groups to understand the need for additional response options for the main language question. We received evidence of the user need for:

• Welsh (separate from English, in Wales only)
• Cornish
• British Sign Language

We evaluated the requests for these response options against the user need for data for resource allocation, service delivery and policy development. We could not identify a strong user need to add tick-box responses for Cornish or British Sign Language. The data for these languages will be captured by the write-in response option.

We identified a user need to split the “English or Welsh” tick-box on the Welsh questionnaire only. This would produce outputs harmonised with England and explicitly count the number of Welsh main language users in Wales. We commissioned an external research agency to conduct cognitive interviews (2017:14) with Welsh speakers to evaluate a question with “English” and “Welsh” as separate response options. We wanted to understand whether the split response affected the acceptability and ease of answering the question, as well as the impact the split responses would have on responses to the Welsh language skills question.

Testing showed that splitting the response options was unlikely to have any impact on the Welsh language skills question. Nearly all participants were happy with the position of the two questions and would not change their response.

Most participants found it easy to choose between the two response options based on the language they speak most often. However, participants who were bilingual in English and Welsh found it difficult to answer the question. This included those who spoke English and Welsh equally, or who were raised speaking Welsh but speak English more often. These participants would have preferred to tick both response options.

As splitting the response could affect the ease of answering for bilingual respondents and impact on data quality, we have not recommended splitting the “English and Welsh” response for Census 2021.

Review and improve the online guidance

The ONS has reviewed the online guidance for all recommended questions. This guidance was used in the 2019 Rehearsal (2019:15) and the findings from this will be used to further refine the online guidance for Census 2021.

References to tests take the form (Year: Test number). “Year” refers to the calendar year the test was undertaken in and the test number is the position of the test within the year considering all testing that took place in that year. For example, the fifth test conducted in 2017 would be referenced as (2017:5).

A full description of each of these items can be found in summary of testing for Census 2021.

Evaluation of the national identity question

The national identity question has been assessed as having a “low” potential for impact on public acceptability and financial concerns.

Potential for impact on data quality: “Medium”

This question collects information that cannot be directly observed, which can cause data quality issues when answering on behalf of another person. This question is also subjective in nature.

The term “national identity” was not well understood. Feedback from the 2019 Rehearsal (2019:15) showed that a small number of respondents (2.24%) used the guidance accordion to help them answer the question. This will be addressed in Census 2021 by rephrasing the definition as an instruction that will be visible at all times. However, because of space constraints this is not present on the paper questionnaire.

Potential for impact on respondent burden: “Medium”

The national identity question is subjective, it collects information that cannot be directly observed, and contains a write-in field, which can increase respondent burden. However, on the electronic questionnaire respondents to this question will have a search-as-you-type function on the write-in screen to improve burden on using write-ins.

Potential for impact on questionnaire mode: “Medium”

The national identity question is subjective in nature, which could cause a potential impact on data captured between the electronic and paper questionnaire. This is because there is a tendency of some respondents to report an answer in a way they deem to be more socially acceptable on paper compared with online. The sensitivity and subjectivity of the question could heighten this difference.

On the electronic questionnaire we provide a guidance accordion to describe what we mean by “national identity”. On the paper questionnaire this is not present because of space constraints. The difference between modes means there is a “medium” potential impact on data. A small proportion of respondents opened the guidance accordion (2.28%) in the 2019 Rehearsal.

Evaluation of the ethnic group question

The ethnic group question has been assessed as having a “low” potential for impact on financial concerns.

Potential for impact on data quality: “High”

The ethnic group question has been assessed as having a “high” potential for impact on data quality. This has changed from the evaluation conducted in 2016, where this topic was evaluated as having a “low” potential for impact on data quality.

In testing, several participants were confused by the response options, as they encompass aspects of nationality, race, ethnicity and religion. A respondent could change their answer to the ethnic group question depending on how they feel about their ethnicity at the time of answering.

In addition, this question collects information that cannot be directly observed, which can cause data quality issues when answering on behalf of another person.

In the 2019 Rehearsal (2019:15), over 7% of respondents did not answer this question.

Potential for impact on public acceptability: “Medium”

In addition to having a high level of non-response, we also received some negative feedback around acceptability in the 2019 Rehearsal (2019:15) for the ethnic group question because of the topic’s subjectivity and sensitivity. Changes to the question design made after rehearsal aim to address these issues, however, they have not been quantitatively tested.

Potential for impact on respondent burden: “High”

On the paper questionnaire, the ethnic group question has an instruction and 19 response options, including open write-in responses. This is burdensome for respondents to read, particularly those who need to read the whole list to find the most appropriate response.

On the electronic questionnaire, the ethnic group question is presented as a multi-stage question. While this approach increases the total number of questions all respondents must answer. Respondents will only see the second-stage response options that are relevant to the category they selected in the first stage. Only those who do not identify with the existing tick-boxes will see the third write-in stage. This reduces the number of instructions and response options that the respondent must read to answer this question.

Potential for impact on questionnaire mode: “High”

The question is presented in stages on the electronic questionnaire, compared with one question with five ethnic group categories with response options on the paper questionnaire. Cognitive testing has found that the two approaches are unlikely to lead to respondents providing different answers, although some expressed a preference for the list design of the paper questionnaire. A guidance accordion has been included in both stages of the online question, but not the paper question because of space constraints. The accordion explains the need for ethnic group data.

Evaluation of the religion question

The ethnic group question has been assessed as having a “low” potential for impact on data quality, public acceptability and financial concerns.

Potential for impact on respondent burden: “Medium”

The religion question collects information that cannot be directly observed, which can be difficult to answer when a respondent is answering on behalf of another person.

In addition, the religion question has a write-in box where respondents are asked to self-identify their religion, adding an extra burden on respondents.

Potential for impact on questionnaire mode: “Medium”

The religion question in the paper questionniare differs to the electronic questionnare design. There is search-as-you-type functionality on the electronic questionnaire “Other” write-in screen, and so there is a potential for impact. The question is subjective and sensitive, so there is potential for impact on questionnaire mode, with the nature of this topic.

Evaluation of the Welsh language skills question

The Welsh language skills question has been assessed as having “low” potential for impact on data quality, public acceptability, respondent burden and financial concerns.

Potential for impact on data questionnaire mode: “Medium”

The Welsh language skills question has differences between the electronic and paper questionnaire versions. Online, there is mutually exclusive functionality where respondents can select their skill levels, or select none of these apply. This reduces respondent error of selecting both no skills and skills.

Evaluation of the main language question

The main language question has been assessed as having a “low” potential for impact on data quality, public acceptability, financial concerns and questionnaire mode.

Potential for impact on respondent burden: “Medium”

The main language question collects information that cannot be directly observed, which can be difficult to answer when a respondent is answering on behalf of another person.

In addition, the main language question has a write-in box where respondents are asked to self-identify their religion, adding an extra burden on respondents.

Evaluation of the English proficiency question

The English proficiency question has been assessed as having a “low” potential for impact on all five assessment criteria; data quality, public acceptability, respondent burden, financial concerns and questionnaire mode.

Throughout the development of the 2021 Census questionnaires, the Office for National Statistics (ONS), National Records of Scotland (NRS) and Northern Ireland Statistics and Research Agency (NISRA) have worked together to ensure harmonisation of the outputs from the national identity, ethnic group, religion and language questions where possible.

National identity

National identity question in Scotland

The national identity question in Scotland lists the same response options as in England and Wales, but with “Scottish” as the first response and “British” as the last response. The response options are listed in two columns on the paper questionnaire in Scotland, while they appear in one column in England and Wales.

National identity question in Northern Ireland

The national identity question in Northern Ireland lists the same response options as in England and Wales, with the addition of a response for “Irish”, which does not appear on the questionnaire in England and Wales. The response options are listed in three columns on the paper questionnaire in Northern Ireland because of space constraints, while they appear in one column in England and Wales.

Ethnic group

Ethnic group question in Scotland

The ethnic group question in Scotland follows a similar design to the question in England and Wales, but with six high-level ethnic group categories and different response options:

• the top-level categories are “White”, “Mixed or multiple ethnic groups”, “Asian, Scottish Asian or British Asian”, “African, Scottish African or British African”, “Caribbean or Black” and “Other ethnic group”
• the “White” category separates the response options for “Scottish” and “Other British”
• the “White” category includes additional response options for “Polish” and “Showperson”
• the “Mixed or multiple ethnic groups”, “African, Scottish African or British African” and “Caribbean or Black” categories include a write-in response with no tick-box response options
• the “Asian, Scottish Asian or British Asian” and “Other ethnic group” categories contain the same response options as the “Asian or Asian British” and “Other ethnic group” categories in England and Wales, but include Scottish and British examples against each response; for example, the “Pakistani” response option in England and Wales appears as “Pakistani, Scottish Pakistani or British Pakistani” in Scotland

Ethnic group question in Northern Ireland

The ethnic group question in Northern Ireland does not have high-level ethnic group categories. Instead, it lists 10 response options, including two write-in responses. The response options are:

• White
• Chinese
• Irish Traveller
• Roma
• Indian
• Filipino
• Black African
• Black Other
• Mixed ethnic group (write in)
• Any other ethnic group (write in)

On the electronic questionnaire, the write-in responses will appear on a separate screen.

Religion

Religion question in Scotland

The religion question in Scotland has a different question stem and response options to the religion question in England and Wales. As in England and Wales, the religion question is voluntary in Scotland.

The question stem is “What religion, religious denomination or body do you belong to?”.

The response options are:

• None
• Church of Scotland
• Roman Catholic
• Other Christian (write in)
• Muslim (write in)
• Hindu
• Buddhist
• Sikh
• Jewish
• Pagan
• Another religion or body (write in)

Religion questions in Northern Ireland

In Northern Ireland, there are two religion questions. The second question is only asked if the response to the first question is “None”.

The question stem for the first question is “What religion, religious denomination or body do you belong to?”. The question stem for the second question is “What religion, religious denomination or body were you brought up in?”.

The response options for both questions are the same:

• Roman Catholic
• Presbyterian Church in Ireland
• Church of Ireland
• Methodist Church in Ireland
• Other (write in)
• None

Language questions

Language questions in Scotland

The Welsh language skills question is not asked in Scotland. There is an additional language skills question “Can you understand, speak, read and write Scottish Gaelic or Scots language?”

The English language skills question features a bank of 16 tick-boxes in four columns and four rows. Across the top are “Understand (spoken)”, “Speak”, “Read” and “Write”. Down the left are “Very well”, “Well”, “Not well” and “Not at all”. This allows respondents to mark each language skill individually. This question is asked to all respondents, rather than only those respondents whose main language is not English or Welsh.

The main language question is the same as in England and Wales, but the example text reads “including ‘BSL’ and ‘TACTILE BSL’”. In England and Wales, it reads “including British Sign Language”.

Language questions in Northern Ireland

The Welsh language skills question is not asked in Northern Ireland.

There are two additional language skills questions “Can you understand, speak, read or write Irish?” and “Can you understand, speak, read or write Ulster-Scots?”. Each question includes five response options for “No ability”, “Understand”, “Speak”, “Read” and “Write”.

Respondents who tick “Speak” are then asked a follow-up question within the same question design: “How often do you speak Irish?” or “How often do you speak Ulster-Scots?”, with the response options “Daily”, “Weekly”, “Less Often”, “Never”.

The English language skills question is the same as in England and Wales.

The main language question is the same as in England and Wales, but the example text reads “including British/Irish Sign Languages”.