Using qualitative and quantitative methods to develop ethnic identity questions in the UK Alita Nandi and Lucinda Platt with thanks to Liz Spencer, Punita.

Using qualitative and quantitative methods to develop ethnic identity questions in the UK

Alita Nandi and Lucinda Platt

with thanks to Liz Spencer, Punita Chowbey, Heidi Mirza, Heather Laurie, Noah Uhrig, Emily Kean, Sarah

Budd and Alison Patterson and our respondents

Ethnic groups?

Ethnic group is a group of people who believe that they share a common descent based on real or imagined shared attributes (Weber, Schermerhorn, Anderson,..)

Defining one ethnic group requires defining the “other” ethnic group

Both commonality within groups and contrast with other groups are key to group recognition

SLLS Inaugural Conference 2010, Cambridge

Ethnic identity?

Ethnic identity: Ethnic identity is conceived of as being part of social identity (Phinney 1992)

“social identity will be understood as that part of an individual’s self-concept which derives from his knowledge of his membership of a social group (or groups) together with the value and emotional significance attached to that membership” (Tajfel 1981)

Components of ‘state of ethnic identity’ (Phinney 1990):– Self-identification as a group member– Ethnicity (parental ethnic background)– Sense of belonging to the group– Attitude towards own group (and towards the other group)– Participation in own group activities



Ethnicity in censuses and surveys

Focus is on measuring affiliation to ethnic categories/groups− Count demographic groups for monitoring discrimination and

inequalities in opportunities (requirement for evaluation of Equality Act 2010, UK Race Relations Act, US Civil Rights Monitoring and Enforcement)

Ethnic categories/groups− Are expected to be mutually exclusive, consistent and

relatively stable− Do not require strong sense of identification – sufficient to

self-categorise

Development of ethnic categories/groups in censuses: has been subject to extensive testing, development and both pre-and post-evaluation as well as ongoing refinement.


But these measures were inadequate as measures of ethnic identity

These measures just require people to self-categorise even if the categories/groups are of no significance (“fictive unities”, Werbner 1990)

At best these measures do not provide information about the components of identity other than affiliation

Measurement issues relating to the self-categorisation census type questions and questions to measure ethnic identity are different


Measurement Issues for ethnic identity

Do people find it easy to attach ethnic group labels to themselves as a way of expressing their ethnic identity [even if allowed to choose multiple groups]?

Do people agree on what constitutes an ethnic group or their ethnic identity?

Do all researchers agree on what constitutes an ethnic group?

How best to ask a question on identity in general and ethnic identity in particular?


What did we do?Used a multi-pronged approach

Literature review

Qualitative: Focus groups to understand some conceptual questions about ethnic identity

Qualitative: Testing of suite of questions emerging from this process

Qualitative/Quantitative: Testing wording of a question using a self-completion on a convenience sample – Identity Quiz


What did we do?Used a multi-pronged approach

Quantitative: Testing some prototype Identity questions in the Innovation Panel

Qualitative: Cognitive Testing of final set of questions

Qualitative: (Used results from the cognitive testing in Understanding Society wave 1 to ascertain how people respond to multiple response questions)


Focus groups: What we did

London Young men and women

Middle to lower social class

Different non-white ethnic groups

London Young men and women

Middle to higher social class

White and Ethnic minority groups

London Older men Middle to lower Different non-white ethnic groups

Colchester Young men and women

Mixed educational levels White

Colchester Older men and women


White

Sheffield Older and middle aged Women


Pakistani

Sheffield Older and middle aged Women


Black African and Black Caribbean


Focus groups: What we found (1)

For minorities, centrality of ethnic identity“Ethnicity for me is as important as my name because it is

my identity. It’s a part- on a larger scale it is my identity.”

For majority (and to a certain extent for white minorities), ethnicity was a property of ‘others’ (typically immigrants)

“I don’t think much about my ethnic group…. It’s the obvious thing for me, I’m white, I cannot change it and probably it influenced who I am at the moment, shaped me somehow, but I just don’t know.”

Difficulty and annoyance in being categorised/pigeon-holed/forced to tick boxes, but pride in ethnicity


Focus groups: What we found (2)

Centrality of skin colour to others’ – and to self perception“So the Black comes with the British for me.”

Strong attachment to regional locations“I wouldn’t see it in terms of nationality, wouldn’t be like I’m British or Irish, I wouldn’t be proud of that. It wouldn’t kind of occur to me. I’d much rather describe myself, for example, as a Londoner”

Importance of values and value systems to ethnic identity

Importance of patterns of association to ethnic identity

Importance of food / diet to identification (spontaneous)


Conclusions up until now…

It is important to

To provide flexibility for researchers in constructing ‘ethnic groups’

To allow respondents to identify with multiple aspects of ethnic identity rather than imposing a choice

To capture various components of ethnicity / ethnic identity that go beyond allocation to a particular group - belonging, identification, attitude, behaviour & practice


Quantitative Testing: of prototype ‘identification’ questions in the Innovation Panel

(Based on the Citizenship survey)

Do these questions ‘work’ for majority population? (extent of item non-response)

Are distributions of responses by socio-demographic characteristics as expected?

Do responses vary systematically by the interview mode?


Quantitative Testing: Innovation Panel suite of identity questions

We’d like to know how important various things are to your sense of who you are. Please think about each thing I mention, and tell me whether you think it is important, not very important or not at all important to your sense of who you are?

READ OUT EACH AND CODE (1) Important (2) Not very important (3) Not at all important

(a) Your occupation? INTERVIEWER: IF DK PROBE: Is that because you are retired?

(ethnic or racial background, religion, national identity, political beliefs, family, father’s ethnic group, mother’s ethnic group (if different from father’s), marital or partnership status, gender, age and life stage, level of education, sexual orientation)


Do these questions ‘work’ for majority population?

% of Don’t Knows are less than 1% for all except for:

− occupation/profession (2%)− gender (1.4%) − sexual orientation (1.1%)

BUT for those over 60 years of age: DK for occupation is 4.6%


Are distributions of responses by socio-demographic characteristics as expected? YES

Ethnic or Racial background is important for – 74% for non-white/mixed groups – 53% for white

Occupation is important for – 49% for 60+ – 80% for others

Marital or Partnership Status is important for – 27% of those who are single– 41% of separated/divorced/widowed – 88% of those who are currently married or in a partnership


Do responses (for IMPORTANT ) vary systematically by the interview mode?

Ethnic or racial background (F2F 56.3% > Tel 51.2%)

Political beliefs (F2F 33.4% < Tel 42.6 %)

Gender (F2F 79% > Tel 73.8%)

Education (F2F 69.3% < Tel 79.8%)

Sexual Orientation (F2F 66.9% > Tel 62.1%)

Occupation/Profession (F2F 3% > Tel 1%)


13 Semi-structured interviews

Explore similar issues around ethnicity, but in more depth

Test wordings of questions about ethnic identity measured

− Across different dimensions such as religion, language, country of birth...

− Across different components such as belonging, pride,...


Suggested ethnic identity questions:

A set of questions to capture affiliation to a group based on language, religion, country of birth, …

And then based on these groups, questions on

Personal identification (Importance to your sense of who you are….) with this group

Shared values and beliefs with members of these groups (Do you share values and beliefs...)

Pride in these groups (Do you feel proud...)

Interaction with members of these groups (How often do you interact...)

Group belonging (Happy to meet someone who...)


How to measure affiliation to a group, groupness? So, initially we…

Carried out an ‘identity quiz’ (short self-completion questionnaire) where we examined this.

We compared responses to sense of belonging, connection, and closeness to different potential domains of ethnic identity

On a non-random (work-based) sample of 46 colleagues


Identity Quiz

1) Do you feel a sense of

belonging to

2) Do you feel a connection to

3) Do you feel a sense of closeness

to

Yes No Yes No Yes No

A those who speak English (or a dialect of English)?

B those who speak the same language you were brought up in?

C those who are of the same religion as you are?

D those who live in the same city/region that you do?

E the country where you were born?

F the region in the country where you were brought up?

G the country where your father was born?


The Identity Quiz: What we found

Generally a gradation between stronger (closeness) and weaker (connection) but quite a lot of variation

Many respondents wanted to be able to answer yes

Giving irrelevant response categories provoked some frustration

Desire for graduated rather than yes/no responses


So, then for the interviews:

We could avoid group affiliation question as we had information about “objective” group affiliation such as country of birth, religion,... from wave 1 (except question on ‘mother tongue’)

Dimensions: language, English, British, current city/region, country of birth, region brought up in, parents/grandparents’ country of birth if different from own, land of your ancestors, skin color or other visible characteristics, Black

Components: identification, belonging/connection, pride, shared values, interaction


Interviews: What we concluded (1)

Expression of ‘importance to sense of self’ appeared to work for personal identification

And so did ‘happy to meet someone who…’ for belonging

Black rarely used in its one-time political sense – predominantly understood as reflecting African or Caribbean heritage.

Colour important – even if self-evident and no indication that discomfort in answering, though could be confusion for those who were not African or Caribbean or South Asian


Interviews: What we concluded (2)

Interaction should be measured directly – not appropriate to measure through subjective appraisal. [DISCARD]

Values and beliefs – an empirical question, not attitudinal, whether common across ‘groups’ [DISCARD]

Pride produced varied responses – suggests it can differentiate – for some obvious for others difficult or too sweeping


Cognitive testing later: What we did

22 cognitive interviews with respondents varying in ethnicity, age and generation

Tested a limited set of responses for potentially longer question sets. Questions covered:

− language (the “missing dimension”)

− Identification: ‘importance to sense of self’− Closeness/belonging: ‘happy when you meet someone who…’− Pride− Food and dress habits


Cognitive testing later: What we did

Probes covered

Ease/difficulty of answering questions

what phrases/terms meant (‘important to my sense of I am’ Vs ‘important to me’, ‘typical food’, ‘meet someone’)

Preferences for different response options

Degree of comfort with some specific questions


Cognitive testing later: What we found

Respondents didn’t have many difficulties with the questions, and didn’t seem to find them uncomfortable

Appeared to understand the questions and be able to interpret them as relating to self – even if hard to explain them back sometimes

Food questions worked better than clothing questions

Tended to prefer range of responses rather than simple yes/no


Combining the quantitative and qualitative findings

Quantitative findings

− Showed the potential of identity questions to be asked regardless of mode

− Showed that these questions worked well, given the responses varied by socio-demographic characteristics in expected ways



Qualitative results reinforced or enhanced the quantitative results

− confirming the acceptability and comprehensibility of identity questions

− indicating the ability of respondents to distinguish between ‘important to sense of who I am’ and ‘importance to me’ – even if difficult to express the difference – reinforced the validity of the questions



Qualitative findings − were reassuring about the low ‘sensitivity’ of such

questions (particularly among minorities)

− indicated that questions on other dimensions of identity – belonging, pride and food – worked well

− confirmed the complexity of the task undertaken but did not appear to render it impossible!

Thank you!http://www.understandingsociety.org.uk/


Lessons learnt from the Literature Review

Ethnic identity formation is a process – change is expected and is of interest in its own right

Attachment to various dimensions of ethnicity, including practices may define what an ethnic group means, rather than a single category

Constructions of groupness, may vary with the analytical questions considered and with the focus on particular dimensions of ethnicity


Lessons learnt from the Literature Review

Avoid census categories: ‘learnt’ responses Vs identification

Multiple response question that allow for people to choose multiple ethnic identity domains are preferred

Avoid a priori assumptions that place people into pre-existing groups: frustations of persons of mixed parentage and recent immigrants

Be careful of particular words that have certain connotations, e.g., ‘describe’, ‘best’

Avoid an open-ended question


Quantative Testing: Prototype questions to be tested

Suite of ‘identity’ type questions from the Citizenship Survey

SHOWCARD

We’d like to know how important various things are to your sense of who you are. Please think about each thing I mention, and tell me how important it is to your sense of who you are? Please choose your answer from the card.

Your occupation?(1) Very important(2) Quite important(3) Not very important(4) Not at all importantDON’T KNOW

(occupation, ethnic or racial background, religion, national identity, where you live, your interests, family, social class (working, middle), The country your family came from originally, gender, age and life stage, level of income, level of education)


Quantitative Testing: Innovation Panel

Household Panel survey of approximately 1500 households

Stratified, clustered sample (with an epsem design): 2730 addresses selected from 120 areas in UK (excl. N. Ireland and N of Caledonian Canal)

1489 Households & 2384 individuals interviewed in wave 1 (1660 individuals in wave 2)

Wave 1 in 2008, Wave 2 in 2009 (we use data from wave 2)


Ideal for inclusion in 2nd wave of Understanding Society

Understanding Society is a multi-topic Household Panel Survey of up to 40, 000 households (respondents: all 16+ household members) with an ethnic minority boost of 4,000 households.

Ideal because:

Large ethnic minority sample (EM in boost+ in main sample): Facilitate research on ethnic identity, heterogeneity across different groups

Longitudinal study: Later waves can draw on (stable) information collected in earlier waves (such as country of birth, religion, language)


(Earlier Cognitive testing)

Tested a multiple response categorical question on ethnic group

Found

Respondents like the opportunity for multiple responses to reflect different aspects of themselves – though they didn’t like the choice of categories

Acceptability of multiple-response / multi-dimensional approach

Using qualitative and quantitative methods to develop ethnic identity questions in the UK Alita Nandi and Lucinda Platt with thanks to Liz Spencer, Punita.

Documents

ethnic groups

ethnic identity questions

measures of ethnic identity

ethnic identity qualitative

ethnic identity phinney

ethnic group labels

components of identity

group of people