Why do researchers ask such complicated ethnicity questions?
I’m sure you are used to filling in forms with great long screeds of options for you to select your ethnicity from, and recently a few people have asked me what that is all about because they find it annoying.
The complicated thing about a person’s ethnicity is that it is usually defined by a combination of a number of factors:
- Skin colour
- Country of birth
- Country of residence
Take me, for example. How do I define myself?
- The case for English: I was born in England, and both of my parents were born in England.
- The case for Scottish: I have lived virtually all of my adult life in Scotland and am married to a Scottish person.
- The case for British: This would cover both England and Scotland so there would be no need for me to pick one or the other.
- The case for UK: My passport says ‘UK citizen’.
- The case for Irish: Two of my grandparents were born in Ireland.
The consequence of this is that sometimes on a form it is difficult to know what to choose. And I imagine it would be even more complicated if I had dual nationality, or wanted to express both where I was born and the colour of my skin (i.e. British Asian), or felt strongly patriotic towards a country which was not represented in the list (i.e. Scottish rather than British).
Here’s an example of an ethnicity question from the Scottish Health Survey:
Q To which of the following groups do you consider you belong?
- White: Scottish
- White: Other British
- White: Irish
- White: Any other white background (WRITE IN)
- Mixed: Any mixed background
- Asian, Asian Scottish or Asian British: Indian
- Asian, Asian Scottish or Asian British: Pakistani
- Asian, Asian Scottish or Asian British: Bangladeshi
- Asian, Asian Scottish or Asian British: Chinese
10. Asian, Asian Scottish or Asian British: Any other Asian background (WRITE IN)
11. Black, Black Scottish or Black British: Caribbean
12. Black, Black Scottish or Black British: African
13. Black, Black Scottish or Black British: Any other black background (WRITE IN)
14. Any other ethnic group (WRITE IN)
15. Prefer not to say
The main reason that you see such long lists is to allow you, the respondent, to self-define your ethnicity. It is widely acknowledged that ethnicity is a contentious and complex issue and the huge lists are there so you can pick the closest thing to suit your identity, or write something in as you prefer. Annoying it may be, but in my experience respondents get even more annoyed when there isn’t a box that they feel comfortable checking.
Sometimes, perhaps in the case of the Scottish Health Survey even, the question is asked because the client or the researcher genuinely wants to know what the ethnic mix of the sample or population is.
In my experience of working in several different research organisations in the UK however, all of this information tends to get compressed down into ‘white’ and ‘non-white’, to be used for analysis purposes. I have rarely seen anything else done with this information.
Now there is certainly no racist undertone intended in this practice, it is more one of practicality. As I have described in a previous blog post, all sorts of demographic information is collected during a survey to allow the researcher to get a picture of patterns between sub-groups in the data across the whole sample. To do this effectively, ideally you want each sub-group to contain at least 50 and ideally 100+ people so that you can make robust statistical conclusions.
According to the 2001 Census, 2% of the Scottish population were ‘non-white’.
- So if you conducted a representative survey of 100 Scottish people, 2 would be non-white.
- If you conducted a representative survey of 1,000 Scottish people, 20 would be non-white.
- If you conducted a representative survey of 5,000 Scottish people, 100 would be non-white.
According to the 2001 Census, 9% of the UK population were ‘non-white’.
- So if you conducted a representative survey of 100 UK people, 9 would be non-white.
- If you conducted a representative survey of 1,000 UK people, 90 would be non-white.
Although the non-white population of the UK is on the rise, you can see that in this country at least it takes a massive (expensive!) survey to generate a robust ‘non-white’ sub-group, and that is before you start breaking it down any further into Chinese, or Caribbean, or whatever.
So is it worth the complicated question to get the broad-brush answer? What is more important, keeping the respondent happy at the time of interview or collecting only data that will be used in the format it is collected?
Make of that what you will…