By: The ORB Research Team
Designing and implementing survey questions surrounding ethnicity can be challenging and often requires careful consideration. Many regions and countries around the world are multi-ethnic, with some countries home to hundreds of ethnic groups. For example, in Sudan, more than 500 ethnic and sub-ethnic groups across its 18 states. Furthermore, ethnicity questions can be extremely sensitive to discuss in some countries due to historical, political or cultural contexts. So, how should researchers design surveys that accurately capture all identities and groups? Several considerations must be considered which we outline below:
- Work with local teams.
Researchers must make an effort to understand local contexts when designing questions surrounding ethnicity in multi-ethnic countries. The best method to do this is to work with local, in-country teams as they are inherently integrated into the research environment. Local teams can ensure proper names are used for ethnic groups and can advise on question sensitivity in order to ensure all questions are safe to field.
- Consider cultural and historical sensitivities.
Questions regarding ethnicity can be a difficult and sensitive subject, especially in countries with colonial histories regarding indigenous groups or identity-driven conflict. For example, in North Africa, the indigenous group Imazighen/Amazigh (sl) are often misnamed as ‘Berber’, which is the term more commonly known in the West. The group has pushed for the use of ‘Amazigh’ more recently. This type of consideration is important to make note of in the design of surveys to ensure questions are inclusive of all groups and their preferred names.
- Adding additional response options.
Inclusive surveys that ask about respondent ethnicity should be optional and have a choice for writing in an answer, usually given as (Other, please specify___). Including this option allows for respondents to be accurately represented if their ethnic group was not listed amongst the options or provides an option if they do not feel comfortable providing an answer.
For example, in a recent survey ORB conducted in Rwanda, 75% of respondents from the country elected to write in or selected ‘don’t know’ for their ethnicity. When looking at the write in options, many respondents wrote ‘Rwandan.’ This trend may be in line with Rwandan reconciliation efforts where citizens are encouraged to not identify by a specific ethnic identity, as these were imposed by Belgian colonial powers, and identify simply as ‘Rwandan’. This may signify that, in future surveys, researchers should not include an ethnicity question in the country and look at another option of capturing similar information if it is pertinent. Additionally, including a write-in option allows for survey designers to take into account additional ethnic identities and add them when conducting future research in that country, as identities can be fluid. This helps to keep research teams up to date on local contexts.
- Ethnicity questions should be ‘select all that apply’
Regions across the world are home to many multi-ethnic countries. Any question asking a respondent to report their ethnicity should thus be programmed as ‘select all that apply’ so that respondents who come from more than one ethnic background can select multiple groups they identify as. For example, a survey respondent in Nigeria might identify as both Hausa and Yoruba; therefore, having an option that allows them to select both would be the most representative option, rather than choosing just one. This allows for a more inclusive approach toward capturing accurate demographic information and is a simple programming addition to a survey.
Conclusion
Research environments are dynamic and must be adaptable to design. Taking the above practices into consideration when designing surveys in multi-ethnic countries will make survey research more inclusive and ensure respondent comfort. They are also a key component to delivering relevant, accurate, and insightful findings. These outlined above points should be considered in the design stage of your survey to guarantee they are implemented before fielding.