Inclusive Insights: Advancing Race & Ethnicity Data Collection Practices

Several CMBers recently had the opportunity to attend the Insights Association North Atlantic Conference to learn and discuss innovations within the industry, particularly what we can do as market researchers to deliver more inclusive insights. Here are some actionable insights to improve your practices around race and ethnicity:

Ask about Race & Ethnicity in a Way that Allows for a Representative Sample

The IDEA Council recently released a report that offers suggestions on asking more inclusive race and ethnicity questions. The multiracial population is growing rapidly in the United States, so displaying a question as multi-select rather than single select is paramount to accurately recording race and ethnicity. In addition, respondents prefer having more detailed categories to choose from—for example, instead of saying Asian, having separate options for East Asian, South Asian, Southeast Asian, and Middle Eastern.

Improve the Respondent Experience

According to the IDEA Council research, 1 in 4 respondents report feeling upset by the question and 1 in 7 say they are offended. What does the IDEA Council advise?

• Change the “Other” option to “I prefer to self-identify” since “Other” has a negative connotation.
• Include an opt-out option as many respondents don’t understand why they need to provide that information and how the data will be used, and therefore they don’t feel comfortable answering
• Add language in the question text about why we’re asking about race and ethnicity.[1]

Define What you Hope to Uncover When Targeting Specific Races and Ethnicities

From 2010 to 2020, the White population has decreased 8.6% whereas all other single racial populations and multiracial populations have increased.[2] Along with the rapidly changing racial landscape, there is also a changing definition of race depending on the person and context. While discussing this at the conference, one person shared that she was born in Korea to Korean parents, but she was adopted and raised in the US by White parents. She identifies as biologically Korean and culturally White, so she never knows how to define her race when asked. When we analyze racial data in a consumer context, we are potentially grouping respondents in a way that doesn’t make sense for the behavior we’re trying to understand.

This also applies to the distinction, or lack thereof, between race and ethnicity. The US Census currently asks these as separate questions, although the IDEA Council research found that respondents may prefer it to be combined in one question.

Consider Omitting Race & Ethnicity Questions

During the conference, there was no clear answer as the room was split on the right thing to do. First, you must ask why it should be included. For respondent experience? Census-balancing? Analysis and reporting? On one side, you should know who is in the sample and do to ensure all ethnicities are represented. On the other hand, if census-balancing is not being done or using for analysis, why ask the question?

Overall, the conversations that were had at the Insights Association North Atlantic Conference are helping to move our industry forward and give our clients the most inclusive insights and ensure survey respondents feel accurately represented. Let’s continue the conversation. Contact Amanda and/or Emily.

[1] https://www.insightsassociation.org/Resources/Inclusion-Diversity-Equity-and-Access/Asking-Questions-on-Race-Ethnicity

[2] https://www.census.gov/library/stories/2021/08/improved-race-ethnicity-measures-reveal-united-states-population-much-more-multiracial.html#:~:text=The%202020%20Census%20shows%20(Figures,in%20combination%20with%20another%20group