Data tells a story. The story can help you sell to, target, and communicate with your customer base. While simplicity is key to representing your data, the reality is, people are not simple; and as our world continuously evolves to become more inclusive, it is necessary that the language we use in survey research reflects that diversity. If a respondent is excluded by not having the proper box to check, not only do they walk away from that experience unvalidated and discouraged, but you lose out on valuable data, values, and inputs from a very real percentage of your target audience. Clients typically want a balanced sample by demographics, including a 50/50 male/female split, as well as appropriate distribution by race and Hispanic origin, but these concepts are complex and evolving, and it is imperative that the respondent’s preferred terminology is respected in this industry. When it comes to identity, there is no such thing as statistically insignificant.
So how are we changing? We have had to adjust the way in which we ask several demographic questions that in the past had unchallenged response categories. One simple solution was to change our gender demographic question from using binary wording to asking how the participant identifies. We have also added an “other” option. It is important that this option is used to validate where our list may be lacking, not simply as a catch all for anyone who exists outside of the male/female gender binary. In most US adult population surveys, about 0.5% of respondents will check the “other” box, equating to 5 out of every 1000 people surveyed. And as society becomes more comfortable with nonbinary lables and pronouns, those statistics are bound to rise. As a matter of fact, in Canada and Europe, the incidence rate of nonbinary gender responses is closer to 1% of participants. Regardless of any belief system, these identities aren’t going away anytime soon, and in survey research, if you aren’t counted, you don’t count. Inclusion isn’t just preferable or sensitive, it is compulsory to remain relevant, trusted, and accurate in an industry that depends on the honesty of human beings.
Demographic information doesn’t end at gender. The way we have traditionally categorized race in US surveys needs to be examined and updated. Since the year 2000, race has been a multiple choice question on the US census. Multi-racial households have always been present in this country and have become increasingly prevalent in recent decades. Asking a participant “which race category do you most associate with?” is a start, but multiple choice options are your best bet for accurate data collection. Furthermore, one of the most excluded racial categories within surveys is the Middle Eastern population, who are often forced to pick between “white” or “Asian.” We’ve added a category labeled as “Arab/Turkish/Persian/Other Middle Eastern.” This takes no additional effort and allows this percentage of Americans to sort themselves into a meaningful category. Hispanic or Latino ethnicity is also complex. Increasing immigration from the Caribbean, central and south America (e.g. the influx of Venezuelans into Florida), as well as Portuguese-speaking Brazilian-Americans, creates a tapestry of Latin American cultures. While there is some debate around what term best suits this diverse group of people, we’ve found that the term “LatinX” is more appropriate than Hispanic. (LatinX is helpful not only for racial diversity but for gender diversity as well, but we understand this community has a vast amount of opinions regarding the term, so we remain open to the voices of those with Latin-American identities for the most helpful and accurate labels).
Finally, religious identification can be helpful to ensure our sample is sociologically representative. Religious identity is equally complex and often tied to cultural and political values. Some political and public affairs surveys have found that under-representation of White evangelical Christians leads to inaccurate results in the United States. Even the label “Christian” is not always sufficient to understand the distinction of attitudes between Catholics, mainline Protestants, and evangelical or non-denominational Christians, to name a few. This same consideration is equally important for Jews, Muslims, and those who do not subscribe to a religious belief. In political polls or other public affairs issues, such as surveys on Covid vaccination attitudes or school masking requirements, religious identity can be a key variable to explain differences in response.
The intersection of diverse identities, demographic information, and meaningful data is the cornerstone of well-conducted market research, and as our society learns more about these concepts, it is important that our industry remains sensitive and attentive to evolving labels. It is our job to learn from people, and we cannot do that if we refuse to accommodate as many unique and inclusive identities as possible; and we invite our client base to keep us accountable and educated to the ever-evolving landscape of identity.