Aggregating data on AAPI groups can be ‘a form of erasure’
Data is the basis of actions and decisions – what gets funded and which issues get attention.
That’s why Asian American, Native Hawaiian, Micronesian and Pacific Islander groups, and a growing number of data scientists, are concerned about the practice of collecting their many identities under one demographic category, often shorthanded as AAPI.
The term “Asian American Pacific Islander” encompasses some 50 nations and ethnic groups who speak 100 languages. Those groups experience stark differences in health status, household income, educational attainment and more.
It’s an issue that Washington State University-Vancouver researchers Connie Kim Yen Nguyen-Truong and Sara Waters have run into in their work with Asian American, Native Hawaiian, Micronesian and Pacific Islander communities in Oregon and Washington. Their studies identify barriers to education, health care and civic participation faced by members of those communities.
Aggregating data for Asian American, Native Hawaiian, and Micronesian and Pacific Islander groups may appear to be well-intentioned as a way to increase political and social recognition for groups that face marginalization and discrimination.
But the implicit message sent when separate national or ethnic groups are considered as one is that their identities don’t matter. The way it’s perceived is, “You don’t see us,” said Nguyen-Truong, an associate professor in the WSU College of Nursing. Meanwhile, because data is used to allocate resources, many communities aren’t getting the help they need.
“Aggregation is a form of erasure of diversity, perspective and also health needs,” added Waters, an associate professor in the Department of Human Development at WSU Vancouver.
Harmful disparities can be hidden in aggregated data.
For example, cancer rates vary dramatically between different subgroups in the AAPI category. Even though cervical cancer among all Asian Americans is lower than among Non-Hispanic White Americans, the rate is much higheramong Vietnamese American women.
Asian-Americans have a lower poverty rate than the United States as a whole, but among Mongolian and Burmese groups it’s more than twice the national average.
Similarly, more than half of the AAPI group has a bachelor’s degree or higher, but among subgroups the percentages range from 74.5% to 5.2%.
To acknowledge such disparities, Nguyen-Truong and Waters approach their research as partners with study participants. Instead of using a standard form with boxes to check to indicate ethnicity, language and national origin, they work with communities to decide what those boxes need to be.
“We do end up having participants check boxes because we can’t have a conversation with each person,” said Waters, “but the list of boxes is constructed with the community so we can accurately represent everyone who’s participating in the ways they want to be represented.”
Said Nguyen-Truong, “We take the time to know who is with us and build trust before we get to the business part of things. We add a statement, built into the instructions in our background document, that we continue to humbly learn about the diversity in communities.”