Saturday, November 12
Pretesting Methods
Sat, Nov 12, 4:00 PM - 5:25 PM
Regency Ballroom-Monroe
Pretesting with Special Populations

Using Focus Group and Cognitive Interview Methods to Develop a Tribal Enrollment Question for American Indians and Alaska Natives (303330)

*Rodney Lamar Terry, U.S. Census Bureau 

Keywords: race, qualitative research, question development, sensitive questions

When used strategically, multiple pretesting methods can help inform the development of a new question on a complex and sensitive topic. As an example, I present 2020 U.S. Census research that used focus group and cognitive interview methods to respectively draft and pretest designs of an enrollment question for American Indians and Alaska Natives (AIAN). Currently, the 2010 U.S. Census race question has a checkbox for the AIAN category, as well as an instruction to print the name of the “enrolled or principal tribe” in a write-in space. However, because data collected from the write-in space likely include both enrolled and nonenrolled tribe identities, the data cannot be used as a specific source of information on the enrolled AIAN population. Thus, various AIAN stakeholders have requested a separate and specific enrollment question to better understand the enrolled AIAN population.

First I present findings from 11 focus groups with AIAN participants that informed initial drafts of three enrollment question designs for the AIAN population. Second are findings from 63 cognitive interviews that investigated respondent reactions to these enrollment question designs. I identify findings that arose across both studies, including opinions about the Census Bureau collecting data on enrollment, and the preferred tribal enrollment question designs. I also discuss how the focus groups’ open discussion of broad enrollment topics produced rich data that informed question drafting, as well as how probing on specific question elements during the cognitive interviews refined question design. Finally, I discuss the implications of this research for questionnaire design and survey data collection on sensitive topics. This research informed the development of enrollment question designs for inclusion in a national, split ballot test planned for 2017. The results of this national test will help inform whether an enrollment question is included in the 2020 U.S. Census.