Cultural Variability in Respondent Processing of Health Survey Questions

The processes by which culture influences survey-related behaviors and measurement error remain poorly understood. We address this problem by investigating race/ethnic variability in the cognitive processing of a set of questions commonly asked in health surveys. Data come from survey interviews with 600 African American, Korean American, Mexican American, and non-Hispanic white adults recently conducted in Chicago. Survey reports were supplemented with behavior coding of respondent and interview behaviors during each interview. Using hierarchical linear modeling (HLM), we examine similarities and differences in question comprehension and response mapping across cultural groups, while controlling for multiple respondent demographics, survey question characteristics, and interactions between question characteristics and respondent race/ethnicity. The presentation will conclude with a discussion of current knowledge regarding cultural variability in the processing of health survey questions and recommendations for best practice.