Highlights
Every year since 2017, SSRS has collected over 20,000 interviews with adults in California to learn about health and healthcare across the state. The study also includes several thousand interviews focused on children and with adolescents age 12-17. In 2019, SSRS worked with UCLA to transition the CHIS from an exclusively telephone methodology to an address-based sample (ABS) design with both online and telephone interviewing modes.
Challenge
Providing representation of California’s large and diverse population including multiple demographic subgroups and smaller geographies without sacrificing data quality.
Approach
SSRS implements sophisticated stratification and statistical modeling to target harder-to-reach groups as well as counties and smaller geographies in California. Models predict which households are likely to contain the desired populations, and these households are oversampled as part of the overall ABS design.
In addition to the ABS, SSRS dials prepaid cell phone numbers to take advantage of this sample frame’s higher incidence of potential respondents with lower incomes, less education, lower English fluency, and higher likelihood of being Hispanic and uninsured. To further enhance representation, the CHIS is conducted in English, Spanish, Chinese (Cantonese and Mandarin dialects), Korean, Tagalog and Vietnamese.
Results
Overall, the statistical modeling and prepaid cell sample dialing produced the desired outcomes in reaching harder-to-reach populations of Californians. The modeling was particularly effective in targeting ethnic groups and respondents who were more comfortable in taking the survey in one of the non-English languages offered. The model was also efficacious in locating households with children. The inclusion of a prepaid cell sample frame also increased number of completed interviews with younger respondents, those who spoke a language other than English, and lower-income Hispanic Californians.