Sequential and Concurrent Mixed-Mode Designs: A Tailored Approach.

This article examines the impact of mode choice sequences—concurrent versus sequential—in self-administered mixed-mode surveys on sample composition and response rates, using data from the 2021 German General Social Survey (ALLBUS). The study finds that while the sequential design (web mode offered first) achieves slightly better representativeness based on register data (R-indicator), the concurrent design (offering web and mail modes simultaneously) yields a higher response rate and better alignment with an external benchmark, the German Microcensus. Tailoring the mode choice sequence based on age, a register-based indicator, modestly improves sample balance and response rates compared to non-tailored designs, though at a slight increase in survey costs. The findings suggest that tailoring mode sequences using known person-level characteristics can enhance sample composition in mixed-mode surveys, but improvements are incremental and involve trade-offs between cost, response rate, and representativeness.