Web-CAPI sequential mixed mode design in a longitudinal survey: effects on participation rates, sample composition and costs

Abstract

Sequential mixed-mode surveys are often now used as an alternative to single-mode interviewer-administered designs, bringing benefits from each mode. We assess the effects of the introduction of a sequential web-CAPI mixed-mode design over three waves of a longitudinal survey in which members were previously interviewed by CAPI. Data come from a large-scale randomized experiment carried out on the Understanding Society Innovation Panel. No differences in cumulative response rates are found between the mixed mode design and CAPI design and only minimal differences are found in sample composition. Potential cost savings with mixed modes are evident.