Does the Timing of the Mode Switch Matter in a Mixed-Mode Survey? Results from an Experiment

• Activity Number: 34
• Type: Contributed
• Date/Time: Sunday, August 9, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): James Wagner* and Heather M. Schroeder and Andrew Piskorowski
• Companies: University of Michigan and Institute for Social Research and University of Michigan
• Keywords: Mixed-mode survey ; Paradata ; Cost models

Abstract

The Army Study to Assess Risk and Resiliency in Servicemembers (STARRS) is a study of mental health risk and resilience among military personnel. One component of this large study -- the Pre- and Post-Deployment Study (PPDS) -- followed a set of about 9,000 soldiers before and after deployment. The final wave of this study (T3) used a mixed mode design to interview individual soldiers. The design included a web survey followed by telephone attempts for nonresponders. In a mixed-mode design such as this, one important design decision is when to switch from web to telephone. A longer duration for the web mode may reduce costs. On the other hand, a longer duration for the web mode may lower overall response rates. The Army STARRS PPDS T3 survey tested the impact of the timing of the mode switch using four different durations for the web mode. This paper will examine differences in cost, response rates, and key estimates across the four experimental treatments. We do not find differences in estimates or response rates, but costs do differ across the four treatments.