Piggyback Survey Respondents and Mode: Lessons Learned from Design and Operations

• Activity Number: 159
• Type: Topic Contributed
• Date/Time: Monday, August 1, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Brad Edwards*+
• Companies: Westat Inc.
• Keywords: survey design ; multi-mode ; piggyback ; sample design

Abstract

"Piggyback" surveys have at least two parts: data collected from an initial sample are used to spawn another sample. Designing the second sample so that it "piggybacks" on the first is typically much more efficient than other sampling approaches. For many piggyback surveys, data from the two parts are collected in different modes (e.g., face-to-face for one, telephone for the other). This presents a number of design and operational challenges. The same questions may be asked in both surveys; they may be designed in one mode, then adapted for the other mode without considering mode effects. Elapsed time between the specific respondent's interview in Sample A and the interview with the Sample B respondent he or she identified may also be a concern. If too much time elapses, the link between the two may be broken. The paper will discuss lessons learned from a number of experiences with thes