CAPI Versus Audio-CASI: Understanding the Role of Privacy

• Activity Number: 195
• Type: Contributed
• Date/Time: Tuesday, August 10, 2004 : 8:30 AM to 10:20 AM
• Sponsor: Section on Survey Research Methods
• Author(s): Mick P. Couper*+ and Eleanor Singer and Roger Tourangeau
• Companies: University of Michigan and University of Michigan and University of Maryland
• Keywords: audio-CASI ; sensitive questions ; CAPI

Abstract

Audio-CASI has been widely adopted by the survey industry for administering questionnaires on sensitive topics. Several field experiments have demonstrated the advantages of audio-CASI over interviewer administration for sensitive questions, but there have been relatively few attempts to understand the mechanisms by which audio-CASI improves the veracity of self-reports. We report on a laboratory-based experiment to explore some of these issues in detail, specifically the possible role of privacy in affecting the answers provided in the two modes. We carried out a 2×3 experiment that crossed the mode of administration (CAPI versus audio-CASI) with three different privacy conditions. In the high privacy condition, the subject completes the survey in the presence of the interviewer, but with no third party present. In the medium privacy condition, the subject is introduced to an observer at the outset, but there is no expectation of later interaction with that observer. In the low privacy condition, the subjects are introduced to the observer, but also told that the observer will debrief them at the end of the interview, raising expectations about later interaction.