Notice: the WebSM website has not been updated since the beginning of 2018.

Web Survey Bibliography

Title Respondent preferences toward audio-CASI and how that affects data quality
Year 1995
Access date 15.08.2004
Full text

pdf (564k)


As the social and economic problems of the United States become more complex -- AIDS, sexual abuse, violence, homelessness, corruption-- social scientists are called upon more frequently to identify and understand behaviors that are private and sensitive. Survey researchers have attempted to obtain this information by asking people to report their own behaviors in interviews, and have concluded, not surprisingly, that the more private the forum for reporting, the more accurate the report (Bradburn and Sudman, 1979; Bradburn, 1983). Moreover, in demonstrated experiments, the self-administered interview has been determined to provide the best environment for the reporting of many types of sensitive behaviors (Hay, 1990; London and Williams, 1990; Aquilino and LoSciuto, 1990; Schwarz, et al, 1991; Turner, et al, 1992). Limitations of the self-administered interview, however, are obvious. The respondent must have adequate reading skills, and must be able to understand and follow the questionnaire format, a novel experience for many (Lessler and Holt, 1987). Consequently, the burden on the researcher is to design a questionnaire void of complexity yet accomplishing the research objectives. Solutions to this dilemma have been evolving. The successful use of a Walkman-type device in the Youth Risk Behavior Survey conducted by the National Center of Health Statistics (NCHS) (Camburn, et al, 1991) eliminated the need for adequate reading skills by the respondent. But this method could not use skip instructions to access questions, thereby restricting the complexity of the instrument. Incorporating audio with a computerized selfadministration of a questionnaire (audio computerassisted self interviewing, or Audio-CASI) has offered an improved solution. With Audio-CASI the respondent listens to a voice-digitized recording of the questions and answer choices over earphones and keys the answers into a microcomputer. While providing privacy and eliminating the need for respondent literacy, this technology also allows the researcher to design complex questionnaires and provides standardized questionnaire administration.

Access/Direct link

Homepage - conference (full text)

Year of publication1995
Bibliographic typeConference proceedings