When an interviewer-administered questionnaire (IAQ) is used in sex research, the human element involved in a face-to-face encounter can influence the respondent to underreport certain sensitive behaviors and to overreport more normative or socially acceptable behaviors (Dolezal et al., 2011; Potdar & Koenig, 2005). A written or self-administered questionnaire (SAQ) provides an alternative survey method that can overcome some of the difficulties of an IAQ by providing a more private and potentially less threatening means of reporting sensitive behavior. However, SAQs can also be limited by the reading ability or literacy of respondents. Failure to understand written survey questions can be a significant limiting factor when people with relatively low literacy are surveyed.
The recent advent of computer-assisted self-interview (CASI) technology for surveying children, adolescents, and adults has provided an excellent tool for overcoming these barriers to successful sex research. With CASI technologies literacy problems and the potentially negative effect of a human interviewer are minimized. Furthermore, researchers can be confident that key elements of questions’ presentation and measurement are standardized for all respondents.
Two varieties of CASI technology are currently used. In video CASI technology, respondents view questions on a monitor and enter their answers by pressing labeled keys on the keyboard. Audio CASI offers a somewhat more advanced technology in which respondents listen through headphones to questions (which may also be simultaneously displayed in print on-screen) and enter their answers by keystrokes. The audio component has voice-quality sound. Unlike video CASI and more traditional survey methods, audio CASI does not require respondents to be literate. Furthermore, because questions are prerecorded, this technology allows multilingual administration without requiring researchers to be multilingual.
The application of CASI technology is becoming more widespread, and numerous studies have demonstrated that this method is effective for collecting sensitive information on a variety of topics (Dolezal et al., 2011; Hollander, 2008a; Midanik & Greenfield, 2008). For example, a recent investigation used two data collection methods—CASI and a traditional telephone survey—to obtain sensitive information about sexually transmitted infection (STI) incidence from a sample of more than 2,000 men and women. Subjects were randomly assigned to one of these survey methods. Respondents who participated via CASI were significantly more likely than those who spoke with a telephone interviewer to acknowledge having contracted a variety of STIs (Hollander, 2008a).