ceres CF, Mendoza W, rate MR, Fernandez P, Maiorana A, Coates TJ; International Conference on AIDS.
Int Conf AIDS. 2002 Jul 7-12; 14: abstract no. C11024.
Cayetano Heredia University, Lima, Peru
BACKGROUND: Audio Computer-Assisted Self Interviewing (ACASI, where a participant listens to questions through earphones and reads them on a computer screen, entering responses directly) has decreased social desirability bias in assessments of sensitive topics in the U.S., but its appropriateness in other countries has not been well assessed. Familiarity with computers, literacy and conceptions of privacy might influence relative gains. We compared answers to a sexual behavior questionnaire using ACASI with those using CAPI (where an interviewer keys respondents' answers directly into the computer). METHODS: Participants (n=69, aged 18-30, both genders) lived in randomly selected households in Lima and Trujillo, Peru. They responded the questionnaire twice (once in each mode), and were randomly assigned to either ACASI or CAPI first; the second interview occurred 2 days later. Both sets of responses were compared. After both modes were completed, an ethnographer conducted an interview with a sample of 20 participants to assess their understanding of the questions and the validity of responses provided. RESULTS: Of all participants, 31 were randomized to start with ACASI and 38 to start with CAPI. The mean duration of ACASI interviews (20 min. when done first and 18.5 when done second) was longer (CAPI: 14 and 12 min, respectively). Responses to questions on sex and drugs/alcohol in both modes (regardless or order) did not differ significantly, but all respondents preferred ACASI over face-to-face CAPI interviews, since they perceived it ensured more privacy. CONCLUSIONS: ACASI is preferred over CAPI to provide sensitive information in Peru, for privacy considerations. Responses provided in both modes, however, did not differ, which suggests that discomfort vis-a-vis sensitive questions in a face-to-face interview does not necessarily produce social desirability bias. Other issues (i.e. logistics, cost) should be considered to define a mode in this context.
Publication Types:
Keywords:
- Computers
- Data Collection
- Humans
- Interviews as Topic
- Peru
- Physical Examination
- Poverty
- Questionnaires
- Sexual Behavior
- Sexuality
- United States
- methods
Other ID:
UI: 102254505
From Meeting Abstracts