Computerized Interviews: More Accurate Data? Sexually transmitted infections (STIs), including infection with human immunodeficiency virus (HIV), the virus that leads to AIDS, have a devastating impact on women’s health worldwide. Understanding who is at risk for STIs/HIV and what behaviors lead to infection is critical for developing and evaluating programs to prevent new infections and treat those already infected. “Many individuals are uncomfortable answering questions about very personal and sensitive behaviors, for example how often they have sex, with whom, and under what circumstances,” says Population Council senior researcher Barbara S. Mensch. “Unfortunately, if reporting of risky sexual activity is inaccurate, the ability to implement and monitor health programs is compromised and estimates of STI risk may be seriously biased.” To address this issue, Mensch, senior associate Paul C. Hewett, and program associate Heidi Jones have conducted research using audio-computerized self-interviewing (ACASI) technologies. Computerized self-interviewing provides individuals greater privacy and confidentiality when responding to sensitive questions (see box). Getting Personal
Instead of answering questions aloud in a face-to-face interview, a respondent using an ACASI computer hears pre-recorded questions through audio headphones and presses numbers on a numeric keypad to answer them. The computer can remain open, allowing the respondent to read the question along with the audio, or may be closed for complete privacy. Council researchers have found that most participants quickly learn how to use the interview program and prefer the computer over face-to-face interviews. |
The critical question is whether computerized interviewing provides more accurate reporting of risky behaviors than traditional face-to-face interviews. Council researchers, in collaboration with Brazilian colleagues at the Centro de Saúde Escola Dr. Alexandre Vranjac, Barra Funda (CSEBF) health center, set out to answer this question by conducting a randomized experiment comparing the reporting of risky behaviors in computerized versus face-to-face interviews. This study—involving over 800 women in a low-income area of São Paulo who were receiving care and treatment at CSEBF—found that those interviewed using the computer were significantly more likely to report high-risk sexual activities, for instance a greater number of lifetime and overlapping sexual partners, as well as less condom use. ACASI respondents were also more likely to reveal that they and their partners had recently used alcohol or drugs, behavior often associated with high-risk sexual activity. The results of this study provide strong evidence of the benefits of computerized interviewing for reporting of sensitive behavior. A broader program of research and evaluation on this topic has been initiated by Council researchers in countries with high rates of STIs and HIV/AIDS, including Kenya, Madagascar, Malawi, and Uganda. The cumulative results of these studies, and studies of alternative methods for collecting accurate information about populations in settings where computers may not be feasible, are helping researchers and program managers design more effective ways of collecting evidence crucial to improving reproductive health for both men and women. Funding was provided by the National Institute of Child Health and Human Development (NICHD), USAID, and the Bill & Melinda Gates Foundation. Future work on this topic will be funded by NICHD. (Return to issue contents)
| |
