SURVEY MEASUREMENT OF SENSITIVE BEHAVIORS USING A-CASI

测量技术的最新进展（音频、计算机辅助、自我测量）

A recent advance in survey technology (audio, computer-assisted, self- interviewing: Audio-CASl) provides a potentially Important tool for improving the validity and reliability of survey measurements used in research on sexual, contraceptive, and related behaviors. Since survey measurements provide important information on behaviors related to pregnancy, contraceptive practices, and transmission of STDs, improvements in the accuracy of these measurements could have important benefits for both research and for policy making. The general aim of the present research is to explore the measurement characteristics of this new technology when applied to such measurements of such sensitive topics. Among our particular aims are the assessment of: * The extent, if any, to which this new technology may increase the validity and reliability of survey measurements of sexual, contraceptive and other related behaviors. * The types of measurements and research contexts, if any, for which this technology is best suited. * The costs and barriers to adoption of this technology (assuming that the technology is found to yield improved measurements). The proposed research program will include: * An Initial series of four laboratory pretests designed to refine operational details of the use of this new technology; * A field experiment testing the impact of Audio-CASI versus "standard" survey practices upon the validity and reliability of measurements of sensitive sexual, contraceptive, and fertility-related behaviors. The field test will randomly assign 1,368 respondents drawn as a probability sample of adults (ages 18-49) residing in the Washington metropolitan area (MSA) to be interviewed using one of the alternative survey methods. Variations in the levels of reporting of sensitive behaviors (e.g., abortion) will be analyzed to assess the impact, If any, of this new technology on reducing the underreporting bias that typically afflict such measurements. The completeness and consistency of reporting will provide an indication of the effect of this new technology on other aspects of data quality. Data on costs and problems in fielding this new technology will also be collected. * A subsequent series of three experiments will be conducted under controlled laboratory conditions to investigate the cognitive and psychological process that account for differences In survey response when different modes of Interviewing are used (Audio-CASI, written self- administered questionnaires, and Interviewer questioning). While the primary motivation for this research program is methodological, it will also provide Important substantive information. In particular, our field test will provide precise estimates of the patterns of AIDS-related risk behaviors in Washington DC and surrounding counties. The annual incidence rate for AIDS In Washington DC (117 per 100,000) is currently more than six times the national average, and It exceeds that of any other state or territory in the United States. There Is presently no reliable data on the prevalence and patterns of AIDS-related risk behaviors for the Washington metropolitan area.