SEX AND AIDS METHODOLOGY SURVEY--MINORITY EXPANSION

性与艾滋病方法论调查--少数群体扩大(SAMS-ME)增加了一个 少数族裔样本过多(700名非洲裔美国人和800名西班牙裔 受访者)对最近完成的一个项目--性与艾滋病方法论 调查(SAMS)。最初的父项目SAMS是一个国家RDD(随机)项目 数字拨号)对2,030名18岁至49岁的连续居民进行调查 美国被设计用来测试三个与 参与偏差和测量误差的方法学问题 提出敏感和具有威胁性问题的电话调查。这 拟议的少数股扩大，SAMS-ME，将使分析成为可能 到目前为止，在非洲裔美国人和西班牙裔人口中的调查误差 在调查研究的方法论研究中被忽视。具体地说， SAMS-ME基于与SAMS相同的研究设计，将涉及 三个实验。第一个测试将测试三个家庭统计 方法：上一个生日，下一个生日，以及Kish方法 确定哪些方面可以在保持合作的同时加强回应 具有代表性的样本。第二个实验将测试被调查者是否 选择男性或女性面试官的控制，或相同或相反的控制 性别问题采访者，将加强受访者合作，减少 与性行为、堕胎和性行为相关的测量误差 虐待。第三个实验将测试询问是否敏感 非判断性环境中的信息将减少 关于性行为、堕胎和性虐待的问题。方法论 调查研究中的研究很少提供可概括为 人口多样性--人口和流行病学的核心 研究。如果没有这方面的知识，对行为的准确评估 对设计和实施有效的预防工作至关重要 是不可能的。这项研究的结果将开始逆转这一趋势 通过在性取向研究中提供调查误差的估计 并通过测试降低这一比例的技术 错误。

The Sex and AIDS Methodology Survey--Minority Expansion (SAMS-ME) adds an oversample of minorities (700 African American and 800 Hispanic respondents) to a recently completed project, the Sex and AIDS Methodology Survey (SAMS). The initial parent project, SAMS, is a national RDD (random digit dialing) survey of 2,030, 18 to 49 years of age in the contiguous United States that was designed to test three experiments related to methodological issues of participation bias and measurement error in telephone surveys that ask sensitive and threatening questions. This proposed minority expansion, SAMS-ME, will make possible analyses of survey error among African American and Hispanic populations heretofore neglected in methodological studies in survey research. Specifically, the SAMS-ME, which is based on the same study design as the SAMS, will involve three experiments. The first will test three household enumeration methods: the last birthday, the next birthday, and the Kish methods to determine those that increase respondent cooperation while maintaining a representative sample. The second experiment will test whether respondent control of selecting a male or female interviewer, or same versus opposite gender interviewers, will increase respondent cooperation and reduce measurement error related to questions on sexuality, abortion, and sexual abuse. The third experiment will test whether asking about sensitive information in a non judgmental context will reduce measurement error in questions on sexual behavior, abortion, and sexual abuse. Methodological studies in survey research rarely provide findings generalizable to the diversity of populations- central to demographic and epidemiological research. Without that knowledge, accurate assessment of behaviors critical to the design and implementation of effective prevention efforts are not possible. The results of this study will begin to reverse that situation by providing estimates of survey error in studies of sexuality among minority populations and by testing techniques that reduce that error.