Doctoral Dissertation Research: Extending Respondent-Driven Sampling to Social Norm Research and Development of Web-Based RDS Software

该项目将通过分析来自美国两所大学的不同选择性本科生的基于网络的RDS （WebRDS）样本，为受访者驱动抽样方法（RDS）做出一些贡献。RDS是一种新的统计方法，目前广泛用于研究隐藏和难以触及的人群。首先，该项目将RDS扩展到社会网络结构背景下的社会规范研究，将知识寻求规范的执行、遵守和感知映射到每个人群的潜在社会网络结构上。RDS特别适合这种分析，因为它提供了社会网络中行为定义关系的随机样本。整体社会网络结构将根据这些关系进行分析，并结合使用RDS同质性和隶属指数对构成这些关系的受访者的社会规范进行调查。因此，该研究将为社会规范与社会网络之间的相互作用提供重要的见解。其次，该项目开发和测试webbrds软件，这是RDS的一个完全自动化的在线变体，用于向一般科学界分发。最后，虽然RDS估计量在分析和计算上已被证明是渐近无偏的，但大多数RDS研究群体的隐蔽性阻碍了大规模、多地点的经验验证。本项目将通过比较RDS估计值与来自两个大型本科生群体的机构数据来提供这种验证。更大的研究界将受益于本项目开发的方法。首先，网络数据库系统软件的开发将为电子连接的机构提供非常快速和有效的抽样。这对于流行病学暴发的病例控制研究尤其有用，因为减少用于收集数据的时间可以预防更多的感染。本项目中发展的社会规范研究方法可以应用于任何可能通过人际接触传播的结果变量。更直接的是，该项目将研究大学生的求知规范，美国教育部最近认定大学生的素质在下降。作为博士论文研究改进奖，该奖项还将提供支持，使有前途的学生建立一个强大的独立研究生涯。作为支持调查和统计方法研究的联合活动的一部分，该研究得到了方法、测量和统计计划和联邦统计机构联盟的支持。

This project will make several contributions to Respondent-Driven Sampling methodology (RDS), a new statistical method now widely used in the study of hidden and hard to reach populations, through analysis of web-based RDS (WebRDS) samples of undergraduate students from two U.S. universities of varying selectivity. First, the project extends RDS to the study of social norms within the context of social network structure by mapping enforcement, compliance, and perception of knowledge-seeking norms onto the underlying social network structure of each population. RDS is especially well suited for such analysis because it provides a random sample of behaviorally defined ties from within a social network. The overall social network structure will be analyzed based on these ties and combined with survey questions regarding the social norms of respondents who make up those ties using the RDS homophily and affiliation indices. Consequently, the research will provide important insights into the interaction between social norms and social networks. Second, the project develops and tests WebRDS software, a fully automated online variant of RDS, for distribution to the general scientific community. Finally, while the RDS estimator has been shown to be asymptotically unbiased analytically and computationally, the hidden nature of most RDS study populations has prevented large scale, multi-site empirical validation. This project will provide such validation through comparison of RDS estimates with institutional data from two large undergraduate populations.The larger research community will benefit from the methods developed in this project. First, the development of WebRDS software will provide for very fast and efficient sampling of electronically connected institutions. This will be especially useful for case control studies of epidemiological outbreaks where a reduction in time spent on data collection can prevent additional infections. The method of social norm research developed in this project can be applied to any outcome variable that is potentially spread through interpersonal contact. More immediately, the project will study knowledge-seeking norms in university students, a population recently identified by the U.S. Department of Education as declining in quality. As a Doctoral Dissertation Research Improvement award, this award also will provide support to enable a promising student to establish a strong independent research career. The research is supported by the Methodology, Measurement, and Statistics Program and a consortium of federal statistical agencies as part of a joint activity to support research on survey and statistical methodology.