An Experimental Evaluation of How Survey Measurement of Sexual Identity Affects Population Estimates of Identity-Based Differences in the Prevalence of Adverse Health Outcomes

摘要 大量的研究已经提出了证据表明，在不同的人群中， 根据不良健康结果的发生率，按性别特征定义， 认同为性少数的个人（例如，女同性恋、男同性恋、双性恋）的风险更高 物质使用障碍，精神健康障碍，自杀，歧视和危险的性行为。 这项研究导致了旨在了解和减少这些疾病的公共卫生政策的实施。 地方、州和国家各级的差异。这项研究的大部分利用了二次分析 收集性别认同和各种健康指标的大型国家调查数据集， 行为，因此这些研究中报告的估计差异的有效性在很大程度上依赖于高- 性别认同建构的质量测量。尽管最近有大量的定性工作， 为了改进这种测量方法，我们现有的关于这些差异的知识大多基于 调查，收集相对简单的封闭式措施，这一建设。这种做法带来了风险 调查受访者在性别认同方面被错误分类，特别是如果受访者没有 认为所提供的少数性别认同选择适用于他们。这种错误的分类将 最终削弱人口对性别认同与健康结果之间关联的估计， 潜在地低估差异或提供差异方向的误导性估计。的 提议的PI和co-I最近发表了一份在国家层面上对这一重大问题的演示， 使用NICHD资助的全国家庭成长调查的新实验数据进行的初步研究 （NSFG）。 这一拟议的二次分析项目的目的是在这一初步工作的基础上， 特别利益通知的目的和范围NOT-HD-20-022。通过对数据进行二次分析 从NSFG的五年（2015-2019）中收集，其中两个大型国家半样本随机 被分配接受关于性别认同的问题的不同版本，并进行平行分析， 从全国药物使用和健康调查（2015-2019年）的相同数据中复制， 该项目将评估性别认同之间差异估计衰减的可能性， NICHD感兴趣的其他健康结局分布的亚组，并扩展了 对社会人口分组的衰减问题，以及通过以下方式可能缓和这些差异： 与保护性少数群体有关的州一级政策。拟议工作的结果将告知 现有的政策和激励未来的实验研究，进一步评估改进的工具， 在国家调查中衡量性别认同。

ABSTRACT A large body of research has presented evidence of significant differences among subgroups of individuals defined by sexual identity in terms of the prevalence of adverse health outcomes, where subgroups of individuals who identify as sexual minorities (e.g., lesbian, gay, bisexual) have been shown to be at higher risk of substance use disorders, mental health disorders, suicidality, discrimination, and risky sexual behaviors. This research has led to the implementation of public health policies designed to understand and reduce these differences at the local, state and national levels. The majority of this research has utilized secondary analyses of large national survey data sets that collect measures of both sexual identity and the various health behaviors, so the validity of the estimated differences reported in these studies therefore relies heavily on high- quality measurement of the construct of sexual identity. Despite a large amount of recent qualitative work on methods for improving this measurement, much of our existing knowledge about these differences is based on surveys that collected relatively simple closed-ended measures of this construct. This practice introduces a risk of survey respondents being misclassified in terms of their sexual identity, especially if respondents do not perceive that the small number of sexual identity options provided apply to them. This misclassification will ultimately attenuate population estimates of the associations between sexual identity and health outcomes, potentially understating differences or providing misleading estimates of the directions of the differences. The proposed PI and co-I recently published a demonstration of this significant problem at the national level in a preliminary study using novel experimental data from the NICHD-funded National Survey of Family Growth (NSFG). This proposed secondary analysis project aims to build on this preliminary work, consistent with the purpose and scope of Notice of Special Interest NOT-HD-20-022. By conducting secondary analyses of data collected from five years of the NSFG (2015-2019), where two large national half-samples were randomly assigned to receive different versions of a question about sexual identity, and also conducting parallel analyses of identical data from the National Survey of Drug Use and Health (2015-2019) for replication purposes, this project will evaluate the possibility of attenuation in estimates of the differences between sexual identity subgroups in the distributions of additional health outcomes of interest to NICHD, and extend the study of the attenuation problem to socio-demographic subgroups and the potential moderation of these differences by state-level policies related to the protection of sexual minorities. Results from the proposed work will inform existing policies and motivate future experimental research further evaluating improved tools for the measurement of sexual identity in national surveys.