Objective: Most studies examining peer influence on drug-use among youth experiencing homelessness (YEH) have relied on perceptions of peer use rather than measuring peers’ self-reported use, an approach that can inflate estimates of peer substance use behavior. Sociometric network data provide an optimal mechanism to compare perceptual data to actual self-reports from peers. Method: Using an event-based approach, we recruited a sample of YEH (N = 241), ages 13–25 years, between October 2011 and February 2012 from 2 drop-in centers—1 in Hollywood, CA, and 1 in Santa Monica, CA. We used multilevel multiple membership modeling to investigate participant-level, network-level, and relationship-level factors associated with the accuracy of respondent perceptions. Results: The accuracy of respondent perceptions of peer substance use was high (70%–90% depending on the substance). The individual- and network-level factors associated with accuracy or inaccuracy varied by substance. Conclusions: Interventions designed to reduce the risk of substance use among YEH may benefit from using a social norms approach that emphasizes changing norms at a community/group level rather than at an individual level.

中文翻译：

---

社会计量学的方法，以了解在经历无家可归的年轻人中物质使用知觉中的一致性

目的：大多数研究同龄人对无家可归的青年人对毒品使用的影响的研究都依靠对同龄人使用的看法，而不是衡量同龄人的自我报告的使用情况，这种方法可以夸大对同龄人物质使用行为的估计。社会计量网络数据提供了一种最佳的机制，可以将感知数据与对等方的实际自我报告进行比较。方法：采用基于事件的方法，我们从两个咨询中心（2011年10月至2012年2月）招募了YEH（N = 241）样本，年龄在13至25岁之间，地点分别来自2个咨询中心，其中1个位于加利福尼亚好莱坞，而1个位于圣诞老人莫妮卡，CA。我们使用多级多成员模型来研究与受访者感知的准确性相关的参与者级，网络级和关系级因素。结果：受访者对同龄物质使用的理解的准确性很高（70％–90％取决于物质）。与准确性或不准确性相关的个人和网络级别的因素因实质而异。结论：旨在减少YEH中物质使用风险的干预措施可能会受益于使用一种社会规范方法，该方法强调在社区/群体层面而不是个人层面上改变规范。