Social networks and networking that put adolescents at high risk

描述(由申请人提供)：本申请涉及广泛的挑战领域(01)行为，行为改变，以及预防和特定挑战主题01-AA-103*为负面健康行为的高危群体捕获社交网络信息研究社交网络对青少年健康行为的影响主要集中在评估朋辈影响和友谊网络内的选择。尽管友谊很重要，也许是最重要的青少年关系，但它们不一定代表将所有青少年置于负面健康行为风险中的关系。例如，可能是一些青少年受到他们想要的恋人或他们认为受欢迎的人而不是朋友的影响，更倾向于尝试负面的健康行为(吸烟、饮酒)。虽然在青少年健康行为领域的一些发现是站得住脚的，但关于收集网络数据的相当基础的科学仍然有待完成。这项研究建议通过比较调查数据收集的几个方面来调查社交网络数据如何被用来识别青少年的不良健康行为，如吸烟、酗酒或吸毒。这些比较对于确定衡量社会影响、从而设计有效干预措施的最佳方法是必要的。为了进行这项研究，我们将在南加州五所高中的10年级学生队列中收集新的纵向数据。虽然有证据表明，网络效应是强大和持久的，但关于如何在在校青少年中收集社交网络数据，仍然存在几个问题。这种情况经常发生，例如因为网络问题很少成为青少年健康研究的重点，而行为才是。在社交网络影响力领域，至少有五个悬而未决的问题将被调查。这些问题是：(1)与同伴影响和/或选择最相关的网络类型；(2)提出网络问题的相关边界；(3)联系强度重要的程度；(4)线上和线下网络之间的对应关系；以及(5)社交网络与社交身份和团队/团队关系的交集。第一个维度、网络类型和边界规格将随调查管理部门系统地变化，以测量该人群中多达25个不同的网络。新的数据和网络将用于在一个学年内的两个时间点评估吸烟、酒精和药物使用的社交网络风险因素。将根据以下分类办法，利用特定的完善的网络指标，利用基线数据来确定高危青年：未使用的学生：(A)使用网络的受欢迎学生；(B)使用友谊网络的受欢迎学生；(C)与使用其社交网络的部分相连接的桥梁或边缘群体。然后，根据网络类型、网络边界和联系强度的不同，分析将确定高风险分类的稳定性和可预测性。这项研究改变了调查问题的措辞和方法，以确定最显著的网络来评估同伴对青少年物质使用的影响和选择。数据是在一个学年的两个时间点收集的10年级学生的数据，还链接到青少年在社交网站上的在线个人资料。评估了基于网络和行为的风险分类的稳定性和可预测性。

DESCRIPTION (provided by applicant): This application addresses broad Challenge Area (01) Behavior, Behavioral Change, and Prevention and specific Challenge Topic 01-AA-103* Capturing Social Network Information for Groups at High Risk for Negative Health Behaviors Research on social network influences on adolescent health behavior has focused primarily on assessments of peer influence and selection within friendship networks. Although friendships are important, and perhaps the most important adolescent relationship, they may not necessarily represent the relationship that put all adolescents at risk for negative health behaviors. For example, it may be that some adolescents are more strongly influenced to try negative health behaviors (smoking, drinking) by their desired romantic partners or those they perceive as popular rather than friends. While several findings in the field of adolescent health behavior are well-established, considerable basic science regarding the collection of network data remains to be done. This study proposes to investigate how social network data may be used to identify adolescents at risk for negative health behaviors such as smoking, alcohol use, or drug use by comparing several aspects of survey data collection. These comparisons are necessary to establish the best methods for measuring social influence and hence designing effective interventions. To conduct this research we will collect new longitudinal data among a cohort of 10th grade students in five southern California high schools. While evidence indicates that network effects are strong and persistent, there remain several questions regarding how social network data should be collected among in-school adolescents. This often occurs for example because the network questions are rarely the focus of adolescent health research, the behaviors are. At least five unanswered questions within the field of social network influences will be investigated. These issues are the: (1) type of network that is most strongly associated with peer influence and/or selection; (2) relevant boundary for asking network questions, (3) extent to which tie strength matters, (4) the correspondence between online and offline networks, and (5) the intersection of social networks with social identity and group/team affiliations. The first dimensions, network type and boundary specification will be systematically varied with survey administration to measure up to 25 different networks in this population. The new data and networks will be used to assess social network risk factors for tobacco, alcohol, and substance use at two time points within one academic year. Baseline data will be used to identify at-risk youth using specific well-established network indicators based on the following categorization schemes: non-using students who are: (a) popular students embedded within using networks; (b) embedded in using friendship networks; and (c) bridges or marginals connected to using segments in their social networks. Analyses will then determine the stability and predictability of high-risk categorizations given variations in network type, network boundary, and tie strength. This study varies survey question wording and methodology to determine the most salient networks for estimating peer influence and selection on adolescent substance use. Data are collected among 10th grade students at two points in time in one academic year and also linked to adolescents' online profiles on social networking sites. The stability and predictability of at risk categorizations based on network and behavior are estimated.