EAGER: CHS: Examining Self-Harm and Suicide Contagion Risks of Viral Social Media Challenges on Youth and Young Adults

该项目将研究潜在有害的社交媒体挑战，参与者记录自己参与特定活动的情况，在社交网络中分享经验，并鼓励其他人参与。一些挑战有积极的影响。例如，ALS冰桶挑战要求人们在头上倒一桶冰水，以提高人们对卢·格里克疾病的认识，并为其提供资金。然而，其他人鼓励人们从事有可能对身体或心理造成伤害的行为。例如，肉桂挑战涉及吃一勺肉桂，这可能会导致严重的呼吸窘迫，而蓝鲸挑战鼓励一系列日益自我伤害的行为，最终导致自杀。尽管围绕这种有害的挑战有很多民间传说和传闻，但关于它们的实际研究和数据很少。为了解决这个问题，项目团队将进行一系列访谈研究，并对围绕肉桂和蓝鲸挑战的社交网络数据进行量化分析。分析将考察导致人们参与和传播这些挑战的个人层面和信息层面的因素。这一结果将为制定预防措施提供信息，以减轻有害的互联网病毒挑战的传播。调查结果还将有助于更好地了解如何保护面临挑战的脆弱个人免受参与的直接潜在风险和间接潜在风险的影响(例如，通过自我伤害的正常化和参与挑战的同行压力)。该项目将解决一些具体的研究问题，包括：1)更好地评估病毒式社交媒体挑战对社交媒体用户造成实际伤害和构成公共健康风险的程度，以及2)确定包含这些挑战的消息以及接收和传播这些挑战的个人的特征，以预测挑战的风险、采用和传播。这项工作被组织成两个主要推动力。首先，该团队将对大约30名青少年和年轻的成年社交媒体用户(年龄在13-25岁之间)和/或他们的家人进行半结构化的回溯性采访，他们报告在参与研究时普遍存在的社交媒体挑战后受到伤害或实施自我伤害。这些问题将被组织起来，以评估已知的导致自我伤害传染效应的因素(如对挑战的图形描述、同伴压力和对自我伤害活动的支持)，以及探索人们参与此类挑战的自身动机和决策。第二个主要推动力涉及对来自五个流行社交媒体网站的关于自我伤害挑战的数字跟踪数据进行更大规模的量化分析。研究的这一部分首先将包括构建和清理数据集(在保护个人信息和捕获与挑战相关的数据方面)。接下来，研究人员将对挑战信息如何偏离自杀预防资源中心的循证安全信息指南进行定性编码。这些分析旨在确定作者用来分散挑战的策略以及每个挑战构成的潜在风险。最后，计算建模将消息特征与消息覆盖范围的度量(如点赞、查看和分享)联系起来。总而言之，这些研究活动将为现实世界的社交媒体行为和围绕这些有害挑战的个人决策提供亟需的经验证据。这些数据可用于为病毒挑战的传播以及减轻其未来危害的干预措施的理论发展提供信息。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This project will study potentially harmful social media challenges, in which participants record themselves engaging in specific activities, share the experience in social networks, and encourage others to participate. Some challenges have positive effects. For example, the ALS Ice Bucket challenge involved people dumping a bucket of ice water over their heads to raise awareness of and funding for Lou Gehrig's disease. Others, however, encourage people to engage in behaviors that risk physical or psychological harm. For example, the Cinnamon challenge involves eating a spoonful of cinnamon, which can cause severe respiratory distress, while the Blue Whale challenge encourages a series of increasingly self-harmful behaviors culminating in suicide. Although there is much folklore and hearsay around such harmful challenges, actual studies and data about them are sparse. To address this, the project team will conduct a series of interview studies along with quantitative analysis of social network data around the Cinnamon and Blue Whale challenges. Analyses will examine both individual-level and message-level factors that lead people to participate in and spread these challenges. The results will inform the development of preventative measures to mitigate the spread of harmful viral Internet challenges. Findings will also provide a better understanding of how to protect vulnerable individuals who are exposed to the challenges from both the direct potential risks of participating and indirect potential risks (e.g., through the normalization of self-harm and peer pressure to engage in it).The project will address a number of specific research questions, including: 1) better assessing the extent to which viral social media challenges have caused real harm and pose a public health risk to social media users, and 2) identifying characteristics of both the messages containing these challenges and the individuals receiving and spreading them that predict risk, adoption, and spread of the challenges. The work is organized into two main thrusts. In the first, the team will conduct semi-structured, retrospective interviews with approximately 30 adolescent and young adult social media users (ages 13-25) and/or their families, who report being harmed or committing self-harm after engaging in social media challenges prevalent at the time of the studies. The questions will be structured to assess factors known to contribute to self-harm contagion effects (such as graphic depiction of the challenge, peer pressure, and support for self-harmful activities), as well as to probe people's own motivations and decision-making around participating in such challenges. The second main thrust involves larger-scale quantitative analysis of digital trace data about self-harmful challenges from five popular social media sites. This portion of the study will first include constructing and cleaning the dataset (both in terms of protecting personal information and capturing data relevant to the challenges). Next, the researchers will employ qualitative coding of how the challenge messages deviate from the Suicide Prevention Resource Center's evidence-based safe messaging guidelines. These analyses are intended to identify strategies that authors use to spread the challenges as well as the potential risk each poses. Finally, computational modeling will relate message characteristics to measures of message reach such as likes, views, and shares. Together, these research activities will provide much-needed empirical evidence of real-world social media behavior and individual decision-making around these harmful challenges. The data can be used to inform theory development regarding the spread of viral challenges, as well as interventions to mitigate their future harms.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.