CHS: Small: Investigating and Designing for Behavioral Improvement in Online Community Moderation

这是一项关于在线审核系统对人类影响的研究，该系统通过向他们确定具有破坏性的用户发布内容删除或帐户暂停等惩罚措施来处理破坏性的在线行为，例如攻击性语言和仇恨言论。这些审核系统通常无法为受惩罚的用户提供足够的支持，解释他们为什么受到惩罚，并建议他们如何改进。公平、问责和透明度方面的严重限制导致在线节制和社区福祉面临巨大挑战。受惩罚的用户可能不理解惩罚背后的原理，并有可能成为屡犯者。当新来者因违反他们以前不知道的社区规范而受到惩罚时，这就更具挑战性。该研究网站是一个人口众多的在线社区，研究将记录和描述人类惩罚互动（HPI），包括用户如何体验惩罚，用户的惩罚后行为，以及用户使用哪些支持资源来更好地了解社区行为标准和行为改善。这项研究有两个目标。首先，实证方法，如访谈和调查将被用来调查和理论化的主要社会技术层面的HPI。这将通过阐明与惩罚和被惩罚的用户相关的交互和体验来扩展现有的适度文献。第二，更集中的实证方法，如叙事访谈和焦点小组将用于确定用户发起的方式来理解社区规范和行为改善，重点是用户已经从现有的支持资源。行为改变的认知和社会理论将用于理解现有的，用户发起的行为改善方式。该奖项优先考虑对破坏性行为的动态、不断发展的看法，追踪用户在线行为的时间发展，并研究和理论化用户内化社区规范和改善行为的关键时刻。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

This is a study of human implications of online moderation systems that deal with disruptive online behaviors, such as offensive language and hate speech, by issuing penalties such as content removal or account suspension to users they determine to be disruptive. These moderation systems usually fail to provide punished users enough support in terms of explaining why they are punished and suggesting how they can improve. Such severe limitations in fairness, accountability, and transparency lead to enormous challenges to online moderation and community wellbeing. Punished users may not understand the rationale behind penalties, and risk becoming repeat offenders. This is even more challenging when newcomers are punished for violating community norms which they were previously not aware of. The study site is a high-population online community, where the research will document and describe human-punishment interaction (HPI) in terms of how users experience punishment, what are users' post-penalty actions, and what support resources users use for a better understanding of community behavioral standards and behavioral improvement. The research has two goals. First, empirical methods such as interview and survey will be used to investigate and theorize major sociotechnical dimensions of HPI. This will extend the existing moderation literature by articulating interactions and experiences associated with punishment and punished users. Second, more concentrated empirical methods such as narrative interview and focus group will be used to identify user-initiated ways of understanding community norms and behavioral improvement, with a focus on existing support resources that users have drawn from. Cognitive and social theories of behavior change will be used to understand existing, user-initiated ways of behavior improvement. This prioritizes a dynamic, evolving view of disruptive behavior, tracing the temporal development of user online behavior and studying and theorizing critical moments where users internalize community norms and improve behavior.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.