To intervene or not to intervene: young adults' views on when and how to intervene in online harassment.

Lay Summary What influences someone to step in when they see harassment taking place online? We asked 67 young adults about their experiences of online harassment using focus groups and identified key themes. We found that bystanders are generally reluctant to intervene unless the harassment is obvious, severe, and when they knew the victim well. When they did intervene, participants preferred to step in discreetly through reporting (to social networking sites) or private messaging to offer emotional support. Reporting was not always seen in a positive light, with participants describing it as an "empty" experience due to the lack of feedback about what happened next. Bystanders were concerned about making the situation worse, becoming a victim themselves, and not having the right skills to intervene. In the case of celebrities and influencers, online harassment was seen as "part of the job." This was amplified by a feeling of detachment due to the fact they cannot physically see the victim online. Our study highlights several barriers that discourage bystanders from intervening and suggests ways of removing these barriers through design and messaging. We propose a new way of conceptualizing online interventions along a "spectrum" varying in levels of visibility and effort. Incidents of online harassment are increasing and can have significant consequences for victims. Witnesses ("digital bystanders") can be crucial in identifying and challenging harassment. This study considered when and how young adults intervene online, with the aim of understanding the applicability of existing theoretical models (i.e., Bystander Intervention Model; Response Decision-Making Framework). Thematic analysis of eight focus groups (UK community sample, N = 67, 18-25 years) resulted in five themes: Noticing and Interpreting the Harassment, Perceived Responsibility for Helping, Consequences of Intervening, Perceived Ability to Make a Difference, and Deciding How to Help. The online context amplified offline preferences, such as greater preference for anonymity and perceived costs of intervention (e.g., social costs). Intervention strategies varied in visibility and effort, preferring "indirect" micro-interventions focused on supporting victims. A new, merged model specific to digital bystanders is proposed, with implications for the design and messaging on social networking sites discussed.