The Use of Negative Interface Cues to Change Perceptions of Online Retributive Harassment.

Online retributive harassment refers to a range of abusive online behaviors targeted at offenders with the intent of sanctioning norm violations. Online retributive harassment is common in online interactions, can be devastating in its effect, and is hard to moderate. We examined the role that negative interface cues (i.e., Dislikes, flags) might have in changing people's harassment perceptions to mitigate this activity. We conducted a 4x2 between-subjects experiment (N = 242) to test the effects of interface cues (control with Likes only vs. Dislikes outnumbering Likes vs. Likes outnumbering Dislikes vs. a flag with Likes) and harassment severity (low vs. high) on how people perceive retributive harassment. We hypothesized that Dislikes and flags, despite the presence of Likes, signal social disapproval and descriptive norms against harassment and thus reduce bystanders' belief that retributive harassment is appropriate, deserved, or justified. We found that Dislikes can be effective when they outnumber Likes in high-severity harassment but backfire when outnumbered by Likes. A flag, contrary to popular practice, does not have a significant mitigating effect on perceptions of retributive harassment. We demonstrate the potential of negative interface cues to signal anti-harassment norms to bystanders and discuss alternatives to social media platforms' one-size-fits-all content moderation approach.