A matter of perception? investigating subjective and objective exposure to hate speech with a survey and mobile longitudinal linkage study

An increasing number of media users report frequent encounters with hate speech on the internet. Content moderation is only effective when the applied criteria align with users’ perceptions of hate speech. To explore what media users perceive as hate speech and which factors influence this perception, we used a multi-methods approach. First, we conducted a survey with a representative sample of the Swiss population (N = 2000). Second, participants who reported frequent exposure to hate speech took part in a two-week mobile longitudinal linkage study, uploading screenshots and answering questions each time they encountered hate speech. We analysed N = 564 screenshots to see if they met common academic definitions of hate speech. Our findings show that impoliteness and insults are more likely to be considered hate speech when they affect one’s social identity, and that self-reports indicate higher exposure to hate speech than what was documented via screenshots.