Preconditions for guardianship interventions in cyberbullying: Incident interpretation, collective and automated efficacy, and relative popularity of bullies

This paper examines the preconditions for direct and indirect interventions by guardians in cyberbullying incidences and, conversely, when automated prevention and detection systems are imperative and likely to be the most useful. A total of 316 young adults read messages of cyberbullying scraped from Twitter with varying degrees of relative popularity status between the sender and receiver. The respondents were then surveyed to measure their willingness to intervene either indirectly or directly in response to these instances of cyberbullying. The results show respondents expressed a greater willingness to intervene both as incidents are interpreted as cyberbullying and when their perceived severity increased. Perceptions of collective and self-efficacy (but not automated efficacy) also mattered for willingness to intervene in cyberbullying. The results also show that participants were more willing to intervene indirectly when the bully was more popular than the victim. Implications of these findings for the guardian and bystander scripts and for automated detection and prevention systems are discussed.