Dynamic simulation of social media challenge participation to examine intervention strategies

Recently, the use of social media by adolescents and young adults has significantly increased. While this new landscape of cyberspace offers young Internet users many benefits, it also exposes them to numerous risks. One such phenomenon receiving limited research attention is the advent and propagation of viral social media challenges. Several of these challenges entail self-harming behavior, which combined with their viral nature, poses physical and psychological risks for the participants and the viewers. In this paper, we show how agent-based modeling (ABM) can be used to investigate the effect of educational intervention programs to reduce participation in social media challenges at multiple levels—family, school, and community. In addition, we show how the effect of these education-based interventions can be compared to social media-based policy interventions. Our model takes into account the “word of mouth” effect of these interventions which could either decrease participation in social media challenge further than expected or unintentionally cause others to participate. We suggest that educational interventions at combined family and school levels are the most efficient type of long-term intervention, since they target the root of the problem, while social media-based policies act as a retrospective solution.