Asymmetrical third-person effects on the perceptions of online risk and harm among adolescents and adults

Although research has identified a range of opportunities, risks, and harms related to online social networking, the public debate on online risks follows a set pattern by which members of older age groups (parents, regulators) hold a picture of members of younger age groups (teenagers, digital natives) at a uniformly high level of risk. Perceptions of online risk, however, are prone to thirdperson effects in which individuals perceive risks to be more apparent in others than themselves. This study investigated third-person effects across age groups to further our understanding of the set positions found in current public debate. Multivariate analysis was used to compare adolescent and adult users’ personal and third-person perceptions of common psycho-social risks associated with social networking engagement in a sample of 506 UK-based Facebook users (53% male; 13–77 years). Results indicated that rates of exposure to online vulnerabilities were similar for both age groups. However, differences in adult and adolescent perceptions of risk highlighted apparent mismatches between reported exposure to risk and an individual’s perceptions, with adults demonstrating lower personal perceptions and higher thirdperson perceptions of risk than their adolescent counterparts. The research considers the implications of risk perception on an individual’s online vulnerability.

Prior experience of online risks and harm is similar for adolescents and adults but older individuals consider themselves to be less likely to experience risk. They might feel safer online due to their life experience. "Online risks and vulnerability do not apply to adolescent users alone. Indeed, adolescents may have a more pragmatic perception of online risks and vulnerabilities than their optimistically biased adult counterparts, a divide that needs to be bridged in order to ensure safe and enjoyable user practices are experienced by all. " (Buglass et al., 2020: 8).