Differential Characteristics and Content Preferences of Media Violence Exposure Among Chinese Adolescents: A Multi-Method Study

To investigate the differential characteristics and content preferences of media violence exposure among Chinese adolescents, this study adopted a multi-method design combining a questionnaire survey and a scenario-based experiment. Study 1 (N = 601) utilized questionnaire data from multiple Chinese provinces to conduct descriptive analyses on adolescents’ primary devices and domains of online media access, as well as comparative analyses of media violence exposure across different demographic groups. Study 2 (N = 261) employed a scenario-based experiment to examine adolescents’ perceived differences in the level of violence, comprehension, and familiarity across three types of media violence videos: news, gaming, and animation. The results indicated that smartphones were the predominant device for adolescents’ online media use (91.00%), and watching videos was the most common online media activity (85.36%). Boys and adolescents from rural areas exhibited significantly higher media violence exposure than girls and their urban peers (t = 5.98, 3.80，ps 0.05). Regarding content preferences, adolescents perceived news videos as significantly more violent than gaming videos, which in turn were perceived as more violent than animation videos (ps < 0.05). In contrast, adolescents reported significantly higher levels of comprehension and familiarity with gaming and animation videos than with news videos (ps < 0.05). The findings indicate that media violence exposure varies by gender and residential context. Adolescents’ subjective evaluations of violent media differed across content types: news videos were perceived as more violent but less comprehensible and familiar than gaming and animation videos. This study provides empirical support for developing targeted, content-sensitive prevention and intervention strategies to reduce Chinese adolescents’ exposure to harmful media violence. Future studies should further refine violent content subtypes and incorporate objective violence intensity controls to extend these findings.