Evaluating a brief self-affirmation intervention to increase college students’ intentions to reduce internet use: A randomized controlled trial

Background: Internet overuse among youth is a significant health problem. Although self-affirmation interventions combined with health risk information have shown promise in addressing unhealthy behaviors, their application to problematic Internet use remains largely underexplored—particularly in Eastern and Chinese contexts where cultural factors may shape intervention effectiveness. This study is the first to evaluate the efficacy of a brief, digital self-affirmation intervention on message acceptance and behavioral intention to reduce Internet use among a large sample of Chinese college students, while exploring the moderating roles of demographics, risk status, and behavioral characteristics. Method: 354 college students were randomly assigned to an intervention group or a control group. Both groups received health risk information on Internet addiction and completed pre- and post-intervention assessments. Intervention efficacy and potential moderators were analyzed using generalized estimating equations and linear regression models. Results: The self-affirmation task did not significantly improve behavioral intention or message acceptance compared to the control task. However, both groups showed significant increases in behavioral intention from baseline (all p < 0.01). Moderation analyses indicated a stronger intervention effect on behavioral intention among participants from higher-income families (B for three-way interaction: 0.76; p = 0.02). Beneficial effects of self-affirmation intervention on message acceptance were observed for individuals with high baseline risk perception (β for two-way interaction: 0.41; p = 0.03) and self-efficacy (β: 0.39; p = 0.04), but potential backfire effects were observed for those with low risk perception and self-efficacy. Conclusions: This study represents the first trial of a digital self-affirmation intervention targeting Internet use among Chinese youth, highlighting the potential influence of cultural context and individual characteristics. Notably, the behavioral outcome assessed was limited to intention to reduce Internet use, rather than actual behavior. Future research should investigate cross-cultural validation, incorporate multidimensional behavioral assessments, and explore the feasibility of large-scale implementation.