Investigating the Persuasive Effects of Testimonials on the Acceptance of Digital Stress Management Trainings Among University Students and Underlying Mechanisms: A Randomized Controlled Trial

Objective: This experiment aims to investigate the influence of narrative information varying in the degree of perceived similarity and source credibility in supplemented testimonials on the acceptance of digital mental health services (digi-MHSs). Methods: In fall 2020, n=231 university students were randomly assigned to an active control group (aCG, n=55, “information only”) or one of three intervention groups (IGs) receiving information plus different testimonials being presented either by nonacademic staff (IG1, n=60), university students (IG2, n=58) or experts (IG3, n=58). We assessed mediation effects of similarity and credibility on acceptance in terms of attitudes and usage intentions. Results: Exposure to testimonials was associated with higher usage intentions (d=0.50) and more positive attitudes toward digi-MHSs (d=0.32) compared to mere information (aCG). Regarding source-related effects, one-way ANOVA showed group differences in intentions ((Formula presented.) =0.13) that were significantly higher after exposure to testimonials targeted at students than in the other groups after adjusting for baseline intentions ((Formula presented.) =0.24). Concerning underlying mechanisms, there were full mediation effects of similarity (IG1 versus IG2) on attitudes [95%CI (0.030, 0.441)] and intentions to use digi-MHSs [95%CI (0.100, 0.528)] and of credibility on attitudes [IG2 versus IG3; 95%CI (−0.217, −0.004)], all favoring students’ testimonials. Conclusion: Overall, this study indicates that the acceptance of digi-MHSs can be substantially increased by providing a simple, context-sensitive information intervention, including testimonials by university students. Since we identified mediating effects of credibility on cognitive attitudes and similarity on affect-driven intentions, a future trial could vary these features using narrative versus statistic information on digi-MHSs.