Abstract
PURPOSE: Web-based tobacco prevention programs for adolescents have stressed human-computer interaction, but they have not yet extensively applied social interactivity (i.e., computer-mediated or face-to-face interactions). This study examines if prior tendencies for positive social influence (PSI), negative social influence (NSI), and having friends who smoke (HFS) moderate the success of a web-based program for smoking prevention. METHODS: Participants were 101 adolescents (aged 12-18 years) from the ASPIRE-Reactions study, a randomized controlled trial comparing a program called ASPIRE with its text-based version. Knowledge of tobacco consequences and intention to smoke were assessed at baseline and end-of-treatment. Tendency for PSI (i.e., avoid tobacco when advised by friends) and NSI (i.e., accept tobacco when offered by friends) were measured at baseline. Repeated-measures mixed-effect modeling was used for hypothesis-testing. RESULTS: While controlling for ASPIRE effects, both NSI and PSI predicted lower intention to smoke. Adolescents with high NSI were more likely to show a group difference with respect to change in intention to smoke, but not knowledge. Although not significant, this moderation effect was observed in the expected direction with PSI, predicting intention to smoke and knowledge. HFS significantly moderated the effect of ASPIRE on knowledge. Associations with depression and internet use are also described. CONCLUSION: The results suggest that adolescents with high tendencies for NSI may particularly benefit from web-based interventions such as ASPIRE. Also, web-based interventions may benefit from peer-to-peer interactions, boosting PSI. While current web-based programs include human-computer interaction as their main feature, this study suggests considering social interactivity.