LAUSR

AIMS Multiple social influences affect cannabis use in adolescents, including parental and peer cannabis use norms. However, the mechanisms of influence underlying these social influences remain unclear. Recent studies have suggested that cognitions about cannabis use and the effects of cannabis may mediate social influences. The current study explored the relationship between automatic self-generated cognitions and their relationship with parental influences on cannabis use in a sample of n = 675 11 to 16-year-old adolescents over three years (Mean Age: 13.96, SD = 0.88, 56.4% female). METHODS Participants reported perceptions of parental cannabis use and completed a cannabis word association task (CWAT), an open-ended cannabis outcome expectancy liking (COEL) task, and measures of cannabis use in the past year. RESULTS Perceived parental use did not directly predict cannabis use two years later. However, a latent construct loading on both CWAT and COEL scores strongly predicted cannabis use over the following year. Structural modelling demonstrated that the association between previous cannabis use and parental cannabis use and adolescents' cannabis use over the next two years was fully mediated by cognitions. CONCLUSION The results of the study are discussed and interpreted through the lens of dual-process theories.