Abstract:

The purpose of the present study was to investigate the mechanisms by and conditions in which friendships predict mental health. Three plausible models were examined. First, the role of adolescent friendship and putative mediators on mental health problems was investigated. A model in which perceived friendship warmth/support leads to attitudes toward self (i.e., self-worth) and others (i.e., prosocial orientation), which in turn predicts mental health problems was proposed. Second, a moderation model in which the relation between perceived friendship warmth/support and mental health is moderated by perceived caregiver warmth/support was examined. Finally, a combined mediation moderation model was proposed in which perceived caregiver warmth/support moderates the impact of friendship on the two mediators (i.e., self-worth and prosocial orientation), which in turn affect mental health problems. Given prevalent findings of gender differences in the field of friendship research, gender invariance of each model was also investigated. Three general findings emerged from this study. First, the effect of friendship on adjustment differed by gender. As expected, girls reported more warmth and support from their same-gender friendships than boys. Furthermore, friendship is related to different outcomes for boys and girls. Second, friendship not only predicted mental health problems directly but also indirectly through self-worth and prosocial behavior. As hypothesized, self-worth mediated the relation between friendship and later loneliness (for girls) and prosocial behavior mediated the relation between friendship and mental health problems (for boys). Third, friend's direct and indirect effect on mental health problems differed as function of level of caregiver warmth/support.