Adolescents' attachment and theory of mind: Evidence for mediating effects of age, gender, and type of attachment relationship

Introduction. Research on the relationship between attachment and 'theory of mind' has so far focused on children, emphasizing the importance of this relationship for the development of both attachment and 'theory of mind' (e.g., Fonagy, Gergely, Jurist, & Target, 2002). By contrast, though adolescence is generally recognized to be a crucial period for the development and differentiation of attachment relationships, on the one hand, and the sophistication of 'theory of mind' abilities, on the other, little is known about the relationship between adolescents' attachment and 'theory of mind'. In particular, there is no research on the relationship between normal adolescents' attachment, as assessed by standard self-report measures, and normal adolescents' 'theory of mind', as assessed by standard measures of 'theory of mind'. Therefore, the aim of the present research was to provide first evidence about how adolescents' self-reported attachment style relates to their performance on a standard 'theory of mind' task, and thereby to test the partly opposing, though not necessarily contradicting theoretical predictions derived from Mikulincer and Shaver's (2007) model of attachment system activation and dynamics, on the one hand, and Dykas and Cassidy's (2011) integrative life-span-encompassing model of attachment-related social information processing, on the other. While Dykas and Cassidy's model predicts that secure individuals, due to their inferior need to recur to defensive exclusion, perform better on 'theory of mind' tasks than insecure individuals, Mikulincer and Shaver's model predicts that this relationship holds only for one type of insecure individuals, i.e. for avoidant individuals, but not necessarily for the other type, i.e. for anxious individuals: in fact, due to their 'hypervigilance', more anxiously (and thus less securely) attached individuals might perform better on 'theory of mind' tasks than less anxiously (and thus more securely) attached individuals. According to both models, the predicted effects should be rather independent from age, gender, and type of attachment relationship.