“Everyone is Mean to Me” : On the Origins, Consequences, and Malleability of Hostile Attribution Biases in Young Children

Abstract

Some children are prone to interpret others’ behavior as stemming from hostile intentions, even when the actual intentions of the other person are unclear. The aim of this dissertation was to extend our knowledge of such hostile attribution biases in young children (aged 3 to 8). The findings in this dissertation were obtained from seven empirical studies, organized in four chapters, and centered around three overarching aims: the origins, consequences, and malleability of hostile attribution biases in young children. Concerning origins, results showed that children with higher levels of hostile attribution bias had less advanced social-cognitive skills (i.e., they were less skilled at reasoning about others’ mental states), had a heightened sensitivity to negative social outcomes, and had parents who modeled less nonhostile attributions when discussing social situations with them. Children’s hostile attribution bias was not related to basic social-cognitive skills (i.e., the ability to understand others’ mental states) or to victimization by peers. Concerning consequences, links were established between young children’s hostile attribution bias and aggression (r = .20, cross-dissertation meta-analysis). This correlation is similar to correlations obtained for middle childhood and adolescence, suggesting that the link between children’s hostile attribution bias and aggression is robust across development. Concerning malleability, two methods were explored that may reduce hostile attribution biases in young children. The first method was parent-child discussion: asking parents to model benign attributions while discussing a picture book about social mishaps with their children. The second method was self-persuasion: asking children themselves to advocate why peers who caused a mishap may have had nonhostile intentions. Results showed that both methods reduced children’s hostile attribution bias, but not their aggression. These methods’ potential for intervention should be established in future research. To conclude, the research in this dissertation has illustrated novel ways to investigate social cognition in young children. We hope its findings will help scholars and practitioners to better understand and treat hostile attribution biases in early childhood.