Cognitive underpinnings of narrative attachment assessment.

Traditional story completion methods used to examine attachment representations in childhood draw heavily on socially significant content and clinical judgment. Using these methods with 37-month-olds, Bretherton, Ridgeway, and Cassidy (1990) found that attachment security scored from story completions were related to a wide range of secure base, personality, intellectual, and family variables. We examined story completions from 24 of Bretherton et al. 's (1990) subjects who had also produced story completions at 54 months, but scored passage length (idea units) and scriptedness. Results captured much of the attachment-related variance associated with the traditional scoring, but had better discriminant validity vis-à-vis general developmental level. These results indicate that analysis of cognitive variables underlying conventional scoring can advance understanding of attachment representations and their relations to the organization and content of attachment-related narratives.