Adult Attachment Interview

The Adult Attachment Interview (AAI) is an interview and coding system developed by Mary Main and colleagues that yields a classification of an adult's current state of mind with respect to attachment. The following entry presents a description of the AAI, its coding, and basic research findings.

John Bowlby hypothesized that individuals develop an “internal working model” or cognitive representation of the functioning and significance of close relationships that guides behaviors, thoughts, and feelings in those relationships. The AAI was developed to assess the security of an adult's general or overarching working model of attachment, not to assess the security or quality of attachment to any particular person. The use of a narrative to assess adult attachment is based on the idea that attachment-related cognitive processes vary as attachment behavioral patterns do in infants and children, and these processes are reflected in the language of older individuals. The AAI is not published; it requires skillful interviewing and intensively trained coders.

The AAI has 18 autobiographical questions, starting with family background, an overview of the relationship with parents or parental figures in childhood and a request for five adjectives describing the relationship with each parent. The individual is asked for a specific experience that illustrates each adjective. The interview asks about ordinary childhood experiences in which the attachment system is activated (e.g., upset, injury, illness), as well as experiences of loss and abuse. The individual is asked how the childhood attachment experiences have influenced their adult personalities and about hopes for their children.

The AAI is audiotaped and transcribed verbatim for scoring. Scoring is based on (a) the coder's assessment of attachment experiences with parents, (b) the language used in the interview, and [Page 40](c) the ability to give an integrated, believable account of experiences.

Experience scales are used to rate the degree to which each parent demonstrated loving behavior; demanded premature independence in the child (rejection); engaged in involving, role-reversing behavior; pressed the child to achieve; and/or neglected the child. The experiences are inferred by the coder and are not necessarily accurate autobiographical information.

Several scales assess states of mind, with the coherence of transcript score being the strongest correlate of overall security. High coherence means that the narrative demonstrates consistency, clarity, and cooperative collaboration with the interview process. Scales associated with insecurity include idealization, insistence on lack of recall, active anger, derogation, fear of loss, and passivity of speech. Two scales identify unresolved or disorganized states of mind associated with experiences of loss and/or abuse.