What you use decides what you get: comparing classificatory procedures for the Adult Attachment Interview in eating disorder research.

Studies of attachment and eating disorders use different types of measures, including different coding procedures for the Adult Attachment Interview (AAI). Generalizability of findings across studies is therefore uncertain. We compare the Main & Goldwyn procedure with the Dynamic Maturational Method, the two most common procedures for classifying AAI in eating disorder research. The sample consists of 20 female patients with a diagnosis of anorexia nervosa (mean age 22.9 (3.5) years). Attachment insecurity is by far most common, regardless of procedure. Within the insecure categories, there is little overlap between procedures in comparable categories. Both procedures discriminate between Anorexia subgroups (restricting vs bingeing), but do so differently. Findings suggest that comparing findings across methods, beyond the secure/insecure dichotomy, should be avoided.