THE WORKING MODEL OF THE CHILD INTERVIEW: STABILITY OF THE DISRUPTED CLASSIFICATION IN A COMMUNITY INTERVENTION SAMPLE

The Working Model of the Child Interview (WMCI; C.H. Zeanah, D. Benoit, & M.L.Barton, 1986) assesses caregiver internal representation of his or her child and the relationship with the child, with a relatively new coding system for representations associated with disorganized attachment—WMCI-Disrupted (WMCI-D; A. Crawford & D. Benoit, 2009). In the present study, we investigated the stability of the WMCI-D classification using a sample of 62 mothers who completed the WMCI twice as part of their involvement in a randomized trial comparing an attachment-focused parent group to home visiting. Demographic information and measures of maternal sensitivity, parenting stress, and infant attachment also were obtained in the randomized trial. There was significant concordance between WMCI-D classifications over 8 months (from pretest to follow-up) (90% agreement; κ = .79), with 61% of mothers remaining disrupted, 29% remaining not-disrupted, 8% becoming disrupted, and 2% becoming not-disrupted. Compared to mothers with not-disrupted representations, mothers classified as disrupted had lower socioeconomic status, more parenting stress, and infants with less attachment security, ps < .05. These results suggest that the WMCI-D classification is stable over 8 months during infancy. The findings are consistent with research demonstrating stability for disorganized/unresolved/disrupted classifications, the validity of the WMCI-D classification, and the lack of intervention impact on disorganized attachment.