Mother-infant interaction during the first eighteen months and attachment security at the end of the second year

This study aimed at reconstructing attachment quality at the end of the sec ond year from quality of mother-infant interaction in the first 18 months o f the child's life. At the infants' ages of 3, 6, 9, 12, and 18 months, 74 mother-infant dyads were videotaped in semi-structured play situations that were subsequently analyed based on the CARE index (Crittenden, 1988). At 2 1 months, the children and their mothers were seen in the ,,Strange Situati on" (Ainsworth & Wittig, 1969). Attachment quality was assessed with the ,, Preschool Assessment of Attachment" (Crittenden, 1995), an extension of the classical systems that allows for the inclusion of more symbolic behaviors of preschool children. At all but one time period of interaction observati on, maternal sensitivity as well as infant cooperativity significantly diff erentiated later secure from insecure attachment. High and stable maternal sensitivity close in time to the attachment assessment was typical for secu rely attached children. Maternal behavior characterized as high in overt an d covert control as well as by unresponsivity early in development and cont inued preponderance of strained adaptation over unstrained cooperativity in the infant signaled later insecure attachment, particularly of the A-type. Individual differences in maternal behavior proved to be more stable than in infant behavior. Diachronic patterns in the interaction components of mo thers and of infants suggest that there are developmentally and attachment- quality specific changes in interaction during the infants' first year of l ife. Results are interpreted with reference to a transactional and dynamic view of early development and socialization.