A picture is worth a thousand words: children's representations of family as indicators of early attachment.

To ascertain whether attachment representations at age 7 are related to early attachment behaviour, family drawings of 123 7-year-olds of known infant attachment status (25 avoidant, 80 secure, 18 resistant) were scored in four ways. Three of these were based in previous attachment research and one was based on a clinical method. The attachment-based coding schemes included specific markers for each attachment pattern (Kaplan & Main, 1985), global ratings (Fury, Carlson, & Sroufe, 1997) and efforts to classify each drawing as belonging to one of the three primary infant attachment groups (secure, avoidant, resistant). In the clinical scheme, children who had been resistant infants were distinguished from the others by use of overlapping and encapsulated figures. For the attachment based schemes, although individual markers were not successful in discriminating attachment groups, the more global approaches (aggregation of markers, global rating scales and judgments of attachment classification) succeeded in this task. In regression analyses controlling for concurrent child and parent measures, infant attachment did not make a significant contribution to predicting insecurity markers in drawings, although child current emotional functioning did. These findings linking attachment relationships with later representations of family relationships were in accord with the conception that avoidant attachment strategies de-emphasize intimate relationships, while resistant attachment strategies are preoccupied with close relationships. These links are most evident in global interpretive strategies rather than those that rely on specific markers.