Your name here

Fury, G., Carlson, E.A. & Sroufe, L.A. (1997).  Children’s representations of attachment relationships in family drawings.  Child Development, 6, 1154-1164.


I. Is attachment (infants’/children’s  emotional attachment to family members) a predictor of later developmental outcomes?  This topic is a major concern of developmental psychologists who are working with children with emotional problems.  Ultimately, the goal is to more fully understand the kinds of emotional problems that may be accompanied with dysfunctions within the family in order to find treatments that will help children (and their families) in the future.

II. Ainsworth and her colleagues categorized infant-mother dyads into 3 attachment patterns:  secure, anxious/avoidant, and anxious/resistant.  A fourth category (i.e., disorganized) has since been added but was not the general focus of the present study.  The technique typically used in measuring attachment behaviors is the Strange Situation which is commonly used with 12 and 14 month olds.  More recently, the focus has turned to children’s drawings as an important/useful indicator of attachment in children around 8 years of age (3rd graders).  Other measures used are play situations and verbal interviews.  Play situations are more suitable for younger age groups and verbal interviews are more useful for older children whose verbal abilities are better developed.  Currently, few studies have been able to distinguish anxious/avoidant and anxious/resistant groups.  Are these groups distinct or do they simply cluster at opposite poles of a continuum (meaning that they are possibly variants of a single category)?

III. A total of 171 8-year-olds who had previously participated in a longitudenal study on attachment were tested.  Family drawings were requested, one from each child.  The experimenter noted children’s identification of all persons in the drawings as well as their relationships to the children (e.g., mother, father, sister).  Five measures were used: 1) Experimenters classified the drawings into one of the 3 attachment categories: secure, avoidant, resistant; 2) A checklist of each component in the drawings was made according to Kaplan & Main (1986) with some changes in scoring criteria.  For example, some measures were difficult to define such as “false smiles”; 3) An integrative evaluation with attention to the context and patterning of drawing signs was made; 4) Previously obtained attachment scores using Ainsworth’s strange situation collected when the participants were 12 and 18 months of age; and 5) Various measure that are not related (at least in theory) to attachment behaviors but may influence drawings: Teacher report form/Child behavior checklist; Maternal life stress; and a subset of the WISC-R (an IQ test).  Several analyses were conducted in order to find which aspects of the drawings were consistent predictors of attachment.  Particular attention was paid to the patterns that would distinguish avoidant and resistant attachment behaviors.

IV. The results show a high correspondence between the attachment scores at 12 and 18 months using the Strange Situation measure and the attachment classifications of the 8-9 year-olds using the Draw a Family assessment scales.  Analyses distinguished avoidant and resistant categories on the following measures:

1. Arms drawn in the downward position was found more often in the drawings of children with avoidant history.

2. Four associations showed up predominantly in the drawings of children with resistant history: 1) Figures separated by barriers; 2) Unusually small figures; 3) Exaggeration of soft body parts; and 4) Crowded figures.

3. Relationships (in general) were consistent when IQ was factored out of the analyses.  That is, IQ by itself was not a significant predictor of the attachment patterns noted.

4. Emotional distance and tension/anger were related to anxious/avoidant but not anxious/resistant histories.

5. Vulnerability was associated with anxious/resistant history.

6. Global pathology was related to insecure, avoidant, and resistant histories.

V. Attachment behaviors in early development are enduring until at least 3rd grade when the child is raised in the same context (family).  Draw a family, particularly when considered in terms of aggregate signs (rather than individual signs), is a useful tool in assessing the emotional relationships between children and their families.

VI. The authors noted that several drawings were particularly striking and bizarre in appearance.  Was there a particular pattern to those drawings?  What might these unusual drawings indicate in terms of future relationships between these children and their families?  How useful are these measures in predicting other social/emotionaspects of development?