Attachment in infancy and at age six, and children's self-concept : a follow-up of a German longitudinal study

In the tradition of prospective and comparative attachment research, this study investigated stability of child-mother attachment from infancy to six years within a cross-cultural context. Additionally, the six-year old children-'s self-concept was assessed. Consistency across self-concept measures was predicted as was a relationship between attachment and self-concept, with securely attached children exhibiting a more positive self-concept than their insecure counterparts.

Subjects were 44 children participating in a representative longitudinal study in West Germany. Attachment stability was computed for 37 children seen with their mother both in infancy and at age six and currently living in two-parent homes.

Six-year attachment was assessed through analysis of child behavior toward the parent following a 90-minute separation. The child’s self-concept was investigated through clinical interviews, Harter's Scale for Perceived Competence (Harter & Pike, 1984), and a small Q-Sort of Specific Domains. Emotional disturbance was indexed through Koppitz.’ Emotional Indicators on the Draw-A-Person Test (Koppitz, 1968).

Results suggest high stability of child-mother attachment. Eighty-nine percent of the-children showed comparable attachment patterns at age six as they had in infancy. Stability was also found for avoidant behavior and security of attachment. Results were identical to those of the Berkeley Social Development Project (Main et al., 1985) and seen as a convergent cross-validation of the six-year attachment method.

Children did not respond consistently across the self-concept measures. Factor analysis suggested two orthogonal factors, one representing the indirect clinical interviews and the other the structured standardized measures. The open interview loaded on both factors. No correlation was found between emotional indicators and self-concept or attachment.

No predictive relationship was found between infancy attachment and self-concept at six years, while the concurrent relationship between attachment and self-concept was moderate. Absence of a stronger relationship between attachment and self-concept is discussed in view of the methodological problems characteristic of self-concept measures. The relevance of these findings for clinical investigation is reviewed. A final section suggests further analysis of these data and a reanalysis of the infancy. attachment data for a more complete study of the two constructs under question.

Note: Abstract extracted from PDF file via OCR.