Attachment, positive affect and competence in the peer group: Two studies of construct validation
Examined the construct validity of attachment, using longitudinal observations of infants who had participated in Minnessota Birth Announcement study.
Study 1
Subjects were 45 white, middle-class infants (23 males, 27 females) who had been videotaped whilst taking part in the Strange Situation Procedure (SSP) at age 18 months and were later videotaped at 24 months in a free-play situation. In this study, attachment style had previously been determined so only preseparation behaviour was examined and coded from the SSP tapes. Within the preseparation period, only spontaneous smiling at mother significantly distinguished attachment style. Nearly all Secure (18/19) infants engaged in this behaviour compared to fewer than half of the Avoidant and Resistant infants. Other measures of affective sharing such as showing toys, giving toys and looking at mother were all nonsignificant. Overall, affective sharing, whether assessed from the preseparation period of the SSP (at 18 months) or the free-play tapes (at 24 months), was found to correlate well with attachment style as measured at 18 months.
Study 2
The second study was intended to assess continuity in individual adaptation from 15 months to 3½ years. Subjects were 32 white, middle-class infants who had been videotaped in a novel situation procedure at age 15 months. These videos were used to assess attachment at this age, with 20 subjects assessed as secure and 12 as anxious. Comparisons made with the Bayley Mental Development Scale (administered at 14 months) and the Stanford-Binet Test (administered at 36 months) yielded no significant differences between attachment groups.
At age 3½, subjects were observed in a classroom situation for 5 weeks. Behaviour was scored on 2 major scales of peer competance (initiative, skill and engagement in interaction with peers) and ego strength (ego resiliency and control). Significant between-groups differences were found for 11 of 12 items on the peer-competance and 5 of the 12 items on the ego strength scale. In all cases, differences favoured the secure attachment group.