Comparison group
AAI transcripts were available from two earlier studies, a group of parents whose
children were being assessed in a community study of emotional development, where the
parents were involved only in being given the AAI, and a group of adults with current
mood disorder identified as part of a long-term follow-up of children in psychiatric
services. From a sample of 39 control transcripts, 20 were selected to match the
participants as closely as possible on the basis of attachment classification and current
mood disorder. The intention of having a control group was to be able to compare
response styles and patterns of scores across the continuous scales on the AAI, so the
controls were matched for attachment classification to ensure that any differences in these
scales were due to factors related to autism rather than the expected variation across
different attachment categories. As AAI classification is affected by mood (Pianta,
Egeland & Adam, 1996; Roisman, Fortuna & Holland, 2006), controls were also matched
for mood disorder. Further details of the control group are given in the Results.
Measures
• Adult Attachment Interview
The Adult Attachment Interview (AAI; George et al., 1996) is a semi-structured interview
that asks participants to recall their childhood relationships with parents and other
attachment figures. Each AAI was audiotaped and subsequently transcribed by the
interviewer. The AAI transcript is scored according to criteria developed by Main and
Goldwyn (1998), on probable experience and state of mind scales. A transcript can usually
be assigned to one of three organised attachment categories - secure/autonomous,
insecure/dismissing or insecure/preoccupied - by comparing that transcript’s pattern of
scores with the prototypical pattern that characterises each category. If a transcript proves