higher scores being given to indications of “the subject’s ability to understand
psychological states, including conscious and unconscious motivations, underlying their
own reactions and those of others, especially their parents” (Fonagy et al., 1991, p.210).
Again, high reflective function is strongly related to overall attachment security (Fonagy
et al., 1991), but unlike coherence scores it is not used to assign interviewees to the
secure/autonomous category.
• Autism Diagnostic Observation Schedule
The Autism Diagnostic Observation Schedule (ADOS; Lord, Rutter, DiLavore & Risi,
1999), module 4, is a semi-structured assessment for use with adults with autism who
have fluent language. It was used in this study to confirm a diagnosis of autism or autism
spectrum disorder, and to provide a continuous measure of symptom severity in the
domains of communication and reciprocal social interaction. Although the latter is not its
intended use, it has been used in this way in other studies (see for example Aldred, Green
& Adams, 2004; Joseph, Tager-Flusberg & Lord, 2002).
• Wechsler Abbreviated Scale of Intelligence
The Wechsler Abbreviated Scale of Intelligence (WASI; Wechsler, 1999) was used as a
brief measure of IQ, to exclude any participants with an IQ below 70 and to explore the
relationship between attachment and cognitive ability. It has been demonstrated that the
WASI can provide a valid estimate of full-scale IQ for adults with autism spectrum
disorders (Minshew, Turner & Goldstein, 2005).
• Theory of mind measures
Two well-established research measures of theory of mind were used. The “reading the
mind in the eyes” task (Baron-Cohen, Wheelwright, Hill, Raste & Plumb, 2001) was
developed with high-functioning autistic adults; participants are asked to match thirty-six