What is MCAST?

The MCAST is a structured doll play methodology, developed by Professor Jonathan Green and Colleagues, which provides an approach to understanding in young school age children (4.0-8.5 years inclusive) their internal (mental) representations of their attachment relationship with a specific caregiver. Clinical development of the MCAST began in 1992, and it has since been validated (Green et al 2000, Goldwyn et al 2000) and used in numerous international studies in low- and high-risk groups (e.g. Green et al 2007, Futh et al 2008, Minnis et al 2009, Barone et al 2009, Green 2009, Wan et al 2010, Del Guidice 2010).

The MCAST is special in its focus on attachment themes in doll play that is dyadic. Administration of the task introduces attachment-specific story stems along with affective arousal to mobilise attachment representations in a way that children find accessible and engaging. The coding system is based on attachment theory and allows the categorisation of classic attachment patterns, as well as disorganisation and coherence scores. Because the MCAST was particularly developed for application in clinical groups, MCAST has a developed coding scheme for capturing attachment disorganisation.

The instrument is suitable for research and clinical application and is currently being used in a large number of international research projects and in clinical settings by psychologists, psychiatrists, social workers and others. We run annual 3-day training workshops in MCAST administration and coding, and occasional 2-day advanced/refresher workshops. Please see training for further information.

Theoretical background

Attachment theory argues that a specific 'behavioural system' has grown up through evolution in relation to the infant experiences of danger, isolation and other threat. This behavioural system has survival value because it results in the infant seeking proximity and safety with a familiar and safe adult.

Elements of the system are activated very early on in development and different developmental skills are recruited to it in later months and years. Whereas we can understand the child's attachment pattern by observing their behaviour around caregivers in infancy, such behaviour is not overtly observable and reliably measured in this way by school age. Doll play or story stem tasks were developed as a way of accessing the mental representations of young children with respect to behavioural strategy, and specifically in the context of attachment, their 'internal working model'.

Structure and administration

The MCAST is a validated method of evaluating the child's internal representation of their attachment with respect to a specific caregiver. As such, doll play in this task involves dolls to represent the caregiver and child (rather than, for example, the whole family). When compared with some other established 'story stem' measures, the MCAST is structured with a standardised script although semi-structured aspects do allow for some flexibility and adaptation for that particular child and for a more natural and engaging storytelling process.

The measure involves four story vignettes (plus a 'control' vignette) that set up scenarios designed to induce arousal related to the attachment relationship, thus invoking the child's 'internal working model' (i.e. mental representations with respect to behavioural strategy). Administration of the MCAST involves setting up scenarios of predefined everyday situations, mostly within a dolls house context, presenting the story (engagement, mood induction), handover followed by vignette completion by the child, structured prompts and probes.

Evaluating attachment representations

The MCAST is used internationally in both research and clinical contexts. The instrument evaluates both narrative and behavioural aspects of the child's vignette completion across the task as rated usually from video recording. The coding system adapts concepts from 'gold standard' measures of attachment in infancy (Strange Situation Procedure) and adulthood (Adult Attachment Interview; AAI) to allow the categorisation of classic attachment patterns (i.e. A,B,C,D with sub-classification). Outcome codes on attachment disorganisation and coherence are also given. As the MCAST was developed with a particular focus on clinical groups (who may show high levels of disorganised attachments), its coding scheme was developed to capture attachment disorganisation in some detail.

The MCAST demonstrates high inter-rater reliability, test-retest reliability, and content validity (Green et al 2000, Goldwyn et al 2000). It has been validated against children's concurrent ratings on the Separation Anxiety Test. Disorganised attachment measured by the MCAST also shows a predicted association with the caregiver's own attachment representations (i.e. MCAST ratings of disorganised attachment were related to Unresolved status on concurrent maternal AAI).

The instrument has been used in a range of cultural, non-clinical and clinical groups, with many studies ongoing. An advantage of the MCAST in its focus on attachment is that it generates a number of ratings that has been used in published studies in various ways which may add further insight into children's attachment representations. Please see bibliography for details of our validation studies and other published work that has used the MCAST.