Exploring disorganised attachment: unravelling developmental pathways and outcomes using data mining

Disorganised attachment is an important assessment of infant mental health. Disorganisation is thought to result from an infant having in some way experienced trauma in the context of the relationship with their caregiver such as observing severe domestic violence. Infants classified as disorganised have an elevated risk of psychological problems, most notably conduct disorders. The possibility of a finer-grained measure emerged from archival research on the original Berkeley dataset from which disorganised attachment was first identified.

We have been offered unprecedented access to a longitudinal dataset to explore how the finer-grained measure fares against the standard construct in predicting a range of negative outcomes, and whether particular forms of disorganisation have specific antecedents. Given that the goal is exploratory and the array of relevant measures is extensive, data mining will be used rather than hypothesis testing.

The research offers the prospect of a significant transformation of research in this area. Three focus groups with clinicians will be conducted to facilitate clinical input and translation.