A move-by-move paradigm for the computational characterization of attachment style and personality disorder

A current direction of personality disorder research strives to identify key behavioural, cognitive, and ultimately computational facets of patient functioning via the use of engaging social paradigms. Thus far, few such paradigms have been put forward. Here, we introduce a novel task in which subjects interact with previously unknown virtual partners in a turn-taking paradigm akin to a dance, and subsequently report on their experience with each. The partners' ``personalities'' differ in the nature and extent of their reactions to the inter-personal distance kept by participants. We show that the plurality of measures produced may help further characterize attachment style and borderline personality disorder (BPD) symptoms. Higher scores on our measures of attachment anxiety, avoidance, and BPD symptoms were all linked to a general negative appraisal of all the interpersonal experiences. Further, the personalities of the partners encountered mattered: for instance, negative appraisal of a partner who displayed the most biasedly negative range of moods was tied with attachment anxiety and BPD symptoms. Finally, our analyses of proxemics data underscored slower movement initiation from anxiously attached individuals throughout all virtual interactions, whereas BPD symptoms were tied with a tendency to react with further distancing from a partner which is too close.