A correlation network analysis of dissociative experiences

Abstract

The interrelationships between the symptom domains of dissociation, such as the loss of continuity in subjective experience, the inability to access personal information, and the distortions about the perception of self and the environment, need to be better understood. In the current study, 2274 adults from Italy completed the Dissociative Experiences Scale-II (DES-II), and their responses were examined within a correlation network analysis framework. Fifteen dissociative experiences showed the strongest associations with the other dissociative experiences included in the measure, and they were selected for further analysis. A partial correlation network was calculated to reveal the associations between such experiences, and a community detection analysis was used to explore whether they formed distinct clusters in the network. Subsequently, a Bayesian network was estimated to examine the direction of the associations among the dissociative experiences, and a directed acyclic graph (DAG) was generated to estimate a potentially causal model of their relationships. The community detection analysis revealed three clusters of experiences that were conceptualized in terms of trance, experiential disconnectedness, and segregated behaviors. Dissociative amnesia was a common denominator of all the three clusters. The analysis of the DAG further suggested that dissociation can be conceptualized as a network in which dissociative experiences are layered into groups of symptoms that interact among them. Cognizance of the configuration and interactions among the dissociative domains and their related symptoms may be critical for better understanding the internal logic behind the dissociative processes and for addressing them effectively in clinical practice.