Hierarchical continuous-time sequential analysis: a strategy for clinical research

Abstract

This article presents a strategy for analyzing interdyadic differences in sequential data on social interactions. The social interactive data could be, for example, a nonverbal behavior such as eye gazes within dyads, with measurement of both the sequence of behaviors and their durations. This article shows (a) how one can statistically describe an interactional structure within each dyad governing the stream of that dyad's social interactive behavior and (b) how scores describing dyadic structures can be related to covariate information about the dyads. The covariates could include, for example, ratings of therapist skill or client psychopathology. Methods for relating measures of within-dyad structures in interactive behavior to between-dyad covariates could be a powerful tool for research on psychotherapy process or interpersonal relationships.