ABSTRACT
This chapter demonstrates cluster analysis – specifically the k-means clustering algorithm – as a technique to discover natural groupings in therapist–client language, leading to a concrete measure of linguistic synchrony for therapist–client dyads. As an important component of interpersonal synchrony, which has been linked to positive treatment outcomes, linguistic synchrony is the extent to which therapists’ and clients’ linguistic choices align across the treatment span. The logic, characteristics, and major cluster analysis algorithms are first introduced with intuitive examples. The case study involved the following steps: (1) computing sessional LIWC variable scores separately for therapist and client language, (2) k-means clustering with model validation to determine (a)synchrony based on cluster memberships, and (3) qualitative analysis to illustrate how synchrony is co-constructed in context. These were performed on three sample dyads from psychoanalysis, cognitive-behavioral, and humanistic therapy. The resulting synchrony measures (psychoanalysis 33.3%, cognitive-behavioral 0%, humanistic therapy 25%) and qualitative analyses appear to reflect the theoretical traits of these therapy types. This approach offers a systematic and replicable tool for research, training, and self-evaluation in psychotherapy and other similar interactional contexts.
