When is therapist metacommunication followed by more client collaboration? The moderation effects of timing and contexts.

(a) There was a significant 3-way interaction of metacommunication, time, and between-session dominance-submission, such that metacommunication significantly predicted subsequent client collaboration in later speaking turns within a session, and when the therapist generally showed low dominance in that session. (b) There was a significant 3-way interaction of metacommunication, time, and between-session approach-avoidance. Specifically, early in sessions therapist approach buffered the negative effect of metacommunication on collaboration, and later in sessions therapist neutrality (less approaching) facilitated a positive effect of metacommunication on client collaboration. (c) There was a significant 2-way interaction of between-speaking-turn metacommunication and within-session time in predicting subsequent client collaboration, such that early in sessions metacommunication did not predict collaboration, whereas later in sessions more metacommunication was associated with higher collaboration. These results were discussed in relation to the therapeutic metacommunication model proposed by Kiesler (1988). (PsycINFO Database Record