Ambivalence and the working alliance in variants of cognitive‐behavioural therapy for generalised anxiety disorder

Abstract Client characteristics are widely understood to influence alliance development. However, few studies have examined the role of client ambivalence about therapeutic change. Building on past research demonstrating associations between ambivalence and therapy relationship variables, such as resistance, this study examined whether greater ambivalence was associated with poorer alliance quality. Further, it examined whether motivational interviewing (MI), which involves strategies for managing ambivalence, moderated this relationship. Using data from a randomised controlled trial of cognitive‐behavioural therapy (CBT) for 71 individuals who completed treatment for generalised anxiety disorder, this study tested whether ambivalence, operationalised as observed motivational language against change (counter‐change talk; CCT), in session 1 was related to client‐rated alliance quality over time, and whether this relationship varied between two treatments: MI integrated with CBT (MI‐CBT) or CBT alone. CCT predicted lower client alliance ratings at the early, middle and late stages of therapy. At the late stage, treatment group was a significant moderator, such that CCT was associated with poorer alliances for CBT alone, but not for MI‐CBT. Results suggest that early ambivalence can predict early and middle phase alliance problems in both treatments and that, without explicit strategies for managing ambivalence, early CCT is strongly associated with poorer alliances during the late stage of CBT treatment. This research highlights the importance for clinician responsivity to early motivational markers.