Dynamic relationships of therapist alliance and group cohesion in transdiagnostic group CBT for anxiety disorders

Peter J Norton, Nikolaos Kazantzis

Research output: Contribution to journalArticleResearchpeer-review

30 Citations (Scopus)


Objective: Little is known about the temporal variability of the alliance-symptom change and cohesion-symptom change relationships over the course of group therapy. These questions were examined in a sample of 373 clients receiving a transdiagnostic cognitive behavior therapy (tCBT), which culled the principle research-supported mechanisms of change for anxiety disorders. Method: The authors examined relationships between the client versions of the Working Alliance Inventory and Group Cohesion Scale in predicting subsequent symptom change, as assessed by the state scale of the State?Trait Anxiety Inventory. Results: Alliance and cohesion were significant predictors of next session anxiety scores. The alliance was consistently associated with anxiety symptoms (rs = -.152 to -.198, ps <.05), but cohesion only showed significant relationships with anxiety symptoms at Sessions 8 and 10 (Session 8, r = -.233, p = .020, and 10, r = -.236, p = .027). Alliance-anxiety relations remained constant, whereas cohesion-anxiety relations substantially increased from earlier to later sessions. Discussion: Differences that were obtained in the relation of alliance and cohesion with anxiety symptoms suggests that these processes have different roles within group tCBT. If replicated, the present findings would suggest that the dynamic relationships between alliance and cohesion and symptoms within group CBT for anxiety disorders have been an important omission in process-outcome studies
Original languageEnglish
Pages (from-to)146-155
Number of pages10
JournalJournal of Consulting and Clinical Psychology
Issue number2
Publication statusPublished - 1 Feb 2016


  • cognitive behavior therapy
  • group cohesion
  • psychotherapy processes
  • therapist alliance
  • transdiagnostic

Cite this