Examining the Alliance-Outcome Relationship: Reverse Causation, Third Variables, and Treatment Phase Artifacts

Psychotherapy research reveals consistent associations between therapeutic alliance and treatment outcomes in the youth and adult literatures. Despite these consistent findings, prospective associations are not sufficient to support the claim that the alliance is a change mechanism in psychotherapy. The current study examined the direction of effect of the alliance- outcome relationship, the contribution of early symptom change in treatment to the development of therapeutic alliance, and the potential for pretreatment interpersonal functioning characteristics to be third variables that account for the association between alliance and outcome. Participants were adolescents with depression and a history of interpersonal trauma that presented to a community mental health center for treatment. Findings demonstrated that a more positive therapeutic alliance predicted greater subsequent symptom improvement, even after removing symptom change occurring before the measurement of alliance. Results also suggested that early change only slightly contributed to alliance development. Finally, though pretreatment interpersonal functioning was related to the first session alliance, these pretreatment client characteristics were not related to later alliance or symptom change. Overall, results provided some support for therapeutic alliance as a mechanism of change in psychotherapy. Methodological and clinical issues are discussed.