OBJECTIVE: Although patients' expectation for improvement correlates with their treatment outcome, there remains limited information regarding the mechanisms through which outcome expectation influences outcome. Although several studies have revealed alliance as a mediator of the expectancy-outcome relation, most have focused on individual psychotherapy only. More research is needed examining mediators, including alliance quality, of the outcome expectation-outcome relation in group therapy.
METHOD: This study focused on such associative chains among 91 depressed outpatients who completed 10 weeks of group cognitive-behavioral therapy. We conducted simple and multiple mediation analyses, accounting for the nested data structure.
RESULTS: As predicted, we found: (i) The relations between baseline outcome expectation and both posttreatment anxiety and depression were mediated by alliance quality; (ii) the early therapy outcome expectation-posttreatment anxiety relation was mediated by mid-treatment alliance; (iii) the relation between early alliance and posttreatment interpersonal problems was mediated by during-therapy outcome expectation; and (iv) the relation between baseline outcome expectation and posttreatment interpersonal problems was mediated by two variables acting in turn, early alliance and during-therapy outcome expectation. All other tested models were not significant.
CONCLUSIONS: Results suggest that bidirectional relations between outcome expectation and alliance, with both directions influencing outcome. Clinical and empirical implications are discussed.