An empirical evaluation of Luborsky's biphasic model of the therapeutic alliance within the context of a randomized controlled trial

The primary goal of this study was to assess the validity of a prominent model of the therapeutic alliance, Luborsky’s Helping Alliance, in which components of the alliance develop differentially in two phases of therapy. This study used secondary data from a randomized, controlled trial (Svartberg et al., 2004) in which patients rated the alliance twice during therapy using a measure based on Luborsky’s model (Helping Alliance Questionnaire; HAq-I). The secondary purpose of this study was to investigate the association between alliance and short and long-term outcome. HAq-I data were factor analyzed to assess the measure’s fit with the biphasic model and regression analyses were performed to investigate the association between alliance and short and long-term outcome. The results of the analyses did not support the biphasic model as measured by the HAq-I. Regression analyses revealed a number of significant associations between HAq-I scores and long-term outcome.