Changes in the therapeutic alliance and symptomatology : use of a time series design to assess which comes first

This aim of this study was to assess patients' expected therapeutic alliance and symptomatology before beginning psychotherapy, and comparing these baseline assessments with measures of alliance and symptomatology obtained on a continuous basis across a minimum of 10 weeks of therapy. Assessments of alliance and symptomatology, were made every half-week for a range of seven to fourteen times during the baseline period and a range of twenty-two to forty-four times during the treatment phase. Seven female patients and one male patient participated in the study. Three questions were the focus of this study: (I) Does the introduction of therapy have an impact on the initial formation of alliance? (2) Does the alliance change over time or does it stabilize? (3) Is there a time-contingent relationship or a simultaneous relationship between alliance and outcome, and if it is time-contingent, does a1liance predict outcome or does outcome predict alJiance? The results indicated that for a majority of subjects the introduction of therapy reduced totals on the Confident Collaboration scale, the Confident Collaboration 2 scale, and the Goal and Task Agreement scale. There was no significant impact on the Bond scale, the Patient Dedication scale, and the Idealized Relationship scale. The results also supported the hypothesis that therapeutic alliance remained stable over time. Small effect sizes (ranging from absolute values of .109 to .157) indicated one instance of alliance predicting symptomatology, three instances of a simultaneous relationship between symptomatology and alliance, and five instances of symptomatology predicting alliance.