Therapeutic alliance as a predictor of psychotherapy process and outcome: The role of expert versus novice raters
This study examined the relationship between therapeutic alliance and therapeutic process and treatment outcome with adolescent substance abusers that received either CBT or MDFT. In particular, it sought to determine whether expert ratings of alliance would predict process and outcome better than novice ratings of alliance. There has been no published work to date that has examined the differences between expert and novice raters of alliance. Accordingly, the first aim of the study was to evaluate if there were any differences in therapeutic alliance ratings between expert and novice raters. Secondly, it was hypothesized that expert ratings of alliance would predict process variables (adherence and competence) better than novice ratings of alliance. Third, it was hypothesized that expert ratings of alliance would predict outcome variables (adolescent functioning, retention, and substance abuse) better than novice ratings of alliance. Participants for this study were 20 observer raters. These raters were split into three groups (CBT expert, MDFT expert, and novice). Seven raters represented the experts in the CBT condition, while eight raters represented experts in the MDFT condition. These experts were chosen because they were seasoned clinicians in their respective fields. The novice raters were five psychology graduate students and they rated both conditions (CBT and MDFT). The results indicated that, as expected, novice ratings of alliance were significantly different from expert ratings of alliance in both the CBT and MDFT conditions. However, no difference was found the in combined treatment condition. Furthermore, expert ratings of alliance predicted adherence and competence better than novice ratings of alliance in the CBT, MDFT, and combined treatment conditions. Beyond that, expert ratings of alliance also predicted externalizing behaviors and retention, better than novice ratings of alliance in the combined treatment condition. These findings were of particular significance considering that a previous study by Hogue and colleagues (2006) failed to report these findings. Hogue and colleagues had used novices to rate alliance and this study may point to the importance of having alliance rated not only by novices, but by experts as well.
Berzins, Peter, "Therapeutic alliance as a predictor of psychotherapy process and outcome: The role of expert versus novice raters" (2008). ETD Collection for Fordham University. AAI3337627.