Date of Award


Degree Name

Doctor of Philosophy



First Advisor

Dr. Scott T. Gaynor

Second Advisor

Dr. C. Richard Spates

Third Advisor

Dr. Cynthia Pietras

Fourth Advisor

Dr. Helen D. Pratt


Therapeutic alliance has long been recognized as an important contributor to successful psychotherapy, and research has demonstrated that it has a moderate and reliable impact on treatment outcome across studies employing a variety of treatment strategies. The current study seeks to build on the alliance literature by examining the degree to which therapeutic alliance is predictive of treatment retention. Two trained raters assessed the alliance between therapists and their patients by watching and rating therapeutic alliance in videotaped psychotherapy sessions. They watched tapes of therapists using two different types of treatment: cognitive modification and supportive therapy. Alliance was assessed via the Working Alliance Inventory-Observer Form (WAI-O) and the Session Evaluation Questionnaire (SEQ). Results were analyzed using a series of 2 x 2 ANOVAs based on treatment status (completer or dropout) and therapy type (cognitive modification or supportive therapy). While no statistically significant differences were detected on the WAI-O, data from a subscale of the SEQ indicated that treatment completers were significantly more aroused than dropouts. Data also indicated that SEQ total scores and multiple SEQ subscale scores were rated as significantly higher in supportive therapy sessions than in cognitive modification sessions, and that participants and therapists tended to appear to like each other more in supportive therapy than in cognitive modification treatment. The present paper discusses potential implications for these findings, and provides suggestive evidence that the level a psychotherapy participant's arousal may be predictive of treatment completion and premature dropout.

Access Setting

Dissertation-Open Access