The Therapeutic Alliance

This article focuses on therapeutic (working) alliance in counseling, a critical component related to successful outcome in counseling.  I examine a small number of studies, providing background related to the general effects of counseling and the Working Alliance Inventory (WAI) and discussing the impact of the alliance in counseling.  I conclude with a discussion of the implications of the information presented, specifically focusing on the contextual model and the importance of general effects in counseling, the importance of developing a collaborative relationship with clients early on, understanding how clients early formative and current relationships affect their ability to form a working alliance, achieving a balance between process related techniques and alliance strengthening skills, and, finally, evaluating client attachment style and how it may affect the working alliance.  

                        Keywords: Alliance, therapeutic, counseling, client, attachment, contextual.

           The therapeutic alliance, or “working” alliance as Bordin (1979) defined it, is widely accepted as a crucial component of successful outcome in counseling and has been studied extensively.  I examine a relatively small number of studies here and with a somewhat limited focus due to time and other constraints related to this assignment.  I will start by providing some background related to the general effects of counseling as well as a widely used instrument in measuring the alliance, the Working Alliance Inventory (WAI).  I will then discuss the impact of the alliance in counseling, focusing on a few key studies, finishing with a discussion of the implications of the information presented.

General Effects and the Therapeutic Alliance

            In Wampold’s book, The Great Psychotherapy Debate, he identified therapist effects as a “critical factor in the success of therapy” (2001, p. 202).  More specifically, in their article on therapist and patient variability in the therapeutic alliance, Baldwin, Imel, and Wampold (2007) pointed out that it is the therapist’s contribution to the alliance that is foremost in determining a successful outcome for the client.  Wampold (2001) supported the contextual model (versus the medical model), contending that is the general effects (common factors – therapeutic alliance, therapist competence, a belief, by the therapist and client, in the effectiveness of the therapy, etc.) of psychotherapy that are key to a successful outcome, not the specific effects (techniques) of any particular therapeutic approach.   That being said, it is widely contended that it is the therapeutic alliance that will account for a great deal of the variability, both positive and negative, in the client’s outcome.

The Working Alliance Inventory (WAI)

            The Working Alliance Inventory (WAI), developed by Horvath and Greenberg in 1968, is a widely used instrument for measuring the therapeutic alliance and was the instrument of choice in the majority of the studies discussed here. The WAI is a 36-item self report survey consisting of three subscales that mirror Bordin’s three components of the working alliance, goals, tasks, and bonds, and uses a 7-point Likert-type scale.  Parallel forms are available for both clients and counselors (Satterfield and Lyddon, 1995).

The Therapeutic (Working) Alliance

            One of the most important tasks that we as counselors have is to form a positive, healthy, nurturing working alliance with our clients.  As we discussed briefly above, Bordin (1979) defined the working alliance as a collaborative process in which client and counselor  (a) mutually endorse goals or counseling outcomes; (b) join in tasks related to the attainment of successful outcomes; and (c) establish positive personal attachments, or bonds, which are characterized by trust, acceptance, and confidence.  A good working alliance is based on two important factors.  The first factor is the relationship that the counselor develops and fosters from the very beginning of counseling.  Kokotovic and Tracey (1990) found that clients who were viewed by their counselors as having poor social relationships in general had greater difficulty in forming working relationships (alliances) with their counselors.  A second factor is the relationship the client has or had with his or her parents, because that relationship will give us insight into how the client relates to their social network and, ultimately, most likely predict how they will relate to their counselor.  In support of that assertion, Mallinckrodt (1991) also reported evidence of a correlation between clients’ recollections of the quality of their childhood bonds with their parents and the strength of the working alliance.

            In Kivlighan’s (1990) study, the relationship between counselor technical activity (use of intentions – set limits, educate, assess, explore, change, restructure, and support) and working alliance (as rated by the client) was analyzed during the course of four counseling sessions.  Two groups of undergraduate students were asked to participate in a study in which sessions were analyzed to see if the use of intentions by the counselor affected the quality of the therapeutic alliance.  The study found that during the four sessions, three of the intentions mentioned above, assess, explore, and support, were negatively correlated with the working alliance as measured by the WAI.  The authors were somewhat surprised by the negative correlation of the support intention (offering support or encouragement) with the alliance, but concluded that this may have occurred because it put the client in a more passive role.  They also proposed the following questions: 1) “Can counselors be trained to decrease their use of the assessment, explore, and support intentions?” and 2) “Would this training affect client-rated alliance?” We are not sure that such training would be either indicated or useful based on a study with such obvious limitations, considering the importance of assessment, exploration, and support in counseling.  At best, the study points to the need to balance such strategies, with the counselor paying particular attention to the use of intentions that may put the client in a more passive role versus those that will enhance the working alliance.

            According to Satterfield (1995), a client whose attachment style is characterized by a lack of trust in the availability and dependability of others (low level of “depend”) may be more likely to evaluate the counseling relationship in negative terms, particularly during the early phase of counseling. The authors recruited ninety-six first-time clients seeking counseling through the university to participate in a study in which they completed the Adult Attachment Scale (AAS) prior to counseling and the WAI (client version) after the third session.  Sixty participants completed the study and the authors concluded that client attachment, particularly the “depend” measure, is in-fact negatively correlated with the working alliance and may lead to unfavorable counseling outcomes. They also recommended further research on the impact of counselor knowledge of client attachment dimensions and the affect they may have on the working alliance.

Conclusions

            What should counselors focus on if we are to accept the hypothesis of the contextual model and the importance of general effects in counseling as proposed by Wampold?  Based on the research, one of the most important things we can do is learn how to better foster a therapeutic, empathic, nurturing alliance with our clients.  As Bordin (1979) recommended, we should focus on developing a collaborative relationship with our clients early on, setting mutually agreeable goals, working together toward successful outcomes, and establishing positive bonds.  Kokotovic and Tracey (1990) and Mallinckrodt (1991), taught us that we should understand how our clients early formative relationships and their ability to form and maintain current relationships affect their ability to form a strong working alliance.  Kivlighan  (1990) emphasized the importance of focusing on a balance between process related techniques (intentions) and alliance strengthening skills, such as those endorsed by Carl Rogers (genuineness, empathy, and warmth).  Finally, Satterfield (1995) points us to the need to assess our client’s attachment style, looking particularly for those clients who may be characterized by a lack of trust in the availability and dependability of others, and how their attachment style may contribute to the alliance.   In conclusion, because alliance effects are so intertwined with outcome, whether positive or negative, we owe it to our clients to “do no harm” and do everything in our power to foster a positive working alliance.

References

Baldwin, S.A., Imel, Z.E., & Wampold, B.E. (2007).  Untangling the alliance-outcome correlation: Exploring the importance of therapist and patient variability in the alliance. Journal of Consulting and Clinical Psychology, 75 (6), 842-852.

Bordin, E. S. (1979).  The generalization of psychoanalytic concept of the working alliance. Psychotherapy, 16, 252-260.

Kivlighan, D. M. (1990). Relation between counselors' use of intentions and clients' perception of working alliance. Journal of Counseling Psychology, 37(1), 27-32.

Kokotovic, A. M. and T. J. Tracey (1990). Working alliance in the early phase of counselor. Journal of Counseling Psychology, 37, 16-21.

Mallinckrodt, B. (1991). Clients' representations of childhood emotional bonds with parents, social support, and formation of the work alliance. Journal of Counseling Psychology, 38, 401-409.

Satterfield, W. A. and W. J. Lyddon (1995). "Client attachment and perceptions of the working alliance with counselor trainees. Journal of Counseling Psychology, 42 (2), 187-189.

Wampold, B. (2001).  The Great Psychotherapy Debate. Mahwah, NJ: Lawrence Erbaum Associates.

© CombatCounselor 2011 – All Rights Reserved