Failures in Group Psychotherapy and Group Research: Looking Forward

Sally Barlow, PhD

I had the pleasure of being part of a symposium, "When Group Therapy Fails: Experienced Therapists Reflect" at the American Psychological Association's Annual Conference in San Francisco last August, along with William Holahan, PhD, CGP, and Allan Elfant, PhD, CGP, FAGPA. We set about the difficult task of talking about our failures as group therapists, as researchers, or as both. In my case, it was as both. My failure was likely the result of trying to wear these two hats at once.
The critical incident that brought this to my attention was when a client brought a gun to group, and I froze. I was busy thinking about the experimental condition and my need to speak in the language of the fourth quadrant of the Hill Interaction Matrix (HIM; Hill, 1973), which classifies interactions in terms of both content and style. (There are 16 categories of interaction, and Quadrant IV verbalizations are highly therapeutic ones.) I was too scared, however, to articulate a therapeutic response and instead got busy getting him to agree to lock the gun in his car. He never returned to group, and eventually the group's potency diminished as members came irregularly, engaging in mostly superficial talk.

I felt caught between a rock and a hard place. On the one hand, I did not want to annoy my colleagues by failing to stay at 80 to 90 percent in Quadrant IV; on the other hand, I was so worried about my safety and the safety of the group members that all I could do was stammer. Since that time, I have often wondered what I could have said to this client that would have qualified as a Quadrant IV statement (discussed in some detail in an earlier The Group Circle article, Barlow, 1997). Perhaps there was leader language that might have raised to articulation the fears of the group, his "communication" to the group that the gun represented, and what it might require to make the group a safe container for "explosive" talk. But these ruminations possess the emptiness of most hindsight. It is unsatisfying simply because it is too late.

Our symposium attempted to bring to light the notion of revisiting failure. Allan Elfant, Bill Holahan, Martha Gilmore and I each told our stories, inviting the audience to come along with us as we joined in an endeavor to make more useful the specter of failure.

As much research attests (e.g., Bednar & Kaul, 1994; Dies, 1986) and recent meta-analyses demonstrate (Burlingame, Fuhriman, Anderson, & Hoag, in press; Tillitski, 1990), group psychotherapy is clearly effective. Occasionally, however, group therapies and group research fail. Certainly we can turn to the empirical literature for the likely culprits. For group therapy, these are selection mistakes, group composition problems, and so on (Dies, 1994). For research, these are the exponential problems inherent in designs that attempt to capture process and outcome on a much larger scale than individual therapy research (Bednar & Kaul, 1994). But as each panelist pointed out in the APA symposium, what is missing from this literature is the personal reflection of the therapist or researcher. 

While there are notable exceptions to this (e.g., Budman, Demby, Soldz, & Merry, 1996; HarPaz, 1994), we are often collectively embarrassed about these failures and unwilling to discuss them in public arenas. Perhaps we experience failure as an itch at our Achilles' heel, fearing that it foreshadows future failures and even our eventual demise. After all, Achilles did have something to fear. This might have occurred, however, because he didn't wear compensatory footwear. That our occasional failures might lead to destruction is just such a myth. In reality, our failures help to expose our vulnerabilities to ourselves so that we can change them by participating in continuing education, ongoing consultation, or collaboration on research projects with colleagues who can gently point out our countertransferences to the particular group member or research design under discussion.

As Dr. Holohan pointed out in his reflections on his least favorite group, "Members bring unpredictable dynamics and behaviors to each group session. No therapist can be aware of all potential destructive, divisive events in advance" (Holohan, 1998, p. 2). He suggested instead that we examine over time "group failure events," which would allow us to discover how we contribute to such failures. That is, while we cannot predict the agendas of individual group members, we can survey patterns and uncover a discernible and "systematic series of decisions" that have likely led up to treatment failures. This scrutiny is likely to face us painfully with "moments of vulnerability, uncertainty, and situational incompetence" (Holohan, 1998, p. 2), yet it is essential if we are to serve our patients and ourselves better. We do this microscopically, when we examine individual interventions, and macroscopically, when we look at patterns over the years. As leaders of our groups, we can realistically model acknowledgment of and recovery from mistake-making for our patients. In supervising groups and while planning research studies, we can model openness to feedback and a genuine acceptance of ourselves as human with our colleagues.

Of course, we all yield to the pressure brought on by our patients (and ourselves, our profession, and our colleagues) to be "infallible" therapists, researchers and the like. As Anthony Joyce, PhD, suggested, "This is the role that patients often want us to take so we are congruent with their idealizations, and, let's face it, it can be an alluring and gratifying role to assume. It is natural-perhaps even required-that our patients at some point are disillusioned with us as 'infallible' helpers.... Modeling acceptance of one's own fallibility can be therapeutic for those patients who suffer greatly when they do not meet the perfection they continually strive for" (1999). There is clearly a difference between demonstrating expertise and chasing perfection. The former helps patients; the latter colludes unwittingly and non-therapeutically with their struggles.

This type of scrutiny represents an important awareness of the impact, intentional or unintentional, of individually harmful interventions in the moment, or collectively harmful unconscious beliefs within us as we look at systematic clinical decisions across time. These unconscious beliefs, including the drive for perfection, represent the all-too-human potential for destructiveness. Then, and only then, can we hope to become more like Wagner's Parsifal rather than Stravinsky's Oedipus Rex (Mondadori, 1978). In the former, an opera about the folly of undergoing isolated crusades, Parsifal finally transforms from his naïve status as "pure fool" to true healer. We all know the fate of Oedipus. I invoke the medium of dramatic operatic theater because I believe our collective failure phobia weakens our profession and dilutes our ability to help patients. Opera stories are full of folly: characters who persist in their treacherous course even when they've been given ample evidence to the contrary. We, too, continue to eschew open forums about failure and elevate instead our successes. Certainly we should note our successes, but as Lorna Benjamin (1996) likes to point out, "reality is our friend," and it is in reality that we will both succeed and fail. The proportion of successes to failures will increase only if we acknowledge and deal with these failures.

Our panelists explored these possibilities, including the narcissistic injury we might experience when we look inward at this wished-for perfection and find it wanting. We cautioned the audience not to turn the symposium into supervision. That is, we were not interested in the "but did you try this..." kind of exchange. Rather, we were interested in a look inward at our own styles of reinvoking the failure myth.

The audience members talked about their own struggles, reliving poignant experiences. I had never before seen this happen at an APA paper presentation. Perhaps it was the combination of a more intimate setting-a smaller room, a packed audience, and the topic of vulnerability-that allowed us to share deeper and more honest revelations. As I listened to others reflect, I began to see my original failure in a new light. I realized I could wear two hats at once. I found a middle ground of truth where Mahler et al. (1975) says it's quite all right to live. Together, as panelists and audience, we had reached the Quadrant IV of the HIM, the place I had longed to be in that earlier group. It was better late than never.

References

Barlow, S. (August, 1998). A therapeutic and research failure in retrospect: Wearing two hats can be dangerous. Paper presentation to the Annual Meeting of the American Psychological Association, San Francisco.

Barlow, S. (Fall,1997). The failed group in retrospect. Invited research report to The Group Circle. American Group Psychotherapy Association newsletter.

Bednar, R., & Kaul, T. (1994). Experiential group research: Can the canon fire? In A.E. Bergin & S.L. Garfield (Eds.), Handbook of psychotherapy and behavior change, (4th ed.), pp. 631-667. New York: John Wiley. 

Benjamin, L. (1996). The interpersonal diagnosis and treatment of personality disorders, 2nd ed. New York: John Wiley.

Budman, S., Demby, A., Soldz, S., & Merry, J. (1996). Time-limited group psychotherapy for patients with personality disorders: Outcomes and dropouts. International Journal of Group Psychotherapy, 46: 357-377.

Burlingame, G., Fuhriman, A., Anderson, E., & Hoag, T. (in press). Group psychotherapy efficacy: A meta-analytic review. Clinical Psychology: Science and Practice. 

Dies, R. (1994). Therapist variables in group psychotherapy research. In A. Fuhriman and G. Burlingame (Eds.). The handbook of group psychotherapy: An empirical and clinical synthesis, pp. 114-154. New York: John Wiley.

HarPaz, N. (1994) Failures in group psychotherapy: The therapist variable. International Journal of Group Psychotherapy, 44: 3-19.

Hill, W. ( 1973). The Hill Interaction Matrix (HIM): The conceptual framework for undertaking groups. New York: University Associates.

Holahan, W. (August, 1998), Reflections on termination in my least favorite group. Paper presentation to the Annual Meeting of the American Psychological Association, San Francisco.

Holahan, W., & Elfant, A. (Co-chairs, August, 1998), When group therapy fails: Experienced therapists reflect. Panel presentation to the Annual Meeting of the American Psychological Association, San Francisco.

Joyce, A. (1999). Personal communication.

Mahler, M., Pine, F., & Bergman, A. (1975) The psychological birth of the human infant. New York: Basic Books. 

Mondadori, A. (Ed). (1978). The book of opera. New York: Simon and Schuster.

Tillitski, C. (1990). A meta-analysis of estimated effect sizes for group vs. individual vs. control treatments. International Journal of Group Psychotherapy, 40, 215-224.

This article was published in the April/May 1999 issue of The Group Circle.