How Being Bad Can Make You Better

Developing a Culture of Feedback in Your Practice

Barry Duncan, Scott Miller, and Mark Hubble
Magazine Issue
November/December 2007
How Being Bad Can Make You Better

Research now shows that a small group of clinicians—sometimes called “supershrinks“—obtain demonstrably superior outcomes in most of their cases, while others fall predictably on the less exalted sections of the bell-shaped curve. But thankfully, even mere mortals can significantly increase their therapeutic effectiveness by implementing the characteristics of these creme-de-la-creme therapists. How? The surprisingly simple answer is conveyed by Mae West’s famous quip, “When I’m good, I’m very good, but when I’m bad I’m better.” Granted, at first blush, these hardly seem like words for therapists to live by; but, as it turns out, they are.

Take the case of Matt, a twenty-something software wizard, frequently on the road, trouble-shooting customer problems. He loved his job. Even so, traveling had become an ordeal. For many, flying long ago lost its luster, but Matt had a personal problem that made flying unpleasant—an inability to urinate in public restrooms. At the outset, it caused only mild discomfort and was easily solved by repeated visits to the bathroom. In time, though, the problem began to cause intense apprehension before each trip, excruciating feelings of pressure in his bladder while on the plane, and hair-raising panic attacks sometimes. Hopeless and demoralized, the young man considered changing jobs. As a last resort, he decided to seek psychotherapy.

He liked his therapist and was glad that he could finally talk about his difficulty. In short order, he was helped to implement relaxation and “self-talk” cognitive-behavioral strategies, which he diligently practiced in session. As agreed, he employed them preceding his next trip and while on board the plane. The results were far from encouraging. The problem intensified and his sense of shame along with it. More alarming, his mood became decidedly depressive.

Now, three sessions into the treatment, Matt was at significant risk for a negative outcome: either dropping out or persisting in therapy without benefit. Unfortunately, his lack of success in therapy isn’t unusual. Current estimates indicate nearly 50 percent of clients prematurely bolt, and from a third to a half of those who remain don’t benefit from our standard strategies.

Learning from the Top Performers

So, what can the field’s supershrinks tell us about the best way to work with a client like Matt?

Psychotherapy’s “top guns” have three lessons to share:

Lesson 1: Step outside your comfort zone and push the limits of the effectiveness of your performance. This means identifying clients who aren’t responding to therapeutic business as usual and addressing the lack of progress in a positive, proactive way that keeps them engaged while you collaboratively seek new directions for the treatment.

To retain clients at risk for slipping through the proverbial crack, we need to remember what we know about positive change in therapy. Time and again, studies have revealed that the majority of clients experience change in the first six visits! This means that clients who report little or no progress early on will show no improvement over the entire course of therapy, or will end up on the drop-out list. Early change predicts engagement in therapy and a good outcome on termination.

This doesn’t mean that if a client reports early change, the problem is “cured” or completely resolved. It means that the client has a subjective sense that therapy has gotten under way and that she’s on the right path. This expectation is predictive of success.

A second robust predictor of positive change, solidly demonstrated by a large body of studies, is the strength of the therapeutic alliance. Clients who highly rate their relationship with their therapist are more apt to remain in therapy and benefit from it.

Lesson 2: Determine as clearly as possible how clients are responding to treatment and their degree of improvement. Assess these known predictors systematically with reliable, valid instruments. Instead of regarding the first sessions as a warm-up period or a chance to try out the latest technique, we should direct special attention to engaging the client from the beginning of therapy. In practical terms, this means asking the client from the start to help us judge whether therapy is progressing positively.

Obtaining feedback on standardized measures provides invaluable data about the prospects for treatment success or failure. Specifically, it gives us information about the match among ourselves, our approach, and the client. This knowledge ensures that we’re as clear as possible about whether we are or aren’t being effective in sessions. The point of seeking feedback is to assess our clinical effectiveness with the goal of improving what we’re doing with clients who aren’t experiencing progress at the beginning, so all our clients improve.

Using standardized measures to monitor outcome may evoke images of laborious torture devices like the Rorschach test or the Minnesota Multiphasic Personality Inventory (MMPI)—testing methods that often deliver a discouraging pathological picture. But the measures developed by Barry, Scott, and colleagues at The Institute for the Study of Therapeutic Change—a group of clinicians and researchers dedicated to translating research into clinical practice and using feedback to truly partner with clients—are simple and oriented to change, rather than assessing illness. They directly involve clients in monitoring progress toward their goals and how well the services they’re receiving fit their needs. Asking them to participate in this way, we’ve found, increases their sense of agency in any decisions about their care.

At this point, a reasonable person would probably say, “Oh great, more paperwork! That’s high on my list for the last thing I need for my practice.” However, finding out who is and isn’t responding to therapy need not be cumbersome. In fact, it only takes a few minutes using the Outcome Rating Scale (ORS) and the Session Rating Scale (SRS), which are available for free at

The four-item Outcome Rating Scale assesses three dimensions: (1) personal or symptomatic distress (individual functioning); (2) interpersonal well-being (how well the client is faring in intimate relationships); and (3) social role (satisfaction with work or school and relationships outside the home). Changes in these three areas are widely considered the most valid indicators of successful outcome. Constructed at an eighth-grade reading level, the ORS is easily understood, and clients have little difficulty connecting the items to their day-to-day experience.

Matt completed the ORS before each session. At the third session, when the ORS reflected no change, it motivated him and his therapist to brainstorm a range of possibilities for the remainder of therapy: changing nothing about the therapy, taking medications, or shifting the treatment approach. During this lively and open exchange, Matt expressed in no uncertain terms how much his problem was interfering with his work. The possibility that he’d have to endure any extended separation from his own bathroom had become almost unthinkable. He became quite animated, in conspicuous contrast to the passive resignation that had characterized previous sessions. When he or the therapist said the words “pissed off,” both broke into raucous laughter.

Later in the visit, the therapist suggested that instead of responding with hopelessness when the predicament occurred, Matt should work himself up into righteous anger about how the problem was sabotaging his life. Matt liked that idea and added, since he was a rock-and-roll buff, that he could also sing Tom Petty’s “Won’t Back Down” during his tirades at the toilet. From then on, he permitted himself, when standing before the urinal, to become thoroughly incensed, “pissed off,” and somewhat amused. His problem soon resolved.

Of course, this kind of collaborative, creative process could have happened with any therapist working with Matt. The difference is that the use of the ORS spotlighted the lack of early therapeutic change. Impossible to dismiss, it brought the risk of a negative outcome front and center. Without the findings from the rating scale, the therapist in this case might have continued with the same treatment for several more sessions, unaware of its ineffectiveness or believing (hoping) that his usual strategies would eventually take hold. As it was, the evidence obtained through the measure pushed him to explore different treatment options by the end of the third visit.

In addition to addressing outcome, pushing the limits of clinical effectiveness requires the therapist to determine if the service fits with the client’s expectations for the therapeutic alliance. By assessing the alliance at every session, therapists are able to identify and correct any weaknesses in their approach to the client’s problems before they exert a negative effect on outcome.

Research repeatedly shows that clients’ ratings of the therapeutic alliance are far more predictive of improvement than the type of intervention or the therapists’ own ratings of the therapeutic relationship. Mindful of these findings, we developed the Session Rating Scale (SRS), a brief alternative to longer alliance measures used for research, to encourage routine conversations in sessions about the client’s perception of the partnership.

The SRS contains four items: (1) A relationship scale, which allows the client to rate the session on a continuum from “I didn’t feel heard, understood, and respected” to “I felt heard, understood, and respected.” (2) A goals-and-topics scale, which rates the conversation from “We didn’t work on or talk about what I wanted to work on or talk about” to “We worked on or talked about what I wanted to work on or talk about.” (3) An approach or method scale, which asks the client to rate the meeting on a continuum from “The approach isn’t a good fit for me” to “The approach is a good fit for me.” (4) A scale that looks at how the client perceives the encounter in total from “There was something missing in the session today” to “Overall, today’s session was right for me.”

Providing feedback to clinicians regarding clients’ experience of the alliance and progress has been shown to lead to significant improvements in client retention and outcome. In our research, clients of therapists who opted out of completing the SRS were twice as likely to drop out. They were three times as likely to have a negative or null outcome. In the same study of more than 6,000 clients, effectiveness rates doubled in cases in which the rating scales were used. As remarkable as the results might appear, they’re consistent with findings obtained by other researchers. In a 2003 metanalysis of three studies, Michael Lambert, a pioneer in the use of client feedback, examined the effects of feedback in therapeutic relationships at risk for failure. Sixty-five percent of those cases in which feedback about progress was solicited ultimately had a more positive outcome at termination than cases in which information hadn’t been solicited.

Think for a moment: research suggests that even the best therapists will have 3 clients go home without benefit for every cycle of 10 clients they see. Over the course of a year, a clinician with a full caseload will have many unhappy customers leaving the office. If a sizeable portion of clients at risk for a negative therapeutic outcome can be recovered by identifying who they are, keeping them engaged, and tailoring services to them based on their feedback, doesn’t it make sense to do that?

Lesson 3: Seek, obtain, and maintain the highest possible level of consumer engagement. Clients usually drop out of therapy for two reasons: the therapy isn’t helping (thus, the need to monitor outcome) or the alliance, the fit between the therapist and client, is problematic. This isn’t rocket science. Clients who don’t feel they’re making progress or feel turned off by their therapist leave. Accordingly, the most direct way to improve effectiveness is keep people engaged in therapy.

A common alliance problem emerges when the client’s goals don’t fit with our sense of what they need. This may be especially true if clients come to us with certain emotionally charged
diagnoses or presenting problems. Consider 19-year-old Sarah, who lived in a group home and received social security disability for mental illness. She was referred for counseling because she was seen as socially withdrawn. People were also worried about her health because she was overweight and spent a lot of time watching TV and eating snack foods.

In therapy, Sarah agreed that she was lonely, but expressed a desire to be a Miami Heat cheerleader. That goal wasn’t taken seriously by the therapist. After all, she’d never been a cheerleader, was diagnosed with schizophrenia, and didn’t have the type of body cheerleaders usually have. No one listened, or even asked her why she had such an interesting goal. So therapy with Sarah understandably floundered. She rarely spoke in her sessions and answered questions as briefly as possible. She wasn’t engaged and was at risk for drop-out or a negative outcome.

The therapist routinely gave Sarah the SRS to complete. Though she usually reported that everything was going swimmingly, in time, a discrepancy began to emerge on the goals scale. Although she gave a 9 or above out of a possible 10 on the rest of the scales, she only gave an 8.7 to the goals scale.

Often it takes more than a bit of work to create the conditions that allow clients to be candid with us. Indeed, the disparity in power between therapist and client, combined with any socioeconomic, ethnic, or racial differences, can make it difficult for our clients to tell us we’re on the wrong track. When was the last time you told your physician, “Listen, you’re making a big mistake with me”? But clients will subtly let us know what they think of us on alliance measures.

At the end of the third session, the therapist and Sarah jointly reviewed her responses on the SRS. At issue were questions like, Did she feel understood? Was the therapy focused on her goals? Did the approach make sense to her? We’ve found such reviews invaluable in fine-tuning the therapy and addressing problems in the therapeutic relationship that have been missed or gone unacknowledged. When asked the question about goals, Sarah repeated her desire to be a Miami Heat cheerleader, all the while avoiding eye contact and nearly whispering.

The counselor looked at the SRS and off came the blinders! The slight difference on the goals scale told the tale. When the therapist finally asked Sarah about her goal, she related a story about being young and watching Miami Heat basketball with her dad. Her father took special delight in her performance of the cheers. Sarah sparkled when she spoke of her dad, who’d
died several years previously. Noticing that it was the most he’d ever heard her talk, the therapist began asking Sarah about her father. He also put the brakes on his efforts to get her to socialize or exercise (his goals), and started focusing instead on Sarah’s interest in cheerleading. It turned out that she regularly watched cheerleading contests on ESPN and enjoyed sharing her knowledge with others. She was also quite informed about basketball.

Sarah’s SRS score improved on the goals scale and her ORS score increased dramatically. After a while, she organized a cheerleading squad for her agency’s basketball team, who played local civic organizations to raise money for the group home. Her involvement with the team successfully addressed the referral concerns about her social withdrawal and lack of activity. Walking the path cut by client goals often reveals alternative routes to improvement that would have never been discovered otherwise.

Failing Successfully

As with Matt and Sarah, it’s often possible to change course and make the therapy experience far more productive for clients. But what happens when the practitioner and the client have implemented new or different strategies and the therapy is still failing? This can be singularly confusing if the therapeutic alliance is strong and therapeutic. The following case was especially instructive for Barry on this theme.

Eighteen-year-old Alina sought help because she felt completely devastated. In her estimation, she’d lost everything she’d worked her whole life to achieve. After captaining her high school volley-ball team, commanding the first position on the debating team, and being named valedictorian, she’d won a full scholarship to Yale. She was the pride of her Guatemalan community—proof of the many benefits her parents had envisioned would come with living in the United States.

Her unqualified success unraveled during her first semester away from home and the insulated environment in which she’d excelled. She began hallucinating and hearing voices. After a visit to the university counseling center, she was admitted to a psychiatric unit and dosed with antipsychotic medications. Despondent, she threw herself down a stairwell, prompting her parents to bring her home. Alina returned home in utter confusion, still hearing voices, and with her self-image badly eroded. Besides seeing herself a failure and major disappointment to her family, she believed she’d let down everyone else in her tightly knit community.

Barry was the 20th therapist the family called, and the first who agreed to see Alina without medication, a precondition she imposed. From the start, it looked as though they hit it off famously. Her investigation of Barry on the web had revealed his consumer-driven philosophy and his leanings away from pharmaceutical solutions, both of which scored high marks with Alina. For his part, Barry admired her humility and intelligence. He was especially taken by her spunk in standing up to psychiatric discourse and asserting her preferences about treatment. Perhaps his history of dropping out of college after a tumultuous first semester, albeit for different reasons, contributed to the strong affinity he felt for Alina and his almost urgent desire to get her back on track.

The two discussed at length her experience of demoralization—how the episodes of hearing voices and confusion robbed her of her dreams, and how her years of hard work had yielded nothing. Barry did what he usually found to be helpful in his clinical practice: he listened, commiserated, validated, and worked hard to mobilize Alina’s resilience to begin anew.

But nothing happened. By session three, she remained unchanged in spite of Barry’s best efforts. The therapy was going nowhere fast and he knew it because of her responses to the ORS. Hard to rationalize these results, since a score of 4 (out of a possible 40) was a rude reminder of just how badly the treatment was progressing.

In the face of this impasse, Alina was specific about what was missing, unlike many clients. She told Barry that she wanted him to be more active. He agreed. She also wanted ideas about how to manage the voices. So, with her participation, he challenged them with thought-stopping, guided imagery, and content analysis. Still no change occurred, and she was increasingly at risk for a negative outcome.

Alina then told Barry she’d read about hypnosis on the Internet and thought that might help her. So he approached her using embedded suggestions, metaphors, and stories intended to build up her immunity to the voices. Though she responded with deep trances, her ORS score stayed firmly fixed at 4.

Barry brought up the possibility of a referral, but they settled instead on a consultation with a team of therapists. Alina again responded well and was more actively engaged than she appeared to be with him alone. On the SRS, she gave the session the highest valuation possible on the measure. The team introduced considerations that Barry hadn’t thought of, including differentiation from her family, as well as gender and ethnic issues. They pursued the ideas from the team for a couple sessions. No progress, though—her ORS score remained a 4.

Now what? They were at session nine, well beyond the number of visits clients typically required for change in Barry’s practice. After collecting data for several years, he knew that 75 percent of the clients who benefited from their work with him would show it by the third session. Ninety-eight percent who profited would do so by the sixth. Considering this data, was it right to continue seeing Alina? Was it ethical?

Despite their mutual admiration society, he decided that a good relationship without change is just another definition of dependence. Barry told her that the lack of movement had nothing to do with either of them; they’d both tried their best, but for whatever reason, it wasn’t the right mix for change. They discussed the possibility of having Alina see another therapist. Finally, after a protracted period during which she insisted she wanted to keep working with Barry, they began exploring who else she might see. She chose someone from the consultation team with whom she’d felt a connection.

By session four with her new therapist, Alina had an ORS score of 19 and had enrolled in a class at a local university. Moreover, she continued progressing and reenrolled at Yale the following year, her scholarship intact! In retrospect, by getting out of her way and allowing her and himself to “fail successfully,” Barry made it possible for Alina to have another chance to get her life back on track.

This client turned out to be a watershed for him. He believed in heeding the results of the measures, what we’ve termed “practice-based evidence.” Nevertheless, he struggled with letting go of those clients who didn’t benefit and yet wanted to continue. This was even more difficult when he really liked the clients and had become personally invested in their lives.

Alina awakened him to the peril of such situations. As many therapists encounter over their careers, for Barry, this young woman represented a client he was tempted to see for as long as she wanted. He cared deeply about her and believed that, in time and with the right amount of persistence, he’d surely find a way to help her.

The wish to save, combined with old-fashioned stubbornness, epitomizes the inattention to the iatrogenic effects of continuing therapy in the absence of benefit that results in “chronic clients.” Just as people don’t marry everyone they date, therapists—no matter how competent, trained, or experienced—can’t be effective with everyone they meet. Regularly using the rating scales our team has developed makes it easier for therapists to determine when their clients aren’t improving and come to more objective decisions about how to strengthen the quality of their services.

In the end, practice-based evidence—the information gathered through measures of progress and the alliance—won’t help with the clients you’re already successfully assisting. It helps with those who aren’t improving by encouraging an open discussion of all options, including stopping a dead-end treatment and moving the client into a more productive relationship

The basic principle behind this approach is straightforward: our daily clinical actions can be informed by reliable, valid feedback about the factors that drive change in therapy. These factors include our clients’ engagement, their view of the therapeutic relationship, and “the gold standard”: the client’s report of real progress. Truly, monitoring the outcome and the fit of our services helps us know that when we’re good, we’re very good, but when we’re bad, we can be even better.

Barry Duncan, Psy.D., Scott Miller, Ph.D., and Mark Hubble, Ph.D., are cofounders of the Institute for the Study of Therapeutic Change (ISTC). Together, they’ve authored and edited numerous professional articles and books, including The Heart and Soul of Change: What Works in Therapy; Escape from Babel; Psychotherapy with Impossible Cases; and The Heroic Client.


References and Resources

Duncan, Barry L., Scott D. Miller, and Jacqueline A. Sparks. The Heroic Client: A Revolutionary Way to Improve Effectiveness, rev. ed. San Francisco: Jossey-Bass, 2004.

Lambert, Michael J., J. L.Whipple, E. J. Hawkins, et al. “Is It Time for Clinicians Routinely to Track Patient Outcome? A Meta-Analysis.” Clinical Psychology: Science and Practice 10 no. 3 (September 2003): 288-301.

Miller, Scott D., Barry L. Duncan, George S. (Jeb) Brown, et al. “Using Formal Client Feedback to Improve Retention and Outcome. Journal of Brief Therapy 5 (2006): 5-22.

Wampold, Bruce E. “The Psycho-therapist.” In John C. Norcross, Larry E. Beutler, and Ronald F. Levant, eds., Evidence-Based Practices in Mental Health: Debate and Dialogues on the Fundamental Questions Washington, D.C.: American Psychological Association, 2006.