The Science Behind T.E.A.M. Therapy

Copyright © 2011 by David D. Burns, MD. Do not cite in whole or in part without the express written permission of Dr. Burns.


I would like to thank Britney Blair, Psy.D, for superb help in editing this manuscript.

I first became involved in Cognitive Behavior Therapy (CBT) in the mid-1970s when I was a research fellow at the University of Pennsylvania Medical School. Initially, I was skeptical that any form of talk therapy could help patients suffering from severe clinical depression, but soon discovered that the methods could be surprisingly helpful, even for patients I’d been stuck with while using more traditional forms of psychotherapy in combination with antidepressant medications. In one of the first controlled outcome studies, CBT without medications proved to be as effective as a popular antidepressant for patients suffering from clinical depression.

Eventually, I decided to give up my full-time academic career researching the “chemical imbalance” theory of depression, so I could pursue full-time clinical work using CBT. However, I quickly realized that while CBT was often effective, it was limited to a few techniques Dr. Aaron Beck had developed to help patients challenge their negative thoughts. These included the Daily Activity Schedule, the Socratic Method, Examine the Evidence, the Experimental Technique, Reattribution, Thinking in Shades of Gray, and a couple of others.

  • Daily Activity Schedule. Record what you do each hour of the day and rate how satisfying each activity was on a scale from 0 (not at all satisfying) to 5 (the most satisfying).
  • Socratic Method. Ask yourself questions that will reveal the inconsistencies in your negative thoughts.
  • Examine the Evidence. Instead of assuming that your negative thought is true, examine the evidence for it.
  • Experimental Technique. Design and carry out an experiment to test the validity of your negative thought, in much the same way that a scientist would test a theory.
  • Reattribution. Instead of blaming yourself entirely for a problem, think about the many factors that contributed to it.
  • Thinking in Shades of Gray. Instead of thinking about your problems in black-and-white categories, you  evaluate them in shades of gray.

Although these techniques were often extremely helpful, they weren’t always sufficient, and sometimes they didn’t seem to work at all. Because of this limitation, I began to develop many new CBT techniques, such as the Paradoxical Double Standard Technique, the Externalization of Voices, the Feared Fantasy, the Acceptance Paradox, the Individual and Interpersonal Downward Arrow Techniques, the Hidden Emotion Technique, the Straightforward and Paradoxical Cost-Benefit Analysis, and a multitude of others that are still evolving today. I also created the list of ten Cognitive Distortions and refined some of the basic tools for helping patients with depression and anxiety disorders, such as the Daily Mood Log and Recovery Circle.

Then I developed a new form of relationship therapy I called Cognitive Interpersonal Therapy (CIT). CIT is  based on these key ideas:

  • We create the very relationship problems that we complain about, but we are unaware that we are doing this, so we play the role of  innocent victim and insist that the other person is entirely to blame for the problem.
  • We are unwilling to examine our own role in the problem, because this is so painful and so shocking.
  • We have far greater power than we think to transform troubled relationships, and often quite quickly, but there’s a price to be paid. We have to be blaming to pinpoint our own role in the problem and focus all of our energy on changing ourselves, rather than try tot change the other person.

I developed powerful new tools for implementing these concepts, including the Relationship Cost-Benefit Analysis, the Relationship Journal, the EAR Checklist, the Good vs. Bad Communication Checklist, the Five Secrets of Effective Communication, Changing the Focus, the One Minute Drill, Intimacy Exercises, and other techniques. If you would like to learn more about these tools, you might want to order the Therapist’s Toolkit or my eBook, Tools, Not Schools, of Therapy. For order forms or more information, feel free to contact me at

I wrote Feeling Good because I wanted to share my excitement with colleagues as well as the general public. In addition, I wanted a guidebook that patients could read between sessions in order to accelerate their recovery. When Feeling Good was first released in 1980, very few people had heard of cognitive therapy. Since that time, the book has sold more than four million copies, and CBT has become one of the most widely used forms of psychotherapy in the world. In addition, a great deal of research over the past 35 years has confirmed the effectiveness of CBT. I am proud that Feeling Good has been so influential in making therapists and the general public alike aware of these important developments.

Some of the research has investigated the effectiveness of the original “Beckian” model of CBT for a wide variety of clinical problems, especially the mood and anxiety disorders, as well as marital problems, habits and addictions, eating disorders, and even schizophrenia. Other research has focused on “bibliotherapy.” This means delivering therapy via a book, such as Feeling Good, without any psychotherapy or medication. And some research has been focused on these  questions: How does CBT (or any form of psychotherapy) actually work? What are the ingredients of therapeutic success or failure? And how can we use this information to develop more powerful and effective treatments?

Controlled Outcome Studies

Many research studies have examined the effectiveness of CBT in the treatment of mood and anxiety disorders. Most of these studies have used the “horse race” strategy. Here’s how it works: Carefully screened depressed patients are randomly divided into different groups, and each group receives a different treatment. For example, one group may receive CBT with no medications, and the other group may receive medication and no CBT. Some studies have included placebo groups or groups that receive CBT plus medications, and some have included groups receiving some other type of psychotherapy. Generally speaking, these studies have confirmed the effectiveness of CBT and other types of psychotherapy (DeRubeis, Hollon, Amsterdam, et al., 2005; Dobson, 1989; Hollon & Beck, 1994; Hollon, DeRubeis, Shelton, et al. 2005; Robinson, Berman, & Neimeyer, 1990).

For example, in one large, well-controlled outcome study, CBT was found to be comparable to the popular antidepressant medication paroxetine (Paxil) in the short-term, and slightly more effective in the long run, when patients were contacted a year or more after treatment (DeRubeis et al., 2005; Hollon et al., 2005). Most researchers and clinicians have concluded that if CBT is at least as good as treatment with antidepressants, then it must be effective. In addition, they have pointed out that antidepressants have limited effectiveness as well as numerous side effects, including significant increases in suicidal urges and completed suicides (Garland, 2004; Healy, 2003; Jureidini, Doecke, & Mansfield, et al., 2004; Khan, Khan, Leventhal, & Brown 2001; Khan, Warner, & Brown, 2000; Whittington, Kendall, Fonagy et al., 2004), so the existence of an equally effective drug-free treatment is exciting.

Researchers have also evaluated the combination of CBT with antidepressants. Some investigators have reported that the combination of CBT plus an antidepressant medication is more effective than either modality alone, but other investigators have been unable to replicate this effect. In an outcome study of patients treated at my outpatient clinic in Philadelphia, Dr. Susan Nolen-Hoeksema and I reported that patients who received CBT plus medication actually recovered more slowly than patients treated with CBT alone (Burns & Nolen-Hoeksema, 1991; 1992). We did not conclude that the antidepressants actually delayed recovery, since patients were not randomly assigned to CBT alone vs. CBT plus medications. However, we saw no evidence that antidepressant medications enhanced treatment with CBT. This finding is entirely consistent with my clinical experience over the last 35 years.

To my way of thinking, all of the outcome studies with CBT as well as antidepressants are somewhat disappointing for a number of reasons. First, many patients do not experience a significant reduction in their depression or anxiety, so these treatments are definitely not panaceas. Second, if you examine the data closely, and understand the rating scales the investigators used, it becomes clear that neither CBT nor antidepressants (nor any form of psychotherapy) appears to be much better than treatment with placebos. In fact, many recent research studies indicate that the so-called “anti-depressant” medications may have few or no significant anti-depressant effect above and beyond their placebo effects (Kirsch & Sapirstein, 1998; Kirsch, Moore, Scoboria, & Nicholls 2002; Kirsch, 2010). So showing that CBT, or any other form of psychotherapy, is comparable to antidepressant drug treatment is rather faint praise because none of these treatments appear to be much better than treatment with inert placebos.

The idea that all of the so-called antidepressant medications and all of the current forms of psychotherapy for depression are barely better than placebos may seem confusing or shocking to you. After all, we all know people who will say that this or that antidepressant “cured” them and changed their lives. Here’s the problem. Approximately 35% to 50% of depressed or anxious individuals who are given a placebo pill will recover if they believe they have received a “real” medication. They will swear that the medication altered their brain chemistry and changed their lives.

In reality, these patients either experienced a spontaneous improvement in their depressed mood or actually healed themselves through a change in their outlook and behavior. In other words, if you strongly believe something will help you, in many cases it will. But it’s your thinking that cured you, and not the pill. This is unfortunate, because people nearly always give the credit to their “pills,” rather than their own efforts. This is called the “placebo effect,” and it confounds research on medications and psychotherapy.

In order for any treatment to be truly deemed “effective” it must provide an effect significantly superior to placebo. Sadly, this is not the case for any of the currently prescribed antidepressant medications or any currently practiced forms of psychotherapy.

The results of the research studies mirrored my clinical experience. Although I observed amazing and inspiring changes in many of the people I was treating with CBT, some patients did not respond nearly as well as I hoped. Many had difficulties engaging with the psychotherapy homework (HW) assignments between sessions which is a central component to the CBT treatment model, and some seemed to resist my efforts. If you have treated significant numbers of patients struggling with depression, anxiety, relationship problems, or addictions, I suspect you’ve encountered similar problems with many of your own patients.

Over time, it became increasingly clear that while the “Beckian” version of CBT was a great tool, it was definitely not the complete answer. I was eager to find out why. What distinguished the patients who got better from those who didn’t? What are the ingredients of therapeutic success or failure? I was hopeful that if I could answer this question, it might be possible to develop a far more robust form of therapy that would build on, but go far beyond, traditional CBT, as well as other forms of psychotherapy.

Process Research

I have spent most of my career as a clinician in private practice. Due to practical and ethical limitations I have focused on process research rather than controlled outcome studies. This means that you analyze data collected during clinical treatment in real world settings in an attempt to learn how therapy actually works so that more effective forms of psychotherapy for depression, anxiety disorders, relationship problems, and habits and addictions could be developed. Ultimately, outcome studies will be needed to confirm the effectiveness of the new treatment methods. My Stanford colleague, Mickey Trockel, MD, PhD, is currently working with my group to see if we can obtain funding for this work.

This research effort began in the 1980s when my private practice and clinic were in full swing in Philadelphia. We studied all the patients seeking treatment at the time of initial evaluation at the clinic and once again 12 weeks later with a wide variety of assessment instruments so we could track changes in depression, anxiety, anger, and other symptoms, as well as other variables of interest, such as therapist experience, the patient’s perceptions of therapist empathy, patient motivation, and compliance with psychotherapy homework assignments, to name just a few. I have published the results in journals such as Journal of Consulting and Clinical Psychology, Journal of Abnormal Psychology, Behavior Therapy, and others. Based on this research, I have developed a new approach called T.E.A.M. Therapy. Recently, I have developed systematic training methods for therapists who want to learn T.E.A.M. Therapy.

Here’s what the acronym stands for:

T = Testing. Therapists learn to test the severity of depression, suicidal urges, anxiety, and anger at the start and end of every therapy session, as well as well as relationship satisfaction and positive feelings such as joy, self-esteem, and productivity. We also assess the patient’s perception of therapist empathy, helpfulness, and satisfaction, using brief, empirically validated and extremely sensitive scales at the end of every session.

Patients complete the scales in the waiting room immediately before and after the session, so the testing requires no therapy time. The testing does not detract from the therapeutic process, but is an incredibly helpful, even essential, therapeutic tool. The data provides patient and therapist alike with precise information about the patient’s progress, or lack of progress, throughout the entire therapeutic journey, and allows for true implementation of the “scientist—practitioner model” in real time.

E = Empathy. Therapists receive rigorous empathy training so they can respond far more skillfully to three common but challenging therapeutic scenarios:

1. The critical, mistrustful patient who claims that the therapist isn’t helping and doesn’t care or understand.

2. The patient who is overwhelmed with depression, profound, despair, suicidal urges, panic, or rage.

3. The patient who makes sexually provocative comments to the therapist or accuses the therapist of inappropriate feelings or behavior.

During role-playing exercises, therapists receive letter grades along with immediate feedback on what they’re doing right or wrong, and specific tips on how to respond more effectively. These exercises are illuminating but require courage, since most therapists receive poor or failing grades initially. This can be a shock to the system, but if you check your ego at the door, it can provide an opportunity for significant personal and professional growth.

A = (Paradoxical) Agenda Setting. Therapists learn to head off resistance at the pass before using any techniques to try to help patients overcome depression, anxiety, relationship problems, or habits or addictions. This is one of the most unique and important aspects of the TEAM Model and accounts for the extremely rapid recovery we are now seeing in many patients who used to require prolonged treatment. The other side of this coin is that nearly all therapeutic failure results from Agenda Setting errors, or the complete failure of the therapist to set the agenda.

Therapists use an eight-dimension system to pinpoint motivational factors that are likely to interfere with effective treatment. There are two unique types of resistance, called Outcome Resistance and Process Resistance, which are completely different for depression vs. anxiety vs. relationship problems vs. habits and addictions. Patients can have Outcome Resistance and/or Process Resistance for each of the four targets. Two times four equals eight—that’s why there are eight completely different types of resistance.

Outcome Resistance means that the patient may fight intensely against a positive treatment outcome, even if no work or effort were required, and all the patient had to do was press a Magic Button and she or he would be instantly cured. Although this may seem counter-intuitive, it actually makes sense once you really comprehend how the patient is thinking and feeling. For example, a woman named Janice had experienced ten years of unsuccessful treatment for severe depression since her abortion at age 19. Extensive psychotherapy had been completely ineffective, and she had taken nearly every known antidepressant medication with no success. She was still experiencing intense feelings of worthlessness, hopelessness, guilt, and self-loathing.

When I asked Janice what negative thoughts went through her mind when she was feeling depressed she said she was telling herself, “I murdered my baby. I deserve to suffer forever.” It’s not surprising that previous therapeutic efforts had failed. Although she’s suffering tremendously, and desperately wants help, there’s a side of her that will fight against getting that help. That’s because she’s simultaneously playing the role of criminal as well as judge, jury, and executioner. She believes she does not deserve to be forgiven for the terrible thing she’s done, so her depression paradoxically allows her to feel pure because she is punishing herself for her bad, unacceptable behavior. In a way, her depression is every bit as much a spiritual issue as it is a psychological issue. And before we can open the doors to recovery, Janice will have to find a spiritual justification for allowing herself to feel joy and self-esteem once again. Therapists who don’t deal with this crucial issue will be doomed to failure.

And that’s only part of the story. In addition, Janice’s depression represents her love for the baby, who hasn’t really died yet. She hasn’t grieved or said goodbye to him. Instead, she thinks about him constantly, thus keeping him alive. Pressing the Magic Button would mean letting him go, grieving, and moving ahead with her life.

Those are just two of a number of Outcome Resistance issues. The Process Resistance issues are different. There is no Magic Button, so if Janice wants to recover, there’s something she’ll have to do, that she probably won’t want to do, to get better. For depression, the outcome Resistance issue always involves psychotherapy homework. If Janice agrees to do daily psychotherapy homework, there is little chance the therapy will fail. If she is unwilling to do this,the prognosis for full recovery is very poor.

To address this issue, we will need to ask Janice what price she’d be willing to pay to overcome her depression. If we agree to work with her, and show her how to defeat the negative thoughts and feelings that have been ruining her life, what would it be worth to her? For example, would she be willing to do psychotherapy homework for thirty minutes a day, five or six days a week, whether or not she’s “in the mood,” and whether or not she’s feeling motivated? This homework might involve recording and challenging her negative thoughts, using the Daily Mood Log, or reading a chapter from a self-help book, such as Feeling Good, and so forth. Getting her explicit agreement to do psychotherapy homework before beginning to work together can have a profound impact on the outcome of the treatment.

Therapist learn to melt away each pattern of resistance using sophisticated techniques such as the Magic Button, Acid Test, Magic Dial, Gentle Ultimatum, Sitting with Open Hands, Paradoxical Inquiry, and several other key techniques which will be described below. We have found that these techniques are absolutely mandatory with patients who are resistant to change, as well as those many who have mixed feelings about change.

M = Methods. Once the therapist has melted away the resistance he or she will select the most effective techniques from a list of 50, to help each patient’s specific issues. Some techniques are especially helpful for mood disorders. Other techniques are more useful for anxiety disorders, relationship problems, or habits and addictions. The methods are not imposed on a patient, based on his or her diagnosis, but rather depend on the precise problem the patient wants help with in a specific session. This approach is radically different from manualized therapy, which has become so popular in recent years.

T.E.A.M. Therapy is not a new “school of therapy,” but rather a basic and flexible approach to understanding and treating human suffering that is based on years of research and clinical experience. Let’s take a look at some of the studies that led to this new approach.

T = Testing

After several years of research at the University of Pennsylvania Medical School on the “chemical imbalance” theory of depression, I decided to go into clinical practice because I was so excited about the newly emerging psychotherapy techniques, including CBT. I was aware, form my research experience, that clinicians’ perceptions of how patients are feeling are very often inaccurate. Therefore, I decided to continue using assessment measures with all my patients and required them to take at least one assessment test between therapy sessions so I could track their progress. I felt it was imperative to track changes in symptoms in order to ensure I was providing the best care possible. In addition, I hoped I might be able to analyze the data I was collecting in order to learn more about how psychotherapy works.

Several years ago I read a startling research study by Hatcher and colleagues, although the investigators did not really explore the implications of one of their key discoveries (Hatcher, Barends, Hansell, & Gutfreund, 1995). The researchers wanted to compare patients’ and therapists’ views of therapist empathy, so they asked patients and their therapists to complete similar scales that assessed the quality of therapist empathy. The patients rated their therapists on items assessing warmth and understanding, and the therapists rated themselves on the same items. Of course, they completed the ratings separately.

The researchers reported a statistically significant correlation between the two scales in the range of .3, indicating there was some overlap in their perceptions as well as some substantial discrepancies in patients and therapists views of the quality of the therapeutic alliance. In fact, a correlation of .3 actually means that only 9% of the variance (or variability) in the therapists’ and patients’ perceptions was shared in common.

Here is the rather shocking conclusion: the study indicated that the therapists’ perceptions of how their patients viewed their warmth and understanding were less than 10% accurate. For the most part, therapists and patients viewed the quality of the therapist empathy in a radically different way; in other words, therapists tended to greatly overestimate or underestimate their own warmth, empathy and helpfulness.

As I reflected on the implications of this study, it dawned on me that my own clinical and research experience had led me to a similar conclusion. Therapists’ misperceptions are not limited to the quality of the therapeutic relationship; they make equally severe errors in their assessments of how depressed, suicidal, anxious, or angry the patient is as well.

For example, I recently conducted a study on the psychiatric inpatient unit at the Stanford Hospital, in which I evaluated how accurate therapists’ perceptions of patients were after an interaction. Student researchers interviewed patients for several hours as part of a research study on psychiatric diagnosis. The interviews focused on how the patients were feeling and what kinds of problems they were experiencing, such as depression, anxiety, and so forth.

At the end of the interview, the patients reported how depressed, suicidal, angry, and anxious they were feeling. They also rated the warmth and empathy of the student researchers who interviewed them, using brief but empirically validated assessment measures. The student researchers completed the same scales at the same time, but were asked to guess how the patients were feeling, based on the extensive interviews they had just completed. The results were similar to the findings by Hatcher et al (1995)—the correlation between the reports of the patients and the researchers was extremely low, and indicated that accuracy of the researchers was, for the most part, less than 10%. In fact, for certain types of feelings, such as anger, the therapists’ accuracy was 0%.

You might think, “Oh, those were just students. I’m a seasoned clinician, and my perceptions are far more accurate.” In extensive informal studies, evaluating the accuracy of experienced clinicians working with their own patients, the findings are similar. In most cases, clinicians’ perceptions of how their patients feel are not accurate. In other words, we think we know how our patients feel, but we are rarely correct. We’re making assumptions that may not reflect reality.

Let me give you a brief example so you’ll know exactly what I mean. We use a five-item anger scale that patients can complete in less than 30 seconds. Patients indicate how angry, frustrated or annoyed they feel at this moment, on a scale from “not at all” (scored as a 0) to “extremely” (scored as a 4). So the total score can range from 0 (no anger at all) to 20 (extreme rage). In some cases, the therapist rated the patient’s anger at zero, but the patient rated his or her own anger at twenty. So we’re talking about obvious and severe errors in perceptions of how patients feel, and not just some statistical finding.

I want to emphasize that therapists from all schools of therapy make these errors, without being aware of it. That’s why all of us, including therapists trained in the new T.E.A.M. Therapy techniques, NEED TO MEASURE, because we simply do not perceive how other people are feeling with any real accuracy.

Understandably, these misperceptions can sometimes have serious consequences. If a patient is feeling suicidal, but the therapist is not aware of those feelings, it could lead to the loss of life. Sadly, this outcome is not uncommon. Published studies have estimated that as many as 10%, or more, of chronically depressed patients eventually commit suicide, even if they have received treatment. If the patient is struggling with Borderline Personality Disorder, substance abuse, or Schizophrenia, the probability of eventual suicide may be even higher.

Of course, there are many reasons that people commit suicide, including feelings of hopelessness as well as feelings of bitterness.  But one of the reasons is rather simple: the therapist did not know the patient was feeling that way. If the therapist had known, she or he could have intervened and probably saved that patient’s life.

I do not mean to put this in a blaming manner, because suicide is an enormous and almost unbearable tragedy for everyone involved—the patient, the family and friends, and the therapist as well. Yes, the therapist is every bit as much a victim as the people who knew and loved the patient. And given the threatening litigious era that we find ourselves in, many suicides lead to lawsuits that can drag on for years. I am convinced that many if not most of these tragedies could be prevented if therapist were armed with accurate information about how patients feel.

That’s one reason, among many, why we require therapists learning T.E.A.M. to use brief, empirically validated assessment scales at the start and end of every therapy session. Patients also rate the quality of the therapeutic alliance in multiple dimensions at the end of every session. You can see an example of these scales by clicking here. This particular patient experienced a fairly dramatic improvement in his depression, suicidal feelings, and anger during the session. His anxiety only improved somewhat (from 13 to 8), and his satisfaction with his marriage remained poor

If you click here, you can see how he rated his therapist at the end of the session. In spite of his improvement, the score on the Empathy Scale was 18 out of 20. Specifically, he did not feel completely cared about or understood. In addition, the Helpfulness Scale score was low, indicating that he did not find the session nearly as helpful as he had hoped. The other ratings indicate additional problems in the alliance that can be productively addressed at the start of the next session.

Many of these problems can be addressed fairly easily if the therapist approaches the feedback with warmth and humility. For example, you might say something like this:

“Ralph, on the empathy scale you indicated that I didn’t always convey as much warmth and support as you might have wanted. You also indicated that I didn’t always understand how you were feeling inside. I greatly appreciate that feedback, and want you to know that we’re on the same page. Last session I also felt that I wasn’t as warm as I wanted to be, and it seemed like I didn’t always understand how you were really feeling. This is incredibly important. Tell me a bit more about where I’ve been missing the boat.”

Of course, your response has to be genuine and will also depend on the specifics of the situation. We’ve found that if we treat negative feedback in a relaxed and interested way, the patients seem to appreciate it enormously, and we frequently end up with perfect empathy scores at the end of the session.

Therapists often ask what I do if patients don’t want to take the tests. In my experience, this rarely happens. At the initial evaluation I inform patients that the testing is not optional, but an inherent and necessary part of the therapeutic process. If patients object, I let them know that although I would love to work with them, if they feel they would prefer not to be treated with the session-by-session testing, I would not be able to accept them as my patient. I’ve never had a patient refuse or drop out of therapy for this reason.

During the late 1980s, I helped create a daily cognitive therapy group program at the Stanford Hospital as part of my volunteer work for the Department of Psychiatry. Once the program was in full swing, I would attend the groups once a week to teach and to see how the program was working. We used scales similar to the ones described above but they were modified somewhat for group work.

At the start of one group, a woman we will call “Missy” reported moderate levels of depression, anxiety, and anger on the Brief Mood Survey, with scores in the range of 15 or 16 on each scale. She also reported fairly strong suicidal urges. During the group, I helped her challenge her negative, self-critical thoughts using a variety of techniques. It seemed like we really clicked, and I was convinced she had experienced a profound improvement in mood during the session. I was eager to see her Brief Mood Survey and Evaluation of Therapy Session scores at the end and was already congratulating myself on what a terrific group leader I had been.

I was surprised to see that Missy put 4′s on every item on the Depression, Anxiety, and Anger scales, with total scores of 20 on each scale. These ratings indicated that she was feeling much worse—in fact, horrible—at the end of the session. I was even more surprised to see that she put 0′s on all the items on the Therapeutic Empathy and Helpfulness scales. I had never seen scores that low before.

I was pretty sure that Missy had made a mistake when she filled out the scales, so I ran out into the hall and asked her to come back into the group therapy room so she could review her answers. I pointed out that on the Brief Mood Scale, the good answers are on the left and on the Evaluation of Therapy Session, the good answers are on the right. Missy glanced at the test and said, “I don’t think there are any mistakes here, Doctor.”

I said, “But I thought we just had a really good session.”

She scowled and said, “Good for you, maybe!”

Missy went on to explain that I had said something during the session that had hurt her feelings. I believe I had used the term “double whammy” in describing her depression, since she had had two devastating losses just prior to her admission to the hospital. To her, my comment sounded sarcastic and lacking in compassion. She felt devastated, but had concealed her feelings from me.

I was able to sit down with Missy and we repaired the rupture. If I had not been using the assessment tests, I would have wrongly concluded that had done a great job and that the group had been extremely helpful for Missy.

Along with exaggerating your own warmth and helpfulness, at times you may also grossly underestimate your effectiveness. At the start of the therapy group the next week, I noticed that a newly admitted patient named Rose had scored 4s on every item on the Depression, Suicide, Anxiety and Anger scales. She explained that she had just been admitted to the locked unit for a nearly successful suicide attempt, and defiantly announced that she intended to complete the job at the first chance she got.

She explained that she had been treated for Borderline Personality Disorder for years, but nothing had helped. She still felt horrible “every minute of every day.” She reported struggling with drug and alcohol addiction for years, and had been living in a recovery community in Los Angeles. After several months of sobriety she relapsed following an argument with one of the administrators. The staff threatened to kick her out of the facility, at which point she drove to the Bay Area and tried to kill herself. She was brought to the hospital by the police and was admitted to the locked unit. She said she felt worthless and hopeless, and had decided it was time to “cash it in.”

I asked Rose if she would like to work on these feelings during the group. She angrily retorted that I sounded like every other stupid “shrink” she had ever wasted time with, insisted that she had no interest in any of my help, and thought psychotherapy was a fraud.

I was taken aback and decided to work with a different patient who also felt hopeless and worthless. Rose did not utter another word during the group. I was afraid to look in her direction but could sense that she was staring at me with anger and disdain. I felt the group had been reasonably productive, but I dreaded having to look at Rose’s Brief Mood Survey and Evaluation of Therapy Session at the end of the session.

I was shocked when I saw that all her scores on the Depression, Suicide, Anxiety and Anger scales had fallen to zero. It was as if all her negative feelings had vanished, and I wondered, again, if she’d filled it out correctly. I was even more surprised to see that she’d given me perfect scores on the Therapeutic Empathy and the Helpfulness scales.

At the bottom, where patients describe what they liked the least about the session, Rose had simply written, “Nothing.” In the section where patients describe what they liked the most about the session, she wrote:

“Doctor Burns, when you worked with that other woman, I felt like you were working with me. Her feelings were just like mine. I’ve never had any cognitive therapy before, but now I can see how distorted my negative thoughts are. I’ve never thought about this before, because I always believed that I really was worthless and hopeless. I can hardly believe it, but I’m suddenly feeling self-esteem. In fact, this is the first happiness I’ve experienced since I was a child. Thank you so much! The last hour and a half has changed my life.”

If I had not been using the assessment tools, I would never have known that the group had been so helpful to Rosie, and might even have gone out of my way to avoid her when I was visiting the unit.

Some therapists are concerned that patients will not be honest when they complete the rating scales. They think patients will just tell them what they think they want to hear. For example, they think patients will report improvement in order to please the therapist, even if they aren’t feeling better. Or they think patients will give them overly high ratings on the Empathy or Helpfulness scales for the same reason.

Our experience indicates that this is not a problem. In the vast majority of instances patients have been extraordinarily honest in the way they fill out the scales. In fact, they seem far more willing to criticize therapists when filling out the feedback forms than when speaking to their therapists face-to-face. Sometimes, the information the patient provides may be upsetting to the therapist.

For example, on the five-item Empathy scale, the total score can range from 0 (the worst possible score) to 20 (a perfect score). On this test, any score below 20 is a failing grade, because it indicates some type of problem in the therapeutic alliance. For example, the patient may not completely trust the therapist, or may not feel understood.

Many therapists become upset when they receive failing grades on the Evaluation of Therapy Session. However, a therapeutic failure is nearly always an opportunity in disguise. If you are willing to be humble and process the feedback with your patients non-defensively and skillfully, the information can lead to a profound deepening of the therapeutic relationship and open the door to therapeutic breakthroughs.

Of course, no test is foolproof, and you always have to use your therapeutic wisdom and shrewdness when interpreting the scores. For example, if the patient is applying for disability or plans to use the information in a lawsuit, he or she may exaggerate symptoms. Or if the patients are hospitalized involuntarily in a locked unit, and they are pushing for discharge, or they are involved in child custody disputes, they may attempt to make themselves look much better than they are. But when patients are coming to us voluntarily for treatment, the scores tend to be amazingly accurate.

E = Empathy

Carl Rogers believed that a warm, accepting alliance was the necessary and sufficient condition for therapeutic change. Although his pioneering work on empathy was tremendously important, his basic hypothesis turned out to be incorrect. A therapist can be incredibly warm and empathic, and the patient may vent for months or even years without any significant change in depression, panic attacks, a troubled marriage, or an alcohol or drug addiction.

Research and clinical experience do indicate that empathy, warmth, and acceptance are necessary for change. Over 100 studies have confirmed that therapeutic empathy is associated with clinical improvement (Orlinsky, Grawe, & Parks, 1995). Patients who report that their therapists are warm and understanding seem to experience the greatest improvement, regardless of the type of therapy they are receiving.

This does not mean that empathy actually facilitates recovery. It is the old chicken vs. the egg problem. What comes first? Does therapeutic empathy actually promote therapeutic improvement and recovery, or does clinical improvement lead to increased empathy?

For example, we know that when severely depressed patients begin to recover, they feel more hopeful, loveable, and worthwhile. Therefore, they might rate their therapists as caring and understanding simply because of the increase in self-esteem. In contrast, patients who have not improved may feel that no one, including their therapists, could possibly care about them or understand them. If this were the case, empathy would still be correlated with improvement in depression, but empathy would not facilitate or lead to recovery. Empathy would simply be a sign of clinical improvement, and there would be no sound reason for therapists to develop outstanding empathy skills.

There’s also the interesting question of whether therapists actually become more empathic as their patients recover and become easier to appreciate and relate to. In this case, empathy would also be correlated with clinical recovery, but might have no causal effects whatsoever on recovery.

Usually, this type of question is resolved using the experimental method. For example, we could randomly assign severely depressed patients to two groups of therapists. One group of therapists would be warm and empathic, and the other group of therapists would have poor empathy skills. Ideally, the therapists in the two groups would be equal in all other respects—in other words, they would have equal “technical” skills. If the patients assigned to the high empathy therapists improved significantly more, we could reasonably conclude that empathy has causal effects on recovery from depression. Of course, a study like this would be unacceptable from an ethical perspective and impossible from a practical perspective.

Structural Equation Modeling can sometimes help researchers sort out these kinds of “circular causality” questions. Consider Figure 1. In this model, there are four possible reasons why therapeutic empathy would be correlated with clinical improvement:

  1. Therapeutic empathy might cause a reduction in depression. In this case, B3 would be negative and statistically significant.
  2. Depression might have direct causal effects on the patient’s perception of therapeutic empathy. In this case, B2 would be negative and statistically significant.
  3. Depression and therapeutic empathy might not be causally related, but might both be influenced by some unknown third variable. In this case, B2 and B3 would not be statistically significant, but B4 and B5 would be positive and statistically significant.
  4. Depression and therapeutic empathy might both have simultaneous direct causal effects on each other. In other words, high therapist empathy might trigger clinical improvement, which triggers an increase in the patient’s perceptions of therapeutic empathy. In this case, B2 and B3 would both be statistically significant.

Technically, the model in Figure 1 is called a “non-recursive model” because of the presence of the circular loop linking depression with empathy (B2 and B3). In other words, Therapeutic Empathy might cause a reduction in depression (B2), but depression might also cause a simultaneous reduction in the patient’s perception of therapist’s empathy (B3).

I tested these four competing models, along with my co-author, Dr. Susan Nolen-Hoeksema, in the 1990s and published the results in the top psychology journal for clinical research (Burns & Nolen-Hoeksema, 1991, 1992). We showed, for the first time, why therapeutic empathy is correlated with recovery from depression. It’s primarily because of a direct causal effect of therapist empathy on recovery from depression. If the therapist is warm and understanding, the patient will tend to recover more rapidly.

We were also able to demonstrate that depression will not have much, if any, effect on how the patient rates the therapist’s warmth and understanding. This means that severely depressed patients will not tend to perceive therapist empathy in a more negative manner than patients who feel happy and completely free of depression.

That study indicated that empathy is important, even in a highly technical form of therapy such as CBT (Burns & Nolen-Hoeksema, 1991, 1992; Burns & Spangler, 1999; Persons, Burns, & Perloff, 1988). Our finding was confirmed by a recent meta-analysis of 59 independent samples involving more than 3,000 clients treated with a variety of therapeutic orientations (Elliott, Bohart, Watson, & Greenberg, 2011). However, the size of the causal effect of empathy on depression was only modest, since empathy could only account for about a 30% reduction in depression, at most. Once that level of improvement has been achieved, nothing more can be gained from a warm and empathic relationship—other technical tools and skills will be necessary to facilitate additional improvement.

This research led to three immediate changes in the way my colleagues and I approached therapy. First, I changed clinic policies so that every patient would be required to rate his or her therapist’s empathy at the end of every session, using an Empathy Scale similar to the one I had used in the research studies. In addition, therapists were now required to review the ratings and process any feelings of dissatisfaction at the start of the subsequent session so that alliance problems could be dealt with immediately.

Second, I developed systematic empathy training techniques that all the clinic therapists practiced once a week so as to develop improved empathy skills, especially with patients who felt angry, mistrustful, or critical of their therapists. I also began including these exercises in workshops I was conducting for therapists throughout the United States and Canada.

Third, I encouraged therapists to check their egos “at the door,” both in training and in therapy sessions, since empathy failures were so common and since mastering the art of empathy is a lifelong process.

A = (Paradoxical) Agenda Setting

A therapist could measure symptoms at the start and end of every session and have superb empathy skills, but if the patient is not motivated to change, the depression, anxiety, marital conflict or addiction will persist. The Jesuit mystic, Anthony de Mello, said that we yearn for change, but cling to the familiar. Many of our patients have “one foot in the water and one foot on the shore.” On the one hand, they say they want help, but on the other hand, they dig in their heels and resist the therapist’s best efforts to help them. Treatment resistance is the cause of nearly all therapeutic failures. Unfortunately, very few therapists understand resistance and fewer yet know how to deal with resistance effectively.

In the 1970s and 1980s, there was a great deal of interest in coping skills. Research by Peter Lewinsohn at the University of Oregon indicated that simply encouraging depressed patients to engage in more productive and pleasurable activities could have significant mood-elevating effects.

To learn more about the effects of various coping skills on recovery from depression, I developed a scale called the Self-Help Inventory (SHI). The SHI listed 45 things that people typically do to cope with feelings of depression, such as going to a movie, talking to a friend, confronting a fear, or getting caught up on something on which you’ve been procrastinating. Patients were asked to rate each coping activity on three separate scales:

1. How often do you do this type of thing when you’re feeling depressed?

2. How helpful do you think it would be if you tried it?

3. How willing would you be to give this a try if suggested by a therapist or trusted friend?

The first scale was a Behavior Scale. People with high scores were the “super-copers” who actively battled negative moods, and people with low scores were more passive when depressed. There was a widespread belief among mental health professionals that “super-copers” would be ideal candidates for an active form of therapy like CBT, whereas the more passive patients would do well with a psychodynamic approach or other therapies that rely more on insight.

The second scale was a Cognitive Scale. People with high scores were optimistic about the effects of their own efforts to fight depression, and people with low scores were more pessimistic. We thought that people with high scores on this scale would also be ideal candidates for CBT, since they anticipated improvement would result from self-help activities. In addition, their optimism might trigger improvement, since the belief that something will help often reduces feelings of hopelessness and tends to work as a self-fulfilling prophecy. This is often called the “placebo effect.”

The third scale was a Willingness Scale. People with high scores were very willing to try a variety of coping behaviors to battle depression, and people with low scores were much more resistant and less willing to try to help themselves. I anticipated that individuals with high scores would be more willing to do psychotherapy homework during the treatment and would recover more rapidly than individuals with low scores.

In a cross-sectional pilot study of a group of depressed women, my colleagues and I reported that all three scales were significantly correlated with depression severity (Burns, Shaw, & Crocker, 1987). As expected, women who were more depressed indicated that they used fewer coping strategies, had more pessimistic expectations that the coping strategies would be helpful, and were substantially less willing to give them a try, even if suggested by a therapist or trusted friend.

Of course, correlations tell us little about causal relationships, and only causal relationships hold the key to valid understanding and therapeutic innovation. Therefore, I initiated a longitudinal study of several hundred patients seeking treatment at my clinic in Philadelphia so I could find out if any of the three coping scales actually played a role in the therapeutic process. In these studies, I administered the Self-Help Inventory along with two measures of depression (the Beck Depression Inventory and the depression subscale of the Hopkins Symptom Checklist-90) at the initial evaluation and tested the patients 12 weeks later, using the same scales.

Once again, all three scales on the SHI were significantly correlated with initial depression. Which scale or scales, if any, actually predicted changes in depression? Put your best guess here before you continue reading.

Did you make your best guess? No? You’re just reading? I thought so! Please ask yourself which group recovered more rapidly? Were they the super-copers, the true believers, or the patients who were willing to try? Or, perhaps they all recovered at about the same rate. What’s your hunch?


Keep in mind that if a scale predicts changes in depression, then one can fairly persuasively argue that the scale, or something that’s correlated with the scale, has causal effects on depression. Since scientists still do not know the cause, or causes, of depression, we’re after some pretty big fish.

The Behavioral Scale did not predict changes in depression. Contrary to expectations, patients who were “super-copers” did not recover more rapidly than patients who were more passive at the initial evaluation. This finding was not consistent with the idea that super-copers were better candidates for CBT or that more passive individuals would be poor candidates for CBT. Both groups recovered at exactly the same rate. This finding also suggests that the passive behavior and “do-nothingism” that we see in depressed patients may simply be a symptom, rather than a cause, of the depression.

The Cognitive Scale did not predict changes in depression either. People who had a strong belief that coping behaviors would help did not recover more rapidly than people who were very pessimistic at the initial evaluation. This finding was not consistent with the widely held belief that highly optimistic patients are the ones who recovery more quickly when treated with CBT. This finding suggests that the negative expectations of depressed patients are also likely to be symptoms, and not causes, of depression.

One of the things I like about research is that our most cherished beliefs can usually be shown to be false. If you’re open to where the science guides you, this can be enormously exciting. But if you have fixed beliefs about how things work, negative results can be deeply disturbing. How about the Willingness Scale (WS)? This scale did have causal effects on recovery from depression, and the magnitude of the effect was large. Patients who indicated that they were more willing to try a variety of coping activities recovered much more rapidly than patients with low scores on the WS at the initial evaluation.

In addition, the effect of the Willingness Scale on changes in depression appeared to be mediated by patients’ compliance with psychotherapy homework assignments. Patients with higher WS scores at the initial evaluation did more psychotherapy homework, which in turn led to changes in depression (Burns & Nolen-Hoeksema, 1991, 1992). This result is exciting because of the rather profound theoretical and practical implications about the etiology and treatment of depression.

Several years later, I replicated these findings in a second study with an entirely new group of patients, and published the findings in the same journal (Burns & Spangler, 2000). Subsequently, other investigators reported that scores on the Willingness Scale also predicted improvement in an outpatient self-help group for depression modeled after my book, Ten Days to Self-Esteem (Kazantzis, Deane, & Ronan, 2000).

The research with the WS echoed what I was observing clinically. Patients who were willing to work hard during sessions and between sessions usually recovered quickly. In contrast, patients who refused or forgot to do their homework between sessions recovered much more slowly, if at all, and many of them dropped out of therapy prematurely.

One of the shortcomings of the SHI was its extreme length—patients had to complete 135 ratings of coping behaviors. This limited the use of the SHI in real-world settings. Therefore, I modified the scale in several ways. First, I eliminated the Behavioral and Cognitive Scales since they did not predict improvement. I also reduced the length of the WS from 45 coping activities to just 10 and improved the response options. Now I had a scale that patients could easily complete and score in less than one minute—but would the new scale be reliable and valid?

To find out, I administered the brief Willingness Scale to two groups of consecutively admitted patients at the Stanford Hospital psychiatric inpatient unit (total N = 160) at the time of admission along with several scales to assess depression severity. The reliability of the brief scale was excellent: 90% or more in both cohorts of patients. We tested the patients again roughly one week later, prior to discharge from the hospital. Once again, patients with high scores on the WS at the time of admission improved substantially, but the patients with low scores on the WS improved very little, if at all.

This was surprising, since the treatment on the inpatient unit is heavily biological, with a primary focus on medications and electroconvulsive therapy. The findings indicate that the association between motivation and recovery from depression is not related to the treatment methodology. The findings also indicate that even in an inpatient setting, one of the major mechanisms of recovery may result from psychosocial, rather than biological, variables.

The WS is the first variable I am aware of that has been shown to have robust, consistent effects on recovery from depression in inpatient and outpatient samples, regardless of the treatment method that is employed. We are now using an even shorter, 5-item version of the WS clinically, since the scores can help clinicians quickly identify new patients who are likely to recover rapidly and those who are likely to resist and recover more slowly. We believe that the latter group will benefit from new interventions we have developed which are designed to boost patient motivation and reduce resistance.

Based on these studies, I radically changed the approach to treatment that I’d been using for years, and developed the Paradoxical Agenda Setting (PAS) techniques that are now at the heart of T.E.A.M. Therapy. The PAS methods are beyond the scope of this document, (see Burns, 2005, for more information), but the basic idea is to melt away the patient’s resistance prior to using any techniques to help patients with their depression, anxiety, relationship conflicts, or habits and addictions.

To facilitate this process, I developed the list of the eight most common patterns of therapeutic resistance, as along with many techniques to reverse each pattern. Certainly more research is needed, but from a clinical perspective, these PAS techniques appear to have revolutionized treatment and have produced high-speed recovery for many patients who had failed to make significant progress after years of more conventional therapy, including “Beckian” CBT.

My colleagues and I are convinced that failures of Agenda Setting represent the most common cause of therapeutic failure. We also believe that the vast majority of therapeutic failures can be quickly and dramatically reversed with the skillful and compassionate use of the PAS techniques I’ve developed.

M = Methods

Over the years, I’ve developed many techniques to help patients challenge and defeat the negative thoughts that trigger depression, hopelessness, and low self-esteem, as well as fears, phobias, anxiety, and feelings of panic. My treatment philosophy has always been to “fail as fast as you can,” since you can never predict what technique will work for what patient or problem.

Although the concept of “failing as fast as you can” might sound negative, it’s actually quite positive, because it rests on the assumption that there is a method that will help each patient. Therefore, the more methods you try, the sooner you will find the one that works. A trained T.E.A.M. therapist can easily try and fail with four, five, or even more techniques each session. Let’s assume that a patient with chronic and severe depression has the thought, “I’m defective at the core.” As long as he continues to believe that thought, he will feel worthless, hopeless, ashamed, inadequate and angry, to name just a few of the negative feelings that are typically triggered by that type of thought. The moment he stops believing that thought, he will immediately feel better. But that’s far easier said than done, because this thought is rooted in deeply-held beliefs about himself. And there’s no doubt that family members, friends, and therapists, have tried hard to convince him that he really is worthwhile—but these efforts have not been effective.

If we want to succeed where others have failed, we’ll have to use many powerful and innovative techniques that no one has used before. For example, we might try Identify the Distortions, the Individual Downward Arrow, the Pleasure Predicting Sheet, Examine the Evidence, the Paradoxical Double Standard, Externalization of Voices, Feared Fantasy, Acceptance Paradox, Paradoxical Cost-Benefit Analysis, and Semantic Method, to name just a few. Most of these techniques will not be effective, and he’ll still believe he’s “defective at the core.”

Let’s assume that the sixteenth technique works, and the patient suddenly stops believing that she or he is defective, and begins to feel a lot better. If you can fail four or five times each session, how long will it take before he recovers? The math is simple: dramatic improvement will probably occur within three or four sessions. But if you only have a small number of techniques in your therapeutic arsenal, and you resolutely continue to use them over and over, then the treatment may drag on for months or years without any tangible improvement.

What is the evidence, if any, that the therapy methods my colleagues and I have been developing are actually effective? This is not a trivial question. Some investigators have suggested that most forms of psychotherapy may only be effective because of non-specific factors, such as the therapeutic alliance, or placebo effects, as opposed to specific techniques that are unique to each school of therapy (Messer & Wamopld, 2002).

In 1985, I collaborated in a study with one of my former students, Dr. Jackie Persons, to see if we could find out how CBT actually works (Persons & Burns, 1985). We examined changes in mood during individual psychotherapy sessions in a group of patients from our clinical practices, and examined two variables of interest: 1) the patient’s rating of therapeutic empathy during the session; and 2) the reductions in the patient’s beliefs in his or her distorted negative thoughts during the session. Did either variable account for the improvement in mood, or the failure of the patient’s mood to improve, during sessions?

Together, empathy and the reduction in negative thoughts accounted for an almost unbelievable 85% of the variance in depression scores at the end of the session, when controlling for depression severity at the start of the session. In addition, the relationship variable (empathy) and the technical variable (reductions in the belief in the distorted negative thoughts) made additive and independent contributions to the improvements in depression severity. I believe this was the first demonstration that cognitive techniques actually work by changing negative thinking patterns with specific techniques, such as Externalization of Voices, but the study also highlighted the importance of trust, warmth, and understanding.

This is one of the main reasons that we measure symptoms at the start and end of every therapy session. We’re looking for large, obvious effects—dramatic reductions in symptoms within a single session—and not just placebo effects. So are we really able to achieve this, or are we fooling ourselves?

A complicating factor is that T.E.A.M. therapists emphasize the importance of psychotherapy homework as well as hard work in sessions, so during the week, patients practice and master the techniques that their therapists have introduced during therapy sessions. Although a number of studies have reported significant correlations between psychotherapy homework and recovery from depression, we are again left with the chicken vs. the egg problem. What comes first, willingness to do psychotherapy homework, or clinical improvement?

This problem is similar to the conceptual and statistical problem we encountered when evaluating the effects of therapeutic empathy on recovery. Here’s why. A loss of motivation is one of the classic symptoms of depression, so we might anticipate that more severely depressed individuals would be less likely to complete psychotherapy homework (HW) assignments between sessions. As patients improve, we would anticipate a boost in motivation should lead to an increase in HW compliance. If this were the case, the correlation between HW compliance and clinical improvement might result from a causal effect of depression on HW, and not vice versa. Then there would be no reason to assign psychotherapy HW, since it would not actually speed up the recovery process.

This, of course, is another problem of circular causality, since there are several possible causal pathways linking depression with HW compliance:

  1. HW compliance might reduce depression severity and facilitate recovery, just as cognitive and behavioral therapists suggest.
  2. Clinical improvement might lead to increases in HW compliance.
  3. HW and clinical improvement might have no causal linkages, and might both result from some known or unknown third variable, such as motivation, with simultaneous causal effects on depression as well as HW compliance.
  4. HW compliance and clinical improvement might be linked in a system of simultaneous circular causality.

One could use an experimental strategy to attack this problem, but since I’m a clinician in private practice, I can’t randomly assign my patients to HW vs. no HW groups. Instead, I used the same non-recursive structural equation modeling techniques I described earlier to examine the causal linkages between homework compliance and recovery from depression in several hundred patients treated at my clinic in Philadelphia during the first 12 weeks of therapy. Here’s what we found (Burns & Nolen-Hoeksema, 1991, 1992):

  1. HW compliance and recovery from depression were significantly correlated, but the magnitude of the correlation was small. As expected, patients who did the most HW also reported the greatest reductions in depression.
  2. The correlation between HW and changes in depression resulted from a direct causal effect of HW compliance on changes in depression.
  3. The magnitude of this effect was large. In fact, nearly all the patients who consistently did HW recovered and nearly all the patients who refused to do HW failed to improve. Many of these patients deteriorated or dropped out of therapy prematurely.
  4. Depression had a reciprocal causal effect on HW compliance, but the magnitude of this effect was so small as to be trivial. For the most part, severely depressed patients did as much HW as patients who were minimally depressed, or not at all depressed.
  5. Psychotherapy HW compliance was associated with motivation, as measured by the Willingness Scale (WS) at the initial evaluation. In fact, much, if not all of the effect of the WS on recovery from depression was mediated by psychotherapy HW compliance.

Once the vital importance of psychotherapy HW became clear to me, I changed clinic policies in several respects. First, I developed a “Concept of Self-Help Memo” that we mailed to every prospective patient prior to the first therapy session. The memo describes the importance of HW and lists ten of the most common types of HW patients are asked to do, such as recording negative thoughts on the Daily Mood Log, scheduling more pleasurable and rewarding activities on the Pleasure Predicting Sheet, or recording a problematic interaction with a friend or family member on the Relationship Journal.

The memo also includes a “Self-Help Contract” patients must complete and sign, indicating whether they agree to do the HW, and if so, how many minutes per day, and how many days per week they are willing to do psychotherapy HW. The memo also includes a list of “25 GOOD Reasons NOT to do the Psychotherapy Homework.” This list includes all of the most compelling excuses patients have given over the years for not doing the HW—such as feeling hopeless, overwhelmed, angry, unmotivated, and so forth. Patients are asked to indicate which reasons most reflect their own feelings.

I trained the staff to negotiate HW compliance with each new patient, after reviewing the memo together, in a paradoxical manner. Essentially, we told individuals with mixed feelings about HW that they could not be accepted as patients at the clinic because our methods were not effective in the absence of the HW. However, we would be happy to provide a referral list to local clinicians who do not require psychotherapy HW if they would prefer a different type of treatment. We also told the patients that we did not want to get rid of them, and hoped they would decide to work with us, but if they did, they should know that the issue of psychotherapy HW was not negotiable.

We were initially concerned that we might lose many patients because of this method, but surprisingly, it seemed to have the opposite effect of boosting patients’ determination to work with us. Out of a group of 700 consecutive patients, fewer than ten declined treatment or requested referral outside of the clinic because of the HW requirement.

Bibliotherapy Research

“Bibliotherapy” (reading from a self-help book like Feeling Good between sessions) has always been an essential part of the therapy I’ve done, and is a formal component of the T.E.A.M. Treatment Model. However, the book the patient reads depends on the type of problem she or he is struggling with. Feeling Good focuses primarily on depression. The Feeling Good Handbook has a much broader range of topics, including depression, anxiety, and relationship problems. When Panic Attacks covers anxiety disorders, and Feeling Good Together brings the interpersonal model and methods to life. And, of course, there are many additional self-help books on the market that therapists routinely “prescribe” for patients. In the late 1980s, Dr. Forrest Scogin, a research psychologist at the University of Alabama, conducted a remarkable series of studies designed to answer two questions: 1. What’s the least expensive way to treat patients suffering from clinical depression? 2. Do self-help books actually help?

In his first study, Dr. Scogin and his colleagues randomly assigned 80 senior citizens seeking treatment for major depressive episodes at the University of Alabama to one of two groups. Patients in the both groups were told they would be placed on a four-week waiting list before beginning treatment. Patients in one group were given a self-help book (either my Feeling Good or Peter Lewinsohn’s Control Your Depression) and asked to read it while waiting for treatment. Patients in the second group were not given any book to read while waiting for treatment.

Every week a research assistant contacted the patients in both groups and administered a depression test to assess the severity of symptoms. At the end of the four weeks, the two groups were compared. Surprisingly, two-thirds of the patients who received a self-help book had improved substantially or recovered. In fact, these patients did not need any further treatment at the medical center. In contrast, there was little or no significant improvement in the patients who did not receive a book. To the best of my knowledge, this was the first scientific study to show that at least some self-help books can have antidepressant effects (Scogin, Hamblin, & Beutler, 1987).

In the next phase of the study, Dr. Scogin and his colleagues told the patients who had not received a book that they would have to wait four more weeks before beginning treatment, but he now gave them one of the two self-help books and asked them to read it while waiting. Once again, two-thirds of these patients also recovered, just like the patients in the first group who had received a self-help book at the beginning of the study. This study provided strong evidence that the rather strong beneficial effects of the “bibliotherapy” were probably not due to chance (Scogin, Hamblin, & Beutler, 1987).

The results suggested that self-help books actually had antidepressant effects comparable to the effects of individual psychotherapy or treatment with antidepressant medications, but the effects appeared more rapid, since most outcome studies of drugs or psychotherapy last 12 to 16 weeks, and not just four weeks. Given the tremendous pressure to cut health-care costs, these findings are of considerable interest, since a mass-market paperback book costs less than one day of antidepressant drug treatment,and is presumably free of troublesome side effects as well.

The researchers still could not be certain that either book had any specific antidepressant effects, since reading any book might have a comparable “placebo effect.” To evaluate this possibility, Dr. Scogin and his colleagues conducted a new study with a slightly different experimental design. Once again, they randomly assigned a group of depressed patients to one of two groups, and both groups were told they would have to wait four weeks before beginning treatment. However, the patients in one group were given my book, Feeling Good, and the patients in the other group were given a “placebo book” that was not expected to have any antidepressant effects. The book they received was Victor Frankl’s classic book entitled Man’s Search for Meaning. If the patients in both groups improved to a comparable degree, it would show that reading could be helpful if you are depressed, but it might not make any difference what you read.

The researchers reported that the patients who received Victor Frankl’s book did not improve significantly, but the patients who read Feeling Good did. This indicated that the improvement from Feeling Good was real and not just a “placebo effect” (Scogin, Jamison, & Gochneaut, 1989).

Finally, the researchers addressed another important concern: would the antidepressant effects last? Motivational speakers can get a crowd of people fired up and feeling optimistic for brief periods of time, but these brief mood-elevating effects may not last. The same problem holds for the treatment of depression with psychotherapy or medications. After a period of time, many patients who initially improved relapse back into depression. These relapses can be devastating and sometimes even more difficult to treat than the initial episodes because patients feel so demoralized and hopeless.

To find out if the effects of Feeling Good would last, Dr. Scogin and his colleagues conducted several follow-up studies. The initial follow-up study was for a two-year period (Floyd, Rohen, Shackelford, et al., 2006; Scogin, Jamison, & Davis, 1990), and the next follow-up study was for a three-year period (Smith, Floyd, Jamison, & Scogin, 1997). These studies revealed that the patients who initially improved after reading Feeling Good did not relapse, but continued to improve. Many indicated that when they were upset, they simply opened up Feeling Good and re-read the section that had been the most helpful initially. The researchers concluded that these self-administered “booster sessions” were important in maintaining a positive outlook following recovery. The public health implications are considerable, since these follow-up bibliotherapy booster sessions were free.

Since his pioneering studies, there have been additional studies of bibliotherapy in a variety of settings and for a variety of patient populations of all ages. Taken as a whole, the results indicate that bibliotherapy can be almost as good, if not better, than the results obtained with antidepressant medications or psychotherapy in controlled outcome studies (Ackerson, Scogin, Lyman, & Smith, 1998; Floyd, Scogin, McKendree-Smith, et al., 2004; Jamison & Scogin, 1995; Mains & Scogin, 2003; McKendree-Smith, Floyd, Scogin, 2003). These studies are somewhat surprising and very encouraging.

Because of this research, the use of self-help books to supplement therapy has become commonplace. Surveys indicate that most therapists now “prescribe” bibliotherapy for patients they are treating. I am very proud that surveys of North American mental health professionals indicate that Feeling Good is the most highly rated and commonly recommended book for patients struggling with depression (Santrock, Minnett & Campbell, 1994).

However, it is important to emphasize that no treatment, including bibliotherapy, is a panacea—some patients will respond and others will not. Once again, that’s a good reason not to fall in love with any one method of treatment, but rather to use a variety of approaches and techniques. In addition, my colleagues and I are not using bibliotherapy as a stand-alone treatment, but simply as one component of our T.E.A.M. Therapy model. Presumably, seeing a skillful and compassionate therapist while reading the book should be significantly more effective than simply reading the book on one’s own, but I am not aware of any research that has addressed this important question as of yet.

Final Comments

In the late-1980s, Dr. Jackie Persons and I published the first uncontrolled outcome study of CBT in private practice outpatient settings working with depressed patients with extensive Axis I and Axis II comorbidity, and compared the efficacy with controlled outcome studies of CBT with highly rarified, homogenous populations of patients with major depressive disorder and little or no comorbidity (Persons, Burns, & Perloff, 1988). This work shed a favorable light on the efficacy of CBT in private practice settings. Of course, the T.E.A.M. treatment model goes far beyond CBT. Those early studies also highlighted the importance of psychotherapy HW compliance as one crucial component of effective treatment.

In the early 1990s, Dr. Susan Nolen-Hoeksema and I published much more extensive psychotherapy outcome and process studies based on several hundred patients treated at my clinic in Philadelphia, and I’ve alluded to some of our findings on Empathy and psychotherapy HW earlier in this manuscript. But there were other interesting findings, including the fact that depressed patients with severe Borderline Personality Disorder (BPD) recovered almost as quickly as depressed patients without BPD (Burns & Nolen-Hoeksema, 1991; 1992).

Approximately 20% to 30% of our patients were diagnosed with BPD at the initial evaluation, usually in conjunction with many other diagnoses (Burns & Nolen-Hoeksema, 1991). In fact, the patients with BPD played a crucial role in the development of all four aspects of the T.E.A.M. model, including testing at each session, empathy, paradoxical agenda setting, and the use of many methods, rather than just a few, to help the more difficult and challenging patient.


Ackerson, J., Scogin, F., Lyman, R.D., & Smith, N. (1998). Cognitive bibliotherapy for mild and moderate adolescent depressive symptomatology. Journal of Consulting and Clinical Psychology, 66, 685-690.

Antonuccio, D.O., Danton, W.G., & DeNelsky, G.Y. (1995). Psychotherapy versus medication for depression: Challenging the conventional wisdom with data. Professional Psychology: Research and Practice, 26, 574 – 585.

Antonuccio, D.O., Burns, D., & Danton, W.G. (2002). Antidepressants: A triumph of marketing over science? Prevention and Treatment, 5, Article 25. Web link:

Antonuccio, D.O., Danton, W.G., DeNelsky, G.Y., Greenberg, R., & Gordon, J.S. (1999). Raising questions about antidepressants. Psychotherapy and Psychosomatics, 68, 3-14.

Burns, D. D. (2005). Tools, Not Schools, of Therapy. Strategies for Therapeutic Success. (also called the Psychotherapy eBook). Los Altos Hills, CA: author.

Burns, D. D., & Auerbach, A. (1996). Therapeutic Empathy in Cognitive-Behavioral Therapy: Does it Really Make a Difference? Chapter 7 in Frontiers of Cognitive Therapy (P. Salkovskis, ed.) New York: Guilford Press, pp. 135 – 164.

Burns, D.D. & Nolen-Hoeksema, S. (1991). Coping styles, homework compliance and the effectiveness of cognitive-behavioral therapy. Journal of Consulting and Clinical Psychology, 59(2), 305 – 311.

Burns, D. D., & Nolen-Hoeksema, S. (1992). Therapeutic empathy and recovery from depression in cognitive-behavioral therapy: A structural equation model. Journal of Consulting and Clinical Psychology, 60(3), 441 – 449.

Burns, D. D., Shaw, B. F., & Crocker, W. (1987). Thinking styles and coping strategies of depressed women: An empirical investigation. Behavior Research and Therapy, 25(3): 223 – 225.

Burns, D.D., & Spangler, D. (2000). Does psychotherapy homework lead to improvements in depression in cognitive behavioral therapy? Or does improvement lead to increased homework compliance? Journal of Consulting and Clinical Psychology, 68(1), 46 – 59.

Burns, D.D., Westra, H., & Trockel, M. (2011). Motivation and Changes in Depression in Two Inpatient Cohorts. Manuscript under review.

DeRubeis, R. J., Hollon, S. D. Amsterdam, J. D., Shelton, R. C., Young, P. R., Salomon, R. M., O’Reardon, J. P., Lovett, M. L., Gladis, M. M., Brown, L. L., & Gallop, R. (2005). Cognitive Therapy vs. Medications in the Treatment of Moderate to Severe Depression. Archives of General Psychiatry, 62: 409-416. Web abstract:

Dobson, K.S. (1989). A meta-analysis of the efficacy of cognitive therapy for depression. Journal of Consulting and Clinical Psychology, 57(3), 414 – 419.

Elliott R, Bohart AC, Watson JC, & Greenberg LS. (2011). Empathy. Psychotherapy (Chic). Mar;48(1):43-9.

Floyd M, Rohen N, Shackelford JA, Hubbard KL, Parnell MB, et al. (2006) Two-year follow-up of bibliotherapy and individual cognitive therapy for depressed older adults. Behavior Modification, 30: 281-294.

Floyd M, Scogin F, McKendree-Smith NL, Floyd DL, Rokke PD (2004) Cognitive therapy for depression: a comparison of individual psychotherapy and bibliotherapy for depressed older adults. Behavior Modification,28: 297-318.

Garland, E. J. (2004). Facing the evidence: antidepressant treatment in children and adolescents. Canadian Medical Association Journal, 170, 489-491.

Hatcher, R. L., Barends, A., Hansell, J. & Gutfreund, M.J. (1995). Patients’ and therapists’ shared and unique views of the therapeutic alliance: An investigation using confirmatory factory analysis in a nested design. Journal of Consulting and Clinical Psychology, 63(4), 636 – 643.

Healy, D. (2003). Lines of evidence on the risk of suicide with selective serotonin reuptake inhibitors. Psychotherapy and Psychosomatics. 72, 71-79.

Hollon, S.D., & Beck, A.T. (1994). Cognitive and cognitive behavioral therapies. Chapter 10 in A. E. Bergin & S. L. Garfield (Eds.), Handbook of Psychotherapy and Behavioral Change (pp. 428 – 466). New York: John Wiley & Sons, Inc.

Hollon, S. D., DeRubeis, R. J., Shelton, R. C., Amsterdam, J. D., Salomon, R. M., O’Reardon, J. P., Lovett, Young, P. R., Haman, K. L., Freeman, B. B., & Gallop, R. (2005). Prevention of Relapse Following Cognitive Therapy vs. Medications in Moderate to Severe Depression. Archives of General Psychiatry, 62: 417-422. Web abstract:

A brief summary with an interview appeared in Medical News Today on Friday, July 8, 2005: Cognitive therapy as good as antidepressants, effects last longer. The link I had is no longer working, however.

Hypericum depression trial study group. (2002). Effect of Hypericum perforatum (St. John’s wort) in major depressive disorder: A randomized, controlled trial. Journal of the American Medical Association, 287, 1807-14. You can read a summary online at: A full text of this article is available at:

Jamison, C., and Scogin, F. (1995). Outcome of cognitive bibliotherapy with depressed adults. Journal of Consulting and Clinical Psychology, 63, 644 – 650.

Jureidini, N., Doecke, C.J., Mansfield, P.R., Haby, M.M., Menkes, D.B., & Tonkin, A.L. (2004) Efficacy and safety of antidepressants in children and adolescents, British Medical Journal, 328, 879-883.

Kazantzis, N., Deane, F.P., & Ronan, K.R. (2000). Homework assignments in cognitive and behavioral therapy: A meta-analysis. Clinical Psychology: Science and Practice, 7, 189–202.

Khan A, Khan SR, Leventhal. RM, Brown WA (2001).  Symptom reduction and suicide risk in patients treated with placebo in antidepressant clinical trials: a replication analysis of the Food and Drug Administration Database. International Journal of Neuropsychopharmacology, 4, 113-118.

Khan, A., Warner, H. A., & Brown, W. A. (2000). Symptom reduction and suicide risk in patients treated with placebo in antidepressant clinical trials: An analysis of the Food and Drug Administration database. Archives of General Psychiatry 57, 311-317.

Kirsch, I. & Sapirstein, G. (1998). Listening to Prozac but Hearing Placebo: A Meta-Analysis of Antidepressant Medication. Prevention and Treatment, 1, Article 0002a.

Kirsch, I., Moore, T. J., Scoboria, A., & Nicholls, S. S. (2002). The Emperor’s New Drugs: An Analysis of Antidepressant Medication Data Submitted to the U.S. Food and Drug Administration. Prevention and Treatment, 5:23. Web link:

Kirsch, Irving. (2010). The Emperor’s New Drugs: Exploding the Antidepressant Myth. New York: Basic Books.

Krupnick, J. L. et al. (1993). The role of the therapeutic alliance in psychotherapy and psychotherapy outcome: Findings in the NIMG treatment of depression collaborative research program. Journal of Consulting and Clinical Psychology, 64(3),636 – 643.

Mains JA, Scogin FR (2003) The effectiveness of self-administered treatments: a practice-friendly review of the research. Journal of Clinical Psychology, 59: 237-246.

McKendree-Smith NL, Floyd M, Scogin FR (2003) Self-administered treatments for depression: a review. Journal of Clinical Psychology, 59: 275-288.

Messer, S.B., & Wampold, B.E. (2002). Let’s face facts: Common factors are more potent than specific therapy ingredients. Clinical Psychology: Science and Practice, 9(1), 21-25.

Norcross, J. C., Santrock, J. W., Campbell, L. F., Smith, T. P., Sommer, R., & Zuckerman, E. L. (2003). Authoritative Guide to Self-Help Resources in Mental Health, Revised Edition. New York: Guilford Press.

Orlinsky, D. E., Grawe, K., & Parks, B. K. (1995). Process and outcome in psychotherapy–Noch einmal. Chapter 8 in A. E. Bergin & S. L. Garfield (Eds.), Handbook of Psychotherapy and Behavioral Change (pp. 270 – 376). New York: John Wiley & Sons, Inc.

Persons, J. B., & Burns, D. D. (1985). Mechanism of action of cognitive therapy: Relative contribution of technical and interpersonal interventions. Cognitive Therapy and Research, 9(5), 539 – 551.

Persons, J.B., Burns, D.D., & Perloff, J.M. (1988). Predictors of dropout and outcome in cognitive therapy for depression in a private practice setting. Cognitive Therapy and Research, 12, 557 – 575.

Robinson, L.A., Berman, J.S., & Neimeyer, R.A. (1990). Psychotherapy for the treatment of depression: A comprehensive review of controlled outcome research. Psychological Bulletin, 108, 30 – 49.

Santrock, J. W., Minnett, A. M., & Campbell, B. D. (1994). The Authoritative Guide to Self – Help Books. New York: Guilford Press.

Scogin, F., Hamblin, D., and Beutler, L. (1987). Bibliotherapy for depressed older adults: A self-help alternative. The Gerontologist, 27, 383 – 387

Scogin, F., Jamison, C., and Davis, N. (1990). A two-year follow-up of the effects of bibliotherapy for depressed older adults. Journal of Consulting and Clinical Psychology, 58, 665 – 667.

Scogin, F., Jamison, C., and Gochneaut, K. (1989). The comparative efficacy of cognitive and behavioral bibliotherapy for mildly and moderately depressed older adults. Journal of Consulting and Clinical Psychology, 57, 403 – 407.

Smith, N. M., Floyd, M. R., Jamison, C., and Scogin, F. (1997). Three-year follow-up of bibliotherapy for depression. Journal of Consulting and Clinical Psychology, 65(2), 324 – 327.

Whittington, C.J., Kendall, T., Fonagy, P., Cottrell, D, Cotgrove, A, & Boddington, E. (2004). Selective serotonin reuptake inhibitors in childhood depression: systematic review of published versus unpublished data. The Lancet, 363, 1341-1345.

[*]     Copyright © 2011 by David D. Burns, MD. Do not cite in whole or in part without the express written permission of Dr. Burns.

*     Copyright ã 2004 by David D. Burns, M.D. Reproduction or download is prohibited. If you would like to use scales like these in your clinical work, contact for ordering information for the Therapist’s Toolkit/

*     Copyright © 2001 by David D. Burns, M.D. Revised, 2004. Reproduction or download is strictly prohibited.