This document is an update of McKeel 1996 ('A clinician's guide to research on solution-focused therapy', in Miller, S.D., Hubble, M.A. and Duncan, B.L. (eds), Handbook of Solution-Focused Brief Therapy (pp. 251-71). San Francisco: Jossey-Bass.). It is copyright and all rights are reserved to the author. It may not be quoted without appropriate citation. For permission to reproduce it in whole or in part please contact the author at firstname.lastname@example.org
selected review of
A. Jay McKeel
Unfortunately, many research consumers will decide that the research of SFBT does not meet their needs. Few outcome studies of SFBT use experimental, quantitative outcome research that can provide some readers with convincing results, and few studies compare SFBT with other therapeutic models. Studies of SFBT techniques can ask more useful questions and use multiple measures to investigate the effectiveness of the interventions. More research asking clients about their experience of SFBT and specific techniques would be welcomed by many clinicians and research consumers.
This review has four parts. First is a review of outcome studies of SFBT. Next, investigations of SFBT techniques are presented. The third section describes information from and about clients about their experience of SFBT. The last section offers ideas for new research of SFBT. Finally, this article offers suggestions for investigating SFBT that many research consumers will find more useful.
brief therapy effective?
Interestingly, most clients begin therapy expecting it to be brief (Garfield, 1994; Koss & Shiang, 1994). Therapists, on the other hand, tend to prefer longer-term therapy (Warner, 1996; Pekarik, 1991).
Research also suggests that SFBT is an effective treatment for a broad range of client problems. De Jong and Berg (1998) report SFBT accomplished 70% or better success rates for many clinical problems, including depression, suicidal thoughts, sleep problems, eating disorders, parent-child conflict, marital/relationship problems, sexual problems, sexual abuse, family violence, and self-esteem problems.
Other research studies have examined the effectiveness of SFBT with specific client populations. For example, de Shazer & Isebaert (1997) report a 74% success rate for 250 inpatient clients and a 73.5% success rate for 72 outpatient clients with alcohol problems. An outcome study by Eakes, Walsh, Markowski, Cain, & Swanson (1997) found SFBT successful in treating five clients with a diagnosis of schizophrenia.
Three research studies have examined whether SFBT is effective in group therapy. These studies conclude that SFBT group therapy successfully addresses couples issues (Zimmerman, Prest, & Wetzel, 1997), parenting issues (Zimmerman, Jacobson, MacIntyre, & Watson, 1996), and anger reduction (Schorr, 1997).
Other research studies have found that SFBT is effective in an Occupation Therapy setting (Cockburn, Thomas, & Cockburn, 1997), in schools (Lafountain, Garner, & Eliason, 1996; Littrell, Malia, & Vanderwood, 1995), in Social Work Agencies (Sudman, 1997), and in prisons (Lindforss & Magnusson, 1997).
The number of sessions of SFBT that a client attends and whether that client accomplishes their treatment goals seem related. In Macdonald's (1994) follow-up study, clients who reported a good outcome had an average of 5.47 sessions, while those who reported their situation was the same or worse than before therapy attended an average of 2.67 sessions. Another study found that clients who received three sessions or less of SFBT had a success rate of 69.4% while clients attending four or more sessions had a success rate of 91.1% (Kiser, 1988; Kiser & Nunnally, 1990).
SFBT better than other therapy models?
The Lindforss & Magnusson (1997) is a good model for comparative outcome studies. Sixty prisoners were randomly assigned to receive either SFBT or the existing treatment (control group) provided to inmates. The outcome measure was re-arrest (recidivism) during two follow-up periods: 12 months after the inmate was released and again at 18 months. At 18 months, 40% of the inmates who received SFBT had not been re-arrested while only 14% of the control group had not been re-arrested, which was a statistically significant difference. The SFBT group had fewer arrests in the 18 months following their release from prison (86 vs. 153) and the SFBT group spent less time in prison for those arrests (86 vs. 136).
Consumers of the Lindforss & Magnusson (1997) research include prison administrators and clinicians who work with prisoners. These investigators compared SFBT with a treatment that these consumers were already using; without this comparison, these clinicians and administrators would have no convincing evidence that SFBT produced better results. Also, the investigators selected an outcome measure that is meaningful to these consumers: recidivism. This type of research design and outcome measures produces results that will interest research consumers and lead to improvements in services therapists provide.
Early in the development of SFBT the Milwaukee team used research to discover that when therapists use solution-talk, clients usually respond by talking about improvements in their situation or view of their situation (Gingerich, de Shazer, and Weiner-Davis, 1988). This study revealed the BFTC team rarely used solution-talk early in their first sessions; because of this research project de Shazer and the team to begin asking about change earlier in the first session.
Another study found that the more a client uses solution-talk in his/her first session, the more likely the client is to continue their therapy. Also, the more clients talk about solutions or goals in their first session, the more likely they are to complete treatment rather than drop out (Shields, Sprenkle, and Constantine, 1991).
Research shows that pretreatment improvement is common. Allgood, Parham, Salts, & Smith (1995) found that 30% of 200 clients reported pretreatment improvement in the situation that led them to seek therapy. A study reviewing data from 2400 clients found that 15% made significant improvement before attending their first session (Howard, Kopta, Krause, & Orlinsky, 1986).
How does pretreatment change affect the process and outcome of therapy? One study found that clients who report pretreatment change are four times more likely to complete therapy than clients who report no pretreatment change (Johnson, Nelson, & Allgood, 1998). Additional studies about pretreatment improvements could ask: What differences exist between clients who do and do not report pretreatment change? Are clients who report pretreatment change more optimistic about accomplishing their goals? Do clients believe doing more of what led to pretreatment improvement will help them accomplish their treatment goals?
Studies show that clients are more likely to report pretreatment improvements when asked presuppositional questions. Weiner-Davis, de Shazer, and Gingerich (1987) asked clients in their first session, "Many times people notice in between the time they make the appointment for therapy and the first session that things already seem different. What have you noticed about your situation?" (p. 360). This question presupposes pretreatment improvement has occurred. Twenty of the thirty clients asked the question reported pretreatment improvements and gave specific examples regarding the problems that led them to seek therapy.
Replications of this study also found that more than 60% of clients who were asked a similar question by their therapist in their first session identified something about their problem that was better (Lawson, 1994; McKeel & Weiner-Davis, 1995). Johnson, Nelson, & Allgood (1998) asked clients a presuppositional question about pretreatment improvement in a written questionnaire completed before the beginning therapy and 53% of these clients reported pretreatment change.
Future research of presuppositional questions may explore issues other than pretreatment change. For instance, does a presuppositional question at the beginning of the second or subsequent session (e.g., "So, what has been better since I saw you last?") increase clients' reports of between-session improvement? Other research could explore if and how presuppositional questions lead clients to develop new views about themselves and their situation.
In a study of clients receiving SFBT, clients explained that the miracle question helped them focus on their goals for treatment and helped them focus on doing something different to accomplish their goals. These clients also felt more hopeful about their situation after answering the question (Shilts, Rambo, & Hernandez, 1997).
In a study of SFBT couples' therapy, researchers interviewed clients at the end of treatment to find out if they remembered their therapist asking the miracle question. Most clients did, but few could remember how they answered. Clients were more likely to recall their spouse's answer than their own (Odel, Butler, & Dilman, 1997).
Nau (1997) observed experienced SFBT therapists conducting first sessions and discovered two factors important in effectively asking the miracle question. First, the therapist must clearly join with the client before asking the question. Second, the technique is more effective if the therapist explores exceptions with the client before asking the miracle question.
First Session Task
One study found that in their second session, 89 percent of the clients assigned the FFST reported that something positive and worthwhile had occurred since their first session and 57% reported that their situation was better (de Shazer, 1985).
Piercy, and Jurich (1991) compared a SFBT first session, which included assigning
the FFST, with a problem-focused first session using a problem-focused first session
task. At the beginning of their second session, the therapist followed up on the
task. Clients assigned the FFST were more likely to have completed the task, more
clear about their treatment goals, and more likely to report improvement in their
presenting problem. However, clients assigned the FFST were not more optimistic
about accomplishing their goals in therapy.
views of SFBT techniques
Of these four SFBT questions, therapists rated the miracle question as the most therapeutic. Scaling questions were the most frequently used and therapists rated these questions as the best way to evaluate a client's progress. Therapists described exception questions as typically leading a client to report exceptions and improvements in his/her problems and to describe what he/she did to achieve the change. Therapists rated questions about pretreatment changes as the least effective of the questions and the most difficult to use successfully use (Skidmore, 1993).
When evaluating SFBT, clients explain that their relationship with their therapist is more important than any specific technique their therapist uses. "[T]hey appreciate those therapists who are respectful and take the time to 'listen' to the families' story. Families consistently report that therapy appears most beneficial when the therapist appears caring and concerned" (Shilts, Rambo, & Hernandez, 1997, p. 129).
On the other hand, clients are critical of SFBT when they do not develop a close relationship with their therapist or do not feel heard and understood. One client noted, "We just never connected . . . he never understood our situation" and another client explained, "We were in the middle of a huge crisis and I don't think he picked up on it" (Odel, Butler, & Dielman, 1997). In their study of SFBT couples' therapy, Odell, Butler, and Dielman (1997) found that "when one member of the couple felt a lack of connection [with their therapist], therapy was over."
Next Step in Researching SFBT
This section suggests strategies researchers can use to address the needs of some SFBT research consumers. Third party payers of psychological services, trainers and educators, and many clinicians want results that well-designed quantitative outcome studies of SFBT provide. Clinicians who want practical and easy-to-gather information about improving services to their clients can use simple qualitative studies to meet their needs.
quantitative outcome research
Define SFBT. Existing outcome studies of SFBT typically fail to provide basic information about the model practiced by the therapists in the study. de Shazer and Berg (1997) urge researchers to "demonstrate that the model of therapy being tested is indeed the model used by the therapists [in the research study]. Otherwise any and all findings are suspect." (p. 123).
However, a rigid protocol of SFBT in research studies conflicts with the "do what works; if it does not work, do something different" philosophy of the model. To remedy this, studies of SFBT can encourage therapists to do what they consider is best for their clients. After treatment is finished, the investigator can review the case to determine if it meets the criteria for inclusion in a study of SFBT. While this research strategy presents some methodological problems, it has three advantages: better representing how SFBT is actually practiced, allowing the investigator to discover if and when SFBT practitioners deviate from or abandon the model, and it best meet clients needs.
Design comparative studies. Future quantitative outcome research needs to randomly assign clients to two or more treatment conditions. Almost all existing research of SFBT has no comparison group, a design that Kerlinger (1986) deems scientifically worthless.
Past useful quantitative research has used several types of comparison groups. A comparative outcome strategy compares the model under examination (the experimental group) with an established model (the control group) to determine if the new model compares favorably or produces better results than the control model.
A dismantling outcome strategy compares two groups to learn what components of the model are necessary or sufficient for treatment success. For example, research could test the importance of SFBT homework assignments in treatment outcome; one group of clients would always be assigned SFBT homework while the other would be assigned either no homework or non-SFBT homework. A constructive outcome strategy could be used to test new techniques or ideas by conducting traditional SFBT with one group while the other group receives SFBT plus the new idea. Finally, outcome studies can compare two groups based on therapist characteristics; for instance, comparing a team approach to therapy conducted by an individual therapist or comparing experienced and novice therapists (Kazdin, 1994).
Researchers debate the merits and ethics of a no-treatment control group; that is, comparing a group of clients receiving treatment with clients who receive nothing. Some argue that this type comparison group is necessary to prove that a particular model produces therapeutic improvement by determining if people who call to request therapy may be as likely to resolve their problems without any clinical help (e.g., Kazdin, 1994).
Other researchers believe using a no-treatment control group is unnecessary and unethical (e.g., Todd & Stanton, 1983). These researchers argue that no-treatment groups are no longer necessary because a sufficient body of research shows that people are significantly more likely to improve their situation if they receive therapy. More importantly, denying services may be unethical and even dangerous for many clients asking for therapy, especially when violence, suicidal thoughts, eating disorders, or substance abuse are problems.
Improve outcome measures. Some outcome studies of SFBT categorize clients who report they met "some goals" as treatment successes. Weak outcome criteria offer little information to practitioners or other consumers or audiences of research.
When researchers use multiple measures, they speak to the various consumers of research studies. Some therapists will be satisfied with relying only on clients' perceptions of their treatment while others will place more confidence in the results of a study that also includes success ratings from the therapist or other observer (this is especially true regarding the treatment of substance abuse or violence). Third-party providers may have more confidence in a model that demonstrates its effectiveness with traditional research measures and instruments.
The success of treatment can be measured in multiple ways. Client self reports can be global (my marriage is better), affective (I am happier), behavioral (we spend more time together), and cognitive (I do not worry about divorce now). Observational data can be collected from therapists and/or others (e.g., a team behind a mirror or an interviewer who meets with the client). Standardized tests are another important tools available for researchers wanting to demonstrate treatment success.
Franklin, Corcoran, Nowicki, and Streeter (1997) offer one method of collecting outcome data that may interest SFBT practitioners. Clients develop scales (0 = client fails to make any movement toward goals; 10 = client is completely satisfied). Johnson (personal communication) has also developed a method for tracking client progress based on a 1-10 rating scale which uses scaling to track clients' progress.
Variance between therapists. Research literature consistently demonstrates that some therapists are more effective than others. Future studies of SFBT should collect data about therapist characteristics, practice, and therapeutic philosophy to identify variances between therapists and to detect which variables relate to better success rates of more effective therapists. Understanding what more effective therapists do may contribute more useful information to clinicians than any other type of research results.
simple qualitative study
Our agency asks clients to complete this questionnaire before their second and subsequent sessions. Clients report this questionnaire motivates them to take more action to change and appreciate that their therapist cares about their progress and what the client considers important to do in their sessions. It helps therapists better understand what our clients want from us. The questionnaire includes the following questions:
What was most useful about your last session?
At the beginning of each session, clients give their questionnaire to their therapist who begins the sessions by responding to clients' answers.
After collecting enough responses, Bowie Youth & Family Services will review the answers to identify patterns or categories of responses. The categories the investigator develops would be guided by the answers clients gave to these questions, but examples of categories that might occur include: developing a plan of action, feeling support from the therapist, using therapy to make important decisions, and better understanding my spouse. The result can describe what clients find useful and what they want from therapy.
Clinical research should offer clinicians information that helps them better serve to their clients. Studies such as the Bowie project accomplish that goal in the short term (by providing relevant, on-the-spot information to the therapist about what the client has and will find useful) and in the long-term by reporting findings of this study to other therapists.
Studies of SFBT techniques have found that: (a) when therapists use solution-talk, their clients are more likely to talk about change. The more clients use solution-talk, the more likely they are to complete their therapy. (b) Pretreatment improvement is common. (c) Presuppositional questions usually accomplish their goal of helping clients notice strengths, improvements, and past successes. (d) The miracle question can help clients focus on their treatment goals and feel more optimistic about their situation. (e) Clients usually complete the FFST and report improvements when assigned that homework. (f) SFBT therapists frequently use scaling questions.
(g) SFBT therapists report that exception questions typically prompt clients to talk about exceptions and improvements regarding their problems. (h) Therapist find questions about pretreatment change to be the most difficult to use effectively.
This body of research also tells us that focusing solely on techniques misses one key to successful SFBT: the client-therapist relationship. The study by Odell, Butler, and Dielman (1997) highlights this: although therapists were asking the miracle question, assigning the FFST, and using other SFBT techniques, if a client did not feel their therapist heard and understood them, they left therapy. Studies of the perceptions of clients highlight the importance of therapists' relationship with their clients and that SFBT is successful when it is both solution-oriented and client-oriented.
Allgood, S. M., Parham, K. B., Salts, C. J., & Smith, T. A. (1995). The association between pretreatment change and unplanned termination in family therapy. The American Journal of Family Therapy, 23,
Bowie Youth & Family Services (1999). Client Feedback Form.
Beyebach, M., & Carranza, V. E. (1997). Therapeutic interaction and dropout: Measuring relational communication in solution-focused therapy. Journal of Family Therapy, 19, 173-212.
Beyebach, M., Morejon, A. R., Palenzuela, D. L., & Rodriguez-Arias, J. L. (1996). Research on the process of solution-focused therapy. In S. D. Miller, M. A. Hubble, & B. L. Duncan (Eds.), Handbook of solution-focused brief therapy (pp. 299-334). San Francisco, CA, USA: Jossey-Bass.
Cockburn, J. T., Thomas, F. N., & Cockburn, O. J. (1997). Solution-focused therapy and psychosocial adjustment to orthopedic rehabilitation in a work hardening program. Journal of Occupational Rehabilitation, 7, 97-106.
De Jong, P., & Berg, I. K. (1998). Interviewing for solutions. Pacific Grove, CA, USA: Brooks/Cole.
de Shazer, S. (1985). Keys to solution in brief therapy. New York: Norton.
de Shazer, S. (1988). Clues: Investigating solutions in brief therapy. New York: Norton.
de Shazer, S. (1991). Putting differences to work. New York: Norton.
de Shazer, S. (1994). Words were originally magic. New York: Norton.
de Shazer, S., & Berg, I. K. (1997). 'What works?': Remarks on research aspects of Solution-Focused Brief Therapy. Journal of Family Therapy, 19, 121-124.
de Shazer, S., & Isebaert, L. (1997). A solution-focused approach to the treatment of alcohol problems: The Bruges Model. Unpublished manuscript.
de Shazer, S., & Molnar, A. (1984). Four useful interventions in brief family therapy. Journal of Marital and Family Therapy, 10, 297-304.
Dine, K. R. (1995). Visions of the future: The miracle question and the possibility for change. Unpublished dissertation; Massachusetts School of Professional Psychology.
Eakes, G., Walsh, S., Markowski, M., Cain, H., & Swanson, M. (1997). Family centered brief solution-focused therapy with chronic schizophrenia: A pilot study. Journal of Family Therapy, 19, 145-158.
Franklin, C., Corcoran, J., Nowicki, J., & Streeter, C. (1997). Using client self-anchored scales to measure outcomes in solution-focused therapy. Journal of Systemic Therapies, 16, 246-265.
Garfield, S. L. (1994). Research on client variables in psychotherapy. In S. L. Garfield & A. E. Bergin (eds.), Handbook of psychotherapy and behavior change (4th ed.). New York: Wiley & Sons.
Gingerich, W. J., de Shazer, S., & Weiner-Davis, M. (1988). Constructing change: A research view of interviewing. In E. Lipchik (Ed.), Interviewing (pp. 21-32). Rockville, MD, USA: Aspen.
Howard, K. I., Lopta, S. M., Krause, M. S., & Orlinsky, D. E. (1986). The dose-effect relationship in psychotherapy. American Psychologist, 41, 159-164.
Johnson, L. N., Nelson, T. S., & Allgood, S. M. (1998). Noticing pretreatment change and therapeutic outcome: An initial study. American Journal of Family Therapy, 26, 159-168.
Johnson, L. D., & Shaha, S. (1996). Improving quality in psychotherapy. Psychotherapy, 33, 225-236.
Jordan, K. B., & Quinn, W. H. (1994). Session two outcome of the formula first session task in problem- and solution-focused approaches. The American Journal of Family Therapy, 22, 3-16.
Jordan, K. B., & Quinn, W. H. (1997). Male and female client perception of session two outcome of the problem- and solution-focused approaches. Family Therapy, 24, 23-37.
Kazdin, A. E. (1994). Methodology, design, and evaluation in psychotherapy research. In A. E. Bergin & S. L. Garfield (Eds.), Handbook of psychotherapy and behavior change (4th ed., pp. 19-71). New York: Wiley.
D. (1988). A follow-up study conducted at the Brief Family Therapy Center. Unpublished
Kerlinger, F. N. (1986). Foundations of behavioral research (3rd ed.). New York: Holt, Rinehart, & Winston.
Koss, M. P., & Shiang, J. (1994). Research on brief psychotherapy. In A. E. Bergin & S. L. Garfield (eds.), Handbook of psychotherapy and behavior change (pp. 664-700). New York: Wiley & Sons.
Lawson, D. (1994). Identifying pretreatment change. Journal of Counseling and Development, 72, 244-248.
LaFountain, R. M., Garner, N. E., & Eliason, G. T. (1996). Solution-focused counseling groups: A key for school counselors. School Counselor, 43, 256-267.
Lee, M-Y (1997). The study of solution-focused brief family therapy: Outcomes and issues. American Journal of Family Therapy, 25(1), 3-17.
Lindforss, L., & Magnusson, D. (1997). Solution-focused therapy in prison. Contemporary Family Therapy, 19, 89-103.
Littrell, J. M., Malia, J. A., & Vanderwood. M. (1995). Single-session brief counseling in a high school. Journal of Counseling and Development, 73, 451-458.
Macdonald, A. J. (1994). Brief therapy in adult psychiatry. Journal of Family Therapy, 16, 415-426.
Macdonald, A. J. (1997). Brief therapy in adult psychiatry: Further outcomes. Journal of Family Therapy, 19, 213-222.
McKeel, A. J. (1996). A clinician's guide to research on solution-focused therapy. In S. D. Miller, M. A. Hubble, & B. L. Duncan (Eds.), Handbook of solution-focused brief therapy (pp. 251-271).
McKeel, A. J., & Weiner-Davis, M. (1995). Presuppositional questions and pretreatment change: A further analysis. Unpublished manuscript.
Metcalf, L., Thomas, F. N., Duncan, B. L., Miller, S. D., & Hubble, M. A. (1996). What works in solution-focused brief therapy. In S. D. Miller, M. A. Hubble, & B. L. Duncan (Eds.), Handbook of solution-focused brief therapy (pp. 335-349). San Francisco, CA, USA: Jossey-Bass.
Nau, D. S. (1997). Before the miracle: The optimum use of the solution-focused miracle question. Doctoral dissertation, NOVA Southeastern University.
Odel, M., Butler, T. J., & Dielman, M. B. (1997). Client experience of solution-focused couple therapy. Presentation for the 55th Annual Conference of the American Association for Marriage and Family Therapy, Atlanta GA, Sept 1997
O'Hanlon, W. H., & Weiner-Davis, M. (1989). In search of solutions: A new direction in psychotherapy. New York: Norton.
Pekarik, G. (1991). Relationship of expected and actual treatment duration for adult and child clients. Journal of clinical Child Psychology, 20, 121-125.
Schorr, M. (1997). Finding solutions in a roomful of angry people. Journal of systemic therapies, 16, 201-210.
Shields, C. G., Sprenkle, D. H., & Constantine, J. A. (1991). Anatomy of an initial interview: The importance of joining and structuring skills. American Journal of Family Therapy, 19, 3-18.
Shilts, L., Flippino, C., & Nau, D. S. (1994). Client-informed therapy. Journal of Systemic Therapies, 13, 39-52.
Shilts, L., Rambo, A., & Hernandez, L (1997). Clients helping therapists find solutions to their therapy. Contemporary Family Therapy, 19, 117-132.
Skidmore, J. E. (1993). A follow-up of therapists trained in the use of the solution-focused brief therapy model. Doctoral dissertation, University of South Dakota.
Schwartz, R. C., & Breunlin, D. (1983). Research: Why clinicians should bother with it. Networker, 7, 24-27, 57-59.
Sundman, P. (1997) Solution-focused ideas in social work. Journal of Family Therapy, 19, 159-172.
Todd, T. C., & Stanton, M. D. (1983). Research on marital and family therapy: Answers, issues, and recommendations for the future. In B. B. Wolman & G. Strickler (Eds.), Handbood of family and marital therapy (pp. 91-115). New York: Plenum.
Warner, R. E. (1996). Counsellor bias against shorter term counselling? A comparison of counsellor and client satisfaction in a Canadian setting. International Journal for the Advancement of Counselling, 18, 153-162.
Weiner-Davis, M., de Shazer, S., & Gingerich, W. J. (1987). Using pretreatment change to construct a therapeutic solution: An exploratory study. Journal of Marital and Family Therapy, 13, 359-363.
Zimmerman, T. S., Jacobson, R. B., MacIntyre, M., & Watson, C. (1996). Solution-focused parenting groups: An empiracle study. Journal of Systemic Therapies, 15, 12-25.
Zimmerman, T. S., Prest, L. A., & Wetzel, B. E. (1997). Solution-focused couples therapy groups: An empiracal study. Journal of Family Therapy, 19, 125-144.