MULTILEVEL-PROCESS MODELS OF THERAPEUTIC CHANGE IN MARRIAGE AND FAMILY THERAPY

BEYOND COMMON FACTORS: MULTILEVEL-PROCESS MODELS OF THERAPEUTIC CHANGE IN MARRIAGE AND FAMILY THERAPY

Sexton, Thomas L

A number of scholars have proposed the common factors perspective as the future direction of marriage and family therapy (MFT). Although intuitively appealing, the case for the common factors perspective is not as clear-cut as proponents portray. In its current form, the common factors perspective overlooks the multilevel nature of practice, the diversity of clients and settings, and the complexity of therapeutic change. In contrast, comprehensive process-based change models are an alternative to the limitations of common factors. In this article, we consider the limitations of the common factors perspective and propose the necessary and sufficient components and processes that might comprise comprehensive, multilevel, process-based therapeutic change models in MFT.

The rise of the common factors perspective, the evidence-based practices movement, the increased emphasis on accountability in clinical practice, and the aging of its founding theories all have pushed marriage and family therapy (MFT) once again to reconsider its conceptual foundations (Alexander, Sexton, & Robbins, 2002; Becvar, 2003). In a sense, the profession is at a critical choice-point or, stated differently, at the proverbial fork in the road at which MFT will make decisions with significant implications for the future unfolding of theory, research, practice, and training. Some scholars (i.e., Duncan, Miller, & Sparks, 2003; Sprenkle, Blow, & Dickey, 1999) have proposed that common factors should define the future direction of MFT. Without question, finding a common core of factors to explain successful therapy would be a major breakthrough. This finding would simplify practice, training, and research. It would unify the theoretical schools of MFT, which often compete against one another and find themselves in contentious struggles. In essence, it would serve as a shorthand explanation for the complexity of practice and the diversity of clients, settings, and the sometimes disparate research findings. Although intuitively appealing, the critical issue is whether or not the common factors perspective has been subjected to the type of rigorous scholarly debate, investigation, and discussion necessary to be declared a breakthrough worthy of serving as the conceptual foundation of MFT.

From our perspective, the debate over common factors as an explanatory model of MFT is a complex one for which the evidence is far from conclusive. On the one hand, there is no question common ingredients contribute to the efficacy/effectiveness of therapy. Any effective model of MFT should include the core common factors shared by many approaches to psychotherapy. On the other hand, common factors oversimplify the complex processes of MFT, the critical research needed to explore therapeutic change, and the theoretical richness required to explain practice. Hence, we do not believe that the common factors perspective is an adequate alternative to the current theoretical and research problems of MFT. In fact, we worry that the common factors perspective overlooks the multilevel nature of practice, the diversity of clients and settings, and the complexity of therapeutic change.

Regardless of one’s position on this issue, we suggest that the time has arrived for rigorous reflection on the relative merits of common factors and its role in the future direction of MFT. Unfortunately, to date, the debate has not been comprehensive. Furthermore, merely raising concerns over common factors is often met with strong emotion, rather than thoughtful reflection. In some arenas, common factors are even pitted against specific intervention models as though there is an either/or choice about what is beneficial in MFT. For example, Sprenkle et al. (1999) admonished builders of models to be humble and to recognize that all approaches achieve similar outcomes. We argue that such either/or thinking does not move the field forward. Instead, it is only through careful reflection that ideas and unsubstantiated claims can give way to more comprehensive ideas or “metatheories” that provide better answers to questions about therapeutic change, stimulate new theoretical revolutions, and provoke better programmatic research.

The purpose of this article is to initiate the scholarly debate and intense reflection needed at this critical choice-point. Our ultimate goal is to assist the profession in making an informed decision about its future conceptual foundation. To understand the debate and the choices that are before us, a context is necessary. In this case, the context is the dynamic and evolutional nature of MFT’s conceptual foundation. Over the years, we have witnessed the rise and abatement of noted authorities and theoretical perspectives. The common factors perspective takes its place among many movements. After explaining the context, we pose a series of questions regarding common factors that are intended to be thought provoking. Finally, we propose a modest alternative that we believe has the potential to guide the future direction of the profession. In making our proposal, we do not claim to have the final answer, but we lay out important parameters and components that heretofore have not been seriously considered.

MFT AS A SYSTEM OF EVOLVING KNOWLEDGE

Marriage and family therapy is a profession with deep theoretical and research roots. Many professionals in the field may find it easy to expect the foundational knowledge of MFT to endure as a way of thinking that remains relatively stable and static. However, like any dynamic system, new information entering the system requires revisions in the conceptual foundation. Some have suggested that knowledge bases, like any meaning system, are always open to change and modification (Alexander et al., 2002; Becvar, 2003). Hence, we should expect theories of today to be replaced by the new ideas of tomorrow. Likewise, MFT scholars have described changes in the profession’s knowledge base as an evolution of ideas or eras of thinking (Alexander et al.; Becvar; Gurman & Kniskern, 1981).

Evolution of MFT Thought

The dynamic evolution of MFT is well documented, having its roots in systems theory and cybernetics. In time, a systems approach to family therapy became widespread, and the field began to acknowledge that every family member perceives the family differently (Becvar, 2003). The 1960s and 1970s produced a variety of new models that were more comprehensive than previous approaches. By the late 1970s and 1980s, distinct schools of family therapy had developed. New challenges eventually forced a focus on integrating gender and power issues into clinical practice, leading to increased interest in eclectic approaches with their respect for the uniqueness of every client system and an emphasis on collaboration between therapist and client (Becvar). Sprenkle et al. (1999) suggested that the early schools were organized around the charismatic personalities of the progenitors, generating both loyalty and opposition that resulted in the development of alternative “schools.” It is of interest that the early schools remain the primary basis of teaching and, in many cases, practice in MFT (Sexton, Weeks, & Robbins, 2003).

The current era is characterized by dissatisfaction and competing perspectives. Among scholars and researchers, dissatisfaction stems primarily from the inability of existing theories to explain adequately the complexity involved in clinical practice. Surveys reveal that once practitioners enter the real world of practice, they abandon theory in favor of “eclectic” or “integrative” therapy (Lebow, 2003). Although these findings are not specific to MFT, they suggest that traditional theory falls short in providing practitioners the necessary guidance to make systematic clinical decisions in dealing with complex clinical problems.

The current era is also characterized by a rapidly changing landscape of MFT practice. The field is now held to a higher level of accountability in which MFTs are required to demonstrate outcomes, use “best practices,” and maintain cost-effectiveness (Alexander et al., 2002; Sexton, Weeks, et al., 2003). Although some lament this situation as an imposition by managed care, we take a positive view of the movement toward accountability. The result has been the development of specific intervention models for specific clinical problems (Sexton, Alexander, & Mease, 2004), the rise of best practices, and the evolution of evidence-based treatment models. Alexander et al. (2002) argued that the new era of MFT has fostered the development of “mature clinical models.” These models are built on the formative concepts in the field, include constructs from first- and second-generation schools or theoretical approaches, and are informed by current process and outcome research.

The growth in the research literature on process and outcome in MFT also has become a factor of increased importance. What began as a need to demonstrate efficacy and justify MFT has evolved into a rich knowledge base involving elaborate and innovative research. The focus of investigation has shifted to change processes and clinical outcomes that mirror the unique clinical challenges facing MFT. In this evolution, the research has broadened its focus from simple questions of outcome (e.g., does it work?) to specific applications of MFT with specific clinical problems in specific settings (e.g., effectiveness). Process research has evolved from identifying broad and general “process events” to specific change mechanisms that promote proximal goals within specific models of therapy (Alexander, Holtzworth-Munroe, & Jameson, 1994; Pinsof & Wynne, 2000; Sexton, Alexander, et al., 2004). Critical reviews have suggested that MFT literature in many ways is far ahead of other specializations in its specificity, elaborateness, and external validity in community-based effectiveness trials.

The MFT research literature has been summarized in a number of qualitative reviews and metaanalyses as well as a recent major contribution published by American Association of Marriage and Family Therapy (Sprenkle, 2002). Recent qualitative reviews (Alexander & Barton, 1995; Lebow & Gurman, 1995; Pinsof & Wynne, 1995; Sexton, Robbins, Hollimon, Mease, & Mayorga, 2003) concluded that couple and family interventions are efficacious in the broadest sense; they are efficacious for a wide range of problems; they produce positive results with different types of families; and positive results endure over long periods of time. More specifically, the current reviews have suggested that MFT intervention programs are effective for the treatment of adult schizophrenia, alcoholism, and drug abuse, and for adolescent conduct/oppositional defiant disorders and drug abuse (Alexander et al., 1994; Estrada & Pinsof, 1995; Pinsof & Wynne, 2000; Sexton, Alexander, et al., 2004; Stanton & Shadish, 1997). The cumulative results of the outcome research have led some to argue that no quantitative differences have been found among the traditional schools of therapy; thus, there must be no differences (Gurman & Kniskern, 1981; Gurman, Kniskern, & Pinsof, 1986; Shadish et al., 1993). Other scholars have maintained that the great success of certain intervention programs with specific clinical problems suggests that differences between approaches do indeed exist and that these differences are substantial (Sexton, Alexander, et al., 2004).

Rise of Common Factors

The common factors perspective evolved out of the apparent inability of traditional schools of therapy or philosophical ideas to explain and guide practice. Seemingly inclusive results of research examining differences between models provided an additional impetus. The common factors movement was further supported by the development of meta-analytic research review methodology. The common factors ideas began in the individual psychotherapy literature. The perspective has gained momentum in MFT because it has intuitive appeal, seems to explain the disparate findings of theory and research, and has the potential to unite the field by identifying common ground. As noted above, common factors are the result of our theoretical and research evolution, and they represent one of the critical choices made in encountering theoretical and research forks in the road. In some ways, it seems that the embracing of common factors is as much an oppositional reaction to the current empirically validated/supported treatment movement and statement of philosophical support of the values of the current post-modern perspectives as it is a specific endorsement of the movement itself.

Duncan et al. (2003) and Sprenkle et al. (1999) were the first to suggest that the field of family therapy should move beyond current paradigms to consider a common factors model that is client-directed and outcome informed. The common factors model that has been proposed is not based fully on research, but in part, on Lambert’s ( 1992) taxonomy of factors in successful therapy. Based on a qualitative review of research, this often-cited taxonomy suggested that certain percentages of the variance in change could be attributed to clients/extratherapeutic factors (40%), relationship factors (30%), placebo, hope, and expectancy factors (15%) and model/technique factors (15%).

Interestingly, the common factors movement is not new. The idea of common factors in counseling dates back to Saul Rosenzweig’s 1936 publication in which he suggested that common factors across schools of psychotherapy are responsible for facilitating change. Since the 1936 publication, we located almost 30 lists of common factors in the psychotherapy literature. Almost two decades later, Frank (1971) proposed six factors: An intense emotionally charged relationship; a rationale that explains that nature of patient’s distress; provision of new information about the sources of the client’s problems; strengthening of the client’s expectation of help through the therapist’s personal qualities; provision of the experience of success; and facilitation of emotional arousal. In recent years, new lists of common factors have emerged that continue to build from these early conceptualizations. Orlinsky and Howard (1986) identified five categories of common factors in the counseling literature: Therapeutic contract, therapeutic interventions, therapeutic bond, patient self-relatedness, and therapeutic realizations. Similarly, Grencavage and Norcross (1990) identified five superordinate categories of common factors in the literature: Client characteristics, therapist qualities, change processes, treatment structure, and therapeutic relationship. Further, they stated that the most consensual commonalities across all categories included the development of a therapeutic alliance (56%), opportunity for catharsis (38%), acquisition and practice of new behaviors (32%), clients’ positive expectancies (26%), beneficial therapist qualities (24%), and provision of a rationale as a change process (24%).

Common factors have also gained support from the conclusions of early qualitative reviews and metaanalyses reviews of both individual psychotherapy and MFT outcome research. In individual therapy domains, the early meta-analytic studies found individual therapy to be successful (when compared to no-treatment conditions) but were unable to identify differences among the broad theoretical approaches studied (Smith, Glass, & Miller, 1980). These early findings have been replicated in more recent metaanalyses, which also have been unable to quantitatively differentiate traditional “schools” of therapy. The most impressive work in this regard was conducted by Wampold (2001) who argued for common factors as an alternative to the “medical models” approach in individual therapy. By using meta-analytic techniques, Wampold was able to demonstrate convincingly that the variance in outcomes in therapy is accounted for best by common rather than specific factors.

The early reviews of MFT resulted in similar findings and thus, a similar conclusion that the “nonspecific” or “common” factors were what made for successful MFT outcome. Qualitative reviews (Gurman & Kniskern, 1981; Gurman et al., 1986) and meta-analyses (Hazelrigg, Cooper, & Borduin, 1987) found strong main effects for MFT in general but were unable to demonstrate or support the differential effectiveness of specific models of practice. The models that did have some degree of research support were active or highly directive in nature (Hazelrigg et al.). The most recent meta-analysis of meta-analyses of MFT research replicated these earlier findings (Shadish & Baldwin, 2002).

COMMON FACTORS AS AN INADEQUATE FOUNDATION FOR MFT

There is no question that the common factors approach has an intuitive appeal for students and clinicians who are attempting to sort out the complexity of practice, MFT educators who are looking for a unifying conceptual framework to use in training their students, and for all who are trying to make sense out of the wide and diverse research findings in the field. There is also an interesting symbiosis with the current values and culture of postmodern philosophy that Becvar (2003) suggested in the new “post modern era” of MFT. In addition, the basic premise of this movement is probably correct: There are central and common factors that contribute to successful outcomes that cut across seemingly different theoretical and practice models. Furthermore, the research findings noted above seem to point toward a “common” rather than specific set of active ingredients in both individual and MFT. Thus, there is an appeal for unity in the midst of diversity.

In this section, we raise pointed questions about the common factors perspective. Answers to these questions pose a challenge to the growing predominance and centrality given to this perspective (Duncan et al., 2003; Sprenkle et al., 1999). These questions reflect on the broad domains of theory, research, practice, and training.

1. Is There Research Support For Common Factors?

Systematic inquiry and research are critical to support the theories that underscore practice. Therefore, it is justifiable to question whether any theoretical framework has research support that establishes it as a basis of practice. The question is particularly relevant in the case of common factors. Research has been a central piece of the argument for using the perspective in MFT.

Early research that failed to find differences between the traditional schools of therapeutic intervention (in both psychotherapy in general and MFT) has been a central feature in the argument for common factors. In particular, most qualitative research reviews and meta-analyses failed to find either trends or specific effect size estimates for the differential effects. The major question is whether the conclusions and implications drawn from these findings are accurate. The answer is both yes and no. When the differential effects of the traditional schools of MFT are compared, there are no differences. However, many scholars do not find this surprising. The primary issue here is whether we should ever expect differences in statistical and clinical significance, given the complexity of variables involved in the outcome variance of any clinical intervention. Consider the nature of the traditional schools. By and large, they are broad and general worldviews that, to various degrees, set forth principles of therapeutic actions rather than specific clinical protocols that guide therapists to act in very different ways. A structural family therapist may see different aspects of the family relational system than a Bowenian family therapist, but there is likely to be little difference in the behavior in which they may engage during therapy.

The strongest support for common factors has come from the findings of meta-analytic studies. Meta-analytic studies are appealing because they produce effect sizes that measure the magnitude of effects. Wampold (2001) and others (Brown, 1987; Rosenthal, 1985; Shadish & Baldwin, 2002) have noted that meta-analysis is an invaluable tool to translate the diverse findings of individual clinical trial studies into trends that have practice implications. The work of Wampold in particular is an outstanding scholarly analysis of the common factors versus specific model factors in therapeutic outcomes. Despite the value of the approach, two important considerations need to be taken into account in regard to these findings. First, effect sizes can only be calculated for those aspects of studies that are coded by the researcher prior to the analysis. Only the most general variables of clinical trial studies have been coded. Thus, the outcome of a meta-analysis is limited to those variables identified and coded and to the operational definition of the coded variables. To date, no meta-analysis has coded for anything other than broad and general approaches to therapy (e.g., structural family therapy). Given that schools of therapy do not likely represent specific practices, this is a serious limitation in regard to identifying differences among approaches. Second, the statistical power to find differences among approaches is lacking in each of the previous meta-analytic studies. Each of the major meta-analyses (Hazelrigg et al., 1987; Shadish et al., 1993) included a number of individual clinical trial studies. However, there were few studies that actually compared different approaches. This is analogous to having a specific clinical trials study in which the number of subjects is quite low. When sample sizes (e.g., subjects) are low, the probability of type II error (not finding something that is there) increases.

In other words, meta-analysis, like any other analytic method, is based on the data that are included. Each of the previous meta-analyses identified and coded for broad and general theoretical schools of therapy. This is not surprising. As noted above, at the time these analyses were conducted, schools were the primary way in which therapeutic approaches were differentiated. However, it is questionable as to whether global therapeutic orientations really represent the important and distinguishing characteristics that differentiate effective and ineffective interventions (Alexander et al., 1994; Sexton, Alexander, et al., 2004). In addition, classifying approaches as a single main effect (e.g., the effect of the approaches) without consideration for the clinical problem of the client is questionable in our view. In actual clinical practice, the interaction between approaches and clinical problems in MFT is quite complex. How an MFT works with an adolescent with an externalizing behavior disorder is quite different from how the therapist works with a depressed child. Furthermore, none of the previous meta-analyses coded for intervention type that captures the current group of comprehensive models of intervention (Sexton & Alexander, 2002).

In light of these difficulties, conclusions regarding a lack of differences between clinical approaches cannot be answered by the current meta-analyses. However, there is other research to reference in an attempt to answer this question. For example, recent individual clinical trials and systematic qualitative reviews have provided a very different picture about the primacy of common factors. These studies have indicated that a number of family therapy and family-based intervention models have enough evidence to suggest they are effective with specific client problems and populations (Alexander et al., 1994; Sexton, Alexander, et al., 2004). Kazdin’s (1997) review identified three promising treatment approaches in the area of adolescent externalizing behavior disorders that produced results that were better than other approaches: Parent management training (PMT; Patterson, 1996), functional family therapy (FFT; Alexander & Sexton, 2002; Sexton & Alexander, 2002), and multisystemic therapy (MST; Henggeler, & Borduin, 1990; Henggeler, Schoenwald, Borduin, Rowland, & Cunningham, 1998). In a review of the literature on adolescent drug use/abuse problems, Sexton and Alexander (2002) found four programs for treating adolescent drug use/abuse problems that consistently produced outcomes greater than other MFT approaches: Multidimensional family therapy (MDFT; Liddle & Dakof, 1995), brief strategic family therapy (BSFT; Szapocznik & Kurtines, 1989), FFT, and MST. Interestingly, these family models address clinical problems often considered to be very difficult to treat (e.g., adolescent conduct disorders, drug abuse, and schizophrenia) (see Sexton, Alexander, et al., 2004, for a more specific review).

Liddle and Dakof (1995) and Rowe and Liddle (2003) found that family therapy is more successful in engaging and retaining youth and their families in drug treatment when compared with peer-group and individual therapy, as well as when compared with treatment as usual. They identified three specific models of practice to account for these differences: MDFT, MST, and FFT. Based on a review of 13 random assignment clinical trials studies investigating the outcomes of family-based treatment for adolescent drug use, Waldron (1997) identified three categories of effective family-based intervention models: Family systems models (structural-strategic models and FFT), behavioral family models, and ecological family-based intervention models (MST) that produced outcomes greater than other approaches. Sexton, Alexander, et al. (2004) and Sexton, Robbins, et al. (2003) reported trends from a study of all clinical trials studies in couple and family therapy over the last decade and found that some models of MFT (e.g., behavioral marital therapy, emotionally focused couple therapy, MST, and FFT) were more effective than others in addressing specific clinical problems. Sprenkle (2002) also provided an impressive compilation of reviews of effective approaches for specific clinical problems.

Two qualitative reviews from outside the traditional research settings shed light on the extent of the evidence for certain specific models of family therapy as well. Elliott (1998) conducted a systematic search of existing intervention programs for youth violence for the Center for the Study and Prevention of Violence (CSPV), which was the basis for the “Blueprint” program. This collaborative venture between CSPV, the Centers for Disease Control, and the Office of Juvenile Justice and Delinquency Prevention (OJJDP) was undertaken to identify best practices in the treatment of adolescent behavior problems. Only 10 (now 11) of 1000 interventions met their strict standards, and three of these approaches were family-based interventions: FFT, MST, and Oregon Treatment Foster Care (Chamberlain & Rosicky, 1995). The second review was conducted by the Surgeon General (U.S. Public Health Service, 2001). Based on stringent standards of evidence, four intervention programs were identified as “Level 1” programs effective for use with the range of clinical problems labeled conduct disorder and violent youth. Of these, three were therapy based, FFT, MST, and OTFC. In a unique study using cost savings as an outcome, The Public Policy Institute of the State of Washington completed an economic analysis of the outcome and cost-effectiveness of various approaches to reduce delinquency (Aos & Barnoski, 1998). Two family-based intervention programs, FFT and MST, had among the strongest effectiveness ratings and highest cost savings when compared to other juvenile offender programs.

Overall, a critical review of qualitative research reviews and individual clinical trial data suggests that some clinical intervention models may be more effective than other approaches, particularly with specific clinical problems. Certainly, there is a legitimate argument that qualitative reviews and even clinical trials research studies are subject to bias and may not reflect broad trends. However, as noted above, meta-analyses have similar limitations.

Despite these consistent trends, one may question whether or not these findings supporting certain intervention programs are nothing more than common factors packaged into new models with different names. Interesting new research investigating the role of specific model adherence and clinical outcomes suggests this is probably not the case. This work suggests that treatment fidelity, or adherence to a specific model, is a critical factor in the delivery and outcome of effective programs (Henggeler et al., 1997; Sexton & Alexander, 2002). Treatment fidelity is most often defined as therapist adherence or the degree to which a given therapy is implemented in accordance with essential theoretical and procedural aspects of the model (Waltz, Addis, Koerner, & Jacobson, 1993). In a recent study of therapist adherence in FFT, Sexton, Sydnor, and Rowland (2003) and Barnowski (2002) found therapist adherence to be the primary mediating factor in the successful delivery of FFT. Those therapists who delivered the model with high rates of adherence had significantly better outcomes than those who did not. Although there is no question that the work on adherence and competence in family therapy is only beginning, these findings would suggest that the demonstrated positive outcome of the current generation of family-based treatment models (e.g., MST and FFT) are dependent on specific theoretical and procedural elements of the model.

Despite these limitations, the fundamental question of interest remains: If there are differences in models, why can we not find them? We argue that a better question is, “why have we not yet found them?” We reassert that the difficulty in identifying differential effects is due more to the nature of the clinical models of MFT practice than to the primacy of common factors. Over the years, schools of therapy have given way to specific, systematic, and well-articulated clinical models (e.g., MST, MDFT, and FFT; Alexander et al., 2002; Becvar, 2003). Many now believe that there are certain systematic intervention models of MFT practice that are more effective than integrative and eclectic forms of family therapy or than the traditional schools of family therapy. To date there has not been a meta-analysis that appropriately compared the impact of these current clinical intervention programs of MFT. In fact, the findings based on recent clinical trials studies and systematic qualitative reviews of the available evidence by scholars within and outside the field of MFT suggest that the confident conclusion of common factors proponents may be premature. With the rise of the systematic approaches to both couple and family therapy, the future will certainly provide a more accurate test of these issues. We suggest that a telling test of common factors versus specific clinical models would be meta-analysis conducted on the current models of practice rather than with historical classification systems (e.g., broad theoretical schools) in which the clinical problem or category of clinical problem was the mediating variable.

2. Do Common Factors Integrate Research Into Practice?

One hallmark of a vital profession is the inextricable link between research, practice, and theory. Theory generates ideas; research seeks to test these ideas, and whenever ideas are verified, a stronger foundation exists for practice. Unfortunately, in MFT there is a well-documented research-practice gap (Gurman et al., 1986; Sexton, Alexander et al., 2004). Despite the considerable knowledge developed through both outcome (efficacy and effectiveness) and process research, these findings continue to have little impact on actual MFT practice (Pinsof & Wynne, 2000; Sexton, Robbins et al., 2003). The gap exists despite the current era of accountability in which research and the real world of practice are considered necessary partners and within the political world in which “appropriate services” for clients are determined. A crucial concern is whether the common factors perspective unites research and practice in a way that will move the field forward conceptually and practically. For instance, will it stimulate questions that serve as the basis for investigating ecologically valid clinical interventions?

We argue that the proposal for common factors in their current form for psychotherapy (Duncan et al., 2003; Hubble, Duncan, & Miller, 1999) or MFT (Sprenkle et al., 1999) does not provide for such integration. Two reasons lead us to this conclusion. First, common factors are not conceptually clear, operationally defined, or contextualized within a clinical process enough to make them either researchable or understandable. Second, as currently described, common factors are independent factors that are decontextualized from the complex process of therapy. For example, an often-cited factor, “the relationship” (between therapist and family/couple), has numerous meanings depending on the perspective of the researchers. For some scholars, the relationship is steeped in the psychoanalytic tradition of transference (Gelso & Carter, 1985), whereas for other scholars, it is a process of alliance (Sexton & Whiston, 1994). Without a more specific definition, this “common factor” cannot be subjected to rigorous research. Furthermore, the critical and important relationships between the factors and their proximal and distal outcomes are not articulated and thus, not open to study.

Along a different line, common factors are often presented as the primary set of active ingredients in therapy. The implication is that all the previous research has been “chasing the tail” and not produced the necessary answer (Duncan et al., 2003). Although certainly unintended, the implication that the sophistication and complexity of current process and outcome research is unnecessary, misguided, and not relevant to practice often comes with the dissemination of the common practices perspective. Such a perspective does little to overcome the research-practice gap and, in our view, actually contributes to the gap. As Alexander et al. (1994) and Sexton, Robbins, et al. (2003) suggested, the ultimate goal should be the convergence of research and practice in the interest of helping clients. Although each domain serves a unique purpose, both are tied to each other and are critical to the dialectic of therapeutic change. We are concerned that the passion for common factors sends an unintended message that further separates rather than unites research and practice.

3. Do Common Factors Provide an Adequate Theoretical or Conceptual Foundation to Explain the Processes or Mechanisms of Change?

Regardless of the diversity of practice settings or the specific theoretical approach of the practitioner, MFT is fundamentally about change-change in families, change in couples, change in individuals, and change in the social, cultural, and organizational settings in which they exist. The clinical impact of a comprehensive theory is pragmatic. In this regard, the question has been: “Do the theoretical constructs explain the intrapersonal (within-client or within-therapist) and interactional (between client and therapist) process of change?” As the efficacy of MFT has become clear, the more interesting and important question has become: “What are the processes that make therapy effective?” In the current era of accountability, this goal is even more specific and requires a systematic understanding of the process of change and the specific clinical and interpersonal mechanisms of change that demonstrate outcomes. Therefore, the field begs for an adequate explanation of the process and mechanisms of the bigger and more complex change process, rather than mere description and identification of “factors” involved in change.

Despite the contributions of the common factors perspective, it does not explain the complexity of change or the process through which change takes place. In their current form, common factors are presented as discrete factors that, as noted above, are neither operationally defined, contextualized within the clinical process into which they might fit, or explicated as to the mechanisms that might promote their outcome. In fact, many of the factors that find their way to common factors lists are actually the outcomes of an undefined relational and interpersonal process rather than the therapeutic change mechanisms they imply. For example, some proponents of the perspective have identified hope and expectation as common factors (Frank, 1971; Hanna, 1996; Snyder, Michael, & Cheavens, 1999). Although these factors probably contribute to change in some way, neither describes their underlying mechanisms of therapeutic change, specific therapist actions that activate them, how they participate in the process, or when in a complex process of change hope and expectation may be most important.

Our analysis is supported by many scholars from within the common factors movement. For example, Arkowitz (1995) stated that an important challenge for future research is to develop theories of the processes of change that include a variety of factors and their interactions. He maintained that instead of developing more listings of discrete common factors, we need to explicate theories of change that address factors that may be common to all therapies. Prochaska (1999) argued that the field of psychotherapy needs an adequate theory of behavior change that helps to explain how people change within and between therapy sessions. Kassel and Wagner (1993) argued that there is variability in perceptions of the most important change mechanisms, and that we need to investigate which mechanisms are most salient at what points during the change process. Hanna (1996) suggested that if researchers are to improve the change potential of psychotherapy, they must address some key aspects of change, which include intensifying client motivation, augmenting client involvement in therapy, enhancing the client’s sense of possibility that change can take place, and removing perceived obstacles to change.

The notion of change process as opposed to common factors also has been advanced in recent years by the process research conducted in MFT. The impressive work of Robbins, Alexander, and Turner (2000) and Alexander, Newell, Robbins, and Turner (1995) related to therapeutic alliance; Gottman, Coan, Carrere, and Swanson (1998) regarding marital interaction; and Szapocznik and Kurtines (1989) on engagement with adolescent drug users are all examples of significant efforts at explaining the moderators (complex preexisting client and context variables) and mediators (process mechanisms that have an impact on outcome) of change.

Many clinical researchers now suggest that “intervention-specific” processes may be critically related to outcome, and that specific intervention processes may be particularly effective with specific couples/families (Rohrbaugh et al., 2001). In their comprehensive review of research on couple and family therapy over the last decade, Sexton, Alexander, et al. (2004) attempted to identify the moderators and mediators of successful MFT outcomes. They focused on identifying “mechanisms” or processes that result in change rather than descriptive factors (such as common factors). In attempting to identify research support for common or core change mechanism constructs that have been identified from research across intervention models, Sexton, Alexander, et al. suggested that there were common change mechanisms in effective therapies, but that these mechanisms could not be understood apart from the systematic change models in which they are activated. The common mechanisms of change included: (a) Redefinition of the presenting problem, (b) impasse resolution, (c) therapeutic alliance, (d) reduction of within-session negativity, (e) improved interactional and behavioral competency, and (f) treatment adherence to the specific model being practiced.

From our perspective, change mechanisms are not the same as common factors. Instead, they are evidence-based mechanisms facultative of different short- and long-term goals that ultimately lead to therapeutic change. The difference lies in the description of common factors as nonspecific or common features or elements of change. This usage is understandable in the current context in which the medical model dominates. In taking a medical model view, nonspecific factors are not the active ingredients of the treatment. A placebo or some contextual feature makes the difference. In MFT or psychotherapy in general, the nonspecific factors are part the active ingredients in the change process. Thus, the critical issue should be the explanation of the mechanism or clinical pathways by which these factors (e.g., hope, relationship) are created and developed. The goal is to identify the mechanisms by which these factors can be made to occur, the process within which they must be placed (e.g,. the temporal order), and the way in which they can help the change agent (e.g., the therapist) follow them in such an order to increase the likelihood of successful change. This perspective would suggest that the context or the model within which these factors exist is critical. It is the model that pulls together, provides the principles for certain links between factors, and gives a temporal and directional dimension.

4. Do Common Factors Advance Theory Development?

Just like research, theory development is integral to any profession. As we argued previously, the theoretical ideas of a profession are not static but evolve through various eras of thinking (Becvar, 2003) and through various shifts in prevailing paradigms (Kuhn, 1964). Implicit in the dynamic evolution of theory is that each era serves as a platform for examining the adequacy of the prevailing views and provides a stepping stone to new ways of thinking. In MFT, the primary theory questions are: What models/theoretical perspectives accurately explain the change processes of clients, the active ingredients of change, and the relational ways in which those change mechanism can successfully be achieved? For example, what are the critical preexisting client factors that must be accounted for and understood for successful therapy? What is the etiology of problems (e.g., what is a theoretical model for the development of family conflict and couple symptoms)? What are the critical short-term (proximal) and long-term (distal) goals of a successful therapy? What is the temporal and systematic relationship among these goals within a “process of change”?

Unfortunately, the current articulation of the common factors perspective is seriously lacking as a comprehensive theoretical foundation to MFT. In our opinion, this is due to the oversimplification of the complex client change processes, the interactional dynamics between client and therapist, and the change mechanisms inherent in the common factors perspective. Because common factors are inherently an oversimplification of a complex set of factors (the diversity of client and therapist) and process (change mechanisms, change process, and client-therapist interactional dynamics), critical elements of a theory are overlooked. Common factors leave too many critical questions unanswered and, in our view do not provide a structure for the explication of these factors. Thus, in their current articulation, common factors can never provide a theoretical platform for further theoretical or research development. This is not to say that further development of these ideas might not result eventually in theory development. However, it is to say that the oversimplification inherent in common factors will need to be replaced by a more accurate representation of the complexity of MFT.

5. Do Common Factors Provide the Guidance Necessary for Successful Clinical Work?

Marriage and family therapy practice is a complex multilevel interpersonal process. The clinical procedures necessary to systematically conduct therapy in ways that promote successful outcome are equally complex. Clinical decisions require a complex understanding of the client, the principles of change, and the mechanisms that facilitate that change. If a perspective like common factors is to be useful, it must not only explain theory and the process of change, but also guide practitioners as they conduct this complex relational process of helping clients.

The question is whether common factors can provide that guidance necessary to navigate the complexity of the process. In our view, it does not, and it never will as it is currently conceptualized. Consider, for example, the therapeutic relationship, which is often referred to as a common factor. Despite the confusion regarding the meaning of the term relationship, it has been identified as an important factor in successful therapy. However, the relationship is really an outcome of some action and activity of the therapist and the client. Thus, the relationship (or expectation, etc.) factor is a goal for some set of interactions, behaviors, or activities that occur between the client and therapist. Common factors do not provide the directional steps that may facilitate change. Thus, common factors do not explain the process of change and therefore leave the clinician at a loss for guidance in the most basic need to know what first, what second, and what third.

Sexton and Alexander (2002) argued that in the midst of facilitating change, clinicians are best served by having comprehensive principles that explain clients and their problems as the basis of a map or set of systematic procedures that describe the process of change. Having conceptual principles allows the therapist to make informed clinical decisions, whereas the procedural map increases the probability of achieving successful outcomes. What a map does is provide direction that increases the efficiency and likelihood of success in the journey. What a map does not do is explain the events of the journey or remove the critical decision making required to overcome unexpected events along the way (e.g., explain all that will happen along the way, where there will be delays). Regardless of the events that occur in any journey, the outcome is greatly facilitated by a directional map. The uniqueness and richness of the journey remain a delightful mystery despite having the road map. There is no way to know the detours, unique signs, sounds, or events in the process, but a map of the process allows clinicians to be efficient, responsive to the needs of the client, and increase the likelihood of successful outcomes.

6. Can Common Factors Serve as the Basis of Clinical Training?

Clinical training models are inherently and inexorably related to the theoretical and research of any domain of clinical practice like MFT. The challenge for the clinical trainer or training program is to provide the foundations of clinical theory and practice while developing an appreciation for the role of research, theory, and the complexities of client and therapeutic change. The question for common factors is whether it can be the basis of teaching and training. Because of the issues noted above, we do not believe that this is possible. It does not provide a comprehensive model or theory that gives a trainee the principles of practice. It does not provide the specific goals and procedures of various phases of change that new therapists can grasp on to when trying to learn to navigate the complex process of clinical change. It does not provide a “map” of the initial structure that over time will provide the support necessary to learn the process of change.

THE NEXT CHOICE-POINT: EMBRACING THE COMPLEXITY OF THERAPEUTIC CHANGE

If the common factors perspective is lacking as the conceptual foundation for MFT, what is a viable alternative? Is there anything that goes beyond traditional broad theory and school approaches to MFT? Is there anything more comprehensive than common factors or any perspective that is anchored in practice and research and at the same time systematic enough to provide clinical guidance? Is there anything that moves the knowledge base forward? We assert that the alternative is to develop systematic and process based models of change. These models should link critical factors of change based on a set of principles and articulate the multiple components and processes inherent in therapeutic change. In fact, we suggest that models of change arc indeed the best choice for MFT at the current choice-point.

The notion of change models has a long history in the field. The very literature that brought forth common factors actually has its roots in change models (Ogles, Anderson, & Lunnen, 1999). Unfortunately, by extracting only the outcomes or the factors, the important models and processes that link these factors are lost. The change models that have been developed to date provide insight into only one process involved in change (e.g., client change as described by Prochaska, 2000; Hanna, 2002; and Snyder et al., 1999), but none provide a comprehensive framework that describes multiple levels, components, and processes of change. Therefore, although each model has something useful to offer, none is sufficient as a foundation for MFT.

There also are a number of systematic models of change already existing in MFT. These models demonstrate the value of and critical elements necessary for a clinical model (e.g., MST, FFT, EFCT, behavioral/integrative couple therapy, MDFT, BSSFT). Alexander et al. (2002) described some of these as “mature clinical models” of practice because of their integration of practice, research, and theory in a way that improves practice. According to Alexander et al. (2002), although they have different theoretical principles and therapeutic process, the emerging models share some commonalities. They are well-articulated, systematic approaches to treatment in which: (a) Clinically meaningful problems are targeted; (b) a coherent conceptual framework underlies the clinical interventions; (c) specific interventions are described in detail with an articulation of the therapist qualities necessary to follow them; (d) process research identifies how the change mechanisms work; and (e) outcome research that demonstrates how well they work. It is beyond the scope of this article to review each of these models. However, each has demonstrated efficacy/effectiveness research support, systematic theory, and clear clinical protocols that are responsive to the individual differences of clients. For the interested reader, there are numerous resources that provide specific information on each of these models (see Sexton, Weeks, et al., 2003).

We assert that the next logical evolution is the development of multilevel-process models of therapeutic change. These models should be grounded in theory, supported by empirical research, and provide a clear framework for practice. We contend that only complex, comprehensive, and systematic models of change have the potential to explain the complexity and diversity of clients and relational processes inherent in MFT practice. Our goal here is not to promote any single model. In fact, we do not believe necessarily that the field should adopt a single model of change. We think that there is potential for multiple models that differ in their specific principles, practices, and types of clinical problems they address. In this section, we make a modest proposal of what we believe are important issues for consideration in developing a multilevel-process model of change. In keeping with our stated purpose, we propose what we believe are the necessary dimensions of change models as opposed to describing an actual model.

The Nature of a Model

To begin the task, we should examine the nature of a model. A theoretical model is a concise explanation of some phenomenon or aspect of reality. Of course, the phenomenon we wish to explain is therapeutic change. A model specifies the salient features or components of a phenomenon and attempts to demonstrate the manner in which the features interact with one another (Dublin, 1978). According to Dublin, most theoretical models represent a complex portion of the real world. On this later point of Dublin, two ideas are relevant for model building: Representation and complexity. First, the value of any model rests on its accuracy in representing the phenomenon being explained. Obviously, models that are inaccurate representations cannot be useful. Second, the phenomenon represented by a model typically is complex. Rescher (1998) defined complexity as having two essential features: (a) The number and variety of constituent elements of a phenomenon, and (b) the elaborateness of the interactions of the elements. Unfortunately, many theoretical models inaccurately represent and oversimplify the complexity inherent in the phenomenon for which they purport to represent. As a result, inaccuracy and oversimplification undermine meaningful theory explication, practice, research, and training.

Criteria for a Model of Change

What would a useful model of change look like? In reviewing the literature, we found no clear, cogent criteria for building models of change. Because of this shortcoming, we established our own criteria based on what we believe a model should accomplish. Our seven criteria follow.

Comprehensive

A model includes all of the constituent elements or components of the process of change. It also includes the interrelationships of the elements. At the same time, the model excludes elements and interrelationships that are irrelevant to understanding change.

Heuristic

A model stimulates further investigation of the process of change. The very nature of the model raises important questions for inquiry that are not provoked by previous models. The ultimate result should be to improve theory explication and scientific inquiry, continually expanding our knowledge base and broadening our understanding of change.

Metatheoretical

A model is not constrained by any theoretical orientation. Instead, it draws on parsimonious principles and concepts as well as scientific knowledge that cut across theoretical lines. In addition, the model includes critical nonparsimonious principles and concepts that are nevertheless integral to understanding change. The metatheoretical perspective supercedes the narrowness of specific theories and models. On this subject, Alexander et al. (2002) argued that models need strong internal consistency that endures over time. Yet, they are dynamic and open to assimilate and incorporate new ideas that explain further aspects of change.

Systematic

In the process of therapeutic change, as in any system, the whole is more than the sum of its parts and components. As Rescher (1998) pointed out, one aspect of complexity is the interrelational elaborateness of the constituent elements of a phenomenon. We propose that a model of change demonstrate how the interaction of constituent components operates in concert to achieve therapeutic change. Specifying a systematic approach would be an advantage over other models whose components are disconnected and whose outcomes of change can only be explained as random events.

Practical

Kurt Lewin (1951) once said, “There is nothing so practical as a good theory” (p. 169). A model of change should have utility. By integrating up-to-date scientific knowledge and theory, the bridge to practice should be apparent. Therapists would be armed with more informed guidance in making treatment decisions and selecting interventions.

Simplistic Without Being Oversimplistic

A model must overcome the significant problem of failing to include all of the constituent elements and interrelationships. We propose that a model of change be manageable and understandable while avoiding the trap of oversimplifying complex phenomena. This criterion is based on the premise that oversimplification leads to confusion in understanding a phenomenon as does the inclusion of irrelevant information.

Clear

A model should provide explicit definitions of key principles, concepts, and components. It avoids the use of vague, ambiguous, and imprecise language, making it difficult for readers to interpret the author’s intended meanings. Clarity also enables participants in a conversation to have meaningful and unambiguous dialogue.

A Modest Proposal: A Multilevel-Process Model

Drawing on Rescher’s (1998) definition of complexity, our proposal suggests that a model of therapeutic change consists of two major components: Foundational principles and multilevel change processes. Figure 1 provides a graphical representation of these major elements and relationship among elements that would comprise a comprehensive change model that meet the criteria noted above.

Foundational Principles

To be a useful model, the components of the model need to integrate theory, research and practice. This integration occurs at the level of principles that form the conceptual basis of therapeutic action. Effective therapy could not proceed without a guiding model, because it is the model that provides the conceptual basis for therapists to understand clients comprehensively and to make systematic immediate and long-term clinical decisions. As illustrated in Figure 1, the change model is brought into the therapeutic encounter by the therapist, the socially sanctioned professional who is charged with the responsibility of facilitating and managing therapeutic change. Therapists acquire the model through training and clinical experience yet, any well-established models also exist independent of individual therapists.

To be comprehensive, the change model is composed of two highly interdependent components: Principles that explain clients, therapy, and change, and a systematic clinical protocol followed by the therapist. The principles of a model answer the questions “How does the client function?” and “What is the nature of change?”, whereas the protocol answers the questions “What actions should I take?” and “When should I take them?” The conceptual principles of the model should be comprehensive theoretical articulations that explain how clients function, how psychological problems develop, explanations of how to help people change, and the interrelationship among these factors. Principles are the theory and conceptual foundation that are the basis of the therapist actions and the logic and reason for their clinical decisions. During therapy, the principles exist in the background as the basis of specific interventions and therapeutic actions that exist in the foreground. The clinical protocol delineates the specific time and place of therapeutic interventions within the process of therapy. The protocol encompasses the specific therapeutic goals and the clinically based change mechanisms that facilitate the successful therapy. The protocol remains in the foreground and functions as a map for the therapist that guides their clinical activities in a systemic way. The clinical protocol is determined by the principles (e.g., the nature of change) that specify and order the activities of the therapist within a temporal process of change.

Multilevel Change Processes

Therapeutic change occurs in three arenas: (a) Therapeutic process/procedures and consequent change mechanisms inherent in the model that are brought to therapy by the therapist, (b) the change experiences of the client(s) as they go through these therapeutic procedures, and (c) the immediate relational interactions between client(s) and therapist through which change takes place. The interaction among these processes is also critical. For example, the change mechanisms of the therapist change model will be conducted within the relational interaction and relationship between the client and therapist. The experiences and change processes of the client(s) will be altered by these interactions. The experiences of the client and the model of the therapist will affect the interaction and relationship between them. The directional arrows in Figure 1 illustrate an example of how the interactions among these processes might occur.

Therapeutic change process. By its nature, therapy is a temporal process in which certain activities, or change mechanisms, may have different places or importance depending on where in this process they occur (Kanfer & Shefft, 1988; Tracey & Ray, 1984). Therapists initiate and facilitate this process based on the principles and protocol they bring to therapy. Similarly, the process serves as the map that the therapist follows in facilitating therapeutic change. This temporal process has been described by some as phase-based (Alexander & Sexton, 2002; Rowe & Liddle, 2003) in which each phase usually has specific process goals and systematically applied change mechanisms that are pursued by the therapist. These process goals are much like the “little o” outcomes proposed by Pinsof and Catherall (1986) as critical to consider in thinking about change in MFT. Proximal or phase-based goals lead to longer term more broad client changes, described by Pinsof and Catherall as distal outcomes (big “O”). It is important to note that the goals suggested here are not client outcome goals but process goals that ultimately contribute to the client achieving the changes they seek.

A number of phase-based models both outside and from within MFT illustrate this process. For example, Kanfer and Schefft (1988) proposed a seven-phase process. The first two phases are concerned with client motivation and engagement. Kanfer and Schefft suggested that early phases of therapy do not focus on changing behavior but on setting the stage for change to occur by developing a working alliance and establishing client motivation to change. The middle phases are concerned with setting goals and making behavior changes. The final phase is concerned with generalization of skills, maintenance of skills in the absence of the therapist, and termination. Alexander and Sexton (2002) and Sexton and Alexander (2002) proposed a similar three-phase model involving engagement/motivation, followed by behavior change, ending with generalization. Change mechanism research in MFT provides a valuable contribution in this regard by suggesting that within these temporal stages or phases, there are specific critical mechanisms that promote the successful accomplishment of therapeutic process goals. Sexton, Robbins, et al. (2003) have identified a number of specific phase-based change mechanisms from the process research of MFT.

Within-client(s) change process. It is the client who seeks professional assistance for the purpose of improving couple, family, or individual functioning. Whether they are individuals, couples, or families, clients have varied clinical presentations, levels of severity, and degrees of motivation to change. Whether in therapy or not, clients experience changes that can also be described as a process that unfolds over time rather than as a static event (Prochaska, 2000). Various aspects of the onset, direction, and outcome of client change process are affected by their environmental and relational context (see Fig. 1). The change process experienced by the client is different than the “process” of therapy. Although successful, therapy should have an impact on the direction, intensity, pace, and/or outcome of client change. Thus the client and therapeutic change process are interdependent in successful therapy, and the phases of therapy initiated by the therapist should match the stage of change of the client (Prochaska, 2000).

The transtheoretical model proposed be Prochaska, DiClemente, and Norcross (1992) is an example of a systemtic articulation of the process clients(s) may go through when experiencing change. The model proposes that change is a spiral rather than a linear process in which individuals often go through periods of relapse in between phases of progression that can be captured by five stages of change: Precontemplation, contemplation, preparation, action, and maintenance. Prochaska (1999; 2000) has also discussed a sixth phase of change concerned with termination. Although not often applied to couple and family relational systems, the client change model of Prochaska and colleagues illustrates the construct of within-client change processes necessary to articulate in a comprehensive model of change.

Therapist-client interactional processes. Therapist-client interaction is the relational stage that brings together the principles and procedures of the therapist change model and the natural change processes of the client(s). The interaction between therapist and client (individual, couple, or family) is in some sense the place where the therapist adjusts and translates the constructs of the change model to the client’s personality, relational system, culture, and specific clinical problem(s). The complex and elaborate interactions that take place between therapist and client(s) are usually identified as the therapeutic relationship (Claiborn & Lictenberg, 1989; Sexton & Whiston, 1994). The qualities of the relationship and the dynamics that transpire between these participants have an important bearing on therapeutic change, because they serve as the forum in which and through which the professional expertise of the therapist and the personal experience of the client(s) interact. Butler and Strupp (1986) suggested moving away from identifying simplistic change ingredients and moving toward identifying principles of human interaction. Traditional descriptions of the therapeutic relationship are most often presented as descriptions of the necessary components and thus as static explanations (Bordin, 1979; Gelso & Carter, 1985; Horvath & Symonds, 1991). When viewed as a process, the therapeutic relationship represented by the interactions between therapist and client(s) can become a process through which change can be promoted rather than merely a factor in change.

Two existing models illustrate the interactional process of the therapeutic relationship. Using the early systemic constructs, Claiborn and Lichtenberg (1989) proposed an interactional model that purports client change results from the way in which the therapist and client negotiate their relationship. Tracey and Ray (1984) proposed a three-stage model of complementarity examining early, middle, and late stages of therapy. Bachelor and Horvath (1999) suggested that the client-therapist complementarity model is an important area of literature to examine. Alexander and colleagues (Alexander & Sexton, 2002; Sexton & Alexander 2003) have described an interactional model of family interaction that describes the development of therapeutic engagement and motivation as the result of within-family alliance and alliance between therapist and family.

CONCLUSIONS

In dynamic systems there are always choice points: decisions that occur along the way impacting the future direction (Alexander et al., 2002). Dynamic systems theory suggests that these choice-point decisions may be made through thoughtful reflection, or they may be made inadvertently and unintentionally. In either case, future directions do not simply unfold. They are determined by the choices made along the way. In both the theoretical and research domains, it is clear that the knowledge of MFT has evolved and expanded considerably since the founding days.

We have suggested that MFT is currently at an important choice-point in its dynamic evolution. The pressures for choosing a direction come from both inside and outside the profession. Internally, there are questions regarding the theoretical foundation and clinical basis of practice (e.g., how do we think about clients?; what is the nature of change?; what is the legitimate basis of therapy and of the profession?; what is a legitimate type and source of knowledge? and what are the most useful ways to teach and prepare new professionals entering the profession?). Externally, the struggle is one of accountability (e.g., are we the best/most efficient treatment for particular clinical problems?), services that will be reimbursed (e.g., is this a legitimate mental health service?), and the type of practices that will be supported in community systems of care (e.g., evidence based models of practice, integrated systems of care). Regardless of the source of the pressures, they are not easy and are often accompanied by debate and dissent. Many such debates currently occur between practitioners and researchers, between empiricists and constructivists, and between those who believe that the therapist is the expert and others who look to the family for guidance.

The debate over common factors is central to the current choice-point in MFT. There is no question that effective models of MFT should and will include the core common factors shared by the various approaches to psychotherapy. In its current form, the common factors perspective is, however, an inadequate solution to the current theoretical and research dilemma of MFT, because it does not integrate research into practice, provide a conceptual or theoretical foundation to understand clients or change, or provide guidance for the practitioner. There are many more dynamics operative in therapeutic change. Unfortunately, the common factors perspective is deceptively appealing in that it oversimplifies the complex processes of MFT, the important research needed to understand therapeutic change, and the theoretical richness required to explain the dynamics of the participants in therapy. We worry that the oversimplification inherent in the common factors perspective undermines the attempt to achieve best practices in the field that address complex and difficult clinical problems. Our analysis suggests that a more productive solution is to appreciate the complexity of MFT and to integrate research into practice by building comprehensive multilevel-process models of practice that provide a theoretical and conceptual foundation (through guiding principles) and that describe systematic clinical procedures that serve as the basis of practice.

The current debate around common factors feels quite familiar. Throughout the history of MFT, ideas have evolved in dynamic ways. It is no surprise that the field once again finds itself facing an important period of scholarly reflection and debate. It is unfortunate that the conversation often becomes a polemic of either/or choosing between common factors and specific clinical models rather than a search for an overarching conceptual schema that captures the complexity of therapeutic change. We believe that after scholarly reflection and debate we will find ourselves heeding Alexander et al.’s (2002) admonishment: savor the dialectic and embrace the unification and integration of common factors and the complexity of specific and different comprehensive models of change. We suggest that can be done when we view common factors as necessary (but not sufficient) ingredients within specific comprehensive models of therapeutic change that guide practice, build theory, and integrate research to promote successful outcomes.

REFERENCES

Alexander, J. F., & Barton, C. (1995). Family therapy research. In R. H. Mikesell, D. Lusterman, & S. H. McDaniel (Eds.), Integrating family therapy: Handbook of family psychology and systems theory (pp. 199-215). Washington, DC: APA.

Alexander, J. F, Holtzworth-Munroe, A., & Jameson, P. B. (1994). The process and outcome of marital and family therapy: Research review and evaluation. In A. E. Bergin, & S. L. Garfield (Eds.), Handbook of psychotherapy and behavior change (4th ed., pp. 595-630). Oxford, UK: Wiley.

Alexander, J. F., Newell, R. M., Robbins, M. S., & Turner, C. W. (1995). Observational coding in family therapy process research. Journal of Family Psychology, 9, 355-365.

Alexander, J. F., Pugh, C., Parsons, B., & Sexton, T. L. (2000). Functional family therapy. In D. Elliott (Series Ed.), Book three: Blueprints for violence prevention (2nd ed.). Golden, CO: Venture.

Alexander, J. F., & Sexton, T. L. (2002). Functional family therapy: A model for treating high-risk, acting-out youth. In F. W. Kaslow (Ed.), Comprehensive handbook of psychotherapy: Integrative/eclectic (Vol. 4, pp. 111-132). New York: Wiley.

Alexander, J. F., Sexton, T. L., & Robbins, M. S. (2002). The developmental status of family therapy in family psychology intervention science. In H. A. Liddle, D. A. Santisteban, R. Levant, & J. Bray (Eds.), Family psychology: Science-based interventions (pp. 17-40). Washington, DC: APA.

Aos, S., & Barnoski, R. (1998). Watching the bottom line: Cost-effective interventions for reducing crime in Washington. Washington State Institute for Public Policy: RCW 13.40.500.

Arkowitz, H. (1995). Common factors or processes of change in psychotherapy? Clinical Psychology: Science and Practice, 2, 94-100.

Bachelor, A., & Horvath, A. (1999). The therapeutic relationship. In M. A. Hubble, B. L. Duncan, & S. D. Miller (Eds.), The heart and soul of change: What works in therapy (pp. 133-178). Washington, DC: APA.

Becvar, D. S. (2003). Era of epistemology: A survey of family therapy thinking and theorizing. In T. L. Sexton, G. R. Weeks, & M. S. Robbins (Eds.), Handbook of family therapy: The science and practice of working with families and couples (pp. 3-20). New York: Brunner Routledge.

Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research, and Practice, 16, 252-260.

Brown, J. (1987). A review of meta-analyses conducted on psychotherapy outcome research. Clinical Psychology Review, 7, 1-23.

Butler, S. F., & Strupp, H. H. (1986). Specific and nonspecific factors in psychotherapy: A problematic paradigm for psychotherapy research. Psychotherapy, 23, 30-40.

Chamberlain, P., & Rosicky, J. G. (1995). The effectiveness of family therapy in the treatment of adolescents with conduct disorders and delinquency. Journal of Marital and Family Therapy, 21, 441-459.

Claiborn, C. D., & Lichtenberg, J. W. (1989). Interactional counseling. The Counseling Psychologist, 17, 355-453.

Dublin, R. (1978). Theory building (Rev. ed.). New York: Free Press.

Duncan, B. L., Miller, S. D., & Sparks, J. A. (2003). Interactional and solution-focused brief therapies: Evolving concepts of change. In T. L. Sexton, G. R. Weeks, & M. S. Robbins (Eds.), Handbook of family therapy: The science and practice of working with families and couples (pp. 101-124). New York: Brunner Routledge.

Elliott, D. S. (1998). Blueprints for violence prevention. University of Colorado, Center for the Study and Prevention of Violence. Boulder, CO: Blueprints.

Estrada, A. U., & Pinsof, W. M. (1995). The effectiveness of family therapies for selected behavioral disorders of childhood. Journal of Marital and Family Therapy, 21, 403-440.

Frank, J. D. (1971). Therapeutic factors in psychotherapy. American Journal of Psychotherapy, 25, 350-361.

Gelso, C. J., & Carter, J. A. (1985). The relationship in counseling and psychotherapy: Components, consequences, and theoretical antecedents. The Counseling Psychologist, 13, 155-243.

Gottman, J. M., Coan, J., Carrere, S., & Swanson, C. (1998). Predicting marital happiness and stability from newlywed interactions. Journal of Marriage and the Family, 60, 5-22.

Grencavage, L. M., & Norcross, J. C. (1990). Where are the commonalities among the therapeutic common factors? Professional Psychology: Research and Practice, 21, 372-378.

Gurman, A. S., & Kniskern, D. P. (1981). Family therapy outcome research: Knowns and unknowns. In A. S. Gurman & D. P. Kniskern (Eds.), Handbook of family therapy (Vol. 1, pp. 742-776). New York: Brunner/Mazel.

Gurman, A. S., & Kniskern, D. P., & Pinsof, W. M. (1986). Research on the process and outcome of marital and family therapy. In S. L. Garfield, & A. E. Bergin (Eds.), Handbook of psychotherapy and behavior change (3rd ed., pp. 565-624). New York: Wiley.

Hanna, F. J. (2002). Therapy with difficult clients: Using the precursors model to awaken change. Washington, DC: APA.

Hanna, F. J. (1996). Precursors of change: Pivotal points of involvement and resistance in psychotherapy. Journal of Psychotherapy Integration, 6, 227-264.

Hazelrigg, M. D., Cooper, H. M., & Borduin, C. M. (1987). Evaluating the effectiveness of family therapies: An integrative review and analysis. Psychological Bulletin, 101, 428-442.

Henggeler, S. W., & Borduin, C. M. (1990). Family therapy and beyond: A multisystemic approach to treating behavior problems of children and adolescents. Pacific Grove, CA: Brooks/Cole.

Henggeler, S. W., Schoenwald, S. K., Borduin, C. M., Rowland., & Cunningbam. P. B. (1998). Multisystemic treatment of antisocial behavior in children and adolescents. New York: Guilford Press.

Horvath, A. O., & Symonds, B. D. (1991). Relation between working alliance and outcome in psychotherapy: A meta-analysis. Journal of Counseling Psychology, 38, 139-149.

Hubble, M. A., Duncan, B. L., & Miller, S. D. (1999). Directing attention to what works. In M. A. Hubble, B. L. Duncan, & S. D. Miller (Eds.), The heart and soul of change: What works in therapy (pp. 407-447). Washington, DC: APA.

Kanfer, F. H., & Schefft, B. K. (1988). Guiding the process of therapeutic change. Champaign, IL: Research Press.

Kassel, J. D., & Wagner, E. F. (1993). Processes of change in Alcoholics Anonymous: A review of possible mechanisms. Psychotherapy, 30, 222-234.

Kazdin, A. E. (1997). Practitioner review: Psychosocial treatments for conduct disorder in children. Journal of Child Psychology and Psychiatry and Allied Disciplines, 38, 161-178.

Kuhn, T. S. (1964). The structure of scientific revolutions. Chicago: University of Chicago Press.

Lambert, M. J. (1992). Implications of outcome research for psychotherapy integration. In J. C. Norcross, & M. R. Goldfried (Eds.), Handbook of psychotherapy integration (pp. 94-129). New York: Basic Books.

Lebow, J. L. (2003). Integrative approaches to couple and family therapy. In T. L. Sexton, G. R. Weeks, & M. S. Robbins (Eds.), Handbook of family therapy: The science and practice of working with families and couples (pp. 201-228). New York: Brunner Routledge.

Lebow, J. L., & Gurman, A. S. (1995). Research assessing couple and family therapy. Annual Review of Psychology, 46, 27-57.

Lewin, K. (1951). Field theory in social science: Selected theoretical papers. Cartwright (Ed.). New York: Harper & Row.

Liddle, H.A., & Dakof, G. A. (1995). Efficacy of family therapy for drug abuse: Promising but not definitive. Journal of Marital and Family Therapy, 21, 511-544.

Ogles, B. M., Anderson, T., & Lunnen, K. M. (1999). The contribution of models and techniques to therapeutic efficacy: Contradictions between professional trends and clinical research. In M. A. Hubble, B. L. Duncan, & S. D. Miller (Eds.), The heart and soul of change: What works in therapy (pp. 201-226). Washington, DC: APA.

Orlinsky, D. E., & Howard, K. I. (1986). Process and outcome in psychotherapy. In S. Garfield, & L. A. Bergin (Eds.) Handbook of psychotherapy and behavior change (3rd ed., pp. 311-381). New York: Wiley.

Pinsof, W. M., & Catherall, D. R. (1986). The integrative psychotherapy alliance: Family, couple, and individual therapy scales. Journal of Marital and Family Therapy, 12, 137-151.

Pinsof, W. M., & Wynne, L. C. (1995). The effectiveness of marital and family therapy. Journal of Marital and Family Therapy, 21, 585-613.

Pinsof, W. M., & Wynne, L. C. (2000). Toward progress research: Closing the gap between family therapy practice and research. Journal of Marital and Family Therapy, 26, 1-8.

Prochaska, J. O. (2000). Change at different stages. In C. R. Snyder, & R. E. Ingram (Eds.), Handbook of psychological change: Psychotherapy processes and practices for the 21st century (pp. 109-127). New York: Wiley.

Prochaska, J. O. (1999). How do people change, and how can we change to help many more people? In M. A. Hubble, B. L. Duncan, & S. D. Miller (Eds.), The heart and soul of change: What works in therapy (pp. 227-255). Washington, DC: APA.

Prochaska, J. O., DiClemente, C. C., & Norcross, J. C. (1992). In search of how people change: Applications to addictive behaviors. American Psychologist, 47, 1102-1114.

Rescher, N. (1998). Complexity: A philosophical overview. New Brunswick, NJ: Transaction.

Robbins, M. S., Alexander, J. F, & Turner, C. W. (2000). Disrupting defensive family interactions in family therapy with delinquent adolescents. Journal of Family Psychology, 14, 769-998.

Rohrbaugh, M. J., Shoham, V., Trost, S., Muramoto, M., Gate, R. M., & Leischow, S. (2001). Couple dynamics of change-resistant smoking: Toward a family consultation model. Family Process, 40, 15-31.

Rosenthal, R. (1985). Meta-analytic procedures for social research. Newbury Park, CA: Sage.

Rosenzweig, S. (1936). Some implicit common factors in diverse methods of psychotherapy. American Journal of Orthopsychiatry, 6, 412-415.

Rowe, C. L., & Liddle, H. A. (2003). Substance abuse. Journal of Marital and Family Therapy, 29, 97-120.

Sexton, T. L., & Alexander, J. F. (2002). Family-based empirically supported interventions. The Counseling Psychologist, 30, 238-261.

Sexton, T. L. & Alexander, J. F. (2003). Functional Family Therapy: A Mature Model for working with at-risk adolescents and their families. In T. L. Sexton, G. R. Weeks, & M. S. Robbins (Eds.), Handbook of family therapy: The science and practice of working with families and couples (pp. 323-350). New York: Brunner Routledge.

Sexton, T. S., Alexander, J. F., & Mease, A. L. (2004). Levels of evidence for the models and mechanisms of therapeutic change in couple and family therapy. In M. J. Lambert (Ed.), Bergin and Garfield’s handbook of psychotherapy and behavior change (5th ed., pp. 590-646). New York: Wiley.

Sexton, T. L., Robbins, M. S., Hollimon, A. S., Mease, A. L., & Mayorga, C. C. (2003). Efficacy, effectiveness, and change mechanisms in couple and family therapy. In T. L. Sexton, G. R. Weeks, & M. S. Robbins (Eds.), Handbook of family therapy: The science and practice of working with families and couples (pp. 292-262). New York: Branner Routledge.

Sexton, T. L., Sydnor, A., & Rowland, M. (2003). The role of therapist model adherence and competence in successful outcomes of family-based interventions. Symposium presented at the American Psychological Association Annual Meeting. Toronto, Canada.

Sexton, T. L., Weeks, G. R., & Robbins, M. S. (2003). A new era of couple and family therapy. In T. L. Sexton, G. R. Weeks, & M. S. Robbins (Eds.), Handbook of family therapy: The science and practice of working with families and couples (pp. xxiii-xxx). New York: Brnnner Routledge.

Sexton, T. L., & Whiston, S. C. (1994). The status of the counseling relationship: An empirical review, theoretical implications, and research directions. Counseling Psychologist, 22, 6-78.

Shadish, W. R., & Baldwin, S. A. (2002). Meta-analysis of MFT interventions. In D. H. Sprenkle (Ed.), Effectiveness research in marriage and family therapy (pp. 339-370). Alexandria, VA: AAMFT.

Shadish, W., Montgomery, L., Wilson, P., Wilson, M., Bright, L, & Okwumabua, T. (1993). Effects of family and marital psychotherapies: A meta-analysis. Journal of Consulting and Clinical Psychology, 61, 992-1002.

Smith, M. L., Glass, G. V., & Miller, T. I. (1980). The benefits of psychotherapy. Baltimore: Johns Hopkins Press.

Snyder, C. R., Michael, S. T, & Cheavens, J. S. (1999). Hope as a psychotherapeutic foundation of common factors, placebos, and expectancies. In M. A. Hubble, B. L. Duncan, & S. D. Miller (Eds.), The heart and soul of change: What works in therapy (pp. 179-200). Washington, DC: APA.

Sprenkle, D. H. (2002). Editor’s introduction. In D. H. Sprenkle (Ed.), Effectiveness research in marriage and family therapy (pp. 9-25). Alexandria, VA: AAMFT.

Sprenkle, D. H., Blow, A. J., & Dickey, M. H. (1999). Common factors and other nontechnique variables in marriage and family therapy. In M. A. Hubble, B. L. Duncan, & S. D. Miller (Eds.), The heart and soul of change: What works in therapy (pp. 329-359). Washington, DC: APA.

Stanton, M. D., & Shadish, W. R. (1997). Outcome, attrition, and family-couples treatment for drug abuse: A meta-analysis and review of the controlled, comparative studies. Psychological Bulletin, 122, 170-191.

Szapocznik, J., & Kurtines, W. M. (1989). Breakthroughs in family therapy with drug abusing and problem youth. New York: Springer.

Tracey, T. J., & Ray, P. B. (1984). Stages of successful time-limited counseling: An interactional examination. Journal of Counseling Psychology, 31, 13-27.

U.S. Public Health Service. (2001). Youth violence: A report of the Surgeon General. Rockville, MD: Department of Health and Human Services, National Institutes of Health, National Institute of Mental Health.

Waldron, H. B. (1997). Adolescent substance abuse and family therapy outcome: A review of randomized trials. In T. H. Ollendick, & R. J. Prinz (Eds.), Advances in clinical child psychology (vol. 19, pp. 199-234). New York: Plenum Press.

Waltz, J., Addis, M. E., Koerner, K., & Jacobson, N. S. (1993). Testing the integrity of a psychotherapy protocol: Assessment of adherence and competence. Journal of Consulting and Clinical Psychology, 61, 620-630.

Wampold, B. E. (2001). The great psychotherapy debate: Models, methods, and findings. Mahwah, NJ: Erlbaum.

Thomas L. Sexton, Charles R. Ridley, and Amy J. Kleiner

Indiana University

Thomas L. Sexton, PhD, Charles, R. Ridley, PhD, and Amy J. Kleiner, MS, Department of Counseling and Educational Psychology, Indiana University.

Correspondence regarding this article should be addressed to Thomas L. Sexton, Counseling Psychology Program, Indiana University, 201 N. Rose Ave., Bloomington, Indiana, 47405. E-mail: thsexton@indiana.edu

Copyright American Association for Marriage and Family Therapy Apr 2004

Provided by ProQuest Information and Learning Company. All rights Reserved