The Importance of Theory in Cognitive Behavior Therapy: A Perspective of Contextual Behavioral Science

Available online at www.sciencedirect.com

ScienceDirect Behavior Therapy 44 (2013) 580–591

www.elsevier.com/locate/bt

The Importance of Theory in Cognitive Behavior Therapy: A Perspective of Contextual Behavioral Science

James D. Herbert Drexel University

Brandon A. Gaudiano Butler Hospital/Alpert Medical School of Brown University

Evan M. Forman Drexel University

For the past 30 years, generations of scholars of cognitive behavior therapy (CBT) have expressed concern that clinical practice has abandoned the close links with theory that characterized the earliest days of the field. There is also a widespread assumption that a greater working knowledge of theory will lead to better clinical outcomes, although there is currently very little hard evidence to support this claim. We suggest that the rise of so-called “third generation”models of CBT over the past decade, along with the dissemination of statistical innovations among psychotherapy researchers, have given new life to this old issue.We argue that theory likely does matter to clinical outcomes, and we outline the future research that would be needed to address this conjecture.

Keywords: theory; cognitive behavior therapy; acceptance and commitment therapy; contextual behavioral science

There is nothing so practical as a good theory. — Lewin (1951, p. 169)

For years, scholars of the family of psychotherapy approaches known under the broad umbrella of

Address correspondence to James D. Herbert, Ph.D., Drexel University, Department of Psychology, 3141 Chestnut St., Stratton 119, Philadelphia, PA 19104; e-mail: james.herbert@drexel.edu. 0005-7894/44/580-591/$1.00/0 © 2013 Association for Behavioral and Cognitive Therapies. Published by Elsevier Ltd. All rights reserved.

cognitive behavior therapy (CBT) have been calling for an increased focus on the theories that underlie applied technologies. The common theme of these appeals is that there has been a gradual erosion of the strong connection between theory and technique that characterized the field’s early days, and that a renewed focus on such links will lead to more rapid and reliable advances in our understanding, devel- opment, testing, implementation, and dissemination of CBT approaches. In his 1984 presidential address of the Association for Advancement of Behavior Therapy (now the Association for Behavioral and Cognitive Therapies; ABCT), Alan Ross lamented that “a reading of the current literature on behavior therapy suggests that the field is at risk of losing its momentum in a preoccupation with technological refinements at the expense of theoretical develop- ments” (Ross, 1985, p. 195). Wilson and Franks (1982) similarly decried the rapid proliferation of clinical techniques decoupled from theory, suggest- ing that this trend could ultimately sow the seeds of the field’s demise. More recently, Beck (2012) noted that “. . . the robustness of a therapy is based on the complexity and richness of the underlying theory. A robust theory, for example, can generate new therapies or can draw on existing therapies that are consistent with it” (p. 6). David and Montgomery (2011) proposed a new framework for defining evidence-based psychological practice that pri- oritizes the level of empirical support of the theoryhttp://dx.doi.org/http://dx.doi.org/http://dx.doi.org/mailto:james.herbert@drexel.edu

581theory in cbt

supporting a treatment. Recommendations that clinicians should develop better working knowledge of the theories underlying CBTs often are presented during discussions of how to maximize treatment outcomes, prevent treatment failures, and ameliorate treatment resistance in complex cases (Foa & Emmelkamp, 1983;McKay, Abramowitz, & Taylor, 2010; Whisman, 2008). An interorganizational task force led by the ABCT recently issued a report on doctoral training in cognitive behavioral psychology inwhich training in theory and even thephilosophyof science underlying CBTs was emphasized (Klepac et al., 2012). The call for greater emphasis on theory within

CBT therefore spans the generations. In fact, if one were tomask the author anddate, itwould be hard to distinguish writings on this subject made by contem- porary authors from thosewritten over 30 years ago. There appears to exist a widespread assumption among many clinicians and researchers alike that better knowledge of theory will bear fruit in terms of improved clinical outcomes across a number of con- texts. Although this notion has considerable face validity, there is a paucity of research that has directly evaluated it. Historically, the desire for empirically supported

treatments led to testing psychotherapies in controlled clinical trials to determine their efficacy, a procedure borrowed from other medical treatments. For example, the seminal study known as the National Institute ofMental Health’s Treatment of Depression Collaborative Research Program (Elkin et al., 1995) randomized patients with major depression to cog- nitive therapy, interpersonal psychotherapy, or anti- depressant medication, and ushered in a new era of evaluating psychotherapies in large-scale and meth- odologically rigorous clinical trials. CBTs, given their empirical basis, inherent structure, and time- limited nature, were particularly well-suited for testing in clinical trials. As a result, CBTs became highly manualized in an effort to ensure treatment fidelity, an important component of the internal validity of such trials (Addis & Krasnow, 2000). Originally CBTs were more principle-driven and theory-dependent in the way that they were concep- tualized and implemented (e.g.,Goldfried&Davison, 1994). With the growth of clinical trials during the 1970s and 80s, however, treatment manuals began to focus more on how to implement specific CBT techniques and strategies and less on interventions derived from case conceptualization based on the ideographic assessment of the patient guided by an underlying theory. We are unaware of data directly comparing the level of theoretical knowledge of early practitioners of behavior therapy relative to modern CBT clinicians. Nevertheless, even a casual compar-

ison of the field’s early books and journals targeting clinicians relative to later works reveals a stark contrast in the degree of emphasis on theory. As the evidence base for CBTs expanded due to the

rapid accumulation of supportive efficacy research, the problem of how best to implement and dissem- inate the treatments emerged as a pressing problem (Addis, 2006). Although novel psychotherapies typically begin in complex and sophisticated forms because they are created by experienced researchers and clinicians, disseminating them to community practitioners exerts pressure to simplify them as much as possible. It is easier to train nonexpert therapists to implement a set of standard techniques than it is to train them to comprehend an underlying theory. Once standard techniques are mastered, clinicians well versed in theory can potentially apply their knowledge to unique cases in order to deduce tailored interventions. The picture is complicated further because there is

no single CBT model, nor single theory underlying it. CBT is a broad umbrella term that encompasses a range of distinct therapy models (Herbert & Forman, 2011). These models share certain features, while also havingdistinct characteristics. The theories underlying these approaches likewise share certain commonalities (e.g., traditional respondent and operant conditioning principles), while also positing unique features.More- over, key theoretical issues, such as the best way to understand the role of cognitive processes in treat- ment, are currently the subject of intense professional debate (Hofmann, 2008; Longmore&Worrell, 2007; Worrell & Longmore, 2008), and have undergone considerable changes over the years (Beck, 2005). We believe that two developments over the past

decade have added a new twist to the long-standing question about the role of theory in guiding psychotherapy. First, the question has been reinvigo- rated by the rise of the so-called “third wave” (also known as “third generation”) models of CBT. These newer CBT approaches such as Mindfulness-Based Cognitive Therapy (Segal, Williams, & Teasdale, 2002), Dialectical Behavior Therapy (Linehan, 1993), and especially Acceptance and Commitment Therapy (ACT; Hayes, Strosahl, & Wilson, 2011) eschew a simplistic focus on specific techniques and strategies in favor of increased attention to the putative principles underlying behavior change, which are in turn linked with basic psychological theories (Ablon, Levy, & Katenstein, 2006; Hayes, 2004; Rosen & Davison, 2003). Second, psychotherapy treatment researchers have increasingly focused on therapy processes using component analysis studies (Borkovec & Sibrava, 2005; Lohr, DeMaio, & McGlynn, 2003) and the identification of treatment-related mediators and moderators (Kraemer, Wilson, Fairburn, & Agras,

582 herbert et al .

2002). These two developments have had synergistic effects, further stimulating discussion of the role of theory in CBTs. For example, calls by proponents of third-generation approaches to focus on psycho- therapy processes, rather thanmerely techniques, has accelerated research on the mechanisms of action in CBTs more generally. Simultaneously, more accessi- ble and advanced statistical procedures have made it easier for researchers to investigate mechanisms of change, and have stimulated therapy innovators to evaluate the proposed theories underlying their approaches. Although the argument that under- standing theory will improve clinical outcomes has been a perennial theme in the field, innovations associated with the development of third-generation models of CBT, along with the development of new statistical tools, have brought this issue back into the forefront of discussion. We should not lose sight of the fact that proponents

of the utility of theory are often themselves theoreti- cians and may thus overestimate the importance of theory. The extent to which improving clinicians’ theoretical knowledge does, in fact, result in improved clinical outcomes is ultimately an empirical question. The best approach to evaluating this supposition is itself complicated and will require clarification of a number of related issues. First and foremost, are there compelling reasons to hypothesize that knowledge of theory will, in fact, improve outcomes? Second, what evidence, if any, currently supports the notion that gaining a better theoretical understanding of a psychotherapeutic approach enhances outcomes over and above mere technical knowledge of the approach? Third, even if theoretical knowledge is found to accrue incremental benefits, does it pass a cost-benefit test? Fourth, if such efforts can indeed be demonstrated to be cost effective, how much theoret- ical knowledge and training is needed to improve outcomes? Fifth, which theory or theories should be prioritized? Presumably, some theories have greater breadth, depth, precision, explanatory power, and incremental efficacy than others, making them more useful guides. Sixth, to what degree is it necessary simply to understand theory in abstract terms versus being able to apply it to individual cases?And seventh, and perhaps most fundamentally, what exactly do we mean by the term “theory”? We will briefly explore these and related questions regarding the role of theory, using ACT in particular as a case in point.1

1We focus on ACT as the prototypical “third-generation” model of CBT for two reasons. First, it has received the most research attention to date of these various approaches. Second, as elaborated below, it is based on a well-developed theory, and it strongly emphasizes the link between theory and technique. The emphasis on ACT is not meant to imply that other approaches are not also theoretically grounded.