Abstract
In this invited commentary, I provide an historical and conceptual context for “The Future of Intervention Science: Process-Based Therapy” by Stefan Hofmann and Steven Hayes (this issue, p. 37). Special attention is paid to the importance of change mechanisms in psychological intervention, the use of functional analysis instead of categorical schemes such as the Diagnostic and Statistical Manual of Mental Disorders and treatment manuals in randomized clinical trials, and the challenge of applying nomothetic principles to individual cases. Historical context is provided for prior examination of these core issues in numerous articles and books extending back for over 60 years.
In their article, “The Future of Intervention Science: Process-Based Therapy,” Stefan Hofmann and Steven Hayes (2019), two highly respected clinical researchers, discuss a number of very important issues in contemporary clinical psychology and offer some interesting ideas on potentially useful approaches that research and scholarship should focus on. At the same time, most of what they write should, I respectfully opine, be viewed more as a restatement of previous scholarship than novel and innovative. For someone like myself who values and finds it professionally necessary to contextualize ideas proposed as new and even revolutionary, one can discern instances in which there is more restating of prior scholarship than offering novel perspectives, namely, the “breakthrough-ism” arising from the ahistorical perspectives of many younger psychologists (Lilienfeld, 2017).
My colleagues do give a nod to earlier work (“Our argument is not new. In fact it brings us back to the very beginning of behavior therapy and its foundational element — functional analysis”), but the thrust of their article is that they are proposing very new ideas. This can be seen in their abstract: Clinical science seems to have reached a tipping point. It appears that a new paradigm is beginning to emerge that is questioning the validity and utility of the medical illness model, which assumes that latent disease entities are targeted with specific therapy protocols. (p. 37)
Simply put, there is little that is new in their article. A cumulative science requires an appreciation and fulsome analysis of prior contributions.
This is not to detract from the value of their exposition, only to argue that it should be more clearly placed in historical and scholarly context. While it can be somewhat awkward if some of that context includes publications of one’s own, a forthright commentary can require a suspension of eschewing self-citation. It is not that there is nothing new under the sun, only that sometimes shiny objects have been there a long time and are therefore not as shiny as one might think.
Principles of Change/Change Mechanisms
The principal thesis of Hofmann and Hayes is that research in psychosocial interventions has concerned itself almost entirely with the comparison of treatment packages in their efficacy and effectiveness in alleviating psychological problems. I agree with their concern and am in good company. Previous writers argued many years ago that such a focus takes clinical psychological scientists away from what could and perhaps should be our primary mission, namely, to explain why a given intervention effects beneficial change (e.g., Bandura, 1969; Davison, 1994, 1997, 2000; Davison, Goldfried, & Krasner, 1970; Goldfried, 1980; Rosen & Davison, 2003). Indeed, the very title of Bandura’s (1969) classic book basically says it all: Principles of Behavior Modification. I went further almost 20 years ago in proposing a research strategy that turns therapy research on its head: Several years ago I commented on the role of basic research in clinical psychology (Davison, 1994) and had occasion to develop the argument further during a conference sponsored by the National Institute on Drug Abuse (NIDA) concerned with untapped opportunities to use basic research in developing clinical procedures de novo (Davison, 1997). Simply put, searching for change mechanisms in existing effective techniques is to work after the fact, and although such process research is very important . . . , working in the other direction may be even better . . . : Moving from experimentally established principles of change to novel and effective clinical application . . . is an inadequately explored strategy for developing new therapeutic procedures that, from the outset, will have known mechanisms of change (because such research begins with principles of change. (Davison, 2000, p. 581)
Functional Analysis Versus the Diagnostic and Statistical Manual of Mental Disorders and Treatment Manuals
Hofmann and Hayes restate the view that what is nowadays called intervention science (itself an unfortunate term, implying that the scientific study of intervention is new) has for too long been locked into the use of the Diagnostic and Statistical Manual of Mental Disorders (DSM) to create putatively homogeneous groups (e.g., people “having” “major depressive disorder” or “generalized anxiety disorder”) and then assigning them randomly to a putatively effective or efficacious treatment group and to one or more control groups. This randomized clinical trial (RCT) research paradigm was modeled after drug research and became the “gold standard” as a result of the classic Elkin, Parloff, Hadley, and Autry (1985) comparison of a treatment package called cognitive therapy with “interpersonal therapy” and two other comparison conditions. Noteworthy in this large and ambitious study was the use of treatment manuals that prescribed or strongly constrained the behavior of clinicians in each of the conditions, thus giving strong impetus to the widespread use of treatment manuals. This research paradigm has held sway for almost 40 years, due perhaps to its designation as a necessary methodology by major funding agencies.
This is all well and good because even undergraduates are taught that it is crucial to operationally define one’s independent variable, else one’s experiment diminishes in internal validity. But in my view, which I believe Hofmann and Hayes agree with, there have always been two central problems with this methodology as implemented in RCTs of psychosocial interventions: (a) the assumption that DSM categories are sufficiently homogeneous to be the basis for randomization of participants, even when experimenters exclude patients carrying other diagnoses, so-called comorbid disorders; and (2) the sometimes major constraints placed on the study’s clinicians in how they interact with participants, referred to by Haaga and Stiles (2000) as the ballistic nature of treatment manuals, something that makes sense in ensuring the integrity of the independent variable but that is (properly) vitiated by the essentially interactive nature of psychological intervention.
The suggestion that Hofmann and Hayes make to use functional analysis rather than categorical diagnostic schemes like the DSM is welcome but, again, has been around for a long time. It hearkens back to the assessment approach proposed earlier by people like Staats and Staats (1963), Kanfer and Saslow (1965), Mischel (1968), Goldfried and Pomeranz (1968), Goldfried and Kent (1972), and Goldfried and Davison (1976, 1994). It may be of interest to note that the latter writers proposed a functional analytic scheme that does not involve DSM or other category-based “illness” nosology at all. Specifically, behavior targeted for change could, we argued, be conceptualized in terms that, unlike the case with the DSM, can be useful in the design of effective interventions. For example, clinical problems can be seen as difficulties in the stimulus control of behavior, deficient behavioral repertoires, aversive behavioral repertoires, difficulties with incentive systems (reinforcers), and aversive self-reinforcing systems (Goldfried & Davison, 1976). Space limitations preclude a full description and critical analysis of this proposal, but the point is that behavioral functional analysis for clinical purposes has been around since at least the 1960s and has been in use by many of us instead of more conventional DSM (type) category systems.
Nomothetic and Idiographic
Related to the foregoing, Hofmann and Hayes are correct in calling attention to the difference between the nomothetic and the idiographic, namely, the scientific focus on general principles and the analysis of individual cases. This distinction and the challenge that clinical work entails in blending these foci was of course highlighted many years ago by Gordon Allport (1937) in his classic book, Personality: A Psychological Interpretation. The intensive study of the single case was articulated by one of the very earliest behavior therapy researchers, Monty Shapiro (1957). With the growth of RCTs, especially in cognitive-behavioral comparative outcome research, this distinction was, I believe, lost in the excitement that the paradigm was generating in our efforts to be regarded as scientific. Lazarus and I focused on nomothetic-idiographic challenges in our 1971 chapter in Bergin and Garfield’s Handbook of Psychotherapy and Behavior Change (Lazarus & Davison, 1971), elaborating on the issue in subsequent publications (Davison & Lazarus, 1994, 1995). In other words, the issue is not new.
Concluding Comment
Sometime in the late 1970s, during one of my last years at Stony Brook, I gave the following take-home final exam to first-year clinical students in my psychological intervention class. It went something as follows: Set aside practical constraints and select any behavior change principle and then create a new clinical procedure that is based on this principle. I recall the students laboring over this assignment during exam period and then coming up with some creative and thought-provoking proposals. It is probably good that Hofmann and Hayes have revisited this issue and the others discussed previously, as these are useful to be reminded of.
Footnotes
Action Editor
Scott O. Lilienfeld served as action editor for this article.
Author Contributions
G. C. Davison is the sole author of this article and is responsible for its content.
Declaration of Conflicting Interests
The author(s) declared that there were no conflicts of interest with respect to the authorship or the publication of this article.