Thursday, March 12, 2009

Narrative and External Validity in Hermeneutic Single Case Efficacy Design

Entry for 12 March 2009:

Note. The good news was that this week, after many years’ delay, the paper on the adjudicated version of the HSCED method was finally accepted for Psychotherapy Research by Clara Hill. The bad news was that she made us cut another 10+ pages from it, including the abridged version of the judges’ opinions that we’d sweated over last January and a cool bit of the Discussion that she didn’t think fit that well with the rest. She’s probably right about it not fitting that well, but I think it’s nice enough and of general interest enough to post here as a blog entry, because it’s where I really develop the idea of a narrative approach to HSCED. And because it’s about the method in general rather than about the particular case, it makes sense out of context. So here is another outtake from “An Adjudicated Hermeneutic Single Case Efficacy Design Study of Experiential Therapy for Panic/Phobia”:

In the course of carrying out this study, we (and our audiences) struggled with the large amount of often-contradictory information generated by the HSCED method. As a result, we began to develop a more narrative view of HSCED. It now seems to us that making causal inferences in general is fundamentally a narrative process. That is, explaining something that has happened, such as the outcome of a client’s therapy, involves constructing a story about how and why it came about. HSCED is therefore a narrative research method. As we now see it, the function of the Affirmative and Skeptic teams is to create persuasive narratives that express a particular view of the data. In addition, the judges each construct their own narratives as they seek to make sense of the case from their own theoretical perspective. These competing narratives are part of a dialectic process that serves to clarify our understanding of whether and how the client changed and what brought any changes about. The goal is not necessarily to construct a single master narrative to bind together all the other narratives, but rather to highlight the main plausible, coherent accounts and the remaining ambiguities. Systematic data collection, argumentation and judging processes offer reasonable methods for eliminating error and estimating where the weight of the evidence falls.

Viewing HSCED as a narrative research method also helps us to think about the perennial issue of generalizing from single cases (cf. Stiles, 1994). It now seems to us that a good narrative provides two pillars for inferring generalizability: First, it supplies orienting contextual information (moderator variables) that helps us judge the relevance and applicability of the results to other, similar cases. In particular, it is important that the formulation include what we think it was that the client brought to therapy that made change possible, as we have done for George. Second, the description of the sequence of primary change processes tells us what in the therapy was relevant to change, that is, the active mediator variables. This allows us to reframe the Kiesler (1966)/Paul (1967) specificity “litany” for single case research: “When a client with these problems, those resources and this history experiences these in-therapy change processes, then this kind of change is possible/likely.” Thus, effective affirmative researcher-constructed narratives and judges’ opinions should include views about what the client brought to therapy that was relevant to his/her successful therapy. (Skeptic teams might similarly try to account for the apparent failure of the therapy.) Establishing this for a single case shows what is possible; in legal terms, it establishes a precedent. Identifying the same factors in multiple similar cases through replication increases the likelihood the same relationships will hold in future similar cases, as well as clarifying further conditions under which this is likely or unlikely to be the case (cf. Sidman, 1960).

Thus, along with Fishman (1999), Bohart and Boyd (1997), Miller (2004) and Stiles (2007), we are proposing a science of single cases, using the quasi-judicial methods illustrated here. Such single case causal studies can be done in standalone fashion as was done here, but they can (and should) also be carried out within group designs to help researchers understand their results, essentially filling in the “black box” in a randomized clinical trial. To those familiar with Sidman’s (1960) classic work in behavioral single case experimental design, this may be seen as coming full circle, but with a broader range of methods that apply to a wide range of therapy approaches.

Furthermore, what do we mean by “therapy” when we make the inference that therapy was a major influence on client change? Surely we are not referring only to the therapist’s interventions but presumably mean the partnership of both people working together. Therefore, “therapy” here refers not only to the opportunities that the therapist offers the client, but also the opportunities the client offers the therapist, and also how client and therapist make use of those opportunities within therapy. Although the driving questions used here in HSCED are phrased in terms that sound like old-fashioned linear causality, it does not take the method long to uncover multiple sets of complex bidirectional causal processes in which client and therapist offer each other opportunities to work together productively within the context of the client’s larger life, which itself continually interacts with therapy process and outcome, with both immediate and delayed effects on the client’s problems and life functioning. From the point of view of HSCED, we now believe that it is enough for therapy (and the therapist) to play an important, but by no means exclusive, role in the client’s continuing change process. Interpretive single case methods such as HSCED offer the possibility of beginning to specify such complex, nonlinear interactive processes.


Bohart, A.C., & Boyd, G. (Dec., 1997). Clients' construction of the therapy process: A qualitative analysis. Poster presented at meeting of North American Chapter of the Society for Psychotherapy Research.

Fishman, D.B. (1999). The case for pragmatic psychology. New York: New York University Press.

Kiesler, D.J. (1966). Some myths of psychotherapy research and the search for a paradigm. Psychological Bulletin, 65, 110-136.

Miller, R.B. (2004). Facing Human Suffering: Psychology and Psychotherapy as Moral Engagement. Washington, DC: APA.

Paul, G.L. (1967). Strategy of outcome research in psychotherapy. Journal of Consulting Psychology, 31, 109-118.

Sidman, M. (1960). Tactics of scientific research. New York: Basic.

Stiles, W.B. (1994). What do you know once you’ve heard a story? Paper presented at meeting of Society for Psychotherapy Research, York, England.

Stiles, W. B. (2007). Theory-building case studies of counselling and psychotherapy. Counselling and Psychotherapy Research, 7, 122-127.

No comments: