|
Article Excerpt
THE CONSTRUCT OF DIRECTIVENESS historically has been given multiple meanings in the psychotherapy and counseling fields. Early psychotherapy research characterized directiveness as leading the client (Ashby et al., 1957) or controlling the agenda during treatment sessions (Rogers, 1942). Later work characterized directiveness as authoritarianism (Rudolph, 1989), teaching and confronting (Patterson and Forgatch, 1985), and the level of therapist activity (Lafferty et al., 1989). Many genetic counselors consider giving advice to be a directive intervention (Bartels et al., 1997). Although similarities can be seen across these definitions in that they reflect aspects of therapist influence, it is also apparent that there is considerable variability in how therapist directiveness can be conceptualized.
In the field of alcohol treatment research, recent work has addressed the role of directiveness in treatment effectiveness. Specifically, research on behavioral treatments for alcoholism has found significant interactions between therapist directiveness and levels of patient anger (Karno and Longabaugh, 2004) and patient reactance (i.e., the tendency to resist relinquishing control in interpersonal interactions; Karno and Longabaugh, 2005). Karno and Longabaugh (2004) studied data from a single clinical research unit of Project MATCH (Matching Alcoholism Treatments to Client Heterogeneity) and found that, for patients high in anger, as measured by the Trait Anger Scale (Speilberger et al., 1983), more directive therapy was associated with more frequent drinking posttreatment. However, for patients low in anger, more directive therapy was associated with less frequent drinking. These results suggested that directiveness was helpful for patients low in anger yet was contraindicated for patients high in anger. In a separate analysis with the same sample, Karno and Longabaugh (2005) found that patients who were high in reactance, as measured by the Reactance subscale of the Clinician Rating Form (Fisher et al., 1999), drank more frequently following more directive treatment, while directiveness did not predict drinking outcomes for patients low in reactance. Thus directiveness was contraindicated for reactant patients but did not affect patients low in reactance.
Given the paucity of significant patient-treatment matching effects observed in the original Project MATCH study (Project MATCH Research Group, 1997a,b), the emergence of these two interactions, both involving therapist directiveness, is surprising and encouraging. The results do, however, beg the question about the extent to which the interactions were unique effects or expressions of a shared, underlying effect. Follow-up analyses found that anger and reactance were only modestly correlated (r = .35). This correlation indicates that the anger and reactance variables measured fairly distinct constructs. Hence, there is a good possibility that the interaction of each of those variables with therapist directiveness explained unique variance in drinking outcomes. However, this conclusion cannot be made without a formal test for the independence of each interaction.
Moreover, even if the interactions are found to be independent of each other, the question would remain whether angry and reactant patients responded to the same aspects of directiveness. The earlier studies by Karno and Longabaugh (2004, 2005) used a global directiveness rating that reflected an average across various items assessing therapist interpretation, confrontation, agenda setting, closed-ended questions and teaching (Fisher et al., 1995). That approach did not permit the analysis of separate aspects of directiveness. It may well be that breaking the directiveness measure down into its component parts would further our understanding of how the interactions are similar and/or unique. Further, this approach could help define more precisely the construct of directiveness in the research literature.
The purpose of the present study is to address these questions in three steps. First, the study provides a formal test for the independence of the interactions between therapist directiveness and the patient variables of anger and reactance. We hypothesize that each interaction will predict unique variance in drinking outcomes while controlling for the other. Second, the study presents an exploratory analysis of the components of directiveness based on the data collected in earlier work by Karno and Longabaugh (2004, 2005). Third, the study tests for the interactions of those components with both patient anger and reactance. We hypothesize that patient anger and reactance will each interact with different components of therapist directiveness.
It is expected that this work will refine our understanding of the directiveness construct and will allow us to assess more fully the similarities and differences in these patient-treatment interactions. Such clarification may well advance patient-treatment matching theory, help guide future research and offer clinicians specific recommendations for treatment practice.
Method
Participants
The sample (N = 139) all received treatment through the Providence Clinical Research Unit (CRU) of Project MATCH. All patients met criteria for a Diagnostic and Statistical Manual of...
|
