|
Article Excerpt
Qualitative research methods were used to elicit master therapists' statements regarding their use and understanding of the therapy relationship. The master therapists were identified and recruited in a previous study (Jennings & Skovholt, 1999) through a procedure used to create a sample of information-rich cases. The result of the analysis is a Model of Relationship Stances. The Safe Relationship Domain is composed of three categories of therapist actions: Responding, Collaborating, and Joining. The Challenging Relationship Domain also is composed of three categories of therapist actions: Using Self, Engaging, and Objectivity. The domains and categories are conceptualized as relationship stances utilized by the master therapists to meet individual client needs.
**********
The long war among theoretical orientations in counseling and psychotherapy has come to a truce. Numerous studies have found that most approaches can produce positive change, but a consistent superiority of one approach has not been found. Instead, much of what is effective in psychotherapy is due to pantheoretical or common factors, those shared by many schools of psychotherapy (Asay & Lambert, 1999).
Although the most important pantheoretical variable appears to be the client (Tallman & Bohart, 1999), the therapist is another promising variable. There is now strong evidence for the therapist effect on client outcome. Luborsky, McLellan, Woody, O'Brien, and Auerbach (1985) found that significant therapy success was determined most by a helpful relationship with the therapist. In research studies, the contribution by the therapist surpassed all but the contribution by the client (Teyber & McClure, 2000). In an exhaustive review of the research, Wampold (2001) affirmed the neglected but critical therapist effect:
We have seen that the particular treatment that the therapist delivers does not affect outcomes. Moreover, adherence to the treatment protocol does not account for the variability in outcomes. Nevertheless, therapists within treatment account for a large proportion of the variance. Clearly, the person of the therapist is a crucial factor in the success of therapy. (pp. 20-21)
A highly effective therapist does seem to make a difference. We know this implicitly as people and consumers when we actively search for a really good practitioner: a doctor, dentist, attorney, or therapist. We know that some are better than others. Yet, this notion of exploring the therapist variable has been less of a focus in contemporary research.
Reviewing the research that pointed out the critical importance of the therapist and the therapeutic relationship, Teyber and McClure (2000) emphasized focusing future research on factors that enhance the therapist-client relationship. The authors stated, "In many studies, what therapists say and do in the therapy hour that promotes a good working alliance has proven to be the most important contributor to change and positive treatment outcome" (p. 70). Therefore, the present research investigated the working alliance as constructed by a sample of those considered the "best of the best" by their professional colleagues, a group of master therapists.
METHODOLOGY
We used a qualitative interview method to investigate therapist contributions to the therapy relationship because in-depth interviews with practicing therapists can provide data for variables that have yet to be identified in the research literature. When the purpose of a study is to describe clinically relevant therapist modes of perceiving and acting, a discovery-oriented approach is deemed most suitable (Marshall & Rossman, 1999). We were interested in what master therapists could tell us about their construction and use of the therapy relationship. In the pursuit of this research question, information-rich informants provided the kind of data this investigation sought.
Respondents
Recruitment. A sample of information-rich cases, a master therapist group, was described in Jennings and Skovholt's (1999) study of the personal characteristics of master therapists. In their study, peer nomination was combined with snowball sampling to find participants. Anastasi and Urbina (1997) supported the psychometric quality of peer nomination, concluding that it is a reliable technique that has been used in a variety of settings. They stated, "When checked against a variety of practical criteria dependent on interpersonal relations, such ratings [peer nomination] usually have been found to have good concurrent and predictive validity" (p. 468).
First, three senior and respected psychotherapists were each asked to nominate three psychotherapists whom they considered to be master therapists based on a list of written criteria. The criteria were as follows:
1. This person is considered to be a "master therapist."
2. This person is most frequently thought of when referring a close family member or dear friend to a psychotherapist considered to be the "best of the best."
3. Because of this person's superior skills as a psychotherapist, one would have full confidence in seeing this psychotherapist for one's own personal psychotherapy. Therefore, this psychotherapist might be considered a "therapist's therapist."
The therapists who received nominations were contacted and asked to nominate three therapists according to the same criteria. This snowball sampling was continued until a select few therapists were repeatedly nominated. A total of 103 psychotherapists were nominated. Seeking a balance between depth and breadth of data, Jennings and Skovholt (1999) decided that four or more nominations was a useful cutoff point, creating a master therapist group of 10 therapists.
In the present study, the 10 master therapists were initially contacted through a letter. A week after the letters were sent out, the first author phoned the therapists and scheduled interviews. All 10 therapists agreed to participate.
An interview-follow-up format was used in the present study. This design enables the respondents to participate in the analysis as derived from the initial interviews. In this way, a "validity through dialogue" (Skovholt & Ronnestad, 1995, p. 147) is established.
Respondent characteristics. The master therapist group (i.e., 7 women and 3 men) included 1 psychiatrist: 3 master's level social workers, one of whom is also a master's level psychologist; and 6 Ph.D. psychologists. Their age range was 53 to 75 years, with an average age of 62 years. The years of therapist experience ranged from 24 to 44 years, with an average of 32.5 years. Theoretical orientations of the therapists were psychodynamic (n = 4), integrative (n = 2), family systems (n = 2), existential (n = 1), and Gestalt and client-centered (n = 1). All of the therapists were Caucasian.
Trends evident in therapists' descriptions of their clients indicated that, early in their careers, these therapists worked with a wider range of clients. Now they were more selective in their clientele. Some clients were other practicing therapists. Generally, these therapists saw clients who were functioning well in some areas of life but not so well in other areas. Diagnostically, depression and anxiety were the most frequently mentioned disorders. This group of therapists generally did not see clients who had an active addiction or were psychotic.
Instrument
The first author developed an initial set of questions after a review of the literature on therapy relationships, particularly the therapy alliance (Bachelor, 1991; Bordin, 1979, 1994; Gaston & Marmar, 1994; Henry & Strupp, 1994; Horvath, 1994; Horvath & Greenberg, 1994; Luborsky, 1994; Marziali, 1984). The second author reviewed and amended the initial set of questions. The first author then further refined the questions by conducting two pilot studies. The content of the questions reflects concern with clinical relevance while also incorporating aspects of alliance theory. Table 1 shows the correspondence of interview questions with alliance concepts.
Procedure and Data Analysis
The face-to-face interviews occurred over a 2-month period. The audiotaped interviews averaged 90 minutes. At the beginning of the interview, the purpose of the study was explained, and an informed consent form was obtained from each respondent. One of the respondents had to be reinterviewed because of an audiotape error. The interviewer checked the verbatim transcriptions for accuracy.
An inductive analysis procedure described by Patton (2002) and Jennings and Skovholt (1999) was utilized in this study. The data analysis proceeded inductively from the smallest unit, concepts, to themes, then to categories, then to the largest unit, domains. A concept comes from analysis of a small part of the transcript, such as a sentence or a paragraph answer supplied by a respondent. All the concepts then are examined and ordered for the purpose of identifying themes. Themes are identified by organizing a variety of concepts to make a coherent whole. Next, themes are examined for the purpose of identifying categories. Categories organize a variety of themes. Finally, the categories are examined for the purpose of identifying domains. Domains, the largest organizing unit, house similar categories. The analysis involved a back-and-forth process in which--
both the meaning of parts of the text and global meaning of the text are continually modified through an analysis of both. One arrives at a better understanding of the parts through analysis of the global meaning, and one arrives at a better analysis of the global meaning through analysis of parts. (Skovholt & Ronnestad, 1995, p. 148)
The third author, an experienced qualitative researcher, served as the auditor. As a validity check, the auditor and the first author separately coded portions of three transcripts from the study for concepts. These researchers compared and discussed concepts derived from the two analyses until they obtained agreement. The first author then completed the identification of concepts, resulting in 567 concepts. Each concept then was written...
|
