Psychotherapy Theory, Research, Practice, Training 2009, Vol. 46, No.

3, 376 389

2009 American Psychological Association 0033-3204/09/$12.00 DOI: 10.1037/a0017000

WHEN THE SHOE IS ON THE OTHER FOOT: A QUALITATIVE STUDY OF INTERN-LEVEL TRAINEES PERCEIVED LEARNING FROM CLIENTS
JESSICA V. STAHL, CLARA E. HILL, TOVA JACOBS, STACEY KLEINMAN, DANIEL ISENBERG, AND ALEXA STERN
University of Maryland, College Park

Twelve trainees (3 men and 9 women) who had recently completed predoctoral internships at a university counseling center were interviewed about what they had learned from clients. Data were analyzed using a consensual qualitative research method. Participants reported learning things about doing therapy, themselves, client dynamics, human nature, the therapy relationship, and the usefulness of supervision. In addition, participants highlighted the importance of consultation and self-reection to help them recognize what they learned. Implications for practice and research are discussed. Keywords: learning from clients, training, supervision, internship
As trainers of psychotherapists, it is incumbent on us to investigate how our students learn about psychotherapy so that we can modify our training programs to make them more effective. Most of
Jessica V. Stahl, Clara E. Hill, Tova Jacobs, Stacey Kleinman, Daniel Isenberg, and Alexa Stern, Department of Psychology, University of Maryland, College Park. Jessica V. Stahl is now at The Massachusetts School of Professional Psychology in Boston, MA. This article is based, in part, on the dissertation of Jessica V. Stahl under the direction of Clara E. Hill. A version of these results was presented at the 2007 annual meeting of the Society for Psychotherapy Research in Madison, WI. Correspondence concerning this article should be addressed to Jessica Stahl, The Massachusetts School of Professional Psychology, 221 Rivermoor St., Boston, MA 02132. E-mail: Jessica_Stahl@mspp.edu

the literature has focused on helping skills training (see Hill & Lent, 2006) and supervision (see Bernard & Goodyear, 2004) as the key factors in training psychotherapists. It is interesting that less attention has been paid to learning from clients, although there is some evidence that therapists attribute the majority of their learning about therapy to their clients. For example, Rnnestad and Skovholt (2003), in their qualitative study of therapist development, stated that:
Counselors/therapists at all levels of education and experience expressed in unison a voice that interacting with clients is a powerful source of learning and development. By disclosing their distress, their developmental histories, and ways of managing and coping with their problems of living, clients inform counselors/therapists of causes and solutions to human distress. The knowledge thus attained not only supplements and expands, but also brings depth and intensity to the theoretical knowledge obtained in formal schooling. (p. 33)

Likewise, Orlinsky, Botermans, and Rnnestad (2001) found that more than 4,000 therapists from a variety of specialties, nationalities, and theoretical orientations identied experience in therapy with clients as the most important inuence on their overall professional development. Other qualitative accounts of therapist development have also highlighted the importance of clients as primary teachers of therapists (e.g., Farber, 1983; Freeman & Hayes, 2002; Goldfried, 2001; Skovholt & McCarthy, 1988; Skovholt & Rnnestad, 1992). These studies are important because they conrm that therapists do learn from clients, but they provide no evidence about what therapists learn from clients. Most of the literature about what therapists learn involves personal accounts of learning from clients. For example, Crawford (1987) delineated lessons he learned from a beginning practicum. Bugental (1991), Strupp (1996), Mahrer (1996), Lazarus (1996), Kaslow (1996), and Ellis (1996) each described what they learned from a lifetime of conducting or researching psychotherapy. In addition, Freeman and Hayes (2002) described their changes on personal, professional, and spir-

376

Interns Perceived Learning From Clients itual levels as a result of experiences with one client; Kahn and Fromm (2001) synthesized 16 hypnotherapists accounts of how each was changed by a particular client. Although these accounts provide foundational information about what therapists learn from clients, their anecdotal nature makes it difcult to obtain a systematic perspective of what therapists learn from clients and what they do with what they learn. In the two published empirical studies on the impact of clients on their therapists, therapists reported that they were affected by their clients in several ways. For example, they described changes in personality and sense of self, such as becoming more psychologically minded and introspective (Farber, 1983) or experiencing more self-esteem and self-condence (Farber, 1983; Myers, 2002). In addition, therapists indicated that they gained a greater appreciation for therapy as a partnership between client and therapist and a greater understanding of the process of therapy (Myers, 2002). Farber (1983) also found some negative impacts, such as acting therapeutically toward others outside the ofce and experiencing less emotionality at home. Unfortunately, Farber (1983) and Myers (2002) did not provide much detail about measures used and coding methods, making it difcult to judge the validity of the ndings. Hence, the purpose of this study was to investigate what therapists learn from clients. Given the paucity of research on this topic, a qualitative approach was used to allow us to collect in-depth data about the experiences of therapists-in-training. More specically, we chose to use consensual qualitative research (CQR; Hill, Thompson, & Williams, 1997; Hill et al., 2005) because it has been used widely for psychotherapy research and because it allowed us to collect consistent data across participants as well as to pursue in-depth unique information from individual participants. Furthermore, given that what therapists learn from clients probably differs on the basis of their level of training, it seemed important to restrict the level of training in this study. Hence, we decided to focus this investigation on predoctoral interns. At this level, students are immersed in training full time for a year and so are deeply into the experience and thus likely to be good observers about what they have learned. In addition, students at this level have been involved for several years in supervision, and thus are likely to have paid attention to their experiences and to be good reporters of those experiences. Because a major principle in qualitative research is to choose participants who are experts and capable of reporting on their experiences (see Hill et al., 1997), predoctoral interns seemed like an ideal population. As is typical in qualitative research, we did not test a theoretical framework in this study because we wanted to allow the theory to emerge from the data. Relatedly, because there was little foundation from the literature to guide the development of hypotheses and because qualitative researchers eschew hypotheses as potentially biasing the investigators, we instead formulated research questions to guide our study. Our research questions were the following: What do interns perceive that they learn from their clients? How do interns come to realize what they learn? How do interns report applying what they learn from clients? Method Participants Interviewees. Participants were 12 (3 men, 9 women; 7 White, 1 biracial, 2 African American, 1 Asian American, 1 Middle Eastern, mean age 34.00 years, SD 7.11; 11 counseling, 1 clinical; 10 PhD, 2 PsyD) doctoral students who had recently completed APA-accredited predoctoral internships at university counseling centers. To provide contextual information about the sample, we asked participants to rate the degree to which they believed in and adhered to the techniques of four major theoretical orientations on 5-point scales (1 not at all, 5 completely). Using these rating scales, participants rated humanistic approaches an average of 3.75 (SD 0.87), psychodynamic approaches an average of 3.71 (SD 0.92), feminist/multicultural approaches an average of 3.58 (SD 0.90), and behavioral cognitive approaches an average of 2.75 (SD 0.97). Thus, this sample was primarily composed of White, female, in their 30s, predoctoral psychology interns who were relatively eclectic in theoretical orientation but leaned toward humanistic, psychodynamic, feminist approaches. Research team. The primary research team included ve White Jewish members (a 27-yearold female advanced doctoral student in counseling psychology, two female and one male postbaccalaureate individuals ages 23 or 24, and one

377

Stahl et al. 22-year-old upper level female psychology major student) who had all completed counselingrelated courses. The auditor was a 58-year-old White female PhD counseling psychologist who had extensive experience in psychotherapy research and developed CQR. Measures Demographics questionnaire. Participants were asked for age, sex, and doctoral program (clinical vs. counseling; PsyD vs. PhD), and endorsement (1 not at all, 5 completely) of humanistic, psychodynamic, behavioral cognitive, and feminist/multicultural theoretical orientations. Semistructured interview. The interview protocols were developed by the rst author on the basis of the existing literature on learning from clients, her clinical experience, two focus groups of graduate students and faculty members, and two practice interviews. Through these practice interviews, we discovered that participants gave more concrete information about lessons if they rst described lessons from one selected client and then talked about other lessons they had learned. Hence, the nal protocol was structured around asking rst about lessons from one particular client and then asking more generally about other lessons that had been learned from clients. The rst interview began by asking therapists to provide background information about the selected client (e.g., demographics, number of sessions, how long ago they saw the selected client, personal reactions). Participants were then asked an openended question about what they learned from the selected client. When participants could not think of anything more that they had learned from the selected client, they were asked about learning from the client about themselves, therapy, clients, and human nature. Participants were next asked to discuss how they came to realize what they learned, and how they had applied or anticipated that they would apply the lessons learned from this client. Participants were then asked about what they had learned from other clients, using the questions described above. A follow-up interview was included, as suggested in CQR (see Hill et al., 1997, 2005) to encourage in-depth data and allow participants to think more about the topic between interviews. The follow-up interview began with questions left over from the rst interview (where appropriate) and any clarication questions that came up for the interviewer on the basis of reviewing the tape of rst interview. Participants were then asked about their reactions to the rst interview, further thoughts about learning from clients, and what they had learned from participating in the interviews. Procedure Participant recruitment. Five doctoral interns in the rst authors counseling psychology program were sent an e-mail asking them to participate and to recommend fellow interns who might be willing to participate. Those individuals, in turn, were asked to recommend other interns they knew who might be willing to participate. All potential participants were sent an e-mail that briey described the purpose of the study, listed the risks and benets of participating, and included a copy of the interview protocol. Once a potential person agreed to participate, the interview was scheduled at a mutually convenient time, and participants were asked to read the protocol and think about their answers prior to the interview. Interviews were audiotaped and conducted by phone by the rst author. The rst interview lasted 1 to 1.5 hr, after which the interviewer listened to the tape to determine whether there were issues that needed clarication in the second interview. The follow-up interview was held in the next 1 to 7 days and lasted 15 to 30 min. After both interviews, the interviewer recorded her impressions of the interview and the interviewee. In total, 31 interns were invited to participate; 12 agreed to participate and completed both interviews (39% completion rate). CQR process. Interviews were transcribed and checked for accuracy. All identifying information was removed from transcripts, and cases were assigned code numbers. Before coding began, team members read the two CQR articles (Hill et al., 1997, 2005). Early meetings focused on creating an atmosphere in which all judges felt comfortable contributing to the consensus process. The rst task involved recording and discussing biases (i.e., personal issues that make it difcult to respond objectively to the data) and expectations (i.e., beliefs about what the data will look like). We report these here given that it is standard practice in qualitative

378

Interns Perceived Learning From Clients articles to describe the expectations and biases of judges so that readers have a context within which to judge the results. The judges expected that therapists would have learned about (a) the diversity of lifes problems and methods individuals employ to conquer adversity; (b) their own issues through working with a client with similar issues; (c) themselves as therapists most when they had trouble connecting with clients, when their approach to therapy was challenged, or as a result of exposure to different presenting problems, diagnoses, and symptoms; (d) technical issues (e.g., self-disclosure, countertransference issues) and separating their own issues from those of their clients; and (e) existential issues. Once they expressed their expectations, judges were asked to set these expectations aside (bracket them) and focus on the data during data coding and analysis. Judges were encouraged to discuss their expectations and biases as they arose and to challenge each other during the process of the study about these so that they could try to keep as objective a perspective as possible. In the rst step of analyzing the data, domains (i.e., topic areas) were derived from reading and coding the rst few interviews. These domains were modied considerably while going through the remaining transcripts. Once the list was stable, team members independently read each transcript and identied domains for each complete thought. The team then met to discuss and agree on the domain codings. After identifying the domains for each thought unit in a case, the team collaborated to construct core ideas (i.e., a concise yet comprehensive summary that stays as close to the participants words as possible) for each thought unit of raw data. In the meeting, a team member read a thought unit out loud (to understand the context of the case) and shared her or his version of the core ideas for that thought unit. The team then discussed and coconstructed the nal core ideas. The auditor read through the consensus version (which included domain titles, core ideas, and all the raw data for each domain) to determine whether the raw material was in the correct domain, that all the important material in each thought unit was included in the core ideas, and that the wording of the core ideas accurately characterized the data. The team then met to consider (accept or reject) all suggestions made by the auditor. Once discussion about the auditors comments was complete and the accepted changes were incorporated, the transcript was returned to the auditor so that she could see how the team responded to her comments and suggest further changes if appropriate. Because this was a dissertation and the auditor (the second author) was helping the rst author learn CQR, great care was taken to ensure that the team clearly understood the CQR process; hence, the auditing process continued for each case until the auditor was satised that the team captured the essence of the data. For the cross-analysis, the judges independently examined the data in each domain across cases to think about the underlying themes. Judges then brainstormed possible categories and arrived at consensus about the nal list. Once they had a nal list, they discussed the placement of core ideas into categories until consensus was achieved. When questions arose, they returned to the raw data to make sure that they were representing the data fairly. The auditor reviewed the cross-analysis to consider wording and representativeness of the categories and whether categories should be collapsed or further subdivided. The research team then considered the auditors comments as described above. For the lessons domains, we compared ndings for the selected client versus additional clients. No differences were found in categories of lessons discussed, indicating that these groups were not distinct (i.e., interns learned the same types of lessons from both groups). Hence, the data were combined across all clients. Once the cross-analysis was complete, categories were labeled to indicate the degree to which they represented the sample. Following Hill et al. (2005), the term general was used for categories that included 11 or 12 cases, typical 7 to 10 cases, variant for 3 to 6 cases. Categories emerging from one or two cases were placed into a miscellaneous category. Results We begin the results by comparing the demographics of our sample with interns in general to help readers place our sample in context. In terms of degree type, our sample differed from the larger intern cohort, as 92% of our sample were earning degrees in counseling psychology (vs. 16% nationally), and 83% of our sample were pursuing PhDs (vs. 64% nationally; Association of Psychology and Postdoctoral Internship Cen-

379

Stahl et al. ters, 2004). For age, gender, and race, we compared our sample with the 20052006 cohort (Association of Psychology and Postdoctoral Internship Centers, 2005) because this information was not collected on the 2004 2005 cohort. In terms of gender, our sample was fairly similar to the national sample (75% of our group were female vs. 78% female nationally). Otherwise, our participants were a bit older (M 34.00 years, SD 7.11 vs. M 30.83 years, SD 6.56), and more diverse (58% of our sample was White vs. 78% nationally) than the national sample. Thus, our sample was ethnically diverse, and composed primarily of women in their mid-30s who were seeking counseling psychology PhD degrees. Three domains emerged from the data: lessons learned from clients, how intern realized lessons from selected client, and how intern applied lessons from selected client. The categories, subcategories, and frequencies of the each of these domains can be found in Tables 1 and 2. In the tables, we list all the ndings to provide the range of topics that participants talked about, but in the text we present only the ndings that were at least typical. In addition, a prototypical case is presented to illustrate the ndings. Lessons Learned From Clients There were ve types of typical lessons learned from clients: doing therapy, self, clients, human nature, and therapy relationship (see also Table 1). Lessons about doing therapy. Participants mostly discussed learning about how to do therapy. First, they described learning about therapy skills plus, which refers to learning the nuts and bolts of how to conduct therapy. Under this rubric, they discussed learning about the need to use and be aware of oneself in sessions (e.g., to pay attention to your feelings and reactions to a client and kind of basic countertransference stuff, that they should use their own feelings and reactions to the client as information about the client, and that it was important to communicate their own reactions or feelings to client appropriately). Second, they described issues related to theoretical orientation, such as the helpfulness of supplementing their theoretical approach with specic techniques (e.g., [This client] reafrmed my trust in paradoxical work. The I know youre probably not going to come back next week because of what happened today, so . . . . But she would come back just to show me, which is a good thing.). Third, they talked about session or case management. They talked here about such things as learning how to work with specic diagnoses as well as the limits and benets of diagnosis; setting and maintaining boundaries (e.g., how to do it, that they struggle with it); the process of making referrals, dealing with outside therapists and agencies, and hospitalizing clients; and the need to make, keep track of, and check on therapy goals. Participants also talked about the challenges and limitations of therapy. For example, one participant said,
I learned maybe that change is slow and gradual and thats okay. I think she did a lot of work in that 6- or 8-month period even though to her or even to people outside of her they might not have noticed a change at all . . . it reminds me that people dont change that quickly and therapy is such a long process of difcult work.

Participants also typically mentioned having learned about the complexity and challenge of doing therapy. For example, one participant noted,
I knew this before . . . theres no recipe for [therapy]. I think this is where the art of therapy comes in. Its really a creative process for the therapist and for the client, but its not creative in the ways we think it is . . . its emotionally creative . . . that makes it both scary but exciting and fun at the same time.

Similarly, another said, Im just always struck by how complex [therapy] is . . . its always a negotiation, so were trained to kind of be with the client where they are and at the same time I had my own notions of what would be helpful. Lessons about self. All participants discussed lessons that they had learned about themselves. These lessons typically involved personal limitations (e.g., I learned about my own comfort and discomfort with boundary setting and really challenged my, sort of need to be liked by a client; [Ive learned that] the idea that as a therapist Im different than my clients is just not true. I think were all people and we all have our problems . . . and we all deal with them in our own ways). On the ip side, participants also typically reported having learned that they were good enough therapists or discussed having felt an increased sense of competence as a therapist (e.g., I learned I could do good enough work with my clients. It doesnt have to be outstanding work or brilliant work, it just has to be good enough work).

380

Interns Perceived Learning From Clients

TABLE 1. Categories, Subcategories, and Frequencies for Lessons Learned From Clients Domain, category, and subcategory 1. Lessons about doing therapy A. Therapy skills plus 1) Process skills a. Need to use and be aware of self in sessions 1. Need to be self-monitoring and self-aware when working with clients 2. Need to use own feelings and reactions to client as information about client 3. Need to communicate genuine or vulnerable attitude to clients b. Importance of being responsive to clients c. Need to give clients responsibility for change d. Importance of trusting the process e. Need to communicate feelings to clients f. Need to sit with own and clients feelings in a session 2) Issues related to theoretical orientation a. Usefulness of [technique] in lieu of (or in addition to) theoretical orientation b. Need to be exible with theoretical orientation, approach, or conceptualization c. Preference for theoretical orientation 3) Session and case management a. Working with and impact of diagnoses b. Skill and strategies for setting and maintaining boundaries c. There is a specic process to make a referral and deal with outside therapists and agencies; hospitalizing clients; learned about system d. Need to keep track of, make therapy goals, and check in with clients on how things are going 4) Specic skills and interventions (e.g., silence, self-disclosure, challenge) B. Challenges and limitations of doing therapy 1) There are limits to what therapy can accomplish; what are realistic expectations and goals for client change (e.g., change is slow) 2) Therapy is hard, unpredictable, complex, and challenging 3) It is hard to tell who (therapist or client) is responsible for therapy outcome 2. Lessons about self A. Participant has personal limitations (idiosyncratic to participant) B. Participant is good enough therapist; felt increased sense of own competence as therapist C. Participant learned that s/he has biases (what they are) and how they affect the work D. Participant has high expectations for self (wants to be perfect), but realized it is okay to not be perfect E. Participant learned something about own personal characteristics, issues, or countertransference triggers (idiosyncratic to participant); gained self-awareness F. Participant prefers to work with certain types of clients (actual preference is idiosyncratic to participant) G. Participant learned about own personal rewards for therapy work 3. Lessons about clients A. Change is hard for client and is very individual to client (e.g., clients motivation to change plays a big role) B. Our knowledge and understanding of clients are limited C. Clients interpersonal style in therapy is a replication of their outside relationships; there are parallels between therapy relationship and clients outside relationships D. Conceptualization of client evolves over time and should be exible E. Clients have reasons for behaving the way they do 4. Human nature A. Our environment, family, culture, and early experiences shape our lives and problems B. Life is unfair and can change in an instant C. People have negative qualities (e.g., selsh, evil, critical, complex, rigid) D. People are resilient and can change when they want to E. People are ambivalent about change, want a quick x, and do not want to take responsibility for their lives F. Therapy plays an important role in peoples lives and society 5. Lessons about therapy relationship A. The therapy relationship is important and curative B. The therapy relationship is complex 6. Supervision, consultation, collaboration, and training are important and useful Frequency G G G T V V V V V V V V G T V V T V V V V V T T T V G T T T V V V V G T V V V V G T V V V V V T V V V

Note. N 12. G general (this category occurred for 11 or 12 participants); T typical (this category occurred for 7 to 10 participants); V variant (this category occurred for 3 to 6 participants).

381

Stahl et al.
TABLE 2. Categories, Subcategories, and Frequencies for How Lessons Were Realized and Applied Category and subcategory 1. How realized lessons A. Talking to others 1) Supervision 2) Consulting with others B. Self-reection C. Readings or class D. Directly from the client him-/herself; from observing the client; from doing therapy with subsequent clients E. From the interview 2. How applied lessons A. Applied to clinical work 1) Used in subsequent work with clients 2) Shared with others (colleagues, supervisees) B. Applied to self Frequency G G T T V V V G G V V mom and dad, just very difcult for her to say anything negative about them . . . . It just really showed me how strong that bond is and how children do that because they have to survive and how deeply positive that can be but also how damaging that can be.

Note. N 12. G general (this category occurred for 11 or 12 participants); T typical (this category occurred for 7 to 10 participants); V variant (this category occurred for 3 to 6 participants).

Lessons about the therapy relationship. Participants typically reported having learned about the therapy relationship. They discussed having learned that the therapy relationship is important or curative (e.g., One thing that I got from her [was] . . . the importance really of having a condential place to get support . . . but then also having somebody to whom one holds oneself accountable), and that the therapy relationship is complex (e.g., I had to learn that would not make me an effective therapist if I just worked at keeping them liking me . . . you can be challenging and confronting . . . but it doesnt take away any sense of attachment or any sense of connection that we have). How Intern Realized Lessons From Selected Client

Participants also typically indicated that they learned about their own biases (i.e., what they are) and how these biases affect their work as therapists. For example, an African American participant noted,
I learned a lot about . . . introducing race and culture into my work . . . . She was White and I didnt do a great job of that, and since then it caused me to think more about how I think about White people and their cultural development . . . . I really tend to focus on culture with people of color and not White people.

Lessons about clients. Participants generally discussed lessons about clients. The only typical subcategory here was learning about clients motivation to change (i.e., participants said they learned that change is hard for some clients, and that change is very individual to clients level of motivation, readiness to change, or specic areas of resistance). According to one participant, Clients have taught me [that] we are all ambivalent about change. And every client is going to manifest that differently . . . [there is a] variety of ways that I get the hand. Lessons about human nature. All participants discussed lessons about human nature. Typically, they reported having learned that background shapes ones life and problems. For example, one participant said,
Her childhood was fairly chaotic and dysfunctional . . . yet she was still incredibly protective and loving toward both

Table 2 lists the ways in which interns realized what they learned from clients. First, all participants described learning from direct discussions with others, such as supervisors (e.g., I felt very comfortable talking with [my supervisor] about my reactions to clients and things like that . . . it was very helpful to be able to talk to him just about my surprise of her willingness to work with me . . . . I think supervision was probably the main thing) or peers (e.g., [My conversations about] my general experiences and who I am as a therapist and how I may have changed and how my perspectives have changed and what I learned . . . I feel like I had those with my peers, my cohort of interns). Self-reection was also typically discussed by interns as a source of learning. For example, one intern noted, I think some of it I learned, probably most of it, just from the amount of time I invested in thinking about this client outside our therapy . . . so I think a lot of self-questioning, a lot of self-reection during the process of meeting with her. Application of Learning Participants described two primary ways in which they had or would in the future apply what they learned from their clients (see Table 2). First, participants generally indicated that they

382

Interns Perceived Learning From Clients had used or will use what they learned from their client in subsequent work with clients. For example, one intern (whose client had a very signicant history of trauma) said that since seeing this client, she has done a better job getting histories from clients; another (whose client elicited strong feelings of anger in her) discussed being able to supervise herself more effectively regarding strong reactions to clients. Second, participants discussed how they apply what they learned from their client to themselves. They described having gained an appreciation for their own lives or used the lessons to foster personal growth. Prototypical Case: I Wanted to KILL Him! To bring the ndings to life, we present one case to illustrate how one intern discussed the lessons she learned from the client she selected, how she realized the lessons from that client, and how she has or would in the future apply the lessons. This particular case was chosen because the research team agreed that it was a particularly good example of the range of lessons participants discussed. To protect condentiality, the identities of the therapist and client have been disguised. The therapist has read and approved this write-up of her case. W was a 34-year-old White female therapist who discussed a 25-year-old African American male law student client who presented with depression and procrastination problems. They worked together for 24 sessions on getting the client to take responsibility for his work and life given that the client was a chronic underachiever. W remarked that this client was brilliant but that his emotional wounds were great (he had grown up in a military family and his father was described as a rage-aholic) and that his emotional resources did not match his intellectual capacity. When asked about her reactions to the client, W said, I wanted to kill him. She described the process of their therapy as very difcult, as the client could not own his anger, and would distance himself from W by irting or trying to engage her in intellectual conversations about therapy (e.g., Do all people struggle like this? Why havent I made any progress yet? Why do we need to talk about feelings?). Things shifted after Session 15 when W asked the client what it was like to not make progress; subsequently, he was able to access some of his anger and be more vulnerable in sessions. W discussed three lessons about doing therapy. First, she learned that she needed to have appropriate expectations for the therapy process and outcome related to how wounded a client is; clients who are more wounded may not be able to go as far as less wounded clients. In addition, she learned about the importance of getting clients to take responsibility for their feelings. The shift happened when she simply asked the client what it was like to not be making progress (vs. trying to reassure the client or explain why things had not changed); this taught her that it is not her job to heal her clients, but to help them heal themselves. Third, she learned how to use her own reactions to conceptualize what is going on for a client and how to direct the work. She conceptualized the anger she felt during sessions as stimulated by the client as his way of connecting to her, such that she was experiencing the anger that he could not own or experience. She needed to experience those feelings to help him connect, have compassion for those feelings, and use the feelings in sessions. From this client, W learned some lessons about herself, primarily that she is not incompetent. After discussing this client in supervision, W recognized that her feelings of incompetence came only after sessions with this particular client. From recognizing this, she was able to feel good about herself and recognize that her evaluation of her competence as a therapist was separate from what the client did. In addition, because this client really wanted her to be directive and give him homework, W also learned that she is not very good at being directive. Although she met him there as best she could in ways that were consistent with who she was as a therapist, their work highlighted for her that she does not like to do directive therapy. Regarding client dynamics, W learned how to conceptualize irting as a defensea way of avoiding connecting with her emotionally rather than as a sign of sexual attraction. By conceptualizing it in this way, she was able to discuss the irting with the client in an ethical manner. She also learned that not all clients of color want to talk about race; even though the clients racial identity was part of her conceptualization (e.g., he was African American but identied as White, which she saw as a way of not accepting himself), he was not ready to talk about it. A third major lesson for W about client dynamics was that her clients angry presentation and the feelings of

383

Stahl et al. anger and incompetence that he elicited in her were a manifestation of his narcissistic wounds. Fourth, this client highlighted for W how limiting it is when clients cut you off from pieces of themselves; this client would not talk about how his family dynamics contributed to what was going on with him. However, W felt that if they did not address his family dynamics, she could not help him the way she wanted to. So, she had to learn to be okay with what she could do with the things he would discuss. Finally, in terms of human nature, W learned that an important part of being human is taking responsibility for oneself, and that doing so is difcult. Although this is something she was aware of before, this client challenged her on that more than had any previous client. W noted that much of what she learned about this client came through self-reection. She further explained that she learned what to do with it (e.g., that it was not her job to take on the full responsibility for the sessions) from supervision. W indicated that her supervisor was pivotal in helping her gain understanding about her strong reactions to the client because she could only get so far when reecting on them by herself. Having gone through the experience with this client, W felt more able to supervise herself when she had strong reactions to clients. Furthermore, she was able to help a fellow intern deal with a similar reaction. W also noted that since working with this client, she found it easier to use her emotional reactions to conceptualize and focus the work. Working with this client also made it easier for her to have more realistic expectations for clients, the kind of work they can do, and what it feels like to put responsibility back on the client. Finally, W learned to accept that although all clients are wounded, their presenting issues and the amount of progress they can make will differ. Discussion These 12 predoctoral psychology interns from university counseling centers reported that they had learned a lot of profound lessons from their experiences with clients, and these lessons helped them become better therapists. Interns reported learning about doing therapy, themselves, clients, human nature, the therapy relationship, and supervision, all of which have been reported at least briey in previous anecdotal literature. Freeman and Hayes (2002) reported that they personally had changed on personal, professional, and spiritual levels as a result of their experiences with specic clients. Similarly, Kahn and Fromm (2001) stated that therapists can be changed by clients in terms of learning new approaches (specic changes on the therapists practice), individuation (intellectual or emotional change in the therapist, such as a perspective shift), and transformation (therapists understanding and worldview are shifted). In this section, we rst discuss the lessons learned and then put the ndings into a developmental framework. Lessons Lessons about doing therapy. As one would expect of individuals who had just completed an intensive therapy training experience such as internship, all participants discussed many lessons they had learned about doing therapy. The range of lessons that participants discussed across this category demonstrates the wide variety of things that interns learn about doing therapy. In addition to learning about therapy at both the micro level (i.e., specic interventions like silence and selfdisclosures) and the big picture level (e.g., learning to trust in the process, the limits of what therapy can accomplish), interns learned about various components of doing therapy (e.g., process skills, session and case management, having and using a theoretical orientation, the challenges and limitations of doing therapy). Many of the lessons about doing therapy discussed by participants are consistent with anecdotal accounts of learning from clients as well as empirical studies on the impact of clients on their therapists. For example, several more experienced authors have discussed learning the importance of using a variety of therapeutic techniques (Kahn & Fromm, 2001) and being exible with ones theoretical approach when working with specic clients (Lazarus, 1996; Norcross, 1996). Lessons discussed by participants have also been reported by similarly experienced or less experienced therapists (e.g., Crawford, 1987, who was writing about what he had learned on completing his rst practicum; Myers, 2002, whose participants experience ranged from 5 to 30 years), such as learning to use ones own feelings or reactions as information about the client (Crawford, 1987) and about the need to give clients responsibility for change (Crawford, 1987;

384

Interns Perceived Learning From Clients Myers, 2002). Perhaps as a result of the sheer number of lessons participants discussed in this category, many of the specic lessons in this category had not previously been reported. Lessons that at least three participants reported that had not been not discussed in the previous literature included: need to communicate genuine or vulnerable attitude to clients, importance of trusting in the process, working with and impact of diagnoses, the process of making referrals and dealing with outside agencies, using silence and challenges, the limits of what therapy can accomplish, the difculty and complexity of doing therapy, and the difculty of knowing who is responsible for therapeutic change. In addition, the lessons participants discussed in the therapy skills plus category mirror a distinction made by Hill and Lent (2006) that in addition to learning specic helping skills (e.g., reections of feeling, restatements, interpretations), therapists-in-training learn additional skills that facilitate their therapeutic endeavors (e.g., theoretical framework, self-awareness, facilitative attitude, responsiveness to clients, case conceptualization skills, case management skills, professionalism, and ethics). The ndings in the present study conrm that Hill and Lents distinction is a useful one for conceptualizing trainees lessons about doing therapy, as interns discussed learning about both helping skills and the plus additional skills. Furthermore, the nding that interns spoke relatively little about learning how to do therapy at the micro level (i.e., learning helping skills, although a few of them did discuss the use of silence, self-disclosures, and challenges) but reported learning a whole lot at higher levels (i.e., everything else in the therapy skills plus lessons subcategory) is consistent with Hill and Lents conclusion that a focus on the additional skills continues throughout training. The present ndings contrast nicely with Hill, Sullivan, Knox, and Schlosser (2007), who found that novice therapists asked to keep a journal of their experiences in a prepracticum course reported a lot of learning about helping skills. These interns were clearly at a more expert status than were the novice therapists. These ndings are consistent with research on the development of therapy expertise. Specically, the therapy expertise literature suggests that as one gains experience with conducting therapy, basic therapy skills become automatized such that they fade into the background and become second nature, allowing the therapist to attend to more complex cognitive tasks such as client conceptualization (e.g., Cummings, Slemon, & Hallberg, 1993; Sakai & Nasserbakht, 1997). Furthermore, it is notable that when participants discussed learning about specic interventions, these lessons concerned using such skills at a complex level (e.g., sitting in silence for 6 or 7 min, disclosing ones sexual orientation to a client in a clinically appropriate time and manner, challenging a client about his marijuana use). Together, these ndings suggest that interns had automatized basic therapy skills (e.g., reections of feeling, restatements, interpretations), opening up space in their working memory to attend to and learn more cognitively complex skills and tasks such as trusting in the process, becoming exible with ones theoretical orientation, and session and case management. This supports the idea that therapists reach some degree of expertise on completion of their training, although research on expertise in the cognitive science eld posits that approximately 10 years of experience are required to be considered an expert (Sakai & Nasserbakht, 1997). Lessons about self. The nding that participants learned about themselves from clients conrms previous literature. In a study of more than 4,000 therapists of diverse backgrounds, Orlinsky, Rnnestad, and Willutzki (2004) found that therapy interactions inuenced therapists lives and personalities. In addition, Freeman and Hayes (2002) and Kahn and Fromm (2001) indicated that therapists changed personally as a result of their work with specic clients; Farber (1983), Myers (2002), and Kaslow (1996) reported that therapists became more psychologically minded, self-aware, and self-assured. Two of the specic lessons about themselves described by at least three participants in this study have not previously been reported. The rst was that the participants had high expectations for themselves as therapists (i.e., wanted to be perfect) but that it was okay to not be perfect. This nding is consistent with the perfectionism associated with participants advanced student status (see Rnnestad & Skovholt, 2003). The second unique lessonabout the types of clients the participants each liked to work withmakes sense given that these participants had just completed internship, where they worked with many different clients, perhaps gaining exposure a wide range of types of clients for the rst time.

385

Stahl et al. Lessons about clients. Trainees learned about clients and client dynamics, which is necessary for therapists to be responsive to clients individual needs (Hill & Lent, 2006). The lesson that change is hard for clients and is very individual to clients (the only lesson in this category discussed by more than half the participants) is similar to ndings reported by Myers (2002) that therapists no longer gave up on clients after their experiences with inuential clients. Similarly, Freeman and Hayes (2002) indicated that therapists learned about the change process from courageous clients. This lesson is also reective of participants transition into the novice professional phase of therapist development, in which they gain an appreciation for both the complexity of therapy and the importance of client feedback in assessing how the work is going (see Rnnestad & Skovholt, 2003). Several lessons about clients and client dynamics have not previously been reported in the literature. These include that our knowledge and understanding of clients is limited, clients interpersonal style is a replication of their outside relationships, conceptualization of clients evolves over time and should be exible, and clients have reasons for behaving the way they do. Lessons about human nature. Participants lessons reected both positive aspects of human nature (e.g., resilience) and about human weaknesses and aws (e.g., people can be selsh or critical; people are ambivalent about change). These ndings are consistent with Yaloms (2002) description of clients and therapists as fellow travelers who learn about the human existence from each other. Similarly, Bugental (1991), Kahn and Fromm (2001), and Freeman and Hayes (2002) have written about the ways in which therapists gained an understanding of the role of therapy in peoples lives and society as a result of their work with clients. Lessons about the therapy relationship. The lesson that the therapy relationship is important and curative is consistent with previous literature on client impact and learning from clients (e.g., Crawford, 1987). The lesson that the therapy relationship is complex, however, has not explicitly been found in previous literature, but this lesson is not surprising given that one might argue that all relationships are complex (particularly when one is trained to examine them at the level to which one does as a therapist). Furthermore, research on the various components (e.g., transference/countertransference conguration, working alliance, real relationship) of the therapy relationship and the many factors that go into each one (e.g., tasks, bond, and goals of the working alliance) certainly reect the complexity of the therapy relationship (e.g., Gelso & Hayes, 1998). Putting the Lessons in a Developmental Framework Our intent with this study was not to test a particular theoretical framework, but to allow theory to emerge from the data. Yet, lessons learned are consistent with what one would expect based on Rnnestad and Skovholts (2003) model of therapist development. Specically, interns are transitioning from the advanced student phase (students beyond the beginning year who are working as counselors and therapists-intraining and receiving regular and formalized supervision) to the novice professional phase (individuals in their rst few years after graduation). According to the model, advanced students often feel pressure to do things perfectly, appreciate the impact of their professional training, seek conrmation and feedback from seniors and peers, and critically evaluate and assess models in an effort to differentiate, accept, and reject model components. Participants lesson about their theoretical orientation (e.g., the importance of incorporating various techniques into their approach) and their realization that they have high expectations for themselves suggests that interns still exhibit features of this stage. In addition, some of the interns lessons (e.g., need to use and be aware of self in sessions, learning that they are a good enough therapist, learning about their own biases and how they affect the work) are consistent with the external to internal shift that Rnnestad and Skovholt (2003) posited. They suggested that as students progress through the advanced student phase, they move from an external focus (e.g., looking at supervisors for how to be a professional) to an internal focus (e.g., looking at more complex issues of personal development, parallel process, transference/countertransference, etc.). Interns discussed several lessons that reect their transition into the novice professional phase. According to Rnnestad and Skovholt (2003), novice professionals experience a sense of being on their own, and they often test the validity of

386

Interns Perceived Learning From Clients what was learned in school only to discover the gaps in their training, leading to a sense of disillusionment. At times, they feel lost without the support of graduate school and supervision, and look to mentors for guidance and support. At this stage, client feedback becomes an increasingly powerful means of assessing what works, and the therapist becomes increasingly aware of the ways in which his or her own personality is present in the work. Several of the participants lessons (e.g., the need to use their own feelings and reactions to the client as information about the client, learning about their own personal limitations as a therapist, and learning about their own characteristics) reected an increased reliance on client feedback and awareness of ones own role in the therapy work. Finally, Rnnestad and Skovholt (2003) noted that the novice professional often experiences an increasing sense of the complexity of therapy work and recognizes the importance of the therapy relationship for client progress, while simultaneously expressing a renewed interest in learning techniques specic to the work being done. These changes were reected in the present study in that participants discussed learning about the challenges and limitations of doing therapy (e.g., there are limits to what therapy can accomplish, that therapy is complex and challenging), learning to trust in the process of therapy, learning that the therapy relationship is important and curative, and learning that specic techniques are useful in addition to (or in lieu of) ones theoretical approach. Thus, these ndings provide some conrmation for Rnnestad and Skovholts model. How Interns Realized Lessons Participants realized that they learned what they did from their clients through a variety of sources (e.g., supervision, consultation with others, readings and class, self-reection, directly from the client, from the interview). Of these sources of realization, supervision and consulting with others were most commonly endorsed by participants, highlighting the importance of discussing with and getting feedback from others to solidify things learned from clients. Similarly, Bernard and Goodyear (2004) highlighted the importance and value of supervision. Self-reection was also endorsed by most participants as responsible for their realization of the things they learned from clients. Given that much Limitations As is common to all qualitative research (Polkinghorne, 2005), the self-report, retrospective nature of the data may have affected the ndings. Participants may have been motivated to respond to questions in a socially desirable manner. In addition, interviews took place 2 to 3 months after the completion of a very intense year of internship. Thus, participants were reporting what they perceived that they learned from their clients. Without having assessed participants knowledge before and after seeing the selected client or gaining corroboration from their supervisors, we cannot know how accurately participants reported their lessons. of therapy training requires that one engage in self-reection (especially in regards to ones work; Hayes, 2002), this nding is not surprising. The primary importance placed by participants on both supervision or consultation and selfreection is also consistent with experiential learning theory, which posits that immediate or concrete experiences must be followed by observations and reections to be assimilated into long-term memory and subsequently used (Abbey, Hunt & Weiser, 1985; Kolb, Boyatzis, & Mainemelis, 2001). Application of Learning The nding that participants generally indicated that they have already applied or will apply what they learned to their current or subsequent clinical work is consistent with the literature about changes clients engender in their therapists practice of psychotherapy (Kahn & Fromm, 2001; Myers, 2002). Participants also occasionally shared what they learned from specic clients with colleagues or supervisees, suggesting that interns gain clinical knowledge not only from their own clinical experiences, but from each others. A few participants also indicated that they applied what they learned from their client to themselves, which is consistent with the literature (Farber, 1983; Kahn & Fromm, 2001). For example, a few participants indicated that as a result of working with clients, they appreciated their own life or fortune more or that they used what they learned from their client to foster personal growth.

387

Stahl et al. Furthermore, although providing participants with a copy of the research protocol ahead of time gave them the opportunity to reect on the topic and potentially provide richer data, the questions and prompts listed in the protocol may have biased participants thinking about learning from clients. Similarly, although steps were taken to minimize raters bias (e.g., writing about and discussing expectations and biases at the beginning of the coding process and checking in throughout), judges expectations and biases may have inuenced ndings (e.g., led us to nd what we expected to nd). Finally, the small sample size limits the generalizability of the ndings. The results of this study may apply only to recent interns at university counseling centers who are willing to talk about their experiences. Interns who completed internships in different settings may have learned different things or may have learned more in certain categories (e.g., someone who interned at a psychiatric hospital may have discussed learning more about psychopathology). Similarly, the results may not be generalizable to therapists at other levels of development (e.g., beginning therapists or expert therapists). Implications for Practice and Training Despite the fact that therapists often say that experience with clients is the most important inuence on their professional development (e.g., Orlinsky et al., 2001), learning from clients has rarely been discussed in therapy training or research. This lack was reected in the surprise that some participants expressed at never having before been asked about what they learn from clients (e.g., I feel like this shouldnt be the rst time Im being asked what Ive learned from my clients). The results of this study thus suggest that supervisors should explicitly ask their supervisees what they learn from clients. The potential utility of incorporating this question into supervision is supported by research on the role of intention to learn on memory acquisition. Although intention does not directly inuence memory acquisition, it inuences the strategy individuals choose to use when encoding new information, which in turn inuences the quality of memory acquisition. Specically, when one is told that one should be learning something specic from a given stimulus, information is processed at a more elaborative level, thus facilitating the quality of memory acquisition (Reisberg, 1997). Furthermore, because reection on learning from clients helps therapists explicitly consider what they have gained personally and professionally from their therapeutic work, such reection may prevent burnout. One might hypothesize that reection on learning from clients could prevent (or ameliorate) the emotional exhaustion component of burnout, which involves feeling drained and emotionally overextended by ones contacts with other people (Maslach & Jackson, 1984). Finally, practitioners might consider telling clients for whom it would be clinically appropriate what they learned from them. This implication stems directly from a comment made by one participant: I think its sad that clients dont know how much we learn from them . . . . I think some clients need to think that we know everything. But to the clients who dont have that need, I think it would mean a lot to them . . . . I wonder if theres some way to talk about it that is helpful . . . . I think a lot of clients would like to know they had an impact on us. Implications for Research Future researchers might investigate what therapists of different experience levels (novices vs. those with more than 10 years of experience), from different settings (hospitals vs. schools), with different client populations (severe psychopathology, trauma, disabilities, etc.) or theoretical orientations learn from clients. Are there some things that therapists within a particular experience level learn regardless of setting (e.g., all interns learn about trusting in the process)? Such a study could validate Rnnestad and Skovholts (2003) phase model of therapist development and the hypothesis posed by expertise theory that novices would focus on individual helping skills, whereas experts focus more on higher order constructs such as case conceptualization (e.g., Cummings, Slemon, & Hallberg, 1993; Sakai & Nasserbakht, 1997). Finally, it could be helpful to develop a measure of learning from clients. The measure could be a paper-and-pencil means of identifying the most important lessons therapists take away from therapy sessions or relationships. Furthermore, assessing other dimensions of the lesson (e.g., newness, likelihood of applying the lesson to other clients, relationship of the lesson to various

388

Interns Perceived Learning From Clients categories of learning, impact of lessons on self) would allow researchers to investigate both qualitative and quantitative dimensions of learning from clients. References
ABBEY, D. S., HUNT, D. E., & WEISER, J. C. (1985). Variations on a theme by Kolb: A new perspective for understanding counseling and supervision. The Counseling Psychologist, 13, 477501. Association of Psychology and Postdoctoral Internship Centers. (2004, March 15). 2004 APPIC match: Survey of internship applicants. Retrieved December 18, 2008, from http://appic.org/match/5_2_2_4_6_match_ about_statistics_surveys_2004.html Association of Psychology and Postdoctoral Internship Centers. (2005, May 8). 2005 APPIC match: Survey of internship applicants. Retrieved December 18, 2008, from http://appic.org/match/5_2_2_4_7_match_ about_statistics_surveys_2005.htm BERNARD, J. M., & GOODYEAR, R. K. (2004). Fundamentals of clinical supervision (3rd ed.). Needham Heights, MA: Allyn & Bacon. BUGENTAL, J. F. T. (1991). Lessons clients teach therapists. Journal of Humanistic Psychology, 31(3), 28 32. CRAWFORD, R. (1987). A students experience of practicum: Learning from clients. Texas Association for Counseling and Development Journal, 15(1), 7176. CUMMINGS, A. L., SLEMON, A. G., & HALLBERG, E. T. (1993). Session evaluation and recall of important events as a function of counselor experience. Journal of Counseling Psychology, 42, 156 165. ELLIS, A. (1996). How I learned to help clients feel better and get better. Psychotherapy, 33, 149 151. FARBER, B. A. (1983). The effects of psychotherapeutic practice upon psychotherapists. Psychotherapy: Theory, Research and Practice, 20, 174 182. FREEMAN, M. S., & HAYES, G. (2002). Clients changing counselors: An inspirational journey. Counseling and Values, 47, 1321. GELSO, C. G., & HAYES, J. A. (1998). The psychotherapy relationship: Theory, research, and practice. New York: Wiley. GOLDFRIED, M. R. (2001). Conclusion: A perspective on how therapists change. In M. R. Goldfried (Ed.), How therapists change: Personal and professional reections (pp. 315330). Washington, DC: American Psychological Association. HAYES, J. A. (2002). Playing with re: Countertransference and clinical epistemology. Journal of Contemporary Psychology, 32, 9399. HILL, C. E., KNOX, S., THOMPSON, B. J., WILLIAMS, E. N., HESS, S. A., & LADAny, N. (2005). Consensual qualitative research: An update. Journal of Counseling Psychology, 52, 196 205. HILL, C. E., & LENT, R. W. (2006). Training for novice therapists: Helping skills plus. Psychotherapy, 41, 1116. HILL, C., SULLIVAN, C., KNOX, S., & SCHLOSSER, L. (2007). Becoming a psychotherapist: The experiences of novice trainees in pre-practicum training. Psychotherapy: Theory, Research, Practice, Training, 44, 434 439. HILL, C. E., THOMPSON, B. J., & WILLIAMS, E. N. (1997). A guide to conducting consensual qualitative research. The Counseling Psychologist, 25, 517572. KAHN, S., & FROMM, E. (2001). Introduction. In S. Kahn & E. Fromm (Eds.), Changes in the therapist (pp. xiii xxvi), Mahwah, NJ: Erlbaum. KASLOW, F. W. (1996). Memorable lessons from a multifaceted career. Psychotherapy, 33, 153170. KOLB, D. A., BOYATZIS, R. E., & MAINEMELIS, C. (2001). Experiential learning theory: Previous research and new directions. In R. J. Sternberg & L. Zhang (Eds.), Perspectives on thinking, learning, and cognitive styles (pp. 227247). Mahwah, NJ: Erlbaum. LAZARUS, A. A. (1996). Some reections after 40 years of trying to be an effective psychotherapist. Psychotherapy, 33, 142145. MAHRER, A. R. (1996). Lessons from 40 years of learning how to do psychotherapy. Psychotherapy, 33, 139 141. MASLACH, C., & JACKSON, S. E. (1984). Burnout in organizational settings. Applied Social Psychology Annual, 5, 133153. MYERS, S. (2002). The other outcome: How therapists integrate their experiences with clients. In P. Shohov (Ed.), Advances in psychology research (Vol. 17, pp. 113130). Hauppauge, NY: NOVA Science Publishers. NORCROSS, J. C. (1996). The lifetime lessons of six psychologists: An introduction. Psychotherapy, 33, 129 130. ORLINSKY, D. E., BOTERMANS, J.-F., & RNNESTAD, M. H. (2001). Towards an empirically grounded model of psychotherapy training: Five thousand therapists rate inuences on their development. Australian Psychologist, 36, 139 148. ORLINSKY, D. E., RNNESTAD, M. H., & WILLUTZKI, U. (2004). Fifty years of psychotherapy process-outcome research: Continuity and change. In M. J. Lambert (Ed.), Bergin and Garelds handbook of psychotherapy and behavior change (pp. 307390). New York: Wiley. POLKINGHORNE, D. E. (2005). Language and meaning: Data collection in qualitative research. Journal of Counseling Psychology, 52, 137145. REISBERG, D. (1997). Cognition: Exploring the science of the mind. New York: Norton. RNNESTAD, M. H., & SKOVHOLT, T. M. (2003). The journey of the counselor and therapist: Research ndings and perspectives on professional development. Journal of Career Development, 30, 5 44. SAKAI, P. S., & NASSERBAKHT, A. (1997). Counselor development and cognitive science models of expertise: Possible convergences and divergences. Educational Psychology Review, 9, 353359. SKOVHOLT, T. M., & MCCARTHY, P. M. (1988). Critical incidents in counselor development. Journal of Counseling and Career Development, 67, 69 72. SKOVHOLT, T. M., & RNNESTAD, M. H. (1992). Themes in therapist and counselor development. Journal of Counseling and Development, 70, 505515. STRUPP, H. H. (1996). Some salient lessons from research and practice. Psychotherapy, 33, 135138. YALOM, I. D. (2002). The gift of therapy: An open letter to a new generation of therapists and their patients. New York: HarperCollins.

389