The relationship between therapist and client, also called therapeutic alliance, joining, or building rapport, is an important aspect of the therapeutic process in helping relationships (Bordin 1979, 1994; Mahaffey and Granello 2007; Sprenkle et al. 2009). The therapeutic alliance is defined as a conscious, collaborative, working relationship between therapist and client along the dimensions of bond, task, and goal (Bordin 1979). It is important for researchers and therapists to consider the process of building an alliance, because a strong relationship between therapist and client impacts the therapeutic process and outcome. Researchers postulate that the therapeutic alliance is a predictor for therapeutic outcome (Laszloffy 2000; Pinsof and Wynne 1995; Quirk et al. 2014; Shade et al. 2015). Specifically, the stronger the therapeutic alliance, the more likely the therapeutic outcome is to be successful (Knerr et al. 2011; Martin et al. 2000; Porter and Ketring 2011). Research has begun to explore how the conceptualization of therapeutic alliance changes in couple and family therapy (Gurman et al. 1986; Knerr et al. 2011; Mahaffey and Granello 2007). Though a strong alliance is critical to a successful therapeutic outcome, how the therapeutic alliance is developed and maintained is not thoroughly understood (Taft et al. 2004). Even less is known about how the process of developing an alliance changes when the client is a couple or family (Knerr et al. 2011). In relational therapy, establishing and maintaining the therapeutic alliance with multiple individuals and relationships is challenging (Bertram 1996; Friedlander et al. 2011), particularly in light of attempting to balance alliance and neutrality with several family members (Brimhall and Butler 2010; Butler et al. 2011). Beginning therapists experience developmental challenges in the task of alliance building (Morrissette 1996). Despite research and education that many beginning therapists may receive on alliance building and its importance, the lack of information and guidance about building relationships with couples and families may leave beginning therapists at a loss as to how to begin therapy effectively with family systems. To begin to address this gap in the literature, the purpose of the present study was to understand the way therapists with differing levels of experience conceptualized rapport with couple and family clients.

Background

Systemic Therapeutic Alliance

A systemic therapeutic alliance is simply defined as a therapeutic relationship with a couple or a family client system. Consistent with systems theory, the family system is greater than the sum of its individual parts (Minuchin and Fishman 1981); therefore, the potential for the alliance with couple or family clients to become strained rises (Morrissette 1996). More recently, researchers are beginning to investigate the process by which therapists build alliances with couple or family clients (Knerr et al. 2011). Historically, marriage and family therapy (MFT) model developers and early theorists intended for the process of building a relationship with relational clients to be a systemically conceptualized process. In the early stages of the MFT profession, Minuchin and Fishman explained that therapists must become part of the family system in order to effect change within that family. Minuchin and Fishman intended for the therapist to become an active member of the family system through the process of building a relationship with the existing family relationships. Haley (1987) also described early therapy sessions as a time to allow therapists to “join the universe of the family” (p. 27). Similarly, Susan Johnson emphasized that therapists must join the couple system in order to effect change (Shade et al. 2015). Carl Whitaker discussed the therapists’ challenge of avoiding entrapment in the family system, often by holding to therapeutic techniques and skills (Napier and Whitaker 1977). Bowen (1976) described the ease with which a family system can emotionally overpower the therapist, making him or her responsible for the success or failure of treatment. Anderson (1997) explained that human beings have a natural curiosity that leads to “being in a relationship” (p. 61) and “experiencing each other” (p. 62) through language, connection, and genuineness. “The more we became immersed in our clients’ language and meanings and positioned ourselves as inquiring learners, the more we acknowledged, encouraged, and heard their voices” (Anderson 1997, p. 63). In developing this alliance, Minuchin and Fishman (1981) explained that a therapeutic alliance requires therapists to “transcend technique” (p. 1).

The writings of MFT model developers discuss therapy as an art form and encourage trainees to master and implement technique in the context of a genuine relationship with their clients. Examining the language used by MFT model developers emphasized their view of the therapeutic alliance as a systemic process. Building an alliance with relationships in the therapy room shows sophisticated systemic conceptualization. Therapists must be able to build an alliance with each individual client, as well as the relational system; if both levels of alliance development do not occur, the therapeutic outcome is often less satisfying for the clients (Bertram 1996).

Despite the clear systemic intention of the MFT model developers, the majority of therapeutic alliance research has been conducted with individual clients. While the body of research on the importance of the therapeutic alliance with individual clients is rich, minimal empirical data has investigated the impact of systemic therapeutic alliances (Knerr et al. 2011; Sprenkle et al. 2009). Notable exceptions exist. Pinsof and Catherall (1986) formally introduced the multiple complexities of relational alliances to the empirical literature in MFT with their development of the Family Therapy Alliance Scale (FTAS). The FTAS was explored further for family’s use by Johnson et al. (2013). Pinsoff and Catherall also developed the Individual Treatment Alliance Scale, Revised-Short Form (ITASr-SF) to measure therapeutic alliance and Owen (2012a, b) conducted a confirmatory factor analysis of the ITASr-SF the with participants in group or couple therapy. Results supported the importance of the systemic nature of therapeutic alliance. Friedlander et al. (2006) developed the System for Observing Family Therapy Alliances (SOFTA) which includes instruments for observational and self-report data collection on four dimensions.

Therapist Experience

Therapists may experience increased difficulty establishing and maintaining alliances with each member of a couple and family system due to the complexity of simultaneously balancing many client variables (Glebova et al. 2011). Beginning therapists often underestimate family therapy case complexity, applying a reductionist lens to family processes (Andolfi et al. 1993). Beginning therapists experience vulnerability, specifically with couple and family clients; therefore, they experience increased challenges (Morrissette 1996; McCollum and Gerhart 2010; Andolfi et al. 1993). Napier and Whitaker (1977) recognized the struggle with developing rapport with couple and family clients that new clinicians may experience. They recommended conducting co-therapy with trainees as a way of training inexperienced therapists to actively participate in treatment without becoming overly involved with the family, which they believe takes years of experience to master. In this recommendation, Napier and Whitaker implied that novice MFTs may conceptualize and experience therapeutic rapport with couple and family clients differently than do experienced clinicians.

Research Questions

The present study aims to understand the perceptions of therapists at various experience levels regarding a therapeutic alliance with relational clients. Our primary aims are: (1) Define how therapists with different levels of experience conceptualize the process of developing a therapeutic alliance with relational clients. (2) Describe the process of creating a systemic therapeutic alliance. (3) Discuss how results inform educators, supervisors, and therapists. The research question guiding this study was: How do therapists with varying levels of experience conceptualize the process of building a therapeutic alliance with couples and family clients?

Method

Participants

Eighty participants provided data via 14 focus groups. Sixty-one (76.6%) participants indicated ethnicities of European descent, with 12 (15%) choosing not to respond. Three (3.8%) participants identified as African-American; two (2.5%) participants identified as multiracial and two (2.5%) as Latino/Hispanic. Sexual orientation and socio-economic information were not collected. Participants were recruited purposefully based on professional level. See Table 1 for experience levels of each group, number of participants in each stage, and basic demographics.

Table 1 Participants’ experience levels, group numbers, and participants’ demographics
Full size table

To be eligible to participate in the MFT or MFC student focus groups, people needed to identify as marriage and family therapy or counseling students or practitioners specializing in family clients. Participants were required to have reached the Pre-Clinical courses in their programs. Data were collected from students at three universities, spanning different states and training levels (two master’s level programs and one doctoral program). Participants were also recruited through professional workshops. One master’s program has a CACREP-accredited marriage and family track, while the other programs are COAMFTE-accredited. We chose to recruit participants from two types of programs, marriage and family therapy (MFT) and marriage and family counseling (MFC) because we were interested in the overarching themes of alliance. Training in both programs relied heavily on systems theory. Data were also collected from practicing therapists who specialized in working MFT. These focus groups were held in the United States, in cities in the northwestern, southeast, and mid-Atlantic.

We pre-planned the number of focus groups based on the research questions and added groups as needed to reach data saturation. Because we initially compared groups with differing experience levels (Mallinckrodt and Nelson 1991), we held a minimum of three focus groups for each level of experience. We set the range of participants per group at n = 3–12. Availability of participants and interest led to a total of 14 focus groups. As we compared field notes and analyzed data, the data reached saturation before the 14 groups were completed, but having multiple groups within each experience level allowed us to analyze data within and between experience levels. Through analysis, the groups fit into two main categories with the lower two levels of experience expressing similar perspectives and the higher two levels (1000 h of clinical experience) expressing similar perspectives.

Recruitment

A variety of recruitment methods were used to solicit participants with varying levels of experience. Recruitment methods included an email sent to students in each program and also through student announcements. Practicing MFTs were recruited in multiple ways. Recent graduates of one of the masters programs received a recruitment flyer with an invitation to an alumni event and asking them to participate in a focus group before or after the event. Experienced MFTs were recruited at a university-sponsored supervision workshop and a MFT-focused training. In both situations, professionals who registered for the workshops were emailed information about the study prior to the workshop, and the study was announced. Therapists who met the criteria were invited to participate in focus groups during lunch.

Procedures

Participants were organized into focus groups based on experience levels. Focus groups allowed participants to share ideas with each other and build on the thoughts of others (Morgan 1997). In addition to the research-based benefits of focus groups, they also created a positive, energetic atmosphere in which participants learned from and supported each other. When each group began, researchers explained the study purpose and asked participants to complete a consent form and a demographic form. Informed consent was obtained from all individual participants included in the study. One moderator and one note-taker conducted focus groups of three to eleven participants. Focus groups were audio recorded, and participants are identified by pseudonym. Focus groups facilitators followed the protocol, and groups generally lasted between 45 and 60 min. Participants were incentivized with games, books, or items for therapeutic practice valued at five to ten dollars. After groups, researchers reflected on their initial impressions and noted their experiences, perceptions, thoughts, feelings, and possible initial themes.

Measures

The focus group protocol that we used was designed for this study. Questions were based on alliance literature and supervision experience and were reviewed by colleagues prior to use. Open ended questions focused on defining therapeutic alliance, perceptions of differences between systemic and individual alliance, assessing systemic alliance, strategies used to establish an alliance, building an alliance with diverse clients, and handling damaged alliances with couple and family clients. Throughout the research process, our terminology for building a therapeutic alliance varied. Our research questions began with “relationship building” as a more casual way of discussing the connection between client and therapist to facilitate discussion. We used the term “rapport” in a few of the protocol questions. We noted that participants used “rapport,” “joining,” “relationship,” and “alliance” interchangeably and we echoed their language in our follow-up questions. Examples of protocol questions include: (1) How do you define “rapport” in the context of counseling or therapy? (2) How does this definition change when the client is a couple or family? (3) What is an example or experience when you felt that you have done well establishing rapport with a client? (4) Talk about building a connection with clients who differ greatly from you. We frequently followed up by asking: Can you think of a specific situation with a client that provides an example of that? Groups were asked specific questions based on their level of experience. For example, new clinicians were asked, “As you think about working with couple and family clients, do you have any concerns about establishing a connection with clients? If so, what are they?” and experienced clinicians were asked, “What advice would you give students or newer professionals regarding establishing rapport with couple and family clients?”

Data Analysis Procedures

Focus group data were transcribed and formatted in Excel. Each transcription was verified by having one researcher listen to the recording while reading the transcript. Any errors in transcripts were corrected. Four researchers in two teams analyzed the data. The researcher teams each included one doctoral level researcher who had completed qualitative projects and had clinical experience in marriage and family therapy. The additional two researchers were one doctoral student and one master’s student who were studying qualitative research. The researchers met multiple times in their dyads and as a complete team to discuss analysis procedures and coding schemes. Analysis consisted of open, axial, and selective coding, in a modified grounded theory technique (Glaser and Strauss 1999). The process involved multiple steps. First, the researchers each used open coding in which we looked for themes in the comments and allowed the data to continually refine the themes. As a theme changed, we re-coded previous data to confirm or adjust the theme. In this step, the researchers worked individually on a few transcripts within one experience level group of participants (e.g., Pre-Practicum groups) and then met to discuss our themes. Our discussions led to multiple iterations of the coding scheme. Once the coding scheme was fairly well established, the team worked in dyads, coding transcripts together until we agreed on the theme which best fit each participant comment. Then the teams coded transcripts individually and compared coding on the same transcripts and discussed themes to ensure consistency. After working with one experience level group of participants, we began the process again with another experience level. However, we soon realized that the themes were almost identical, with only a few ideas distinguishing between levels of experience. Therefore, we created an overall coding scheme and allowed the differences between experience levels to emerge within the themes.

Once we were coding consistently across researchers, we divided the transcripts between researchers and each transcript was assigned a researcher to do the main coding and two researchers to verify the coding. Any discrepancies were noted and discussed until an agreement was met. We then organized the data by theme and read all the data for each theme to ensure consistency in thought and message. In this process, we further refined the themes and better understood the participants’ essential messages. Finally, we focused on the relationships between themes. We did this by re-reading the original transcripts, reviewing our field notes, and paying careful attention to any themes that seemed to relate consistently to each other. We also divided the data by participants’ experience level and looked for variations by experience. Themes which preceded other themes, themes which consistently promoted a follow-up theme, themes described as mutually exclusive or opposing themes, themes which were described differently by novice vs. experienced participants—these relationships between themes elucidated the overall perceptions of participants and enabled the data to emerge as a grounded theory. The final coding scheme is provided in Table 2.

Table 2 Final coding scheme for MFT alliance analysis
Full size table

Qualitative rigor is established through trustworthiness (Patton 2002). To support trustworthiness and credibility we used the following strategies: the researcher who was present during the focus group verified the transcription, the researcher team frequently debriefed after focus groups and during analysis, and we each wrote and shared our field notes (Anfara et al. 2002). Field notes contained notes about unique ideas participants presented in the focus groups, ideas introduced in focus groups that were confirmed in subsequent groups, points which participants stated briefly which were quickly confirmed or rejected by other focus group members but may not have stayed the focus of the group. We discussed these field notes and referred to them when we developed the coding scheme and during analysis. We also triangulated transcripts and field notes by analyzing for similarities and differences. We collected data throughout a year, allowing time to reflect and process as we collected and analyzed it. In writing the project report, we attempted to bring readers close to the data by using several participant quotes to illustrate each theme (Anfara et al. 2002).

Findings

The themes of the findings applied across all groups of participants. Therefore, most of the findings are presented with quotes from participants of all experience levels. Where there are distinctions based on participants’ experience levels, those are noted within the themes.

As a basis of the findings, focus groups began with the facilitator asking participants how they defined rapport. Their responses presented a foundation for understanding therapeutic alliance and centered on vulnerability, trust, and openness. James, a doctoral student, said, “I describe [the alliance] as an ongoing positive relationship that’s built on trust and mutual shared belief in—that therapy can work.” Hank, a newly licensed counselor, shared his experience of positively aligning with a client: “when you see that healthy balance of disclosing and their vulnerability, … that can … indicate that there has been trust that’s developed.” An experienced clinician defined systemic alliance building as “a two-way thing; they’re sizing me up, I’m sizing them up. [They are deciding] if they feel that I can be trusted, if I have solutions or knowledge that will be helpful to them.” As these quotes show, therapeutic alliance has a human element and an intuitive connection. Participants were adamant that therapeutic work could not progress until an alliance was established—the therapeutic relationship was the pre-requisite for engaging therapeutically.

Emergent themes in the data report participants’ understanding of how systemic alliance was created, monitored, and maintained. Emergent themes spanned three major concepts. Two of these themes were: (1) Create the alliance and (2) Monitor the system’s complexity. The third theme, (3) Balance individual and systemic needs overlapped with the previous themes. Every focus group discussed each of the three themes. The findings section details participants’ perspectives of each of these themes and subthemes.

Create the Alliance

Participants discussed their experiences creating therapeutic alliance and offered techniques and conceptualizations. Three subthemes emerged: (1) developing a sense of trust, (2) relating genuinely, and (3) basing all techniques in a systemic understanding.

Developing a Sense of Trust

Creating a safe and trusting therapeutic context emerged as the first step in creating a therapeutic alliance. This included a safe physical and emotional environment for each individual client and the participant. Trust was seen as an essential prerequisite for effective therapy. Sean, an experienced MFT, asserted, “There has to be safety for any therapeutic relationship to really progress.” Participants at all levels of experience embraced the importance of creating trust. Louisa, a Pre-Practicum student explained the value of trust, “I think of building trust so that [the clients] will feel free to share openly what they’re thinking and feeling, even if they’re negative in their views.”

Participants viewed the process of developing trust as beginning from the outset of therapy. Jackie, doctoral student, explained:

Making sure clients and families feel safe, that it is a warm and inviting environment for them to be honest and hope [to develop] that trust. Within a first session, taking a lot of questions, if they are really confused about how things work. With attire, if I have a family with kids, I will wear like funky earrings [to connect] with them.

This illustrates the importance of attending to details early as a way to build therapeutic alliance.

Relating Genuinely

Relating genuinely involves using one’s self in the therapeutic relationship. Being genuine included: being open, using one’s personality, connecting personally with clients, and using one’s intuition to understand how to establish a human connection.

Jenna, a 1st-year Master’s student described the importance of being genuine:

Clients will pick up on not being genuine and having that front will get in the way. If they have a sense that you are not being real with them, then there is no chance they are going to be open and real with you.

The idea of being genuine connected therapist’s interests and experiences with the clients, through self-disclosure at times. A pre-practicum student in a counseling program explained the connection between genuineness, openness, and client sense of safety:

You have to really make sure that you’re portraying yourself to be nonjudgmental, and that you’re interested in learning about them, and … trying to understand their point of view, because it is different from your own.

When participants talked about developing rapport with diverse clients, genuineness and the use of self-disclosure was frequently discussed.

The use of personality characteristics was also often part of being genuine. Art, a 1st year master’s student in a COAMFTE program, described how he used elements of his personality to add to the comfort clients experience in the therapy room. He stated:

I have only seen a couple family clients, but both times, I tried to make them laugh because I didn’t want it to be so serious, I wanted it to be kind of a fun, you are trying to enjoy each other and enjoy this process and not trying to be so dry about your emotions.

Basing all Techniques in a Systemic Understanding

A final subtheme that emerged from this theme of connection related participants’ systemic conceptualization of building therapeutic alliances with relational clients and included emphasis on specific concepts like boundaries and subsystems. Creating boundaries for therapy helped clients know what to expect and feel more comfortable in sessions. As Jackie, a 2nd-year master’s student explained,

When I think of therapeutic rapport, I think … set clear boundaries like in the very beginning so that the relationship can be established so you know where things are going to go. There’s kind of like a set up, you kind of know how this is going to work…

An experienced MFT expanded on conceptualizing boundaries with couples, “[Clients are] reassured that I’m not going to be playing favorites, that I’m not … holding a secret. … [t]here are … some rules about the openness and about the boundaries.”

Participants recognized the challenge of building therapeutic alliances with a system while not damaging the alliance with sub-systems. Lola, a 2nd-year student in a CACREP program, described,

You have to try to have an equal relationship with all members of the family or a couple, … they both can trust you. [You don’t] … take sides –you will be objective and hear both of them. … [T]hat, obviously, is harder than with just one person, because you’re dealing with different sides of the same story. So maybe you’ll have to work harder to have a good relationship with all members of the family.

Participants set the stage for what they hoped would be a trusting and intimate therapeutic relationship. They used the human connection in the room and their understanding of systems to create and maintain rapport.

Monitor the System’s Complexity

Maintaining systemic therapeutic alliances involved an assessment and monitoring process. Participants noted that when clients differed from therapists (e.g., age, culture, ethnicity, sexual orientation, religion, etc.), clients may perceive therapists differently than the therapists intended and that therapists may misinterpret client non-verbal cues. To mediate this, participants more overtly and/or frequently assessed diverse clients’ perception of rapport. Within assessment of the alliance, three subthemes emerged: (1) maintaining a systemic conceptualization, (2) accessing informal and formal assessments, and (3) developing one’s own “intuitive barometer.” Participants continually acknowledged the complexity of working with relational systems, especially as it related to their role in building an alliance with the system. This awareness increased as participants began the later phases of therapy. This assessment was often highlighted by an intuitive moment, an “ah-ha” moment, where the therapist knew that the therapeutic alliance had been established and “work” could begin.

Maintaining a Systemic Conceptualization

The use of systemic terminology and understanding was the underlying way that therapists assessed the therapeutic alliance on a continual basis. Danielle, a 1st year MFT Master’s student shared her understanding of this systemic lens:

[Building a relationship is] complicated, there are a lot of people, where connections and bonds, that you have to be aware of and conscious about, not just yourself and the client, but between the clients. It just makes the situation much more complicated.

Another student echoed Danielle’s comments by building in systemic language. He stated:

[Building a relationship with a system is] more difficult, because they each have different expectations. A common theme is, “Fix my spouse,” and so if they sense that you’re not going to do that, then, [they think] “Whoa, I’ve got some work I’ve got to do on this, too;” it makes it more difficult to establish rapport … they want to build a triangle with you against that spouse sometimes.

Jane (a recently licensed counselor) advanced Danielle’s comments to the next stage of therapist skill when she explained that part of balancing those dynamics involved attending to her role, “… it’s a balance, and you have to figure out how to manipulate it, because sometimes you can end up being that third—that triangulation …” Systemic understanding helped participants recognize the importance of the therapeutic relationship to therapeutic success and implementing this systemic understanding in therapy grew as participants gained experience.

Accessing Informal and Formal Assessments

Every group talked about therapeutic alliances as continually changing, and emphasized the importance of monitoring alliance. James, a doctoral student, illustrated the fragility of therapeutic alliances using systemic language:

I don’t ever want to get comfortable, because chances are … you can always do something more to increase a person’s confidence in you, or increase their comfort with you. I try not to ever go into a session where I think that I don’t have to continue doing the work of building a therapeutic alliance. … continually checking in with clients about that relationship is the most important thing.

At times monitoring the development of therapeutic alliances included recognizing when alliance was broken or had to be re-established. Pam, a recently licensed counselor, related this story that highlighted a broken alliance:

I had a couple … and [to] the husband I said that he wasn’t a very good listener. He got really upset and he actually … stormed out. And I definitely had an “Oh, crap” moment.

Kirk, an experienced MFT, looks for these signals of a strong therapeutic alliance, “Do they show up on time? Are they there even early? When they walk in, are they eager to get to things? Does it look like they’re glad to see me?” Participants used informal assessments to monitor the alliances by directly asking clients how therapy was going, such as when one participant described, “[S]imply asking… “How are we doing?” I mean, … if you expect your clients to be honest with you, I think it’s fair for you to be open with them … just asking them.”

In addition to using informal observation and direct questioning to monitor therapeutic alliance, several participants explained that they used formal assessment tools to invite the client’s perception of the therapeutic alliance. Shaune, a 2nd year Master’s student described,

I use my gut, but I also use a scale, I have them rate the session and we talk about it at the end of the session, so I kind of have a sense if what I am doing is clicking with them. It has been very helpful.

Part of assessing and getting feedback focused on the participants’ willingness to be open to that feedback and to adjust as needed. The monitoring of a therapeutic alliance was a consistent process of connection to families and understanding families from one’s “gut” while observing and implementing assessments.

Developing an “Intuitive Barometer”

Participants actively assessed therapeutic alliances throughout the process of couple or family therapy using their intuitive sense, their “gut.” When one participant called this an “intuitive barometer” the other participants in that group readily nodded in recognition and agreement. Participants seemed to recognize this as important but struggled to clearly define this experience and turned to metaphors or examples to explain their thoughts. Molly, a doctoral student, shared her understanding of the therapeutic relationship as

a gut feeling that I know that I can say those unspeakable things and that they will come back and that it won’t be a bomb dropped that will be destructive… I can’t describe [how to assess the therapeutic alliance] any other way than that.

This intuitive barometer was based on participants’ human connection with their clients. Rather than using interventions to create this human connection, participants described it as being “in sync” with clients. Marisol described connecting on a human, intuitive level with a family of an ethnicity different than hers: “We were all laughing, we were making jokes, and I was included… [I]t felt like I was kind of part of that family, for just that half hour. … I felt I had a relationship with them.” Kate, a 1st-year master’s level student in a COAMFTE program, described a human element of humor and increased client disclosure when she stated, “[Y]ou can see if they start joking or their personalities start coming out, you can feel it in the room sometimes, and they start sharing more information as well, like a secret or something personal.” Kate assessed therapeutic alliances intuitively, by attending to a feeling in the room. This connection to instincts and human elements was frequently discussed as a critical component of systemic relationship development. Layla, a 1st-year Master’s student in a COAMFT program shared:

As we gain experience, we also gain the intuition and the ability to sense that something is not being said and sense kind of where the connection level is. … I do believe it comes with experience.

Balance the Individual and Systemic Needs

Participants in every group discussed the importance of connecting with each individual as well as the system. Differences between groups of therapists with different experience levels emerged most drastically along this theme. Less clinically experienced participants conceptualized balancing the system as building relationships with each individual in the client system. These participants spoke about building a systemic alliance as if they were jugglers attempting to keep all the props in the air at the same time. Jennifer, a newly licensed clinician, explained building an alliance individually and looking at individual characteristics, rather than the system as a whole when she explained that her strategy is to “… identify what the dynamics are and maybe some of the personality traits, and [figure] out how to balance that … as you’re building the rapport.” Sunny shared, “each client has to be … treated differently. You use the same set of skills, but each client is approached differently.”

In contrast, therapists with years of experience conceptualized developing a relational alliance by focusing on systemic patterns of interaction. As therapists gained experience, they shifted their focus from the individual to the entire system. The struggle of making this shift can be seen in Lola’s quote, used in a previous section. As a 2nd-year student, she was realizing the need for systemic rapport but did not yet have the ability to see the system as a unit and was still juggling rapport with individuals. Once they were able to develop a systemic focus, experienced therapists tolerated small rifts in the alliance with individuals. Experienced therapists did not at all neglect the importance of alliance with the individuals. Rather, they held a macro level perspective that led them to feel confident that individual clients could tolerate small ruptures in the therapeutic alliance for the longer-term benefit of the system. Experienced therapists maintained focus on the system’s needs rather than the individuals’ needs. One experienced therapist bluntly said, “you can’t do it all at once all the time.” Therefore, they focused the majority of their energy on joining with the system. Barbara, a doctoral student, sounded more like the group of experienced therapists when she explained the concept of this clearly. She explored building an alliance with individuals, subsystems, and the family unit:

You can have a relationship with the individual, but also the couple, you can make them annoyed at the same time as a couple, but you can still have that union with them. … And, you may not have potentially individually connected with all of them, but like in a family, you may have parental rapport and children rapport, like dual bonds within that family. I think sometimes the unit of rapport carries you through those low points of individual rapport.

Participants balanced the systemic alliance by moving between challenging individuals and adhering to the rules and boundaries of the system; all while maintaining a systemic focus. Participants sensed that they needed to be part of the systemic elements in order to build the therapeutic alliance. A newly graduated therapist shared, “I want to be accepted into that system, into their group.” As they discussed joining the client system, participants also recognized the need to balance joining the system with ethics (e.g., not accepting unhealthy or harmful behaviors even when the system did), maintaining personal boundaries (e.g., inappropriate rapport could include personal friendships), and therapeutic goals (e.g., rejecting the systems’ overt and covert rules when those rules inhibited the client in progressing toward goals).

Discussion

Participants’ ideas about developing a systemic therapeutic alliance were consistent with Minuchin and Fishman’s (1981) perspective of the therapist as a healer, a human being concerned with engaging other human beings about things that cause them pain while retaining great respect for their values, areas of strengths, and preferences. Minuchin and Fishman (1981) discuss that one goal of therapists is to transcend technique. While the findings of this study confirm the fundamental nature of the therapeutic alliance (Blow et al. 2007; Bordin 1979, 1994; Hubble et al. 1999; Mahaffey and Granello 2007; Sprenkle et al. 2009), they also show that relational therapeutic alliance is fundamentally different than individual alliance (Owen 2012a, b; Karam et al. 2014b, c). While there are similarities between alliance-building skills with individual clients and systemic clients (e.g., nonverbal skills, active listening, setting goals), participants who focused on developing an individual alliance expressed difficulties implementing these strategies effectively with couple and family clients, and their attempts distracted them from smoothly moving forward in therapy with couple or family clients. Perhaps the most unique findings of the present study are the distinctions in therapeutic alliance building techniques between novice and experienced therapists. The present study identified tangible distinctions between the techniques and thought processes of novice therapists and experienced therapists.

A Model of Systemic Alliance in Therapy

Findings of the present study support the importance of conceptualizing systemic therapeutic alliance as distinct from individual therapeutic alliance and put systemic alliance within the context of couple and family therapy. The findings of this study blend with previous literature on alliance and rapport to create a model of systemic alliance within the context of couple and family therapy. The model is presented in Fig. 1.

Fig. 1
figure 1

Model of systemic alliance in therapy

Full size image

In this model, first consider the contextual aspects of the model. Systemic alliance is the underlying context in which the work of therapy occurs, as visualized by the shadowed background. The shape at the bottom of the model, which states “Consistent systemic conceptualization in implementing treatment” signifies that the consistent systemic conceptualization is the foundation for building the systemic alliance and for the work of implementing therapy. The arrow on the foundation shape conveys the context of therapy progressing toward goals. Systemic goals and shared purpose are represented on the right of the model. The development of systemic alliance and therapy implementation are dynamic and the strengthening of systemic alliance is indicated by the darkening of the shade of the systemic alliance background shape and the tightening of the shape as the therapist and client are working more closely together. As the foundation shape indicates, systemic alliance building is occurring as therapy is working toward shared goals or purpose. Note the relationship between the systemic alliance and the systemic goals and purpose on the right. As the therapeutic system focuses on shared purpose (Friedlander et al. 2006), clients and therapists agree on goals (Bordin 1979), and clients make progress toward their goals, the systemic alliance is strengthened. Participants from the current study confirmed that shared goals and client progress strengthens the systemic alliance.

Having established the contextual aspects of this model, consider the components in the center of the model. The components (a) create, (b) balance, and (c) monitor, work together and occur simultaneously. When therapists create systemic alliance, therapist genuineness and care for clients as well as an agreed upon structure that values the uniqueness of the clients help create the essential sense of safety and trust (Minuchin and Fishman 1981). Friedlander et al. (2006) included client sense of safety and trust as one of the four components of systemic alliance, which participants from this study emphasized as vital from the beginning of the first session of therapy. The agreed upon structure includes reviewing the informed consent with clients and ensuring everyone has the same expectations of the logistics of therapy. The sense of safety also requires therapists to be culturally competent and honor client diversity (Friedlander et al. 2006). In addition, a sense of emotional connectedness between the therapist and the client system is an aspect of creating the systemic alliance (Friedlander et al. 2006) and incorporates Bordin’s (1979) emphasis on bond between therapist and client as part of the therapeutic alliance.

As participants in this study highlighted, balance is another component central to systemic alliance in therapy. Therapists balance the needs of the individual and the system and navigate this balance with clients. Researchers have discussed the challenges therapists face when balancing alliance and needs of individuals and systems with couple and family clients (Glebova et al. 2011; Morrissette 1996; Napier and Whitaker 1977). Brimhall and Butler (2010) and Butler et al. (2011) discussed the tension and balance of “alliance” and “neutrality.” In this context, the authors explain that “alliance … suggests an engagement between the therapist and system members that is emotionally empathic and in some meaningful way, validating and ‘partial’ (Brimhall and Butler 2010, p. 29). In therapy with couples and families, this balance requires particular focus because validating one client in the system may damage rapport with another. Additionally, the therapist balances the therapy role and deliberately moves in and out of the system to facilitate systemic alliance and to promote progress toward goals (see Haley 1987; Napier and Whitaker 1977; Shade et al. 2015).

The third component, monitor, includes informal and formal methods of assessment to monitor the therapeutic alliance and client progress toward goals. Findings of this study emphasize the results of D’Aniello et al. (2019), who explored why MFT clients quit or continue therapy. D’Aniello and colleagues found three primary factors: (1) Client motivation, (2) Therapeutic relationship, and (3) Progress toward goals. Therefore, monitoring these aspects of therapy not only supports the development of the systemic alliance but also alerts therapists to cases at high risk of dropout. To monitor the therapeutic alliance, informal methods of assessment, which include any methods that do not use standardized tools, may be used. For example, when monitoring the therapeutic alliance, participants in the current project mentioned observing client nonverbals (e.g., clients who perceive the therapist as aligning with another person in the system may avoid eye contact with the therapist, turn away physically, and/or shrug or roll their eyes), asking clients their thoughts on the alliance, and listening for clients to disclose new, personal information. Formal methods of alliance assessment are addressed and strongly recommended in literature, particularly because clinician and client ratings of the alliance differ and clients’ ratings consistently predict therapy outcome (Horvath and Symonds 1991). Formal assessments of therapy alliance or rapport are available. For example, when the Session Rating Scale (SRS; Duncan et al. 2003) was used in couples therapy, couples who used the SRS and the Outcome Rating Scale (ORS, a standardized assessment for client functioning) (Miller et al. 2003) progressed toward change significantly more effectively and maintained changes longer than couples whose therapy did not include the SRS and ORS. Friedlander et al. (2006) developed the System for Observing Family Therapy Alliances (SOFTA) to formally assess systemic alliance and Owen (2012a, b) applied the Individual Treatment Alliance Scale, Revised-Short Form (ITASr-SF) to couple therapy. The FTAS was developed specifically for family clients (Pinsof and Cathrall 1986). Artigas et al. (2017) provided an example of using the SOFTA to assess couples therapy throughout the process and explored the four dimensions of therapeutic alliance measured by the SOFTA.

Additionally, therapists use informal and formal methods to monitor client progress toward goals. Informal methods could include reviewing case notes, discussing progress with the client, and observing clients’ changing affect and interactions over time. Standardized assessments focus on symptom reduction, functioning, and satisfaction scales and should be selected based on clients’ individualized goals. Participants in this projects noted that client progress strengthens the alliance. Therefore, if clients are not progressing toward goals, the therapist could hypothesize that the systemic alliance may be weak and address this in therapy either directly or indirectly. Monitoring client engagement is also a signal of systemic alliance and is needed for the client to progress toward goals (Friedlander et al. 2006). This aspect of client engagement overlaps with Bordin’s (1979) inclusion of “task” in therapeutic alliance; where therapists and clients are aligned on the tasks of therapy, client engagement is likely to occur. In addition to what is happening in therapy, therapists need to monitor the impact of outside influences or extratherapeutic factors, as extratherapeutic factors are linked to therapy outcome (Sprenkle et al. 1999; Perkins 2010). This could include things like cultural support or distain for therapy as well as resources available that make it possible for clients to access therapy (Perkins).

The monitoring component creates a feedback loop, indicated by the arrows in the center of the model. Monitoring is a continual, recursive feedback process. The arrows under the “monitor” component indicate what therapists would do with the information about alliance. As systemic alliance decreases, or there are rifts in the alliance, the therapist would challenge clients less and rebuild alliance using strategies such as those described by Karam et al. (2014c), Friedlander et al. (2006), and Safran et al. (2011). As systemic alliance strengthens, the therapist can challenge clients therapeutically to increase progress toward goals. This model captures the findings of the current study and key findings in previous literature and can guide therapists and educators in developing systemic alliance.

Clinical and Training Implications

Clinical Practice

This model can assist clinicians’ understanding of how systemic therapeutic alliances impact clients and therapy outcome, considering both the individuals and the system, and the impact that alliance building has on treatment options. At the forefront of clinical work is the reminder of the importance of therapeutic alliance. Additionally, the present study highlights that when it comes to the therapeutic alliance, experienced and novice therapists do things differently. This study confirmed that experience is in fact a teacher when it comes to the craft of doing therapy (Doherty 2012). Attending to the model of systemic alliance in the context of therapy can help therapists deliberately address each component of systemic alliance. Therapists can integrate into therapy informal and formal methods of assessing the components of systemic alliance and client progress. This feedback data will offer guidance and support for therapists as they work with couple and family clients.

Training

Finding that there were differences in understanding of systemic therapeutic alliance between novice and experienced clinicians suggested that students would benefit from a training module for building systemic therapeutic alliances (e.g., Carpenter et al. 2008). Although there is no substitute for experience, focused training may help the novice more quickly know what to look for and capture the lessons of practice and experience. This could begin with literature providing the evidence of the importance of the therapeutic alliance for client outcome. Faculty and supervisors could teach conceptualization of and specific skills for building a therapeutic alliance with couple and family clients. Building on the founders of MFT, students should learn about joining and identify when to and how it looks to “enter” and “exit” family systems. Students would benefit from recognizing and practicing informal and formal methods for assessing systemic therapeutic alliances throughout therapy, and beginning to develop their own intuitive barometer. Formal methods of assessing the relational alliance could provide opportunities for integrating research in training (Hodgson et al. 2005; Friedlander et al. 2006). Additionally, instructors could teach how to discuss therapeutic alliances with systems and how to repair broken alliances, perhaps using information from Butler and Bird (2000) about recognizing and responding to client resistance or struggle.

Therapeutic training often consists of the understanding of “self” in the context of therapy (Aponte et al. 2009; Lutz and Irizarry 2009). Building on this training by looking at what the therapist brings into the room (e.g., humor, calmness, instinct) would strengthen the understanding of how to build and maintain the therapeutic alliance. Because several participants said they used self-disclosure as a way to build therapeutic alliances, particularly with diverse clients, MFT educators should explicitly teach about the use of appropriate self-disclosure. The module could include ideas for building therapeutic alliances with clients who are very different than the clinician. This certainly includes multicultural considerations, but also involves awareness of relational dynamics that are uncomfortable for the clinician.

Karam et al. (2014a) supported the importance of including common factors in MFT training. In addition to emphasizing MFT models, common factors draw our focus to the curative elements of the therapy process (Davis et al. 2012). The focus on common factors also aligns with and supports the emphasis on honing general therapy skills (Sprenkle and Blow 2004). The present findings support emphasizing basic therapy skills, such as building the therapeutic alliance, in addition to the emphasis on learning models.

Supervision

Much of the couple and family therapists’ growth takes place through experience with systemic clients and supervision that focuses on understanding that system (Imber-Black 2014; Rampage 2014). One way to build in identifying systemic alliance building is in the early stages of supervision, where students are observing cases. In this teaching process, the supervisor would verbally discuss the case conceptualizations and include the systemic alliance, selecting strategies for developing systemic alliance with specific clients, and assessment of the alliances. Supervisors would recognize the developmental component of systemic alliance building and help supervisees move from attempting to juggle individual therapeutic alliances with several individuals at once to joining with the system as a whole. This could raise anxiety in some supervisees, as they then must prioritize the system over discomfort of individuals. Supervision would guide clinicians in assessing when this is appropriate and when to spend time regaining therapeutic alliance with an individual for the good of the system. Helping supervisees apply systemic concepts in ways that include the alliance can aid supervisees as they learn appropriate boundaries and move in and out of the system deliberately.

Limitations

While the focus and findings of this study contribute to the literature on the therapeutic alliance with couple and family clients, it is not without limitations. Though all potential participants were contacted at each campus, the differing recruitment methods used on the different campus may have appealed to potential participants differently. Though the present sample is large by qualitative standards, it is not diverse with race, culture, or gender. Further, the sample has more student and novice therapists than experienced therapists. Experienced therapists were more difficult to recruit and less likely to participate in research. This may occur for several reasons: not wanting to cancel clients, preferring to use lunch as a break, and limited interest in research. Additionally, there was a greater proportion of males in the experienced group than in the other groups, which could result in gender confounding the findings about differences between experience levels. Another possible limitation of this study is that we sampled both MFTs and MFCs; though the data did not appear to differ between groups, there may be variation in their training of systemic alliance building. Finally, this study relied on clinician perception rather than client perception. Research has shown that therapists’ experience of therapy may not match clients’ experience of therapy (Sundet 2011). Future research could include client perspective of systemic therapeutic alliances.

Conclusions and Implications for Future Research

Future research examining the process of therapists go about developing the therapeutic with systemic clients could provide a solid base for this essential component of MFT. Process research is of great importance to the MFT field; it is an important step in narrowing the research practice gap (Oka and Whiting 2013; Pinsoff and Wynne 2000). Further research on the client perception of systemic therapeutic alliance would be a useful way to continue this research. Additionally, studying therapist development to examine whether and how this occurs might be helpful in training.

The growing emphasis on client outcome assessment and feedback informed treatment encourages deliberate, frequent assessment of the therapeutic alliance (Knerr et al. 2011). Although some studies have assessed alliance in treatment of couples (Knerr et al. 2011), individual tools were used and each person in the client system was asked to complete a therapeutic alliance measurement individually. This approach, while logical, neglects the possibility that an individual client in a treatment system may not have strong relationship with the therapist, but the therapist could still have positive relationship with the couple or family unit. MFTs would benefit from tools that could assess the therapeutic alliance with from a systemic perspective. The SOFTA (Friedlander et al. 2006) may be a useful tool for a systemic purpose and would benefit from exploring how the data from each person in the client system and the therapist could be combined to create a measure of overall and nuanced systemic alliance.

The findings of this study add information to the field’s understanding of systemic alliance from the perspective of therapists at various stages of development. The findings of this study were combined with previous research on therapeutic alliance and a model for systemic alliance in the process of therapy was presented. This model can be used to support clinical work, training, and supervision of systemic clients.