(p. 351) The Real Relationship
The real relationship in psychotherapy has been scientifically neglected for many decades. However, the idea of a real relationship between therapists and patients has been around for a long time, probably since the beginning of the “talking cure.” This element of the therapeutic relationship emerged from psychoanalysis. Freud (1919, 1937) referred to it when he remarked that “not every relation between an analyst and his subject during and after analysis was to be regarded as transference; there were also friendly relations which were based on reality and proved to be viable” (1937, p. 222). Other early psychoanalysts also commented on a personal, non-work connection between the analyst and analysands.
Perhaps the most incisive comment from these early analysts was offered by Anna Freud (1954), when she wrote:
[W]ith due respect for the necessarily strictest handling and interpretation of the transference, I still feel that somewhere we should leave room for the realization that patient and analyst are two real people, of equal status, in a real relationship with each other. I wonder whether our at times complete neglect of this side of the matter is not responsible for some of the hostile reactions we get from our patients and which we are apt to ascribe to “true transferences” only. (p. 372)
Anna Freud’s comment was typical of analysts’ views in the sense that the real relationship was seen as the counterpoint to transference. In other words, the word “real” was taken to mean “realistic” in that the real relationship involved each participant perceiving and experiencing the other in ways that befit the other, rather than through the lenses of transference. By contrast, the latter would involve a displacement of past unresolved conflicts with significant others onto the present therapeutic relationship. Thus, when transference was happening, the therapeutic relationship was experienced (p. 352) in a distorted way, whereas the real relationship represented the realistic, nondistorted element of the relationship.
A second key element of the real relationship was highlighted by the psychoanalyst Greenson (1967). This aspect concerned the participants in the analytic dyad being genuine with each other, that is, being themselves in the relationship rather than holding back or being artificial. In this way, the real relationship is connected to the humanistic/experiential therapies that placed genuineness or congruence as the centerpiece of treatment (Perls, 1969; Rogers, 1957). Greenson’s conceptualization focused both on realism and genuineness as the key elements of the real relationship. This two-part conception has been a fundamental element of current thought and empirical research on the real relationship (Gelso, 2014; Wampold & Budge, 2012).
Although the real relationship has its roots in psychoanalysis, the current literature focuses on the real relationship as transtheoretical, applying to all theoretical orientations in psychotherapy (Gelso, 2009, 2011; Gelso & Carter, 1985, 1994; Gelso & Silberberg, 2016; Wampold & Budge, 2012). Theoretically, the strength of the real relationship should not vary according the therapist’s theoretical orientation, and existing empirical evidence supports this expectation (Gelso, 2011).
Clarifying and extending Greenson’s (1967) psychoanalytic conceptualization, the modern transtheoretical definition of the real relationship has been “the personal relationship between therapist and patient marked by the extent to which each is genuine with the other and perceives/experiences the other in ways that befit the other” (Gelso, 2009, p. 119; emphasis added). Thus, the real relationship consists of two fundamental elements: realism, or realistic perception/experience of the other, and genuineness. The more realistically the participants experience and perceive each other, and the more genuine they are with each other, the stronger the real relationship is within the overall therapeutic relationship.
These two elements, realism and genuineness, may be further divided into what has been termed magnitude and valence. Magnitude refers to how much realism and genuineness exist in the therapeutic relationship, both overall and on a moment-to-moment basis. Valence is a bit more complicated. The real relationship varies on a positive versus negative dimension. Thus, one’s genuine and realistic feelings toward another may be negative. For example, the therapist may not like the patient with whom he or she is genuine and who he or she perceives/experiences realistically. In other words, one may have a high magnitude of realism and genuineness vis-à-vis another, but still feel negatively toward the other. Naturally, however, in what may be considered a good or strong real relationship (see below), the participants’ feelings for one another would be largely positive.
The additive combination of realism and genuineness, including their valence and magnitude, yields the strength of the real relationship. This construct of strength has been its main measure in research studies.
(p. 353) The Who and When
The real relationship is a bipersonal phenomenon, and as such it is contributed to by both the therapist and patient. The therapist contributes by direct self-disclosures of thoughts, feelings, and information, as well as indirectly. For example, the therapist reveals who he or she is not only by what he or she says to the patient but also through the therapist’s sense of humor, attire, office décor, facial expressions, body posture, and the like. These all enable the patient to build an image of the therapist as a person. The therapist also contributes to the strength of the real relationship by being genuine and nonphony with the patient, as well as by experiencing/perceiving the patient as he or she is rather than as a projection based on the therapist’s past and present unresolved conflicts.
From the patient’s side, the enactment of the patient role itself contributes to the formation and development of the real relationship. That is, the real relationship is built and strengthened by the patient’s getting in touch with inner experiences and through both verbal and nonverbal self-exploration and communication that share who he or she is. These communications are a part of the patient’s genuineness. In addition, the patient contributes to the real relationship through experiencing and perceiving the therapist in ways that befit the therapist. This is not to say that misperceptions, often referred to as transference, will not occur, for better or worse, as described in the following discussion.
Regarding when the real relationship manifests itself and unfolds, it has been theorized (Couch, 1999; Gelso, 2009, 2011, 2014; Greenson, 1967) to be present from the first moment of contact between therapist and patient. Each participant perceives and experiences the personhood of the other immediately, although to probably varying degrees. It has been theorized that as the therapeutic relationship deepens, the strength of the real relationship increases throughout the work (Gelso, 2014). At least two studies on the unfolding of the real relationship support this suggestion (Fuertes, et al., 2013; Gelso et al., 2012)
Probably all relational constructs are connected to one another. However, for the real relationship, three related constructs are the working alliance, patient transference, and attachment are especially relevant. In addition, both therapist self-disclosure (Chapter 11, this volume) and patient self-disclosure are related to the real relationship. The person-centered concept of therapist congruence (Chapter 10, this volume) is, in certain ways, synonymous with the therapist genuineness element of the real relationship.
The Real Relationship and the Working Alliance
The real relationship and the working alliance have been theorized to be highly interrelated, and they have indeed been referred to as sister concepts (Gelso, 2014). (p. 354) Beginning with the theoretical work of Greenson (1967), the real relationship was considered more foundational, as it was seen as existing in all relationships from the first moment of contact. The working alliance, on the other hand, was theorized as an artifact of treatment, existing solely for the purpose of getting the work accomplished. Presently, both are more likely to be viewed as occurring from the beginning of treatment, and as highly related, while still separate to the extent that each contributes independent variance in its relation to treatment process and outcome.
The real relationship may be thought of as the personal or person-to-person, nonwork connection between two or more persons, whereas the working alliance is the work connection. The working alliance is usually framed as the bond that exists between patient and therapist, the extent of their agreement on the goals of treatment, and their degree of agreement on the tasks that will accomplish those goals (Bordin, 1979). The bond part of the working alliance is particularly overlapping with the real relationship.
The bond that is part of the working alliance is a working bond, that is, the connection between the participants’ that directly reflects their therapeutic work (Gelso, 2011, 2014). For example, when the patient communicates his or her confidence in the therapist’s skills and competence, or feels a connection to the therapist as an effective professional, that may be considered part of the working alliance bond. Similarly, when the therapist expresses liking of the patient as a patient, this reflects more the working alliance than the real relationship. However, when either of the participants feels a connection to the other on a person-to-person basis, or feels liking or caring for the other as a person, this bond resides more in the realm of the real relationship.
Given the substantial overlap between these two constructs, it has been important to study whether they contribute independently to process and outcome. The empirical findings provide strong support for the real relationship and working alliance as both overlapping and independent elements. Several studies indicate that each contributes independently to the prediction of treatment progress and outcome (see review by Gelso, 2014). In the following text, we extend this research literature by, in addition to reporting the relation between the real relationship and treatment outcome, examining the size of the real relationship working alliance correlation. The purpose of that analysis is to see how much, and under what circumstances, these sister constructs overlap.
Transference and the Real Relationship
Transference entails the patient’s perceiving, experiencing, and reacting to the therapist in ways that do not befit the therapist, in ways that are not realistic. Given this fundamental tenet of transference, one can see where it might be viewed as opposite of the real relationship. And, indeed, the real relationship has been theorized by some to be the nontransference element of the therapeutic relationship.
However, most current conceptualizations consider the real relationship as more than merely the opposite of transference (Gelso, 2011, 2014; Wampold & Budge, 2012). First, the real relationship includes two elements, and realism or realistic perception is only one of these two (the other being genuineness). Second, any given session, (p. 355) part of a session, or any single patient expression contains a mixture of transference projections and realistic perception. When the patient is expressing transference-based reactions, he or she is also likely expressing reality-based reactions. A case example of this inevitable mixture follows (Gelso, 2009):
In my first session following a rather serious surgery, my patient, John, expressed concern by asking “How are you doing today, buddy?” I replied honestly, “I am doing well, thanks.” As I began to pursue how some of his concern was transferentially related to the material with which we had been dealing, John replied. “Well, that may be so, but I also was just concerned about you as a person.” As I pondered the expression of concern, it seemed clear to me that this single expression was both very rich with transference and very deeply reflective of real relationship. (p. 257)
The real relationship and patient transference, while not opposites, are at least partially and negatively overlapping (e.g., strong real relationship reflects less transference, and vice versa). The research evidence fits this expectation, indicating a small to moderate negative correlation between the two constructs (Gelso, 2014).
Attachment and the Real Relationship
Patients and therapists bring relationship patterns (relationship schema) into their clinical relationship that can enhance or detract from the strength of the real relationship. A key relationship pattern that has received some empirical attention are the attachment patterns of both parties (see Chapter 19 in Volume 2 for more details).
Two dimensions, attachment avoidance and anxiety, are typically used to describe peoples’ attachment behaviors and perceptions. Gelso and Hayes (1998) theorized about the role of therapists’ and patients’ attachment styles in the formation of the real relationship. Specifically, they contended that insecure attachment, low levels of attachment anxiety or avoidance, would be related to a stronger real relationship. In the following discussion, we summarize the research literature on the attachment–real relationship association by reporting the results of a meta-analysis examining the size of the real relationship and attachment correlation.
Only in recent years has empirical research been conducted on the real relationship. A major reason for this slow development is that no reliable measures had been created prior to the mid-1990s. The first measure of the real relationship was developed by Eugster and Wampold (1996), who created patient-rated, eight-item scales of both therapist and patient real relationship. These scales assessed patients’ feelings and reactions toward their therapists, as well as patient’s perceptions of their therapists’ feelings and reactions to them. Eugster and Wampold also developed parallel therapist-rated, 8t-item scales of therapist and patient real relationship. The items were identical to the patient-rated items, except they tapped the therapist’s views. The items mostly pertain (p. 356) to the genuineness and liking elements of the real relationship, with little attention to the realism element.
Two studies (Eugster & Wampold, 1996; Kelley et al., 2010) found rather modest internal reliability of this measure, with Cronbach’s alpha coefficients ranging from the .60s to the mid .70s. Still, the measure did correlate significantly with several other measures to which it theoretically should relate, providing support for its construct and convergent validity. Despite its psychometric limitations, the Eugster and Wampold measure represents a brief, convenient assessment of the real relationship in terms of genuineness and their liking of each other.
Virtually all of the quantitative research on the real relationship since that 1996 study have used two measures that were subsequently developed: the Real Relationship Inventory–Therapist Version (RRI-T; Gelso et al., 2005) and the Real Relationship Inventory–Client Version (RRI-C; Kelley et al., 2010). As implied by their names, the RRI-T taps the therapists’ perceptions of the real relationship, whereas the RRI-C assesses the client’s perceptions. Both measures consist of a total score and subscale scores for Realism and Genuineness. Within the subscales, items assess the sub-elements of magnitude and valence as previously defined. Furthermore, both the RRI-C and RRI-T examine the rater’s assessments of self, other, and the relationship. For example, in the RRI-C, the client rates real relationship items pertaining to himself or herself, the therapist, and their relationship. The valence and magnitude of realism and genuineness combine to yield scores for strength of the real relationship. Solid evidence of reliability and validity of these measures has accumulated since their initial appearance.
The RRI-T and RRI-C have been abbreviated to a 12-item measure, with six items in each of the two subscales, Realism and Genuineness (Hill et al., 2014). Items were derived by the authors selecting the 12 items that they believed best represented the two components within the longer measures. The abbreviated version correlated .91 and .94 with the longer RRI-C and RRI-T, respectively. The shortened RRI has been used in several studies, and support has been found for its reliability and validity (e.g., Kivlighan et al., 2017). This measure would appear to be a good choice when ease of usage is crucial and when the total score is what is being assessed, rather than subscale scores. Examples of items from both the full-length and abbreviated versions of the Client Form are (a) “I had a realistic understanding of my therapist as a person,” and (b) “My therapist did not see me as I really am” (negatively stated item). Examples of items from the Therapist Form are (a) “There is no genuinely positive connection between us” (negatively stated item), and (b) “My client’s feelings toward me seem to fit who I am as a person.”
What does the real relationship look like, or how does it manifest, in the psychotherapy hour? The real relationship shows itself at three levels, independent of the therapist’s theoretical orientation. At the first level, the real relationship exists in the background of patient–therapist transactions. “It shows itself through the participants’ ongoing sensing (p. 357) and understanding of one another and in their feelings toward each other. . . . These inner states simply exist as the therapist and patient explore the patient’s inner conflicts and outward behavior about matters other than the relationship” (Gelso, 2011, p. 88). However, the real relationship is expected to come into the foreground when there is a rupture in some aspect of the relationship, when disruptive transferences occur, and when the patient needs the support of a strong real relationship.
At the second level, the real relationship may manifest itself either subtly or explicitly in the behaviors of the participants that reflect realism and genuineness and the valence of these. At the third level, the real relationship can show itself in each and every communication between the participants. This follows Greenson’s (1967) suggestion that there is a real relationship aspect in all communications, regardless of how fantastical they may seem.
The psychotherapist can foster the real relationship by sharing aspects of himself or herself and his or her feelings toward the patient. Examples of sharing (from Kasper et al., 2008) occurred between a 51-year-old White male therapist and a 24-year-old White female patient involved in therapy with a 12-session limit. Throughout treatment the therapist expressed a range of feelings in the moment, for example, caring, sadness, disappointment, and connection. One example that highlights the real relationship has to do with the therapist’s feelings about the patient’s reactions to the forthcoming termination:
Therapist: Last week when we were talking about when we’re gonna end and how long we’re gonna go. . . . I was sort of wondering what’s going in that it doesn’t seem like it matters to you one way or the other how long we meet.
Patient: No; actually, the truth of the matter is, like I mean I think I mentioned to you that I would like to go on more than 12.
The patient then goes on to say that she thinks it would be selfish to ask for more sessions. The therapist follows with:
Therapist: So you didn’t in some ways wanna hurt me or upset me. For me it hurt, that it felt like it didn’t matter [to you] how long [we met].
In subsequent research interviews, the patient expressed to the researchers that because of the therapist’s self-disclosure, she felt this was a powerful session resulting in greater closeness and understanding of the impact she has on others.
Another example of deep sharing that is reflective of the real relationship is taken from the 17th and terminating session (from Hill et al., 2008). The therapist is a 55-year-old White heterosexual man and the patient is a 29-year-old African American, lesbian female with long-standing anxiety and depression. This vignette demonstrates how the therapist’s genuine sharing facilitates the patient’s genuineness:
Therapist: Jo, I respect you so much.
Patient: Do you?
Therapist: I respect you so much. The way you go at these huge issues and face them with such courage . . . the work you’ve done with me since November has been so hard and so challenging and you have been so strong and capable and successful. I respect your integrity, I respect your courage.
(p. 358) Patient: Thank you, but I’m glad I met you because there’s no telling if I met with someone else. Not to say that it would have been. . . . It’s probably more of a feeling you know, with you than let’s say somebody else who is just kind of like “so how do you feel about that?” . . . You really talk about issues and . . . it matters what we talk about in here. I always reflect back and say “Oh that makes sense” or then I’ll jot it down. . . .
Therapist: You surprised me right from the get-go, Jo, you just got in the driver’s seat and you put your foot on the gas pedal and you went to work. You initiated and you led me and us to such profound conversations at times. It’s been a deep sharing.
Patient: Oh, yeah, it really has.
The real relationship may also be fostered by the therapist working with the patient to see the therapist as he or she is. Yalom (2002), for example, describes a case in which he and his patient shared impressionistic notes of each session. The patient idealized Yalom, and Yalom wanted to diminish this idealization so that the patient could seem him realistically. Thus, in Yalom’s notes of the sessions that he shared with the patient, he deliberately tried to reveal his most human feelings, including his frustrations, irritations, insomnia, and vanity.
In a more moderate manner, Dr. Patrice Duquette (personal communication, May 2009) provided an example of using the real relationship to help her patient who was suffering from intense anxiety and withdrawal. Duquette is a highly experienced White psychodynamically oriented psychiatrist, and the patient is a 36-year-old White, heterosexual female, who is married and has one child. In the following vignette, Duquette’s observations are in parentheses following the patient’s verbalizations:
Patient: I can hear myself being tight. I don’t want to really feel, don’t want to have those feelings. But I don’t want to be like this either. (Her mouth is set tight, her forehead is raised and eyes tightened, her voice is very tight and has a cry it in.)
Therapist: Can you feel your throat at all?
Patient: Not really. Can’t feel it separate from all the other tightness. Not quite. Don’t want to almost. Feels like I can’t feel or think, like my brain has just stopped.
Therapist: Try a bit. Can you feel anything? Can you direct your energy, to feel where you are? Can you look at me?
Patient: A little bit. (Tears well up obviously in her eyes, her mouth twists more, her eyes go to an almost vacant look, with her eye contact less intensely focused.) Now I just . . . just . . . it just feels like there is a big gaping hole. (Silence, and she continues looking, but is appearing more frightened by the second.)
Therapist: Stay here. Can you look at me? At me . . . here?
Patient: A little . . . (More tears, eyes are still looking vague and fading.)
Therapist: What do you see? Here? Are you along?
Patient: I see you.
Therapist: (Nods and gestures, as if to say, “More.”)
Patient: I can see you (mouth still twisting, patient is limiting her verbal output.)
Therapist: What do you see in me? Can you see my eyes? What do you see?
(p. 359) Patient: That you are present.
Therapist: Who is? Who am I?
Patient: Double D. (Patient chuckles, as if it were a private joke. Therapist recalls that this is how the patient writes her name in her appointment book and smiles slightly in recognition.) Dr. Duquette, that’s who you are. (She says this in a firmer voice, with a moment of eye contact. Such moments had come up before in group situations, when I had directed her and others to use my full name by way of fuller recognition of me and our relationships in a given moment.)
Therapist: And so?
Patient: That you see me, care about me, are listening. And that as you see me you are like 99.9% accurate about what you see. (Her voice is settling, she speaks more spontaneously, her eye contact is more directed.)
Research on the real relationship is of recent vintage (although research on one of its elements, therapist genuineness, has existed for many years). The aforementioned Eugster and Wampold (1996) investigation was the first empirical study on the topic. The next study was published nearly a decade later (Gelso et al., 2005) and represented the first instrument development piece, a therapist-completed measure of the real relationship. This measure, along with the subsequent client-completed measure (Kelley et al., 2010), provided psychotherapy researchers with reliable and validated instruments that facilitated empirical efforts. In this section, we summarize the Eugster and Wampold study, the subsequent Gelso et al. investigation, and then a more recent investigation by Kivlighan and colleagues (2017), which represents a multilevel modeling effort separating effects into therapist, client, and session effects. Finally, we mention qualitative work and underscore its near absence.
The First Study
Eugster and Wampold’s (1996) study is a landmark because it represents the first empirical study on the real relationship. These researchers surveyed 114 highly experienced therapists and 119 of their patients, using an 8-item measure of the real relationship. Therapists and patients made ratings on a 6-point scale of the quality of the therapist-offered real relationship and the patient-offered real relationship. Four items of the 8-item measure elicited ratings of the self, and four items elicited ratings of the other. Thus, four scores resulted: therapist rating of therapist-offered real relationship (4 items) and patient-offered real relationship (4 items) and patient ratings of patient- and therapist-offered real relationship (4 items each). The researchers developed a battery of nine measures (real relationship was one of these) thought to be relevant to treatment outcome. The criterion was the evaluation of sessions.
Although a range of complex statistical analyses was employed, the simple bivariate correlations with outcome may prove the most revealing. Both therapist and patient ratings of therapist-offered and patient-offered real relationship correlated moderately (p. 360) but uniformly positively with session evaluation (rs ranging from .28 to .64). The correlations with session evaluation were especially notable for patient’s ratings of their therapist-offered real relationship (r = .60) and their own enacted real relationship (r = .64). It was also found that patients’ and therapists’ ratings of the real relationship were significantly associated with patient and therapist ratings of involvement in treatment, comfort, patient progress, therapist expertness, and therapist interpersonal style (e.g., friendliness, empathy, warmth). In essence, this first study demonstrated that the real relationship was a promising variable in terms of its potential influence on treatment outcome, as well as its connection to other important process variables.
The RRI-T (Gelso et al., 2005) was the first psychometrically sound measure of the real relationship. The results of the instrument-development study for the RRI-T have been described in the Measures section of this chapter. In sum, the RRI-T consisted of 24 items, 12 assessing realism and 12 assessing genuineness. Items were written to tap magnitude and valence within each of the two subscales and to elicit ratings of self (therapist), other (client), and the relationship within realism, genuineness, valence, and magnitude.
In the development of the RRI-T, 130 items were originally written and then, through a series of steps, including the sampling of 80 psychologists, reduced to 24 items. Then 130 psychologists and psychology trainees completed the reduced measure, along with several other measures aimed at assessing the validity of the RRI-T. It was found that RRI-T was significantly related to several measures to which it theoretically should relate (see Measures section), supporting the construct and convergent validity of this instrument. Since the appearance of the RRI-T, all studies of therapist-rated real relationship have employed this measure.
One of the major steps forward in psychotherapy research has been the development of advanced statistical analyses that can clarify treatment effects. One such technique is often referred to as multilevel modeling, and this technique seeks to determine the independent contribution of different elements (e.g., patient, therapist, treatment, duration) to treatment process or outcome.
In a recent study, multilevel polynomial regression and response surface analysis were used to examine how the real relationship and the working alliance were related to session quality (Kivlighan et al., 2017). Real relationship and working alliance ratings of 2,517 sessions of 144 clients and their 23 therapists were partitioned into therapist-level, client-level, and session-level components. This important study determined how therapy proceeds when the sister concepts of the working alliance and the real relationship are similar or dissimilar from one another in the treatment hour. Multilevel analysis provides a potent method of analyzing these relationships at differing levels.
(p. 361) For both therapists and clients at most levels of analysis (therapist, client, and session), client-rated session quality was highest when both the real relationship and working alliance were strong and lowest when both relational factors were weak. However, for some clients (within a therapist’s caseload), session quality across sessions was greatest when working alliance was stronger than the real relationship, whereas for other clients, session quality was best when the real relationship was stronger than the alliance. Within clients, some sessions had the highest quality when working alliance was strongest, whereas other sessions were rated most favorably when the real relationship was strongest. Thus, at both the client and session levels, the most important relational ingredient varied. This finding suggests that therapists need to pay close attention to which relational ingredient is most needed by each client and within each session.
One limitation of research on the real relationship is the absence of qualitative studies. This is surprising because the construct seems so fitting for qualitative analyses.
There have been some qualitative studies in which implications were drawn for the real relationship. For example, Knox and associates (1997) conducted a qualitative study on 12 psychotherapy patients to determine their views of what were helpful therapist self- disclosures. They found that such disclosures allowed patients to see their therapists as more real and human and influenced clients to be more open and honest. These findings suggest that the genuineness element of the real relationship is strengthened by therapists sharing more about themselves. Similarly, Curtis and colleagues (2004) discovered that psychoanalysts viewed their own analysts’ warmth, emotional availability, genuineness, and willingness to explore the patient’s feelings about him/her (the analyst) as most helpful. This supports the genuineness aspect of the real relationship. Qualitative work is needed on the real relationship to supplement the quantitative research.
Because this was the first meta-analysis of the real relationship literature, we included all studies (published and unpublished), regardless of publication date, that reported data allowing calculation of the correlation between the strength of the real relationship and: (a) treatment outcome (pretest–posttest change, treatment progress, and session quality or session outcome), (b) working alliance, (c) transference, and (d) attachment. Pretest–posttest change was defined as studies that used a psychometric instrument (e.g., Outcome Questionnaire-45) that patients completed prior to commencing treatment and at the completion of treatment. For treatment progress, (p. 362) patients and/or therapists assessed the progress to date in their treatment. These progress reports were often, but not always, completed at the conclusion of the patient’s treatment. For session quality or outcome, patients reported on the quality or outcome of the session in which the real relationship was assessed. Studies were excluded if they did not have the information necessary to calculate a correlation between real relationship and any of the predictor variables or if the data set was not independent of other studies included in the meta-analysis.
We searched the database PsycInfo for published studies on the real relationship and the pre-identified process and outcome variables. We also contacted researchers known to conduct research on the real relationship for unpublished studies and student theses or dissertations. All except one study were published in English. The exception was a South Korean dissertation in which the key information was translated by the doctoral student.
We used the search term “real relationship” paired with the terms, “therapy/counseling outcome,” “session quality/outcome,” “treatment progress,” “working alliance,” “transference,” and “attachment.”
Titles and/or abstracts of potential studies were independently coded by the third author and two advanced graduate students in counseling psychology. Disagreements were discussed among the coders until a consensus was reached.
Data Collection Process
We developed spreadsheets for coding both study-level and effect size-level data. The second author and an advanced graduate student in counseling psychology independently extracted the targeted data. Disagreements were again discussed and resolved by the two judges. When sufficient data for computing standardized effect sizes were unavailable, study authors were contacted.
Along with data necessary for computing standardized effect sizes (Pearson’s r), the following data were extracted: (a) type of outcome assessed (see Eligibility Criteria section), (b) sample size, and (c) who (client or therapist) made the ratings.
(p. 363) Summary Measures
The Pearson’s correlation coefficient (r) was the effect size measure used in this research. In addition, we report the d for each relationship examined. Methods described by Cooper et al. (2009) were used to compute this effect size and its variance. The Comprehensive Meta-Analysis (v3) statistical software was used to conduct the analyses.
Synthesis of Results
When studies contained multiple effect sizes, we aggregated data within studies and then between studies, based on the specific comparisons from our different analyses. We computed Pearson’s r and 95% confidence intervals as summary statistics. The heterogeneity among effect sizes in an analysis was assessed using the Q-statistic (assessing whether between-study heterogeneity exceeds that expected by chance alone). All analyses used random effects models.
For the correlation between real relationship and outcome, type of outcome (pretest–posttest change, treatment progress, or session quality) was assessed as a between-study moderator of the correlation between real relationship and outcome. For the real relationship outcome analysis, within-study moderators included (a) source of the real relationship rating (client or therapist), (b) source of the outcome rating (client or therapist), and (c) rater match (same rater for both real relationship and outcome; e.g., client-rated real relationship and outcome) or different rater for real relationship and outcome (e.g., therapist-rated real relationship and client-rated outcome).
For the correlations between (a) real relationship and working alliance, (b) real relationship and transference, and (c) real relationship and attachment, within-study moderators included (a) source of the real relationship rating (client or therapist), (b) source of the outcome rating (client or therapist), and (c) rater match (same rater for both real relationship and outcome; e.g., client-rated real relationship and outcome) or different rater for real relationship and outcome (e.g., therapist-rated real relationship and client-rated outcome). For the real relationship and transference analysis, type of transference (positive or negative) was an additional within-study moderator. For the real relationship and attachment analysis, attachment dimension (anxiety or avoidance) was an additional within-study moderator.
Real Relationship and Outcome Results
Five studies reported the correlation between real relationship and psychotherapy progress, five studies reported the correlation between real relationship and pretest–posttest (p. 364) outcome, and six studies reported the correlation between real relationship and session quality. The omnibus effect size was significant (r = .38, 95% CI [.30, .44], p < .001, d = 0.80, N = 1,502 participants). There was significant heterogeneity across the studies (Q = 31.90, p = .007). The fail-safe N was 759, and the trim-and-fill analysis did not suggest evidence of publication bias. The funnel plot for this analysis is displayed in Figure 10.1. This result shows a moderate association between real relationship and outcome. The real relationship–outcome relation is of a larger magnitude (moderate) than the working alliance–outcome relation (small, r = .28), found in a previous meta-analysis (Horvath et al., 2011).
A moderator test was conducted to determine whether the type of outcome assessed was related to the magnitude of the correlation between real relationship and psychotherapy outcome. For this analysis, we created two dummy variables: (a) whether or not the outcome assessed represented progress and (b) whether or not the outcome assessed represented session quality. Therefore, whether or not the outcome assessed was a pretest–posttest change was the reference group. Neither the progress dummy variable (coefficient = 11.69, z-value = .90, p = .371) nor the session quality dummy variable (coefficient = .08, z-value = .62, p = .538) was significant. Therefore, the type of outcome assessed was not related to the strength of the real relationship outcome correlation.
(p. 365) Within-study moderator tests revealed that (a) who assessed the real relationship (client or therapist) was unrelated to the strength of the real relationship outcome correlation (coefficient = –.09, z-value = –.51, p = .614), (b) who assessed outcome (client or therapist) was unrelated to the strength of the real relationship outcome correlation (coefficient = .20, z-value = 1.33, p = .183), and (c) whether the assessor of real relationship and outcome was the same (e.g., client assessed both real relationship and outcome) or different (e.g., therapist assessed real relationship and client assessed outcome) was unrelated to the strength of the real relationship–outcome correlation (coefficient = –0.18, z-value = –1.19, p = .235). Therefore, rater source did not affect the size of the real relationship–outcome correlation. The role relationship moderator is important because it suggests that single-rater bias does not affect the size of the real relationship–outcome correlation.
Relation with Other Process Variables
As suggested earlier, the real relationship probably interconnects with other key elements of the therapeutic relationship. Below we summarize the meta-analytic results on the association between the real relationship and the working alliance, transference, and attachment style.
Real Relationship and Working Alliance
Nine studies reported the correlation between real relationship and working alliance. The omnibus effect size was significant (r = .58, 95% CI [.51, .64], p < .001, d = 1.42, N = 1,070 participants). There was significant heterogeneity across the studies (Q = 19.67, p = .012; Figure 10.2). The fail-safe N was 4,016, and the trim-and-fill analysis did not suggest evidence of publication bias. This result supports Gelso’s characterization of the real relationship and the working alliance as “sister” constructs; medium to large correlation but not identical constructs. The funnel plot for this analysis is displayed in Figure 10.2.
Within-study moderator tests revealed that: (a) who assessed the real relationship (client or therapist) was unrelated to the strength of the correlation (coefficient = –.17, z-value = –1.68, p = .092), (b) who assessed WA (client or therapist) was unrelated to the strength of the real relationship–working alliance correlation (coefficient = .01, z-value = .14, p = .889) but (c) whether the assessor was the same (e.g., client assessed both WA and RR) or different person (e.g., therapist assessed WA and client assessed RR) was related to the strength of the real relationship–working alliance (coefficient = –.52, z-value = –5.02, p < .001). The assessor match explained 54% of the variance in the size of the real relationship–working alliance correlation. When the person was in the same role, the real relationship–working alliance correlation was .83, but when different people assessed real relationship and the working alliance, the real relationship–working alliance correlation was .31. This correlations for the role analysis suggest (p. 366) that single-rater bias may, under certain circumstances, inflate the size of the real relationship–working alliance correlation.
Real Relationship and Transference
Four studies reported the correlation between real relationship and transference. The omnibus effect size was significant (r = –.17, 95% CI [–.25, –.10], p < .001, d = –.35, N = 619 participants). There was not significant heterogeneity across the studies (Q = 0.07, p = .995; Figure 10.3). The fail-safe N was 129, and the trim-and-fill analysis did not suggest evidence of publication bias. The real relationship is not the opposite of transference but does evidence a small to moderate negative relationship with transference. The funnel plot for this analysis is displayed in Figure 10.3. Due to the small number of studies, we do not report on any moderator analyses.
Summary and Conclusions
The meta-analytic results show that there is a moderate correlation between real relationship and treatment outcome. This relationship is not moderated by how outcome is assessed or by who does the assessment of the real relationship and outcome; in particular, there is no evidence of single-rater bias inflating the size of the real relationship–outcome correlation. These results confirm the importance of the real relationship and provide support for Gelso’s (e.g., 2014) theoretical writings. As previously noted, the real relationship–outcome meta-analytic correlation is larger than the working alliance–outcome correlation. However this size difference needs to be (p. 367) placed in context. There have been far more studies reporting the working alliance correlation than studies reporting the real relationship–outcome correlation. Therefore, the working alliance correlation in meta-analyses should be more stable than the real relationship–outcome correlation found in this study.
As previously noted, the real relationship is a component of Gelso’s (2014) tripartite model of the therapeutic relationship. As hypothesized in that model, the real relationship and the working alliance should have a moderate to large correlation, and our meta-analytic results confirm this hypothesis. There is evidence that mono-rater bias may inflate the size of the real relationship–working alliance correlation. Therefore, researchers should be careful when interpreting the size of the relationship–working alliance correlation when both ratings are coming from either the client or therapist.
Gelso (2014) argued that the real relationship should not be considered the opposite of transference because the real relationship is not entirely composed of the opposite of distortion (reality). Therefore, he argued that the correlation between the real relationship and transference should be small to moderate and negative. The meta-analytic results support this expectation. Taken together, the results of these three meta-analyses provide support for the propositions that Gelso laid out in his tripartite model of the therapeutic relationship.
Patient and Therapist Contributions
Patient attachment theoretically contributes to the strength of the real relationship. Alas, only four studies reported the correlation between real relationship and attachment. For these studies, a positive effect size indicated that a secure attachment (low attachment anxiety and avoidance) was related to a stronger real relationship. The omnibus effect size was significant (r = .17, 95% CI [.06, .28], p < .0003, d = 0.35, (p. 368) N = 303 participants). Therefore, a secure attachment relationship was associated with a stronger real relationship. The funnel plot for this analysis is displayed in Figure 10.4.
Both the patient and the therapist bring differing capacity to form a therapeutic relationship in the psychotherapy encounter. These different capacities are embedded in patients’ and therapists’ relationship histories. As hypothesized, the meta-analytic results of the three available studies show that the anxious and avoidant attachment orientations of both patients and therapists are negatively related to the strength of the real relationship.
Do differences between therapists or differences between clients in their abilities to form and grow real relationships best predict treatment success? There are two parts to this question. First, are some therapists (clients) better than other therapists (clients) in forming real relationships with their clients (therapists) and are these between-therapist (between-client) differences in real relationships related to psychotherapy outcome? Second, if between-therapist and between-client differences are related to outcome, which type of difference is a more important predictor of outcome? Three studies (Kivlighan et al., 2014; Li et al., 2016) examine this important question.
Kivlighan et al. (2015) found that therapists who, on the whole, were rated as having stronger third session real relationships had clients who reported more treatment progress. However, for clients within any single therapist’s case load, there was no association between the client’s real relationship with their therapist and treatment progress. Although complicated, this difference between therapists as a whole and individual therapists as rated by clients on their caseloads implies that the therapist’s contribution to the real relationship and its connection to session progress is stronger than is the patient’s contribution.
Kivlighan et al. (2014) reported a similar finding. They found that therapists who, across all of their clients, were rated as having increasing real relationship strength as treatment progressed, had clients who reported a greater decrease in symptoms. Within any given therapist’s case load, however, changes (increases or decreases) in (p. 369) client-rated real relationship with their therapist were unrelated to change in client symptoms across treatment. Again, this difference between therapists as a whole and any given therapist’s caseload suggests that therapists are more significant contributors than are patients to the role of the real relationship in symptom improvement as treatment progresses.
Another study extended these two investigations by examining between-therapist differences in the dyadic real relationship (Li et al., 2016). Theoretically, the real relationship is a two-person construct with both the therapist and client contributing to the dyad’s real relationship. However, most of the research literature treats client and therapist real relationship perceptions as discrete entities.
Researchers (Li et al., 2016) used the common fate model (CFM; Ledermann & Kenny, 2012) to model the dyadic real relationship and to examine how this dyadic real relationship was related to session evaluation. Psychotherapists who, across all of their clients and sessions, had a stronger dyadic real relationships had clients who reported better session evaluations; whereas therapists who across all of their clients and sessions had weaker dyadic real relationships also had clients who reported worse session evaluations. When clients within a therapist’s case load reported a stronger dyadic real relationship with their therapist, there was no statistical association with session evaluation.
Taken together, the three statistically sophisticated studies show that between-therapist differences in early real relationship, between-therapist differences in the growth of the real relationship, and between-therapist differences in the dyadic real relationship are better predictors of treatment outcome than between-client differences in the real relationship. That is, in plain English, the individual therapist makes more difference than the individual client when it comes to the real relationship. These studies suggest that researchers should continue to study therapist differences in the real relationship, as they may prove key to successful therapeutic change. For example, what characteristics do some therapists possess that allow them to form and grow stronger real relationships with their clients? Or, what do some therapists do or not do with their clients to form and grow stronger real relationships?
Evidence of Causality
It is critical to understand the causal relationships between the real relationship and treatment outcome, but unfortunately there are no purely experimental studies that allow us to draw strong causal inferences. It is difficult, if not impossible, to conduct true experiments in field studies of organismic variables, such as the real relationship. This is so because in the real world of psychotherapy, one cannot ethically assign clients randomly to differing levels of real relationship offered by the same or equivalent therapists. Experiments in the form of laboratory analogues of psychotherapy may allow for causal inferences, and such studies would be a useful addition to the field studies that have been done to date. While laboratory experiments are in themselves not directly relevant to clinical practice, when combined with field studies, this sort of methodological triangulation may allow for meaningful causal inferences.
(p. 370) At the same time, longitudinal field studies have been done, and these do allow for stronger causal statements than cross-sectional correlational investigations. For example, three growth curve studies have provided strong suggestions of causality. All three studies collected client- and therapist-rated real relationship ratings longitudinally and examined how linear change in real relationship was related to treatment outcome.
In one of the studies (Gelso et al., 2012), it was found that linear increases in therapist-rated and client-rated real relationship predicted therapists’, but not clients’, ratings of treatment progress. Another of these studies (Lee, 2017) found that linear increases in client-rated, but not therapist-rated, real relationship were related to decrease in distress. Finally, the third study (Kivlighan et al., 2014) revealed that when clients of a given therapist reported increasingly strong real relationships, they also experienced decreasing emotional distress.
Taken together, these studies show that an increasing real relationship is related to better treatment outcome. However, there are inconsistencies in findings, depending on who rates the real relationship and who rates outcome. Still, these growth curve studies provide some tentative evidence that change in real relationship is related to better outcomes. It is important, however, for more studies to examine the causal ordering of the real relationship and outcome using other methodological approaches.
Limitations of the Research
The main limitations of the research conducted to date on the real relationship pertain directly to the few studies that have been conducted. Thus, although the findings of this meta-analysis are highly promising regarding the relation of the strength of the real relationship to differing outcomes (session outcome, treatment progress, and treatment outcome), the number of studies is small, especially for the three outcome types taken separately. More studies are needed associating the real relationship to treatment outcome, both immediately following termination and at follow-up. No studies of which we are aware have been done on the latter.
Similarly, no studies have examined the real relationship in psychotherapies of varying theoretical orientations. Is the real relationship different with therapists who practice psychodynamic, cognitive behavioral therapy, experiential, integrative, and other therapies? Does the relation of the strength of the real relationship to outcome differ for dissonant treatments? What is the relation of real relationship and outcome for brief versus longer-term therapies? Nearly all of what we now know is based on brief treatments, and work on longer treatments is sorely needed.
Another limitation of the research is that nearly all of the studies on real relationship have been conducted by a small number of researchers from a small number of laboratories. It is well documented throughout the history of psychotherapy research that findings emerging from laboratories that espouse the theoretical construct being studied have somewhat different, and usually more favorable, findings than those in which the investigators are not proponents. In the case of the real relationship, nearly all studies have been conducted by those who have theorized about the importance (p. 371) of the real relationship and their colleagues and students. Findings will probably be modified as a larger number of investigators, including those who are not proponents, study this construct and its impact.
We have already lamented the paucity of qualitative research on the real relationship. It would be helpful to study from a qualitative perspective what patients and therapists believe make for a strong real relationship and what they believe weakens it in particular treatments. It would be especially useful to ground these opinions in specific experiences, for example, of patients commenting on their own therapy experiences or of therapists commenting on specific cases. Such work could lead to further refinements in practice, training, and theory.
Because empirical research on the real relationship is still in its early stages, there are many questions that await empirical scrutiny. Among the major ones are:
a. What therapist–patient interactions occur in sessions in which there are stronger and/or more salient real relationships?
b. Which therapist and patient factors facilitate stronger real relationships? Although some such factors have been studied, many therapist and patient factors either require further study or have not been studied at all.
c. How do cultural factors such as race, ethnicity, and sexual orientation relate to the strength and salience of the real relationship?
d. Similarly, what factors, for example, multicultural competence and orientation, may moderate or mediate the relation of such cultural factors and the real relationship?
e. How does the real relationship manifest itself in treatment in other countries? To date, the only other countries in which the real relationship has been studied are Italy (e.g., LoCoco et al., 2011) and South Korea (Eun Ju, 2015), and the findings have been consistent with research in the US.
f. How does the real relationship vary in strength and impact for patients with differing personalities and disorders?
These are but a few of the many questions that may be fruitfully examined in future studies.
Little is known at this point about how diversity, such as race, ethnicity, and sexual orientation, influence the real relationship, or how the strength and salience of the real relationship may moderate or mediate the relation of diversity to treatment outcomes. To date, three studies have examined such factors, and these present intriguing findings.
One such study examined how therapists’ multicultural orientation (MCO; values about the salience of cultural factors in the lives of therapists and clients) was associated with the strength of the real relationship, working alliance, and treatment outcome (Owen et al., 2011). One hundred seventy-six clients rated the MCO of their 33 psychotherapists and also rated their working alliances, real relationships, and (p. 372) treatment outcomes. Most central to the present chapter, clients’ ratings of the real relationships with their therapists were positively correlated with treatment outcomes for both White (n = 95) and racial-ethnic minority (n = 81) clients. Also, client ratings of their therapists’ MCO indicated that multicultural orientations of therapists were positively associated with ratings of the strength of the real relationship. The greater therapists’ MCO in the eyes of their clients, the stronger the real relationship between therapist and client.
In another such study involving 144 clients and their 19 therapists, clients rated some therapists as having stronger real relationships and working alliances with their REM clients, whereas other therapists were seen to have stronger working alliances and real relationships with their White clients (Morales et al., 2018). These differences were independent of the race of the therapists. Nuanced findings such as these may prove valuable as they take us closer to understanding which therapists work best with which clients regarding race and ethnicity.
A third diversity-related study focused on the real relationship with lesbian (n = 76) and gay (n = 40) clients (Kelley, 2015). The real relationship accounted for significant variance in clients’ positive feelings about their therapists above and beyond months in therapy, therapists’ helpful and unhelpful therapy practices, and the working alliance. This study offers evidence of the importance of the real relationship in the eyes of lesbian and gay clients.
No studies to date have examined how therapist and/or patient gender bear upon the role of the real relationship. Although virtually all studies on the real relationship break down samples by gender, none of these investigations has actually studied gender and the real relationship. This absence is especially striking in light of the long-standing, replicated findings about differences in men and women in various relational characteristics.
The evidence presented in this chapter suggests that a strong real relationship is an important ingredient of successful psychotherapy. From this, we encourage psychotherapy training programs and psychotherapy supervisors to focus on the real relationship and its strengthening and weakening over time.
Teaching the real relationship to trainees is a complicated endeavor because the concept is often experienced as highly nebulous, nonverbal, and difficult to grasp in concrete terms. In comparison, for example, the working alliance, is more easily explained and tangible to trainees, as it is often defined as agreement on the goals and tasks of therapy and the working relationship. The real relationship (although it has been operationalized for research purposes) is something that must be experienced to be truly known.
We suggest that training in the real relationship consists of both a didactic and experiential component. For example, in practicum and internship classes, readings on the real relationship can be accompanied by students completing a case studies on a patient with whom they have experienced a strong real relationship and one with (p. 373) whom they have experienced a weak one. On a more basic level, it is often helpful to have students think of a relationship in their personal lives in which they shared a high degree of genuineness and have felt truly known, and how this experience, in turn, affected their sense of self and the overall relationship.
Another factor that complicates training students on the real relationship is that it is often difficult for students to separate transference/countertransference reactions from the real relationship and, at the same time, understand that both these phenomena can be true at the same time. That is, client and therapist may be smack in the middle of a transference–countertransference reenactment and at the same time have real and genuine feelings for one another’s authentic self. We have found it helpful for supervisors and instructors to share with students their own experiences with patients when it was, for example, difficult for them to separate a countertransference from real relationship reaction to a patient, or when they had a strong feeling of both at the same time. Along these lines, we have seen many trainees unintentionally dismiss patients’ feelings toward them as purely transferential, when doing so (although it was partly true) minimized the client’s real experience of connection to the therapist. Thus, we believe it a critical part of training to underscore to trainees that the therapeutic relationship is complex and multifaceted, and to use real life case examples as demonstrations.
Lastly, we believe it is essential to teach trainees that while self-disclosure is one avenue toward a strong real relationship, it is not the only avenue, and sometimes taking this road may get you lost. On the one hand, depending on the specific therapeutic situation and the therapist’s theoretical orientation, self-disclosure, or therapeutic immediacy, may of course strengthen the real relationship, as the therapist makes himself or herself vulnerable and human and models genuineness and authenticity for the patient. On the other hand, we have seen many trainees rush to self-disclosure out of their own performance anxiety, need to be liked, or confusion over their role and boundaries in the therapeutic relationship. Here, self-disclosure may actually hinder the development of a strong real relationship.
The real relationship may be facilitated by what the therapist says, but it is also a product of who the therapist is. It is vital to teach trainees that much of what we communicate to our patients (and vice versa) about who we are is in fact not verbal at all. A useful skill for trainees to develop is to be able to accurately assess the strength of the real relationship. In this regard, the real relationship inventories can be used as a training tool. For example, trainees can watch a therapy session putting themselves in the place of the client or the therapist and rate the session using the real relationship inventory. Trainees can compare ratings and discuss the differences among their ratings. It may be particularly helpful if sessions ratings of the real relationship by the patient and therapist are available for comparison.
How can practitioners foster and develop a strong real relationship? From one vantage point, the real relationship simply is. The participants in the therapy relationship are whom they are, and nothing can be manipulated to strengthen the real relationship. (p. 374) The empirical evidence suggests that the real relationship does change across sessions, usually strengthening (Gelso et al, 2012; Kivlighan et al., 2016), but this change may represent a natural evolution that comes about through the participants getting to know each other more deeply. This strengthening may be a result of the similarities between therapist and patient in terms of human qualities, for example, sense of humor, basic interests and values, and sense of attunement to others. When therapist and patient are “part of the same tribe” (Gelso & Silberberg, 2016) in these ways, they more likely take to each other as persons, thus paving the way for a strong real relationship. From a second vantage point, although a portion of the real relationship may simply exist and naturally unfold, there are therapist actions that facilitate a strong real relationship. These include:
◆ Seek to grasp empathically the patient and he or sher inner experience. Evidence indicates that therapist empathy is significantly related to the strength of the real relationship (Fuertes et al., 2007). The therapist’s successful understanding of the experience of the patient facilitates the realism element of the real relationship on the therapist’s side, and because feeling seen and understood accurately can be so intimate, it fosters the patient’s sense of personal connection to his or her therapist. In addition, it is likely that therapist empathy begets the strengthening of empathy in the patient, and patient empathy fosters seeing the therapist as he or she truly is, thus deepening the realism aspect of the real relationship on the patient’s side.
◆ Manage countertransference. Fundamental elements of the therapist’s effectiveness in managing how his or her own unresolved conflicts bleed into the treatment (managing countertransference) include self-understanding, managing one’s own anxiety, and grasping the boundaries between oneself and the patient (see Chapter 15, this volume, on managing countertransference). All of these behaviors serve to deepen seeing the patient as he or she is (rather than as projections of the therapist’s issues) and being genuine with the patient. They also, in turn, foster the patient seeing the therapist as he or she is and being genuine with the therapist.
◆ Share reactions with the patient. Although therapist self-disclosure is certainly an imperfect indication of genuineness, it does relate modestly to the strength of the real relationship (Ain, 2008, 2011). Well-timed disclosures (including disclosures of feelings within the therapeutic relationship and about the patient) highly relevant to the patient’s needs (rather than the therapist’s needs) foster therapist genuineness in the patient’s eyes.
◆ Explain when not sharing. Despite the connection of therapist self-disclosure to the real relationship and genuineness, we know from clinical experience and research evidence that therapists can be genuine while being relatively nondisclosing. One of the factors that fosters the patient’s perceptions of therapist genuineness when not disclosing involves telling the patient just why the therapist is not disclosing. When the therapist clarifies why he or she is not disclosing, the therapist is, in fact, disclosing at a different level, what might be seen as a meta-level.
◆ Be consistent and constant. At the most fundamental level, the patient’s sense that he or she could count on his or her clinician to simply be there, and be there on time, (p. 375) fosters a sense that the therapist can be personally trusted and that the therapist is genuinely interested in the patient as a person as well as a patient. This seems particularly crucial for patients who are highly vulnerable. In addition, consistency is a key factor in helping the patient trust the therapist as a person, and this includes consistency between the therapist’s verbal and nonverbal behavior, as well as between each of these over time. This consistency also provides credibility to the real relationship that the therapist is offering to the patient. It fosters the patient’s sense that the therapist can be counted on as a person and is congruent as a person.
References marked with an asterisk indicate studies included in the meta-analyses.
Ain, S. C. (2008). Chipping away at the blank screen: Therapist self-disclosure and the real relationship (Unpublished doctoral dissertation). University of Maryland, College Park.Find this resource:
*Ain, S. C. (2011). The real relationship, therapist self-disclosure, and treatment progress: A study of psychotherapy dyads (Unpublished doctoral dissertation). University of Maryland, College Park.Find this resource:
*Baumgartner, S. (1995). Systematic effects of participant role on evaluation of the psychotherapy session. Madison: University of Wisconsin.Find this resource:
*Bhatia, A., & Gelso, C. J. (2017). The termination phase: Therapists’ perspective on the therapeutic relationship and outcome. Psychotherapy, 54(1), 76–87. https://www.doi.org/10.1037/pst0000100Find this resource:
*Bhatia, A., & Gelso, C. J. (2018). Therapists’ perspective on the therapeutic relationship: Examining a tripartite model. Counselling Psychology Quarterly, 31(3), 271–293.Find this resource:
Bordin, E. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research and Practice, 16, 252–260.Find this resource:
*Cooper, H., Hedges, L., & Valentine, J. (2009). The handbook of research synthesis and meta- analysis (2nd ed.). New York, NY: Russell Sage Foundation.Find this resource:
Couch, A. S. (1999). Therapeutic functions of the real relationship in psychoanalysis. The Psychiatric Study of the Child, 54(1), 130–168.Find this resource:
Curtis, R., Field, C., Knaan-Kostman, I., & Mannix, K. (2004). What 75 psychoanalysts found helpful and hurtful in their own analyses. Psychoanalytic Psychology, 21(2), 183–202.Find this resource:
*Eugster, S. L., & Wampold, B. E. (1996). Systematic effects of participant role on evaluation of the psychotherapy session. Journal of Consulting and Clinical Psychology, 64(5), 1020–1028.Find this resource:
*Eun Ju, L. (2015). The unfolding of the real relationship, working alliance and the outcome (Doctoral dissertation). Sookmyung Women’s University,Seoul, South Korea. Manuscript in preparation.Find this resource:
Freud, S. (1919). Lines of advance in psychoanalytic therapy. In J. Stratchy (Ed.), Standard edition of the complete works of Sigmund Freud (pp. 157–168). London, England: Hogarth.Find this resource:
(p. 376) Freud, S. (1937). Analysis terminable and interminable. In J. Stratchy (Ed.), Standard edition of the complete works of Sigmund Freud (pp. 209–253). London, England: Hogarth.Find this resource:
Freud, A. (1954). The widening scope of indications for psychoanalysis discussion. Journal of the American Psychological Association, 2(4), 607–620.Find this resource:
*Fuertes, J. N., Ganley, J., & Moore, M. (2016). Real relationship, therapist self-disclosure, attachment, and outcome in psychotherapy. Manuscript submitted for publication.Find this resource:
*Fuertes, J. N., Gelso, C. J., Owen, J. J., & Cheng, D. (2013). Real relationship, working alliance, transference/countertransference and outcome in time-limited counseling and psychotherapy. Counselling Psychology Quarterly, 26(3–4), 294–312. https://www.doi.org/10.1080/09515070.2013.845548Find this resource:
*Fuertes, J. N., Mislowack, A., Brown, S., Gur-Arie, S., Wilkinson, S., & Gelso, C. J. (2007). Correlates of the real relationship in psychotherapy: A study of dyads. Psychotherapy Research, 17(4), 423–430.Find this resource:
Gelso, C. J. (2009). The real relationship in a postmodern world: Theoretical and empirical explorations. Psychotherapy Research, 19(3), 253–264.Find this resource:
Gelso, C. J. (2011). The real relationship in psychotherapy: The hidden foundation of change. Washington, DC: American Psychological Association Press.Find this resource:
Gelso, C. J. (2014). A tripartite model of the therapeutic relationship: Theory, research, and practice. Psychotherapy Research, 24(2), 117–131.Find this resource:
Gelso, C. J., & Carter, J. A. (1985). The relationship in counseling and therapy: Components, consequences and theoretical antecedents. The Counseling Psychologist, 13, 155–243.Find this resource:
Gelso, C. J., & Carter, J. A. (1994). Components of the psychotherapy relationship: Their interaction and unfolding during treatment. Journal of Counseling Psychology, 41, 296–306.Find this resource:
Gelso, C. J., & Hayes, J. A. (1998). The psychotherapy relationship: Theory, research, and practice. Hoboken, NJ: Wiley.Find this resource:
*Gelso, C. J., & Johnson, D. H. (1983). Explorations in time-limited counseling and psychotherapy. New York, NY: Columbia University Press.Find this resource:
*Gelso, C. J., Kelley, F. A., Fuertes, J. N., Marmarosh, C., Holmes, S. E., Costa, C., & Hancock, G. R. (2005). Measuring the real relationship in psychotherapy: Initial validation of the therapist form. Journal of Counseling Psychology, 52(4), 640–649.Find this resource:
*Gelso, C. J., Kivlighan, D. M., Busa-Knepp, J., Spiegel, E. B., Ain, S., Hummel, A. M., . . . Markin, R. D. (2012). The unfolding of the real relationship and the outcome of brief psychotherapy. Journal of Counseling Psychology, 59(4), 495–506. https://www.doi.org/10.1037/a0029838Find this resource:
Gelso, C. J., & Silberberg, A. (2016). Strengthening the real relationship: What is a psychotherapist to do? Practice Innovations, 1(3), 154–163.Find this resource:
Greenson, R. R. (1967). The technique and practice of psychoanalysis. New York, NY: International Universities Press.Find this resource:
*Gullo, S., Lo Coco, G., & Gelso, C. (2012). Early and later predictors of outcome in brief therapy: The role of real relationship. Journal of Clinical Psychology, 68(6), 614–619. https://www.doi.org/10.1002/jclp.21860Find this resource:
Hill, C. E., Gelso, C. J., Chui, H., Spangler, P. T., Hummel, A., Huang, T., . . . Miles, J. R. (2014). To be or not to be immediate with clients: The use and perceived effects of immediacy in psychodynamic/interpersonal psychotherapy. Psychotherapy Research, 24(3), 299–315.Find this resource:
(p. 377) Hill, C. E., Sim, W., Spangler, P., Stahl, J., Sullivan, C., & Teyber, E. (2008). Therapist immediacy in brief psychotherapy: Case study II. Psychotherapy: Theory, Research, Practice, Training, 45(3), 298–315.Find this resource:
Horvath, A. O., Del Re, A. C., Flückiger, C., & Symonds, D. (2011). Alliance in individual psychotherapy. Psychotherapy, 48(1), 9–16.Find this resource:
Kasper, L. B., Hill, C. E., & Kivlighan, D. M. Jr. (2008). Therapist immediacy in brief psychotherapy: Case study I. Psychotherapy: Theory, Research, Practice, Training, 45(3), 281–297.Find this resource:
*Kelley, F. A., Gelso, C. J., Fuertes, J. N., Marmarosh, C., & Lanier, S. H. (2010). The real relationship inventory: Development and psychometric investigation of the client form. Psychotherapy: Theory, Research, Practice, Training, 47(4), 540–553.Find this resource:
Kelley, F. A. (2015). The therapy relationship with lesbian and gay clients. Psychotherapy, 52, 113–118. http://dx.doi/10.1037/a0037958Find this resource:
*Kivlighan, D. M., Gelso, C. J., Ain, S., Hummel, A.M., & Markin, R. D. (2015). The therapist, the client, and the real relationship: An actor-partner interdependence analysis of treatment outcome. Journal of Counseling Psychology, 62(2), 314–320. https://www.doi.org/10.1037/cou0000012Find this resource:
*Kivlighan, D. M., Hill, C.E., Gelso, C. J., & Baumann, E. (2016). Working alliance, real relationship, session quality, and client improvement in psychodynamic psychotherapy: A longitudinal actor partner interdependence model. Journal of Counseling Psychology, 63(2), 149–161. https://www.doi.org/10.1037/cou0000134Find this resource:
*Kivlighan, D. M. Jr., Baumann, E. C., Gelso, C. J. & Hill, C. E. (2014). Symptom change and between therapist and within therapist variability in third session intercepts and linear change slopes for longitudinal ratings of alliance and real relationship. Paper presented at the meeting of the Society for Psychotherapy Research, Copenhagen, Denmark.Find this resource:
Kivlighan, D. M. Jr., Kline, K., Gelso, C. J., & Hill, C. E. (2017). Congruence and discrepancy between working alliance and real relationship: Variance decomposition and response surface analyses. Journal of Counseling Psychology, 64(4), 394–409.Find this resource:
Knox, S., Hess, S. A., Petersen, D. A., & Hill, C. E. (1997). A qualitative analysis of client perceptions of the effects of helpful therapist self-disclosure in long-term therapy. Journal of Counseling Psychology, 44(3), 274–283.Find this resource:
Ledermann, T., & Kenny, D. A. (2012). The common fate model for dyadic data: Variations of theoretically important but underutilized model. Journal of Family Psychology, 26(1), 140–148.Find this resource:
*Lee, E. J. (2017). The unfolding of the working alliance, the real relationship and the outcome (Unpublished doctoral dissertation). Sookmyung Women’s University, Seoul, South Korea.Find this resource:
Li, X., Kivlighan, D. M. Jr., Gelso, C. J., & Hill, C. E. (2016, November). Longitudinal relationships among the real relationship, working alliance and session outcome: A common fate model. Paper presentedat the 2016 North American Society of Psychotherapy Research Conference, Berkeley, CA.Find this resource:
*Lo Coco, G., Gullo, S., Prestano, C., & Gelso, C. J. (2011). Relation of the real relationship and the working alliance to the outcome of brief psychotherapy. Psychotherapy, 48(4), 359–367.Find this resource:
*Marmarosh, C. L., Gelso, C. J., Markin, R. D., Majors, R., Mallery, C., & Choi, J. (2009). The real relationship in psychotherapy: Relationships to adult attachments, working alliance, transference, and therapy outcome. Journal of Counseling Psychology, 56(3), 337–350. https://www.doi.org/10.1037/a0015169Find this resource:
(p. 378) Morales, K. S., Keum, B. T., Kivlighan, Jr., D. M., Hill, C. E., & Gelso, C. J. (2018). Therapist effects due to client racial/ethnic status when examining linear growth for client-and therapist-rated working alliance and real relationship. Psychotherapy, 55(1), 9–19.Find this resource:
*Owen, J. J., Tao, K., Leach, M. M., & Rodolfa, E. (2011). Clients’ perceptions of their psychotherapists’ multicultural orientation. Psychotherapy, 48(3), 274–282.Find this resource:
*Perez Rojas, A. (2015). Does the acculturation of international student therapists predict the process of psychotherapy with U.S. clients? An exploratory study (Unpublished doctoral dissertation). University of Maryland, College Park.Find this resource:
Perls, F. (1969). Gestalt therapy verbatim. Lafayette, CA: Real People Press.Find this resource:
Rogers, C. (1957). The necessary and sufficient conditions of therapeutic personality change. Journal of Consulting Psychology, 21(2), 95–103.Find this resource:
Wampold, W. E., & Budge, S. L. (2012). The relationship and its relationship to the common and specific factors of psychotherapy. The Counseling Psychologist, 40(4), 601–623.Find this resource:
Yalom, I. (2002). The gift of therapy. New York, NY: HarperCollins.Find this resource: