ABSTRACT Title of Dissertation: THE REAL RELATIONSHIP, THERAPIST IMMEDIACY, AND CLIENT EXPERIENCING LEVEL: A DYAD STUDY OF PSYCHOTHERAPY PROCESS AND CONNECTION Tamara Tashiko Walden, Doctor of Philosophy, 2013 Dissertation directed by: Professor Charles J. Gelso, Department of Psychology The present study examined the strength of the real relationship (RR) in relation to session quality, therapist immediacy (T-IMM), and client experiencing level (C-EXP) in six psychotherapy dyads over the course of time-limited treatment. Six clients and three therapists completed measures of RR and session quality immediately after each session. Trained judges rated the amount, appropriateness, depth, quality and resolution of T-IMM and the peak and mode level of C-EXP using audio recordings and transcripts post-treatment. This study used the Actor Partner Interdependence Model (APIM; Kenny & Cook, 1999), Structural Equation Modeling (SEM), and Hierarchical Linear Modeling (HLM) to analyze the data due to nesting and interdepence. No significant associations were detected between therapist or client ratings of RR and session quality, between therapist ratings of RR and T-IMM, or between therapist or client ratings of RR and C- EXP. Client perceptions of RR were significantly related to T-IMM rating dimensions, though not significantly related to T-IMM amount or occurrence. The results indicate that regardless of the amount, when therapist immediacy is used appropriately, with depth, quality or resolution, clients perceive stronger real relationships with their therapists. Implications for practitioners and researchers regarding the findings are discussed. In addition, due to the unique racial/ethnic minority (REM) composition of the majority of the dyads, cultural implications for research and practice are considered. THE REAL RELATIONSHIP, THERAPIST IMMEDIACY, AND CLIENT EXPERIENCING LEVEL: A DYAD STUDY OF PSYCHOTHERAPY PROCESS AND CONNECTION by Tamara Tashiko Walden Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2013 Advisory Committee: Professor Charles J. Gelso, Chair Assistant Professor Julia Bryan, Professor Clara E. Hill Professor Dennis M. Kivlighan Assistant Professor Sharon Kirkland-Gordon ? Copyright by Tamara Tashiko Walden 2013 ii Dedication To Helen Walden Eason, my mother, who continues to be my soil, water, and sun ? without your love, faith, and courage I would not have succeeded. To Xander Akiro Walden-Regenold, my son and bright light and my Mr. Incredible ? without your joy and spirit I would not have realized fully who I am and what I am capable of doing. To Nyla Nyoko Walden-Regenold, my daughter, You are with me always ? my love and loss of you connects me to the preciousness and the grace in life. Mommy, Xander-Bean and Nyla-Bean, you have been deeply-needed, daily inspirations of love, faith, and never giving up. You have made me a better clinician, scientist, mother and human being. Thank you. I love you all with all my heart and soul. iii Acknowledgments I would first like to thank my advisor Charlie Gelso for his mentorship and personal support. I am profoundly grateful for his willingness to accommodate the twist and turns of my journey throughout this process. I would also like to express my sincere appreciation for the commitment, faith and support he has provided me and thank him for the awareness, insight, and focus his guidance has given me. I could not have completed this project without his encouragement and his belief in me. I feel very fortunate to have had him as an advisor, as his wisdom throughout our time together has more than prepared me to be a successful psychologist and person. I would also like to thank the other members of my committee, Julia Bryan, Clara Hill, Dennis Kivlighan, and Sharon Kirkland-Gordon, who have been extremely gracious with their time and expertise throughout this project and whose grace and brilliance in my proposal and defense meetings were remarkable and tremendously educational to me as a clinician and researcher. I would especially like to thank Dennis Kivlighan whose generosity, compassion, and expertise went above and beyond any call of duty. His generously- given statistical efforts with my project have made it a substantial contribution to the literature. I would also like to thank Amy Zontek, Melissa Thevenot, Shari Pincus, Christina Sakala, Laura Dochtherman, and Graham Knowles, who comprised my rating teams. Without their dedication, reliability, work ethic, and flexibility I would not have successfully collected my data. I am deeply grateful for their belief in me and my research, and their willingness to see the project through despite any adversity. In addition, working with them was personally and professionally rewarding, as our journey iv was collaborative and stimulating. I would also like to thank my colleagues, everyday poets and best friends who have provided me with precious help when I needed it most that ranged from editing to encouragement to laughter. In particular, I would like to thank Will Hayden, who always found the most extraordinary words during some of my most painful times. I would like to also thank Megan MacNamara-Monday, whose words and actions affected me with a kindness and compassion that was often a rope leading me over of the highest walls. I would like to thank Margit Berman, who?s brilliant grace struck with such power and compassion that has left me humbled and inspired. I would also like to thank Chinyere Mathewson, whose humor and extraordinary life sense has been a reliable and nourishing compass for me. I also want to thank my sister, Ayanna Miyako Walden, who has continued to be the ?Ying? to my ?Yang? as her editing and encouragement complimented my work tremendously. Finally, I would like to thank my mother, whose tireless support, endless patience, and essential role in taking care of my son and myself, has allowed me to successfully complete this project, but has also empowered me to be a successful person and psychologist. I am grateful beyond words for her everyday effort and sacrifice to help me during this project and for the lifetime of love and nourishment she has given me. v Table of Contents Table of Contents???????????????????????????.?v! Chapter 1: Introduction and Review of Literature???????????????...1! Introduction???????????????????????????..1! Review of Literature????????????????????????5! The Real Relationship????????????????????.6! The Therapy Relationship ? the Tripartite Model??????...6! The History and Controversy of the Real Relationship????.10! Measuring the Real Relationship????????????...12! Research on the Real Relationship????????????14! Therapist Immediacy????????????????????.21! Form and Function of Therapist Immediacy????????.22! Subtypes of Therapist Immediacy????????????.23! Theoretical Context of Immediacy????????????25! Rationale for Using Immediacy?????????????.26! Research on Therapist Immediacy????????????.27! Therapist Immediacy and the Real Relationship??????...33! Client Experiencing Level??????????????????.35! History and Context of Client Experiencing Level??????35! Basic Tenets of Experiencing??????????????36! Definition?????????????????????...38! Measuring Experiencing Level?????????????..38! Research on Client Experiencing Level??????????.40! vi Client Experiencing Level and the Real Relationship??...??42! Chapter 2: Statement of the Problem and Hypotheses?????????????...45! Statement of the Problem?????????????????????...45! Hypotheses???????????????????????????.46! Chapter 3: Method???????????????????????????.52! Design and Data Set???????????????????????...52! Participants???????????????????????????.53! Measures???????????????????????????....58! Procedures for Data Collection???????????????????..66! Procedures for Data Coding. ????????????????????.69! Chapter 4: Results???????????????????????????..77! Preliminary Analyses???????????????????????.77! Analysis of Hypotheses and Research Questions????????????..86! Real Relationship and Session Quality??????????..88! Real Relationship and Therapist Immediacy????????.91! Real Relationship and Client Experiencing Level??????.98! Immediacy and Client Experiencing Level?????????99! Chapter Five: Discussion????????????????????????.110! Descriptive Findings. ??????????????????????..110! Main Findings?????????????????????????..114! The Real Relationship and Session Quality????????.114! The Real Relationship and Therapist Immediacy??????117! The Real Relationship and Client Experiencing Level (C-EXP).120! vii Additional Analyses ? Therapist Immediacy and Client Experiencing Level?.121! Implications for Practice and Research????????????????122! Limitations??????????????????????????...126! Appendix A?????????????????????????????..131! Appendix B?????????????????????????????..132! Appendix C?????????????????????????????..133! Appendix D?????????????????????????????..134! Appendix E?????????????????????????????..135! Appendix F?????????????????????????????..136! Appendix G?????????????????????????????..137! Appendix H?????????????????????????????..138! Appendix I?????????????????????????????...139! Appendix J?????????????????????????????...140! Appendix K?????????????????????????????..141! Appendix L?????????????????????????????..142! References??????????????????????????????143! List of Tables?????????????????????????????viii viii List of Tables Table 1. Therapists (TH) & Clients (CL) Descriptive Information. ................................ 55! Table 2. Means and Standard Deviations for SQ-T, SQ-C , RRI-T and RRI-C .............. 80! Table 3. Means and Standard Deviations for RRI-T, RRI-C, and IMM-Amt. ................ 81! Table 4. Means and Standard Deviations for RRI-T, RRI-C, IMM-A, IMM-D, IMM-R, and IMM-Q. ...................................................................................................................... 82! Table 5. Means and Standard Deviations for RRI-T, RRI-C, C-EXP-M, and C-EXP-P 83! Table 6. Means and Standard Deviations for C-EXP-M, C-EXP-P, IMM-A, IMM-D, IMM-R, IMM-Q ............................................................................................................... 84! Table 7. Frequency Summary of Immediacy Speaking Turns for each Dyad. .............. 137! Table 8. Correlation Matrix for Therapist Immediacy Dimensions. ............................... 86! Table 9. Real Relationship & Therapist Immediacy Amount ........................................ 139! Table 10. Real Relationship & Therapist Immediacy Occurrence ................................ 140! Table 11. RR & Immediacy Depth .................................................................................. 95! Table 12. RR & Immediacy Appropriateness .................................................................. 96! Table 13. RR & Immediacy Resolution ........................................................................... 97! Table 14. RR & Immediacy Quality ................................................................................ 98! Table 15. RR & Peak Client Experiencing Level .......................................................... 141! Table 16. RR & Client Experiencing Level - Mode ...................................................... 142! Table 17. Immediacy Depth & Post-Experiencing Mode .............................................. 102! Table 18. Immediacy Appropriateness & Post-Experiencing Mode ............................. 103! Table 19. Immediacy Resolution & Post-Experiencing Mode ...................................... 104! Table 20. Immediacy Quality & Post-Experiencing Mode ........................................... 105! ix Table 21. Immediacy Depth & Post-Experiencing Peak ............................................... 106! Table 22. Immediacy Appropriateness & Post-Experiencing Peak ............................... 107! Table 23. Immediacy Resolution & Post-Experiencing Peak ........................................ 108! Table 24. Immediacy Quality & Post-Experiencing Peak ............................................. 109! Table 25. Correlation Matrix for RR,R-IMM and C-EXP. ............................................ 138! DYAD STUDY OF PSYCHOTHERAPY PROCESS 1 Chapter 1 Introduction In recent years, researchers are giving more attention to the therapeutic relationship in efforts to find the ?what? and ?how? of effective psychotherapy (Norcross, 2002). Although the centrality of the therapeutic relationship in psychotherapy varies across theoretical orientations, the therapeutic relationship remains one of the strongest predictors of outcome (Lambert and Barley, 2002). However, what in the relationship is healing? Carl Rogers (1961) made the profound claim that: ?What is most personal and unique in each one of us is probably the very element which would, if it were shared or expressed, speak most deeply to others? ( p.26). Despite Rogers?s (1961) propitious claim about the nature of authentic and intimate connection, the majority of research on the therapeutic relationship has focused mostly on the less personal or less authentic aspects of the connection between client and therapist. There has been extensive research on the working alliance (Bordin, 1979; Greenson, 1967, 1978; Greenson & Wexler, 1969) and some research, too, on transference and countertransference (Gelso & Hayes, 1998, 2002, 2007; Graff & Luborsky, 1977; Multon, Patton, & Kivlighan, 1996) in the investigation of the role of the therapy relationship in treatment outcome. However, two issues have limited this body of literature. First, the wide range of definitions used in working alliance research introduced what Norcross (2002) called ?conceptual fuzziness.? Specific ?contents? such as behaviors, interventions, and conceptualizations from diverse schools of thought have not been easily integrated in research and clinical practice because there has been little common terminology across different approaches. In addition, different psychotherapy DYAD STUDY OF PSYCHOTHERAPY PROCESS 2 methods from different theoretical approaches can engender the same therapeutic process, while the same methods can engender different therapeutic processes. As a result, it can be difficult to determine what is influencing client process and outcome. Thus, it has been challenging for researchers to integrate findings and feel confident they are examining the same construct. Secondly, researchers are faced with the challenge of combining enough breadth to integrate diverse psychological perspectives while maintaining enough depth to discover insight that is experience-near and clinically relevant. Psychotherapy efficacy researchers (especially in the empirically-supported or evidence-based treatment area) have often over-emphasized techniques and methods to such an extent that it can portray psychotherapy as not involving a person at all. Norcross (2002) warned that empirically- supported treatment research can result in lists and guidelines that portray psychotherapy as the work of ?disembodied therapists performing procedures on Axis 1 disorders? (p. 4). Considering these challenges, a more integrative conceptual framework for the therapeutic relationship is needed to improve process and outcome research. In addition, as Rogers?s (1961) insightful words suggest, there is a need for scientific examination of the therapeutic relationship that takes the person as well as the more intimate and authentic contributions of the therapist and client into account. Over the past six years, Gelso and his collaborators have been engaged in a program of research that theoretically and empirically explores a promising, and in some respects controversial, construct that addresses those challenges (Ain & Gelso, 2008; Fuertes, Mislowack, Brown, Gur-Arie, Wilkinson, & Gelso, 2007; Fuertes, Gelso, DYAD STUDY OF PSYCHOTHERAPY PROCESS 3 Perolini, Walden, Kasnakian, & Parsons, 2008; Gelso, Kelley, Fuertes, Marmarosh, & Holmes, Costa & Hancock, 2005; Hummel & Gelso, 2010; Kelley, Gelso, Fuertes, Marmarosh, & Lanier, 2010; LoCoco et al., 2011; Marmarosh, Gelso, Markin, Majors, Mallery, & Choi, 2009; Moore & Gelso, 2011; Spiegel, Busa-Knepp, Ma, Markin, Ain, Hummel, ...Gelso, 2008). The construct is called the real relationship. Gelso and his colleagues (Gelso, 2009a, 2009b, 2011; Gelso & Hayes, 1998; Gelso & Carter, 1985, 1994) proposed the tripartite model of the therapeutic relationship that is comprised of three interrelated components: the working alliance, the transference-countertransference configuration, and the real relationship. This model integrates previous theory and research on the therapy relationship that has been at the forefront of the empirical lens (e.g. working alliance and transference). At the same time, the model illuminates a third component of the overall relationship: the personal or real relationship that, until recently, has been empirically ignored and theoretically overlooked (Gelso 2009a, 2009b). The real relationship is a construct that focuses on the authentic and non-distorted quality of the personal connection between the therapist and client (Gelso, 2009a, 2009b, 2011; Gelso & Hayes, 1998; Gelso & Carter, 1985, 1994). Gelso (2009a) defined the real relationship as the ?personal relationship existing between two or more people as reflected in the degree to which each is genuine with the other, and perceives and experiences the other in ways that befit the other?(p. 254-255). Thus, the extent or amount of genuineness and realism in the relationship between the therapist and client determines the strength of the real relationship. As such, the real relationship is argued to be an essential concept and experience of the overall therapy relationship. Moreover, it is an important factor in effective psychotherapy that captures something beyond the work DYAD STUDY OF PSYCHOTHERAPY PROCESS 4 alliance that is not merely the opposite of transference (Gelso, 2009a, 2009b, 2011; Gelso & Hayes, 1998; Gelso & Carter, 1985, 1994). Investigation of the real relationship enables research to examine how the non-distorted, authentic relationship between the therapist and the client influences psychotherapy process and outcome independent of tasks, goals, or any ingredients specific to a particular treatment or theoretical approach. Therapist immediacy has been theorized to be an effective way to process the therapeutic relationship by acknowledging here-and-now interactions with the client and therapist (Hill and Knox, 2009). Furthermore, therapist immediacy has been argued to be a valuable mechanism of change in psychotherapy that can strengthen the relationship and lead to significant client gain (Bordin, 1979; Hill and Knox, 2009; Teyber and McClure, 2011). One of the most compelling aspects of therapist immediacy in regards to the real relationship is that it is likely experienced by the therapy participants as an invitation to interact with each other in an intimate, self-revealing manner with more authenticity and less facade, defense, and repression. Thus, therapist immediacy may capture an essential process, which helps each real relationship to develop its unique, authentic and intimate character. As one of the most researched psychotherapy process variables in the literature, client experiencing level has robustly predicted success and failure in treatment outcome for over 50 years. Client experiencing level is thus an integrative and empirically established change process in psychotherapy (Gendlin, 1968; Orlinsky and Howard 1986). It refers to the manner of client self-exploration and in-session, bodily-felt engagement. Higher levels of experiencing equate to greater self-involvement but also to less intellectualizing, event reporting, or mere cartharting ?about? one?s self. Regarding DYAD STUDY OF PSYCHOTHERAPY PROCESS 5 the real relationship, client experiencing level is a very compelling construct because higher levels of experiencing and the lack of the ?about me? phenomena are likely to affect the genuineness and realism in the therapy participants? real relationship (Gendlin, 1968, 1996). The purpose of the current study, then, is to create a clearer picture of how psychotherapy is successful by examining how the strength of the real relationship relates to psychotherapy outcome and progress, as well as to therapist immediacy and client experiencing, two constructs that offer integrative and personal understandings of psychotherapy process. Review of Literature The review of literature has three main sections. The first section reviews the real relationship. This section examines the history of the real relationship including its conceptualization within an overall model of the therapy relationship, definition, measurement, and related research. The second section reviews therapist immediacy. This section examines the definition of therapist immediacy, the form and function of therapist immediacy including its pan-theoretical context and description of its subtypes, and concludes with relevant research that pertains to treatment outcome and its connection to the real relationship. Lastly, client experiencing level is reviewed. In this section, the history, definition, and theoretical model underlying client experiencing level are briefly reviewed. This section concludes with a review of its measurement, empirical highlights and a discussion of its connection to the real relationship. DYAD STUDY OF PSYCHOTHERAPY PROCESS 6 The Real Relationship The Therapy Relationship ? the Tripartite Model Gelso and Hayes (1998), extending the original work of Greenson (1967) and Gelso and Carter (1985, 1994), conceptualized a model of the therapeutic relationship that is comprised of three interrelated components: the working alliance, the transference- countertransference configuration, and the real relationship. Gelso and Carter?s (1985) improvement to Greenson?s theory was a result of explaining how the model was relevant to all types of therapy, thus making it integrative across several theoretical approaches. Gelso (2009a, 2009b, 2011), like Greenson, proposed that the working alliance emerges from the real relationship. Furthermore both components impact the extent to which the client or therapist works with his or her respective transference ultimately affecting the therapeutic relationship as it unfolds. As a result, Gelso argued that the real relationship is an essential and fundamental component of the therapeutic relationship that is not merely equal to the work alliance or the opposite of transference. Rather, the real relationship captures important phenomena that go beyond the working relationship (e.g. alliance) and transference and thus needs to be studied. Below, each component of the tripartite model will be briefly discussed. However, the real relationship component will be given the most focus because it is the component of the therapeutic relationship that will be examined in the present study. The working alliance. The working alliance is the component of the tripartite model of the therapeutic relationship that solely functions to further the therapeutic work (Gelso 2009b; Gelso 2009a; Gelso 2010). Gelso and Carter (1994) defined the working alliance as ?the alignment or joining of the reasonable self or ego of the client and the DYAD STUDY OF PSYCHOTHERAPY PROCESS 7 therapist?s analyzing or ?therapizing? self or ego? (p. 297). The alignment of the reasonable sides of the participants is proposed to facilitate the following three ingredients conceptualized by Bordin (1979) to characterize an effective working alliance. The strength of the working alliance is affected by the degree to which the therapist and client (a) agree on the therapeutic goals of their work, (b) agree on the tasks that will be useful in attaining those goals, (c) and experience a sound bond around the work itself. The working alliance has been one of the most clearly defined and heavily researched components of the therapeutic relationship for the past 30 years. As a result, there is strong research evidence suggesting that across different theoretical approaches to therapy, a strong working alliance contributes significantly to successful psychotherapy (Horvath & Bedi, 2002; Horvath, 1991; Samstag, 2006). The transference-countertransference configuration. The transference- countertransference configuration in the therapy hour is the component of the therapeutic relationship that is based on unrealistic perceptions. Transference is defined as the ?client?s experience of the therapist that is shaped by the client?s own psychological structures and past, and involves displacement onto the therapist of feelings, attitudes, and behaviors belonging rightfully in earlier significant relationships? (Gelso and Hayes, 1998, p. 51). In contrast, Gelso and Hayes (2007) define countertransference as, ?the therapist?s internal and external reactions that are shaped by the therapist?s past or present emotional conflicts and vulnerabilities (p. 25). The manner or configuration of the transference from either the client or therapist can be subtle and is often complex. Transference configurations can play positive, negative, or neutral roles in therapy, depending on their content and how they are processed in the therapy session. Thus DYAD STUDY OF PSYCHOTHERAPY PROCESS 8 countertransference can provide valuable information about the interaction, but managing it effectively often depends on the ability to process the way it is experienced and used therapeutically. The real relationship. Although the real relationship is theorized as the most basic and fundamental of the three components, the real relationship has received the least empirical scrutiny. Most of the recent scholarship and empirical investigation of the real relationship has used Gelso and his colleagues? definition of the real relationship (Gelso 2002, 2009a; Gelso & Samstag, 2008). As stated earlier, Gelso (2009a) defined the real relationship as ?the personal relationship existing between two or more people as reflected in the degree to which each is genuine with the other, and perceives and experiences the other in ways that befit the other? (p. 254-255). Thus, there are two main aspects to this definition: realism and genuineness. Realism denotes realistic perceptions and reactions to another person in the moment. Realistic perceptions refer to the perceptions of the client or therapist that are not distorted by transference or other defenses, allowing the client and therapist to view each other realistically (Gelso 2009). The second feature of the real relationship is genuineness, which is the willingness and ability to be who one truly is in the here-and- now. In a real relationship that contains a high level of genuineness, the client and therapist will experience themselves and the other as authentic, open, honest, and congruent as opposed to phony, fake or defended (Gelso 2009a). In addition, there are two sub-elements of the real relationship: magnitude and valence. Magnitude refers to how much of the real relationship is present and can vary from low to high. Valence refers to how positive or negative the therapist?s and client?s feelings are toward one DYAD STUDY OF PSYCHOTHERAPY PROCESS 9 another, and this dimension is best understood as a continuum (Gelso 2009a, 2009b, 2011). Gelso and his collaborators offer important theoretical propositions regarding the real relationship that may not be commonly held views in the field of psychotherapy as of yet (Gelso 2009a, 2009b, 2011; Gelso and Hayes 1998). First, the real relationship is not limited to therapeutic relationships or specific types of interactions. There are real relationship aspects in every relationship and within each interaction inside that relationship. Whereas the working alliance exists solely for the purposes of the work aspect of therapy (Gelso, 2007), the real relationship is based on the experiences between two people separate from the working aspects of the relationship and begins to develop from the first moment of interaction. In addition, the real relationship is not limited to verbal exchanges. Affective experiences and non-verbal communication (e.g., tone of voice, eye contact, etc.) also contribute to the development of genuineness and realism occurring between the client and therapist. Thus, the real relationship is theorized to develop independent of theoretical orientation, technique, or any other specific ingredient of therapy, whereas the working alliance develops solely in connection to the therapeutic work. Within the transference-countertransference configuration, the therapist?s and client?s transference reactions may be genuine, but they are based on unrealistic perceptions (Gelso 2009a, 2009b, 2011; Gelso and Hayes 1998). This is assumed to be separate from the real relationship because the real relationship is based on genuine and realistic perceptions. At the same time, compared to the countertransference-transference configuration, the real relationship is not conceptualized as merely the opposite of DYAD STUDY OF PSYCHOTHERAPY PROCESS 10 transference nor mutually exclusive. Both can be high or low during the same time period, session or communication. This is because every perception has a real relationship element and a transference element. In sum, although the working alliance and transference-countertransference configuration may each relate to aspects of the real relationship, the real relationship is an important and theoretically distinct component of the overall therapeutic relationship. The History and Controversy of the Real Relationship Even though the real relationship has not been in the forefront of clinical theory and empirical examination, the existence of a real or personal relationship has been acknowledged since psychotherapy began in its early psychoanalytic roots. Sigmund Freud (1937) noted that ?Not every good 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 which proved to be viable? (p.222). Many psychoanalysts wrote about the real relationship, particularly emphasizing realism, perceiving and experiencing the other without transference distortion (Gelso 2009a). However, Ralph Greenson?s (1967) ideas differed from his predecessors in that he believed that the real relationship consisted of genuineness as well as realism. The concept of genuineness was challenging to the traditional psychoanalysts of the time because it was associated with self-disclosure and thus, in their eyes, requiring a departure from the ?blank-screen? neutrality central to psychoanalytic practice (Gelso 2009a). Unlike their psychoanalytic colleagues, humanist clinicians like genuineness as it is a central tenet of almost all humanistic theories. In conjunction, most humanists DYAD STUDY OF PSYCHOTHERAPY PROCESS 11 (and many post-modern analysts) take issue with realism because of their underlying constructivist and/or social constructionist beliefs (2009a). In a similar vein, amidst the current postmodern zeitgeist of psychotherapy, the term ?real?, the concept of ?realism?, and the capability of an empirical measure to determine what is real continues to be controversial. Gelso (2009a, 2009b) addressed the skepticism and criticism by subscribing to the philosophical stance of constructive realism (Neisser, 1967). Gelso asserted that, ?one can believe there is a reality of the therapist and a reality of the client, but also believe that the participants? perceptions or constructions of these realities are extremely important? (p.256). Thus, the therapist need not be the sole arbiter of truth and reality in the real relationship. Ultimately, for the real relationship to be strong, the client and therapist must strive to grasp each other?s reality. In response to challenges that the real relationship seems superfluous because everything is real in the therapeutic relationship, Gelso (2009a, 209b) has countered that although the feelings, attitudes and behaviors of the relationship exist and thus are real, the concept of the real relationship is referring to more than the mere existence of relationship phenomena. The concept is addressing the manner or quality of the existing feelings, attitudes and behaviors; namely the realism of the participants? perceptions and experience and the genuineness of their expressions. Thus, the real relationship is likely to add different but essential aspects of the therapeutic relationship that the other relationship components (e.g. working alliance an transference) do not capture. Gelso?s (2009a, 2009b) response also clarifies complaints about the use of the term ?real? which some have understood to mean that there were unreal aspects of the relationship by implication. Gelso discourages considering any phenomena in the DYAD STUDY OF PSYCHOTHERAPY PROCESS 12 relationship as unreal. However, if the term ?real? is not made synonymous with the term ?existence? and is instead understood as the genuine and realistic quality of the relationship between the client and therapist, then the label is less problematic. Gelso has acknowledged that any term will have its strengths and limitations, and thus these challenges do not detract from the value or viability of the real relationship construct. Measuring the Real Relationship The first study to examine the role of the real relationship in psychotherapy was completed by Eugster and Wampold (1996) in their investigation of which psychotherapy processes predicted client- and therapist-rated session evaluation. In this study the first measures of the real relationship were presented as a part of a larger battery that examined a total of nine process variables rated by either the therapist or patient. The patient real relationship measure assessed the patient?s feeling of liking for the therapist, feeling of personal connection to the therapist, and genuine or authentic manner of relating. Differing slightly, the therapist real relationship measure focused on the therapist?s willingness to be known through transparency or disclosure (i.e. genuineness), the therapist?s liking for the patient, and the therapist?s role-defined versus natural and spontaneous behavior. Results of the study showed that therapist-rated session evaluation was best predicted by therapist expertness, while patient-rated session evaluation was best predicted by therapist real relationship. The results indicated that a patient would evaluate the session positively when she or he perceives that the therapist is not relating to him or her in a prescribed, role-bound manner and instead is experienced by the patient as natural, spontaneous and authentic. DYAD STUDY OF PSYCHOTHERAPY PROCESS 13 It is important to note that the findings may have also indicated that the patients in this study valued therapist real relationship over any of the other psychotherapy processes assessed: Interestingly, regression analysis indicated that therapist-rated real relationship negatively predicted therapists? session evaluation. However, this finding needs to be replicated as an additional bivariate analysis indicated the opposite result (i.e. RR positively predicted session evaluation). In addition, this study exhibited some important limitations related to the real relationship measure development. Despite the authors? efforts to tailor the real relationship measures after the theoretical formulations of Gelso and Carter (1994), the measures exhibited only minimally acceptable reliability. In addition, the items on the patient form did not tap into realism as much as features of genuineness. Instead, they loaded heavily on genuineness and liking or caring about the other. Regardless of the psychometric limitations of the patient form, the Eugster and Wampold study served as an important initial validation study providing direction and insight for later measurement efforts. Building upon Eugester and Wampold?s (1996) efforts, more psychometrically sound and theoretically-rooted therapist (RRI-T; Gelso, et al., 2005) and client (RRI-CL; Kelley, et al., 2010) forms were developed to measure the strength of the real relationship. Gelso and his colleagues wrote 130 items that eventually were developed into a client and a therapist measure each consisting of 24 items that break down into two 12-item subscales, one for realism and one for genuineness. Each scale also assesses the magnitude and valence of the real relationship. For each of the two forms of the measure, the rater is asked to rate the self, the other, and their relationship with the other. The DYAD STUDY OF PSYCHOTHERAPY PROCESS 14 therapist form (RRI-T; Gelso, et al., 2005) has demonstrated convergent and discriminant validity. It has been found to positively relate to therapist ratings of their working alliance (with clients), clients? level of insight (emotional and intellectual), and session outcome. The therapist form also exhibited a negative relation to transference and was not correlated with social desirability. Similar to its counterpart, the client form (RRI-CL; Kelley, et al., 2010) also exhibited convergent and discriminant validity. It was found to correlate positively to Eugster and Wampold?s (1996) measures of the real relationship, a client?s capacity to observe, reflect and understand oneself (observing ego strength), and patient ratings of their therapists? genuineness on the Barrett-Lennard Relationship Inventory (BLRI; Barrett-Lennard, 1964). The client form also negatively correlated to a measure of one?s tendency to hide true feelings and behave in accordance to others? expectations (other- directedness) but had no significant correlation to social desirability. Thus, both the therapist and client forms that assess the strength of the real relationship are highly reliable and have strong early evidence of their validity. . Research on the Real Relationship As a result of the development of these new measures, a number of important studies on the real relationship have been undertaken. At this point, twelve studies have examined the real relationship (RR) during psychotherapy using these measures (Ain & Gelso, 2008; Fuertes, et al. 2007, Fuertes et al., under review; Gelso, et al. 2005; Gelso et al., 2012; Hummel & Gelso, 2010; Kelly et al., 2010; Owen, Imel, Tao, et al. 2011; Spiegel et al., 2008; LoCoco et al., 2011; Marmarosh et al., 2009; Moore & Gelso, 2011). These studies have varied in methodologically important ways (relevant to this study) and DYAD STUDY OF PSYCHOTHERAPY PROCESS 15 include both retrospective studies and prospective studies sampling dyads in naturalistic settings. Although many of the studies? findings apply to more than one area, this research can be classified along three main areas of inquiry: the strength of the real relationship in connection to (1) other components of the therapeutic relationship, (2) progress or outcome, and (3) therapist and/or client variables. This collection of studies is a promising beginning of the empirical investigation of the real relationship, as it demonstrates the real relationship?s incremental validity and how it is correlating in theoretically consistent ways with other variables and psychotherapy outcome. However, there are also limitations to these investigations that suggest directions for future empirical examination. Real Relationship, Transference, and Work Alliance. Seven quantitative studies (Fuertes et al., 2007, Fuertes et al., under review; Gelso et al., 2005; Kelley, et al., 2010; LoCoco, et al., 2011; Marmarosh et al., 2009; Owen, Tao, Leach, & Rodolfa, 2011) provide supportive data of the real relationship?s (RR) association to the working alliance (WA) and transference. The findings of these studies provide mixed evidence that therapists and clients are able to differentiate between RR and WA even though the two constructs are related to each other. For example, therapists? ratings of the real relationship and the working alliance have exhibited a moderate range of overlap (23% to 43%) in the above correlational studies. On the other hand, the RR and WA ratings of the clients in those studies have shown a much higher range of overlap (52% to 43%), suggesting that clients may have more difficulty distinguishing the real relationship and the working alliance. Two of the studies (Marmarosh et al., 2009; Gelso et al., 2005) also tested the association between the strength of RR and transference. Consistent with what DYAD STUDY OF PSYCHOTHERAPY PROCESS 16 was theoretically predicted by Gelso (2009a), both studies found therapist ratings of RR (but not patient ratings) to be moderately and negatively associated to therapist ratings of the patient?s transference. In addition, these findings illustrate the importance of the real relationship inventory?s utility in examining how therapist and client perceptions of the strength of RR differentially relate to other constructs (in this case, therapist perceptions of the client?s negative transference). Real Relationship, Client Factors and Therapist Factors. Several recent studies have found relationships between the real relationship and attachment style, observing ego, attending to one?s feelings, other-directedness, therapist empathy, and therapist self-disclosure (Ain & Gelso, 2008, 2011; Fuertes et al., 2007; Gelso et al. 2005; Kelly et al., 2010; LoCoco et al., 2011; Marmarosh et al., 2009; Moore & Gelso, 2011). In addition to the instrument-development studies mentioned above (i.e. observing ego, insight, other-directedness, therapist empathy, etc.), two dyad studies (Fuertes et al., 2007; Marmarosh et al., 2009) and one recent retrospective study (Moore & Gelso, 2011) produced findings indicating a significant relationship between RR and attachment style, as well as attachment to therapist. Despite these findings with related constructs, no study to the author?s knowledge has examined therapist immediacy or client experiencing level in relation to the strength of the RR. Real Relationship, Progress, and Outcome. The empirical scholarship exploring the real relationship and outcome has been an important part of construct validity for the real relationship. Eight studies (Ain and Gelso, 2008, 2011; Fuertes et al., 2007; Fuertes et al., under review; Gelso et al., 2012; LoCoco et al., 2011; Marmarosh et al., 2009; Spiegel et al., 2008) to date have found significant connections between the real DYAD STUDY OF PSYCHOTHERAPY PROCESS 17 relationship, including its two defining elements (realism and genuineness), and treatment progress and outcome. In order for the real relationship construct to have any meaning and measurement value, it must predict relevant phenomena differently than the working alliance. The results of these studies provide key supporting data that the real relationship predicts outcome significantly and differently than the working alliance. However, at the same time, the studies? findings were inconclusive regarding which sampling point best predicts outcome (i.e. early-, middle-, late-treatment, therapist- or client- ratings). As one of the first dyad studies to use the new real relationship measures, Fuertes et al., (2007) examined the association among client and therapist perceptions of the real relationship (RR) and working alliance (WA), clients? attachment to their therapist, therapist empathy, and treatment progress in ongoing psychotherapy. The results of the study showed that client ratings of RR predicted progress in treatment above and beyond their ratings of WA, attachment to therapist, and therapist empathy. This was a key finding as it evidenced that client-rated RR predicted treatment progress beyond key relational variables. Unfortunately, in order to maximize participation in a naturalistic setting, this study only measured ratings of RR at one random point in time. In addition, the participants could not rate treatment outcome, only treatment progress (at one point in time), as the researchers could not control if measures were completed at the end of treatment or during ongoing treatment. Another dyad study by Marmarosh et al. (2009) replicated and extended the Fuertes et al. (2007) findings by investigating the relationships among client and therapist ratings of RR and WA early in treatment (3rd session), client adult attachment and therapy DYAD STUDY OF PSYCHOTHERAPY PROCESS 18 outcome (rather than treatment progress). These investigators found that both clients? and therapists? early ratings of RR predicted pre to post symptom change above and beyond their ratings of WA and adult attachment, although therapist ratings were a stronger predictor than client ratings of RR. Although the studies by Marmarosh et al., (2009) and Fuertes et al. (2007) added some support regarding the relationship between early ratings of RR and treatment outcome (symptom change over the course of treatment), they did not rule out whether session ratings occurring in other parts of the treatment predicted outcome at all or in significantly different ways (e.g. late ratings negatively correlate or more strongly positively correlate to outcome than early). In an Italian dyad study by LoCoco et al. (2011), the researchers extended the tests of the previous studies by measuring clients and therapists? ratings of RR after the third and eighth session of treatment and testing them in relation to treatment outcome. The investigators found that client ratings of RR significantly related to treatment outcome but did not find the same connection between outcome and therapists? ratings of RR, or either participants? ratings of their working alliance. They also found that clients? ratings of RR later in treatment (8th session) more strongly predicted outcome than client ratings early in treatment (3rd session). Though the findings only occurred with client ratings of RR, this study?s data provided support that ratings taken at different points in treatment differentially relate to treatment outcome. These findings may indicate that ratings of the real relationship vary in important ways in more or less successful treatments. In order to detect such patterns the phenomena needs to be sampled with more frequency over the course of treatment. DYAD STUDY OF PSYCHOTHERAPY PROCESS 19 Two studies, one preliminary unpublished study (Spiegel et al., 2008) presented at a national conference and another study by Fuertes et al., (under review), gathered data from clients and therapists who rated the real relationship, outcome and other variables after every session. The studies provided some provisional yet promising results regarding how the real relationship may manifest over the course of treatment. Spiegel et al. examined 28 dyads of psychotherapy and found that overall client ratings of real relationship (RRI-C) after all sessions were significantly related to treatment outcome. However, they found no relationship between early ratings of RR and outcome from either the therapist or client perspectives. Clients who fared best in therapy (moderate/high-outcome group) started treatment with a strong real relationship that became stronger by the end of therapy. In contrast, clients who experienced less successful therapy (low-outcome group) started with a strong real relationship that weakened by the end of treatment. It is important to note that these patterns emerged only in the RRI-C ratings (from the client perspective). Fuertes et al. (under review), examined six brief psychotherapy dyads and found that overall, both the clients? and therapists? ratings of the real relationship were strong from the beginning of treatment and increased in strength across all four quarters of treatment. However, when treatment outcome was taken into consideration, in the more successful dyads, increases in the clients? ratings from quarter to quarter did not exceed the reliable change index (RCI), while increases in the therapists? ratings in the first, third and fourth quarters of therapy did exceed the RCI. Thus, in more successful dyads, client ratings were not clinically significant, but therapist ratings of RR were clinically significant. In less successful dyads neither the clients? nor the therapists? ratings DYAD STUDY OF PSYCHOTHERAPY PROCESS 20 exceeded the RCI in any quarter. These findings suggest that how the real relationship unfolds over the course of treatment relates to outcome. Interestingly, these initial and preliminary findings indicate that for successful cases, there are no declines in the strength of RR between client and therapist. Unlike the working alliance and the impact of alliance ruptures (Kivlighan and Shaugnessy, 2000; Stiles et al., 2004), successful therapy dyads begin with a strong real relationship that strengthens linearly over time. Although the lack of inferential statistical analysis of the patterns in the Fuertes et al. (under review) data limits any concrete conclusions, the combination of both studies? results do provide grounds for examining the overall pattern of RR session ratings after every session over the course of psychotherapy treatment. The connection of the pattern of RR ratings to outcome may provide important insights regarding how the strength of the real relationship develops in more or less successful psychotherapy. A recent study by Gelso, Kivlighan, Busa-Knepp et al., (2012) followed-up on these preliminary investigations and examined how the real relationship unfolded in relation to outcome in 42 dyads. The findings indicated that client and therapist ratings of RR related to outcome very differently. Client RR ratings after the first session, first quarter and after all sessions combined related to outcome, while therapist RR ratings at these points in time did not relate to outcome. Instead, from the therapist perspective, increasing convergence with clients? ratings of RR as well as increases over time in therapists? ratings of RR strength related to outcome. Taken all together, the emerging real relationship research findings indicate that ratings of the real relationship sampled from various points of treatment and rater DYAD STUDY OF PSYCHOTHERAPY PROCESS 21 perspective (e.g. therapist or client) relate to other components of the therapeutic relationship, relevant therapist and client factors, session evaluation, treatment progress and outcome in ways that are theoretically consistent. These findings also indicate it is important to test which real relationship rating vantage point best relates to treatment process and outcome when therapy participants are sampled after every session in treatment. Therapist Immediacy Psychotherapy theorists and researchers have proposed that self-revealing techniques such as therapist immediacy bring the real relationship into the foreground, and if done appropriately, strengthen the real relationship (Bordin 1979, 1994; Crits- Christoph and Gibbons 2002; Gelso, 2011; Hill & Knox, 2009; Kiesler, 1996; Teyber and McClure, 2011; Safran et al., 2002). As such, the current study will be examining the connection between the real relationship and therapist immediacy. The next section will provide a discussion of therapist immediacy, including its theoretical definition and empirical history, clinical and training implications, and finally its connection to the real relationship. Definition Therapist immediacy is a term that refers to a type of therapy intervention that serves to process the relationship between client and therapist in the here-and-now. The dynamic nature of this therapist-initiated exchange makes a clear definition difficult because the use of therapist immediacy is often situation specific and often used on an ?as needed? basis (depending on the therapist?s theoretical orientation). The present study will examine therapist immediacy using Hill?s (2009) definition of immediacy, as DYAD STUDY OF PSYCHOTHERAPY PROCESS 22 her conceptualization includes a clear account of form, function, and applies to a wide range of theoretical approaches and client problems. Hill defines therapist immediacy as disclosures within the therapy session of how the therapist is feeling about the client, about him/herself in relation to the client, or about the therapeutic relationship. Not directly stated in its definition, but indicated in Hill?s descriptions of some of its subtypes, immediacy also refers to the therapist?s explicit attempts to get the client to engage in similar disclosures about how they are feeling about the therapist, about him/herself in relation to the therapist, or about the therapeutic relationship. Thus, the aim of therapist immediacy is to initiate processing the relationship, through both the client?s and therapist?s here-and-now experience of each other in the session. In all, Hill?s definition provides a heuristic framework, that can add clarity to both psychotherapy theory and research. Form and Function of Therapist Immediacy Although it overlaps with many other skills, Hill (2009) highlights therapist immediacy as a separate skill because of its effectiveness in facilitating client exploration and insight, especially regarding interpersonal patterns and behaviors. For example, consider the following example of therapist immediacy. ?I am feeling bored and frustrated right now that despite how much you are talking, we are spinning our wheels. What do you sense is going on between us?? If other (less advanced) counseling skills were considered, the therapist speaking turn above could be categorized as a sequence consisting of a self-disclosure, followed by an interpretation, and ending with an open question. However, this description does not capture the important relational process unfolding between the therapist and client. As will be further discussed next, the DYAD STUDY OF PSYCHOTHERAPY PROCESS 23 subtypes of therapist immediacy illustrate important nuances in its possible forms and the various ways therapist immediacy functions to process the relationship. Subtypes of Therapist Immediacy Based on previous empirical work (Hill, Sim et al. 2008; Kasper, Hill et al. 2008) and interpersonal theory (Teyber and McClure 2011; Safran, Muran et al. 2002), Hill (2009) suggested four subtypes of therapist immediacy: inquiries about the relationship, statements of therapists? reactions to clients, making the covert overt, and drawing parallels with outside relationships. Inquiries about the relationship occur when the therapist invites the client to share in-the-moment perceptions and reactions about the therapeutic relationship as in the example above (e.g. ?What do you sense is going on between us??). Frequently these queries involve the therapist gently probing the client about feelings that are currently being sensed but have yet to fully emerge in the client?s verbal content. With the second subtype of immediacy, statement of the therapist?s reaction to the client, the therapist is expressing his or her here-and-now reactions to what the client has just done or said which reveals the impact of the client on the therapist. For example, ?As I listen to you describe this, I find myself feeling anxious, too.? The third subtype of immediacy occurs when the therapist makes the covert overt. In this subtype, the therapist is attempting to reveal any unspoken interpersonal feelings, conflicts, or dynamics as they are occurring in the moment between the client, herself or himself. Making the covert overt involves tentatively offered suggestions based on the therapist?s immediate observations, as an attempt to acknowledge what may be happening between both therapy participants. For example, a therapist might say, ?You DYAD STUDY OF PSYCHOTHERAPY PROCESS 24 were unusually late today, and you seem to be looking at your coat more than making eye-contact with me. I wonder if you are uncomfortable being here with me right now.? The fourth immediacy subtype that Hill (2009) describes occurs when the therapist draws parallels with outside relationships. The therapist, again, tentatively and sensitively wonders out loud whether the client?s current reactions to him or her are similar to those the client has to other people. An example would be, ?You said you don?t let yourself get attached because everyone has let you down. I wonder if that is keeping you disconnected from me right now.? The therapist immediacy subtypes and examples above highlight three essential aspects of therapist immediacy that comprise Hill?s (2009) definition and augment what other theorists have theorized about immediacy. First, during an immediacy event, both participants of the relationship contribute their reactions to the discussion (Crits- Christoph and Gibbons 2002; Hill 2004, 2009; Kiesler 1996; Teyber and McClure, 2011). In fact when discussing how to use immediacy, Hill (2009) encouraged therapists to use ?I? statements to take responsibility for their feelings and to acknowledge their role in relationship problems. Such behaviors demonstrate the therapist?s willingness to explore and work through potential misunderstandings, inaccurate perceptions, or other interpersonal problems, which, if successful, can strengthen the therapeutic relationship and move clients closer to their key concerns and therapeutic gains (Bordin 1979, 1994; Hill and Knox 2009; Kiesler 1996; Teyber and McClure 2011). A second important feature of therapist immediacy, is its emphasis on the therapist revealing to the client his or her ?here-and-now? experience (i.e. perceptions and reactions) of the client. Teyber and McClure (2011) described immediacy as a self- DYAD STUDY OF PSYCHOTHERAPY PROCESS 25 involving statement as compared to a self-disclosing statement that refers to the therapist?s own past or personal experiences. Consider the following two examples that illustrate a self-disclosing statement and a self-involving statement (i.e. therapist immediacy), respectively. ?I struggle with anxiety too, sometimes.? Versus: ?As I listen to you describe this, I find myself feeling anxious, too.? As the examples illustrates, in contrast to self-disclosing statements, therapists? statements of their reactions to the client (i.e. self-involving statements), keep the focus on the client and expose how the client?s current words or behaviors are impacting the therapist in the moment. Furthermore, interpersonal theorists suggest that when therapists effectively share personal reactions to what their clients have just expressed or done in the immediate moment, therapists convey personal involvement and emotional resonance with clients (Crits-Christoph and Gibbons 2002; Hill 2004; Kiesler 1996; Teyber and McClure 2011). The third defining feature of therapist immediacy is that it brings attention to the process dimensions of the communication in the therapist-client interaction (Hill 2009; Keisler 1988; Teyber and McClure 2011; Yalom 2005). Accordingly, the therapist facilitates a perceptual shift away from the content of what is being explicitly discussed, and begins to collaboratively focus on the relational process of how she/he and the client are interacting. Theoretical Context of Immediacy Therapist immediacy is a specific intervention that belongs to a larger group of diverse psychotherapy approaches and strategies that function to process the therapeutic relationship (Hill and Knox, 2009). Key relational features are prominent across several major, but diverse theoretical perspectives. These interpersonal elements consist of DYAD STUDY OF PSYCHOTHERAPY PROCESS 26 strategies that enable therapists to use their current interaction with their client to strengthen the therapeutic relationship and help clients grow. For example, in classic psychoanalysis, therapeutic work is facilitated by therapist neutrality and transference interpretations that function to process the relationship and help the client gain therapeutic insight into the origins of their displaced interpersonal patterns (Hill and Knox, 2009). While therapists working from an interpersonal theoretical approach use the therapeutic relationship to attempt to provide clients with a corrective relational experience (also referred to as a corrective emotional experience in some literature). It is important to note that theoretical approaches vary in two ways related to processing the relationship: 1) how central repairing or strengthening the relationship is to the therapeutic work, and 2) the amount of focus given to the therapist?s contribution to the relationship (Hill and Knox, 2009). However, these differences do not negate the need for and effectiveness of processing the therapeutic relationship to enhance diverse forms of treatment (Hill and Knox, 2009). Rationale for Using Immediacy Processing the therapeutic relationship has been argued to be an essential mechanism of change for some clients across a diverse array of psychotherapy approaches and client problems (Hill and Knox 2009). The basis of this claim is that the therapeutic relationship gets built and strengthened through a ?tear and repair? process (Bordin 1979, 1994; Safran and Muran 2002; Hill and Knox 2009). In other words, throughout the course of therapy the client and therapist will experience inevitable conflicts or problems in their relationship (i.e. ?tears?). If the therapist and client directly address their feelings about each other and these problems as they occur in the moment DYAD STUDY OF PSYCHOTHERAPY PROCESS 27 (i.e. processing the relationship, therapist immediacy, relational work, etc.), the therapy relationship will be enhanced and the client will experience positive therapeutic growth. As will be further described below, preliminary research investigating immediacy and interventions similar to immediacy that serve to process the relationship, provide evidence that it enhances client outcome. Research on Therapist Immediacy Empirical investigation of immediacy is in its early stages, with the majority of the research providing rich descriptive information through qualitative designs (Kasper et al. 2008; Hill et al. 2008; Mayotte-Blum et al., 2012; Rhodes, Hill et al. 1994; Hill et al. 2003). Several studies provide evidence that therapist immediacy is an effective intervention that is related to positive client outcomes and enhanced therapeutic relationships. First, several studies provide findings that indicate that behaviors that comprise the current study?s definition of therapist immediacy are related to enhanced therapy relationships and improved client outcome. Foreman and Marmar (1985) examined therapist actions over the course of treatment for six cases of brief psychotherapy with bereaved clients that began therapy with poor therapeutic alliances. Upon comparison of the cases at termination, the therapists? actions most similar to the current definition of therapist immediacy distinguished the cases with improved alliances and positive outcomes from the unimproved-alliance group of cases that had less positive outcomes. However, one of the limitations of this study was its use of only one judge to review clinical videos and determine which therapist actions were present and associated to improvement or decline. DYAD STUDY OF PSYCHOTHERAPY PROCESS 28 Several in-depth qualitative examinations of difficult events or instances in therapy have also provided evidence that behaviors that are key components of therapist immediacy improve therapy outcomes for clients. In a study by Rhodes, Hill, Thompson and Elliot (1994), instances in which clients felt misunderstood by their therapist were investigated. The researchers found that when the misunderstanding was resolved, clients recalled therapist behaviors that were much more client-responsive and immediate than the therapist behavior recalled by clients with unresolved misunderstandings. For example, clients reported that they were able to assert their negative feelings or lack of satisfaction with the therapist, experienced collaboration with the therapist to make sense of the misunderstanding, and experienced an enhanced relationship with the therapist as a result. In sharp contrast, clients with unresolved events did not report good relationships, nor did a large number of the clients report asserting their negative feelings or reactions to the therapists. Of the few unresolved cases that reported disclosing their dissatisfaction, they described their therapist as unresponsive and not open to the client?s point of view and feelings. The generalizability of this study is limited because it was retrospective in nature and only looked at single events from the client?s perspective. Hill, Kellems et al. (2003) investigated problematic events in therapy from the perspective of the therapist by qualitatively examining 13 therapists? recollections of being the target of a client?s anger (hostile or unasserted, passive). The results of the study indicated that resolution was associated to several elements of therapist immediacy. Therapist contributions associated to resolution of hostile anger events included turning the negative feelings outward, having a goal of connecting with clients, therapists exploring the anger with clients and explaining their behaviors, and conceptualizing DYAD STUDY OF PSYCHOTHERAPY PROCESS 29 anger as a problem of the therapy relationship rather than the client?s personality problems. For unasserted anger events, therapist contributions included attempts to help the client gain insight and explore the anger, and a strong therapy relationship. Safran, Muran, Samstag, and Stevens (2002), reviewed 10 years of research on alliance ruptures, and found evidence for four stages of rupture repairs that reflected components of therapist immediacy and its importance in resolving problems in the therapy relationship. Therapist behaviors in the four stages that denoted immediacy included: encouraging the client to express negative feelings toward the client, accepting responsibility for his or her contribution to the interaction, probing for any fears that might be blocking the client?s negative expression of feelings toward the therapist, and encouraging the client to express the underlying wish/need and primary emotion associated with that need. A comparison of the results of this research on the resolution of difficult events within therapy and the (Foreman and Marmar, 1985) research examining the actions of therapists with clients that began therapy with poor alliance but had positive outcomes indicates that the following therapist behaviors strengthens the therapeutic relationship and enhances client outcomes: (1) encouraging clients to talk about immediate, negative feelings toward their therapist, (2) acknowledging and exploring the difficult event and collaboratively making meaning of it, and (3) focusing on the interaction between client and therapist as opposed to only the client?s personality dynamics, which also includes the therapist?s contribution. In response to the retrospective nature and single event methodology of several of the studies, two case studies by Hill, Sim et al. (2008) and Kasper et al. (2008) examined DYAD STUDY OF PSYCHOTHERAPY PROCESS 30 therapist immediacy in depth as it occurred over the course of ongoing brief interpersonal therapy. Comparison of the two case studies revealed rich information about therapist immediacy including potential markers for when it might or might not be therapeutic to use it and areas for further investigation. In both cases, both clients were non-White females paired with older, gentle, White male therapists that completed short-term psychotherapy (12 and 17 sessions). Despite superficial similarities in the two cases regarding gender and minority status, the cases exhibited important differences regarding the ways immediacy was experienced and expressed, presenting client issues, client symptom severity and vulnerability, effects of immediacy, and outcome of treatment. In the first case (Kasper, et al. 2008) the client, ?Lily,? was a 24-year-old female, described by the authors as a well-functioning and well-defended graduate student in mental health that had volunteered to participate in a research project involving therapy for problems in interpersonal relationships. Lilly had few symptoms of distress (OQ; d = 1.10) and had a high level of self-understanding (SUIP-R; d = .76), but was not functioning well interpersonally (IIP-32; d = .64). Throughout treatment, her therapist, Dr. N, used immediacy in about a third of his speaking turns (34%; M = .34, SD = .12) each session with Lily. The most frequent subtypes of immediacy Dr. N used were drawing parallels from outside relationships to the therapy relationship and encouraging Lily to express immediate feelings to him. His efforts were aimed at confronting and challenging her defenses, and encouraging her to deal with the therapeutic relationship and live more in the moment. The outcome of this case was mixed. Post treatment, Lily worsened in terms of symptomatology and interpersonal functioning but improved in terms of self-understanding (OQ, IIP, and SUIP-R). At the same time, Lily and Dr. N DYAD STUDY OF PSYCHOTHERAPY PROCESS 31 noted that she appeared to ?soften? and open up in her manner of relating throughout the course of treatment so that it was easier to connect to her. On one hand, therapist immediacy helped Lily to open up, express feelings she normally restricted herself from, feel closer to Dr. N, feel cared for and feel satisfied with the session. On the other hand, immediacy sometimes made Lily feel uncomfortable, pressured to respond, vulnerable, challenged and hurt. In the second case by Hill, Sim et al. (2008), the client, ?Jo,? was a 29-year-old female. The authors reported that Jo had survived a traumatic abuse history, presented severe symptoms of distressing personal issues at intake, and exhibited a high level of vulnerability in her current functioning. Jo?s therapist, Dr. W, used immediacy in only 12 % of his total speaking turns each session even though he was much more verbally active and directive than Dr. N, the therapist in the Kasper, Hill et al. case (44%. vs. 13% total words in the sessions). Also in contrast to Dr. N?s challenging and confrontational type of immediacy, the subtypes Dr. W used most frequently were supportive in nature in an effort to help stabilize the client. They included reinforcing the client for in-session behavior, encouraging the client to collaborate, inquiring about the client?s reactions to therapy, reminding the client it was okay to agree with him, and indicating pleasure at seeing the client. At the end of therapy, Jo exhibited improvement on all the outcome measures (OQ, IIP, and SUIP-R) and showed observable and dramatic improvements in behavior, optimism, autonomy and interpersonal relationship skills. Across the two cases, therapist immediacy exhibited similar positive effects despite noticeably different therapist factors (e.g.. usage frequency, usage style, etc.) and client factors (e.g. presenting issue, symptom severity, amount of defense or DYAD STUDY OF PSYCHOTHERAPY PROCESS 32 vulnerability, etc.). In both cases immediacy led to the clients experiencing and expressing their feelings, helped the dyads effectively negotiate the rules of their relationships, and enabled the clients to have corrective relational experiences. These effects provide empirical support for the four functions theorized for therapist immediacy in the literature (e.g. Cashdan, 1988; Hill, Sim et al. 2008; Hill and Knox 2009; Ivey, 1994; Kasper, Hill et al. 2008; Kiesler, 1988, 1996; Safran and Muran 2001; Teyber and McClure 2011; Yalom, 1995). However, the differences in the two cases, especially regarding treatment outcome (i.e. mixed outcome vs. fully positive outcome), suggests that effective use of the subtypes of immediacy depends on the therapist and client as well as what is going on in therapy. A third case study (Mayotte-Blum et al., 2012) was more recently conducted to examine therapeutic immediacy over the course of long-term psychodynamic psychotherapy. The investigators found similar themes to those that emerged in the previous case studies by Hill et al. (2008) and Kasper et al. (2008). These themes included an increased ability in the client to tolerate and explore deeply painful and shameful feelings, the experience of a new relational experience for the client, and for both therapy participants, an increased ability to communicate care, concern and general positive feelings towards one another. However it is important to note that this study differed from the two previous case studies of immediacy in that the long-term treatment was completed before any of the case studies had begun and the case was selected on the basis of its interpersonal components. Furthermore, the length of the case allowed for more process and outcome phenomena to be gathered and assessed, though only certain portions (relating to high quality immediacy events) of the overall treatment were DYAD STUDY OF PSYCHOTHERAPY PROCESS 33 studied. Thus the generalizability of the study?s findings regarding immediacy should be regarded cautiously. Overall, the findings of these three case studies and the subsequent studies on the components of therapist immediacy indicate that therapist immediacy is not formulaic and suggests important implications for clinical practice and future research. First, therapist immediacy appears to effectively increase the authenticity and relevancy of the therapist?s and client?s interaction. This suggests that the use of immediacy requires that therapists have the interpersonal range to modify their interventions so they can provide responses that clients find most useful. It also may require that therapists be receptive enough to detect and stay in touch with clients? immediate responses, in addition to their own intentions behind each response or turn in conversation. In terms of future research, due to the retrospective nature and single event or single-case methodology of several of the studies, the findings need to be replicated, as the generalizability of the findings is limited. Additionally, more research is needed to examine the amount and timing of therapist immediacy throughout the course of treatment (i.e. overall use and frequency) and its relationship with other relevant process variables such as the real relationship. Therapist Immediacy and the Real Relationship Therapist immediacy seems highly relevant to real relationship because it 1) initiates a self-revealing exchange about the therapist and the therapist-client interaction and is thus likely to affect perceptions of realism by reducing distortion and 2) is likely experienced by the client as an invitation for both people to be more authentic with less facade, defense, and repression, thus affecting perceptions of genuineness (Hill, 2009; DYAD STUDY OF PSYCHOTHERAPY PROCESS 34 Kasper, et al. 2008; Mayotte-Blum). Furthermore, as discussed above, when a time arises in which the real relationship is needed for the client to move forward, such as crises in which the client needs support, ruptures in a part of the relationship, or when challenging transference feelings surface, therapist immediacy can be used to increase the strength of the real relationship and enhance the client?s therapeutic gain (Gelso, 2011; Hill, 2009; Hill and Knox, 2009; Safran and Muran, 2001). The findings in the three in-depth case studies discussed above (Hill, Sim, et al. 2008; Kasper, Hill & Kivlighan, 2008; Mayotte-Blum), provide support of a connection between the real relationship and therapist immediacy. Despite differences in the three cases, both clients felt that immediacy helped them to express immediate feelings and have a different (corrective) interpersonal experience than what they had experienced in other relationships. Particularly, all three clients expressed that their therapist?s immediacy had facilitated being able to interact in ways that were more authentic and genuine, despite inhibitions resulting from past abandonment or rejection in other significant relationships. For two clients it was being ?fragile and vulnerable? and experiencing feelings that had been actively avoided, and for the other client it was expressing disagreement or being confrontational. Based on these findings, it is likely that therapist immediacy and the real relationship are strongly connected as the clients expressed changes in their experiences of genuineness and realism. To date, no study has been completed examining the association between the real relationship and therapist immediacy. Thus, the connection between the two variables is important to study. DYAD STUDY OF PSYCHOTHERAPY PROCESS 35 Client Experiencing Level The personal relationship between the therapist and client is theorized to be inherently involved in the experiencing level of the client (Gendlin 1961, 1968). Furthermore the real relationship, due to its features of genuineness and realism, is also theoretically linked to client experiencing level (Gelso, 2011). However, no empirical investigation of a relationship between the two constructs has been done. Thus, the current study will be investigating the relationship between the real relationship and client experiencing level. The next section will discuss client experiencing level including its theoretical definition and background, a brief empirical history, and its connection to the real relationship. It is important to note that in the current study the terms ?client experiencing level? and ?experiencing level? as seen in the theoretical and empirical literature that will be cited, are synonymous. History and Context of Client Experiencing Level The concept of Experiencing Level and its accompanying Experiencing Scale originated from Gendlin?s philosophical ?explication model? of how words function in relation to experiencing (Gendlin 1962). This direction was inspired by the innovative research being done at the time using audio recordings and transcripts of therapy sessions. Most of the research during that time period regarding the effectiveness of psychotherapy was examining ?what? was being disclosed by clients -- the content of their verbalizations. Gendlin?s explication model refocused the investigative lens on the ?how? of client verbalizations -- the manner in which content was being discussed -- as opposed to the ?what? of client-therapist dialogue. DYAD STUDY OF PSYCHOTHERAPY PROCESS 36 Gendlin was inspired by the work of Carl Rogers, and was interested in formulating what was occurring when a therapeutic response was described as effective in therapy. For example, in a client-centered response such as a reflection, Gendlin believed that the therapist response was more than mere repetition, which adds nothing to what the client says. He theorized that any therapist response that was effective in therapy caused essential therapeutic change processes to develop in the client. Accordingly, an effective client-centered response referred to and facilitated the client?s immediate ?felt experiencing?. He believed this true of any therapist response regardless of theoretical orientation. Thus, Gendlin?s aims were integrative and his concepts pan- theoretical. Gendlin?s emphasis on process dimensions made it possible to look at the manner in which any content is being discussed. As such, his paradigm enabled more fundamental questions and hypotheses regarding personality change in psychotherapy to be investigated across diverse theoretical orientations. Basic Tenets of Experiencing Though the current study is using the construct experiencing level, it is important to explain the experiencing variable that is a component of it. Experiencing is theorized as an essential intrapersonal process that can be facilitated by certain kinds of interpersonal interactions (especially those in therapy) and that ultimately lead to therapeutic change and personal growth (Gendlin 1961). Two essential features that are central to Gendlin?s theoretical formulation accounting for how experiencing is an essential change process in psychotherapy are: (1) immediate experiencing centered within the body, and (2) the interpersonal context (of the bodily-felt change process). DYAD STUDY OF PSYCHOTHERAPY PROCESS 37 These features are key elements in the following theoretical propositions regarding experiencing and therapeutic change. First, experiencing is an ongoing flow of events that are felt in the body, as opposed to being thought, intellectually understood, or verbalized (Gendlin, 1961). Gendlin explained that this type of ?felt meaning? occurs because a person ?is? a bodily interaction with others (and the environment) much like breathing is an interaction with the cardiorespiratory system and air in the environment. Thus how a person lives, reacts, perceives, and behaves is a bodily process occurring in situations and felt in the body (Gendlin, 1968). Secondly, in contrast to generalized aspects of an individual such as traits or dispositions, experiencing is the immediate changing flow of feeling which enables every individual to feel something in any given moment. Next, while experiencing itself is subjective, private, and unable to be externally observed, a client can refer directly to his or her experiencing in ways that are observable by others. Gestures, tone of voice, or manner of expression are often observable indicators that a client is referring directly to the felt data of their experiencing. This pointing to one?s immediate experiencing is called ?direct reference? (Gendlin, 1961). Lastly, the client uses experiencing to guide her- or himself toward increasingly accurate self- understanding and growth -- a process he called ?carrying forward one?s implicit felt shift? (Gendlin, 1961). Considering these propositions together, Gendlin (1961, 1968) asserted that for therapeutic change to be achieved in session for the client, the therapeutic work must include direct reference to the client?s immediate experiencing. It is important to clarify that direct reference is not mere inward attention to any kind of experience or process. DYAD STUDY OF PSYCHOTHERAPY PROCESS 38 Direct reference is a mode of experiencing that is distinct from the emotions a client might feel about an aspect of his or her experience. Direct reference connects a person to the fuller and more complex experience that underlies the emotion. Similarly direct reference is not what Gendlin (1968) describes as an inward recitation of circumstances (e.g. what was done or should have been done or said) or circumstantial explanations and conceptualizations (e.g. ?I am trying to make up for the lack of relationship with my dad?). Such mental replay and re-enactment lacks indications that a person is connecting to bodily-felt experience occurring in the immediate moment. Definition Client experiencing level indicates the quality or manner of a client?s congruence between their immediate, physically-felt, inner experience and their verbal expression (Gendlin, 1962, 1984, 1996; Hendricks 2009; Klein, Mathieu, Kiesler, & Gendlin, 1969). In other words, it assesses the manner in which what a client says relates to their bodily- felt sense. Even though a person?s experiencing is subjective and dynamic, the concept of (client) experiencing level provides a systematic way of accounting for and observing the client?s experiencing process as it emerges and evolves over the course of therapy. Each manner of process, categorized into levels, has precise linguistic and somatic characteristics that can be observed. As a result, the client?s first person process can be researched with specificity and precision. Measuring Experiencing Level The Experiencing Level Scale, developed by Klein, Mathieu, Kiesler, and Gendlin, emerged in 1969 as a reliable measure of the manner in which what a client says relates to their subjectively felt experience (i.e. bodily-felt sense). There are 7 levels or DYAD STUDY OF PSYCHOTHERAPY PROCESS 39 stages of the scale described below. The authors theorize that progressing through the levels of the scale reflect greater elaboration and integration of emotions and experience, which consequently results in resolution of client problems. Stage 1. At this level, the content of the client?s verbalizations and manner is impersonal, abstract and externalized. There is no explicitly expressed association between the speaker?s content and the speaker, which is reflected by a communication style that has an absence of personal involvement and an avoidance of feelings. The narrative is expressed in a manner in which it could belong to any person since nothing unique, or personal is expressed. Stage 2. When this level is reached by a client, the association between the speaker and his or her content becomes explicit, but only to the extent that it serves to convey the speaker?s narrative or idea (as opposed to a feeling or inner experience). The speaker?s manner of communication is externally descriptive with any emotional involvement circumscribed only to the specific situation or content (versus inner parallels to one?s self across situations). The speaker?s feelings are thus implied but never explicitly expressed. Stage 3. The speaker begins to add explicit comments of feelings or emotional reactions to his or her narrative content at this level. However, the speaker?s self- description is limited to circumscribed behavioral terms and the personal remarks about his or her private experience are parenthetical to the speaker?s communication. Stage 4. When client is at this level, the quality of involvement in speech content reflects a shift in the speaker?s attention to the subjective felt flow of his or her experience rather than to events or abstractions. Rather than objective and analytical, the DYAD STUDY OF PSYCHOTHERAPY PROCESS 40 speaker?s style is subjective, descriptive, and emotional. The speaker?s verbal effort centers around expanding and elaborating the details of her or his inner experience. Stage 5. Using their internal elaboration of emotions and personal meanings, the speaker constructs a problem or question about the self when they reach this level. Stage 6. At this level, new feelings and meanings emerge in the client from ongoing explorations to resolve emotional problems related to the self. Stage 7. At this highest level, the client goes through the process of expanding awareness of immediately present feelings and internal processes more consistently and steadily. Thus the speaker immediately links and integrates felt nuances of experience when and as it occurs in the present moment. The last three levels reflect a more distinct qualitative shift from subsequent levels. At levels four or higher, the client attends directly to a bodily-felt sense of the situation and allows words to emerge directly from that sense. Individuals focus mainly on exploring their inner, personal felt meanings; utilize the present tense; pause and wait for words or images to arise; and use language metaphorically to point to what is sensed but not yet defined. The personal perspective becomes clearer until, at the highest levels, clients are actively processing their subjective experience in the moment towards therapeutic gain. Thus, scores of 4 and above are regarded as productive process. Research on Client Experiencing Level Lambert and Hill (1994) reported that experiencing level was one of the most researched process variables to date. This is likely due to its pan-theoretical nature and its emphasis on the process of productive therapeutic client change rather than its content (Rogers, 1950; Gendlin, 1958). In Orlinsky and Howard?s (1986) review of over 30 DYAD STUDY OF PSYCHOTHERAPY PROCESS 41 process and outcome studies, they reported that in-session emotional experiencing as measured by the Experiencing Scale (EXP; Klein, Mathieu, Kiesler, & Gendlin, 1969) was significantly related to positive therapy outcome. EXP has been shown to be related to treatment outcome in person-centered (Hill et al., 1988), experiential (Goldman et al., 2005), cognitive behavioral (Castonguay, Goldfried et al., 1996), and Rational-Emotive Therapy (Stalikas, Fitzpatrick 1995, 1996; Fitzpatrick, Peternelli, 1999) psychotherapy. In addition, EXP has been found to have a positive connection with diverse measures of outcome including creativity, ego strength and psychological differentiation (Hendricks 2009). A study by Lutgendorf et al. (1994) that used biophysical indices of treatment outcome with trauma patients disclosing traumatic events is an illustrative example of the value of the experiencing level construct and its diverse application. The results of the study revealed that greater experiential involvement during disclosure, as measured by EXP summed up across three disclosure sessions, was associated with increased immune function (measured by increased EBV-VCA antibody titres) over the course of the experiment. In contrast, disclosure alone did not affect the EBV-VCA antibody titres in any significant way. The authors concluded that it is the manner in which one engages in the expression of trauma that makes the difference in immune functioning rather than just talking or writing about the content of the trauma. In sum, research in a diverse array of areas in psychotherapy and mental health indicate that when it comes to therapeutic gain it is the manner in which clients experience their verbalizations (i.e. experiencing level) during treatment that has more influence than the content of those disclosures. Thus research on client experiencing DYAD STUDY OF PSYCHOTHERAPY PROCESS 42 level indicates that it is an essential psychotherapy process highly related to treatment progress and outcome. Furthermore, use of the experiencing scale appears to have the capacity to capture the associations between important psychotherapy processes and outcome in ways that help clarify the impact of various therapeutic processes on clients? progress and outcome. Client Experiencing Level and the Real Relationship As stated earlier, when a client manifests a high experiencing manner she or he attends directly to a bodily felt sense of the situation and allows words to emerge directly from that sense (Gendlin, 1996; Hendricks, 2002). Therefore, by definition it is likely that higher levels of client experiencing would be associated with a strong real relationship: as high genuineness reflects the tendency to be honest, open, authentic, and congruent in one?s inner experiencing, and realism reflects the lack of distortion (Gelso 2009a, 2009b, 2011). However, to the author?s knowledge, no empirical investigation has been done to test this relationship between the real relationship and client experiencing level. There have been some empirical examinations of experiencing level and variables related to the real relationship construct. In one study, Van der Veen (1967) found that clients with higher experiencing levels perceived their therapist as more congruent than clients with lower levels of experiencing. In another study by Castonguay, Goldfried et al. (1996), examining cognitive-behavioral improvement in 30 depressed patients, client?s emotional experiencing and working alliance were related to improved client outcome. As well, in a similar study by Goldman, Greenberg and Pos (2005), theme-related experiencing level and working alliance were found to be correlated to each other as well as significant predictors of change in experiential therapy with depressed clients. DYAD STUDY OF PSYCHOTHERAPY PROCESS 43 However, these studies only indirectly suggest a possible relationship between the real relationship and client experiencing level. Though the focus was on therapist immediacy, the two case studies mentioned earlier (Hill et al. 2008; Kasper, Hill & Kivlighan, 2008) provide more direct evidence of a connection between the real relationship and client experiencing level. The results of the studies indicated that the clients had positive relationships with their therapists and that the clients? abilities to express their immediate feelings had expanded. Both participants noted a positive personal salience to their relationship and an effortless inner intensity in their interaction that manifested with a non-defensive and open quality. Particularly in the Kasper et al. (2008) study, the client and therapist explicitly acknowledged the mutual, real relationship they had experienced. The client communicated to her therapist she experienced him as a gift from God. With similar fervor, the therapist expressed how moved he was by the non-defended, courageous and vulnerable manner of disclosure the client manifested throughout the treatment. The therapist reported to the researchers that he rarely felt the need to process their relationship because the client courageously engaged in intense work and was open and mutually collaborative with him in session. These observations by the therapy participants strongly suggest a relationship between the strength of the real relationship and client experiencing level. Overall Summary Overall, then, previous research and theory suggests there is a connection between the real relationship and treatment outcome and progress (e.g. Ain and Gelso, 2008, 2011; Fuertes et al., 2007, under review; Gelso et al., 2012; LoCoco et al., 2011; Marmarosh et DYAD STUDY OF PSYCHOTHERAPY PROCESS 44 al., 2009; Spiegel et al., 2008). Findings also hint at a connection of the real relationship and therapist immediacy (Gelso, 2011; Hill, Sim et al. 2008; Kasper, Hill et al. 2008; Mayotte-Blum et al., 2012) and the real relationship and client experiencing level (e.g. Hill et al., 2008; Kasper et al., 2008; Rogers, 1950; Gendlin, 1958). Despite these findings linking processes that promote or reflect authentic connection and self-revealing person-to-person interaction with enhanced therapeutic relationships and client outcome, no study to date has investigated the relationship between the real relationship and therapist immediacy and client experiencing level. DYAD STUDY OF PSYCHOTHERAPY PROCESS 45 Chapter 2 Statement of the Problem Over the past several years, recent empirical efforts have provided evidence that the strength of the real relationship in psychotherapy relates to treatment progress and outcome above and beyond other process variables (e.g. Ain & Gelso, 2008, 2011; Fuertes et al, 2007; Fuertes et al., under review; LoCoco et al., 2006; Gelso et al., 2012; Marmarosh et al., 2009; Spiegel et al., 2008). In order to build upon these studies, it is important to both corroborate these findings as well as to investigate therapist and client contributions that may relate to the strength of the real relationship. Due to sampling limitations, some of the real relationship investigations may not have sampled enough of each case to capture how the real relationship is theorized to vary over the course of therapy (Gelso, 2009; Gelso & Hayes 1998; Gelso & Carter, 1994; Greenson, 1967). Sampling more points across treatment may likely provide a better account of the unfolding of the real relationship and have important implications for its relationship to therapy outcome and other factors in the therapeutic process. Two variables that reflect separate client and therapist contributions to the psychotherapy process but that also seem to relate to the strength of the real relationship are client experiencing level (Klein et al., 1969) and the therapist intervention immediacy (Hill et al., 2008; Hill & Knox, 2002; Kasper et al., 2008). Theoretical and empirical examination also suggests that therapist immediacy and client experiencing may be related to treatment outcome as well as the real relationship (Castonguay et al. 1996; Gendlin, 1962; Goldman, Greenberg, & Pos, 2005; Hendricks, 2009; Hill et al., 1988, 1992, 2008; Kasper et al., 2008; Kiesler, 1971; Orlinsky & Howard, 1986; Rogers, 1959). DYAD STUDY OF PSYCHOTHERAPY PROCESS 46 However, since much of what is empirically known about the real relationship and outcome has stemmed from some investigations that only sample RR ratings at one or two points in treatment (e.g. early- or late-ratings, or both), more information is needed on the relation between therapist immediacy, client experiencing and the strength of the real relationship from a sample that better reflects the entire course of therapy (Lambert & Hill, 1994). In the present study I will examine the strength of the real relationship in 54 sessions of brief psychotherapy that occurred in 6 different psychotherapy dyads. I will examine how the client and therapist ratings of the strength of the real relationship over the course of treatment relate to treatment outcome, session quality, session frequency of therapist immediacy, and the level of client experiencing per session. Hypotheses A strong real relationship is also theorized to facilitate better treatment outcomes by enhancing factors that positively relate to the therapist?s and client?s functioning during sessions (Gelso, 2011). As such, it is likely that therapist and client evaluations of sessions (e.g. quality) relate to the real relationship. A strong real relationship is suggested to enhance the therapist?s experience of the session and increase therapeutic gain (Gelso, 2011). Previous findings regarding the real relationship and session evaluation indicate that there will be a significant association between client and therapist perspectives of the real relationship and session quality from both perspectives (Eugster & Wampold, 1996; Gelso et al, 2005; Kelley, LeBeouf-Davis, & Weiss, 2008; Kelley et al., 2009). In the present study, it is expected that these findings will be replicated. DYAD STUDY OF PSYCHOTHERAPY PROCESS 47 Hypothesis 1: There will be a positive correlation between the strength of the real relationship and session quality. The stronger the real relationship, the better the session quality. This relationship is expected individually for both client and therapist ratings of the real relationship and for both therapist and client rated session quality. As such four sub-hypotheses, 1a, 1b, 1c, and 1d, are as follows. Hypothesis 1a: The stronger the real relationship from the client?s perspective, the better the session quality from the therapist?s perspective. Hypothesis 1b: The stronger the real relationship from the client?s perspective, the better the session quality from the client?s perspective. Hypothesis 1c: The stronger the real relationship from the therapist?s perspective, the better the session quality from the therapist?s perspective. Hypothesis 1d: The stronger the real relationship from the therapist?s perspective, the better the session quality from the client?s perspective. Gelso (2011) theorized that when used at appropriate times, higher amounts of immediacy will strengthen the real relationship and improve treatment outcome. Recent empirical investigations show support of a connection between immediacy, the real relationship and treatment outcome. Three case studies (Hill et al., 2008; Kasper et al., 2008), rich in descriptive report from therapist and client participants, indicated that therapist immediacy was related to the therapeutic relationship and therapeutic outcome. In two related quantitative studies on therapist self-disclosure, a subtype of immediacy, Ain and Gelso (2008 and 2011) found significant relationships among self-disclosure, the real DYAD STUDY OF PSYCHOTHERAPY PROCESS 48 relationship, and treatment outcome. In the Ain and Gelso (2008) study, the correlation between amount of self-disclosure and the stregnth of the real relationship was .32 (p < .01). Therefore, on the basis of existing theory and research, I predict that the strength of the real relationship will be positively related to the amount of therapist immediacy. Hypothesis 2: There will be a positive relationship between the strength of the real relationship and the amount of therapist immediacy. The stronger the real relationship, the higher the amount of of therapist immediacy in appropriate relational contexts. This relationship is expected individually for both client and therapist ratings of the real relationship. As such, two sub- hypotheses, 3a and 3b, are as follows. Hypothesis 2a: The higher the amount of therapist immediacy, the stronger the real relationship from the client?s perspective. Hypothesis 2b: The higher the amount of therapist immediacy, the stronger the real relationship from the therapist?s perspective. Gelso (2011) also proposed that for self-disclosures and immediacy to have a healthy impact on the strength of the real relationship, immediacy usage should be well-timed, relevant, and particularly attuned to the client?s needs and dynamics, the dynamics of the therapeutic relationship, and the patient?s actual concerns. Recently, case- and event analyses of immediacy using a consensus rating method have effectively examined dimensions of immediacy that evaluate immediacy usage in terms of appropriateness, depth, quality, and resolution (Hill et al., 2008; Hill et al., under review Kasper et al., 2008; Mayotte-Blum et al., DYAD STUDY OF PSYCHOTHERAPY PROCESS 49 2012). These dimensions are consistent with the usage described in Gelso?s (2011) proposition mentioned above, though they have never been studied in relationship to the strength of the real relationship. However, recent research (Ain & Gelso, 2008; 2011) has examined self-disclosure (recall immediacy is a type of self-disclosure) and RR strength and the results of these studies provide some support that the strength of the real relationship strength is related to the relevance of self-disclosure. In the first study, relevance was significantly related to RR strength while amount of self-disclosure was not related. In the second study, amount was significantly related to RR strength while relevance was not related. Based Gelso?s theory and the Ain & Gelso (2008) findings regarding the relevance of self-disclosure and RR strength, it I propose that the dimensions of immediacy (i.e. depth, appropriateness, quality, and resolution) that capture how immediacy is used will also be related to the strength of the real relationship. Hypothesis 3: There will be a positive relationship between the strength of the real relationship and each of the therapist immediacy dimensions. The stronger the real relationship, the higher the therapist immediacy dimension will be rated. This relationship is expected individually for both client and therapist ratings of the real relationship and for each of the four dimensions of therapist immediacy (i.e. depth, appropriateness, resolution, and quality). As such, eight additional sub-hypotheses, 3a, 3b, 3c, 3d, 3e, 3f, 3g, and 3h are as follows. Hypothesis 3a: The higher the depth of therapist immediacy, the stronger the real relationship from the client?s perspective. DYAD STUDY OF PSYCHOTHERAPY PROCESS 50 Hypothesis 3b: The higher the depth of therapist immediacy, the stronger the real relationship from the therapist?s perspective. Hypothesis 3c: The higher the appropriateness of therapist immediacy, the stronger the real relationship from the client?s perspective. Hypothesis 3d: The higher the appropriateness of therapist immediacy, the stronger the real relationship from the therapist?s perspective. Hypothesis 3e: The higher the resolution of therapist immediacy, the stronger the real relationship from the client?s perspective. Hypothesis 3f: The higher the resolution of therapist immediacy, the stronger the real relationship from the therapist?s perspective. Hypothesis 3g: The higher the quality of therapist immediacy, the stronger the real relationship from the client?s perspective. Hypothesis 3h: The higher the quality of therapist immediacy, the stronger the real relationship from the therapist?s perspective. As persons increase their openness to their feelings and become more engaged in the process of their lived experience, their personality functioning will be altered in positive ways (Gendlin, 1962; Rogers, 1959). Gelso (2011) proposed that being in a relationship with a genuinely carring person and being accurately understood will tend to create sense of saftety in exploring vulnerable and risky feelings. Considering these theoretical assumptions regarding the impact of the real relationship, the client is likely to exhibit a high level of experiencing during a session in which the client and therapist perceive their relationship characterized by (a) the ability to express (and be) themselves as they DYAD STUDY OF PSYCHOTHERAPY PROCESS 51 truly are, (b) the experience of being perceived and accepted without distortion, and (c) positive feelings and liking toward one another. No empirical study has investigated the relationship between real relationship and client experiencing level. However, based on this theoretical literature, it is proposed that a strong real relationship is evidence of high congruence in both the client and the therapist. Therefore I predict that the strength of the real relationship will be positively related to client experiencing level. Hypothesis 4: There will be a positive relationship between the strength of the real relationship and client experiencing level. The stronger the real relationship, the higher the client experiencing level. This relationship is expected individually for both client and therapist ratings of the real relationship. As such two sub-hypotheses, 4a and 4b, are as follows. Hypothesis 4a: The higher the client experiencing level, the stronger the real relationship from the client?s perspective. Hypothesis 4b: The higher the client experiencing level, the stronger the real relationship from the therapist?s perspective. DYAD STUDY OF PSYCHOTHERAPY PROCESS 52 Chapter 3 Method Design and Data Set A quantitative study of six brief individual psychotherapy dyads was conducted in a nested time series design to examine the relationships (within sessions and across treatment) among real relationship, therapist immediacy, client experiencing, and therapy outcome. This was a correlational field study of psychotherapy dyads that were originally examined in a study by Fuertes et al. (under review). In order to examine the relationships among real relationship, therapist immediacy, client experiencing level, and session quality, I analyzed six cases of brief therapy conducted within one university counseling center and one university health center that provided counseling to students. Because I wanted to examine hypotheses regarding phenomena that happened at the speaking-turn level, session level, and dyad level we employed a nested time-series design. This enabled the investigation to follow six dyads in brief therapy from beginning to end and to obtain completed measures from the clients and therapist of the real relationship and quality of sessions after each session. Data were collected over a one-year period, though all of the cases only spanned one semester in duration (i.e. two cases conducted in a fall semester, and four cases conducted in a spring semester). Number of sessions per case ranged from 5 to 12 (M = 9.17, SD = 2.71). In total, information for 55 sessions of therapy was obtained. All of the sessions were audiotaped but due to tape degeneration only 50 sessions were transcribed in order for five judges to rate therapist immediacy and three judges to rate client experiencing level. Thus for analyses that involved post-session measures, the number of DYAD STUDY OF PSYCHOTHERAPY PROCESS 53 cases was 55. However, for the analyses that involved the judge-rated measures, the number of cases was limited to 50 since five sessions were unable to be transcribed. Participants Overview of Cases (see Table 1). Therapist A was a 28 year-old, female, advanced doctoral student (completing her internship), Chinese Asian American, who had 4 years of experience, and was currently working at a Northeastern university counseling center. She rated her theoretical orientation as moderately psychodynamic (3 on a 5-point scale), moderately cognitive-behavioral (3 on a 5-point scale), and moderately high in the humanistic/existential orientation (5 on a 5-point scale). Her clients were: Client A, a single, 19 year-old, Latina American, low income (less than $20,000 annual family income), female, with presenting concerns of relationship aggression with her mentally-ill boyfriend, academic performance and career development, and depression; and her Client B, a single, 20 year-old, Indonesian, Asian American, low income (less than $20,000 annual family income), female with presenting concerns of relationship issues with her father and step-mother, anxiety regarding her living arrangements and depression. Therapist B was a 26 year old, Chinese, Asian International, female, advanced doctoral student (completing her internship), who had 2 years of experience, and was currently working at a Northeastern university counseling center. She rated her theoretical orientation as moderately psychodynamic (3 on a 5-point scale), highly cognitive-behavioral (5 on a 5-point scale), and moderately humanistic/existential (3 on a 5-point scale). Her clients were: Client C, a married, 39 year-old, African American, middle-income ($50,000 to $80,000 annual family income), with presenting concerns of DYAD STUDY OF PSYCHOTHERAPY PROCESS 54 family-work life balance and career development; and her Client D, a single, 18 year-old, Indonesian-Chinese, Asian American, low to middle income (between $20,000 and $50,000 annual family income), male, with presenting concerns of writing procrastination, overall academic performance, and career development concerns. Therapist C was a 56 year-old, European American, female, social worker (LCSW), who had had 15 years of experience, and was currently working at a Mid- Atlantic university health center. She rated her theoretical orientation as moderate to highly psychodynamic (4 on a 5-point scale), moderately cognitive-behavioral (3 on a 5- point scale), and moderate to highly humanistic/existential (4 on a 5-point scale). Her clients were: Client E, a single, 20 year-old, European American, high-income (over $100,000 annual family income), female, with presenting concerns of disordered eating, depression, feelings of homesickness, and relationship concerns with men; and Client F, a single, 19 year-old, European American, high-income (over $100,000 annual family income), female, with presenting concerns of procrastination and school performance, career development issues, depression and relationship issues with her mother (See Table 1 below). DYAD STUDY OF PSYCHOTHERAPY PROCESS 55 Table 1. Therapists (TH) & Clients (CL) Descriptive Information. TH Age Gender Therapist level Race/Ethnicity Exp. Workplace Orientation CL A 28 Female Advanced doctoral student Chinese Asian American 4 years University counseling center 3 = PD 3 = CB 5 = Hm/Ex A (12 sessions) B (12 sessions) B 26 Female Advanced doctoral student Chinese Asian International 2 years University counseling center 3 = PD 5 = CB 3 = Hm/Ex C (5 sessions) D (10 sessions) C 56 Female LCSW European American 15 years University health center 3 = PD 5 = CB 3 = Hm/Ex E (8 sessions) F (8 sessions) CL Age Gender Race/Ethnicity Income Level Treatment Duration & Presenting Concerns TH A 19 Female Latina American Low (< $20,000) 12 sessions ? Relationship aggression with her mentally- ill boyfriend, academic performance and career development, and depression A B 20 Female Indonesian Asian American Low (< $20,000) 12 sessions ? Relationship issues with her father and step- mother, anxiety regarding her living arrangements and depression A C 39 Female (married) African American Middle ($50K ? $80K) 5 sessions ? Family-work life balance and career development B D 18 Male Indonesian- Chinese Asian American Low-middle ($20K ? $50K) 10 sessions ? Writing procrastination, overall academic performance, and career development concerns B E 20 Female European American High (>$100,000) 8 sessions ? Disordered eating, depression, feelings of homesickness, and relationship concerns with men C F 19 Female European American High (>$100,000) 8 sessions ? Procrastination, school performance, career development, relationship issues with her mother, and depression C DYAD STUDY OF PSYCHOTHERAPY PROCESS 56 Client Summary. The 6 clients (5 female, 1 male; 2 Euro-American, 1 African- American, 2 Asian, 1 Latino) averaged 23.8 (SD = 8.7) years of age and were all university students (being seen at either a Northeastern university counseling center or Mid-Atlantic university health center). Clients were selected on the basis of a single criterion, i.e., that the potential client was seen for at least 4 sessions of time-limited therapy. Presenting problems described during intake included disordered eating issues (1 client), relationship concerns (3 clients), anxiety and depression (4 clients), procrastination and performance in school concerns (2 clients), career development issues (4 clients), family-work balance concerns (1 client), and family of origin concerns (2 clients). Some clients described more than one presenting problem and no formal diagnoses were given. Therapist Summary. The six dyads consisted of three therapists whom each saw two clients. The three therapists (3 female; 2 Asian-American, 1 Euro-American) were two advanced doctoral students in clinical psychology (completing their doctoral internship) aged 26 and 28 years old and an experienced LCSW therapist age 56. They had 2, 4, and 15 years of experience practicing psychotherapy respectively (as opposed to post-degree experience). Using 5-point Likert scales ranging from low (1) to high (5) therapists answered the following question: ?Please rate the extent to which you believe in and adhere to the theory and techniques of the following therapies? and rated themselves on average as 3.33 (SD = 0.57) on psychoanalytic/dynamic theoretical orientation, 3.66 (SD = 0.57) on humanistic/existential theoretical orientation, and 3.66 (SD = 1.15) on cognitive/behavioral theoretical orientation. DYAD STUDY OF PSYCHOTHERAPY PROCESS 57 Judges. For the research team that rated therapist immediacy, judges included the first author who was an advanced doctoral student (female; African American) and 4 undergraduate upper-level psychology students (3 female, 1 male; 4 European American; 4 seniors). All four of the undergraduate judges had completed an upper-level course in basic counseling skills and research (Helping Skills) prior to the study and thus had didactic training specific to the research project (i.e. coding and rating speaking turns). It is important to note that in terms of biases regarding the research, the course (and didactic training each undergraduate judge completed) is typically taught from a multi-theoretical approach (Hill, 2009) including skills from psychodynamic, humanistic, interpersonal, and cognitive-behavioral orientations. In addition, the course taught immediacy based on the same definition (Hill, 2009) used in the current study. In individual interviews, all of the undergraduate judges expressed a value and interest in interpersonal aspects of therapy and immediacy during the interviews and thus were assigned to the coding team for therapist immediacy. Lastly, the first author (and fifth judge) had completed internship, described her orientation as psychodynamic and interpersonal, and regarding possible biases, valued immediacy, and expressed a moderately high level of comfort using immediacy. For the research team that rated client experiencing level (C-EXP), judges included the first author who was an advanced doctoral student (female; African American) and 2 undergraduate upper-level psychology and sociology students (2 female; 2 European American; 1 senior, 1 junior). Similar to the therapist immediacy judges, one of the undergraduate judges on this C-EXP rating team had completed the upper-level course in basic counseling skills and research (Helping Skills) prior to the DYAD STUDY OF PSYCHOTHERAPY PROCESS 58 study and thus had didactic training specific to the research project (i.e. coding and rating speaking turns). The second undergraduate judge had not taken the same class but had a year of prior research experience coding behavioral data in a psychology lab and thus had relevant training and experience for the project. The two undergraduate judges had no prior experience or knowledge of C-EXP, but during individual interviews expressed a larger interest in examining client behavior (as opposed to therapist behavior) during therapy and consequently were assigned to the team coding client experiencing level. Measures The Real Relationship Inventory-Client Form (RRI-C; Kelley et al., 2010; Appendix C). The RRI-C contains 24 items and the total score of the inventory was used as a measure of the strength of the real relationship from the client?s perspective. The RRI-CL consists of two 12-item subscales: Genuineness and Realism. Using a scale ranging from 1 (strongly disagree) to 6 (strongly agree), respondents rate items pertaining to the self, the therapist, and their relationship. Genuineness is the willingness and ability to be authentic, honest, and open ? in other words, to be who one truly is in the relationship. An example of an item from the RRI-CL that measures genuineness is ?I was able to be myself with my therapist.? Realistic perceptions are defined by the perceptions of the client or therapist that are not distorted by transference or other defenses. Realistic perceptions between the client and therapist enable them to view each other realistically (e.g., ?I was able to separate out my realistic perceptions of my therapist from my unrealistic perceptions?). Within the Genuineness and Realism subscales, items refer to the magnitude (how much) of the real relationship, as well as the valence (how positive or negative) of the real relationship. The authors assume that DYAD STUDY OF PSYCHOTHERAPY PROCESS 59 magnitude, the amount of genuineness and realism, can fluctuate over the course of therapy. Regarding valence, the authors assume that a client or therapist may genuinely like or dislike the other based on realistic perceptions and thus within the context of the real relationship, their feelings for one another may vary from positive to negative over the course of therapy. Higher scores on the RRI-CL reflect stronger real relationships as perceived by the client. In particular, high scores reflect higher perceived levels of genuineness and realism in the client and therapist, and greater perceived magnitude and positive valence. In Kelly et al.?s (2010) study, researchers obtained the following internal consistency alpha for the scale: .90 for Realism, .91 for Genuineness, and .94 for the total score. In addition, the measure also demonstrated strong test-retest reliability with coefficients of stability of .84 for Realism, .88 for Genuineness, and .87 for the total score. The authors also found support for the construct validity as the measure correlated in theoretically predicted ways to measures of the working alliance, observing ego, other- directedness, and therapist ratings of the real relationship. In addition, the RRI-CL did not correlate significantly with social desirability. Thus this study provided strong evidence of the RRI-CL scale?s reliability and validity. The Real Relationship Inventory?Therapist Form (RRI-T; Gelso et al., 2005; Appendix D). The RRI-T was used to measure the strength of the real relationship from the therapist?s perspective. Like the client version, RRI-T contains 24 items, comprised of two 12-item subscales: Genuineness and Realism; and the total score of the inventory is seen by the authors as a measure of the strength of the real relationship from the therapist?s perspective. As with the client version, respondents rate items pertaining to DYAD STUDY OF PSYCHOTHERAPY PROCESS 60 the self, the client, and their relationship using a scale ranging from 1 (strongly disagree) to 5 (strongly agree). An example of an item from the RRI-T that measures genuineness is ?My client and I were honest in our relationship.? An example of an item from the RRI-T that measures realism is ?My client was able to see me as a real person separate from my role as a therapist.? Magnitude and valence are also examined in the same manner as the client version. Thus, similarly to the RRI-C, higher scores on the RRI-T reflect stronger real relationships as perceived by the therapist. High RRI-T scores reflect therapist perceptions of the relationship as more real and genuine, with greater perceived magnitude and positive valence. Gelso et al. (2005) found the following internal consistency alphas for the scale: .89 for the total scale, .79 for the Realism subscale, and .83 for the Genuineness subscale. The RRI-T was also found by Gelso et al., to correlate in theoretically predicted ways with measures of the working alliance, session evaluation, client insight, and negative transference. Similar to the RRI-C, the RRI-T did not correlate significantly with social desirability. Session Quality (SQ: Appendix E). Clients and therapists rated the overall quality of the recently completed session on a 5-point scale (1=very poor and 5=very good) using one item: ?Using the scale above, please rate the overall quality of today?s session?. Session quality items such this have been used in previous process research (Bhatia & Gelso, 2013; Gelso, Hill, Kivlighan, 1991; Markin, Kivlighan, & Gelso, under review; Orlinsky & Howard, 1986; Elliot, 1986). Most recently, Bhatia and Gelso (2013) found a high correlation (estimated around .90) of this item with multi-item measures of session outcome (e.g., Session Evaluation Questionnaire ?Depth, Stiles & Snow, 1984). In addition, Markin et al., (under review) used a procedure developed by Muran et al. DYAD STUDY OF PSYCHOTHERAPY PROCESS 61 (1995) that is the equivalent of to retest reliability and found moderately stable coefficients (.67 and .70) for therapist-rated and client-rated session quality/outcome. The authors did not expect higher stability coefficients because session outcomes vary to some extent from session to session. Markin et al., (under review) also addressed the validity of the session quality/outcome and found significant associations between client- and therapist-rated session outcome and client and therapist treatment outcome. Considering these results together, these analyses suggest that the session quality/outcome ratings possess satisfactory reliability and validity. The Experiencing Level Scale (EXP; Klein, Mathieu, Kiesler, & Gendlin, 1969). This scale was used to measure the nature of the client?s personally and subjectively felt experiencing using observable linguistic and somatic markers of the client?s verbalizations. The EXP scale reliably measures the manner in which what a client says relates to their personally and subjectively felt experience (?bodily felt sense?). A client?s current manner can range from an externalized description of events with no reference to their bodily felt experiencing to an immediate exploration of meanings arising directly from felt experiencing. There are 7 levels or stages of experiencing and each have precise linguistic and somatic characteristics that can be observed. The authors assume that progressing through the levels of the scale reflect greater elaboration and integration of emotions and experience, which consequently results in resolution of client problems. At Stage 1, the content of the client?s verbalizations and manner is impersonal, abstract and externalized. There is no explicitly expressed association between the speaker?s content and the speaker, which is reflected by a communication style that has an absence of personal DYAD STUDY OF PSYCHOTHERAPY PROCESS 62 involvement and an avoidance of feelings. The narrative is expressed in a manner in which it could belong to any person since nothing unique, or personal is expressed. When Stage 2 is reached by a client, the association between the speaker and his or her content becomes explicit, but only to the extent that it serves to convey the speaker?s narrative or idea (as opposed to a feeling or inner experience). The speaker?s manner of communication is externally descriptive with any emotional involvement circumscribed only to the specific situation or content (versus inner parallels to one?s self across situations). The speaker?s feelings are thus implied but never explicitly expressed. At Stage 3, the speaker begins to add explicit comments of feelings or emotional reactions to his or her narrative content at this level. However, the speaker?s self-description is limited to circumscribed behavioral terms and the personal remarks about his or her private experience are parenthetical to the speaker?s communication. When client reaches Stage 4, the quality of involvement in speech content reflects a shift in the speaker?s attention to the subjective felt flow of his or her experience rather than to events or abstractions. Rather than objective and analytical, the speaker?s style is subjective, descriptive, and emotional. The speaker?s verbal effort centers around expanding and elaborating the details of her or his inner experience. At Stage 5, the speaker constructs a problem or question about the self when they reach using their internal elaboration of emotions and personal meanings. At Stage 6, new feelings and meanings emerge in the client from ongoing explorations to resolve emotional problems related to the self. At Stage 7, the client goes through the process of expanding awareness of immediately present feelings and internal processes more consistently and steadily. Thus the speaker immediately links and integrates felt nuances of experience DYAD STUDY OF PSYCHOTHERAPY PROCESS 63 when and as it occurs in the present moment. At levels of 4 or higher, the personal perspective becomes clearer until, at the highest levels, clients are actively processing their subjective experience in the moment towards therapeutic gain. Thus, scores of 4 and above are regarded as productive process. Interrater reliability using Ebels class correlations has been reported for the EXP scale ranging from .75 to .99 (Castonguay, Goldfried, Wiser, Raue,& Hayes, 1996; Goldman, Greenberg & Pos, 2005; Hill, Helms, Tichenor, O?Grady, Spiegel, & Perry, 1988; Klein, Mathieu, Kiesler, & Gendlin, 1969). In addition, the EXP scale has demonstrated validity as studies show that it correlates with client variables such as introspectiveness, creativity, ego strength and cognitive complexity (Hendricks, 2009; Klein et al., 1986) as well as physiological and attentional variables such as body relaxation indicators and EEG alpha frequencies (Bernick et al., 1969; Don, 1977; Gendlin & Bergin, 1961). The Speaking Turns Therapist Immediacy Measure (Kasper, Hill & Kivlighan, 2008). This measure was used to assess therapist immediacy during speaking turns, including the dimensions of frequency and type. Therapist immediacy is defined as disclosures in the therapy of how the therapist is feeling about the client, him- or her- self in relation to the client, or about the therapy relationship (Hill, 2004). In addition, based on previous empirical and theoretical research and the results of Kasper et al.(2008) and Hill et al., (2008) case studies, the following characteristics further define therapist immediacy. Therapist immediacy is a disclosure: (1) involving both the therapist and the client (i.e. excluding feedback solely focused on the client), (2) about what is happening in the here-and-now client-therapist interaction, and (3) that involves DYAD STUDY OF PSYCHOTHERAPY PROCESS 64 at least moderate engagement between the participants (i.e. minimal social pleasantries that often start the session are excluded). These three characteristics were used in the current study as the criteria for evaluating whether therapist immediacy occurred. Additionally, in the current study, therapist immediacy was categorized using the following four types of immediacy that emerged from previous case analyses of immediacy (Hill et al., 2008; Kasper et al., 2008): (1) negotiation of the tasks and goals of therapy, (2) exploration of unexpressed feelings or making the covert overt, (3) drawing parallels between other relationships and the therapy relationship, and (4) attempts to repair ruptures by talking about what is going on between the therapist and client. Kasper et al. reported that the average kappa between pairs of raters was .72 for therapist categories. Lastly, in the current study the proportion with which therapist immediacy occurred was determined using the Kasper et al. (2008) formula ? dividing the number of times a therapist immediacy action occurred in a speaking turn divided by the number of total speaking turns for the therapist in the session. The therapist in the Kasper et al. study used immediacy in about 33% of his speaking turns (M = .34, SD = .12), most frequently using the immediacy subtype inquiry about the relationship (M = .25, SD = .11), and least often using the subtypes intimately self-involving statements (M = .05, SD = .04) and feedback (M = .05, SD = .05). In addition, the authors found a significant relationship between the therapist?s use of immediacy and the use of the immediacy by the client in the subsequent speaking turn. This relationship was significant for the subtypes inquiry about the relationship r(10) = .63 p < .05 and intimate self-involving statements r(10) = .58 p < .05, but not for feedback. DYAD STUDY OF PSYCHOTHERAPY PROCESS 65 The Counseling Outcome Measure (COM; Gelso & Johnson 1983; Appendix F). This measure will be used to assess the outcome of treatment from both the client and therapist perspectives. The shortness of this measure makes it ideal for field studies in which longer measures may discourage client and therapist participation. All clients and therapists completed this measure after the last session in the treatment. Using a scale from 1 (much worse) to 4 (no change) to 7 (much improved), the four-item measure asks the participant(s) to evaluate the amount of the client?s improvement since the beginning of therapy. The items assess improvement in feelings, behavior, self-understanding, and overall functioning. The scores on each item are summed to obtain one total score. Likely due to the anchors used, the therapist and client ratings have tended to be high for example, greater than 5.0 (Gelso & Johnson, 1983; Gelso et al., 1997). Yet, despite this skew, excellent reliability of the COM has been established with studies examining test- retest reliability and internal consistency (Ain & Gelso, 2008, 2011; Fuertes et al., 2007; Gelso & Johnson, 1983; Gelso et al., 1997; Geslo et al.., 2012; Tracey, 1987). Gelso and Johnson assessed the test-retest reliability for individual items at three weeks and found the reliability to range from .63 to .81. Another study by Tracey testing the measure?s internal consistency found the measure to have an alpha estimate of .89. A later study by Gelso, Kivlighan, Wine, Jones, and Friedman replicated these results and found the form to have an internal consistency of .89. Generally, the coefficient alphas ranged from the high .80s to low .90s. The validity of the COM has also been established by the data in research examining outcome estimates for structured interviews between clients, counselors, and independents judges. Gelso and Johnson (1983) found these outcome intercorrelations to be very high. In addition, Patton, Kivlighan, and Multon (1995) DYAD STUDY OF PSYCHOTHERAPY PROCESS 66 found that client COM scores correlated significantly with outcome scores of the Brief Symptom Inventory (Derogatis & Melisaratos, 1983). In a similar vein, Patton et al. (1995) also reported significant correlations between client COM scores and outcome scores on the Inventory of Interpersonal Problems (Horowitz, Rosenberg, Baer, Ureno, & Villasenor, 1988). The construct validity of COM as an outcome measure has been supported by its associations to theoretically expected predictors such as the real relationship (Ain & Gelso, 2008; Fuertes et al., 2007), elements of time-limited psychotherapy (Gelso & Johnson, 1983) and client transference and insight (Gelso et al., 1997). The COM will be used in the current study as opposed to more established treatment outcome measures because, in addition its sound reliability and validity, it only consists of four items. Keeping the client and therapist measures short served to minimize the impact of the data collection on the therapeutic work, which enabled researchers to collect data from actual clinical settings and increase the likelihood of client and therapist participation. Procedures for Data Collection Therapist recruitment. All therapists were approached personally by the primary investigators of the original study by Fuertes et al., (under review) and invited to participate. One therapist was recruited from the Mental Health Services Center at the University of Maryland. The other two therapists were recruited from the Counseling Center at Baruch College. Client recruitment, screening, pre-therapy assessment and intake. Therapists were instructed to ask the first new client they met and that fit the research inclusion criteria to participate. They were instructed to recruit two clients in this manner. In order DYAD STUDY OF PSYCHOTHERAPY PROCESS 67 to keep the selection as random as possible, therapists were asked to use the following criteria for selection: (a) client has never been a client of the therapist prior to intake, and (b) client would very likely participate in at least 4 sessions of time-limited psychotherapy. Therapists were instructed not to select clients for any other reasons (e.g. good demeanor, cooperative, interested in research). At the beginning of the intake session with a new client, therapists were instructed to ask him or her if she or he would like to participate. During this invitation, therapists were instructed to inform the client of the following: the general purpose of the study to improve the psychotherapy process by looking at the therapist-client relationship; the required 5-10 minute post-session time demand to complete measures; sessions would be audio-taped; client and therapist responses were anonymous and not a part of therapy. Lastly, therapists were instructed to emphasize the steps taken by the researchers to ensure confidentiality (e.g. measures completed in the lobby and dropped off to the front desk staff to lock into a box for collection by the investigators, lack of personal identifying labels). After clients agreed to participate, therapists handed the clients a pre-session packet of measures that included a brief demographic questionnaire prior to the first session (See Appendix A). Clients were instructed to complete the packet of measures in the lobby or waiting area, seal it in the envelope provided, and return it to the front desk receptionist. While clients completed their pre-session packet in the lobby, therapists completed a brief demographic questionnaire in their offices (See Appendix B). Therapists enclosed the completed form in a coded envelope and also gave it to the receptionist immediately. After both participants finished the pre-session packet, DYAD STUDY OF PSYCHOTHERAPY PROCESS 68 therapists brought clients back into their offices and conducted their sessions in their normal style. No restrictions were placed upon any of the participating therapists. Treatment. The six psychotherapy cases occurred at two different University settings. One therapist from Mental Health Services (MHS) provided time-limited psychotherapy treatment to two clients at the University of Maryland?s Health Center. Typically the duration limit for the MHS is 6-8 sessions, but the therapist agreed to provide at least 12 sessions if needed. Each of the two clients at MHS received one 55- minute session once a week for eight weeks. The other four psychotherapy dyads were treatment cases from the Counseling Center at Baruch College. Two therapists provided weekly treatment two clients each (total of four dyads). The Counseling Center at Baruch College has a 12-session limit, however only two of the four cases received 12 sessions. The other two cases received 5 and 10 sessions respectively. At both universities, the psychotherapy was generally humanistic-experiential and ecclectic, with no specific guidelines or methods prescribed or introduced by the researchers. Thus the approaches and techniques reflected the client and therapist preferences and goals. As described earlier, the therapists rated their use of multiple theoretical approaches (i.e. psychoanalytic/dynamic, humanistic/existential, and cognitive/behavioral theoretical orientation) as moderately high reflecting that the psychotherapy engaged in by the participants was flexible and integrative. All of the sessions were 45-60 minutes in duration and were audiotaped. Post-session assessment. After each session (including the first session), clients took a post-session packet back to the lobby and completed the RRI-C and Session Quality measures. They sealed the envelope and turned it in to the receptionist. DYAD STUDY OF PSYCHOTHERAPY PROCESS 69 Therapists completed a post-session packet that included measures of the RRI-T and Session Quality. Similarly to the clients, therapists sealed the completed measures in an envelope and gave them to the front desk receptionist. Other brief measures that were not related to the present study were also included in the post-session packet. These measures included: the Working Alliance Inventory-Short-Therapist (T-WAIS) and Client Forms (C-WAI; Tracey & Kokotovic, 1989), Transference Scale (TS), the Inventory of Countertransference Behavior (ICB, Friedman & Gelso, 2000). All measures were ordered randomly inside the packets. Post-therapy assessment. After the final session, therapists gave clients a post- treatment packet of measures to complete in the lobby that included the RRI-CL, Session Quality, and the COM. After completing the measures, clients sealed them in the envelope and handed them in to the receptionist. Therapists completed a post-treatment packet of measures in their offices that included the RRI-T, Session Quality, and the COM. When finished, therapists sealed the measures in the coded envelope and turned them into the receptionist. The receptionist collected the packets and stored them in a locked file box that the investigators would retrieve and empty each week. Procedures for Data Coding. Session transcription. All of the sessions of the six case studies combined for a total of 55 sessions. A team of researchers working under Fuertes et al. (under review), which included two doctoral graduate students in counseling psychology and one master?s level counseling psychology student, transcribed 36 of the 39 sessions from Baruch College from audiotapes. There were 14 out of 16 sessions from the University of Maryland that were transcribed from audiotapes by two undergraduate research assistants DYAD STUDY OF PSYCHOTHERAPY PROCESS 70 in psychology and checked by a doctoral student in counseling psychology working with the principal investigator to collect the data at the University of Maryland for the Fuertes et al. study. Five sessions (two from Maryland, and three from Baruch) were unable to be transcribed due to audiotape deterioration. The remaining 50 transcripts were edited so that all clients? identifying information was removed. Training judges and coding immediacy. All judgments regarding immediacy were made using a rating method called consensual qualitative research-cases (CQR-C; Jackson, Chui, & Hill, 2012). In CQR-C, judges are encouraged to use clinical intuition, reasoning, and thoughtful multifaceted discussion in order to build a final shared viewpoint (consensus). When utilizing CQR-C it is important to make considerable effort to ensure each person feels comfortable talking and sharing his or her viewpoint in order to construct an effective consensus. This method was chosen for my study for two main reasons: 1) it enabled the effective evaluation of immediacy, which is a complex clinical phenomenon, and 2) it allowed for multiple clinical perspectives to be voiced and ultimately integrated into a collaborative and more clinically relevant immediacy judgment. Four advanced undergraduate students and the first author (an advanced doctoral student) comprised the research team. Prior to the first team meeting, the undergraduate judges read about immediacy (Hill & Knox, 2009; Kiesler, 1996; Teyber, 2006; Yalom, 1995, 2002) and noted any questions regarding the definition of immediacy. Then the team met for three hours and observed several immediacy events (using multiple role- plays and listening to practice video-taped sessions) in order to clarify and refine the definition of an immediacy event (i.e. it had to be about the immediate therapy DYAD STUDY OF PSYCHOTHERAPY PROCESS 71 relationship in the moment, both people had to participate in the discussion, had to be more than social chit chat, and did not include feedback regarding only the client) and to develop criteria for judgments. During this meeting, the team also practiced identifying an immediacy event, conceptualizing its type and purpose, and rating the immediacy event on depth, appropriateness, resolution, and quality, using 5-point scales from 1 (poor) to 5 (excellent). Utilizing audiotapes and written transcripts, each session was pre-screened by the advanced doctoral student in order to identify whether a session contained an immediacy event. If an immediacy event was not identified by the doctoral student in a particular session, then the session was independently screened again by two of the other judges to confirm the lack of immediacy. In order for the session to be rated using CQR-C, two out of the three judges had to observe an immediacy event in a session during the screening process. Thus all sessions were pre-screened by at least two-judges independently to identify immediacy events. Fifteen sessions were identified as containing immediacy events and evaluated using CQR-C. Three-person teams were then randomly created with one of the three judges always being the advanced doctoral student. For 8 weeks, each team coded between three and eight sessions occurring in three cases, and the change in team membership over time ensured the teams were coding consistently. Overall, the research team worked for 15 weeks ? reviewing 50 sessions, and coding 16 sessions, which contained an immediacy event. The proportion of sessions that had an immediacy event (32%) fell in between the frequency proportions of therapist immediacy in previous studies that ranged from 12% - DYAD STUDY OF PSYCHOTHERAPY PROCESS 72 38% (Hill, Gelso et al. under review; Hill et al., 2009; Kasper et al., 2009; Mayotte-Blum, 2012). During each meeting, teams listened to (and followed typed transcripts of) one of the sessions involving an immediacy event. As the team listened to each session, they were encouraged to briefly note and record any strong reactions they experienced. Whenever an immediacy event occurred, the team stopped the tape. They then reviewed the event using the transcript (and tape playback if necessary) and wrote a description of the event noting the specific speaking turn in which the event began and ended and the duration of the event. At times, considerable discussion was required to decide when an event began. The team then discussed the event to conceptualize how it was used within the entire session, and then coded the type of event (negotiation of the tasks and goals of therapy, exploration of unexpressed feelings or making the covert overt, drawing parallels between other relationships and the therapy relationship, attempts to repair ruptures by talking about what is going on between the therapist and client). Next, the judges independently rated depth, appropriateness, resolution, and quality, using 5-point scales from 1 (poor) to 5 (excellent), and then discussed their ratings until they reached consensus. The anchors used in the 5-point scale for depth were identical to those to the Hill et al. (2008) and Mayette-Blum (2012) case studies: 1 = mundane, one-sided; 2 = minimal two-person exchange; 3 = longer two-person exchange lacking depth; 4 = prolonged two-person exchange; 5 = prolonged exchange with both participants actively expressing genuine immediate feelings. Anchors were also created for the other immediacy dimensions by the judges during the training. For appropriateness, 1 = no (clinical or interpersonal) relevance, clear client discomfort, DYAD STUDY OF PSYCHOTHERAPY PROCESS 73 offensive or distancing to client; 2 = minimal relevance, little to no client comfort during the interaction (but not offensive), minimally tailored to client (i.e. used client?s language, or intellectual level of understanding, etc.); 3 = some relevance, some client comfort, some tailoring to client; 4 = mostly relevant, comfortable, and tailored to client; 5 = highly relevant, comfortable, and tailored to client. The anchors for resolution were 1 = no awareness of new/ different reactions and feelings, no here-and-now elaboration; 2 = minimal awareness of new/different reactions and feelings, no elaboration; 3 = some new/different awareness, minimal elaboration; 4 = more new/different awareness, some elaboration; 5 = more new/different awareness, extended elaboration, and made new/different connection(s) during the elaboration. There were no anchors created for quality and to some degree reflected a summary of the other three dimensions. However, there were instances in which the judges experienced the event as a higher or lower quality event than would be expected by their other dimensions numbers due to various factors not accounted for in the other dimensions. For example, after one event (A) the judges discussed feeling more compassionate or inspired (or the opposite, shut-down and highly critical) compared to another event (B), though the other dimension ratings were very similar for the (A and B) events. Thus the judges rated event A of higher quality. The judges were also encouraged to write down brief notes describing the reasoning for their ratings using the anchors. During the discussion the raters were also directed to describe their rationales explaining why they did not choose a higher or lower number and to discuss any reactions that may have positively or negatively influenced their evaluations. These efforts were done to promote discussion during the meetings and to ensure consistency during and between team meetings. DYAD STUDY OF PSYCHOTHERAPY PROCESS 74 Training judges and coding client experiencing level. Three judges were trained according to the formal training procedures and materials given in The Experiencing Scale: A Research and Training Manual (Klein et al., 1969). The research team was comprised of two undergraduate students and one advanced doctoral student that led the team. All of the judges read and familiarized themselves with the theoretical and research background on the scale, the description of the stages or levels of experiencing, and instructions for the rating task provided in the manual. The overall training program consisted of eight 2-hour sessions. Judges used sample excerpts from sessions included in the training manual to practice rating. Judges also used the Experiencing Decision- Tree (Rogan et al. 1999) to facilitate rater accuracy and speed. Each session involved rating 10 practice segments, which were then compared with criterion ratings and justifications-explanations provided in the manual (Klein et al., 1969). Inter-rater reliability was assessed at the end of training using a block of 20 segments. Prior to the first team meeting, the advanced doctoral student trained with a researcher from McGill University already highly trained (3 years experience) on the scale until their ratings attained an inter-rater reliability of .90. Similar to the training for the undergraduate students, the advanced doctoral student trained utilizing the formal training procedures and materials given in the training manual (Klein et al.). After every 2 practice sessions, the doctoral student and experienced rater met to discuss statements they did not reach agreement upon. The doctoral student completed the training after 8 practice sessions. The doctoral student then led two undergraduate psychology majors through a similar training process for 8 weeks. The judges worked independently during each DYAD STUDY OF PSYCHOTHERAPY PROCESS 75 practice session, but the research team met twice a week to discuss statements that they did not reach agreement upon. Inter-rater reliability was assessed at the end of training using a block of 20 segments. The kappa between raters for the 20-segment block was .88. Then for 10 months the research team rated client experiencing level for 51 sessions (6 months of active rating). Each judge rated every session using ?running ratings? (Fitzpatrick, et al., 1999; Klein et al., 1969). Running ratings were continuous ratings of each phrase or statement within a speaking turn. A judge recorded a rating from the beginning of each turn and then only noted a another rating if and when the level changed from the previous statement or phrase. Running ratings allowed for a mode and peak client experiencing level to be determined for each speaking turn. Judges worked independently using a transcript in order to make their ratings, but met as a group to explore and discuss problematic or unclear segments. In order to facilitate accuracy and speed, judges used the Experiencing Decision-Tree (Rogan et al. 1999) and noted rating justifications that explained why a particular speaking turn segment did not get a higher and/or lower number. Moreover, the justifications had to stem from the actual stage descriptions given in the manual rather than the rater?s own conceptualizations or ideas. The group leader compiled each week?s ratings. Any statement that was classified identically by two out of the three raters was accepted as the final rating for that statement. Any rating that did not get a two out of three level of agreement was subjected to the consensus procedure. The consensus procedure occurred when the raters met to discuss statements that had not reached the two out of three level of agreement. The DYAD STUDY OF PSYCHOTHERAPY PROCESS 76 group listened to the audio recording of the disputed statement while following along with the transcript and each member discussed their rationale for their rating until consensus was reached. Inter-rater reliability was also checked every 5th or 6th session. For the 13 sessions checked, intraclass correlations among the judges ranged from .79 to .98 (M = .91, SD = .05). DYAD STUDY OF PSYCHOTHERAPY PROCESS 77 Chapter 4 Results Preliminary Analyses Intercorrelations. An intercorrelation matrix involving all of the main variables of the study: therapist-rated real relationship (RRI-T), client-rated real relationship (RRI- C), therapist-rated session quality (SQ-T), client-rated session quality (SQ-C), amount of therapist immediacy (IMM-Amt), therapist immediacy depth (IMM-D), therapist immediacy appropriateness (IMM-App), therapist immediacy resolution (IMM-R), therapist immediacy quality (IMM-Q), mode of client experiencing level (EXP-M), and peak of client experiencing level (EXP-P) is presented in Table 27 in Appendix H. In considering these correlational results, it is important to recall that the correlations do not take into account the interdependence and nesting of the data. From the client vantage point, there were several significant correlations between the real relationship and the other variables. Corresponding to Hypothesis 1a, RRI-C was significantly positively correlated to SQ-C. Consistent with Hypotheses 3c, 3e, and 3g, RRI-C also was significantly positively correlated to immediacy appropriateness, immediacy resolution, and immediacy quality. Lastly, RRI-C and mode of client experiencing level were significantly positively correlated which supports Hypotheses 4a. The client?s perspective of session quality (SS-C) was also significantly related to immediacy appropriateness, resolution, quality, as well as mode of client experiencing level, though no hypotheses were predicted for these variables. From the therapist vantage point, RRI-T was significantly correlated to RRI-C as well as session quality ratings from both the client and therapist. Interestingly, the DYAD STUDY OF PSYCHOTHERAPY PROCESS 78 correlations between RRI-T and SQ-C and SQ-T were lower than the correlation between RRI-T and RRI-C. There were no other correlations between RRI-T and therapist immediacy and client experiencing level. Regarding the associations between variables in which no hypotheses were made, neither SQ-T nor immediacy amount were not significantly correlated to any other variables. In addition, there were moderate and positive correlations between immediacy appropriateness and EXP-M and EXP-P, and immediacy quality and EXP-M and EXP-P. And lastly, all of the immediacy dimensions were highly correlated to each other (also in Table 8) discussed further below. Means and standard deviations. There were four main sets of HLM analyses done. It is important to note that three of the variables (real relationship, client experiencing level, therapist immediacy) were involved in multiple analyses where the number of cases for the variable in each analysis may have varied. The number of cases depends on many factors, including the level of the variable within the HLM model and the unit of analysis for each level. As a consequence, real relationship (RR), client experiencing level, and therapist immediacy have different means and standard deviations depending on the analysis. For example, in one set of analyses where n = 16, RR was examined in relationship to immediacy and RRI-T had a mean of 4.38 while RRI-C had a mean of 4.81. In another set of analyses where n = 50, RR was examined in relationship to experiencing level and RRI-T had a mean of 4.05 while RRI-C had a mean of 4.29. Each set of analyses is represented in each of the five subsequent tables below. It is also important to note that no statistical tests were done to determine the reliability and significance of the differences between the means, so the comparisons reported DYAD STUDY OF PSYCHOTHERAPY PROCESS 79 below are estimates regarding where the means and standard deviations fall within the analyses and regarding other studies, but are not inferences to the population, which would require significance testing. Overall means and standard deviations for clients? (RRI-C) and therapists? (RRI- T) ratings of RR for each set of analyses are presented in Tables 1, 2, 3, and 5. Although the overall mean scores on the RRI-C and RRI-T varied in this study according to analysis, there were two main trends that applied to the RRI-I and RRI-T means across all analyses: 1) all of the RR mean scores were above 4 on a 5-point scale where 1 = strongly disagree and 5 = strongly agree indicating that the clients and therapists in this study perceived a strong real relationship, and 2) in every analysis conducted, client ratings on average were greater than therapist ratings. In Table 2, the means and standard deviations from the analyses of RR in relation to session quality are reported. In these analyses the number of cases was 55. The means of RRI-T scores appear higher than RRI-T means from previous studies (Ain & Gelso, 2008, 2011; Gelso et al., 2012; Lo Coco et al., 2011; Marmarosh et al., 2009). The RRI- C means were slightly higher than the RRI-C means in previous studies (Ain & Gelso, 2008, 2011; Gelso et al., 2012; Lo Coco et al., 2011; Marmarosh et al., 2009). The means for client-rated and therapist-rated session quality were higher than the judge-rated session quality mean in a previous study (Gelso, Hill and Kivlighan, 1991) and clients rated session quality higher than the therapists. Again, these comparisons were not tested to determine their reliability or statistical significance, but the above trends appeared occurred in the data set. DYAD STUDY OF PSYCHOTHERAPY PROCESS 80 Table 2. Means and Standard Deviations for Clients? and Therapists? Ratings on Session Quality Measure and the Real Relationship Inventory Clients Therapists Variable N M SD M SD Session Quality 55 4.32 0.53 3.84 0.37 Real Relationship 55 4.07 0.40 3.96 0.30 The means and standard deviations for real relationship and therapist immediacy amount within a session and occurrence within a session are presented in Table 3. In this set of analyses the n for immediacy was 55 (unit = session), and the n for RR was 6 (unit = dyad). The RRI-C and RRI-T ratings of the 6 dyads were above 4.0 on a 5-point scale and appeared similar to previous studies and other analyses in the current study, though no inferential statistics were done to determine whether these trends were reliable or applied to the population. The average amount of speaking turns that contained immediacy within a session was 8.13 (SD = 13.47). The average percentage of sessions in which immediacy occurred across all the dyads (each dyad) was 29% (SD = 46%). DYAD STUDY OF PSYCHOTHERAPY PROCESS 81 Table 3. Means, Standard Deviations, Minimums and Maximums for Clients? and Therapists? Ratings on the Real Relationship Inventory and for Amount of Speaking Turns that Contained Therapist Immediacy within a Session and Percentage of Sessions in which Therapist Immediacy Occurred. Variable N M SD Min Max Amount of Immediacy STs within a Session 50 8.13 13.47 0.00 70.00 % of Sessions with IMM Occurrence 50 32% 47% 0% 100% Client Real Relationship 6 4.35 0.49 3.87 4.97 Therapist Real Relationship 6 4.13 0.34 3.70 4.58 The means and standard deviations for real relationship and therapist immediacy are presented in Table 4. In this set of analyses the n for immediacy was 363 (unit = speaking turn), and the n for RR was 16 (unit = session). The RRI-C and RRI-T ratings of the 16 sessions in which therapist immediacy occurred had the highest mean score compared to all the other analyses in this study. In Table 2, it is important to note that a mean and standard deviation is reported for each of the four immediacy rating dimensions (i.e. immediacy depth, immediacy appropriateness, immediacy resolution and immediacy quality) because each dimension was examined in its own HLM analysis. Therapist immediacy depth, appropriateness, resolution, and quality were all 5-point Likert scales, ranging from 1 (poor) to 5 (excellent). Of the four rating dimensions of immediacy, appropriateness had the highest mean rating while the resolution had the lowest mean rating. Additionally, the average immediacy event was rated above the midpoint on all the immediacy dimension scales -- thus, exhibiting moderate resolution and depth, and moderately high quality and appropriateness, likely indicating a moderately to highly effective immediacy event. A previous study by Hill, Gelso et al. (under review) combined all four immediacy dimensions into one construct called quality of immediacy and had a lower mean rating of 2.61 (SD = .42, range = 2.00 to 3.46), DYAD STUDY OF PSYCHOTHERAPY PROCESS 82 though no inferential testing was done to determine whether this trend was reliable or statistically significant. Table 4. Means, Standard Deviations, Minimums and Maximums for Clients? and Therapists? Ratings on the Real Relationship Inventory and for Judges? Ratings of Immediacy Depth, Appropriateness, Resolution, and Quality. Variable N M SD Min Max Immediacy Depth 363 3.17 0.89 1.00 1.00 Immediacy Appropriateness 363 3.90 0.71 1.50 1.50 Immediacy Resolution 363 2.89 0.86 1.00 1.00 Immediacy Quality 363 3.37 0.93 1.00 1.00 Client Real Relationship 16 4.81 0.42 3.71 3.71 Therapist Real Relationship 16 4.38 0.38 3.71 3.71 In Table 5, the means and standard deviations for real relationship and client experiencing level (mode and peak) are reported. For this set of analyses the number of cases was 4443 (unit = speaking turn) for experiencing level and for RR the number of cases was 50 (unit = session). Recall, that for the judge-rated measures, five sessions did not get assessed due to technical difficulties. Again, in regards to real relationship, all of the mean ratings were above 4.0 and the client-rated mean was higher than the therapist- rated mean. Client experiencing level was rated on a 7-point scale, ranging from 1 (where client disclosure is abstract, impersonal, and externalized) to 7 (where client awareness and engagement of feelings is immediate, expansive, and exploratory and experiencing serves as the basic referent for problem resolution and self-understanding). The peak client experiencing level mean was only .07 points higher than the mean mode of client experiencing level, which indicates that high levels of experiencing were infrequently reached within this study. In fact out of 4,443 speaking turns, only 63 (1.42% of all speaking turns) reached an experiencing level of 4 or higher in which the DYAD STUDY OF PSYCHOTHERAPY PROCESS 83 quality of the client?s involvement in his or her disclosure shifts to the more subjective, personal, and emotional and ?felt? flow of his her experience rather than to events or abstractions (Klein, et al., 1969). Table 5. Means, Standard Deviations, Minimums and Maximums for Clients? and Therapists? Ratings on the Real Relationship Inventory and for Judges? Peak and Mode ratings of Client Experiencing Level. Variable N M SD Min Max Experiencing Mode 4,444 2.07 0.34 1.00 6.00 Experiencing Peak 4,444 2.14 0.43 1.00 6.00 Client Real Relationship 50 4.29 0.55 3.46 5.00 Therapist Real Relationship 50 4.05 0.42 3.21 5.00 The overall average peak experiencing levels in the present study appeared lower than previous studies? average peak levels (Castonguay et al.,1996; Goldman et al., 2005; Hill et al. 1988), though no significance testing was done to determine whether this result was significant or reliable. Yet, the means of those three studies still fell below 4.0 (M = 2.91, SD = 0.35; M = 2.31, SD = 0.59; and M = 3.81, SD = 0.49, respectively) suggesting that experiencing levels of 4.0 or higher are not frequently reached. Thus, the means for peak and mode experiencing level in the current study indicate that in the typical speaking turn the client was disclosing in an externalized manner with limited to no involvement of any inner self-referential process or exploration of feelings. Furthermore, in only about 1% of speaking turns did clients display peaks of high experiencing levels (> 4.0) and resulting productive process. Means and standard deviations for therapist immediacy and client experiencing level (prior- and post-experiencing level) are presented in Table 6. The trends in these analyses followed very similar patterns to the other analyses in this study. In these DYAD STUDY OF PSYCHOTHERAPY PROCESS 84 analyses the number of cases for both therapist immediacy and client experiencing level (prior- and post-experiencing level) was 363 (unit = speaking turn). Of the four immediacy dimensions, appropriateness had the highest mean rating and resolution had the lowest mean rating. Again, the peak experiencing level mean was only slightly higher than the mean for mode experiencing level (for both prior- and post-experiencing level). Similarly, the mean for the prior peak experiencing level was slightly higher than the mean for prior mode of experiencing level. Table 6. Means, Standard Deviations, Minimums and Maximums for Judges? ratings of Therapists Immediacy Depth, Appropriateness, Resolution, and Quality and for Judges? Peak, Mode, Session-Prior Peak and Session-Prior Mode ratings of Client Experiencing Level. Variable N M SD Min Max Immediacy Depth 363 3.18 0.89 1.00 4.67 Immediacy Appropriateness 363 3.91 0.71 1.50 5.00 Immediacy Resolution 363 2.89 0.85 1.00 4.33 Immediacy Quality 363 3.38 0.92 1.00 4.67 Post-Experiencing Mode 363 2.15 0.47 2.00 6.00 Post-Experiencing Peak 363 2.17 0.49 2.00 6.00 Prior Experiencing Mode 363 2.16 0.50 2.00 6.00 Prior Experiencing Peak 363 2.18 0.52 2.00 6.00 Frequency Summary. Unlike the other speaking-turn level variable (experiencing level), therapist immediacy did not occur during every speaking turn. Therefore, a frequency summary for therapist immediacy is presented in Table 7 in Appendix G. Out of 50 total sessions, 16 sessions had immediacy events. A total of 46 therapist immediacy events took place in those 16 sessions and the events consisted of 363 speaking turns. Overall, the therapists used immediacy in only 8% of the 4,444 total speaking turns in the study. However, the therapists did not have uniform usage DYAD STUDY OF PSYCHOTHERAPY PROCESS 85 proportions. Therapist A did not use immediacy at all. Therapist B used immediacy very rarely, in only 1% of her speaking turns during the course of treatment for only one of her clients. On the other hand, the therapist C, used immediacy on average in about 16% of her speaking-turns with her clients (15% for Client E; and 25% for Client F). Although a total of 46 therapist immediacy events (ranging from 0 to 24 per dyad) occurred only in three of the six dyads, for the descriptive data we computed the means for each dyad and then computed the averages across all dyads in order to control for the different number of sessions across cases. Across the six dyads, the average number of therapist immediacy events per session was 1.08 (SD = 1.56, ranging from 0 to 3.17 per session per dyad). This is slightly higher than the average (M = .61) in the study by Hill, Gelso et al. (under review), though no inferential tests were done to determine if this is significant. Regarding types of immediacy events (this includes primary and secondary descriptions), an average of 26% (SD = 31%, range = 0% to 71%) involved an exploration of unexpressed or covert feelings, 25% (SD = 29%, range = 0% to 67%) involved discussion of parallels to other relationships, 17 % (SD = 19%, range = 0% to 42%) involved discussion of ruptures, 16% (SD = 17%, range = 0% to 33%) involved negotiation of tasks and goals. In comparison to Hill, Gelso et al., (in press) where discussion of covert feelings about the therapy relationship was generally the focus of therapist immediacy events (occurring in 59% of immediacy events), the therapists in this study appeared to use the immediacy types with similar frequency as no immediacy type was used on average more than 26% of the time. DYAD STUDY OF PSYCHOTHERAPY PROCESS 86 In summary, the typical immediacy event in this study was similar to previous studies in that it was brief and at least moderately effective. However, the therapists used the immediacy types about equally. In addition, the average immediacy event was rated above the midpoint on all the immediacy dimension scales, exhibiting moderate resolution and depth and moderate to high quality and appropriateness, likely indicating a moderate to highly effective immediacy event. Table 8. Correlation Matrix for Therapist Immediacy Dimensions. Depth Appropriateness Resolution Quality Depth 1 Appropriateness .75*** 1 Resolution .94*** .81*** 1 Quality .91*** .90*** .96*** 1 Note: ***p < .01 Therapist Immediacy Dimensions. Table 8 presents the bivariate intercorrelations for the four dimensions of therapist immediacy: depth, appropriateness, resolution and quality. All of the dimensions were highly and positively correlated with each other with intercorrelations ranging from .75 to .96 (p < .01). These correlations were similar to a previous immediacy case study (Hill et al., under review) that also found high positive intercorrelations (ranging from .76 to .92) for each of the four dimensions of therapist immediacy. Analysis of Hypotheses and Research Questions HLM (Raudenbush, Bryk, & Congdon, 2008) and Structural Equation Modeling (SEM) were used to conduct the data analyses due to the nested nature of the data. HLM was the preferred analysis because the variables were measured at and nested within DYAD STUDY OF PSYCHOTHERAPY PROCESS 87 different levels (i.e. speaking turn, session, and dyad) for each analysis. As a result of this nesting, there is an increased probability of dependent observations and an increased potential for fallacious aggregation or disaggregation of the data. For example, for the first hypothesis, the strength of the real relationship and session quality are session-level variables. However, the real relationship and session quality ratings were collected from both the client?s and the therapist?s perspectives and are nested within psychotherapy dyad. Thus, they are not independent. HLM takes into account the fact that there are correlated error terms between clients who have the same therapist. Ignoring this hierarchical structure of the data could cause overestimation of sampling variances, exaggerated degrees of freedom, overly narrow confidence intervals, and an increase in the likelihood of Type I error (Croninger, 2010). For these reasons, multivariate analyses were conducted using HLM 6.0 (Raudenbush, Bryk, & Congdon, 2000). Four sets of HLM analyses were conducted to test two hypotheses and one research question, as well as to conduct an additional exploratory analysis. The first set of HLM analyses were 2-level models (real relationship in relation to therapist immediacy) conducted to test Hypotheses 2a and 2b. The second set of HLM analyses were 3-level models (real relationship in relation to client experiencing level) conducted to test Hypotheses 3a and 3b. Again, a three-level model was used because the observed data were nested within speaking turn, session, and dyad. The third set of HLM analyses were 2-level models (therapist immediacy versus client experiencing level) conducted to explore possible relationships between therapist immediacy and client experiencing level. The last set of HLM analyses were 2-level models (between counseling outcome and real relationship) conducted to test the research question. DYAD STUDY OF PSYCHOTHERAPY PROCESS 88 Real Relationship and Session Quality Hypothesis 1a. The stronger the real relationship from the client?s perspective, the better the session quality from the therapist?s perspective. Hypothesis 1b. The stronger the real relationship from the client?s perspective, the better the session quality from the client?s perspective. Hypothesis 1c. The stronger the real relationship from the therapist?s perspective, the better the session quality from the therapist?s perspective. Hypothesis 1d. The stronger the real relationship from the therapist?s perspective, the better the session quality from the client?s perspective. As stated earlier, the client?s and therapist?s real relationship and session quality ratings are nested within the psychotherapy dyad and are not independent. Consequentially, the first set of hypotheses was tested by 1) constructing the variables into an actor-partner interdependence model (APIM) that takes into account the dependencies of the nested data, and then 2) analyzing the data using structural equation modeling (SEM). The APIM is advantageous because it assumes interdependence between clients and therapists and tests the data for this interdependence. Using the APIM to examine the effect of an actor on their partner and the effect of the partner on the actor was recommended by Kenny et al. (2002). The APIM removes the actor?s (client or therapist) real relationship ratings from the calculation of their partner?s (client or therapists) session quality ratings. The APIM accounts for the nesting of clients and therapists by specifying a correlation between error terms associated with the partner?s session quality dimensions. This correlation models the nonindependence DYAD STUDY OF PSYCHOTHERAPY PROCESS 89 of errors between an actor?s (client or therapist) and partner?s (client or therapist) session quality ratings. SEM was used instead of HLM because it has several advantages over MLM approaches that address nesting. First, SEM makes of use of a more simple data structure. In the current data set, the actor and partner are distinguishable making it suitable for SEM. Unlike SEM, MLM would still require additional variables and product terms to distinguish the actor from the other group members. Lastly, the APIM conceptual model can be translated in very straightforward ways into the path-analytic model. Thus, in the current dyad study, a path analysis within an SEM framework was used to analyze the APIM (Kenny, Mannetti, Pierro, Livi, & Kashy, 2002) and test the first set of hypotheses. The basic version of APIM is applicable in the current study and is displayed in Figure 1. Within this model, it is important to note that, consistent with the APIM literature, the term effect is used to describe different actor-partner relationships and not to indicate causation. The APIM that was constructed includes two predictor variables: the therapist?s and client?s real relationship (within his/her dyad), and the therapist?s and client?s session quality (within his/her dyad). There are two types of actor effects: the effect of the client?s rating of the real relationship on his or her own rating of session quality (CAE; client-actor effect), and the effect of the therapist?s rating of the real relationship on his or her own rating of session quality (TAE; therapist-actor effect). In addition, there are two types of partner effects: the effect of the client?s real relationship rating on the therapist?s rating of session quality (CPE; client-partner effect), and the effect of the therapist?s real relationship rating on the client?s rating of session DYAD STUDY OF PSYCHOTHERAPY PROCESS 90 quality (TPE; therapist-partner effect). Because the APIM model is saturated, fit statistics are not relevant. Figure 2 displays the APIM results for real relationship and session quality using the standardized solution. As shown in the figure, none of the actor effects (therapist or client) or partner effects (therapist or client) were significant. The therapist?s actor (b = .09, p > .05) and partner (b = .01, p > .05) effects were not significant, nor was the client?s actor (b = .13, p > .05) and partner (b = .09, p > .05) effects. Together, RRI-C and RRI-T only accounted for 2% of the variance in session quality. The relationship between RRI-C and RRI-T was almost zero and not significant (r = .03, p > .05). Similarly, the relationship between therapist-rated session quality and client-rated session quality was small and not significant (r = .12, p > .05). Therapist-rated Real Relationship Client-rated Session Quality Therapist-rated Session Quality Client-rated Real Relationship PC TAE TPE CPE CAE Figure 1. Schematic of actor-partner independence model for therapists? and clients? real relationship and their session quality ratings. CAE = client-actor effect; TAE = therapist-actor effect; TPE = therapist-partner effect; CPE = client-partner effect; PC = the correlations between the predictor variables, CC = the correlation between the criterion variables. CC DYAD STUDY OF PSYCHOTHERAPY PROCESS 91 In sum, the hypotheses were not supported as CRR was not related to client- or therapist- rated session quality (hypotheses 1a and 1b), and TRR was also not related to client- or therapist-rated session quality (hypotheses 1c and 1d). Real Relationship and Therapist Immediacy Hypothesis 2a. The higher the amount of therapist immediacy, the stronger the real relationship from the client?s perspective Hypothesis 2b. The higher the amount of therapist immediacy, the stronger the real relationship from the therapist?s perspective. To determine whether therapist and client ratings of RR in a session was related to amount of immediacy in a session, six HLM analyses were conducted with real relationship and therapist immediacy. Immediacy amount was defined as the number of times (turns) in a session that the therapist made an immediacy response. In the first Therapist-rated Real Relationship Client-rated Session Quality Therapist-rated Session Quality Client-rated Real Relationship b = .09 b = .01 b = .09 b = .13 Figure 2. Actor-partner independence model (APIM) for therapists? and clients? real relationship and their perceptions of session quality. r = .12 r = .03 DYAD STUDY OF PSYCHOTHERAPY PROCESS 92 HLM analysis, the independent variable was the real relationship as rated by both the therapist and client (level-2 variable; unit = dyad) and the dependent variable was therapist immediacy amount (level-1 variable; unit = session). This analysis tested both hypotheses 2a and 2b. The two-level model was specified as follows. Level 1 Model Y (amount of immediacy) = ?0 + ?1 ? (therapist RR within session) + ?2 ? (client RR within session) + r. Level 2 Model ?0 = ?00 + ?0 ?1 = ?10 + ?1 ?2 = ?20 + ?2 In the second HLM analysis, the independent variable was the client and therapist rating of the real relationship (level-2 variable; unit = dyad) and the dependent variable was a dichotomous variable that indicated whether (coded 1) or not (coded 0) the therapist used immediacy in the session. This analysis tested both hypotheses 2a and 2b by testing whether the occurrence or non-occurrence of immediacy (Immediacy Yes/No) and if it did occur whether it was related to RR strength for clients and therapists. The two-level model was specified as follows: Level-2 Model Probability (Immediacy ? Yes/No = 1|?j) = ? log[?/(1 - ?)] = ? ? = ?0 + ?1 x (therapist RR within session) + ?2 ? (client RR within session) DYAD STUDY OF PSYCHOTHERAPY PROCESS 93 Level-2 Model ?0 = ?00 + u0 ?1 = ?10 + u1 ?2 = ?20 Level-1 variance = 1/[ ? (1- ?)] Hypothesis 3a: The higher the depth of therapist immediacy, the stronger the real relationship from the client?s perspective. Hypothesis 3b: The higher the depth of therapist immediacy, the stronger the real relationship from the therapist?s perspective. Hypothesis 3c: The higher the appropriateness of therapist immediacy, the stronger the real relationship from the client?s perspective. Hypothesis 3d: The higher the appropriateness of therapist immediacy, the stronger the real relationship from the therapist?s perspective. Hypothesis 3e: The higher the resolution of therapist immediacy, the stronger the real relationship from the client?s perspective. Hypothesis 3f: The higher the resolution of therapist immediacy, the stronger the real relationship from the therapist?s perspective. Hypothesis 3g: The higher the quality of therapist immediacy, the stronger the real relationship from the client?s perspective. Hypothesis 3h: The higher the quality of therapist immediacy, the stronger the real relationship from the therapist?s perspective. DYAD STUDY OF PSYCHOTHERAPY PROCESS 94 To determine whether therapist and client ratings of RR in a session was related to each of the four dimensions of immediacy in a session, four HLM analyses were conducted. The independent variable for these analyses was the real relationship (level-2 variable; unit = session) and the dependent variable was each of the four immediacy rating dimensions (level-1 variable; unit = speaking turn). Thus a separate HLM was created for each of the four immediacy-rating dimensions (i.e. real relationship in relation to immediacy depth, real relationship in relation to immediacy appropriateness, real relationship in relation to immediacy resolution, and real relationship in relation to immediacy quality). The four models for each analysis were identical and tested hypotheses 3a ? 3h by examining whether the amount of immediacy that was appropriate, in-depth, of good quality, and good resolution was related to the strength of the real relationship from the client and therapist perspectives. The two-level model was specified as follows. Level 1 Model Y (immediacy dimension) = ?0 + ?1 ? (speaking turn) + r. Level 2 Model ?0 = ?00 + ?01 ? (therapist RR) + ?02 ? (client RR) + ?0 ?1 = ?10 + ?11 ? (therapist RR) + ?12 ? (client RR) + ?1. Coefficients, standard errors, and t-ratios for fixed effects are presented in Tables 9, 10 for Hypotheses 2a and 2b, and in Tables 11, 12, 13, 14, and 15 for Hypotheses 3a ? 3h. Table 9 in Appendix I and Table 10 in Appendix J present the results for Immediacy Amount and Immediacy Occurrence. There were no statistically significant or meaningful results in either analysis. DYAD STUDY OF PSYCHOTHERAPY PROCESS 95 Table 11 presents the results for RR and Immediacy Depth. There were only two significant findings. The first finding was that the average level of Immediacy Depth was 2.9 and was significantly different from zero. However, this result is not meaningful because the scale does not contain zero. Thus for the remaining analyses it will not be noted. Secondly, while therapist rated RR was not related to average level of immediacy depth, client-rated RR was found to be significantly related to immediacy depth. Accordingly, for every 1 point rise in client-rated RR, there is almost a .06 rise in Immediacy Depth (?02 = 0.06, SE = 0.03, t(12) = 2.13, p < .054). Thus when the client is reporting a stronger RR, the depth of immediacy is higher than when the client is reporting a weaker RR. Dividing this gamma (?02) by the standard deviation of the dependent variable provides an estimate of d, the effect size (ES). Thus the effect size for the relationship between client-rated RR and immediacy depth is .07, which is considered a small effect size. Table 11. RR & Immediacy Depth Effect Coefficient SE t(12) p d Depth Intercept 2.90 0.25 11.41 0.000 3.26 Therapist RR -0.008 0.03 -0.28 0.787 -0.009 Client RR 0.06 0.03 2,13 0.054 0.07 Depth Slope 0.001 0.0006 0.23 0.823 0.001 Therapist RR 0.0001 0.0007 0.22 0.831 0.0001 Client RR 0.001 0.0007 1.54 0.149 0.001 Table 12 presents the results for RR and Immediacy Appropriateness. Similar to the results for Immediacy depth, there was only one significant and meaningful finding. Client-rated RR was significantly related to immediacy appropriateness (?02 = 0.08, SE = DYAD STUDY OF PSYCHOTHERAPY PROCESS 96 0.02, t(12) = 3.72, p < .003). Accordingly, for every 1 point rise in client-rated RR, there is almost a .08 rise in Immediacy Appropriateness. Thus when the client is reporting a stronger RR, immediacy is being used more appropriately than when the client is reporting a weaker RR. The effect size for the relationship between client-rated RR and immediacy appropriateness is .11, which is only slightly larger than the ES for immediacy depth and still considered a small effect size. Table 12. RR & Immediacy Appropriateness Effect Coefficient SE t(12) p d Appropriateness Intercept 3.83 0.20 19.1113 0.000 5.39 Therapist RR -0.022 0.02 -0.96 0.359 -0.03 Client RR 0.08 0.02 3.72 0.003 0.11 Appropriateness Slope 0.002 0.004 0.43 0.677 0.003 Therapist RR -0.00004 0.0004 -0.10 0.922 -5.63 Client RR 0.0006 0.0004 1.36 0.200 0.0008 Table 13 presents the results for RR and Immediacy Resolution. Similar to the previous analyses, there was only one significant and meaningful finding. Only client- rated RR was found to be significantly related to immediacy Resolution. Accordingly, for every 1 point rise in client-rated RR, there is a .06 rise in Immediacy Resolution (?02 = 0.06, SE = 0.02, t(12) = 2.79, p < .017). Thus when the client is reporting a stronger RR, immediacy has higher resolution than when the client is reporting a weaker RR. The effect size (d) for the relationship between client-rated RR and immediacy resolution is .07, which is considered a small effect, similar to the effect sizes of the relationships between client-rated RR and immediacy depth and immediacy appropriateness. DYAD STUDY OF PSYCHOTHERAPY PROCESS 97 Table 13. RR & Immediacy Resolution Effect Coefficient SE t(12) p d Resolution Intercept 2.62 0.19 13.74 0.000 3.05 Therapist RR -0.01 0.02 -0.59 0.568 -0.01 Client RR 0.06 0.02 2.79 0.017 0.07 Resolution Slope 0.004 0.007 0.54 0.599 0.005 Therapist RR 0.00005 0.0007 0.07 0.943 5.81 Client RR 0.0004 0.0008 0.43 0.673 0.0005 Table 14 presents the results for RR and immediacy quality. Again, there was only one significant and meaningful finding. Therapist-rated RR was not related to average level of immediacy quality, while client-rated RR was found to be significantly related to the average level of immediacy quality. For every 1 point increase in client RR, there is an 0.08 point increase in Immediacy Quality (?02 = 0.08, SE = 0.02, t(12) = 3.865, p < .003). Thus when the client is reporting a stronger RR, the therapist is using higher quality immediacy than when the client is reporting a weaker RR. The effect size (d) for the relationship between client-rated RR and immediacy quality is .09, which is considered a small effect, similar to the effect sizes of the relationships between client- rated RR and immediacy depth, immediacy appropriateness, and immediacy resolution. DYAD STUDY OF PSYCHOTHERAPY PROCESS 98 Table 14. RR & Immediacy Quality Effect Coefficient SE t(12) p d Quality Intercept 3.13 0.19 16.09 0.000 3.37 Therapist RR -0.03 0.02 -1.42 0.182 -0.03 Client RR 0.08 0.02 3.87 0.003 0.09 Quality Slope 0.0002 0.006 0.04 0.971 0.0002 Therapist RR 0.000009 0.0006 0.01 0.990 9.68 Client RR 0.0002 0.0007 0.32 0.758 0.0002 In sum, for real relationship and therapist immediacy, only Hypotheses 3a, 3c, 3e, and 3g were supported among all of the different categories of immediacy (depth, appropriateness, resolution, and quality). In addition, the effect sizes for these relationships are considered small, ranging from .07 to .11. Real Relationship and Client Experiencing Level Hypothesis 3a. The higher the client experiencing level, the stronger the real relationship from the client?s perspective. Hypothesis 3b. The higher the client experiencing level, the stronger the real relationship from the therapist?s perspective. To test these hypotheses, two 3-level HLM analyses were conducted examining the real relationship as the independent variable (level-2 variable; unit = session) and client experiencing level as the dependent variable (level-1 variable; unit = speaking turn). The third-level unit for this model was dyad. One HLM analysis tested the real relationship and peak experiencing level, and the other HLM analysis tested the real relationship and mode experiencing level. The models for each analysis were identical and all tested both hypotheses 3a and 3b. The specific three-level model was: DYAD STUDY OF PSYCHOTHERAPY PROCESS 99 Level 1 Model Y (mode experiencing level) = ?0 + ?1 ? (speaking turn) + e. Level 2 Model ?0 = ?00 + ?01 ? (therapist RR) + ?02 ? (client RR) + r0 ?1 = ?10 + ?11 ? (therapist RR) + ?12 ? (client RR) + r1 Level 3 Model ? 00 = ?000 + ?00 ? 01 = ?010 ? 02 = ?020 ? 10 = ?100 ? 11 = ?110 ? 12 = ?120 Coefficients, standard errors, and t-ratios for fixed effects are presented in Tables 15 and 16 in Appendix K, and L, for Hypotheses 3a and 4b. Table 15 presents the results for RR and Peak Experiencing Level. Average level peak experiencing was 2.17, which was statistically significant, but not meaningful. However, no other results were significant. Table 16, presents the results for RR and Mode Experiencing Level. Similarly, with peak experiencing level, there were no statistically significant or meaningful results. In sum, for real relationship and client experiencing level, neither hypothesis 3a nor 3b were supported. Immediacy and Client Experiencing Level To determine whether change within session in therapist immediacy were related to change in client experiencing level, two 2-level HLM analyses were conducted. The DYAD STUDY OF PSYCHOTHERAPY PROCESS 100 independent variable was therapist immediacy (level-1; unit = speaking turn) and the dependent variable was client post-experiencing level (level-1; unit = speaking turn). An additional variable, prior experiencing level (level-1; unit = speaking turn), was also added to the model. Since these variables were nested, the second-level unit for this model was session. Eight separate HLM analyses were conducted within this set because an analysis was done for each of the four immediacy rating categories separately for mode ratings of client experiencing level (i.e. depth in relation to mode, appropriateness in relation to mode, resolution in relation to mode, and quality in relation to mode) and peak ratings of client experiencing level (i.e. depth in relation to peak, appropriateness in relation to peak, resolution in relation to peak, and quality in relation to peak). However the models for each analysis were identical. The two-level model was specified as follows. Level 1 Model Y (post-experiencing level) = ?0 + ?1 ? (Immediacy) + ?2 ? (prior experiencing level) + r. Level 2 Model ?0 = ?00 + ?0 ?1 = ?10 ?2 = ?20 Coefficients, standard errors, and t-ratios for fixed effects are presented in Tables 17, 18, 19, 20, 21, 22, 23, and 24 for the additional exploratory analysis. It is important to note for clarification, that for all of the additional analyses the client experiencing level that occurred in the speaking turn that occurred immediately before the speaking turn that DYAD STUDY OF PSYCHOTHERAPY PROCESS 101 contained therapist immediacy was labeled ?prior-experiencing? and the experiencing level in the subsequent speaking turn(s), which contained the immediacy event, was labeled ?post-experiencing level?. To further clarify, consider for example this set of speaking turns containing a therapist immediacy event: CL Turn #3: ?I am not sure, but I might be able to find a different day next week to reschedule.? (C-EXP = 2) TH Turn #4: ?I am concerned you are afraid to disappoint me by saying no to me right now, just like with your mom when you didn?t want to go the mall yesterday? What is it like for you to say no to me?? (IMM) CL Turn #4: ?Umm? Yea, I do not want to disappoint you, so I feel some dread and guilt about telling you I cannot reschedule for next Thursday?? (C-EXP = 3) For the client in speaking turn #3, the experiencing rating was a level 2 and the experiencing rating for the client in speaking turn #4 was a level 3. Thus in this example, the therapist immediacy event occurred in speaking turn #4 (initiated by the therapist), so the client?s prior-experiencing was a level 2 and the client?s post-experiencing was a level 3. Mode of client post-experiencing level. Table 17 presents the results for depth of therapist immediacy and mode of client post-experiencing level. All of the results of the analysis were significant. First, the prior speaking turn?s mode of experiencing level is positively related to post-experiencing level (the following speaking turn?s mode of experiencing level) (?20 = 0.58, SE = 0.04, t(359) = 14.81, p < .000). In other words, if the prior mode was higher, then the experiencing mode in the following speaking turn (i.e. post-experiencing level) was higher. Thus, it is likely that if a client was experiencing at a high level, then the following speaking turn was also likely to be high in experiencing level. The effect size for the relationship between prior mode of client DYAD STUDY OF PSYCHOTHERAPY PROCESS 102 experience level and mode of client post-experiencing level is 1.23, which is considered a large effect. Secondly, therapist immediacy depth was also significantly related to mode of client post-experiencing level when controlling for the prior speaking turn?s mode of client experiencing level (?10 = 0.05, SE = 0.02, t(359) = 2.42, p < .016). These results indicate that the higher the depth of immediacy, the greater the post-experiencing level mode even after controlling for the prior mode (of experiencing level). Moreover, the effect size for the relation between immediacy depth and mode of client experiencing level when controlling for the prior speaking turn?s mode of experiencing, was 0.11, which is considered a small effect. Table 17. Immediacy Depth & Post-Experiencing Mode Effect Coefficient SE t(359) p d Experiencing Mode Intercept 2.16 0.03 69.76 0.000 4.60 Immediacy Depth Slope 0.05 0.02 2.42 0.016 0.11 Prior-Experiencing Mode Slope 0.58 0.04 14.81 0.000 1.23 A similar pattern was found for immediacy appropriateness and mode of client post-experiencing level. Table 18 presents the results for appropriateness of therapist immediacy and mode client post-experiencing level. Again, two findings of the results were significant. The prior mode of experiencing level was positively related to the mode of post-experiencing level (?20 = 0.58, SE = 0.04, t(359) = 14.86, p < .000). Therefore, if the prior speaking turn?s mode was higher then the experiencing mode in the following speaking turn was higher. The effect size for this relationship between immediacy appropriateness and mode of client post-experiencing level is 1.23, which is considered a large effect. And secondly, immediacy appropriateness was also DYAD STUDY OF PSYCHOTHERAPY PROCESS 103 significantly related to mode of client post-experiencing level when controlling for the prior speaking turn?s mode of client experiencing (?10 = 0.07, SE = 0.03, t(359) = 2.63, p < .009). Again, these results indicate that the higher the appropriateness of immediacy, the greater the mode of post-experiencing level even after controlling for the prior speaking turn?s mode (of experiencing level). The effect size for this effect is .15, which is considered a small effect. Table 18. Immediacy Appropriateness & Post-Experiencing Mode Effect Coefficient SE t(359) p d Experiencing Mode Intercept 2.16 0.03 80.07 0.000 4.60 Immediacy Appropriateness Slope 0.07 0.03 2.63 0.009 0.15 Prior-Experiencing Mode Slope 0.58 0.04 14.86 0.000 1.23 Table 19 presents the results for resolution of therapist immediacy and mode of client post-experiencing level. Similar to the depth and appropriate analyses, two major findings of the analysis were significant. First, the prior speaking turn?s mode of experiencing level was positively related to client post-experiencing mode (?20 = 0.58, SE = 0.04, t(359) = 14.77, p < .000). Thus, if the prior speaking turn?s mode was higher, then the experiencing mode in the following speaking turn was higher. Similar to the previous immediacy and experiencing level analyses, the effect size for this relation is considered large (d = .13). Secondly, therapist immediacy resolution was also significantly related to mode of client post-experiencing level when controlling for the prior speaking turn?s mode of client experiencing (?10 = 0.06, SE = 0.02, t(359) = 2.46, p < .014). Yet again, this result indicates that the higher the resolution of immediacy, the greater the post-experiencing level mode even after controlling for the prior mode (of DYAD STUDY OF PSYCHOTHERAPY PROCESS 104 experiencing level). The effect size for this relationship between immediacy appropriateness and mode of client post-experiencing level when the prior speaking turn?s mode of client experiencing level is controlled, is .13, a small effect size. Table 19. Immediacy Resolution & Post-Experiencing Mode Effect Coefficient SE t(359) p d Experiencing Mode Intercept 2.16 0.03 84.59* 0.000 4.60 Immediacy Resolution Slope 0.06 0.02 2.46 0.014 0.13 Prior-Experiencing Mode Slope 0.58 0.04 14.77 0.000 1.23 Table 20 presents the results for quality of therapist immediacy and mode of client post-experiencing level. All of the results of the analysis were significant and had the same pattern as depth, appropriateness, and resolution. First, the prior speaking turn?s mode of experiencing level is positively related to the mode of client post-experiencing level (?20 = 0.58, SE = 0.04, t(359) = 14.88, p < .000). Hence, if the prior mode was higher then the experiencing mode in the following speaking turn was higher. Again, the effect size for this relation was large (d = 1.23). Second, therapist immediacy quality was also significantly related to mode of client post-experiencing level when controlling for the prior mode of client experiencing (?10 = 0.05, SE = 0.02, t(359) = 2.31, p < .022). Once more, this finding indicates that the higher the quality of immediacy, the greater the post-experiencing level mode even after controlling for the prior speaking turn?s mode. And again, the effect size for this relation was small (d = .11). DYAD STUDY OF PSYCHOTHERAPY PROCESS 105 Table 20. Immediacy Quality & Post-Experiencing Mode Effect Coefficient SE t(359) p d Experiencing Mode Intercept 2.16 0.03 80.82 0.000 4.60 Immediacy Quality Slope 0.05 0.02 2.31 0.022 0.11 Prior-Experiencing Mode Slope 0.58 0.04 14.88 0.000 1.23 Peak client experiencing level. Table 21 presents the results for quality of therapist immediacy and peak client post-experiencing level. The results for these subsequent four analyses with peak post-experiencing level as the dependent variable were identical to the results of the previous four analyses in which mode of client post- experiencing level was the dependent variable. Again, two of results of the analysis were significant. First, the prior speaking turn?s peak of experiencing level was positively related to the peak client post-experiencing level (?20 = 0.60, SE = 0.04, t(359) = 15.51, p < .000). Thus, if the prior peak was higher then the experiencing peak in the following speaking turn was higher. The effect size of this relation was large (d = 1.22). Secondly, immediacy quality was significantly related to peak post-experiencing level when controlling for the prior speaking turn?s peak experiencing level (?10 = 0.06, SE = 0.02, t(359) = 2.60, p < .012). These results indicate that the higher the quality of immediacy, the greater the post-experiencing level peak even after controlling for the prior speaking turn?s peak. The effect size for this relation was .12, which is considered small. DYAD STUDY OF PSYCHOTHERAPY PROCESS 106 Table 21. Immediacy Depth & Post-Experiencing Peak Effect Coefficient SE t(359) p d Experiencing Peak Intercept 2.18 0.03 67.72 0.000 4.45 Immediacy Depth Slope 0.06 0.02 2.60 0.010 0.12 Prior-Experiencing Peak Slope 0.60 0.04 15.51 0.000 1.22 Table 22 presents the results for appropriateness of therapist immediacy and peak client post-experiencing level. Once again, all of the results of the analysis were significant. First, as in the previous analyses, the prior speaking turn?s peak of experiencing level was positively related to peak client post-experiencing level (?20 = 0.59, SE = 0.04, t(359) = 15.84, p < .000). Hence, the higher the experiencing peak in the prior speaking turn, the higher the experiencing peak in the speaking turn that follows. The effect size for this relationship is 1.20 and it is considered to be a large effect. Secondly, therapist immediacy appropriateness was also significantly related to peak of client post-experiencing level when controlling for the prior speaking turn?s peak of client experiencing (?10 = 0.08, SE = 0.03, t(359) = 2.91, p < .016). Similar to the previous analyses, these results indicate that the higher the appropriateness of immediacy, the greater the post-experiencing level peak even after controlling for the prior speaking turn?s peak. The effect size for this association is .16, which is slightly larger than the one found with depth and peak, but it is still considered to be a small effect. DYAD STUDY OF PSYCHOTHERAPY PROCESS 107 Table 22. Immediacy Appropriateness & Post-Experiencing Peak Effect Coefficient SE t(359) p d Experiencing Peak Intercept 2.17 0.03 77.83* 0.000 4.43 Immediacy Appropriateness Slope 0.08 0.03 2.91 0.004 0.16 Prior-Experiencing Peak Slope 0.59 0.04 15.48 0.000 1.20 Table 23 presents the results for resolution of therapist immediacy and peak client post-experiencing level. Similar to the depth and appropriate analyses, all of the results of the analysis were significant. First, the prior speaking turn?s peak experiencing level was positively related to the client?s post-experiencing level peak (?20 = 0.60, SE = 0.04, t(359) = 15.45, p < .000). Thus, if the prior speaking turn?s peak experiencing level was higher, then the experiencing peak in the following speaking turn was higher. The effect size for this relation is 1.22, which is considered a large effect. Second, immediacy resolution was also significantly related to peak post-experiencing level after controlling for the peak of prior-experiencing level (?10 = 0.06, SE = 0.02, t(359) = 2.58, p < .012). Yet again, these results indicate that the higher the resolution of immediacy, the greater the post-experiencing level peak even after controlling for the peak of the prior speaking turn. The effect size for this relationship was small (d = .12) DYAD STUDY OF PSYCHOTHERAPY PROCESS 108 Table 23. Immediacy Resolution & Post-Experiencing Peak Effect Coefficient SE t(359) p d Experiencing Peak Intercept 2.17 0.03 83.25 0.000 4.43 Immediacy Resolution Slope 0.06 0.02 2.58 0.011 0.12 Prior-Experiencing Peak Slope 0.60 0.04 15.45 0.000 1.22 Table 24 presents the results for quality of therapist immediacy and peak client post-experiencing level. All of the results of the analysis were significant and had the same pattern as depth, appropriateness, and resolution. The two major findings were as follows. First, the prior speaking turn?s peak experiencing level (i.e. prior-experiencing peak) was positively related to post-experiencing level peak (?20 = 0.60, SE = 0.04, t(359) = 15.59, p < .000). Hence, if the experiencing level peak in the prior speaking turn was higher, then the experiencing peak in the following speaking turn was higher. Again, the effect size for this relationship was large (d = 1.22). Second, therapist immediacy quality was also significantly related to peak post-experiencing level after controlling for prior- experiencing level peak (?10 = 0.05, SE = 0.02, t(359) = 2.46, p < .010). Once more, these results indicate that the higher the quality of immediacy, the greater the post- experiencing level peak even after controlling for the prior speaking turn?s peak. The effect size for this relation was .10, which considered a small effect. DYAD STUDY OF PSYCHOTHERAPY PROCESS 109 Table 24. Immediacy Quality & Post-Experiencing Peak Effect Coefficient SE t(359) p d Experiencing Peak Intercept 2.17 0.03 79.53 0.000 4.43 Immediacy Quality Slope 0.05 0.02 2.46 0.015 0.10 Prior-Experiencing Peak Slope 0.60 0.04 15.59 0.000 1.22 Overall for immediacy and experiencing level, two major findings emerged in each sub-analysis. First, for both mode and peak client experiencing level, the prior- experiencing level significantly and positively related to the post-experiencing level. This relation had a consistent and large effect size throughout the sub-analyses. Secondly, all of the associations between therapist immediacy (depth, appropriateness, resolution, and quality) and client post-experiencing level (peak and mode) were significant and positive but had small effect sizes (ranging from .10 to .16). DYAD STUDY OF PSYCHOTHERAPY PROCESS 110 Chapter Five Discussion This chapter will include an overview and discussion of the major findings from the present study, followed by implications for research and practice, as well as limitations. All findings were derived from six clients and their three therapists who had completed at least four sessions of ongoing brief psychotherapy. Given the relatively high mean outcome scores for participants overall, results should not be generalized to participants who had a poor outcome in treatment. Lastly, as this study was a naturalistic, correlational field study, causal inferences or inferences about directionality are not offered. Instead, I describe the relationships between variables observed in the six cases treated at a Mid-Atlantic university health center and a Northeastern university counseling center. Descriptive Findings. Frequency of therapist immediacy. Therapist immediacy was used relatively infrequently, occurring from 0% to 25% of speaking turns in the six cases of brief psychotherapy. The vast majority of therapist immediacy events (43 out of 46) were initiated by Therapist C, the experienced therapist in the study. The remaining three therapist immediacy events were initiated by therapist B in dyad 4, an Asian international therapist-trainee working with an Asian-American male client (Client D). Therapist B did not initiate any immediacy events with her other client (Client C) and Therapist A, also an Asian international therapist-trainee, did not initiate any immediacy events with any of her clients. This is consistent with previous research, which suggests that experienced therapists use immediacy more frequently and effectively than therapist- DYAD STUDY OF PSYCHOTHERAPY PROCESS 111 trainees because trainees were less confident of their ability to use immediacy and less comfortable using immediacy (Hill, Gelso et al. under review; Hill et al., 2009; Kasper et al., 2009; Mayotte-Blum, 2012). In addition to inexperience, the cultural background of the therapist-trainees in this sample may have also affected how frequently immediacy was used. As the two therapist-trainees in this sample were Asian International therapist-trainees, it has been suggested by Hill, Gelso et al. (under review) that immediacy may be contrary to cultures that discourage direct communication about maladaptive behaviors and thus is uncomfortable to use. As such, the trainees may have relied on interventions other than immediacy that were in better harmony with their own cultural experiences and values. However, it is important to note that culture involves many factors and dimensions of a person?s experience and furthermore, the international therapist-trainees? psychotherapy experience using immediacy, theoretical orientation and cultural background was formally assessed with little depth or not at all, so the following cultural consideration is only speculative regarding immediacy usage by the therapist-trainees in this study and highlights only one of many possible factors influencing its use. Therapist immediacy types. Therapist C was responsible for almost all of the immediacy events (43 out of 46 events) in this study. As such, the data on types essentially reflects only therapist C?s usage and allow for some comparison between her and the experienced therapists in previous case studies (Hill et al, 2008; Kasper et al., 2008; Mayotte-Blum et al., 2012). The most common type of immediacy used by therapist C was an exploration of unexpressed or covert feelings (e.g. ?You were unusually late today, and you seem to be DYAD STUDY OF PSYCHOTHERAPY PROCESS 112 looking at your coat more than making eye-contact with me. I wonder if you are uncomfortable being here with me right now?). Used less often was discussion of parallels to other relationships (e.g. ?You said you don?t let yourself get attached because everyone has let you down. I wonder if that is keeping you disconnected from me?). Used least often were discussion of ruptures (?I apologized for my mistaken assumption about your intentions last week. What do you sense is going on between us in today?s session?), and negotiation of tasks and goals (?How do you feel about exploring the pressure you feel to please me instead of answering any difficult questions?). In comparison to the therapists in the Hill et al. (2008) and Kasper et al. (2008) case studies, Therapist C used her types of immediacy in supportive ways (reinforcing the client for something she did, indicating that she wanted to partner/collaborate with her, inquiring about client?s reactions to therapy, or reinforcing that it was okay to disagree with her) similar to Dr. W in the Hill et al. case, even though she used drawing parallels to other relationships and making the covert overt -- possibly more challenging and confrontational types of immediacy ? in a similar frequency to the other more supportive types (that were used by Dr. N in the Kasper et al., case study). Thus, Therapist C used the different types of immediacy with more diversity than experienced therapists in previous research, not highly favoring one type over the other. In addition, in comparison to the experienced therapists in the previous case studies (Hill et al, 2008; Kasper et al., 2008; Mayotte-Blum et al., 2012), Therapist C seemed to use all of the types of immediacy in supportive ways that increased the client?s comfort, despite the challenging and confrontational nature of certain aspects of the immediacy types. For example, the judges often noted Therapist C?s knack of incorporating the client?s DYAD STUDY OF PSYCHOTHERAPY PROCESS 113 wording and manner of expression into her immediacy statements, while positively framing (or reframing) the client?s response in terms of their immediate interaction. To the judges, this seemed to set the client at ease despite the challenging nature of the immediacy event. Thus, her high ratings of appropriateness and diverse use of immediacy types suggests a flexibility with immediacy that helped Therapist C to tailor her use of immediacy to her client?s needs and the needs of their therapeutic relationship. Therapist immediacy dimensions. Almost all of the immediacy events in the current study were judged as a moderately to highly effective. On average, the therapists in the study implemented immediacy events that were moderate in depth (M = 3.15, on a 5-point scale), reached a moderate level of resolution (M = 2.89), were moderately high in quality (M = 3.37), and were highly appropriate (M =3.90). The current sample?s judges? scores reflect that none of the events were considered harmful and that on average the immediacy events were considered valuable and had impact. Client experiencing level (C-EXP). In a typical speaking turn in this study, clients exhibited low levels of experiencing (< 3.0, on a 7-point scale). Thus, clients disclosed in an externalized manner with limited to no involvement of inner self- referential process or exploration of feelings. High levels of experiencing (> than 4.0), which are considered to reflect productive process for a client (Klein et al., 1969), only occurred in about 1% of clients? speaking turns. It is not clear as to what proportion of experiencing should be above 4, as many aspects of the therapy process vary in importance within each client (like building rapport, etc.). Goldman et al. (2005) examined the experiencing level of clients in isolated portions of therapy that involved thematic emotional problems. In that study, the average experiencing level in the later DYAD STUDY OF PSYCHOTHERAPY PROCESS 114 portions of treatment was 4.13. So in building a picture of the how the client?s experiencing emerges in a session, it is possible that aggregating experiencing data that samples all portions of the treatment (both clinically central and non-central aspects) may mask the occurrence of productive process. Furthermore, client experiencing level may best relate to measures of variables that similarly observe and assess phenomena within speaking turns (Hill et al., 1988). Main Findings The Real Relationship and Session Quality The strength of the real relationship was expected to positively relate to session quality because a strong real relationship is theorized to enhance the therapist?s experience of the session and increase therapeutic gain for the client (Gelso, 2011). None of the Hypotheses (1a ? 1d ) regarding real relationship (RR) and session quality were supported. This finding was unexpected because in eight previous studies (Ain & Gelso, 2008, 2011; Fuertes et al., 2007; Gelso et al., 2012; Gullo et al.,2012 ; LoCoco et al., 2011; Markin, Kivlighan, Gelso et al., under review; Marmarosh et al., 2009) RR was found to significantly relate to treatment progress and outcome from either the therapist?s perspective and/or the client?s perspective. However, the mean scores on RR for both clients and therapists were high (> 4.0 on a 5-point scale) and the mean ratings of session quality were high for clients and therapists (>6.0 on a 7point scale). As a result, it is possible that the very high scores indicated a ceiling effect, due to many scores at the upper limit of the RR and session quality measures (Cramer, 2005; Vogt, 2005). This ceiling effect may have reduced the variance of RR and session quality, and reduced variance tends to result in attenuated correlations. (Cramer, 2005; Vogt, 2005). DYAD STUDY OF PSYCHOTHERAPY PROCESS 115 Consequently, the reduced variance may have reduced the sensitivity of the analyses to detect significant correlations between the real relationship and session quality (Cramer, 2005; Vogt, 2005). Another important consideration regarding the lack of detected associations between RR and session quality in the current study is that most other studies that measured RR and session progress or outcome did not use APIM to control for the actor and partner interdependence in client and therapist ratings of RR and session progress/outcome (Ain & Gelso, 2008, 2011; Fuertes et al., 2007; Gullo et al., 2012; LoCoco et al., 2011). Marmarosh et al. (2009) used HLM to control for the nested nature of the data, similar to the current study, but this only allowed the authors to address actor effects. The current study conceptualized the therapy dyads as couple-oriented dyads similar to the previous study by Gelso et al. (2012). As Gelso et al. pointed out, therapy dyads are more likely to be couple-oriented dyads, in which each participant?s perception (therapist or client) of outcome is related to both participants? perceptions (therapist and client) of the real relationship, due to nonrandom dyad composition, sharing the same therapeutic environment, and mutual influence. Perhaps had the other studies controlled for actor-partner independence, they may not have found a significant relationship between RR and outcome. However, recent studies by Gelso et al. (2012) and Markin et al. (under review) used the APIM and found significant results between RR and session quality and treatment progress/outcome, so it is not clear how actor-partner interdepence may or may not have affected results of the previous studies. In addition to the above measurement and analysis considerations, the racial/ethnic minorities (REM) composition of this sample was noteworthy. In DYAD STUDY OF PSYCHOTHERAPY PROCESS 116 comparison to the previous studies (Ain & Gelso, 2008, 2011; Fuertes et al., 2007; Gelso et al., 2012; Gullo et al., 2012; LoCoco et al., 2011; Markin, Kivlighan, Gelso, under review; Marmarosh et al., 2009) that detected significant associations between RR and session quality or treatment progress/outcome, the majority of this sample?s clients and therapists were racial/ethnic minorities (REM). Despite sample differences such as numbers of cases, setting, presenting problems, experience level and theoretical orientation of therapists, in the previous studies the majority of the clients and therapists were European American (ranging from 74% to 100% in each studies? client sample and ranging from 71% to 100% of each studies? therapist sample), whereas in this study only 33% (2 out of 6) of the clients and 33% (1 out of 3) of the therapists were European American. As a result, four of the six dyads in the current study were similar-REM and cross-REM dyads. Multicultural scholars recommend examining the effects of REM psychotherapy dyads given that the interpersonal and sociopolitical contexts of psychotherapy dyads mutually influence each other (Gaztambie, 2011; La Roche, 2005; Owen, Tao, Leach et al., 2011). Hence it is questionable whether and how previous findings regarding RR strength and session progress and outcome should generalize to the present sample with the majority of REM participants. In sum, within this sample, there were no significant associations between the strength of the real relationship and session quality from either the therapist or client perspectives, which is inconsistent with previous investigations. Measurement and analysis differences between the current study and the previous studies regarding a potential ceiling effect in the RR and session quality measures as well as the use of an APIM to address interdependencies in the data may relate to this inconsistency. DYAD STUDY OF PSYCHOTHERAPY PROCESS 117 However, other differences in this sample such as its culturally diverse dyad compositions could also be a factor that reflects a lack of generalizability of the previous significant findings regarding RR strength and session quality. As such, additional research is needed to address these measurement, analysis, and REM sample composition differences between these studies to determine whether and in what samples the significant associations between RR and session progress/outcome are reliable and consistent. The Real Relationship and Therapist Immediacy The strength of the real relationship was expected to positively relate to amount of therapist immediacy because therapist immediacy initiates a self-revealing exchange that likely reduces distortion, enhancing perceptions of realism; and is likely experienced as an invitation to be interact with less fa?ade, defense, and repression, increasing genuineness. The findings of the study indicated mixed support for this prediction. Neither client nor therapist RR in a session was significantly related to either the amount of immediacy or whether immediacy occurred in a session. However, client perceptions of RR strength were significantly associated to the average rating of immediacy quality, appropriateness of immediacy, immediacy depth, and resolution of immediacy. As discussed above in the descriptive findings, the typical immediacy event in the current study was rated moderate to high in depth, appropriateness, resolution and quality ? indicating that in the sessions in which it was used, immediacy was generally effective. Thus, these findings suggest that for clients? perceptions of the real relationship, the amount of immediacy did not matter as long as the immediacy was conducted effectively in terms of quality, depth, appropriateness and resolution. DYAD STUDY OF PSYCHOTHERAPY PROCESS 118 To date, no previous study has examined therapist immediacy in relation to the strength of the real relationship. However, two studies have examined the real relationship in connection to the amount and relevance of self-disclosure in general (recall that therapist immediacy is a specific type of self-disclosure). Ain and Gelso (2008) examined self-disclosure retrospectively from the client?s perspective, and Ain and Gelso (2011) examined self-disclosure from the therapist and client perspectives in ongoing therapy. In the first study (Ain & Gelso, 2008) only relevance of self-disclosure related to the strength of the real relationship from the client?s perspective, while in the second study (Ain & Gelso, 2011) amount of self-disclosure and not relevance of self- disclosure related to the strength of the real relationship for clients (and therapists). The current study provides evidence that for clients? perceptions of RR, at least for a specific type of self-disclosure, namely immediacy, the amount of self-disclosure can vary as long as the self-disclosure was conducted effectively in terms of quality, appropriateness (which involved evaluation of relevance), depth and resolution. In regards to the therapist?s perspective of the strength of the real relationship, no relationship between RR strength and immediacy was detected in the analyses of amount or in relation to any of the immediacy dimensions. There are no previous studies directly investigating therapist RR strength and therapist immediacy that might provide clues as to what might account for the lack of significant association between these two constructs. However, these findings are somewhat inconsistent with previous case studies of immediacy that found that when experienced, interpersonal therapists processed the therapeutic relationship in-session, the therapist rated the work-related aspects of the therapeutic relationship strongly and had positive reactions when DYAD STUDY OF PSYCHOTHERAPY PROCESS 119 immediacy was used (Hill et al., 2008; Kasper et al., 2008; Mayotte-Blum et al., 2012). In those studies, therapy participants provided descriptive reports of what sounded like strong real relationships (e.g. being able to act in ways that were more authentic and genuine), however, RR strength was not quantitatively assessed. Consequently, even though those therapists rated sessions more favorably when immediacy occurred, it is unclear as to whether immediacy affected the therapists? perceptions of the strength of the real relationship. Regarding the current study, I speculate that, considering that 97% of the sessions involving immediacy were conducted by the highly experienced therapist (Therapist C), it is possible that her effective immediacy usage was not more noticeable than the effective usage of other interventions in the session. As a result, her immediacy usage may not have altered her perceptions of RR strength. Perhaps for experienced therapists who thus have a high level of skill with a variety of interventions, immediacy usage may alter how central or prominent the real relationship becomes within the therapy hour, which Gelso (2011) terms real relationship salience? but immediacy may not alter the strength of the real relationship because other interventions (not central to the real relationship) are used with equal effectiveness. However, experience level was only one of many factors that may have confounded the findings. Thus future research is left to determine whether experience level, RR salience, or any other potential factors affected the association between therapist immediacy and therapists? perceptions of real relationship strength in the study. DYAD STUDY OF PSYCHOTHERAPY PROCESS 120 The Real Relationship and Client Experiencing Level (C-EXP) The strength of the real relationship was expected to positively relate to client experiencing level because it is theorized that a client is likely to exhibit a high level of experiencing (i.e. engagement in inner, lived process and openness to their feelings) when the client and therapist perceive their relationship characterized by genuineness, realism, and positive feelings toward one another ? all characteristics of a strong real relationship (Gelso, 2011; Gendlin, 1996; Rogers, 1959). Surprisingly, none of the Hypotheses (4a and 4b) were supported. To my knowledge, no previous research has examined the real relationship and client experiencing level. However, the current study?s lack of significant association between RR strength and C-EXP are inconsistent with previous research examining constructs related to C-EXP, which found the strength of the real relationship significantly related to the tendency to attend to one?s feelings (Fenigstein, Scheier, & Bus, 1975). In addition, the current statistically insignificant results between RR and C- EXP are inconsistent with previous research that found C-EXP to be significantly associated to constructs related to the real relationship: working alliance (Goldman, Greenberg and Pos, 2005), and self-disclosure (Hill et al., 1988). And finally, the lack of significant relationship between RR and C-EXP in the current study is inconsistent with three case studies (Hill et al. 2008; Kasper, Hill & Kivlighan, 2008; Mayotte-Blum et al., 2012) in which the therapy participants provided descriptive reports of what sounded like strong real relationships (e.g. being able to act in ways that were more authentic and genuine). In those studies, what sounded like a strong RR related to what sounded like increased client experiencing (e.g. an increased openness in and ability to express their DYAD STUDY OF PSYCHOTHERAPY PROCESS 121 immediate feelings). Despite, this related literature, I cannot account for the lack of significant associations between the strength of the real relationship and client experiencing level. Additional Analyses ? Therapist Immediacy and Client Experiencing Level In order to shed some light on questions that emerged during this investigation, but were beyond the original scope of this study, additional analyses were conducted examining the association between therapist immediacy and client experiencing level. HLM analyses revealed two major findings that shed light on the use of immediacy and the clients? experience level (i.e. engagement in inner, lived process and openness to their feelings) during treatment in the current sample. First, if a client was experiencing at a high level in the current speaking turn, then the experiencing level in the subsequent speaking turn was also likely to have a high experiencing level. The effect size for this finding was very large. Secondly, between speaking turns, the higher immediacy was in terms of depth, appropriateness, resolution and quality, the greater the client?s experiencing level in the corresponding speaking turn(s), after controlling for the prior speaking turn?s experiencing level (small effect size). It is important to recall that the current speaking turn?s experiencing level was labeled ?prior-experiencing? and the subsequent speaking turn?s experiencing level was labeled ?post-experiencing level? (see Results section, pg. 93-94). As a result, therapist immediacy that was higher in depth, appropriateness, resolution or quality was related to higher post-experiencing levels after controlling for prior-experiencing levels. These findings therefore indicate that despite the low frequency of therapist immediacy and the low overall average of client experiencing level in the current study, therapist immediacy was typically used DYAD STUDY OF PSYCHOTHERAPY PROCESS 122 effectively in this study and its use had immediate impact in the speaking turns in which it occurred by increasing clients? experiencing level. Though no previous study has examined therapist immediacy and client experiencing level, these findings are consistent with previous immediacy case studies (Hill et al. 2008; Kasper, Hill & Kivlighan, 2008; Mayotte-Blum et al., 2012), that qualitatively examined the consequences of immediacy for the client and found descriptive evidence of strong real relationships between the therapist and client and an increased openness in and ability of the clients to express their immediate feelings. Moreover, these findings are also consistent with Hill et al.?s (1988) study, in which clients attained higher levels of experiencing when their prior experiencing level was higher. Hill et al. found that previous client experiencing level accounted for more variance in outcome than therapists? response modes, which is consistent with the effect sizes of the current findings. Implications for Practice and Research The results from the present study provides mixed support for the theoretical (Gelso & Carter, 1985, 1994; Gelso, 2009a; Gelso 2011) and empirical (Ain & Gelso 2008; Eugster & Wampold, 1996; Fuertes et al., 2007; Gelso et. al., 2005; LoCoco et al., 2011; Gullo et al., 2012; Marmarosh et al., 2009; Spiegel et al., 2008) literature on the real relationship and the important role that it plays in psychotherapy. Psychometric issues regarding a ceiling effect in the RR and session quality measures and analysis inconsistencies in the empirical literature regarding potential actor-partner interdependencies, and the lack of time-varying RR measures that can better relate to phenomena observed at the speaking turn level, may at least partially account for the lack DYAD STUDY OF PSYCHOTHERAPY PROCESS 123 of positive findings regarding RR strength and session quality and client experiencing level in the present study. Furthermore, additional development, refinement and validation of the RR measures may improve these issues. However, the overall lack of significant findings in this study between RR strength and most of the other variables in this study (i.e. session quality, and client experiencing level) suggests that within this sample, the associations between the real relationship and other process and outcome variables may have been influenced by a combination of factors not present in previous investigations. Though I can only speculate, my research highlights the need to include racial/cultural themes as an important construct in the real relationship. Multicultural scholars and practitioners have argued that race, ethnicity, gender, socioeconomic status, sexual orientation, and many other sociocultural values and identities of therapists and clients influence the interpersonal interaction in the therapy hour (Gelso, 2011; Gatzambie, 2011; Owen et al., 2011; La Roche, 2012). The racial, ethnic, and SES diversity of the client and therapist sample in my study (compared to samples in previous related research) and the lack of significant findings between RR and session quality and RR and client-experiencing level suggest that some combination of these factors affect how RR strength relates to aspects of process and outcome. In future research, it would be fascinating to examine the strength of the real relationship in dyads that have cultural similarities and differences that are overt (i.e. race, gender, body-image, disability) and covert (i.e. acculturation, sexual orientation, racial/ethnic identity, gender identity, social ally identity). Perhaps culture examined in this way, cultural factors would moderate the associations between RR and other process and outcome variables. DYAD STUDY OF PSYCHOTHERAPY PROCESS 124 The present study also provides support to recent literature that suggests that self- disclosure in general and immediacy in particular, relates to the strength of the real relationship (Ain & Gelso, 2008, 2011; Gelso, 2011). Based on the results from the current study, it appears that from the client perspective, the more that therapist immediacy was appropriate, in-depth, high quality or higher in resolution, the stronger the real relationship regardless of the amount of immediacy used. Though one cannot make causal inferences from correlational data, these results lend support to Gelso?s (2011) proposition that self-disclosures and immediacy not only make the real relationship more salient, but they affect therapy in a healthy way when they are relevant to the client?s needs, well-timed, infrequent, and brief. Regarding therapist perceptions of the real relationship, in particular for the experienced therapist (Therapist C) who used immediacy, the use of immediacy was not associated to RR strength any more than her use of other interventions, though I speculate that it may have been associated to RR salience. Thus, future research should investigate how immediacy usage affects both RR salience and RR strength in therapists that differ in experience level and comprise dyads with diverse training backgrounds. In addition, therapist immediacy should be compared to other interventions that are similar to and different from immediacy to examine how they differentially relate to RR strength. One remarkable distinction between this sample and those of previous studies was the cultural composition of the dyads. Four out of the six dyads were REM similar- and cross-cultural dyads. However (remarkably), the only significant results in the entire study involved therapist immediacy and came from the cases that were European American similar-culture dyads. The results are thus not generalizable to culturally DYAD STUDY OF PSYCHOTHERAPY PROCESS 125 diverse dyads. Given these differences between the cases where I had different cultural pairings, it would be fascinating to look at cultural implications for immediacy. Owen, Imel, Tao et al. (2011) have found evidence that an orientation toward cultural issues in therapy (MCO) is positively related to the strength of the real relationship and that MCO may be a factor in addressing what they term ?cultural ruptures? in cross- or similar culture psychotherapy dyads with REM backgrounds. The authors suggest that therapists disclosing and revealing their own cultural values and acknowledging their clients? cultural background strengthens the real relationship between the therapy participants. Perhaps this is also an example of what cultural immediacy may look like with culturally different psychotherapy dyads. Such cultural applications of therapy process and outcome research seems particularly important when evaluating the effectiveness of therapist interventions and other potential moderating factors, in an attempt to better tailor therapy to fit specific ingredients, circumstances, points in treatment, and outcomes for diverse clients and therapists (Gelso & Palma, 2011; Hill et al., under review). In addition to the future implications that the present study has for subsequent research, the results of the present study also have implications relevant to practitioners. For practitioners, the most important implication is that therapist immediacy is positively associated to the strength of the real relationship in the eyes of clients. Furthermore, amount of immediacy is not what counts for clients? perceptions of the strength of their real relationship. Rather, it is the use of immediacy with any of the four features of appropriateness, depth, quality or resolution that is highly related to client?s perceptions of stronger real relationships. In previous case- and event-analyses of immediacy used in varying amounts (5% to 38%), therapist immediacy was found to help clients negotiate DYAD STUDY OF PSYCHOTHERAPY PROCESS 126 the therapeutic relationship, express feelings about the therapist and therapy, facilitate a corrective relational experience, and help clients feel validated and cared for (Hill et al., 2009, under review; Kasper et al., 2009; Mayotte-Blum, 2012). Considering these results in conjunction with the findings of this study, it is likely that therapist immediacy, if done appropriately, or with depth, or quality or resolution, will increase the strength of the real relationship for clients. Furthermore, considering the amounts of frequency usage in the current and previous studies (ranging from 0% to 38%), in order to have a strong real relationship, the therapist does not have to be frequently disclosing or highly revealing. For therapists who are hesitant to use immediacy due to experience level, training program culture or cultural background incongruent with being more self-revealing, this may be encouraging and helpful to understand. In general, the current study?s findings support recommendations that therapists should use immediacy or self-disclosure judiciously (Hill & Knox, 2009; Gelso, 2011) to best impact the strength of the real relationship for clients. Limitations One of the main limitations of the study was that given the time frame and methods, only a small number of brief psychotherapy cases could be tracked, which limited the power of certain analyses due to variables with low numbers of cases. As it was a naturalistic study, I chose measures that were short, set minimal criteria for client selection (able to complete at least 4 sessions), allowed the therapy to terminate or continue as the therapeutic needs dictated, and asked therapists to simply ask their next client to participate until therapists had each completed two psychotherapy cases. In this way, within one year, dyads formed, developed, and completed their treatment with DYAD STUDY OF PSYCHOTHERAPY PROCESS 127 minimal research influence but with a small number of cases. Thus, though the results may be especially representative of actual therapy conducted by therapist-trainees (doctoral interns) and highly experienced therapists (LCSW) with diverse adult clients, the number of cases limited the power to detect significant associations in the processes and outcome of the therapy sampled in this study. Equally, there are some generalizability limitations within this sample due to the methodological efforts. Typical reasons for why a student would not be able to receive psychotherapy services at the health center and counseling center (and thus could not participate in the study) was that the client was under 18 years of age, was not registered for classes, was suicidal or a danger to others, or if he or she had symptoms so severe they needed hospitalization. Thus, the results may not generalize to those populations or settings outside the university setting. In addition, the therapists for the study consisted of two psychology pre-doctoral interns from a counseling center in the Northeast region, and one highly experienced LCSW from a health center at a Mid-Atlantic university. Therefore the results of this study may only generalize to social workers and psychology interns with similar years of experience and not apply to psychiatrists, counselors, therapists, therapist-trainees or other providers of psychotherapy that come from different training models. Furthermore, the real relationship inventories are still relatively new and not yet extensively validated. Future research is needed to continue to refine and validate these measures to determine whether the significant associations between real relationship strength and therapist immediacy are reliable and consistent patterns. In addition, the real relationship measures have always been given post-session. Perhaps having the measures DYAD STUDY OF PSYCHOTHERAPY PROCESS 128 completed seconds before the session would provide a more accurate assessment of the therapy participants? perception of RR strength. A pre-session completion of the RR inventories would assess the perception of the RR that the participants carried with them between sessions and the perception that most likely will affect the current session processes and session outcomes. One limitation regarding the consensus rating of therapist immediacy was that other than the first author, the judges only watched the sessions in which immediacy occurred. This limited the judges? contextual understanding of what was happening in the session. Often, phenomena in therapy happen in the short-hand, and thus the judges may have missed important pieces of therapy that was happening when no immediacy occurred and not fully understand what was happening in the sessions they observed and rated. Another limitation was the session quality measure. It is clearly a limitation that the measure was single-item. However, as indicated earlier, it had been used in several studies and yielded theoretically reasonable results. Still it was a single item, and the limitations of a single item are well known. The correlational nature of this study is another limitation. Consequently, conclusions about causality cannot be drawn. Furthermore, although some of the discussion above implies that relationships among variables work in one direction, there is also the possibility that they work in the opposite direction. Thus when interpreting the findings, it is also important to consider different ways the variables relate to one another. Another limitation regarding data collection, was that apart from a brief demographic questionnaire, therapists were not assessed pre-or post-treatment in an in- DYAD STUDY OF PSYCHOTHERAPY PROCESS 129 depth manner about their background as a part of this study. This prevented the collection of detailed information about their therapeutic approach, cultural background, and experience and skill level. Thus no in-depth conclusions involving these factors could be drawn. Similarly, clients were not assessed in an in-depth manner about their presenting problems, cultural background or previous experience in therapy before or after treatment, so the results cannot be interpreted in terms of problem severity and client improvement. Again, though this had the advantage of encouraging clients and therapists to readily participate, it limited the scope of the study?s conclusions. In particular, in-depth cultural information for the clients and therapists would have provided relevant insight into how the real relationship, therapist immediacy, client experiencing level, and session quality related to cross- and similar-cultural therapy dyads in this sample. Recent research has shown a link between multicultural orientation (MCO) and the therapeutic relationship. Specifically, MCO was associated to stronger real relationships and to stronger working alliances. Even further, in the Owen et al. (2011) study stronger alliances were shown to ?buffer? the negative effects of cultural impasses on treatment outcome. Since my sample contained cross- and similar cultural dyads, more detailed information regarding cultural background would have been of great interest in order to examine how cultural factors may have moderated or mediated the associations between real relationship strength and session outcome as in the Owen et al. (2011) study. In addition, the real relationship and therapist immediacy appear to be promising areas to examine how ?cultural ruptures? (in which both the interpersonal and sociocultural patterns of the therapy participants conflict and the therapeutic trust and safety for the client is decreased) can be reduced and/or repaired DYAD STUDY OF PSYCHOTHERAPY PROCESS 130 Despite the above limitations, the present study presents important information about the real relationship, therapist immediacy, client experiencing level, and session quality. Although the results may not generalize to all populations, it is still important to learn from the experiences, outcomes, and culturally diverse backgrounds of the clients and therapists in the cases. Furthermore, this study provides new information about the real relationship and therapist immediacy that can inform researchers and practitioners. Specifically, given the major findings that emerged in this study and previous studies (Hill et al., 2009; Hill, Gelso et al., under review; Kasper et al., 2009; Mayotte-Blum, 2012), how therapists use immediacy (i.e. depth, quality, resolution, and appropriateness), appears to be more important than how much (i.e. amount) therapists use immediacy in fostering strong real relationship perceptions in their clients. Future research will hopefully provide further examination of whether or not these results extend to additional populations. DYAD STUDY OF PSYCHOTHERAPY PROCESS 131 Appendix A Pre-Session Client Demographic Questionnaire Instructions: Please respond by checking off answers in the space provided. Gender ___ a. Female ___ b. Male Age: ____ Please indicate the racial or ethnic group that best describes you ___a. Black/African-American (non-Hispanic) ___b. Native American/American Indian/Alaska Native ___c. Asian American /Pacific Islander ___d. Latino/Hispanic ___e. White/Caucasian (non-Hispanic) ___f. Ethnicity/race not included above (please specify: ________________________) Marital Status ___ a. Married ___ b. Divorced ___ c. Separated ___ d. Widow ___ e. Committed relationship, but not married Total annual family income ___ a. Less than $20,000 ___ b. $20,000-$50,000 ___ c. $50,000-$80,000 ___ d. $80,000-$100,000 ___ e. More than $100,000 Education ___ a. High School ___ b. College ___ c. Graduate School ___ d. Technical Training DYAD STUDY OF PSYCHOTHERAPY PROCESS 132 Appendix B Pre-Session Therapist Demographic Questionnaire Instructions: Please respond by checking off answers in the space provided. 1. Gender ___ a. Female ___ b. Male 2. Age: _____ 3. Please indicate the racial or ethnic group that best describes you: ___ a. Black/African-American (non-Hispanic) ___ b. Native American/American Indian/Alaska Native ___ c. Asian American /Pacific Islander ___ d. Latino/Hispanic ___ e. White/Caucasian (non-Hispanic) ___ f. Ethnicity/race not included above (please specify): ___________________ 4. Education ____ a. MS/MA ____ b. MSW ____ c. PhD, PsyD, EdD Psychologist ____ d. Other (please specify) _________________ 5. How many years of experience do you have as a psychotherapist? _____ years 6. Please rate the extent to which you believe in and adhere to the theory and techniques of the following therapies: Low High Psychoanalytic/Psychodynamic 1 2 3 4 5 Humanistic/Existential 1 2 3 4 5 Cognitive/Behavioral 1 2 3 4 5 DYAD STUDY OF PSYCHOTHERAPY PROCESS 133 Appendix C Real Relationship Inventory-Client Form (RRI-C) Instructions: Please use the following scale to evaluate your perceptions of yourself, your therapist, and your relationship with your therapist, placing your rating in the space next to the item. Strongly Disagree Disagree Neutral Agree Strongly Agree 1 2 3 4 5 ____ 1. I am able to be myself with my therapist. ____ 2. My therapist and I have a realistic perception of our relationship. ____ 3. I hold back significant parts of myself. ____ 4. I appreciate being able to express my feelings in therapy. ____ 5. My therapist likes the real me. ____ 6. It is difficult to accept who my therapist really is. ____ 7. I am open and honest with my therapist. ____ 8. My therapist's perceptions of me seem colored by his or her own issues. ____ 9. The relationship between my therapist and me is strengthened by our understanding of one another. ____ 10. My therapist seems genuinely connected to me. ____ 11. I am able to communicate my moment-to-moment inner experience to my therapist. ____ 12. My therapist holds back his/her genuine self. ____ 13. I appreciate my therapist?s limitations and strengths. ____ 14. We do not really know each other realistically. ____ 15. My therapist and I are able to be authentic in our relationship. ____ 16. I am able to see myself realistically in therapy. ____ 17. My therapist and I have an honest relationship. ____ 18. I am able to separate out my realistic perceptions of my therapist from my unrealistic perceptions. ____ 19. My therapist and I have expressed a deep and genuine caring for one another. ____ 20. I have a realistic understanding of my therapist as a person. ____ 21. My therapist does not see me as I really am. ____ 22. I feel there is a significant holding back in our relationship. ____ 23. My therapist?s perceptions of me are accurate. ____ 24. It is difficult for me to express what I truly felt about my therapist. DYAD STUDY OF PSYCHOTHERAPY PROCESS 134 Appendix D Real Relationship Inventory-Therapist Form (RRI-T) Instructions: Please complete the items below in terms of your relationship with your client or patient. Use the following scale in rating each item, placing your rating in the space adjacent to the item. Strongly Disagree Disagree Neutral Agree Strongly Agree 1 2 3 4 5 ____ 1. My client is able to see me as a real person separate from my role as a therapist. ____ 2. My client and I are able to be genuine in our relationship. ____ 3. My client feels liking for the ?real me.? ____ 4. My client genuinely expresses his/her positive feelings toward me. ____ 5. I am able to realistically respond to my client. ____ 6. I hold back significant parts of myself. ____ 7. I feel there is a ?real? relationship between us aside from the professional relationship. ____ 8. My client and I are honest in our relationship. ____ 9. My client has little caring for who I ?truly am.? ____ 10.We feel a deep and genuine caring for one another. ____ 11. My client holds back significant parts on him/herself. ____ 12. My client has respect for me as a person. ____ 13. There is no genuinely positive connection between us. ____ 14. My client?s feelings toward me seem to fit who I am as a person. ____ 15. I do not like my client as a person. ____ 16. I value the honesty of our relationship. ____ 17. The relationship between my client and me is strengthened by our understanding of one another. ____ 18. It is difficult for me to express what I truly felt about my client. ____ 19. My client has unrealistic perceptions of me. ____ 20. My client and I have difficulty accepting each other as we really are. ____ 21. My client distorts the therapy relationship. ____ 22. I have difficulty being honest with my client. ____ 23. My client shares with me the most vulnerable parts of him/herself. ____ 24. My client has genuinely expressed a connection to me. DYAD STUDY OF PSYCHOTHERAPY PROCESS 135 Appendix E Session Quality (SQ) Post-Session Questionnaire: Instructions: Using the scale above, please rate the overall quality of today?s session. 1 2 3 4 5 Very Poor Poor Neutral Good Very Good DYAD STUDY OF PSYCHOTHERAPY PROCESS 136 Appendix F Counseling Outcome Measure (COM) Post-Therapy Questionnaire: (To be completed after last session) Instructions: Please use the following scale to evaluate the amount of self-improvement in the following areas since the beginning of therapy. 1. Compared to when I entered therapy, my feelings are ______ much worse moderately worse slightly worse about the same slightly better moderately better much better 1 2 3 4 5 6 7 2. Compared to when I entered therapy, my behavior is ______ much less effective moderately less effective slightly less effective no change slightly more effective moderately more effective much more effective 1 2 3 4 5 6 7 3. Compared to when I entered therapy, my self-understanding is _____ much worse moderately worse slightly worse about the same slightly better moderately better much better 1 2 3 4 5 6 7 4. Compared to when I entered therapy, I am ____ much worse moderately worse slightly worse about the same slightly better moderately better much better 1 2 3 4 5 6 7 DYAD STUDY OF PSYCHOTHERAPY PROCESS 137 Appendix G Table 7. Frequency Summary of Immediacy Statements for each Dyad. Dyad 1 2 3 4 5 6 Total No. of Sessions 12 12 3 9 6 8 50 No. of Immediacy Events 0 0 0 3 19 24 46 No. of IMM-STs 0 0 0 13 86 265 363 IMM-STs % per Dyad 0% 0% 0% 1.1% 14.9% 25.1% 8.2% IMM-STs % per Therapist 0% 1.06% 15.02% No. Total Statements 624 960 171 1058 577 1054 4444 DYAD STUDY OF PSYCHOTHERAPY PROCESS 138 Appendix H Table 25. Correlation Matrix for Real Relationship, Therapist Immediacy and Client Experiencing Level. RRI-T RRI-C SQ-T SQ-C IMM-Amt IMM- D IMM- App IMM- R IMM- Q EXP- M EXP- P RRI-T 1 RRI-C .54*** 1 SQ-T .31** .24* 1 SQ-C .34** .49*** .13 1 IMM-Amt -.44* .34 -.70 .23 1 IMM-D .04 .39 -.09 .36 .27 1 IMM-App -.05 .63*** .07 .49* .06 .75*** 1 IMM-R .14 .55** .34 .57** .15 .94*** .81*** 1 IMM-Q .05 .64*** .22 .61** .18 .91*** .90*** .96*** 1 EXP-M .16 .40*** .07 .38*** -.25 .38 .52** .49* .52** 1 EXP-P -.18 .17 .07 .27* -.24 .39 .53** .49* .52** .85*** 1 Note: RRI-T = therapist-rated real relationship; RRI-C = client-rated real relationship; SQ-T = therapist-rated session quality; SQ-C = client-rated session quality; IMM-Amt = amount of immediacy; IMM-D = depth of immediacy; IMM-App = appropriateness of immediacy; IMM-R = resolution of immediacy; IMM-Q = quality of immediacy; EXP-M = mode of client experiencing level; EXP-P = peak of client experiencing level *p < 1.0, **p < .05, ***p < .01 DYAD STUDY OF PSYCHOTHERAPY PROCESS 139 Appendix I Table 9. Real Relationship & Therapist Immediacy Amount Effect Coefficient SE t(5) p d Immediacy Amount Intercept 7.50 5.42 1.38 0.225 0.54 Therapist Real Relationship Slope -0.12 0.27 -0.43 0.684 -0.009 Client Real Relationship Slope 0.40 0.35 1.13 0.31 0.03 DYAD STUDY OF PSYCHOTHERAPY PROCESS 140 Appendix J Table 10. Real Relationship & Therapist Immediacy Occurrence Effect Coefficient SE t(5) p d Immediacy Occurrence Intercept -1.57 1.70 -0.92 0.399 -3.41 Therapist Real Relationship Slope -0.05 0.21 -0.25 0.813 -0.11 Client Real Relationship Slope -0.14 0.16 -0.88 0.385 -0.41 DYAD STUDY OF PSYCHOTHERAPY PROCESS 141 Appendix K Table 15. RR & Peak Client Experiencing Level Effect Coefficient SE t(47) p d Peak Intercept 2.17 0.06 35.64* 0.000 5.05 Therapist RR -0.0004 0.005 -0.09 0.930 -0.0009 Client RR 0.0003 0.004 0.07 0.943 0.0007 Peak Slope -0.0006 0.001 -0.57 0.571 -0.001 Therapist RR 0.00009 0.0001 0.71 0.484 0.0002 Client RR 0.00009 0.00009 1.00 0.322 0.0002 DYAD STUDY OF PSYCHOTHERAPY PROCESS 142 Appendix L Table 16. RR & Client Experiencing Level - Mode Effect Coefficient SE t(57) p d Mode Intercept 2.09 0.05 44.35 0.000 6.15 Therapist RR 0.002 0.004 0.42 0.673 0.006 Client RR 0.002 0.003 0.71 0.480 0.006 Mode Slope 0.00009 0.0009 0.10 0.918 0.0003 Therapist RR 0.0001 0.0001 0.97 0.337 0.0003 Client RR 0.00009 -0.00008 -1.10 0.278 0.0003 DYAD STUDY OF PSYCHOTHERAPY PROCESS 143 References Ain, S. & Gelso, C. J. (2008). Chipping away at the blank screen: Self-disclosure, the real relationship, and therapy outcome. Poster submitted for presentation at the Annual Convention of the North American Society for Psychotherapy Research, New Haven, Connecticut, September 2008. Ain, S., & Gelso, C. J. (2011). Client and therapist perceptions of the real relationship and therapist self-disclosure: A study of psychotherapy dyads. Paper presented at the 2011 Convention of the North American Society for Psychotherapy Research, Banff, Canada. Beck, A., Freeman, A., & Associates (1990). Cognitive therapy of personality disorder. New York: Guilford Press. Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research, and Practice, 16, 252-260. Bernick, N. and Oberlander, M.(1969), Effect of verbalization and two different modes of experiencing on pupil size. Perception and Psychophysics, 3, 27?39. Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research, and Practice, 16, 252-260. Cashdan, S. (1988). Object relations therapy: Using the relationship. New York: Norton. Castonguay, L., Goldfried, M. R., Wiser, S., Raue, P. J., & Hayes, A. (1996). Predicting the effect of cognitive therapy for depression: A study of unique and common factors. Journal of Consulting and Clinical Psychology, 64, 474-481. Cramer, D., Howitt, D.L. (2005). The SAGE Dictionary of Statistics: A Practical Resource for Students in the Social Sciences (Third ed.)(pp.21). New York: Sage DYAD STUDY OF PSYCHOTHERAPY PROCESS 144 Crits-Christoph, P., & Gibbons, B.B.C. (2002. Relational interpretations. In J.C. Norcross (Ed.). Psychotherapy relationships that work: Therapist contributions and responsiveness to patients (pp.285-300). New York: Oxford University Press. Derogatis, L.R., & Melisaratos, N. (1983). The Brief Symptom Inventory: An introductory report. Psychological Medicine, 13, 595-605. Don, N.S. (1977). The transformation of conscious experience and its EEG correlates. Journal of Altered States of Consciousness, 3, 147?68. Elliot, R.R. (1986). Intrapersonal process recall (IPR) as a psychotherapy process, research method. In L. Greenberg & W. Pinsoff (Eds.), The psychotherapeutic process (pp. 503- 528). New York: Guilford. Eugster, S. L. and B. E. Wampold (1996). Systematic effects of participant role on evaluation of the psychotherapy session. J Consult Clin Psychol 64(5): 1020-1028. Fitzpatrick, M.R., Peternelli, L., Stalikas, A., & Iwakabe, S. (1999). Client emotional involvement and occurrence of in-session therapeutic phenomena. Canadian Journal of Counselling, 33(3), 179-194. Fitzpatrick, M.R., Stalikas, A., & Iwakabe, S. (2001). Examining counselor interventions and client progress in the context of the therapeutic aliance. Psychotherapy, 2(38), 160-170. Foreman, S. A., & Marmar, C.R. (1985). Therapist actions that address initially poor therapeutic alliances in psychotherapy. American Journal of Psychiatry, 142, 922-926. Freud, A. (1954). The widening scope of indications for psychoanalysis. Journal of the American Psychoanalytic Association, 2, 607-620. DYAD STUDY OF PSYCHOTHERAPY PROCESS 145 Fuertes, J. N., A. Mislowack, et al. (2007). Correlates of the real relationship in psychotherapy: A study of dyads. Psychotherapy Research 17(4): 423-430. Fuertes, J.N., Gelso, C.J., Owen, J.J., Cheng, D. (under review). Real relationship, working alliance, transference/countertransference and outcome in time-limited counseling and psychotherapy. Manuscript submitted for publication. Gaztambie, D.J. (2012) Addressing Cultural Impasses With Rupture Resolution Strategies: A Proposal and Recommendations. Professional Psychology: Research and Practice. 43(3) 183?189. Gelso, C. J. (2002). Real relationship: The ?something more? of psychotherapy. Journal of Contemporary Psychotherapy, 32, 35 ? 40. Gelso, C. J. (2003). Measuring the real relationship: Theoretical foundation. Paper presented at the Annual Convention of the American Psychological Association, Toronto, Ontario, Canada, August 2003. Gelso, C. J. (2004). A theory of the real relationship in psychotherapy. Paper presented at the International conference of the Society for Psychotherapy Research, Rome, Italy, June 2004. Gelso, C. J. (2006). Emerging findings about the real relationship in psychotherapy: Introductory remarks. Paper presented at the international conference of the Society for Psychotherapy Research, Edinburgh, Scotland, June 2006. Gelso, C. (2009a). The real relationship in a postmodern world: Theoretical and empirical explorations. Psychotherapy Research, 19(3), 253-264 DYAD STUDY OF PSYCHOTHERAPY PROCESS 146 Gelso, C. (2009b). The time has come: The real relationship in psychotherapy research. Psychotherapy Research, 19(3), 278-282. Gelso, C.J. (2011). The real relationship in psychotherapy: The hidden foundation of change. Washington, DC: American Psychological Association. Gelso, C. J., & Bhatia, A. (2012). Crossing theoretical lines: The role and effect of transference in nonanalytic psychotherapies. Psychotherapy, 49, 384-390. doi: 10.1037/a0028802 Gelso, C. J., Kivlighan, D. M., Busa-Knepp, J. Spiegel, E. B., Ain, S., Hummel, A. M., Ma, Y. E., and Markin, R. D. (2012). The unfolding of the real relationship and the outcome of brief psychotherapy. Journal of Counseling Psychology, 59. 495-506. doi: 10.1037/a0029838 Gelso, C. J. & Carter, J. A. (1985). The relationship in counseling and psychotherapy: Components, consequences, and theoretical antecedents. The Counseling Psychologist, 13, 155 ? 244. Gelso, C. J. and Carter J. A. (1994). Components of the psychotherapy relationship: Their interaction and unfolding during treatment. Journal of Counseling Psychology 41(3): 296-306. Gelso, C. J. & Hayes, J. A. (1998). The psychotherapy relationship: Theory, research, and practice. NY: Wiley. Gelso, C. J. & Hayes, J. A. (2002). The management of countertransference. In J. Norcross (Ed.), Psychotherapy relationships that work (pp. 267-284). New York, NY: Oxford University Press. DYAD STUDY OF PSYCHOTHERAPY PROCESS 147 Gelso, C. J. & Hayes, J. A. (2007). Countertransference and the therapist?s inner experience: Perils and possibilities. New York, NY: Routledge. Gelso, C.J., Hill, C.E., Kivlighan, D.M. (1991). Transference, insight, and the counselor?s intentions during a counseling hour. Journal of Counseling and Development 59, 428- 433. Gelso, C.J. & Johnson, D.H. (1983). Explorations in time-limited counseling and psychotherapy. New York: Columbia University, Teachers College Press. 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. Gelso, C. J., Kivlighan, D. M., Wine, B., Jones, A., & Friedman, S. C. (1997). Transference, insight, and the course of time-limited therapy. Journal of Counseling Psychology, 44, 209?217. Gelso, C. J. & Palma, B. (2011). Directions for research on self-disclosure and immediacy: Moderation, mediation, and the inverted U. Psychotherapy, 48, 342-348. doi: 10.1037/a0025909 Gelso, C. J. & Samstag, L. W. (2008). A tripartite model of the therapeutic relationship. In S. Brown & R. Lent (Eds.), Handbook of counseling psychology. (pp. 267 ? 283). NY: Wiley. Gendlin, E.T. (1962). Experiencing and the creation of meaning. New York: Free Press. Gendlin, E.T. (1968). The experiential response. In Hammer (Ed.), Interpretation in therapy: Its role, scope, depth, timing and art. New York: Grune and Stratton. DYAD STUDY OF PSYCHOTHERAPY PROCESS 148 Gendlin, E.T. (1996). Focusing-oriented psychotherapy: A manual of the experiential method. New York: Guilford Press. Gendlin, E.T. & Berlin, J.I. (1961). Galvanic skin response correlates of different modes of experiencing. Journal of Clinical Psychology, 17 (1), 73?77. Goldman, R.N., Greenberg, L.S & Pos, A. (2005). Depth of emotional experience and outcome, Psychotherapy Research, 15(3): 248-60. Graff, H., & Luborsky, L.L. (1977). Long-term trends in transference and resistance: A report on a quantitative method applied to four psychoanalyses. Journal of the American Psychoanalytic Association, 24, 471-490. Greenson, R. R. (1978). Explorations in psychoanalysis. New York, International Universities Press. Greenson, R. R. (1967). The technique and practice of psychoanalysis. New York, International Universities Press. Greenson, R. R. , & Wexler, M. (1969). The non-transference relationship in the psychoanalytic situation. International Journal of Psycho-Analysis, 50, 27-39. Hendricks, M. (2009). Experiencing level: An instance of developing a variable from a first person process so it can be reliably measured and taught. Journal of Consciousness Studies, 16(10-12), 129-155. Hendricks, M. (2002). Focusing-oriented/experiential psychotherapy. In Cain, David J. (Eds.): Humanistic psychotherapies: Handbook of research and practice. Washington, DC: American Psychological Association (pp. 221-251). Hill, C. E. (2004). Helping skills: Facilitating, exploration, insight, and action (2nd ed.) DYAD STUDY OF PSYCHOTHERAPY PROCESS 149 Washington, DC: American Psychological Association Hill, C. E. (2009). Helping skills: Facilitating, exploration, insight, and action (3rd ed.). Washington, DC US, American Psychological Association. Hill, C. E., Helms, J. E., Tichenor, V., Spiegel, S. B., O'Grady, K. E., & Perry, E. S. (1988). Effects of therapist response modes in brief psychotherapy. Journal of Counseling Psychology, 35, 222-233. Hill, C. E., Gelso, C.J., Chui, H., Spangler, P., Hummel, A., Huang, T., Jackson, J., Jones, R., Palma, B., Bhatia, A., Shudarshana, G., Ain, S., Klingaman, B., Lim, R.H., Liu, J., Hui., K., Jezzi, M., & Miles, J.R. (Under Review). To be or not to be immediate with clients: The use and perceived effects of immediacy in psychodynamic/interpersonal psychotherapy. Hill, C. E., I. S. Kellems, et al. (2003). The Therapist Experience of Being the Target of Hostile Versus Suspected-Unasserted Client Anger: Factors Associated with Resolution. Psychotherapy Research 13(4): 475-491. Hill, C. E. and S. Knox (2009). Processing the therapeutic relationship. Psychotherapy Research 19(1): 13-29. 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. Horowitz, L.M., Rosenberg, S.E., Baer, B.A., Ureno, G., & Villasenor, V.S. (1988). Inventory of interpersonal problems: Psychometric properties and clinical applications. Journal of Consulting and Clinical Psychology, 56, 885-892. DYAD STUDY OF PSYCHOTHERAPY PROCESS 150 Horvath, A.O., & Bedi, R.P. (2002). The alliance. In J.C. Norcross (Ed.), Psychotherapy relationships that work: Therapist contributions and responsiveness to patients (pp.37- 70). New York: Oxford. Horvath, A.O., & Symonds, B. D. (1991). Relation between working alliance and outcome in psychotherapy: A meta-analysis. Journal of Counseling Psychology, 38, 139-149. Ivey, A.E. (1994). Intentional interviewing and counseling: Facilitating client development in a multicultural society (3rd Eds.). Pacific Grove, CA: Brooks/Cole. Kasper, L. B., Hill, C. E., Kivlighan, D.M. (2008). Therapist immediacy in brief psychotherapy: Case study I. Psychotherapy: Theory, Research, Practice, Training 45(3): 281-297. Kafuman, A.S. (2009). IQ Testing 101. New York: Springer Publishing. pp. 151?153 Kelley, F. A., Gelso, C. J., Fuertes, J. N., Marmarosh, C., & Lanier, S. (2010). The Real Relationship Inventory: Development and psychometric investigation of the Client Form. Psychotherapy Theory, Research, Practice, Training.47(4), 540-553. Kenny, D. A., & Cook, W. (1999). Partner effects in relationship research: Conceptual issues, analytic difficulties, and illustration. Personal Relationships, 6, 433?448. doi:10.1111/j.1475-6811.1999.tb00202.x Kiesler, D. J. (1971). Patient experiencing level and successful outcome in individual psychotherapy of schizophrenics and psychoneurotics. Journal of Clinical and Consulting Psychology, 37, 370-385. Kiesler, D.J. (1988). Therapeutic metacommunication: Therapist impact disclosure as feedback in psychotherapy. Palo Alto, CA: Consulting Psychologists Press. DYAD STUDY OF PSYCHOTHERAPY PROCESS 151 Kiesler, D. J., & Watkins, K. (1989). Interpersonal complementarity and the therapeutic alliance: A study of relationship in psychotherapy. Psychotherapy: Theory, Research, and Practice, 26, 183-194. Klein, M.H., Mathieu, P.L., Gendlin, E.T., & Kiesler, D.J. (1969). The Experiencing Scale: A research and training manual (Madison, WI: Wisconsin Psychiatric Institute). Klein, M.H., Mathieu-Coughlan, P. & Kiesler, D.J. (1986). The Experiencing Scales in The psychotherapeutic process: A research handbook. pp. 21-71. 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, 274-283. Lo Coco, G., Prestano, C., Gullo, S., & Gelso, C.J. (2011). Relation of the real relationship and the working alliance to symptom change in brief psychotherapy. Psychotherapy 48(4), 359?367. La Roche, M. J. (2005). The cultural context and the psychotherapeutic process: Toward a culturally sensitive psychotherapy. Journal of Psychotherapy Integration, 15, 169?185. Lambert, M. & D. Barley (2002). Research summary on the therapeutic relationship and psychotherapy outcome. New York, Oxford University Press. Lambert, M.J., & Hill, C.E. (1994). Assessing psychotherapy outcomes and proceses. In A.E. Bergin & S. L. Garfield (Eds.). Handbook of psychotherapy and behavior change. New York: John Wiley & sons, Inc. DYAD STUDY OF PSYCHOTHERAPY PROCESS 152 Lutgendorf, S.K., Antoni, M.H., Kumar, M. & Schneiderman, N. (1994). Changes in cognitive coping strategies predict EBV-antibody titre change following a stressor disclosure induction. Journal of Psychosomatic Research, 38, 63-78. Markin, R., Gelso, C.E., Kivlighan, D.M., Busa-Knepp, J., Spiegel, E. B., Ain, S., Hummel, A. M., Ma, Y. E. (Under Review). Clients? and therapists? real relationship and session outcome: An actor partner interdependence analysis. 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. Mayotte-Blum, J., Slavin-Mulford, J., Lehmann, M., Pesale, F., Becker-Matero, N., & Hilsenroth, M. (2012). Therapeutic immediacy across long-term psychodynamic psychotherapy: An evidence-based case study. Journal of Counseling Psychology, 59, 27-40. doi: 10.1037/a0026087 Moore, S., & Gelso, C.J. (2010). Attachment in psychotherapy: Considerations of the real relationship. Psychotherapy 48(4), 368?373. Multon, K.D., Patton, M.J., & Kivlighan, D.M. (1996). Development of the Missori Identifying Transference Scale. Journal of Counseling Psychology, 43, 243-252. Neisser, U. (1967). Cognitive psychology. New York, NY: Appleton-Century-Crofts. Norcross, J. C. (2002). Psychotherapy relationships that work: Therapist contributions and responsiveness to patients. New York, Oxford University Press. DYAD STUDY OF PSYCHOTHERAPY PROCESS 153 Orlinsky, D.E., & Howard, K.I. (1986). Process and outcome in psychotherapy. In S. L. Garfield & A. E. Bergin (Eds.), Handbook of psychotherapy and behavior change (3rd Ed., pp. 311-381). New York: Wiley. Owen, J., Imel, Z., Tao, K. W., Wampold, B., Smith, A., & Rodolfa, E. (2011). Cultural ruptures in short-term therapy: Working alliance as a mediator between clients? perceptions of microaggressions and therapy outcomes. Counseling and Psychotherapy Research, 11, 204?212. Owen, J., Tao, K., Leach, M. M., & Rodolfa, E. (2011). Clients? perceptions of their psychotherapist?s multicultural orientation. Psychotherapy, 48, 274?282. Owen, J., Tao, K., & Rodolfa, E. (2010). Microaggressions and women in short-term psychotherapy: Initial evidence. The Counseling Psychologist, 38, 923?946. Patton, M.J., Kivlighan, D.M., Jr., & Multon, K.D. (1995). Missouri psychoanalytic counseling research project: First findings. Unpublished manuscript, University of Missouri- Columbia. Rogers, C. (1961). On becoming a person: A therapist's view of psychotherapy. Houghton Mifflin. Rhodes, R.H., Hill, C.E., Thompson, B.J., & Elliott, R. (1994). Client retrospective recall of resolved and unresolved misunderstanding events. Journal of Counseling Psychology, 41, 473-483. Safran, J., J. C. Muran, et al. (2002). Repairing alliance ruptures. In J.C. Norcross (Ed.), Psychotherapy relationships that work: Therapist contributions and responsiveness to patients (pp. 235-255). New York, Oxford University Press. DYAD STUDY OF PSYCHOTHERAPY PROCESS 154 Safran, J. D., J. C. Muran, et al. (2001). Repairing alliance ruptures. Psychotherapy: Theory, Research, Practice, Training 38(4): 406-412. Samstag, L.W. (2006). The working alliance in psychotherapy: An overview of the invited papers in the special section. Psychotherapy: Theory, Research, Practice, Training, 43(3). Spangler, P., Hill, C. E., Dunn, M. G., Hummel, A., Liu, J., Walden, T., Jackson, J., Ganginis, H., & Salahuddin, N. (Under review). Helping in the here-and-now: Teaching undergraduates to use immediacy. Spiegel, E.B., Busa-Knepp, J., Ma, E., Markin, R.D., Ain, S., Hummel, A., ... Gelso, C.J.(2008, August). Unfolding of the real relationship and its connection to outcome. Paper presented at the 116th Annual Convention of the American Psychological Association, Boston, MA. Stalikas, A., & Fitzpatrick, M. (1995). Relationships between counsellor interventions, client experiencing, and emotional expressiveness: An exploratory study. Canadian Journal of Counseling, 30(4), 262-271. Stalikas, A., & Fitzpatrick, M. (1995). Client good moments: An intensive analysis of a single session. Canadian Journal of Counseling, 29, 160-175. Stiles, W.B., Glick, M.J., Osatuke, K., Hardy, G.E., Shapiro, D.A., Agnew-Davies, R., ... Barkham, M. (2004). Patterns of alliance development and the rupture-repair hypothesis: Are productive relationships U shaped or V shaped? Journal of Counseling Psychology, 51, 81-92. DYAD STUDY OF PSYCHOTHERAPY PROCESS 155 Teyber, E. and F. McClure (2011). Interpersonal Process in Therapy: An Integrative Model. Belmont, CA, Brooks/Cole. Tracey, T. J. (1987). Stage differences in the dependencies of topic initiation and topic following behavior. Journal of Counseling Psychology, 34, 123-131. Van der Veen, F. & Stoler, N., (1965). Therapists Judgment, Interview Behavior and Case Outcome. Psychotherapy, 2, 158. Vogt, W. Paul (2005). Dictionary of Statistics & Methodology: A Nontechnical Guide for the Social Sciences (Third ed.). SAGE. p. 40 Yalom, I (1995). The theory and practice of group psychotherapy (4th ed.). New York. Basic.