ABSTRACT Title of Thesis: DISENTANGLING THE THERAPIST?S CONTRIBUTION TO THE THERAPEUTIC RELATIONSHIP: ATTACHMENT STYLE, COUNTERTRANSFERENCE, AND THE REAL RELATIONSHIP. Beatriz Isabel Palma Orellana, Master of Science, 2012 Thesis directed by: Professor Charles J. Gelso Department of Psychology The current study investigated the relations among therapists? attachment, therapists? countertransference behaviors, and the real relationship therapists established with a client, as perceived by the therapists and their supervisors. Data were gathered from 32 therapist-supervisor dyads. Therapists completed measures of attachment and of the real relationship. Supervisors completed measures on therapist?s countertransference behaviors and the real relationship. Real relationship and countertransference measures were completed based on the work of the therapist with an identified client. Results showed that therapists? attachment security was positively and significantly related to therapists? ratings of the real relationship, but not to supervisors? ratings. Negative countertransference was related to supervisors? ratings of strength of the real relationship, but not to counselors? ratings. Contrary to expectations, positive countertransference was not related to supervisors? ratings of the real relationship. Finally, attachment security was not related to countertransference behaviors. DISENTANGLING THE THERAPIST?S CONTRIBUTION TO THE THERAPEUTIC RELATIONSHIP: ATTACHMENT STYLE, COUNTERTRANSFERENCE, AND THE REAL RELATIONSHIP. by Beatriz Isabel Palma Orellana Thesis 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 Master of Science 2012 Advisory Committee: Professor Charles J. Gelso, Ph.D, Chair Professor Mary Ann Hoffman, Ph.D Professor Matthew J. Miller, Ph.D ii Acknowledgements The present research project was possible thanks to the help of many people. First, I want to thank my advisor and mentor, Charlie Gelso, whose guidance and feedback through the process was invaluable. Charlie was both supportive and challenging, and he knew how to balance these two aspects with what I needed. I also want to express my gratitude to Mary Ann Hoffman and Matt Miller, members of my committee, for providing insight on how to improve this project. In the thesis defense, Charlie, Mary Ann and Matt created an environment that was intellectually stimulating and also infused with humor, and I really appreciated that. I also want to thank Kristen, my research assistant, for her diligent and persistent work in the data collection process. I am also appreciative to all the participants who gave some of their time to be part of this study. I am also truly grateful to Cristian, my husband, for always being there in spite of the physical distance; Thanks for your care and patience through the process. Finally, I want to thank my mom, dad and siblings, for believing in me and for their support as I conducted this study. iii Table of Content Chapter 1: Introduction?????????????????????............1 The Three Components of the Therapeutic Relationship 2 The Working Alliance 2 The Transference Configuration 3 The Real Relationship 5 Attachment 6 Chapter 2: Literature Review?????????????????????.10 Countertransference 10 Countertransference Over the Years 11 Important Distinctions in Relation to Countertransference: Quality and Amount 14 Challenges to the Definition and Operationalization of Countertransference 16 Research on Countertransference 18 Attachment 22 Historical Overview of Attachment 23 Adult Attachment 26 Therapist Attachment Organization and Psychotherapy Process and Outcome: Empirical Findings 29 Measurement of Attachment and its Classifications 31 The Real Relationship 34 Currents Perspectives on the Real Relationship 35 Is the Real Relationship Absence of Transference/ Countertransference? 36 Controversial Aspects of the Real Relationship 38 Research on the Real Relationship 38 Chapter 3: Statement of the Problem and Hypotheses???????????...44 Statement of the Problem 44 Hypotheses 45 Chapter 4: Method?????????????????????????. 51 Participants 51 Measures 55 The Experience in Close Relationships Scale (ECRS) 56 Inventory of Countertransference Behavior (ICB) 57 Countertransference Behavior Measure (CBM) 58 The Real Relationship Inventory-Therapist Form (RRI-T) 59 The Real Relationship Inventory-Supervisor Form (RRI-S) 60 Procedure 61 Recruitment of Therapists. 61 Recruitment of Supervisors. 65 iv Measure Application 66 Chapter 5: Results ?????????????????????????. 67 Descriptive Data 67 Test for Normality and Supervisor Effect 70 Analysis of Hypothesis 71 Chapter 6: Discussion ????????????????????????80 Attachment and Real Relationship 80 Countertransference and Real Relationship 82 Countertransference and Attachment 89 Limitations 91 Future Directions 96 Appendix B: Consent Form for Therapist-in-training???????????..103 Appendix C: Consent Form for Supervisors???????????????106 Appendix D: Demographic Questionnaire for Therapist Trainees??????...109 Appendix E: Demographic Questionnaire for Supervisors?????????..111 Appendix F: The Experience in Close Relationships Scale (ECR)??????. 113 Appendix G: Inventory of Countertransference Behavior (ICB) and Countertransference Behavior Measure (CBM)????????.115 Appendix H: The Real Relationship Inventory-Therapist Form (RRI-T) ???..116 Appendix I: The Real Relationship Inventory-Supervisor Form (RRI-S)???...117 Appendix J: First Email for Therapists?????????????.???...118 Appendix K: Second Email for Therapists?????????????.??.120 Appendix L: First Email for Supervisors?????????????.???123 References????????????????????????????125 v List of Tables Table 1: Ethnicity of therapists and supervisors (Appendix A)???...???....101 Table 2: Comparison Table on the ways supervisors knew about the trainees work, as reported by therapist and by supervisor (AppendixA)....???102 Table 3: Mean Item Scores and Standard Deviations for Real Relationship Inventory (RRI) for therapists (T) and supervisors (S), Experiences in Close Relationships Scale (ECR), Inventory of Countertransference Behavior (ICB), and Countertransference Behavior Measure (CBM)?...68 Table 4: Intercorrelations for Real Relationship Inventory (RRI) for therapists and supervisor, Experiences in Close Relationships Scale (ECR), Inventory of Countertransference Behavior (ICB), and Countertransference Behavior Measure (CBM)??????????..74 Table 5: Intercorrelations for Real Relationship Inventory (RRI) Subscales (Genuineness and Realism) for therapists and supervisor, Experiences in Close Relationships Scale (ECR), Inventory of Countertransference Behavior (ICB), and Countertransference Behavior Measure (CBM)?...76 1 Chapter 1: Introduction It has been estimated that the therapeutic relationship ?? accounts for as much of the [therapy] outcome variance as particular treatments? (Norcross, 2002, p. 5). Furthermore, recently it has been underscored that ??the value of a treatment method is inextricably bound to the relational context in which it is applied? (Norcross & Lambert, 2011, p. 5). In addition, it has been suggested that the relationship between clients and therapists can be curative in and of itself (Wampold, 2001). For years, authors have been discussing the centrality of the therapeutic relationship in psychotherapy (e.g., Gelso & Carter, 1985, 1994; Gelso & Hayes, 1998, Strupp & Binder, 1984). Furthermore, in 2001, the Steering Committee of the Division 29 (Psychotherapy) of APA task force recognized the pivotal role of the therapeutic relationship in psychotherapy by concluding that this relationship contributes to therapy outcome independently of the type of treatment. This group of leaders in the psychotherapy field also determined that knowledge is scant in relation to how the relationship is established, maintained, and why it works (Ackerman et al., 2001). Such findings and conclusions not only highlight the importance of the relationship between client and therapist in clinical work, but they bring to the table the need of discovering what variables might influence the therapeutic relationship. The present work seeks to be an addition to the empirical body of knowledge about the therapeutic relationship by illuminating the therapist?s contributions. Specifically, this study aims at discovering the relations between the therapist?s attachment style, the therapist?s countertransference behavior, and the therapist and supervisor?s ratings of the real relationship between the client and therapist. 2 Before delving into the specific concepts to be examined, it will be central to define what is meant by the therapeutic relationship. The therapeutic relationship is the relationship between therapist and client, which entails both the feelings and attitudes the client and therapist have toward each other and how these feelings and attitudes are expressed (Gelso & Carter, 1985, 1994; Gelso & Hayes, 1998). Furthermore, aligning with Greenson?s conceptions (1967; Greenson & Wexler, 1969), Gelso & Carter (1985, 1994) propose that all therapeutic relationships will have three components: a working alliance, a transference-countertransference configuration and a real relationship (Gelso & Carter, 1985, 1994; Gelso & Hayes, 1998; Gelso & Samstag, 2008). These three components can be considered separate, yet they influence each other. In order to better understand the variables considered for the current study, it is central to first define each of the components of the therapeutic relationship. The Three Components of the Therapeutic Relationship The Working Alliance Gelso and Carter (1994), define the working alliance as ?? the alignment or joining of the reasonable self or ego of the client and the therapist?s analyzing or ?therapizing? self or ego for the purpose of the work? (p.297). Bordin (1979, 1994) stated that the working alliance has three components: the bond between the participants, the extent to which they agree on the goals of therapy, and the extent to which they agree that the tasks the therapist uses will effectively attain those goals. Together, these elements determine not only the quality of the alliance but also how strong it will be (Ligi?ro & Gelso, 2002). Since its introduction, this construct has been widely studied, as it is considered to be the pivotal component of the therapeutic relationship (Castonguay , 3 Constantino & Holtforth, 2006; Gelso and Hayes, 1998). Furthermore, there is substantial empirical evidence that the working alliance is a demonstrably effective construct in therapy (Bordin, 1979; Ackerman et al., 2001), and that it is related to therapy outcome (e.g., Crits-Christoph, Connolly Gibbons, & Hearon, 2006; Horvath & Symonds, 1991; Horvath & Bedi, 2002; Martin, Garske, & Davis, 2000; Safran & Muran, 2006). The Transference Configuration A second component of the therapeutic relationship is the transference configuration, which includes both, the client?s transference and the therapist?s countertransference. Transference refers to ??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 & Hayes, 1998, p. 51). According to Gelso and Hayes (2007), countertransference refers to the reactions that the therapist has, both, at an internal level and at an observable level, which are related to the therapist?s own difficulties and vulnerabilities. Recently, there have been significant efforts in addressing both transference and countertransference in research (Gelso & Hayes, 1998). In relation to countertransference, efforts toward testing the relation between countertransference and outcome have been infrequent (Gelso and Hayes, 2007), yet a recent meta-analysis showed a modest but significant inverse relationship between countertransference and outcome (Hayes, Gelso & Hummel, 2011). It is relevant to mention that for years countertransference had been mainly considered as therapist?s underinvolvement, avoidance, withdrawal or misperceptions from the part of the therapist (e.g., Cutler, 1958; Hayes & Gelso, 1993; Latts & Gelso, 4 1995: Peabody & Gelso, 1982; Yulis & Kiesler, 1968). Yet, overinvolvement (e.g., befriending the client, talking too much in session) can also be indicative of countertransference, and can be as detrimental for the therapeutic work as underinvolvement. Friedman and Gelso (2000) developed the Inventory of Countertransference Behavior (ICB), to assess countertransference behaviors. In their study, these authors found that the theoretical perspective of overinvolvement or underinvolvement of the therapist did not accurately explain the factors that they found in the development of their measure. Friedman and Gelso (2000) presented an alternative way to conceptualize countertransference, as negative (i.e., aggressive, avoidant or punitive behaviors from the therapist) and positive (i.e., approaching the client in inappropriate ways), which allows capturing different aspects of countertransference in a more nuanced way. For example, Ligi?ro and Gelso (2002) found that ??negative countertransference was associated with poorer working alliances, and positive countertransference was related to a weak bond within the working alliance? (p. 3). Therefore, empirical work directed at discovering the relations between countertransference, both positively and negatively valenced, to other constructs that are relevant to psychotherapy process and outcome can be highly valuable. Considering the relationship constituents, there has been research on the relationship between countertransference and working alliance (e.g., Ligi?ro & Gelso, 2002), yet, to this author?s knowledge, no study up to date has seen the associations between countertransference and real relationship. 5 The Real Relationship A third constituent of the therapeutic relationship has been termed the real relationship. This component is theorized to exist from the first moment of contact between client and therapist, and it refers to the personal relationship between the two (Greenson & Wexler, 1969; Gelso & Carter, 1985, 1994; Gelso & Hayes, 1998; Gelso, Kelley, Fuertes, Marmarosh, Holmes, Costa, & Hancock, 2005; Gelso, 2011). The real relationship is comprised of two defining aspects: Genuineness and realism. The first feature, genuineness, corresponds to ?? the ability to be what one truly is in the relationship- to be authentic, open and honest? (Gelso and Carter, 1994, p. 297). The second characteristic, realistic perception and reactions, reflects the idea of perceiving the other, and therefore reacting towards the other, in an accurate and realistic way (Gelso & Carter, 1994). Gelso and Hayes (1998) state that ?within the real relationship, perceptions of and experiencing with the other are largely realistic or nontransferential? (p. 109). The real relationship also has two subcomponents: Magnitude (how much) and valence (positive or negative) (Gelso, 2011). According to Gelso and Carter (1994), the strength of the real relationship (determined by the magnitude and valence of the genuineness and realism of the relationship) will influence the effectiveness of therapy. Of all the components of the therapeutic relationship, the real relationship has been the least studied. Since the introduction of a reliable measure of the real relationship from the therapist?s view (Gelso et al., 2005) and from the client?s view (Kelley, Gelso, Fuertes, Marmarosh, & Lanier, 2010), there have been important efforts in empirically studying the real relationship (e.g., Fuertes, Mislowack, Brown, Gur-Arie, Wilkinson, & Gelso, 2007; Gelso et al., 2005; Marmarosh, Gelso, Markin, Majors, Mallery & Choi, 6 2009; Lo Coco, Gullo, Prestano & Gelso, 2011). Findings include positive associations of real relationship with working alliance and client insight (Gelso et al., 2005), among others. In addition, there is a positive relation between strength of the real relationship and therapy outcomes (Ain, 2008; Marmarosh et al., 2009; Lo Coco et al., 2011). Also, it has been found that the real relationship is negatively associated with measures of client negative transference (Gelso et al., 2005; Marmarosh et al., 2009). Therefore, although the association between real relationship and countertransference has never been tested, studies considering client?s transference could lead to expect that therapists? negative countertransference would be negatively associated with real relationship. It would be relevant to determine if such association holds also for the relations of countertransference with real relationship. As expected, several personal characteristics of the therapist and client might affect the different components of the therapeutic relationship. One of the variables that has been theorized to influence the therapeutic relationship is attachment. Recently there has been an increased focus on the empirical study of psychotherapy and attachment (Slade, 2008). For example, Moore and Gelso, (2011) found that the strength of the real relationship as perceived by the client was positively related to client?s secure attachment and security of client?s attachment to therapist. In addition, there have been empirical efforts to relate attachment to countertransference and/or transference, to the working alliance and to the real relationship. Attachment Bowlby (1969/1982) was the first to introduce the idea of attachment. He maintained that when a child is born, he/she comes equipped with several in-built 7 behavioral systems (e.g., exploratory, fear, attachment, care giving). From birth, the child presents attachment behaviors (e.g., crying, smiling), which ??promote proximity to the attachment figure? (Cassidy, 2008, p. 12). Such behaviors will be organized into an attachment behavioral system, which will be a unique product of the child?s response to internal processes and external stimuli (Cassidy, 2008). When the infant perceives need for care, the attachment system is activated to keep proximity to the caregivers (Bowlby, 1969/1982; Mohr, Gelso, & Hill, 2005). The attachment behavior stops when the stimulus (e.g., perception of danger) ends (Bowlby 1969/1982; Cassidy, 2008). Bowlby also distinguishes attachment bonds, which refers to ??an affectional tie? (Cassidy, 2008, p.12) that a person has with other person who is perceived as wiser and stronger, such as the mother (Cassidy, 2008). It is relevant to mention that all attachment bonds are affectional bonds, but not all affectional bonds are attachment; what distinguishes them is that only in the attachment bond does one seek a wiser and stronger person as a secure base in times of trouble (Cassidy, 2008; Cassidy, 2010). An important distinction between attachment behavior and attachment bond is that although a child might not be exhibiting attachment behaviors to a parent (e.g., crying), he/she is still attached to such caregiver (Cassidy, 2008), and that is the attachment bond. Such bonds are therefore established at an early age, and ??exist consistently over time, whether or not attachment behavior is present? (Cassidy, 2008, p. 13). In addition, such experiences with attachment figures are internalized (i.e., working models) and will guide the way we relate to others during our life course (Bartholomew & Horowitz, 1991), and what we expect from them. 8 As adults, these ways of relating will be manifested in specific patterns or attachment styles. Bartholomew and Horowitz (1991) developed an adult attachment classification, based on whether the person is low or high in both anxiety and avoidance. Such classification has been widely utilized in research, and it has proven to be easily operationalized and adequately assessed by attachment measures (e.g., Experiences in Close Relationships Scale - ECR). Psychotherapy studies considering attachment have shown that client attachment, counselor attachment and/or the interaction of both can be related to other variables relevant to client?s treatment and affect therapy process and outcome. For example, dismissing attachment style of the therapist was positively related to hostile countertransference, rated by supervisors (Mohr et al., 2005). Nevertheless, although there is important evidence about the relevance of these constructs (i.e., countertransference behavior, real relationship, therapist attachment style) for psychotherapy research, there are many questions not yet answered. Specifically, two studies that assessed countertransference behavior and therapist attachment did not show any significant relation between the two variables (Ligi?ro & Gelso, 2002; Martin, Buchheim, Berger & Strauss, 2007). In line with the previous findings from the literature, the present study seeks to increase the body of literature related to the psychotherapy relationship, focusing on the relations between the two least studied constituents of the therapeutic relationship in relation to the therapist (countertransference and real relationship) and their relation to therapists? attachment patterns. As it was previously presented, no study to date has focused on all these relations; however, Ligi?ro & Gelso?s (2002) work on therapist attachment style, working alliance, and countertransference behavior is a pivotal study to 9 consider. Theoretically, it has been stated that the working alliance emerges from the real relationship (Greenson, 1965, 1967); therefore, one could expect similar relations between the variables used in the present study and the ones that these authors found. Ligi?ro & Gelso (2002) did not find a relationship between attachment style of the therapist and countertransference behaviors. One possible explanation to such results can be that the measure that these authors used to assess attachment (i.e. the Relationship Questionnaire) did not measure the construct adequately. The current study seeks to expand Ligi?ro and Gelso?s results by studying a new construct (i.e., real relationship), and using a measure of attachment that has proven to have higher internal consistency, greater validity, and has been used more extensively in adult attachment research. In sum, the present study aims at clarifying the relations between a) therapists? attachment and therapists? perspective of the real relationship, b) therapists? attachment pattern and countertransference behaviors, as rated by the supervisors, and c) the therapist?s rating of the real relationship and the supervisor?s rating of the therapist?s countertransference behavior. In addition, supervisors? rating of the real relationship will be correlated with countertransference behavior, therapists? attachment style and therapist?s rating of the real relationship. 10 Chapter 2: Literature Review The present study seeks to contribute to the body of research that focuses on the therapist?s influence on the therapeutic relationship, by illuminating the relationship between countertransference behavior, therapist attachment patterns, and real relationship. Theoretically, it has been suggested that countertransference is negatively related to the real relationship (Gelso, 2011), yet no study has yet tested such relation. Some authors have done empirical work on the relation between real relationship and the counselor?s attachment (e.g., Mohr et al, 2005) and countertransference and therapist attachment (e.g., Ligi?ro and Gelso, 2002), yet none have considered real relationship, therapist attachment and countertransference behavior. The review in this chapter will focus on the bodies of literature related to the three main variables of the present study: Countertransference, attachment and real relationship. The first construct reviewed in this chapter is countertransference, the second section will focus on attachment, and the third section will explore real relationship. The literature considered includes both, theoretical perspectives and empirical findings, including a brief historic review of the constructs. Countertransference For years it has been theorized that countertransference manifestations could affect the therapeutic relationship and therefore, therapeutic outcomes, in negative ways (Freud 1910/1959; Gelso and Carter, 1994: Gelso and Hayes, 1998, 2007). Yet, few empirical efforts have been directed at testing the relation between countertransference and outcome (Gelso and Hayes, 2007). The current subsection reviews 11 countertransference definition over time, some central distinctions of the concept and challenges for measurement and operationalization, and a review of some key studies. Countertransference Over the Years Since its first introduction by Freud (1910/1959), the concept of countertransference has been defined in different ways, which also reflect dissimilar views regarding its utility in psychotherapy. Based partly on Epstein and Feiner?s (1988) work, Gelso and Hayes (2007) present four conceptions of countertransference, which shed light on how the concept has evolved over time, even though all these views are still present today. The Classical View. The first perspective on countertransference is what is known as the classical view, and it has its origins in Freud?s introduction of the concept of countertransference in the early 1900s (Gelso & Hayes, 2007). From this perspective, countertransference refers to the therapist?s reactions to the patient?s transference, which stem from the therapist?s unresolved conflicts from early childhood (Gelso & Hayes, 2002; Gelso & Hayes, 2007). From this viewpoint, countertransference is undesirable, and something that the therapist needs to eliminate by overcoming his/her own difficulties. The classical conception of countertransference highlights the need for therapist?s to solve his/her own problems as they might influence their work (Gelso & Hayes, 2007); however, its focus on reactions only to transference and its negative valence on countertransference makes this position a restrictive perspective (Gelso & Hayes 1998, 2002, 2007; Epstein and Feiner, 1979). A new perspective on countertransference, the totalistic view, developed, which was a response in part to the 12 narrowness of the classical view, and also grew as a result of the transformation that psychoanalysis was going through (Gelso & Hayes, 2007). The Totalistic View. This conception of countertransference originated in the 1950s (Gelso & Hayes, 2007). According to this perspective, countertransference encompasses all the feelings and attitudes a therapist has toward a patient, and therefore, any reaction is worthy of being studied (Epstein & Feiner, 1988; Gelso & Hayes, 2007). The totalistic view brought to the table the utility of using countertransference as a tool in the treatment of a client, as the therapist?s internal reactions can be similar to the ones other people have towards the client, can help the therapist understand the client?s transference, and can illuminate the client?s internal world (Gelso & Hayes, 2007). Therefore, from this perspective, countertransference is not only about the therapist vulnerabilities, but it can also illuminate the client dynamics and the experience others have to the client?s way of being-in-the-world. Several authors have considered this double aspect of countertransference (e.g., Kiesler, 2001; Winnicott, 1949; Heimann, 1950). An important distinction in this perspective corresponds to objective (reactions generated by the client in most people) and subjective (related to the therapist?s own issues) countertransference. According to Gelso and Hayes (2007), the totalistic view became more prevalent as therapists (such as Kernberg) worked with borderline and severely regressed patients, as such populations might trigger strong reactions in the therapist. The therapist?s reactions could give valuable information about the client; however, the extensive scope of such conception makes it seem that everything is countertransference (Gelso and Hayes, 2007). 13 The Complementary View. In the third perspective of countertransference, the complementary view, countertransference is considered to be the complement to the way the patient tends to relate to others, or the counterpart to his/her transference (Gelso & Hayes, 2007). From this conception, in their interaction, patient and therapist affect each other continuously. Specifically, the client will get the therapist to interact with him/her as others normally do (commonly know as ?pull?), to which the client will then respond. A central defense mechanism within this view is projective identification (Ogden, 1986/1990), and exponents of this view are Racker (1957) and Kiesler (1996, 2001), among others. The shortcoming of such conception is that it does not focus on what the therapist contributes to the relationship (Gelso, 2004). The Relational View. This perspective of countertransference emanates from a two-person psychology within psychoanalysis, emphasizing the co-construction between therapist and patient (Gelso and Hayes, 2007). From this viewpoint, the therapist?s and the client?s ?? needs, unresolved conflicts, and behaviors? add to the countertransference manifestations (Hayes, Gelso & Hummel, 2011, p. 89). Gelso (2004) brings to the table the fact that such emphasis on co-construction might miss the point of the reality that each member of the therapeutic dyad brings to the encounter with the other. The Integrative View. Besides the previous four views on countertransference, Gelso and Hayes (2007) present their own perspective on countertransference. From such perspective, countertransference can be considered to be the therapist reactions, both at an internal (e.g., feelings, thoughts, bodily sensations and emotions) and external (verbal and non-verbal behavior) level, which stem from the therapist?s own vulnerabilities 14 (Gelso and Hayes, 2007). From this perspective, countertransference ??is seen as a potentially useful phenomenon if therapists successfully understand their reactions and use them to help understand the patient? (Hayes, et. al., 2011, p. 89). Other perspectives on countertransference. Although the main theoretical perspective that has addressed the concept of countertransference has been psychoanalysis, there have been some efforts to address it from other points of view. In 2001, the Journal of Clinical Psychology prepared a special issue on countertransference, inviting professionals from different theoretical perspectives to address the construct. The perspectives represented were: rational-emotive therapy (Ellis, 2001), feminist social constructionism (Brown, 2001), constructive brief therapy (Hoyt, 2001), interpersonal therapy (Kiesler, 2001), couples and family therapy (Kaslow, 2001), experiential perspective-calling it ?the therapist?s personal reaction? (Mahrer, 2001), Kohut?s self- psychology (Guy & Brady, 2001), and contemporary psychoanalysis (Gabbard, 2001). It is relevant to mention that Hayes and Gelso (2001) contributed with empirical perspectives. Important Distinctions in Relation to Countertransference: Quality and Amount Besides the specific theoretical distinctions of countertransference, there are other important aspects to consider when thinking about countertransference, such as quality of countertransference and amount of countertransference (Gelso & Hayes, 2007). In relation to quality, countertransference can be either positive or negative. When displaying positive countertransference, a therapist experiences positive feelings toward the client, which can thwart the therapeutic work, for example, by caring too much for the client, or befriending the client. When negative countertransference is present, the 15 therapist feelings towards the client are negative, and therefore he/she presents negative reactions towards the client, such as hostility. As it was previously presented, the ICB is a measure of countertransference behavior that has two subscales that differentiate between the two qualities of the construct, positive or negative, factors which seem to represent the theory better than the previous overinvolvement and underinvolvement constructs used in measurement. Since Friedman and Gelso developed this measure in the year 2000, the ICB has been used in different studies of countertransference. When either positive or negative countertransference is present, the feelings and reactions stem from the therapist?s personal issues, and interfere in the therapeutic work. This idea has been empirically supported by some research findings. As it was previously presented, Ligi?ro and Gelso (2002) found that there was a negative association between positive countertransference and the bond component of the working alliance as rated by supervisors. In addition, negative countertransference was related to weaker working alliances (overall and for each of its components: bond, tasks and goals). These authors used the ICB to measure countertransference behaviors. Countertransference also varies in relation to amount. In a sense, countertransference refers to excess or shortage of reactions. A therapist can become overinvolved, for example, by overpraising the client (Gelso, Hill, Mohr, Rochlen, & Zack, 1999) and therefore, overgratifying him/her. On the contrary, a therapist might withdraw, or become bored in session, which might reflect a therapist?s underinvolvement (Gelso & Hayes, 2007). 16 Challenges to the Definition and Operationalization of Countertransference From the previous sections, it can be inferred that due to such different definitions of countertransference, operationalizing this construct can be very challenging, which presents a pivotal dilemma for research (Gelso, Fassinger, Gomez & Latts, 1995; Fauth, 2006; Najavits, 2000). In the early 50?s, Cutler (1958) highlighted the difficulties on countertransference research, mainly due to the operationalization of psychoanalytic concepts, and the fact that usually therapists do not want to allow researchers to ??. examine their conflicts and inner feelings? (p. 349). In addition, Najavits (2000) discussed the complexities of researching therapists? emotions and countertransference, including some suggestions to consider in order to improve future research. Recently, Fauth (2006) brought to the table the fact that currently there is no conceptual clarity in relation to the term of countertransference, and that there are central measurement issues in relation to the construct. In order to start addressing countertransference in a clearer way, Fauth (2006) recommends to use a model on countertransference originally presented by Hayes (1995), which categorizes countertransference into 5 components: origins (i.e., areas of unresolved issues in the therapist), triggers (i.e., the events in therapy that trigger the therapist unresolved difficulties), manifestations (i.e., affective, behavioral and/or cognitive responses in the therapist, such as feelings of anger, or avoidant behaviors), management (i.e., strategies that the therapist develop to cope with the countertransference), and effects (i.e., the influence that countertransference has on the psychotherapy and outcome). The different instruments one uses will assess different components, therefore, it would be central to use measures that are directed at the component we aim to assess. 17 As countertransference research presents so many challenges, and considering Najavits (2000) and Fauth?s (2006) analysis, it will be important to clearly define the construct and operationalize the concept of countertransference that will be considered in the present study. The countertransference definition used in the current empirical work is the integrative view presented by Gelso and Hayes (2007). As these authors mention, all the different definitions add unique views yet have limitations (Hayes et al., 2011), and the integrative perspective incorporates ??lessons from each? (Hayes et al., 2011, p. 89; Gelso & Hayes, 2007). Furthermore, Hayes et al. (2011) mention that just like the classical view, the integrative definition focuses on vulnerabilities, yet allows considering countertransference as potentially useful. In addition, the integrative perspective isn?t over encompassing as the totalistic view (as it would bring empirical and clinical challenges), but shares with it the idea of countertransference as inevitable (Hayes et al., 2011). Lastly, Hayes et al. (2011) mention that the integrative view goes beyond the therapist?s response to the client?s transference by considering the therapist?s reaction to ?all clinically relevant material? (p. 89), which is in line with the relational and complementary views. The chosen definition was operationalized as the behavioral manifestations of countertransference (including both, positive and negative manifestations). In line with Gelso and Hayes (2007), it is considered that if internal countertransference is not acted out, it can be valuable information to understand the client?s dynamics, and therefore, is not necessarily an unwanted reaction. Countertransference becomes unwanted when such reactions are acted out, that is, manifested in behaviors. In addition, focusing on the behavioral aspects of countertransference allows the possibility of having external raters 18 assessing countertransference (e.g., supervisors), which is central when one is trying to assess therapist?s reactions that stem from his/her own vulnerabilities. Research on Countertransference A final yet central aspect to address in this section on countertransference is research, which has illuminated how this construct is related to several other variables that have proven to affect therapy effectiveness (e.g., therapeutic alliance, empathy, Ackerman et al., 2001). Cutler (1958) examined the transference reactions of therapists to their clients, finding that therapists will have internal and external manifestations of countertransference when listening to material that client presents that is conflict relevant for the therapist. Peabody and Gelso (1982) conducted an analogue study, to discover the relationship between counselor trainees? empathic ability and countertransference feelings and behaviors. These authors used as clients a seductive, a hostile, and a neutral female client, and found that the trainees? empathy was negatively related to countertransference behavior with the seductive client, and that empathic ability was related in a positive way to openness to feelings of countertransference. Hayes and Gelso (1991) studied the relation of state anxiety in therapists? in training and their countertransference behavior, considering the trainees? empathy as a possible moderator of this relationship (i.e., ?the adverse effects of anxiety would influence only the less empathic trainees?, p. 284). Results showed that, as expected, there was a positive relation between state anxiety and countertransference behavior; however, this result just holds for male trainees. In addition, the hypothesis of empathy as a moderator was not supported. 19 Another investigation that focused on counselors? countertransference was the one conducted by Gelso et al. (1995), who used an analogue study (which included two conditions, lesbian and heterosexual) to examine trainee?s countertransference reactions to lesbian clients. In this study, counselor?s level of homophobia was considered a possible moderator, and the researchers also observed gender differences. Countertransference was considered as having behavioral, affective and cognitive components, which were operationalized as avoidance, state anxiety, and cognitive recall (i.e., greater recall problems are due to countertransference), respectively. Results showed that trainees did not exhibit more countertransference towards lesbian than heterosexual clients. Also, in the lesbian condition there was a negative and significant relation between state anxiety and two of the countertransference management subscales: anxiety management and self-integration. In the lesbian condition, there was a positive correlation between counselor?s homophobia and counselor?s avoidance behavior, and women therapist-trainees presented more recall problems with the lesbian client than did men. Latts and Gelso (1995) also conducted an analogue study to assess countertransference behavior with rape survivors. Trainees? responses to a videotape were considered as either ?approach? or ?avoidance?, which accounted for countertransference behaviors. Other variables of interest were gender and countertransference management, given by a two-step model (i.e., awareness of the feelings of countertransference and using a theoretical perspective to understand them). Results showed that the male therapists gave more avoidant responses than the females, and there was an interaction effect between awareness of feelings and theoretical 20 perspective. Specifically, low awareness and high theoretical framework generated the most avoidance, whereas high awareness and high use of theory resulted in least avoidance (Latts and Gelso, 1995). As it can be seen from the previous studies, countertransference used to be operationalized as underinvolvement or avoidance (e.g., Cutler, 1958; Hayes & Gelso, 1993; Latts and Gelso, 1995: Peabody and Gelso, 1982; Yulis & Kiesler, 1968). Yet, overinvolvement (e.g., befriending the client, talking too much in session) can also be indicative of countertransference, and can be as detrimental for the therapeutic work as underinvolvement. Considering that countertransference can take many forms, Friedman and Gelso (2000) developed the Inventory of Countertransference Behavior (ICB), to assess countertransference behaviors. The ICB measures positive and negative countertransference. Ligi?ro and Gelso (2002), in a study upon which the present investigation was based, provided a step forward in the empirical work on countertransference by examining the relations between attachment styles of the therapist, countertransference behavior and working alliance. Therapists in training completed self-report measures on attachment style and working alliance with a particular client. The trainee?s supervisor completed measures of trainee?s countertransference behavior and working alliance, taking into account the same client for which the trainee assessed his/her work. As it was previously presented, results revealed that positive countertransference was negatively related to the bond component of the working alliance, as rated by supervisors. Negative countertransference was negatively related to both therapists? and supervisors? ratings of the working alliance. Finally, these authors found that differences of therapists and 21 supervisors? ratings on the bond component of the working alliance predicted countertransference behaviors. Such findings are very relevant for psychotherapy research, as the working alliance has shown to be a central variable influencing therapy effectiveness (Horvath & Greenberg, 1994; Horvath & Luborsky, 1993). The authors used the ICB to measure countertransference, which proved to effectively assess countertransference behaviors, both in their positive and in their negative form. An important contribution to psychotherapy research was the work of Mohr et al. (2005), who studied clients? and therapist trainees? attachment style (variable that will be discussed later on in greater detail) and their relation to countertransference behavior and session evaluation. These researchers found that counselor attachment predicted countertransference, and that interactions between counselors? and clients? attachment predicted distancing and hostile countertransference; the highest levels of countertransference occurred when the client had a preoccupied attachment pattern, and the counselor?s attachment pattern was either fearful or dismissing. In addition, they found a positive association between counselor?s dismissing attachment and supervisor?s rating of hostile countertransference. In a different study, Fauth and Hayes (2006) investigated the utility of Lazarus?s transactional model of stress for countertransference, finding that stress appraisal predicts countertransference behavior. Moreover, there was a link between negative appraisals and increased hesitance with and distance from the client, whereas there was a positive relation between positive appraisals and positive client?s diagnostic evaluations. Other studies related to countertransference have found that therapist who are low on anxiety presented less countertransference ? based on their being more personally 22 involved with their clients- than the highly anxious therapists (Yulis & Kiesler, 1968); counselor?s homophobia predicted counselor?s discomfort with gay male clients (Hayes & Gelso, 1993); a relation between patient?s reported interpersonal problems and specific countertransference behaviors in the therapists (Rossberg, Karterud, Pedersen, & Friis, 2008); client?s assessment of deeper session and more expert therapists when the therapist made general self-disclosure if the alliance was positive ?there was an interaction effects, so the results were different when alliance was negative (Myers & Hayes, 2006); therapist?s attitude of liking/disliking clients and their countertransference (McClure & Hodge, 1987), among others. Recently, Hayes et al. (2011) conducted a meta-analysis and found an inverse and modest relation between countertransference reactions and psychotherapy outcome. Therefore, it can be seen that countertransference has shown relevant influences in the therapeutic relationship. Considering the studies that relate countertransference to attachment, it would be central to determine whether the lack of relation found by Ligi?ro and Gelso holds in other samples, and if countertransference findings are repeated when considering real relationship instead of the working alliance. In addition, such results could provide empirical support for the direct relationship between these two distinct concepts (real relationship and countertransference). Attachment As it has been previously stated, attachment is a variable that has been related to psychotherapy research. Bowlby himself brought to the forefront the idea of therapists assuming the role of attachment figures with their clients, in order to be a secure base for them, which in turn allows the clients to further explore and change their working models 23 (Bowlby, 1988; Ainsworth, 1989). However, just like mothers respond differently to their child due to their own attachment patterns, we could expect that different therapists attachment patterns will influence the work of therapy and relate to the clients in different ways. In this sub-section, a general overview of attachment theory is presented, followed by more specific information on adult attachment. In addition, theoretical and empirical information on therapist attachment styles is offered. Finally, the challenges of measurement in adult attachment are addressed. Historical Overview of Attachment Theory John Bowlby (1969/1982) developed a new perspective from which to understand personality development, based on the attachment bond that an infant establishes with his/her caregiver. Drawing heavily on findings from the field of ethology (Bowlby, 1969/1982), and with a strong evolutionary underpinning (Cassidy, 2010), Bowlby?s framework is built upon the idea that observing the infant?s behavior towards the mother, while she is present or absent, will illuminate our knowledge of human socio-emotional development (Bowlby, 1969/1982). According to Bowlby, babies are born with a repertoire of attachment behaviors (e.g., crying, babbling), which are organized in ?attachment behavioral systems? (Cassidy, 2008, p. 5), and will be activated when the infant experiences vulnerability and distress. This will activate caregiving behaviors from the mother (or caregiver) (Ainsworth, 1989), who will provide safety for the infant. Thus, attachment behaviors have the biological function of protection, as they are displayed to seek or maintain proximity with another person who is seen as stronger, wiser, and thus more able to cope with the world (Bowlby, 1969/1982, 1988). These attachment behaviors also serve to increase an individual?s survival and reproduction possibilities 24 (Mikulincer & Shaver, 2007). Once proximity is reestablished for the infant, the attachment system is terminated (Cassidy, 2008) and he/she is able to continue exploring the world. This attachment behavioral system involves both external manifestations and internal organizations (Ainsworth, 1989). During the first year, the baby experiences certain regularities in his/her relation with the world, especially in the interaction with the mother (i.e., main caregiver). The infant starts organizing such experiences into an ?internal working model? (or representations) of the self, the mother and the relationship (Ainsworth, 1989; Thompson, 2008). Such mental representations or internal working models of him/herself, of his/her mother, and of the infant-parent relationship are related to what the infant can expect from the interactions, and therefore, will guide the child?s assessment of the situations he/she experiences, will determine his/hers attachment plans and will guide future interactions with the world (Bowlby, 1969/1982). Such mother- child ways of interacting are internalized by the infant and will continue to guide the relationships one has beyond infancy (Bowlby, 1988; Sable, 2007). Not all the times the caregiver?s responses will generate security in the child, making it necessary to understand the individual differences in attachment quality, and how these differences relate to one?s survival. Such understanding was possible due to Mary Ainsworth?s efforts to advance attachment theory. Ainsworth worked with Bowlby, and she is credited with generating methodologies that allowed the empirical testing of attachment theory. Ainsworth (1989) mentions that her contributions to the theory were based mainly in two aspects: normal development of attachment in infants and the qualitative differences in their attachment. Ainsworth described the normative 25 development of attachment during the infant?s first twelve months, based on her observations of infant-mother interactions in their natural environment (Ainsworth, 1989): for example, observation of mother and child interactions in Uganda (Mikulincer & Shaver, 2007; Bowlby, 1969/1982). In addition, her contribution to the empirical study of individual differences in the quality of attachment took an unprecedented step when she developed the Strange Situation, a laboratory procedure that allowed assessment in a controlled environment of the different organizations of attachment behaviors that 12- month-old infants have towards their mothers (Bowlby, 1969/1982; Mikulincer & Shaver, 2007). Such procedure involves a couple of separations and reunions between mother and child, among others, and relevant information on the child?s attachment is attained from every aspect of the procedure (e.g., how does he/she behave upon reunion with the mother? Does the child play with the toys? etc..). Based on her identification of three distinct groups of infants due to their behaviors in the Strange Situation, Ainsworth postulated the existence of three types of attachment patterns in infants: a group that is securely attached (Group B; children who are active in play, seek contact after separation of mother, and are readily comforted), and two that are considered insecurely attached (Bowlby, 1969/1982). The first group of insecurely attached infants was identified as anxious and avoidant (Group A; avoid mother upon reunion, at times treats stranger in a friendlier manner than mother), and the second insecure group was the anxious and resistant (Group C; oscillate between seeking for contact and proximity and resisting interaction with mother. Some may be angry) (Bowlby, 1969/1982). 26 The examination of infant?s attachment differences was furthered advanced by Mary Main, who identified a fourth attachment category: disorganized. Such pattern of attachment corresponds to children who present ?dazed behavior on reunion with parent? (Main, Kaplan and Cassidy, 1985, p. 79), confusion, contradictory behavior patterns, among others, and has been associated to trauma in their own attachment stories (Main, Kaplan & Cassidy, 1985). Further studies have shown that the mothers of these infants may have a history of trauma and/or unresolved loss. In addition, Main and her collaborators further developed attachment theory by assessing attachment in 6-year-old children, and developed an attachment interview (AAI) to classify adults in relation to attachment (Ainsworth, 1989). Finally, Main and her collaborators have done a remarkable job in advancing Bowlby?s ideas on internal working models of attachment by better defining and clarifying these internal representations (Main, et al., 1985). Adult Attachment As it can be seen in the historical overview, the precursors of attachment theory based their efforts mainly at describing the first years in the life span. However, according to Bowlby (1979) attachment behavior is ??from the cradle to the grave? (p. 129) and attachment relationships continue to be central all through the life cycle (Ainsworth, 1989; Bowlby, 1969/1982, 1988). Although Bowlby did not devote much attention to attachment beyond infancy, he cemented the ground for others to further develop attachment theory across the life span. Bowlby (1969/1982) mentioned that with age, the intensity and frequency of attachment behaviors would diminish. In her work ?Attachment beyond Infancy?, Mary Ainsworth (1989) discusses some aspects of the attachment bond through the life cycle. A central aspect that is relevant to note is that 27 attachments are a type of affectional bond (i.e. a tie with a partner that is long and enduring, and in which the partner is non-interchangeable and is valued as unique), in which one feels comfort and security in the relationship and is able to use the partner as a secure base from which one can explore the world with confidence (Ainsworth, 1989). This highlights the centrality and impact that the quality of our attachments will have in the relationships we establish through our life span. Ainsworth (1989) directly addressed the child-parent attachment bond during adolescence, and other affectional bonds: bond of father to child, sexual pair bonds, friends, companions and intimates, and bonds with siblings and other kins. The study of adult attachment further developed with the work of Hazan and Shaver (1987), who introduced the idea of romantic love as an attachment process, and started to focus on the ties between infant attachment and adult attachment. To assess adult attachment, these authors developed a single-item measure of three adult attachment styles, which corresponded with Ainsworth?s description of infant?s attachment (Hazan & Shaver, 1987). The three styles that these authors presented were secure, avoidant and anxious/ambivalent, and results showed that the distribution of their sample (56% secure, 25% avoidant and 19% anxious/ambivalent) was similar to the one found in research on infant attachment at the time. It is relevant to mention that the measure was developed to address working models, and it showed that secure lovers, avoidant lovers, and anxious/ambivalent lovers had different love experiences. For example, secure lovers experienced happiness and trust in their relation, avoidant lovers feared intimacy and had emotional highs and low, and anxious /ambivalent lovers experienced obsession and extreme sexual attraction, among others. 28 Such empirical findings about adult romantic love as attachment opened the door to an important line of studies in adult attachment. Bartholomew and Horowitz (1991) introduced a new model of adult attachment styles, based on their empirical findings. Specifically, these authors postulate four prototypical attachment styles (secure, preoccupied, dismissing and fearful), which are given by the combination of the individual?s self image (positive or negative) and the image of others (positive and negative). Bartolomew and Horowitz (1991) found that each style was associated with its own interpersonal problems. Later on, adult attachment theory was advanced by Fraley & Shaver (2000), who re-revised Hazan and Shaver?s perspective by furthering the discussion on adult attachment, based on the accumulated research at the time. A relevant aspect that they present is that ?the attachment system, a system originally adapted for the ecology of infancy, continues to influence behavior, thought, and feeling in adulthood? (Fraley and Shaver, 2000, p. 147). Considering that the different attachment styles or patterns have been associated to specific interpersonal problems, and that each style represents specific working models, it is central to focus on therapist attachment patterns, as they can directly influence the client and the therapeutic relationship. Although one wouldn?t expect that the therapist activates his/her attachment behaviors with the clients (the client is not seen as a secure base), the internal working models will determine several aspects of how we relate to others. In addition, some exchanges, content or experiences in session might activate the therapist?s attachment system. The following sub-section will address the literature related to therapist attachment patterns. 29 Therapist Attachment Organization and Psychotherapy Process and Outcome: Empirical Findings During the past years, there has been a rise of research on attachment and psychotherapy (Slade, 2008). Such studies have focused on both, therapist and client?s attachment organization and their influence in psychotherapy process and outcome (e.g., Woodhouse, Schlosser, Crook, Ligi?ro, & Gelso, 2003; Kivlighan, Patton, & Foote, 1998). For example, results of a recent meta-analysis showed that ??individuals with more secure attachment styles demonstrated stronger alliances, whereas individuals with more insecure attachment styles demonstrated weaker alliances? (Diener & Monroe, 2011). In spite of the increased research on attachment, fewer studies have focused on therapists? attachment patterns? contributions in therapy than in the clients? effects (Daniel, 2006). In addition, there have been efforts in the theoretical advancement of attachment theory and psychotherapy (e.g., Mallinckrodt, 2010, who conceptualizes the psychotherapy relationship as an attachment bond). In relation to therapists? attachment patterns, Black, Hardy, Turpin & Parry (2005) found that therapists? self-reported secure attachment style was positively and significantly related to the therapist report of a good alliance. On the contrary, therapists? anxious attachment style was negatively and significantly related to good alliance and positively and significantly associated to number of problems reported in therapy by the clinician. Also, self-reported attachment style explained a significant portion of the variance beyond general personality variables. In addition, several studies have found an interaction between therapists? attachment pattern and clients? attachment pattern in relation to variables that address 30 therapy process and outcome. Dozier, Cue & Barnett (1994) found that secure therapists were better at managing the dependency needs of dismissing clients, thus were less vulnerable to countertransference reactions. These therapists also managed better the needs of preoccupied clients. On the contrary, insecure therapists responded to the preoccupied clients based on the clients? overt behaviors instead of their needs. Fuertes et al. (2007) found a significant and negative relation between therapists? attachment avoidance and clients? secure attachment to therapist. Also, therapists? attachment avoidance and therapists? attachment anxiety were negatively and significantly related to clients? ratings of progress. Among several interesting results about clients? attachment, Romano, Fitzpatrick, & Janzen (2008) found that high to moderate levels of counselor global attachment avoidance, together with high levels of client global attachment anxiety, predicted lower levels of session depth, as perceived by the client. Very few studies have addressed research-connecting attachment with the other constructs of interest in the present study (i.e., countertransference and real relationship). As has been stated, Ligi?ro and Gelso (2002) studied the relations between attachment styles of the therapist, countertransference behavior, and working alliance. In their study, therapist attachment style did not relate to working alliance or countertransference behaviors. A possible explanation for the lack of relation between attachment style and countertransference could be the measure that the authors used to assess attachment. In the same vein, Martin et al. (2007) did not find the expected relation between attachment style and countertransference. This could also be related to the countertransference measure the authors used, as it seems that some of the items do not assess 31 countertransference necessarily: ?I would like to work with this patient?, ?I would like to learn something about the patient?. Slade (2008) brings to the forefront the need to continue developing such type of research, stating that it would be relevant to study the link between countertransference and therapist attachment organization, as different clients will evoke the therapist attachment representations, opening the space for possible problematic reactions. It can be seen that studies show an interaction effect between countertransference and attachment behavior for the therapist and client (e.g., Mohr et al, 2005, reviewed in the previous section). However, results about correlations between therapist attachment and countertransference behaviors have been more equivocal. The present study, then, can help in the clarification of whether results on countertransference and therapists? attachment behavior exist independent of interaction effects (and therefore, not detecting them depends on the way they are measured). Measurement of Adult Attachment and its Classification A final aspect to address in relation to the construct of attachment is the issue of adult attachment measurement. According to Sable (2008), a central challenge for adult attachment is that although there have been important research advancements in this area, there is no general agreement on what can be considered to be attached in adulthood (e.g., what it means to be attached, what relationships are adult attachment, what are the functions of these relationships, among others). This issue can be directly related to the way that adult attachment has been defined, operationalized, and therefore, the focus of the measures developed to assess the construct. 32 As previously mentioned, Ainsworth was the first person who systematically assessed individual differences in attachment, by using coding scales to rate the infant?s behaviors during the Strange Situation. Ainsworth analyzed how the three attachment patterns related to the coding scales, finding that two linear combinations could accurately assign the infants to one of the three attachment patterns she had established: Function I, or Avoidance, and Function II, or Anxiety (Mikulincer and Shaver, 2007). Each of the groups would be located within the graph formed by these two functions: for example, infants with secure attachment would be low in anxiety and low in avoidance, whereas infants assessed as avoidant would be high in avoidance and low in anxiety. Mikulincer and Shaver (2007) state that although Ainsworth?s three categories of attachment patterns (i.e. avoidant - A, secure ? B, and ambivalent - C) could be considered from a two-dimension model (i.e., Anxiety and Avoidance), researchers have focused on the A-B-C category topology instead of the dimensions. Such situation caused that ?attachment theory came to be seen as a topological theory from then on, even though Bowlby had not formulated it as such? (Mikulincer and Shaver, 2007, p. 84). In their analysis on self-report measures of adult attachment, Mikulincer and Shaver (2007) bring to the forefront several issues. The first one is related to the distinction between categories versus continuous scores. The authors state that the problem with categorical measures is that they assume that individual variation within a category is non-existent or non-important. Furthermore, research has shown that it is better to use dimensions when assessing adult attachment via self-report (Fraley and Waller, 1998). Therefore, as a way of addressing all the limitations that categorical assessment of attachment present, researchers have been measuring attachment and 33 related constructs using continuous rating measures (Mikulincer & Shaver, 2007). Another challenge for empirical work in attachment is that adult attachment has been measured with different instruments, which might be measuring dissimilar constructs (Slade, 2008). Adult attachment has been assessed mainly via self-report measures, and relative to romantic relationships, although some measures assess attachment to parents (e.g., PAQ, RAQA). A few measures are in interview mode (e.g., Bartholomew created an interview to assess whether people are Secure, Fearful, Preoccupied or Dismissing, which are the 4 categories of attachment that she determined for adults). Mikulincer and Shaver (2007) provide an extensive and historical review of adult attachment measures, which includes the following assessments: Hazan and Shaver?s adult attachment prototypes, Adult Attachment Questionnaire (AAQ), Adults Attachment Scale (AAS), Attachment Style Questionnaire (ASQ), different measures created by Bartholomew (e.g., Relationship Styles Questionnaire-RSQ), Experiences in Close Relationships Scale (ECR), Inventory of Parent and Peer Attachment (IPPA), Parental Attachment Questionnaire (PAQ), Reciprocal and Avoidant Attachment Questionnaires for Adults (RAQA), among others. Two of the most widely used self-report measures are the ECR and the ECR-R (developed from the same item pool than the ECR, based on item response theory, Fraley, Waller & Brennan, 2000; Crowell, Fraley & Shaver, 2008). According to Mikulincer and Shaver (2007), a substantial amount of research (including ?both experimental manipulations and behavioral observations?, p. 91) has shown the validity of the ECR, which make this authors highlight the value of this measure. Furthermore, Mikulincer and Shaver manifest an inclination for the ECR, as the scales in 34 the ECR-R correlate more with each other, and these authors do not like the new wording of some of the items. Additionally, in relation to the differences between the ECR and the ECR-R, Fraley (2010), one of the developers of the ECR-R mentions ?we are not sure if there are any advantages at this point?, ?I suspect that the ECR and the ECR-R are, for all practical purposes, identical measures of attachment?. A final aspect to address is why assess adult attachment in the context of romantic relationships. Attachment theory proposes that the attachment system continues influencing feelings, thoughts and behaviors through the life span (e.g., Bowlby, 1979; Zeifman and Hazan, 2008). In their review of pair bonds as attachments, Zeifman and Hazan (2008) presented different empirical evidence on the similarities between the pair- bond relationship (?in which sexual partners mutually derive and provide security?, p. 438) and the infant-caregiver bond. Furthermore, the different adult attachment styles have shown to have distinctive love experiences (e.g., secure lovers describe romantic relations as happy and trusting, whereas the avoidant lovers? experience included fear of intimacy and jealousy; Hazan & Shaver, 1987). In addition, Zeifman and Hazan (2008) found that most adults preferred to seek emotional support from partners or friends, instead of their parents. Therefore, a measure of romantic attachment, in this study the ECR, might be an adequate means to assess attachment in adulthood. Therapists internal working model will be reflected in the way they approach relationships, and romantic attachment might be a proxy for such templates of relationship. The Real Relationship As it was previously stated, the real relationship is considered by some to be one of the three therapy relationship constituents (i.e., transference-countertransference 35 configuration, working alliance and real relationship). Such classification of the real relationship was first introduced by the psychoanalyst Ralph R. Greenson, who referred to the real relationship as the non-transference part of the therapeutic relationship (Greenson 1967; Greenson & Wexler, 1969). Furthermore, Greenson and Wexler (1969) indicated that it is pivotal to foster the non-transferential relationship or ?real? interactions between therapist and patient in order to resolve the patient?s transference reactions. Although Greenson pointed out the importance of the construct, researchers have not considered it as central. During the past few years, however, advances in relation to this construct have been lead mainly by Gelso and his collaborators, who have focused their efforts on clarifying and studying the real relationship, both at a theoretical and an empirical level. Current Perspective on the Real Relationship According to Gelso (2009), the real relationship is the personal relationship that exists from the first moment in which two or more people are in contact, and it is a central element in the relationship between client and therapist. In the therapeutic context, the real relationship refers to the personal relationship between therapist and client, (Gelso & Hayes, 1998; Gelso et al., 2005), and is marked by the ??degree to which each is genuine with the other and perceives and experiences the other in ways that befit the other? (Gelso, 2009, p. 255). As it can be seen, the real relationship comprises two defining aspects: Genuineness and Realism. The first feature, genuineness, corresponds to ?? the ability to be what one truly is in the relationship- to be authentic, open and honest? (Gelso and Carter, 1994, p. 297). The second characteristic, realistic perception and reactions, refer to see and experience the other person in ways that suit the other 36 person (versus perceptions that are tainted by our own fears and wishes connected on other people from our past). Furthermore, Gelso and Hayes (1998) state that ?within the real relationship, perceptions of and experiencing with the other are largely realistic or nontransferential? (p. 109). In addition, there are two other sub-elements which are central in the real relationship: Magnitude, which refers to the quantity aspect of the real relationship (i.e., ?how much of a real relationship exists?; Gelso, 2009, p.255, that is, if the levels of genuineness and realism are high or low), and valence, which refers to the extent to which the feelings and attitudes one has toward the other(s) involved in the relationship are positive or negative (Gelso, 2009). Thus, a strong real relationship is indicated by high levels of genuineness and realism, and by positive feelings and attitude towards the other. Considering all the previous information, one could consider the real relationship as the ?authentic relationship? between therapist and client. Moreover, one could argue that in the therapeutic realm, the real relationship is the part of the relationship that relates to the encounter between the self of therapist and the self of the client in the here-and- now, as two human beings who seek an authentic connection, each perceiving him or herself, the other and the relationship in a realistic way. Is the Real-Relationship Absence of Transference-Countertransference? As previously presented, Greenson stated that the real relationship is the personal, non-transferential aspect of the therapeutic relationship. According to Gelso and Hayes (1998), ??all experience contains elements of transference, and the main question pertains to how much transference, how much nontransference? (p. 109). Furthermore, 37 these authors state that if we presume that any relationship will be purely transference or all realistic, we are making an erroneous assumption (Gelso and Hayes, 1998). Gelso (2011), states that although ?realism and transference may occur simultaneously, it seems equally true that realism -or, more broadly, the realistic relationship- represents the transference-free part of every relationship, treatment hour, and communication? (p. 43). The author goes even further, clarifying that although not mutually exclusive, ??the real relationship is the part of any communication (etc.) that is free of transference, or nearly so? (Gelso, 2011, p. 43). Furthermore, Morgan et al. (1998) bring to the forefront the fact that the real relationship between therapist and client allows the client to relate to the therapist in ways that depart from the ways he /she might have related to others in the past. All these statements raise some challenging conceptual questions: Why don?t we restrict the definition of the real relationship in terms of absence or presence of countertransference and/or transference? Do we need a new concept; can?t we only talk about presence or absence of countertransference or transference? As previously stated, the real relationship cannot be defined as the absence of countertransference/transference. First, the concept of countertransference refers to the reactions elicited in the therapist, whereas the real relationship refers to the realistic perception of the other and being genuine in the presence of the other. In addition, the real relationship comprises much more than just a realistic, uncontaminated perception of the other and being genuine; a positive real relationship includes several other aspects, e.g., liking each other and a shared sense of respect (Gelso & Hayes, 1998), and an ?empathic attunement? (Gelso, 2011, p. 81), which cannot be accounted for by the presence or absence of countertransference or transference. 38 Controversial Aspects of the Real Relationship It is relevant to mention that the concept of real relationship has been a highly controversial one. According to Gelso and Hayes (1998), some authors, especially psychoanalysts, resonate more with the realistic aspect of the construct, and do not consider that genuineness is part of the real relationship, whereas authors more in the humanistic line resonate more with the genuineness aspect of it, as they consider that there is no reality without distortion. In addition, it can also be argued that part of the controversy might be related to differences in paradigms; from a constructivist?s point of view, the distinction between transferential vs. realistic-non-transferential might not be relevant, as from this paradigm, ?? notions of ?truth? and ?reality? are abandoned in favor of the notion that ideas about the world, especially in the social world, are constructed in the minds of individuals? (Heppner, Wampold and Kivlighan, 2008, p.11). Gelso (2009, 2011) presents a compelling possible solution to this later dilemma, by introducing the term ?constructive realism?. This concept refers to the idea that, although both, therapist and client, have a reality, the therapist can only access the reality that was co-constructed with the client. The introduction of such concept could be seen as an important synthesis of the different perspectives in debate. Research on the Real Relationship Considering all the constituents of the therapeutic relationship, the real relationship has been the least studied; however, the work including this concept has yielded very interesting results. Overall, empirical studies have illuminated the connection of the real relationship and other variables that are influential in therapeutic process and outcome (e.g., working alliance). Some studies have focused on the clients? 39 perspective. For example, Moore and Gelso (2011) studied the relations among clients? recollection of real relationship, clients? attachment security, and clients? attachment to therapist. The sample consisted of college students who had terminated therapy, and had been in treatment for at least five sessions. Results showed that the strength of the real relationship was positively related to clients? secure attachment and to clients? secure attachment to therapist. In addition, real relationship strength was negatively and significantly related to attachment avoidance, but not to anxiety. Research considering the clients? perspective on the real relationship is possible due to Kelley et al.?s (2010) work, who developed the real relationship inventory-client form. In the development of the therapists? measure of the real relationship, Gelso et al. (2005) found that the therapists rating of the real relationship were positively and significantly related to Working Alliance, evaluation of the session as deep and smooth, and clients? intellectual and emotional insight. In addition, real relationship was negatively and significantly related to clients? negative transference. Fuertes et al. (2007) studied the association of the real relationship perceived by both, therapist and client, with their ratings of working alliance, client progress, client ratings of therapist empathy, and attachment style. These authors studied all these variables at one point of ongoing therapeutic treatment, and found that the therapists? rating of the real relationship was positively related to therapist ratings of the working alliance and of the clients? progress. Additionally, there was a significant positive relation between therapist rated real relationship and clients? ratings of therapists? empathy, and client?s ratings of therapy progress after the third session. These authors also found an interesting relation between real relationship and therapists? attachment style. 40 Specifically, Fuertes et al. (2007) found a negative and significant relation between therapists? ratings of the real relationship and therapists attachment avoidance. There was no significant relationship between real relationship and therapists? attachment anxiety (Fuertes et al., 2007). In relation to the client, Fuertes et al. (2007) found that clients? rating of the real relationship was significantly and positively correlated to clients? perception of working alliance, clients? secure attachment to therapist, client-rated therapist empathy and clients? ratings of therapy progress. When considering clients? insecure attachment styles, there was a significant correlation between clients? rating of real relationship and clients? avoidant-fearful attachment to therapist; however, the correlation between the clients? rating of real relationship and clients? preoccupied-merger attachment to therapist was not significant (Fuertes et al., 2007). The finding that real relationship is related to both therapist and clients? rating of clients? progress is quite remarkable, and highlights the importance of the real relationship in therapy. In addition, the positive relation between therapist and clients? perception of real relationship and perceptions of working alliance highlights the pivotal role that real relationship might have in the therapeutic work. Marmarosh et al. (2009) took the study of the therapeutic relationship further, by investigating how the real relationship relates to therapy process and outcome variables. Specifically, they measured therapist and clients? perceived real relationship after the third session and at the end of treatment, and these investigators also examined its relation to other constructs of interest (e.g. working alliance, transference, outcome measures). A positive relation was found between therapists? ratings of the real 41 relationship and therapists? rating of working alliance, both assessed after the third session. The therapists? rating of the real relationship was negatively related to negative transference (rated after the third session), intake symptoms reported by clients, and to symptoms at termination while partialling out symptoms at Intake. Considering clients? variables, Marmarosh et al. (2009) found that the real relationship perceived by the client was positively related not only to clients? ratings of working alliance, like Fuertes et al. (2007) found, but also therapists? perceived working alliance (all rated after third session). Marmarosh et al. (2009) also found that clients? attachment avoidance correlated negatively with clients? ratings of real relationship after the third session of psychotherapy. Recently, Lo Coco, Gullo, Prestano and Gelso (2011) conducted a study on real relationship, working alliance and therapy outcome. Results showed that for the client, the bond aspect of the working alliance (as measured by the WAI-C) and the genuineness component of the real relationship (as measured by the RRI-C) were related to outcome. Also, hierarchical regression analysis showed that ?the client-related real relationship, especially the Genuineness element, did predict outcome and, moreover, added significantly and substantially to the working alliance in predicting outcome? (Lo Coco et al., 2011, p. 359). In addition, therapists? rating of the real relationship was positively related to clients? rating of the bond in the working alliance. These authors also found a positive and significant relation between clients? and therapists? ratings of the real relationship. Interestingly, neither Fuertes et al. (2007) nor Marmarosh et al. (2009) found a significant correlation between therapist perception of the real relationship and client?s 42 perception of the real relationship. A closer a look at the subscales of the real relationship measures in Marmarosh?s et al.?s (2009) study showed a correlation between client genuineness and therapist genuineness. An interesting empirical challenge could be to discover whether the rating of the real relationship done by an external observer correlates to the rating of the real relationship done by a therapist or client. A trainees? supervisor could be a relevant rater to assess the perceived real-relationship, as she/he has discussed the case with the trainee, having extra information on the case. One might even expect that the discussion of the case with a supervisor influences the perception of the real relationship that the therapist has, especially if as part of the supervision work possible distortions of the client are worked through. Therefore, the current study could contribute to discover what is the relation between the real relationship and countertransference, which has never been studied, if the results of therapist?s attachment and real relationship hold, and if there are significant differences in the measurement of supervisors and therapists. As was established in the literature review on countertransference, attachment, and real relationship, some of the research results in relation to the variables of interest are inconclusive. One of the aspects that could have influenced such results are the measures used, which in measuring attachment, might have fall short in the assessment of the relations of the constructs of interest (e.g., relationship between therapist attachment and countertransference behavior). In addition, to the author?s knowledge, there is no study up to date that analyzes the relation between countertransference and real relationship, and that analyzes the difference between therapists and external raters of the 43 real relationship. Therefore, the present study could be an addition to the literature on therapists? contribution to the therapeutic work. 44 Chapter 3: Statement of the Problem and Hypotheses Statement of the Problem Although the centrality of the role of the therapeutic relationship in psychotherapy has been well established (Gelso & Carter, 1985, 1994; Gelso & Hayes, 1998; Norcross, 2002; Strupp and Binder, 1984; Ackerman et al., 2001; Wampold, 2001, among others), the literature is rather scant in relation to the establishment of the relationship, its maintenance, and knowing why it works (Division 29 Steering Committee, 2001). Lately, there have been efforts directed at ??identifying elements of effective therapy relationships? (Norcross & Lambert, 2011). For example, in Psychotherapy there was a recent special issue on ?Evidence-based psychotherapy relationships?, which underscores the idea that research that examines what influences the therapeutic relationship can be highly valuable for the field of psychotherapy research and practice. There has been a proliferation of studies focusing on the client-therapist dyad, which has illuminated how the ratings of both members of the dyad are connected in relation to different variables, such as attachment (Mohr et al., 2005; Marmarosh et al., 2009, among others). When considering therapist variables, some studies have focused on the effects of countertransference in the relationship, or the therapists? attachment, but few have related how therapist factors can relate to the strength of the established real relationship, which is the least studied component of the therapeutic relationship. In addition, some authors have addressed the relation of attachment and countertransference (Mohr et al., 2005), attachment and transference (Woodhouse et al., 2003), real relationship and working alliance (Lo Coco et al., 2011), real relation and attachment (Fuertes et al., 2007; Marmarosh et al, 2009) and attachment, countertransference and 45 working alliance (Ligi?ro & Gelso, 2002), among others. There is no study to date that relates countertransference and real relationship. In addition, although attachment theory has been a fertile field for research in the past years, research on attachment within the psychotherapy encounter has generated more questions than answers (Slade, 2008). In order to contribute to knowledge about both the therapeutic relationship and the person of the therapist, the proposed study seeks to shed light on the relations between attachment styles of therapists-in-training, the countertransference behavior of these therapists, and ratings of the real relationship they establish with a client. Specifically, the goal of the present study is to illuminate the relationships between a) therapists? attachment pattern and the therapists? perspective of the real relationship; b) therapists? rating of the real relationship and supervisors? rating of therapists? countertransference behavior; and c) therapists? attachment and countertransference behaviors, as rated by the supervisors. In addition, supervisors? rating of the real relationship will be correlated with countertransference behavior, therapists? attachment, and therapists? rating of the real relationship. Hypotheses As previously mentioned, and based on the review of the literature, there are several goals for the present study. The first one is to assess the relations between trainees? attachment patterns or behavior and ratings of the real relationship, specifically, to examine the relationship between trainees? attachment and their ratings of strength of the real relationship. Considering the previous question, it can be hypothesized that: Hypothesis 1: There will be a positive relation between degree of attachment security of the counselor and strength of real relationship, such that 46 Hypothesis 1.a. The greater the attachment security of the trainee, the stronger the real relationship as rated by the therapist trainee Hypothesis 1.b. The greater the attachment security of the trainee, the stronger the real relationship as rated by the supervisor As was previously presented in the literature review, it has been shown that therapists rating of the real relationship are related to therapeutic outcome. Specifically, there is a positive correlation between the therapist?s rating of the real relationship and his/her rating of the client?s progress (Fuertes et al, 2007) and outcome (Marmarosh et al, 2009, p. 337; Lo Coco et al., 2011). However, there have been few efforts at clarifying the specific associations between attachment of the therapist and their ratings of real relationship. Fuertes et al. (2007) found that there was a negative relation between therapist?s attachment avoidance and his/her rating of the real relationship. Considering such finding, one could think that the more secure a therapist is, the stronger would be the real relationship with his/her client, as stated in hypothesis 1. Indirect support for such a hypothesis comes from the work of Black et al. (2005), who found that the therapist?s secure attachment was significantly and positively associated to good alliance, as reported by the therapist. As presented in the literature review, one of the three components of Bordin?s (1979, 1994) model of the working alliance, is the bond between therapist and client, which is considered as the emotional relationship established between therapist and client for the purpose of the work. One could extrapolate such findings and expect the same association when considering real relationship instead of working alliance. It is relevant to mention that Ligi?ro and Gelso (2002) did not find a relation between therapist attachment style and working alliance; however, as it was 47 previously mentioned, such lack of connection could be related to the measure of attachment that they used. For the present study, a different measure of attachment (the ECR) will be used, which has been one of the most widely used measures in adult attachment (Mikulincer and Shaver, 2007). Finally, subhypothesis 1a and 1a address the fact that there will be two different ratings on the real relationship, both of which are expected to positively relate to therapists attachment. A second aspect to explore in this study is the relation between real relationship and countertransference, specifically the relation between the trainees? rating of strength of real relationship and their countertransference behaviors. Also, what is the relation between supervisors? rating of real relationship and trainees? countertransference behavior? Considering the first question, it can be expected that: Hypothesis 2: The counselors? rating of the strength of the real relationship will be negatively correlated to the supervisors? rating of negative countertransference behavior, such that the stronger the real relationship, the fewer the countertransference behaviors. To the author?s knowledge, there are no studies up to date that relate countertransference and real relationship; therefore, there are no empirical findings to directly support a specific relationship between these two variables. Support for the specified relationship between countertransference and real relationship comes indirectly from research involving a similar construct. Ligi?ro & Gelso (2002) found that ??negative countertransference was associated with poorer working alliance? (p. 3), for both, counselor and supervisors? ratings of the working alliance. Such significant negative association between negative countertransference and working alliance holds for the working alliance as a whole and for each of its components (i.e., bond, task, goal). 48 As previously stated, in Bordin?s (1979, 1994) model of the working alliance, such bond refers to the emotional connection between therapist and client, which is established for the purpose of the work. If we think of the real relationship between the therapist and a client as the personal relationship among them, it could also be expected a similar relation between countertransference and real relationship. In addition, Ligi?ro & Gelso (2002) found that positive countertransference was negatively related to the bond aspect of the working alliance, for supervisors rating of working alliance. Considering such findings, it can be hypothesized that: Hypothesis 3: There will be a negative correlation between real relationship rated by supervisors and countertransference behaviors, such that Hypothesis 3.a. Supervisors? rating of the real relationship established by the trainees with their client will be negatively correlated to the supervisors? rating of therapists? negative countertransference behaviors. Hypothesis 3.b. Supervisors? rating of the real relationship established by the trainees with their client will be negatively correlated to the supervisors? rating of therapists? positive countertransference behaviors. In addition, considering the association between countertransference and real relationship, another aspect that might be relevant to explore is the disagreement between trainees and supervisors? rating of the real relationship, and countertransference. Does the rating of the real relationship determined by a therapist in training correlate with the rating that the supervisor of the trainee would give to the real relationship? If so, how does that correlation relate to countertransference behaviors? In light of such questions, it could be expected that: 49 Hypothesis 4: The higher the level of disagreement between the trainee and supervisors? rating of real relationship, the more countertransference behaviors from the therapist. Support to such hypothesis can be found in Ligi?ro & Gelso?s (2002) work, who found that a predictor of countertransference was the discrepancy between therapist and their supervisors in ratings of the bond component of the working alliance. Finally, a third goal for the present study is to explore the relationship between the trainees? level of secure attachment and countertransference behavior. Therefore, it can be expected that: Hypothesis 5: The level of secure attachment will be negatively related to amount of countertransference as rated by supervisors, such that the higher the level of security, the fewer the countertransference behaviors of the trainee. As previously presented, empirical findings considering the relationship between attachment security and countertransference behavior have been inconclusive. Ligi?ro & Gelso (2002) did not find a correlation between countertransference and attachment style of the therapist. In this same line, Martin et al. (2007) found no association between therapists and medical students? attachment style and their countertransference reactions. On the contrary, Mohr et al. (2005) found that the counselor trainee?s attachment was related to certain aspects of countertransference. In addition, the interaction of the therapist and client attachment was related to countertransference. Such discrepant findings could be related to the measurement of the constructs of interest. Specifically, the instruments used to measure attachment in Ligi?ro and Gelso (2002) and countertransference in Martin et al. (2007) might have not detected the relationship between these two variables. For the current study, the measure that will be used to assess 50 attachment is the same one that Mohr et al. (2005) used (i.e., ECR), and the countertransference measure is the same one that Ligi?ro and Gelso (2002) used (i.e., ICB) and Mohr et al. (2005) used (i.e., CBM); therefore, it can be expected that there might be a relationship between these variables. Hypothesis 5 can help us start clarifying whether there is a relationship among therapists? attachment security and countertransference, or not. 51 Chapter 4: Method Participants The participants were 32 therapists in training who were receiving psychotherapy supervision at the time of the study, and their 28 clinical supervisors. Trainees were recruited from Ph.D. level graduate programs in counseling and clinical psychology at two large eastern public Universities. In addition, in one of these universities invitation to participate was also extended to students in the rehabilitation and counselor education Master?s and Ph.D. programs, and interns at the university?s counseling center. Therapists in training were recruited via email, and a total of 120 email invitations were sent. From this total, 33 trainees declined to participate. The primary reasons students gave to decline participation were that they were working with families, were not seeing clients at the time, or were working with clients in no more than three sessions. 39 people did not respond after repeated emails (two to three follow-ups). Therapists. Of the total of trainees, 25 were females (78.1%) and seven males (21.9%). Their mean age was 27.69 years old (SD=3.04). In relation to race, eight trainees self-identified as Asian (25%), two as Black (6.3%), 21 as Caucasian (65.6%) and two marked other (Hispanic and Indian, 6.3%). Percentages were calculated based on N = 32, but participants could mark more than one race; therefore, percentages add up to more than 100%. Participants were also asked about ethnicity. Details on ethnicity can be found on the Table 1 presented in Appendix A. In relation to their most advanced degree, six participants said B.A/B.S (18.75%), 19 stated that an M.A. or M.S. was their most advanced degree (59.38%), six had an M.Ed (18.75%), and one had an M.S.W. (3.12%). From the total participants, only one was enrolled in Master?s level training; all the other 52 participants (31) were enrolled in Ph.D. programs in either counseling or clinical psychology. Twenty-eight trainees were enrolled at one university, and four at the other. Trainees were also asked about their theoretical orientation. Specifically, participants had to rate on a scale from 5 (Strongly Representative) to 1 (Not at all), how representative of their work were several theoretical approaches. Therefore, trainees might have included more than one theory as part of their personal approach to therapy. The mean for representativeness of Humanistic Experiential theory was 3.59 (SD = 1.19), for Psychodynamic/Psychoanalytic theory the mean was 4.06 (SD = 1.27), the mean representativeness of Cognitive/Behavioral theory was 3.22 (SD = 1.29), for Systemic theory was 2.41 (SD = 1.10), and other was 2.19 (SD = 1.47). The different theories/perspectives that participants wrote when they chose ?other? included: feminist, multicultural, interpersonal, and gestalt. Trainees were also asked to report their years of clinical experience in general. On average, the trainees had been providing therapy for 3.62 years (SD = 1.97). Specific data in relation to the case considered for this study was also collected. Average number of sessions with the client identified to complete the measures was 15.91 (SD = 14.61), and session number ranged from 3 to 64 sessions. Therefore, there was great variance in the amount of sessions that the participating therapists in training had with their clients. In addition, is relevant to note that the measure could be completed at any point in treatment, as long as the therapist and client had met at least three sessions. The number of sessions trainees had with their supervisors was 11.26 on average (SD = 8.25, ranging from 3 to 36). Trainees also completed information on how the supervisor knew about 53 their work with the client. Such information can be found in the Table 2 presented in Appendix A. Supervisors. The supervisor sample consisted of 28 supervisors, four of whom rated two supervisees each (thus, 32 therapist-supervisor dyads). Of the supervisors, 22 were females (78.6%) and six males (21.4%), and their mean age was 42.5 years old (SD = 12.48). In relation to race, two supervisors self-identified as Asian (7.1%), two as Black (7.1%), 23 as Caucasian (82.1%) and one marked other (3.6%; Hispanic/mixed). Supervisors were also asked about ethnicity. Details on ethnicity can be found on the Table 1 presented in Appendix A. In terms of the most advanced degree attained, one supervisor said that M.A. or M.S. was his/her most advanced degree (3.6%), two had a M.S.W. (7.1%), 24 reported having a Ph.D. (85.7%), and one marked ?other?, reporting having a Psy.D (3.6%). In terms of clinical experience, supervisors had an average of 17.14 years providing therapy (SD = 12.05, ranging from 5 to 41 years). In terms of supervision experience, supervisors had an average of 10.16 years providing supervision (SD = 11.77, ranging from 1 to 41 years). In addition, based on a scale ranging from 5 (Strongly Representative) to 1 (Not at all), supervisors were asked to identify how representative of their work were different theoretical approaches. In relation to representativeness of different theoretical orientations, the mean representativeness for Humanistic Experiential theory was 3.43 (SD = 1.29), for Psychodynamic/Psychoanalytic the mean representativeness was 3.79 (SD = 1.32), for Cognitive/Behavioral was 3.5 (SD = 1.04), for Systemic theory was 2.82 (SD = 1.16), and for ?other? the mean was 2.18 (SD = 1.49). The different theories/perspectives that supervisors wrote when they chose other included: feminist, 54 gestalt, multicultural, interpersonal, narrative, relational, existential, solution focused, object relations, and attachment. Supervisors were also asked different information in relation to the particular case on which they would base their assessments. In this particular area, data was analyzed considering an N=32, as each trainee-supervisor dyad was unique (as mentioned, there were 28 different supervisors but 32 supervisor-therapist dyads). On average, supervisors estimated that their supervisees had worked 13.50 sessions with their clients (SD = 9.89, ranging from 2 to 45 sessions). In relation to number of supervision sessions related to the identified client, the mean number of supervision sessions determined by supervisors was 10.43 (SD = 8.73, ranging from 3 to 42 sessions). The amount of sessions with the supervisor directed to discuss the identified case was 11.26 on average (SD = 8.25, ranging from 3 to 36). Supervisors also completed information on how they knew about the trainees? work with their client. Such information can be found in the Table 2 presented in Appendix A. Finally, it should be mentioned that both members of the participating dyad (i.e., therapist and supervisor) were monitored to ensure they completed measures within no more than two weeks from each other. This was to ensure that therapists and supervisors were completing measures considering the same sessions between the therapist and the client. Power analysis An a priori power analysis was conducted to determine the amount of participants needed for the current study to detect the anticipated effects. Based on the results of a previous study that examined a very similar topic as the present one (Ligi?ro & Gelso, 2002), correlations with medium to large effect sizes were expected in the present study. 55 According to Cohen (1992), if alpha is set up at .05 and we want an 80 percent likelihood of detecting effects, the sample needed to detect a medium size effect would be N = 85, and for a large effect N = 28. The current sample consisted of 32 dyads. Calculations on G*Power 3 show that for such sample size, and setting alpha at 0.05, there is a 64 percent likelihood of detecting a medium size effect for a one-tailed test, and a 51 percent likelihood of detecting a medium size effect for a two-tailed test. Measures Consent Form. For the current study, two consent forms were developed, one for therapists in training and one for supervisors. Such forms were adapted from Ligi?ro (2000) and Ain (2011), and provided a brief description of the study and the procedures to follow. In addition, they presented information on confidentiality and the rights to withdraw at any time (See Appendices B and C). Demographic questionnaire for therapist trainees. A self-report demographic questionnaire for therapist trainees was developed. This questionnaire was based on Ligi?ro?s (2000) paper and pencil demographic questionnaire for therapists, and Ain?s (2011) online demographic questionnaire for therapists. The question?s included asked about therapist?s sex, age, race, ethnicity, type of degree, theoretical orientations, amount of sessions with the client to consider for ratings, and amount of sessions of supervision about the client, among others (See Appendix D). Demographic questionnaire for supervisor. A self-report demographic questionnaire for supervisors was developed based on Ligi?ro?s (2000) paper-and- pencil demographic questionnaire for supervisors and Ain?s (2011) online demographic questionnaire for therapists. The questions included targeted general information such as 56 sex, age, race, ethnicity, most advanced degree, theoretical approach, years of clinical experience, years of experience as supervisor. In addition, some questions addressed the work of the supervisor with the particular supervisee, such as amount of supervision sessions with the trainee, and how did the supervisor know about the case identified for the study (i.e., discussed in depth, heard audio of the session, watched video of the session), among others (See Appendix E). Experience in Close Relationships Scale (ECR; Brennan, Clark and Shaver, 1998). The ECR was used to assess therapist?s attachment. In order to construct this scale, Brennan et al. (1998) conducted a literature search of all the self-report measures of attachment at the time of their study, and then created a pool of items that assessed 60 attachment constructs. These items were presented to approximately 1000 undergraduate students. Then, the authors conducted a factor analysis of those 60 subscale scores, which produced two factors that corresponded to the avoidance and the anxiety dimensions that had been previously described in the literature (Brennan et al., 1998). The end result was the ECR, a 36-item self-report scale, in which each item is rated on a 7-point scale (1 = disagree strongly, 4 = neutral/mixed, 7 = agree strongly). This scale assesses the two dimensions of adult romantic attachment: Avoidance (18 items) and Anxiety (18 items). The first subscale, avoidance, assesses the level of comfort in being close to others, intimacy and self-reliance, among others (e.g, ?Just when my partner starts to get close to me I find myself pulling away?). The second subscale, anxiety, assesses fear of abandonment and/or rejection, jealousy, desire of more closeness than the partner, among others (e.g., ?I worry a lot about my relationships?). Instructions ask participants to respond considering how one experiences romantic 57 relationship in general, not just related to a current relationship. For the current study, scores on the ECR were used to assess attachment security of the trainees within adult romantic relationships. Attachment security was calculated following the procedures utilized by Fraley and Shaver (1997), Mohr, Gelso and Hill (2005), and Moore and Gelso (2011), by adding the anxiety and avoidance scores, and taking the inverse of this additive combination to calculate degree of security. Thus, higher scores reflected higher security. In relation to validity, Brennan et al. (1998) found the theoretically expected associations between their subscales and measures of touch, and sexuality in romantic relationships (e.g., sexual preferences and emotions after sexual activity). In relation to internal consistency, both, the anxiety and the avoidance subscales have demonstrated high internal consistency estimates (alpha greater than .90; Brennan et al., 1998). For the current sample Cronbach alpha was .95 for the ECR (i.e., what we termed security), .93 for the avoidant scale, and .91 for the anxious scale (See Appendix F). Inventory of Countertransference Behavior (ICB; Friedman & Gelso, 2000). The ICB is a 21-item measure, created to assess countertransference behaviors in counselor-trainees as perceived by their supervisors. To construct the scale, the authors created a 32-item scale and sent it to a group of countertransference experts who were asked to judge face validity of each item. In addition, the 32-items scale was completed by 126 psychologists and counselor educators who were conducting supervision at the time. The data was analyzed using an exploratory principal component factor analysis with oblique rotation, which showed a two-factor solution, termed positive countertransference and negative countertransference. 58 The format of the ICB is a 5-point Likert scale, with responses ranging from 1 (to little or no extent) to 5 (to a great extent), where the higher the score, the more countertransference behavior is being displayed in the sessions. As indicated, this scale includes two factors: Positive countertransference behaviors (e.g., - the counselor - ?Befriended the client in the session?) and Negative countertransference behaviors (e.g., - the counselor- ?Was critical of the client during the session?). The measure yields three scores: a negative countertransference behavior, a positive countertransference behavior score, and an overall score. Regarding validity, as it was previously stated, experts in the area of countertransference (11 Ph.D.-level psychologists) evaluated the face validity of the items. The experts had to rate each item in relation to how representative it was of CT behavior. Based on an apriori determined cut-off score, Friedman and Gelso (2000) found that all the items seemed to represent CT behaviors. Experts also provided feedback, which resulted in one item being dropped from the measure, due to its openness for interpretation. In addition, the ICB was found to possess adequate convergent validity, relating negatively to a measure of countertransference management ability and positively to a single-item measure of countertransference behavior in a session. The reported alpha coefficient by Friedman and Gelso (2000) is of .83 for the total subscale, and of .79 for each subscale. In the current study, the Cronbach alpha values obtained were as follows: Total scale = .85, Negative Countertransference = .88, and Positive Countertransference = .59. (See Appendix G) Countertransference Behavior Measure (CBM; Mohr, Gelso, & Hill, 2005). This measure is a 10-item scale, which assesses supervisees? countertransference behaviors as perceived by their supervisors. The CBM allows the assessment of ?specific 59 interpersonal behaviors? (Mohr et al., 2005, p. 301). To develop the CBM, Mohr et al. (2005) conducted maximum likelihood factor analysis with oblique rotation, with the original items used for the development of the ICB. This analysis resulted in three subscales: Dominant Countertransference Behavior (5 items), Distant Countertransference Behavior (2 items), and Hostile Countertransference Behavior (3 items). Then the authors conducted a confirmatory factor analysis with the original data used in the construction of the ICB, which supported the three-subscale structure. It should be noted that the Mohr et al. (2005) analysis showed strong positive skewness. Thus, logarithmic transformations were used. The coefficient alpha values obtained were: Dominant = .89, Distant = .82, and Hostile = .82. For the current study, the coefficient alpha values were as follows: Dominant = .87, Distant = .93, and Hostile = .21. It is relevant to mention that the CBM only adds four items to the ICB (21 item measure). Thus, the countertransference behaviors measures were presented as just one (See Appendix G). The Real Relationship Inventory-Therapist Form (RRI-T; Gelso, Kelley, Fuertes, Marmarosh, Holmes, Costa & Hancock, 2005). This is a 24-item, self-report measure that assesses how the therapist evaluates the strength of the real relationship established with a client. To develop this scale, Gelso et al. (2005) created items that theoretically would reflect the construct of real relationship, capturing genuineness and realism (the two theoretical components of the real relationship), and that incorporated magnitude (how much) and valence (how positive or negative). These items were sent to randomly selected members of Division 29 (Psychotherapy) and Division 42 (Independent Practice) of APA. The sample was divided into an item - development subsample and a validation 60 subsample. The item - development subsample completed a 44-item measure, and psychometric analysis of the results lead to a 24-item measure. In parallel, there was a back translation of such items. The validation subsample, consisting of the previously mentioned practicing therapists plus students in counseling graduate programs, completed the 24-item measure. Data from the 79 practicing therapists and 51 graduate students who completed the measure were analyzed using Confirmatory Factor Analysis (CFA), which supported a one-factor model. However, the authors maintained the two subscales based on theory and on differential correlations that the subscales might have with specific constructs. The RRI-T is composed by two subscales, the Realism subscale and the Genuineness subscale. These two factors, realism and genuineness, are the ones theoretically proposed as components of the real relationship (Gelso and Carter, 1994; Gelso & Hayes 1998; Gelso & Hayes, 2007). Considering validity, the RRI-T was related in theoretically expected ways to measures of working alliance, session depth and smoothness, client?s insight (emotional and intellectual), and client?s negative transference. In addition, as expected, the RRI-T did not correlate with social desirability, which was used to determine discriminant validity. In relation to the reliability, the coefficient alpha values obtained were Realism = .79, Genuineness = .83, and Total score = .89. The coefficient alpha values obtained in the current study were as follows: Realism = .78, Genuineness = .77, and Total score = .88. (See Appendix H) The Real Relationship Inventory-Supervisor Form (RRI-S). This is a 24-item measure designed to assess a supervisor?s evaluation of the strength of the real relationship between a supervisee and his/her client as perceived by the trainee?s 61 supervisor. The RRI-S was developed for the current study, and was based on the RRI-T. Specifically, the same items of the RRI-T were rephrased in a way that reflected the items, but from an external-observer perspective. The items for the scale were reworded by the researcher and her advisor, who extensively studied the real relationship. Once the items were clearly phrased and seemed to reflect the idea captured in the original RRI-T items, the resulting measure was completed by a group of four graduate students in a Counseling Psychology program and the previously mentioned professor. From this application came new suggestions and rewordings, which were incorporated in the final measure. Reliability was calculated using coefficient alpha, and the values obtained were as follows: Realism = .75 Genuineness = .78, and Total score = .87. Validity data have not yet been gathered for the RRI-S. (See Appendix I). Procedure Recruitment of therapist. Participants were recruited mainly from the University of Maryland ? College Park, but also from the Pennsylvania State University ? University Park. Invitations to participate were sent to therapists- in- training in different programs at both universities (further detail in the next section). Identification of potential participants at University of Maryland. In order to get in contact with the potential trainee participants for the study, the first step in the recruitment process was to contact the academic programs at the University of Maryland where students are receiving counseling/clinical training. The departments and programs contacted were: Psychology Department (Counseling Psychology Ph.D. and Clinical Psychology Ph.D. programs), and Department of Personnel and Counseling Services (Rehabilitation Counseling program -Masters and Ph.D.-, School Counseling program - 62 Masters-, and College Student Personnel program-Masters and Ph.D.). The Family Science Department (M.S. in Couple and Family Therapy) was contacted, but was excluded from the study, as they mentioned that their students do not provide individual therapy. In each of these programs, the investigator contacted a person who had regular contact with students (e.g., person in charge of graduate students), discussed the nature of the study with him/her, and asked for a list of therapist trainees (names and contact information) who, at a minimum, were enrolled in their first counseling practicum and who were in ongoing supervision. The researcher also contacted some professors who were teaching courses with a clinical component. In one case, the researcher spoke to a class and invited the students to participate, and in the other, the professor forwarded an email about the project to his students. In addition, the researcher contacted three clinics at the University of Maryland, College Park campus, where students work as externs or interns. These clinics were the Counseling Center, the Center for Healthy Families, and the Psychology Clinic. Some of the trainees in the Counseling program that agreed to participate had supervisors at the Counseling Center, so the Center asked the investigator to apply to their own IRB process, and finally accepted participation in the study. The Center for Healthy Families works from a family therapy perspective thus was excluded. Identification of potential participants at The Pennsylvania State University. The researcher contacted two professors at Penn State: One in the Clinical Psychology program and one in the Counseling Psychology program, and asked for a list of the students in their program who were currently seeing adult clients under supervision. These professors shared the names and email addresses of the students in their program 63 who would fulfill the requirements for the study. IRB approval was obtained from The Pennsylvania State University. Direct contact with potential participants. First contact with potential participant trainees. Once contact information of the potential participants (including name, email and/or phone number) was obtained, a personalized email was sent to the student, including an invitation to participate and a general overview of the study (See Appendix J). Students were told to please reply through a phone call or email if they were interested or had questions. Also, as previously mentioned, the researcher talked directly to a class, and handed in a printed copy of the initial email, in case the therapists-in-training wanted to participate. Second contact with trainees who agree to participate. Once a student agreed to participate, the researcher sent him/her a new email, including a link to complete the measures online (See Appendix K). This email also contained a code, which was unique for each trainee-supervisor dyad (e.g., 001). The trainee was asked to access the link and to complete all the measures with the assigned code. Links varied based on whether the code was even or odd, due to counterbalance of the measures (explained later). Trainees were told that they must have met with the supervisor for at least five times before they both completed the measures. Trainees were assured of confidentiality, and were informed that the only way to track a particular trainee and his/her completed measures was to go to the database in which codes and names were matched. Trainees were also asked to send the researcher the name of their supervisor and the contact information for them. Finally, besides the code number and the link for the measures, this new email also included the specific criteria to choose the client that trainees needed to have in mind 64 when completing the RRI-T (and for supervisors to complete their measures), and how to proceed with the supervisor. Choosing a client to have in mind while completing the measures. In relation to choosing the client, the researcher adhered to requirements that Ligi?ro and Gelso (2002) stipulated. These requirements were: Current client, attending at least three sessions, the case had been discussed in supervision, and that the audiotapes of the sessions were listened by the supervisor (or the videotapes of the sessions were watched). In case that the supervisor did not use audio nor videotape, the requirement was that supervisors were familiar with the case (i.e., in-depth discussion of the case). In order to standardize the procedure, in the second email sent to each therapist, he/she was asked to follow a specific process: First, the therapist needed to discuss the study with his/her supervisor (as stipulated in first email). If the supervisor agreed to participate, the therapist and the supervisor had to review the guidelines to identify the client that both would consider to complete the measures. The client needed to be the first client the therapist would see right after meeting with the supervisor, and who met the requirements that were previously presented. After the trainee had identified with the supervisor a specific client to have in mind for rating, the trainee could go online to complete the measures. In addition, the final step in the process was that the trainee had to send an email to the researcher with the supervisor?s contact information, so that the researcher could send the supervisor the information of the study and the link to the measures. Completing the measures. Once the trainee accessed the link, there was a general presentation of the study, in which the trainee was reminded of adding the specific code 65 he/she was given to complete all the measures. In addition, there was a reminder for the trainee to complete the measures having in mind the client identified jointly with the supervisor. Then, the trainees completed an Informed Consent. This was followed by a demographic questionnaire, and then the ECR and the RRI-T. As was previously stated, the presentation of the study with the Informed Consent, the demographic survey, and the measures can be found in the appendices. Finally, it should be noted that trainees knew that the current study included the supervisors? perspective on the client case. They were told that it will entail their assessment of the relationship between the trainee and the client, and some of the trainee?s behaviors, but were never be told that supervisors were measuring countertransference. Recruitment of supervisors. Once the researcher received the supervisor?s contact information from a trainee, an email was sent to the supervisor (See Appendix L). The email mentioned that the trainee working in supervision with him/her had agreed to participate. In addition, the email explained the supervisor?s role in the study, and contained detailed instructions to complete the measures. Just like with the trainees, a code was given to the supervisors (same as that for trainees), and they were ensured that the only way to track a particular supervisor and his/her completed measures was to go to the database in which codes and names are matched. As with the therapists, the links varied based on whether the code was even or odd, which allowed a counterbalance of the measures (explained later). Completing the measures. Once the supervisors accessed the link, there was a general presentation of the study, in which supervisors were reminded of adding the 66 specific code they were given to access the measures. In addition, there was a reminder for them to complete the measures having in mind the client that him/her and his/her supervisee agreed upon, due to the criteria for choosing a client. Then, the supervisors completed an Informed Consent. This was followed by a demographic questionnaire, and then the RRI-S and the Countertransference measures. The presentation of the study with the Informed Consent, the demographic survey, and the measures can be found in the appendices. Measure Application. Trainees first completed an informed consent and demographic questionnaire, which was followed by the measures. To control for order effects, there were two links sent differently to trainees: the trainees with the odd numbers received a link in which they rated first the ECR, and then completed the RRI-T. The even numbers completed the RRI-T first, and then the ECR. The same counterbalanced procedure was followed for supervisors (i.e., odd numbers=ICB/CBM and then Real Relationship measure; even numbers=Real relationship measure and then ICB/CBM). Supervisors also completed an informed consent and demographic questionnaire at the beginning. Completion of the measures took approximately 15 minutes total for each, supervisor and therapist. 67 CHAPTER 5: RESULTS Descriptive Data Descriptive data were calculated for all the measures used in the present study. The means and standard deviations for each measure and its subscales are presented in Table 3. Table 3 also includes indices of internal consistency of the measures, which were estimated using Chronbach?s alpha. As seen in Table 3, alpha coefficients are above .75, except for Countertransference Behavior Measure (CBM)-Hostile and Inventory of Countertransference Behavior (ICB)-Positive. Results including the CBM-Hostility and ICB-Positive subscales are included in the present section. These should be interpreted with caution. For the most part, the means found in the present study, as presented in Table 3, were highly similar to those found in other published studies on related topics. For example, the means scores obtained for the strength of the real relationship as rated by the therapist were similar to the ones obtained by Marmarosh et al. (2009) in their study (RRI-T Total M = 3.61, RRI-T Realism Subscale M = 3.68, RRI-T Genuineness Subscale M = 3.58). The mean item scores obtained for therapists? attachment avoidance and attachment anxiety are similar to those in Mohr?s (2001) study (ECR Avoidance M = 2.46, ECR Anxiety M = 3.49). The mean scores obtained for the ICB as rated by supervisors were similar to those in the study by Ligi?ro and Gelso (2002; ICB-Negative M = 1.40, ICB-Positive M = 1.73) and Mohr (2001; ICB-Negative M = 1.34, ICB- Positive M = 1.59). Finally, the CBM scores obtained are slightly higher than those obtained by Mohr (2001; CBM-Dominant M = 0.08, CBM-Distant M = 0.17, CBM- Hostile M = 0.00). 68 Table 3 Mean Item Scores and Standard Deviations for Real Relationship Inventory (RRI) for therapists (T) and supervisors (S), Experiences in Close Relationships Scale (ECR), Inventory of Countertransference Behavior (ICB), and Countertransference Behavior Measure (CBM) Measure M SD Alpha RRI-T 3.80 0.45 .88 RRI-T - Realism 3.82 0.45 .78 RRI-T - Genuineness 3.79 0.49 .77 RRI-S 3.63 0.43 .87 RRI-S -Realism 3.67 0.41 .75 RRI-S -Genuineness 3.59 0.49 .78 ICB 1.31 0.33 .85 ICB -Positive 1.42 0.32 .59 ICB - Negative 1.22 0.42 .88 CBM - Dominant 1.26 0.48 .87 CBM - Hostile 1.16 0.33 .21 CBM - Distant 1.23 0.68 .93 ECR - Security 2.85 0.83 .95 ECR - Avoidant 2.15 0.86 .93 ECR ? Anxiety 3.55 0.98 .91 Absolute Value RRIT-RRIS 0.46 0.33 69 Note. RRI-T = Real Relationship Inventory as rated by Therapists; RRI-T-Realism = Realism Subscale of the Real Relationship Inventory as rated by Therapists; RRI-T- Genuineness= Genuineness Subscale of the Real Relationship Inventory as rated by Therapists; RRI-S = Real Relationship Inventory as assessed by Supervisors; RRI-S- Realism = Realism Subscale of the Real Relationship Inventory as rated by Supervisors; RRI-S-Genuineness = Genuineness Subscale of the Real Relationship Inventory as rated by Supervisors; ICB = Inventory of Countertransference Behavior; ICB - Positive= Positive Countertransference Subscale of the ICB; ICB - Negative= Negative Countertransference Subscale of the ICB; CBM-Dominant= Dominant Scale of the Countertransference Behavior Measure; CBM-Hostile= Hostile Scale of the Countertransference Behavior Measure; CBM-Distant= Distant Scale of the Countertransference Behavior Measure; ECR-Security= Attachment security assessed by the inverse of the avoidance plus anxiety scores; ECR-Avoidance=Avoidance scale of the Experiences in Close Relationships Scale; ECR-Anxiety=Anxiety scale of the Experiences in Close Relationships Scale; Absolute Value RRIT-RRIS= Difference in absolute value of therapist and supervisor ratings of the Real Relationship. 70 Test for Normality and Supervisor Effect The first step in the data analysis was to test for normality. Due to the small sample size (N=32), normality was tested using the Shapiro-Wilk test of normality. Both this quantitative test of normality and the Normal Q-Q plots showed that some scales were violating normality assumption (such as skewness and kurtosis). Specifically, skeweness coefficients showed that all the countertransference variables presented substantial positive skewness; therefore a logarithmic transformation (Friedman & Gelso, 2002; Mohr et al. 2005) was applied to the scales that presented a skewness coefficient greater than one (in terms of absolute value). In all cases skewness diminished, yet in only one case the absolute value fell below 1. Therefore, based on such analysis, Spearman?s Rho was used to run correlational analysis that included countertransference measures. Rho is a non-parametric statistic; therefore, can be used when there is violation to normality, as it does not make assumptions about the distribution of the population (Pallant, 2010). The only exception was in relation to ICB positive countertransference, as the logarithmic transformation resulted in a skewness value below 0.5, thus Pearson?s correlation was used when analyzing this countertransference variable. In addition, in the sample, there were four supervisors who rated two different supervisees each. Analyses were run to determine whether there was a supervisor effect in the data. The model failed to converge, and it appears that no effect is due to supervisors. Mixed models were run to examine the possibility that there exists an effect on each variable. Such an effect could violate the assumption of independence of observations. In all cases variance attributable to supervisor was 0 or close to 0, suggesting that little variance was due to supervisor effect. 71 Analysis of the Hypotheses To test predictions, bivariate correlational analyses were conducted. Table 4 presents a summary of all correlations for supervisor and therapist. As was previously mentioned, correlations involving countertransference measures were computed using Spearman?s coefficient, therefore, results are given using Spearman?s Rho. The exception was the positive subscale of the Inventory of Countertransference Behavior. In addition, test of significance was one-tailed for the cases were a specific direction for the relationship was hypothesized (Field, 2005). In cases where no a priori relation was stipulated, correlations were run with two-tailed test of significance. Finally, in line with Ligi?ro and Gelso (2002), analyses included scales and subscales of the different measures. Also, due to a concern about Type II errors in this early-stage research, p values at the .10 level are reported, even though alpha was set at .05. Hypothesis 1: There will be a positive relation between degree of attachment security of the counselor and strength of real relationship, such that Hypothesis 1.a. The greater the attachment security of the trainee, the stronger the real relationship as rated by the therapist trainee Hypothesis 1.b. The greater the attachment security of the trainee, the stronger the real relationship as rated by the supervisor To test the first hypothesis, the relationship between degree of attachment security (as measured by the reverse of the sum of avoidance and anxiety subscales of the ECR, Moore and Gelso, 2011) and real relationship as rated by the therapist trainee (as measured by the RRI-Therapist) was investigated using the Pearson product-moment 72 correlation coefficient. Preliminary analyses were performed to ensure no violation of the assumptions of normality, linearity and homoscedasticity. There was a positive and significant correlation between the two variables, r(30) = .370, p = .018, with higher levels of security associated with stronger real relationships as perceived by the therapists. In addition, the non-hypothesized relationships between the subscales of the real relationship assessed by the therapists, attachment security, and the avoidance and anxiety subscales of the ECR were also examined (See Table 5). Results showed that for the realism and the genuineness subscales of the therapist rated real relationship inventory, there was a positive and significant correlation with attachment security (r(30) = .331, p = .032 for Realism, and r(30) = .377, p = .017 for Genuineness). Therapists ratings of the realism subscale were also significantly and negatively correlated with avoidant attachment (r(30) = -.356, p = .023). The genuineness subscale of the RRI-T also showed a negative significant correlation with the avoidant subscale of the ECR (r(30) = -.398, p = .012). Also in line with the first hypothesis, the relationship between degree of attachment security (as measured by the reverse of the sum of avoidance and anxiety subscales of the ECR) and real relationship as rated by the supervisor (as measured by the RRI-Supervisor) was explored. Contrary to expectations, there was no significant correlation between the two variables (r(30) = -.096, p = .30). Further exploration of the correlations between Realism and Genuineness subscales in the Supervisors? RRI, and attachment security, anxiety and avoidance did not show any significant relation. Security 73 was related to the overall real relationship measure and the two real relationship subscales in the opposite direction than it was expected. Hypothesis 2: The counselors? rating of the strength of the real relationship will be negatively correlated to the supervisors? rating of negative countertransference behavior, such that the stronger the real relationship, the fewer the countertransference behaviors. In accordance to the second hypothesis, the relations between real relationship as rated by the therapist trainee (as measured by the RRI-Therapist) and negative countertransference behaviors assessed by supervisors (as measured by the ICB- negative), were tested using the Spearman?s Rho correlation. Contrary to prediction, there was no significant correlation between these two variables. Further exploration between countertransference behaviors and real relationship rated by the therapist in training showed that there was a significant and negative relation between total countertransference and real relationship as rated by the trainee (rs(30) = -.373, p = .018), such that the more countertransference, the weaker the real relationship. Also, there was a negative significant correlation between real relationship as rated by the therapist trainee and hostile countertransference (rs(30) = -.366, p = .02), yet as previously stated, such results need to be interpreted with caution. In addition, the Realism subscale of the RRI- Therapist was negatively and significantly correlated to negative countertransference (rs(30) = -.326, p < .034). Thus, the stronger the realism element of the real relationship as perceived by therapists, the less the negative countertransference as rated by supervisors. 74 Table 4 Intercorrelations for Real Relationship Inventory (RRI) for therapists and supervisor, Experiences in Close Relationships Scale (ECR), Inventory of Countertransference Behavior (ICB), and Countertransference Behavior Measure (CBM) Variables 1 2 3 4 5 6 7 8 9 10 11 1. RRI - Therapist 2. RRI - Supervisor .225 3. ICB - Total -.373** -.134 4. ICB - Positive -.379** -.208 .854??? 5. ICB - Negative -.175 -.303?? .790??? .437** 6. CBM - Dominant -.136 -.149 .578??? .328* -.704*** 7. CBM - Hostile -.366** -.227 .594??? .239 .731*** .402?? 8. CBM - Distant .060 -.175 .443??? .142 .644*** .551??? .488??? 9. ECR - Security .370?? -.096 -.247? -.243? -.012 -.209 .164 -.124 10. ECR - Avoidant -.395?? .050 .243 .186 .042 .192 -.014 .145 -.885*** 11. ECR - Anxiety -.280? .119 .190 .247 -.012 .177 -.283 .073 -.913*** .619*** 12. Absolute Value .089 -.492** -.047 -.077 .053 -.146 -.035 .069 .192 -.299* -.063 RRIT-RRIS Note. RRI-Therapist = Real Relationship Inventory as rated by Therapists; RRI-Supervisor = Real Relationship Inventory as assessed by Supervisors; ICB-Total = Inventory of Countertransference Behavior; ICB-Positive= Positive Countertransference Subscale of the ICB; ICB-Negative= Negative Countertransference Subscale of the ICB; CBM-Dominant= Dominant Scale of the Countertransference Behavior Measure; CBM-Hostile= Hostile Scale of the Countertransference Behavior Measure; CBM-Distant= Distant Scale of the Countertransference Behavior Measure; ECR-Security= Attachment security assessed by the inverse of the avoidance plus anxiety scores; ECR-Avoidance=Avoidance scale of the Experiences in Close Relationships Scale; 75 ECR-Anxiety=Anxiety scale of the Experiences in Close Relationships Scale; Absolute Value RRIT-RRIS= Difference in absolute value of therapist and supervisor ratings of the Real Relationship. * p < 0.10, two-tailed. ** p < 0.05, two-tailed. *** p < 0.01, two-tailed. ? p < 0.10, one-tailed. ?? p < 0.05, one-tailed. ??? p < 0.01, one-tailed 76 Table 5 Intercorrelations for Real Relationship Inventory (RRI) Subscales (Genuineness and Realism) for therapists and supervisor, Experiences in Close Relationships Scale (ECR), Inventory of Countertransference Behavior (ICB), and Countertransference Behavior Measure (CBM) Variables RRI-Therapist RRI-Therapist RRI-Supervisor RRI-Supervisor Realism Genuineness Realism Genuineness 1. ICB - Total -.533??? -.223 -.038 -.193 2. ICB - Positive -.453??? -.280? -.136 -.248? 3. ICB - Negative -.326?? -.057 -.177 -.381?? 4. CBM - Dominant -.256? -.025 -.069 -.190 5. CBM - Hostile -.434??? -.316?? -.128 -.249? 6. CBM - Distant -.075 .112 -.191 -.151 7. ECR - Security .331?? .377?? -.094 -.089 8. ECR - Avoidant -.356** -.398** .085 .015 9. ECR - Anxiety -.247 -.288 .084 .137 Note. RRI-Therapist Realism = Realism subscale of the Real Relationship Inventory as rated by Therapists; RRI-Therapist Genuineness= Genuineness Subscale of the Real Relationship Inventory as rated by Therapists; RRI-Supervisor Realism = Realism Subscale of the Real Relationship Inventory as rated by Supervisors; RRI-Supervisor Genuineness = Genuineness Subscale of the Real Relationship Inventory as rated by Supervisors; ICB-Total = Inventory of Countertransference Behavior; ICB-Positive= Positive Countertransference Subscale of the ICB; ICB-Negative= Negative Countertransference Subscale of the ICB; CBM-Dominant= 77 Dominant Scale of the Countertransference Behavior Measure; CBM-Hostile= Hostile Scale of the Countertransference Behavior Measure; CBM-Distant= Distant Scale of the Countertransference Behavior Measure; ECR-Security= Attachment security assessed by the inverse of the avoidance plus anxiety scores; ECR-Avoidance=Avoidance scale of the Experiences in Close Relationships Scale; ECR-Anxiety=Anxiety scale of the Experiences in Close Relationships Scale; * p < 0.10, two-tailed. ** p < 0.05, two-tailed. *** p < 0.01, two-tailed. ? p < 0.10, one-tailed. ?? p < 0.05, one-tailed. ??? p < 0.01, one-tailed 78 Hypothesis 3: There will be a negative correlation between real relationship rated by supervisors and countertransference behaviors, such that Hypothesis 3.a. Supervisors? rating of the real relationship established by the trainees with their client will be negatively correlated to the supervisors? rating of therapists? negative countertransference behaviors. Hypothesis 3.b. Supervisors? rating of the real relationship established by the trainees with their client will be negatively correlated to the supervisors? rating of therapists? positive countertransference behaviors. The relationship between real relationship as rated by the supervisors and countertransference behaviors assessed by supervisors was tested. As predicted, there was a negative and significant relationship between negative countertransference and supervisors? ratings of the real relationship (rs (30) = -.303, p = .046), with higher levels of real relationship rated by supervisors associated with lower levels of negative countertransference. Contrary to prediction, there was no significant correlation between supervisor?s ratings of real relationship and positive countertransference behavior. Further analysis of the real relationship inventory rated by supervisor showed that the genuineness subscale was negatively and significantly correlated with negative countertransference (rs (30) = -.381, p = .016). There were no significant correlations between supervisors? realism subscale of the RRI and positive or negative countertransference. 79 Hypothesis 4: The higher the level of disagreement between the trainees? and supervisors? rating of real relationship, the greater the amount of countertransference behavior from the therapist. The relationship between the absolute difference between therapist and supervisors ratings of the real relationship and countertransference behaviors (as measured by the ICB and CBM) was explored, and contrary to prediction, there was no significant correlation between these variables. Hypothesis 5: The level of secure attachment will be negatively related to amount of countertransference as rated by supervisors, such that the higher the level of security, the fewer the countertransference behaviors of the trainee. Finally, the relationship between level of secure attachment of the trainee (as measured by the reverse of the sum of avoidance and anxiety subscales of the ECR), and trainees? countertransference behaviors (as measured by the ICB and CBM), was investigated using Spearman?s Rho correlation. There was no significant relationship between secure attachment and countertransference behaviors. In addition, further examination of the relationship between attachment anxiety and attachment avoidance (as measured by the Anxiety scale and the Avoidant scale of the ECR) and countertransference (as measured by the ICB and CBM) showed that these variables were not significantly related. 80 CHAPTER 6: Discussion As has been previously stated, the aim of this study was to examine the role of the therapist in the therapeutic relationship. Specifically, the purpose was to investigate the relations between three variables: therapist attachment, therapist countertransference behaviors assessed by supervisor, and the real relationship the therapist established with a client (rated by the therapist and by the supervisor). The current chapter presents a discussion of the relevant findings, the study?s limitations, and future research directions. Attachment and Real Relationship The first hypothesis addressed the relations between real relationship and therapist attachment. Specifically, it was expected that degree of attachment security in the therapist would be positively related to strength of the real relationship, as rated by the therapist and as rated by the supervisor. Results indicated that this first hypothesis was partly supported. Considering attachment security and therapists? ratings of real relationship, there was an expected positive and significant relationship between trainees? attachment security and their perception of real relationship, such that the greater the attachment security of the trainee, the stronger the real relationship with a client as perceived by the trainee. Further exploration of the therapists? attachment showed that the trainees? avoidant attachment was negatively and significantly related to real relationship as rated by therapists. There was no significant relationship between therapist anxious attachment and real relationship (however, it should be mentioned that although alpha was set at .05, there was a negative relationship between therapist anxious attachment and real relationship that attained significance at p < .10). Such results are in line with Fuertes et 81 al.?s (2007) findings involving the relation between therapist attachment anxiety, attachment avoidance and real relationship as rated by the therapist. Fuertes et al. (2007) found that the therapist?s avoidant attachment was negatively related to the strength of the real relationship as perceived by the therapist, and there was no significant relation between attachment anxiety and therapists? ratings of the real relationship. The attachment literature can be helpful in explaining the positive relation between trainees? attachment security and their perception of the strength of the real relationship with a patient. According to Mikulincer and Shaver (2007), a secure therapist can utilize several skills, ?? such as gradually transforming a professional acquaintanceship into an intimate therapeutic relationship? (p. 422). According to these authors, secure people are more cognitively open and are more positively oriented towards searching new information and maintaining compassion and empathy towards other people. All these characteristics might allow secure therapists to see the clients in ways that befit them and to be more genuine in the interaction with the client, thus, establishing a stronger real relationship with them (as defined by Gelso, 2011). Mikulincer and Shaver also illuminate the negative relation that was found between therapist avoidant attachment and strength of real relationship as rated by the trainee. These researchers offer that therapists with avoidant attachment ?may lack the skills needed to provide sensitive care and promote emotional bonds with clients? (Mikulincer and Shaver, p. 422). In addition, Mikulincer and Shaver state that people with avoidant attachment favor interpersonal distance and might generate emotional detachment from those with whom they interact. Considering such characteristics of attachment avoidance 82 then, it can be understood that the more avoidant attachment the weaker the real relationship established with the client. On the other hand, the expected relation between degree of secure attachment of therapists in training and strength of the real relationship as rated by supervisors was not supported. Moreover, although not significant, the direction of the correlation between these two constructs was in the opposite direction than expected. In addition, neither avoidant attachment nor anxious attachment was related to the supervisor?s perception of the real relationship of the trainee and his/her client. To the author?s knowledge, there is only one other study that considers cross-sourced ratings of therapists? attachment and real relationship. Fuertes et al. (2007) found that therapists? attachment avoidance was related to therapists? ratings of real relationship, but there was no relation between therapists? attachment and clients? ratings of real relationship. Therefore, it can be speculated that the source of the real relationship rating might have something to do with the relation between attachment and real relationship. In addition, this is the first study relating the real relationship from a supervisor?s perspective and therapist variables. Further research in the area is needed. Countertransference and Real Relationship Several hypotheses on the present study were directed at assessing the relation between real relationship and countertransference. Thus, it was expected that the strength of the real relationship as rated by the trainee would be negatively correlated to negative countertransference behavior, such that fewer negative countertransference behaviors would be associated with stronger real relationships. Contrary to expectations, there was no significant relation between negative countertransference and trainees? rating of the 83 real relationship. However, a further exploration of the relation between these two constructs reveals that there was a negative and significant relation between negative countertransference and the realism subscale of the real relationship-therapist. Therefore, from therapists? perspective negative countertransference is associated to the element of the real relationship that relates to ??perceiving the other in ways that befit the other? (Gelso, 2011, p. 13). It can be considered that in order to see the client in ways that fit the client, therapists need to be unaffected by personal distortions (low countertransference) or have some awareness and management of any potential projections onto the other (good countertransference management). In the cases of higher negative countertransference behaviors, the therapist?s reactions towards the client are stemming from the therapist?s own vulnerabilities (Gelso & Hayes, 2007). Therapists might not be aware of the unconscious motives that generate such distortions, but still might realize that they are not reacting to the client based on the client?s problems but rather they are projecting some of their own issues into the relationship with the client. Therefore, the more the therapist?s reactions are shaped by his/her conflicts (i.e., countertransference, Gelso & Hayes, 2007), the less the therapist is seeing the client as the client is. Then, it seems that both of these constructs (i.e., realism element of the real relationship and countertransference) might intersect in a common component of distortion/accuracy in the perception of the other. In addition, further analysis showed that trainees? ratings of the real relationship were significantly and negatively related to overall countertransference and positive countertransference. Again, both positive countertransference and overall countertransference reflect therapists? internal and external reaction that are not about the 84 client but rather related to the therapists? unresolved conflicts and vulnerabilities (Gelso & Hayes, 2007). Such reactions might interfere with the establishment of a relationship with a client in which the therapist perceives the client accurately and reacts to him/her in genuine ways (i.e., real relationship). Therefore, the more countertransference behaviors from the therapists in the work with a client the weaker the real relationship established with such client. It is relevant to remember that reliability of the positive countertransference subscale was low; therefore, findings related to it should be taken with caution. The third hypothesis stated that supervisors? rating of the real relationship established by the trainees with their clients would be negatively related to both, positive and negative countertransference. This hypothesis was also partially supported, as supervisors? rating of the real relationship was negatively and significantly related to negative countertransference, but not to positive countertransference. There might be different plausible explanations for the lack of significance between positive countertransference and supervisors? perception of real relationship. Supervisors could perceive the behaviors that make up the positive countertransference scale as being supportive of the client (e.g., by befriending the client), and not necessarily seeing its dependent and enmeshed quality (Friedman and Gelso, 2000). Thus, perhaps supervisors did not consider such behaviors as reflective of countertransference. On the other hand, negative countertransference might be seen as punitive (Ligi?ro and Gelso, 2002), and supervisors might be more attuned to it than to positive countertransference. In addition, further exploration of the perception of the real relationship by the supervisor showed that the significant relation holds for the genuineness element of the real relationship, and 85 not for the realism element (which is the inverse of what was found from the therapist?s perspective). Thus, the more negative countertransference behaviors from the trainees, the less genuineness as perceived by the supervisors. A potential explanation for such findings is that usually supervisors have more clinical experience than the therapists, and thus may be more knowledgeable than the therapists in detecting less genuine reactions of the therapist in session. Similarly, supervisors may be more likely to use the therapist?s non-genuine reactions as markers of negative countertransference behavior. For example, when observing the videotape/DVD of a session between a trainee and his/her client, a supervisor might identify that the trainee is avoiding certain content in session by not inquiring in depth about the client?s experience in session and/or by not sharing his/her own reactions to the client?s material (aspects that might reflect genuineness). Further analysis could show that the therapist might also present behaviors such as distancing from the client in session or questioning the client?s motives in an inappropriate way, which reflect negative countertransference. Finally, the lack of relation between realism subscale and countertransference might be due to the fact that, if therapists are seeing the client?s in ways that don?t befit the other (i.e., low realism), they might not share such internal experiences with the supervisors. If the supervisors don?t see specific behaviors that they can explain from a framework of seeing the other in ways that don?t befit him/her, then the relationship between these constructs might not be accounted for. To the author?s knowledge, this is the first study to relate countertransference and real relationship. Thus, there is no specific empirical work to which the findings between real relationship and countertransference can be related. Ligi?ro and Gelso (2002) found a negative relation between negative countertransference and working alliance, as rated 86 by therapists and their supervisors, such that the stronger the alliance the less the negative countertransference. These authors also found that positive countertransference was negatively related to the bond aspect of the working alliance for supervisors rating of working alliance (Ligi?ro & Gelso, 2002). Considering that the sample size is relatively small in the current sample, and smaller than in Ligi?ro and Gelso (n = 32 in present study and 50 in Ligi?ro and Gelso), it is difficult to determine whether the relationship between a supervisor?s rating of real relationship and positive countertransference does not exist or if it is a matter of not enough power to detect such relationship in the present study. The fourth hypothesis of this study stated that the greater the disagreement between supervisors? and therapist trainees? real relationship ratings, the more countertransference behaviors will be exhibited by trainees. Ligi?ro and Gelso?s (2002) results showed that the discrepancy between therapists? and their supervisors? ratings of the bond component of the working alliance was a predictor of countertransference. In the present study, such relationship did not hold for countertransference and real relationship. A plausible explanation to the lack of significant relationships might be that such relationship exists when considering the bond in the context of the therapeutic work (i.e., working alliance), but not when we enter the realm of a personal relationship (i.e., real relationship). It might be that when considering the bond within the therapeutic work, there are specific markers different people can detect, and therefore the lack of agreement among raters might be reflecting a relation of this bond with other variables. However, when entering the personal realm of a relationship, the nuances of the relationship are harder to detect and therefore, each rater has a different perspective on it. 87 Support to such idea might be seen in the fact that there was a strong correlation (r = .55, p < .001) between supervisors? and therapists? ratings of working alliance in Ligi?ro and Gelso?s (2002) sample, but no significant correlation between therapist and supervisors? ratings of real relationship in the present study (r = .225 p < .108). In addition, results on clients? and therapists? cross-ratings on the real relationship are inconclusive. Some previous research involving cross-rating sources on the real relationship (i.e., therapists? and clients? ratings), have also shown lack of significant relationship between these ratings. Fuertes et al. (2007) and Marmarosh et al. (2009) did not find a correlation between therapists? and clients? ratings of the real relationship. On the contrary, Ain (2011) found a significant correlation between clients? and therapists? ratings of the real relationship (r = .48, p < .01). In line with Ain?s findings, Lo Coco, Gullo, Prestano and Gelso (2011), found a significant correlation between real relationship ratings of therapists and clients (r = .36, p < .01). It is relevant to note that Lo Coco et al. (2011) did not find a correlation between working alliance ratings of clients and therapists. On the contrary, Marmarosh et al. (2009) found a significant correlation between therapist and clients? ratings of the working alliance (r = .33, p < .05). Further studies might be needed to figure out the reasons for such cross-rating discrepancies. For example, there might be some cultural differences in the samples, as Lo Coco et al.?s study was conducted in Italy, whereas the other studies were conducted in the US. Additionally, Ligi?ro and Gelso (2002) studied a specific segment of therapy (between three and nine sessions), whereas the upper limit of therapy was left open in the present study. Perhaps the discrepancy between therapists? and supervisors? ratings is related to countertransference in the beginnings of therapy, but when therapy advances, 88 such differences lose significance. It could also be considered that participants of the relationship might have a shared relational experience, which presents as a correlation between the therapists? and clients? ratings of the real relationship. Such experience might not be completely captured by an external observer. Again, further research might help in clarifying the relation between cross-ratings of the real relationship as perceived by therapists and supervisors. Finally, exploratory analysis showed that the difference between therapists? ratings and supervisors? rating of the real relationship, in terms of absolute value, was negatively related to the supervisors? rating of the real relationship. Therefore, the more similar the ratings of therapist and supervisor of the real relationship, the stronger the supervisor sees the relationship. A plausible reason for such finding could be that when there is a strong real relationship between a therapist and his/her client, the therapist might be more open to talk about such relationship, and explore how this relationship relates to the therapeutic work. Thus, in these cases, therapists and supervisors might have a similar experience and perception of such relationship between therapist and client. On the contrary, when therapists have a weaker real relationship with their client, therapists might defensively perceive the relationship as stronger than it really is. Therapists and supervisors might not be addressing this directly in supervision, but supervisors might be detecting that the relationship is weaker than what the therapists believe it is. Thus, it might be that when supervisors perceive the therapists as less genuine and real in their interaction with a client than what the therapists perceive themselves to be, they use it as a marker of a weaker relationship between therapist and client. 89 Countertransference and Attachment The fifth and final hypothesis of the current study stated that there would be a negative relation between secure attachment and amount of countertransference as rated by supervisors. Contrary to expectation, there was no significant relationship between attachment security and countertransference at p < .05 level. It is relevant to mention that there was a relation between total and positive countertransference with attachment security at the p < .10 level. A potential explanation for not finding the expected relationship might be the small sample size, and therefore, not enough power. It might be helpful to further explore whether such a relationship emerges as significant if the sample increases in size. Also, Ligi?ro and Gelso (2002) did not find a significant correlation between attachment style of therapist and countertransference. These authors offer as a plausible explanation the fact that therapist?s attachment is not activated during session (i.e., client?s are not seen as attachment figures), and therefore, attachment is not related to countertransference behaviors. It could also be argued that a therapist?s attachment might get activated in a session with a client (e.g., when the client discusses termination, or gets angry with the therapist during a session), and the therapist might have internal reactions related to it. However, the therapist does not use the client as an attachment figure. Therefore, the therapist does not direct attachment behaviors towards the client, behaviors that could be linked to countertransference by an external observer. Such explanations are in accordance with attachment theory, which states that an attachment bond is a connection of a person with someone that is seen as stronger and wiser (Cassidy, 2008), and therapists might use romantic partners as attachment figures, but not their clients (Ligi?ro and Gelso, 2002). 90 In a different vein, Mohr, Gelso and Hill (2005) found significant interaction effects when considering attachment patterns and countertransference behaviors. These authors state ??countertransference is most likely to occur when the client and counselor differ in their pattern of attachment insecurity? (p. 306). Such findings can also be understood in light of Mikulincer and Shaver?s (2007) model of attachment ?system activation in adulthood. According to these authors, a subjective appraisal of threat (which can be internal or external) might activate the attachment system. The preconscious-activation of the attachment system causes an ?automatic heightening of access to attachment-related thoughts and action tendencies? (p. 33) including proximity ?seeking behavior (if mental representations are not enough). Such mental representations/proximity seeking behavior either soothe the adult and therefore effectively deactivates the attachment system, or fail to give satisfying comfort (due to figure unavailability, lack of responsiveness, etc?) which triggers attachment insecurity, related to either deactivating or hyperactivating strategies to deal with the threat. Considering such a model, certain material presented by a client might be perceived as threatening by a therapist, depending on the therapist?s attachment history (and therefore, internal working models). The therapist?s previous relational experiences might be a template from which to face such threatening material, determining if he/she can comfort him/herself, or if he/she might manifest behavioral reactions that could be related to his/her own vulnerabilities (i.e., countertransference). Therefore, considerations of the unique combination of client and therapists might be central to detect the relation between therapist and countertransference behavior. This highlights the fact that when considering the relations between attachment and countertransference, the mere rating of 91 therapists? attachment does not represent the intricate experience of attachment as it plays out in therapy. Finally, it is relevant to note that there was no relationship between avoidant attachment and distant countertransference. Although this relationship was not hypothesized as part of the current study, one could have expected these two variables to be related; however, results showed otherwise. Limitations The current study has several limitations that need to be considered when reviewing the results. The first limitation can be related to the sample size; as the sample was small, it is hard to know whether a result is non-significant due to not having the power to detect it or if in fact there is no relationship between two constructs. In order to address this last point, the author is continuing to collect data. Another limitation is related to the way in which some constructs were measured. First, therapists? attachment style and the therapists? perception of the real relationship are both self-report measures. A potential problem that needs to be considered is the possibility of mono-method bias: ?If two constructs are measured in the same way (for instance, self-report), the correlation between variables may result from method variance rather than any true correlation between constructs? (Heppner et al., 2008, p. 99). In addition, though the ECR has been widely used and has demonstrated adequate validity and reliability (Mikulincer & Shaver, 2007), recent empirical work has brought to the forefront the issue that it might assess security only as absence of avoidance and anxiety (Mikulincer & Shaver, 2007). When thinking about therapists, it is highly likely that we are dealing with a population that tends to be mainly secure and also, ?people 92 tend to be relatively secure on average? (Fraley et al., 2011, p. 623). Considering the attachment spectrum from secure to insecure, a problem with most attachment measures that are self-report is that they discriminate among people with insecure attachments, but have difficulty distinguishing among securely attached people (Fraley et al., 2011). Thus, the need for measures to target secure attachment seems to be central for attachment research in relation to the therapist. Also, in this study the anxiety and the avoidant subscales of the ECR were highly correlated (r = .619, p < 0.01, two-tailed test). It has been theoretically proposed that attachment anxiety and avoidance are two separate constructs. Thus, it would not be expected for these subscales to correlate highly. Furthermore, according to Mikulincer and Shaver (2007), the correlation of the anxiety and avoidance scales of the ECR is ?often close to zero? (p. 91). In addition, these authors mention that they have observed that these scales ??seem to be more highly correlated when they are administered to members of long-term couples? (p. 91). Some empirical work, on the other hand, has shown that these two subscales tend to correlate highly. Fuertes et al. (2007), found a positive and significant correlation between the attachment avoidance and attachment anxiety scales of therapist?s ECR (r = .54, p < 0.001). Fraley et al. (2011) mention that meta-analysis has shown a correlation of around .20 (either in the ECR or in the ECR-R), and, in line with Mikulincer and Shaver (2007), they state that that there is a stronger correlation when the sample consist of ?people in committed relationships? (p. 624). Fraley et al. (2011) bring up a relevant point, when they mention that ?conceptually distinct things need not be statistically independent? (p. 624). Future studies are needed to illuminate whether the two scales of the ECR are truly separate constructs, how these 93 two scales relate, under what circumstances they tend to correlate highly (i.e., what variables influence their relation), and how each scale uniquely contributes to other variables. In addition, there is an increasing body of work that is bringing to the forefront the issue of attachment being relationship specific (which is in a different line of thought than Bowlby?s idea of internal working models as presented in the literature review). The ECR is oriented towards experience in romantic relationships. It could be argued that the attachment involved in the relationship with romantic partners (and its representations) might present different characteristics than the ones in connection to other attachment figures (e.g., parents). Recently, Fraley et al. (2011) developed the Relationship Structures questionnaire of the Experiences in Close Relationships-Revised, which measures attachment avoidance and anxiety in different types of relationships (e.g., partners, parents). This measure might be a useful assessment tool in future research involving therapists. Third, the supervisor?s perception of the real relationship was assessed with a measure that was a modification of the therapist?s real relationship inventory, and was created for the present study. To assess criterion validity of the measure, supervisors? and therapists? ratings of the real relationship would be correlated, and a significant and positive correlation between these two measures was indeed expected. Results showed, however, that there was no expected correlation between supervisors? and therapists? perception of the real relation between the therapists and their clients. Such a result raises questions about the explanation of this non-relation. Specifically, is it that these two concepts are not related (thus, a theoretical issue), or is it that the supervisor?s version of 94 the Real Relationship Inventory is not assessing what it purports to assess (a validity issue)? In addition, the issue of cross-source ratings was previously addressed, and this issue can possibly be playing a role in the lack of a significant relation. Moreover the lack of significant relations between therapists? and supervisors? ratings of the real relationship is a topic that warrants additional research. In relation to countertransference measurement, there are several aspects relevant to consider. First, both Ligi?ro & Gelso (2002) and Mohr et al. (2005) changed the anchors of the lower end if the ICB-CBM measures in order to increase variability of the responses (i.e., 1= to no extent, versus the original anchor that states 1= to little or no extent). For the present study we maintained the original anchors of the ICB. Maintaining the anchors might have restricted the representation of countertransference behaviors by lumping the lower-end responses in one group, without differentiating among no countertransference and a little bit. Such differentiation can be very relevant, as the present sample of supervisors reported only a small amount of countertransference (see table of means). In addition, Mohr et al. (2005) raised the issue that the CBM, one of the two measures of CT, might be related to counselor?s competence and experience, more than unconscious issues as expected in a CT measure. In response to the difficulties of capturing countertransference behaviors, Friedman and Gelso (2000) stated, ?even behaviors motivated by inexperience may be viewed as fundamentally countertransferential? (p. 1231), as the lack of experience could trigger ??unresolved feelings of inadequacy or a desire to please? (p. 1231). As Mohr et al. (2005) mention, further research is needed to clarify the extent to which unconscious dynamics are represented in the assessed behaviors. 95 Also, as it was mentioned in the results, data from the countertransference measures presented skewness and kurtosis, violating normality assumptions. Data skewness from countertransference measures has also been found in previous studies: Mohr et al., (2005), Ligi?ro and Gelso (2002), and Friedman and Gelso (2000). In all cases, data was subjected to logarithmic transformations. In the present study, in spite of such transformations, the data still presented positive skewness, and therefore, data analysis for the countertransference measures was performed with non-parametric statistics. As was previously stated, data collection still continues for the current study. Thus, it will be important to consider if logarithmic transformations result in less skewness (reaching values below 1) with a larger size sample, which would allow analysis using parametric statistics. In addition, some limitations might be related to the nature of the study design. In order to complete the measures, therapists and supervisors had to identify a client who needed to have had at least three sessions with the therapist. Theoretically it makes sense that if there are high amounts of countertransference behavior from trainees, or a very weak real relationship in the first session(s), client might not continue treatment. Thus, part of the countertransference spectrum may have been excluded. Also, in the current study therapists decided whether to participate or not, and therefore therapists who perceived too much countertransference with their clients and or weak real relationships might have decided not to participate. Finally, as previously stated, two of the subscales used in the current study presented low reliability (i.e., CBM-Hostile and ICB-Positive); therefore, conclusions based on results attained with these two measures must be drawn with caution. 96 Future Directions The current study brings to the forefront the need for further research involving the constructs of interests (i.e., real relationship, countertransference and attachment). Considering the limitations previously presented, a first area of future research could be to continue assessing whether the Real Relationship Inventory ? Supervisor form captures the real relationship between a therapist and his/her client. The supervisor?s perspective on the real relationship between a therapist and his/her client could introduce the possibility of assessing whether the real relationship is a phenomenon that can be detected by external observers, and if such perception matches the perception of those directly involved in the relationship (i.e., therapist and client). In the present study, supervisors completed measures based on discussing the case in depth with the supervisee and/or listening to an audio of the session and/or watching a video of the sessions of the therapist with the client. Supervisors did not observe the session in-vivo, as has been the case in other research situations (i.e., Mohr et al., 2005). Also, in many cases supervisors might have relied only in the therapist?s account of the session. Future research could require the supervisor to do in-vivo observation or watch the DVDs (as some supervisor in the present study did) of the therapy session. Such direct observation could introduce a different perspective on countertransference and real relationship, and assess specific nuances that might not be detected when other means of informing about the session are used. The study here presented has addressed the relations between therapist attachment, countertransference and real relationship, thus focusing on the therapist? role in the relationship. Yet several questions can be raised on how the other participant in the 97 relationship (i.e., the client) might influence the work. Also as previously mentioned, Mohr et al. (2005) showed that there are important interactions in the relation between clients? and therapists? attachment, on the one hand, and countertransference on the other. It might be relevant to consider such interaction effects in relation to real relationship. In addition, it might be significant to assess how the supervisor?s attachment is influencing the relationship. Therefore, future research could assess three levels of influence in the therapeutic relationship: client, therapist and supervisor in relation to the variables of interest (i.e., real relationship, attachment and countertransference). Such assessment would allow the study of moderation, helping to address the ?who, what, when, and where? questions of what works in therapy (Gelso & Palma, 2011). For example, moderation could permit studying whether therapists who have a weaker real relationship display less countertransference behavior with their clients when the supervisor is secure than when the supervisor is anxious. It could also facilitate studying parallel processes in the therapeutic and supervisory relationship: e.g., cases of stronger real relationship of trainee with supervisor associated to stronger real relationship between client and therapist. All these questions might be highly relevant for psychotherapist?s training. In addition, Gelso and Palma (2011) highlight the usefulness of studying mediation in relation to self-disclosure and immediacy for the advancement of the self-disclosure and immediacy field. Future research could also consider mediation among the presented variables. For example, is countertransference a mediator between therapist?s attachment style and strength of real relationship? Or is therapist?s attachment a mediator between countertransference behaviors and real relationship? 98 Further studies are needed to clarify the nature of the therapists? attachment style in therapy. It has been stated that therapists need to act as a secure base for their clients (Bowlby, 1988). Yet what happens with the therapist?s attachment style when meeting their clients? One could expect that the therapists? attachment system be not activated when seeing clients, as hopefully clients are not seen as attachment figures. But in the face of tension and conflict in session, therapists? attachment system may well be activated, and it might be relevant to study how such a system manifests itself in therapy, if it varies according to attachment style, and how therapists? manage it in ways such that allow productive work. Ligi?ro (2000) introduced an important area for research when she proposed that ?is important to examine if a counselor?s attachment style is active during counseling, and when and how this activation occurs? (p. 103). Such questions were raised more than 10 years ago, and are still not sufficiently addressed, in spite of presenting central points for the therapeutic work. In addition, future research could involve the investigation of the same variables as in the present study, but employ repeated measurements over time. Repeated measurement would allow for an understanding of whether these relations evolve across therapy. For example, it is theorized that the real relationship strengthens over time (Gelso and Hayes, 1998). Repeated measures might help assess if such proposition is true for different attachment styles. Repeated measures may also allow detecting patterns in the relation between countertransference and real relationship. For example, perhaps it is true that the real relationship strengthens over time for cases in which therapists display little countertransference. However, when strong countertransference is present, the strength of the relationship might weaken over time. Also, repeated measures over time 99 might allow detecting crucial times in therapy were strong countertransference behaviors might impact the strength of the real relationship in a more negative way (e.g., beginning of therapy). Additionally, a combination of measures over time and three sources of ratings (i.e., client, therapist, supervisor) would allow a richer perspective on how the three variables addressed in the current study unfold in treatment. In addition, these variables could be related to other variables (e.g., working alliance, outcome measures), to increase knowledge on how does therapy work. Finally, data analysis in the current study consisted of bivariate r?s. Although simple correlations can be relevant to determine whether two constructs are related, they miss multivariate relationships that could be accounting and/or influencing the phenomena under observation. In the current investigation the focus was on the relationships between attachment and countertransference, attachment and real relationship, and countertransference and real relationship. The advancement of the study of these variables might benefit from addressing multivariate relationships, such as mediated relationships (e.g., attachment on countertransference and countertransference on real relationship), the influence of two variables on a third one (e.g., attachment and countertransference effects in real relationship), moderation (e.g., the relationship between attachment and real relationship might be moderated by countertransference), among others. In spite of focusing on the three previously mentioned variables, one cannot forget that this is just one aspect of the psychotherapy process, and the chosen variables are going to be influenced by other variables. For example, one could wonder about the role that a client diagnosis has in the strength of the real relationship and 100 therapist?s countertransference reactions. Further work could focus on discovering the multivariate relationships that might be related to the chosen variables. In spite of the limitations previously presented, the current study may help advance knowledge of the therapist contributions to the therapeutic relationship. Specifically, it has been indicated that the person of the therapist is entwined with outcome of psychotherapy (Norcross, 2002). Considering that the therapeutic relationship ??accounts for as much of the outcome variance as particular treatments? (Norcross, 2002, p. 5), efforts directed at illuminating what of the person of the therapist affects such relationship can be highly valuable. In addition, detecting how the person of the therapist affects the therapeutic relationship can be also central for the understanding of the therapeutic relationship. This study is the first to address the relations of countertransference and real relationship. In addition, the study of the relation between therapist attachment and real relationship is rather scant, and there is inconclusive evidence about the relation between therapist attachment and countertransference. Thus, the current study may take a useful step toward disentangling the therapist?s contribution to the therapeutic relationship. 101 Appendix A Table 1 Ethnicity of therapists and supervisors ________________________________________________________________________ Therapists Supervisors Number Percentage Number Percentage ________________________________________________________________________ African American 2 6.3% 2 7.1% Asian American 6 18.8% 0 European American 17 53.1% 20 71.4% Hispanic/Latino 6 18.8% 3 10.7% Other 4 12.5% 3 10.7% ________________________________________________________________________ Note. As with race, the percentage calculated for therapists was based on N=32, and for supervisors on N=28, but participants could mark more than one option. 102 Appendix A Table 2 Comparison table on the ways the supervisor knew about the trainee?s work, as reported by therapist and by supervisor ________________________________________________________________________ Therapists Supervisors Number Percentage Number Percentage ________________________________________________________________________ Discussed the case in depth 31 96.9% 32 100% Supervisor listened to audio 10 31.3% 9 28.1% Supervisor watched DVDs 17 53.1% 17 53.1% _______________________________________________________________________ Note. Number =the amount of therapists or supervisors that selected that option. Therapists and supervisors could endorse more than one option. Percentages for therapists and supervisors were calculated based on N=32. Again, supervisors could mark more than one option. 103 Appendix B Consent Form for Therapists-in-Training Project Title: Therapists characteristics that influence the therapeutic relationship. This project is conducted by Dr. Charles Gelso and Beatriz Palma, M.Ed., at the University of Maryland, College Park. We are inviting you to participate in this research because you are a therapist in training, who is currently seeing clients, and whose work is being supervised. The purpose of this research project is to learn more about the psychotherapy relationship, and about characteristics and behaviors of the therapist that can influence such a relationship. This knowledge could be helpful in understanding and improving the practice of psychotherapy. The procedure involves filling out three questionnaires. The first one asks for some background information. The other two questionnaires address your view about close relationships, and your perception of the relationship that you have with a specific client. Completing all the questionnaires will take approximately 10 minutes total. In order to complete the questionnaires, you and your supervisor must have identified a client based on this study?s criteria (i.e., the first current client that you will see after you discussed this study, attending at least three sessions, the case has been discussed in supervision, and if you use audio or videotapes of the sessions, that that your supervisor has watched the videotapes or listened to the audiotapes). You will fill out the questionnaire pertaining to the relationship with that client considering the most recent sessions (2 to 4 last sessions) combined. Also, please complete the measures only after you and your supervisor have met for at least five times. In addition, your supervisor will also be filing out measures. We will do our best to keep your personal information and responses strictly confidential. You are assigned a code identification number to answer your questionnaires. Please use 104 that code at the beginning of the questionnaires. Once you complete all the measures, your signature of the informed consent will be separated from the rest of your responses, to ensure that your responses and data are not directly matched with your name. The electronic data file will contain no identifiable information. In addition, any data from this study will be kept securely stored in a locked office in a locked suite. Electronic data will be kept securely in a protected file that is stored within a user login to which only the researcher will have access. Only investigators of the project will have access to the data. In addition, your supervisor will never know about your responses. Finally, any analysis or report of the data will use a combination of the data that is collected; no individual responses will be reported. There are no known risks to participating in this study. In addition, this study is not designed to help either you or your supervisor directly (i.e., no individual, specific feedback will be given), but the results might help the investigators learn more about the therapeutic relationship and therapist characteristics and behaviors that can affect the strength and quality of such relationship. In addition, your participation is voluntary. You may choose not to take part at all. If you decide to participate in this research, you may stop participating at any time. If you decide not to participate in this study or if you stop participating at any time, you will not be penalized or lose benefits to which you otherwise qualify. This research is being conducted by Dr. Charles Gelso and Ms. Beatriz Palma at the University of Maryland, College Park. If you have any questions about the research study itself, please contact Dr. Gelso at gelso@umd.edu, or Ms. Palma at bpalma@umd.edu. If you have questions about your rights as a research participant or wish to report a research-related injury, please contact: Institutional Review Board Office, University of Maryland College Park, 0101 Lee Building, College Park, Maryland, 20742. Their email is irb@umd.edu, and the telephone number is 301-405-0678. 105 This research has been reviewed according to the University of Maryland, College Park IRB procedures for research involving human subjects. 1. Do you agree to the above consent form? __ Yes __ No 2. Statement of Age of Subject and Consent Your signature indicates that you are at least 18 years of age; you have read this consent form, your questions have been answered to your satisfaction and you freely and voluntarily agree to participate in this research study. Signature : ________________________ Print name here : _________________________ Date : _________________________ 106 Appendix C Consent Form for Supervisors Project Title: Therapists characteristics that influence the therapeutic relationship. This project is conducted by Dr. Charles Gelso and Beatriz Palma, M.Ed., at the University of Maryland, College Park. We are inviting you to participate in this research because your supervisee has nominated you as a supervisor for the above mentioned study. The purpose of this research project is to learn more about the psychotherapy relationship, and about characteristics and behaviors of the therapist that can influence such relationship. This knowledge could be helpful understanding and improving the practice of psychotherapy. The procedure involves filling out three questionnaires. The first one asks for some background information. The other two questionnaires address your supervisee?s behaviors in sessions with a specific client, and your perception of the relationship between your supervisee and that client. Completing all the questionnaires will take approximately 10-15 minutes. In order to complete the questionnaires, you and your supervisee must have selected a client based on this study?s criteria (i.e., the first current client that your supervisee will see after you discussed this study, attending at least three sessions, the case had been discussed in supervision, and if you use audio or videotapes of the sessions, that that you watched the videotapes or listened to the audiotapes). You will fill out the questionnaires having in mind your supervisee?s work with that client for the most recent sessions (2 to 4 last sessions) combined. Also, please complete the measures only after you and your supervisee have met for at least five times. In addition, your supervisee will also be filing out measures about his or her sessions with that client. 107 We will do our best to keep your personal information and responses strictly confidential. You are assigned a code identification number to answer your questionnaires. Please use that code at the beginning of the questionnaires. Once you complete all the measures, your signature of the informed consent will be separated from the rest of your responses, to ensure that your responses and data are not directly matched with your name. The electronic data file will contain no identifiable information. In addition, any data from this study will be kept securely stored in a locked office in a locked suite. Electronic data will be kept securely in a protected file that is stored within a user login to which only the researcher will have access. Only investigators of the project will have access to the data. In addition, your supervisee will never know about your responses. Finally, any analysis or report of the data will use a combination of the data that is collected; no individual responses will be reported. There are no known risks to participating in this study. In addition, this study is not designed to help neither you nor your supervisee directly (i.e., no individual, specific feedback will be given), but the results might help the investigators learn more about the therapeutic relationship and therapist characteristics and behaviors that can affect the strength and quality of such relationship. In addition, your participation is voluntary. You may choose not to take part at all. If you decide to participate in this research, you may stop participating at any time. If you decide not to participate in this study or if you stop participating at any time, you will not be penalized or lose benefits to which you otherwise qualify. This research is being conducted by Dr. Charles Gelso and Ms. Beatriz Palma at the University of Maryland, College Park. If you have any questions about the research study itself, please contact Dr. Gelso at gelso@umd.edu, or Ms. Palma at bpalma@umd.edu. If you have questions about your rights as a research participant or wish to report a research-related injury, please contact: 108 Institutional Review Board Office, University of Maryland College Park, 0101 Lee Building, College Park, Maryland, 20742. Their email is irb@umd.edu, and the telephone number is 301-405-0678. This research has been reviewed according to the University of Maryland, College Park IRB procedures for research involving human subjects. 1. Do you agree to the above consent form? __ Yes __ No 2. Statement of Age of Subject and Consent Your signature indicates that you are at least 18 years of age; you have read this consent form, your questions have been answered to your satisfaction and you freely and voluntarily agree to participate in this research study. Signature : ________________________ Print name here : _________________________ Date : _________________________ 109 Appendix D Demographic Questionnaire for Therapist Trainees 1. Sex 2. Age 3. Race ___ Asian ___ Black ___ Caucasian ___ Pacific Islander ___ Other (Specify) 4. Ethnicity ___ African American ___ Asian American ___ Hispanic/Latino ___ Other (Specify) 2. Most Advanced degree ___ BA/BS ___ MA/MS ___ M.Ed. ___ MSW ___ Other (specify) 3. Your Theoretical Approach For each of the following theoretical approaches, write the number that states how representative of your work they are: Strongly Moderately Neutral Just a Little Not at all Representative 5 4 3 2 1 ___ Humanistic/Experiential ___ Psychodynamic/Psychoanalytic 110 ___ Cognitive/Behavioral ___ Systems ___ Other 4. If you answered ?other? in relation to theoretical approach, please specify 5. Years providing therapy (your best estimate) 6. Approximate number of sessions with the client (your best estimate) 7. Approximate number of sessions with supervisor about this client (your best estimate) 8. How does your supervisor know about your clinical work with this client (check all that apply) ___ We have discussed this case in depth ___ My supervisor listens to the audiotapes of the sessions ___ My supervisor watches the DVDs of the sessions 111 Appendix E Demographic Questionnaire for Supervisors 1. Sex 2. Age 3. Race ___ Asian ___ Black ___ Caucasian ___ Pacific Islander ___ Other (Specify) 4. Ethnicity ___ African American ___ Asian American ___ Hispanic/Latino ___ Other (Specify) 2. Most Advanced degree ___ MA/MS ___ M.Ed. ___ MSW ___ Ph.D ___ Other (specify) 3. Your Theoretical Approach For each of the following theoretical approaches, write the number which states how representative of your work they are: Strongly Moderately Neutral Just a Little Not at all Representative 5 4 3 2 1 ___ Humanistic/Experiential 112 ___ Psychodynamic/Psychoanalytic ___ Cognitive/Behavioral ___ Systems ___ Other 4. If you answered ?other? in relation to theoretical approach, please specify 5. Approximate number of sessions of supervisee with the client (your best estimate) 6. Approximate number of sessions of supervision about this client (your best estimate) 7. Time working with this supervisee 8. How do you know about your supervisee?s clinical work with this client (check all that apply) ___ We have discussed this case in depth ___ My supervisor listens to the audiotapes of the sessions ___ My supervisor watches the DVDs of the sessions 9. Years of clinical experience (to your best estimate, write how many years you have been providing therapy) 10. Years of clinical supervision (to your best estimate, write how many years you have been providing supervision) 113 Appendix F Experiences in Close Relationships (ECR) Scale The following statement concern how you feel in romantic relationships. We are interested on how do you generally experience relationships, not just in what is happening in a current relationship. Respond to each statement by indicating how much you agree or disagree with it using the following seven-point scale. 1 ----------- 2 ----------- 3 ------------ 4 ---------- 5 ----------- 6 ------------ 7 Disagree Neutral/ Agree Strongly Mixed Strongly 1. I prefer not to show a partner how I feel deep down. 2. I worry about being abandoned. 3. I am very comfortable being close to romantic partners. 4. I worry a lot about my relationships. 5. Just when my partner starts to get close to me I find myself pulling away. 6. I worry that romantic partners won?t care about me as much as I care about them. 7. I get uncomfortable when a romantic partner wants to be very close. 8. I worry a fair amount about losing my partner. 9. I don't feel comfortable opening up to romantic partners. 10. I often wish that my partner's feelings for me were as strong as my feelings for him/her. 11. I want to get close to my partner, but I keep pulling back. 12. I often want to merge completely with romantic partners, and this sometimes scares them away. 13. I am nervous when partners get too close to me. 14. I worry about being alone. 15. I feel comfortable sharing my private thoughts and feelings with my partner. 16. My desire to be very close sometimes scares people away. 17. I try to avoid getting too close to my partner. 18. I need a lot of reassurance that I am loved by my partner. 19. I find it relatively easy to get close to my partner. 20. Sometimes I feel that I force my partners to show more feeling, more commitment. 21. I find it difficult to allow myself to depend on romantic partners. 22. I do not often worry about being abandoned. 23. I prefer not to be too close to romantic partners. 24. If I can't get my partner to show interest in me, I get upset or angry. 25. I tell my partner just about everything. 114 26. I find that my partner(s) don't want to get as close as I would like. 27. I usually discuss my problems and concerns with my partner. 28. When I'm not involved in a relationship, I feel somewhat anxious and insecure. 29. I feel comfortable depending on romantic partners. 30. I get frustrated when my partner is not around as much as I would like. 31. I don't mind asking romantic partners for comfort, advice, or help. 32. I get frustrated if romantic partners are not available when I need them. 33. It helps to turn to my romantic partner in times of need. 34. When romantic partners disapprove of me, I feel really bad about myself. 35. I turn to my partner for many things, including comfort and reassurance. 36. I resent it when my partner spends time away from me. 115 Appendix G Inventory of Countertransference Behavior and Countertransference Behavior Measure Please complete this form on the counselor-trainee you have are rating. On the following scale, please rate the counselor?s reaction considering the last session with a client that your supervisee had. 1 2 3 4 5 to little or to a moderate to a great no extent extent extent The counselor: _____ 1. Colluded with the client in the session. _____ 2. Rejected the client in the session. _____ 3. Oversupported the client in the session _____ 4. Befriended the client in the session. _____ 5. Was apathetic toward the client in the session. _____ 6. Behaved as if she or he were ?somewhere else? during the session. _____ 7. Talked too much in the session. _____ 8. Frequently changed the topic during the session. _____ 9. Was critical of the client during the session. _____ 10. Spent time complaining during the session. _____ 11. Treated the client in a punitive manner in the session. _____ 12. Inappropriately apologized to the client during the session. _____ 13. Acted in a submissive way with the client during the session _____ 14. Acted in a dependent manner during the session. _____ 15. Seemed to agree too often with the client during the session. _____ 16. Inappropriately took on an advising tone with the client during the session. _____ 17. Distanced him/herself from the client in the session. _____ 18. Engaged in too much self-disclosure during the session. _____ 19. Behaved as if she or he were absent during the session. _____ 20. Inappropriately questioned the client?s motives during the session. _____ 21. Provided too much structure in the session. _____ 22. Dominated the session. _____ 23. Was hostile towards the client in the session. _____ 24. Acted parental during the session. _____ 25. Directed the client inappropriately in the session. 116 Appendix H The Real Relationship Inventory?Therapist Form Please complete the items below in terms of your relationship with your client or patient. Use the following 1?5 scale in rating each item, placing your rating in the space adjacent to the item. Strongly Agree Neutral Disagree Strongly Agree Disagree 5 4 3 2 1 ____ 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 feel 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 genuinely expresses a connection to me. 117 Appendix I The Real Relationship Inventory?Supervisor Form Please complete the items below in terms of your the relationship that you have seen your supervisee established with his/her client or patient. Use the following 1?5 scale in rating each item, placing your rating in the space adjacent to the item. Strongly Agree Neutral Disagree Strongly Agree Disagree 5 4 3 2 1 ____ 1. The client is able to see my supervisee as a real person separate from his/her role as a therapist. ____ 2. The client and my supervisee are able to be genuine in their relationship. ____ 3. The client feels liking for the ?real me? of my supervisee ____ 4. The client genuinely expresses his/her positive feelings toward my supervisee ____ 5. My supervisee is able to realistically respond to his/her client. ____ 6. My supervisee holds back significant parts of him/herself. ____ 7. I feel there is a ?real? relationship between my supervisee and his/her client aside from their professional relationship. ____ 8. The client and my supervisee are honest in their relationship. ____ 9. The client has little caring for who the supervisee ?truly is.? ____ 10. The client and my supervisee feel a deep and genuine caring for one another. ____ 11. The client holds back significant parts on him/herself. ____ 12. The client has respect for the supervisee as a person. ____ 13. There is no genuinely positive connection between the client and my supervisee. ____ 14. The client?s feelings toward my supervisee seem to fit who my supervisee is as a person. ____ 15. My supervisee does not like his/her client as a person. ____ 16. My supervisee values the honesty of the relationship with his/her client. ____ 17. The relationship between the client and my supervisee is strengthened by their understanding of one another. ____ 18. It is difficult for my supervisee to express what he/she truly feels about his/her client. ____ 19. The client has unrealistic perceptions of my supervisee. ____ 20. The client and my supervisee have difficulties accepting each other as they really are. ____ 21. The client distorts the therapy relationship. ____ 22. My supervisee has difficulty being honest with the client. ____ 23. The client shares with my supervisee the most vulnerable parts of him/herself. ____ 24. The client genuinely expresses a connection to my supervisee. 118 Appendix J First Email for Therapists Subject: Study about therapist characteristics and the therapeutic relationship Dear ___________, My name is Beatriz Palma, and I am a current student in the Counseling Psychology PhD program at the University of Maryland, College Park, working under the supervision of my advisor, Dr. Charles Gelso. I am writing to you because of your involvement in psychotherapy. If you are not currently seeing clients or patients, please respond to this email to let me know, and I will not contact you further. If you do currently see clients or patients for individual psychotherapy, please read on. Dr. Gelso and I would like to invite you to participate in a study that will examine therapist characteristics that influence the therapeutic relationship. We are asking therapists-in-training who are currently undergoing supervision to fill out some brief measures about themselves and their relationship with a particular client. In addition, the trainee?s supervisor will also fill out brief measures about personal reactions the trainee has had to the client in question. The supervisor will also rate the quality of the therapeutic relationship that the trainee has with that same client. We would very much appreciate your participation in this study. This research would involve approximately 10 minutes of your time and 10 minutes from your supervisor to complete some measures online. We are aware that your time is extremely important, but believe that the nature of this research will make your participation worthwhile. All participants will receive a summary of our findings and be notified of any publications that result from this study. 119 If you are agreeable to participating in the study please do contact me by phone or email to let me know. In addition, if you agree, I will ask you to briefly discuss this study with your supervisor, and ask him/her if he/she is willing to participate. This study has received IRB approval from The University of Maryland. If you have any questions regarding this study, please contact me at bpalma@umd.edu or (240) 393- 6973. Thank you. Sincerely, Beatriz Palma, M.Ed. Doctoral student Phone: 240-393-6973 Email: bpalma@umd.edu Charles Gelso, Ph.D. Professor of Psychology 120 Appendix K Second Email for Therapists Subject: Study about therapist characteristics and the therapeutic relationship Dear ________, Thank you so much for your interest in our study. Your participation will be extremely helpful. This is a two-step process, and this is how it will work: First, you and your supervisor need to identify a client for both of you to have in mind to complete the measures. The client needs to meet the following criteria: Be the first client that you are scheduled to see after you discuss this study with your supervisor; has attended at least three sessions; the client has been discussed in supervision, and if you use audio or videotapes of the sessions, that your supervisor has watched the videotapes or listened to the audiotapes. Next time you meet with your supervisor be sure to review the criteria with him/her, and to identify the client you both will consider. We can?t stress enough the importance of both of you having the same client in mind to complete the measures. Also, you can tell your supervisor that his/her participation involves completing three brief questionnaires online: a demographic questionnaire, a brief measure of your personal reactions in session and the therapeutic relationship you have established with the client in question. The questionnaires take approximately 10 minutes total. You will also complete three measures (this will be addressed more in detail in step two of this process). In addition, it is central that you and your supervisor meet at least five times in supervision in order to complete the measures. After meeting with your supervisor and identifying the client, please send me an email with your supervisor?s name, phone number, email and preferred way of contacting 121 him or her. I will address with your supervisor any details needed to complete their part of the study. Second, you can go online and complete the measures. The questionnaires take approximately 10 minutes total. We are aware that your time is extremely important, but believe that the nature of this research will make your participation worthwhile. You will complete a demographic questionnaire, a questionnaire on close relationships, and a questionnaire on your perception of your relationship with the specified client. In addition, confidentiality is central for us. Therefore, we are assigning the following code _____. This number will be the only way to identify your responses. In order to ensure confidentiality, we will keep all the data with identifying information separate from the file with your responses. In addition, although your supervisor knows that you are completing some measures, he or she will not have access to your responses. If you agree to participate in the study, please go to the attached link to complete the measures: _____________ It is relevant to let you know that all participants will receive a summary of our findings and be notified of any publications that result from this study. This study has received IRB approval from The University of Maryland. If you have any questions regarding this study, please contact me at bpalma@umd.edu or (240) 393-6973. Thank you. Sincerely, Beatriz Palma, M.Ed. Doctoral student Phone: (240) 393-6973 122 Email: bpalma@umd.edu Charles Gelso, Ph.D. Professor of Psychology 123 Appendix L First Email for Supervisors Subject: Study about therapist characteristics and the therapeutic relationship Dear __________, My name is Beatriz Palma, and I am a current student in the Counseling Psychology Ph.D program at the University of Maryland, College Park. I am writing to you because your supervisee _________ agreed to participate in the study ?Therapists characteristics that influence the therapeutic relationship,? that I am conducting for my master?s thesis under the supervision of my advisor, Dr. Charles Gelso. ________ has referred me to you as his/her supervisor to participate in this study. As you might know, it is most helpful to consider the perspectives of both a therapist in training and his/her supervisor about the therapeutic relationship. So, we would greatly appreciate your participation in this project. For this, we request that you to complete some questionnaires online. The questionnaires take approximately 10 minutes. We are aware that your time is extremely important, but believe that the nature of this research will make your participation worthwhile. In addition, all participants will receive a summary of our findings and be notified of any publications that result from this study. After reviewing the informed consent form, please complete three brief questionnaires. The first one asks for some background information. The other two questionnaires address your supervisee?s behaviors in sessions with a specific client, and your perception of the relationship between your supervisee and that client. In order to complete the questionnaires, you and your supervisee should identify a client who meets the following criteria: the first current client that your supervisee will see after you discussed this study; has attended at least three sessions; the client has been discussed in 124 supervision, and if you use audio or videotapes of the sessions, that you have watched the videotapes or listened to the audiotapes. For each questionnaire, you will find specific information on the variables of interest and instructions on how to complete the measures. After you complete the informed consent, you will need to add the following code: ______. This number will be the only way to identify your responses. In order to ensure confidentiality, we will keep all the data with identifying information separate from the file with your responses. In addition, although your supervisee knows that you are completing some measures, he or she will not have access to your responses. If you are agreeable to participating in the study, please go to the attached link to complete the measures: _________________ This study has received IRB approval from The University of Maryland. If you have any questions regarding this study, please contact me at bpalma@umd.edu or (240) 393-6973. Thank you. Sincerely, Beatriz Palma, M. Ed. Doctoral student Phone: 240-393-6973. Email: bpalma@umd.edu Charles Gelso, Ph.D. Professor of Psychology 125 References Ackerman, S., Benjamin, L. S., Beutler, L. E., Gelso, C. J., Goldfried, M. R., Hill, C., Lambert, M. J., Norcross, J. C., Orlinsky, D. E. & Rainer, J. (2001). Empirically supported therapy relationships: Conclusions and recommendations of the division 29 task force. Psychotherapy: Theory, Research, Practice, Training, 38, 495-497. doi: 10.1037/0033-3204.38.4.495 Ain, S. (2008). Chipping away at the blank screen: Therapist self-disclosure and the real relationship (Master thesis). Retrieved from http://hdl.handle.net/1903/8722 Ain, S. (2011). The Real Relationship, Therapist Self-Disclosure, and Treatment Progress: A study of Psychotherapy Dyads. (Unpublished doctoral dissertation). University of Maryland, College Park. Ainsworth, M. D. S. (1989). Attachments beyond infancy. American Psychologist, 44, 709-716. doi: 10.1037/0003-066X.44.4.709 Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61, 226-244. Black, S., Hardy, G., Turpin, G., & Perry G. (2005). Self-reported attachment styles and therapeutic orientation of therapists and their relationship with reported general alliance quality and problems in therapy. Psychology and Psychotherapy: Theory, Research, and Practice, 78, 373-377. doi: 10.1348/147608305X43784 Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research & Practice, 16, 252-260. doi:10.1037/h0085885 126 Bordin, E. S. (1994). Theory and research on the therapeutic working alliance: New directions. In A. O. Horvath & L. S. Greenberg (Eds.), The working alliance: Theory, research and practice. New York, NY: Wiley. Bowlby, J. (1979). The making and breaking of affectional bonds. London: Tavistock. Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd Ed.). New York, NY: Basic Books. (Original work published 1969). Bowlby, J. (1988). A secure base: Parent-child attachments and healthy human development. New York, NY: Basic Book. Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 46?76). New York, NY: Guilford. Brown, L. S. (2001). Feelings in context: Countertransference and the real world in feminist therapy. Journal of Clinical Psychology, 57, 1005-1012. doi: 10.1002/jclp.1068 Cassidy, J. (2008). The nature of the child?s ties. In J. Cassidy & P. R. Shaver (Eds.), Handbook of Attachment: Theory, Research and Clinical Applications (pp.762-782). New York, NY: The Guilford Press. Cassidy, J. (2010). Advanced Developmental Course, Class Notes, University of Maryland, Psychology Department. Castonguay, L. G., Constantino, M. J., & Holforth, M. G. (2006). The working alliance: Where are we and where we should go. Psychotherapy: Theory, Research, Practice and Training, 43, 271-279. doi: 10.1037/0033-3204.43.3.271 127 Crits-Christoph, P., Connolly Gibbons, M.B., & Hearon, B. (2006). Does the alliance cause good outcome? Recommendations for future research on the alliance. Psychotherapy: Theory, Research, Practice and Training, 43, 280-285. doi: 10.1037/0033-3204.43.3.280 Crowell, J., Fraley, R. C., & Shaver, P. R. (2008). Measurement of individual differences in adolescent and adult attachment. In J. Cassidy & P. R. Shaver (Eds.), Handbook of Attachment: Theory, Research and Clinical Applications (pp.599-634). New York, NY: The Guilford Press. Cutler, R. L. (1958). Countertransference effects in psychotherapy. Journal of Consulting Psychology, 22, 349-356. Daniel, S. I. F. (2006). Adult attachment patterns and individual psychotherapy: A review. Clinical Psychology Review, 26, 968-984. Diener, M. J. & Monroe, J. M. (2011). The relationship between adult attachment style and therapeutic alliance in individual psychotherapy: A meta-analytic review. Psychotherapy, 48(3), 237-248 Dozier, M., Cue, K. L., & Barnett, L. (1994). Clinicians as caregivers: Role of attachment organization in treatment. Journal of Consulting and Clinical Psychology, 62, 793- 800. doi: 10.1037/0022-006X.62.4.79 Ellis, A. (2001). Rational and irrational aspects of countertransference. Journal of Clinical Psychology, 57, 999-1004. doi: 10.1002/jclp.1073 Epstein, L. & Feiner, A. H. (Eds.). (1979). Countertransference. New York, NY: Aronson. 128 Epstein, L. & Feiner, A. H. (1988). Countertransference: The therapist contribution to treatment. In B. Wolstein (Ed.), Essential papers on countertransference (pp. 282- 303). New York, NY: New York University Press. Faul, F., Erdfelder, E., Lang, A.G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavioral Research Methods, 39 (2), 175-191. Fauth, J. (2006). Towards more (and better) countertransference research. Psychotherapy: Theory, Research, Practice, Training, 43, 16-31. doi: 10.1037/0033- 3204.43.1.16 Fauth, J. & Hayes, J. A. (2006). Counselor?s stress appraisals as predictors of countertransference behavior with male clients. Journal of Counseling and Development, 84, 430-439. Field, A. (2005). Discovering statistics using SPSS. London: Sage Fraley, R. C. (2010). Information on the Experience in Close Relationships-Revised (ECR-R) adult attachment questionnaire. Retrieved from http://internal.psychology.illinois.edu/~rcfraley/measures/ecrr.htm Fraley, R. C., & Shaver, P. R. (2000). Adult romantic attachment: Theoretical developments, emerging controversies, and unanswered questions. Review of General Psychology, 4, 132-154. doi: 10.1037/1089-2680.4.2.132 Fraley, R. C., & Waller, N. G. (1998). Adult attachment patterns: A test of the typological In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 77?114). New York, NY: Guilford. 129 Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response-theory analysis of self-report measures of adult attachment. Journal of Personality and Social Psychology, 78, 350- 365. Freud, S. (1959). The future prospects of psychoanalytic counseling. In J. Reviere (Ed.), Sigmund Freud: Collected papers II (pp. 285-296) (J. Reviere, Trans.) New York: Basic Books (Original work published 1910). Friedman, S. M., & Gelso, C. J. (2000). The development of the Inventory of Countertransference Behavior. Journal of Clinical Psychology, 56, 1221-1235. doi: 10.1002/1097-4679 Fuertes, J. N., Mislowack, A., Brown, S., Gur-Arie, S., Wilkinson, S., & Gelso, C. J. (2007). Correlates of the real relationship in psychotherapy: A study of dyads. Psychotherapy Research, 17, 423-430. doi: 10.1080/10503300600789189 Gabbard, (2001). A contemporary psychoanalytic model of countertransference. Journal of Clinical Psychology, 57, 983-991. doi: 10.1002/jclp.1065 Gelso, C. J. (2004). Countertransference and its management in brief dynamic therapy. In D. P. Charman (Ed.), Core processes in brief psychodynamic psychotherapy (pp. 231-250). Mawhaw, NJ: Laurence Erlbaum Associates. Gelso, C. J. (2009). The real relationship in a post modern world: Theoretical and empirical explorations. Psychotherapy Research, 19, 253-264. doi: 10.1080/10503300802389242 Gelso, C. J. (2011). The real relationship in psychotherapy: The hidden Foundation of change. Washington, DC: American Psychological Association. 130 Gelso, C.J., & Carter, J. A. (1985). The relationship in counseling and psychotherapy: Components, consequences and theoretical antecedents. The Counseling Psychologist, 13,155-243. doi: 10.1177/0011000085132001 Gelso, C. J., & Carter, J. A. (1994). Components of the psychotherapy relationship: Their interaction and unfolding during treatment. Journal of Counseling Psychology, 41, 296-306. doi: 10.1037/0022-0167.41.3.29 Gelso, C. J., Fassinger, R. E., Gomez, M. J., & Latts, M. G. (1995). Countertransference reactions to lesbian clients: The role of homophobia, counselor gender, and countertransference management. Journal of Counseling Psychology, 42, 356-364. doi: 10.1037/0022-0167.42.3.356 Gelso, C. J., & Hayes, J. A. (1998). The psychotherapy relationship: Theory, research and practice. New York: Wiley. Gelso, C. J. & Hayes, J. A. (2002). The management of countertransference. In J. C. Norcross (Ed.), Psychotherapy relationships that work (pp. 3-16). New York, NY: Oxford University Press. Gelso, C. J., & Hayes, J. A. (2007). Countertransference and the therapist?s inner experience: Perils and possibilities. Mahwah, NJ: Lawrence Erlbaum. Gelso, C. J., Hill, C . H., Mohr, J. J., Rochlen, A. B., & Zack, J. (1999). Describing the face of transference: Psychodynamic therapists? recollections about transference in cases of successful long-term therapy. Journal of Counseling Psychology, 46, 257- 267. doi: 10.1037/0022-0167.46.2.257 131 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, 640-649. doi: 10.1037/0022-0167.52.4.640 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). New York, NY: Wiley. Greenson, R. R. (1965). The working alliance and the transference neurosis. Psychoanalytic Quarterly, 34, 155-181. Greenson, R. R. (1967). The technique and practice of psychoanalysis (Vol. 1). New York, NY: International University Press. Greenson, R., & Wexler, M. (1969). The non-transference relationship in the psychoanalytic situation. International Journal of Psychoanalysis, 50, 27-39. Guy, J. D., & Brady, J. L. (2001). Identifying the faces in the mirror: Untangling transference and countertransference in self psychology. Journal of Clinical Psychology, 57, 993-997. doi: 10.1002/jclp.1066 Hayes, J. A. (1995). Countertransference in group psychotherapy. Waking a sleeping dog. International Journal of Group Psychotherapy. 45, 521-535. Hayes, J. A., & Gelso, C. J. (1991). Effects of therapist-trainees? anxiety and empathy on countertransference behavior. Journal of Clinical Psychology, 47, 284-290. doi: 10.1002/1097-4679 Hayes, J. A. & Gelso, C. J. (1993). Male counselor?s discomfort with gay and HIV- infected clients. Journal of Counseling Psychology, 40, 86-93. 132 Hayes, J. A. & Gelso, C. J. (2001). Clinical implications of research on countertransference: Science informing practice. Journal of Clinical Psychology, 57, 1041-1051. doi: 10.1002/jclp.1072 Hayes, J. A., Gelso, C. J., & Hummel, A. M. (2011). Managing countertransference. Psychotherapy, 48, 88-97. doi: 10.1037/a0022182 Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52, 511-524. doi: 10.1037/0022- 3514.52.3.511 Hazan, C., & Zeifman, P. (1999). Pair bonds as attachment: Evaluating the evidence. In J. Cassidy & P. R. Shaver (Eds.), Handbook of Attachment: Theory, Research And Clinical Applications. New York, NY: The Guilford Press. Heimann, P. (1950). Countertransference. British Journal of Medical Psychology, 33, 9- 15. Heppner, P. P., Wampold, B. E., & Kivlighan, D. M. (2008). Research design in counseling. Belmont, CA: Thomson Brooks/Cole. Horvath, A. O. & Bedi, R. P. (2002). The alliance. In Norcross, J. C. (Ed.), Psychotherapy relationships that work (pp. 37-70). New York, NY: Oxford University Press. Horvath, A. O., & Greenberg, L. S. (1994). The working alliance: Theory, research and practice. Oxford, England: John Wiley & Son. Horvath, A. O., & Luborsky, L. (1993). The role of the therapeutic alliance in psychotherapy. Journal of Consulting and Clinical Psychology, 61, 561-573. doi: 10.1037/0022-006X.61.4.561 133 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. doi: 10.1037/0022-0167.38.2.139 Hoyt, M. F. (2001). Connection: The double-edged gift of presence. Journal of Clinical Psychology, 57, doi: 10.1002/jclp.1069 Kaslow, F. W. (2001). Whither countertransference in couples and family therapy: A systemic perspective. Journal of Clinical Psychology, 57, 1029-1040. doi: 10.1002/jclp.1071 Kelley, F., Gelso, C. J., Fuertes, J., 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. doi: 10.1037/a0022082 Kiesler, D. J. (1996). Contemporary interpersonal theory and research: Personality, psychopathology, and psychotherapy. New York, NY: Wiley. Kiesler, D. J. (2001). Therapist countertransference: In search for common themes and empirical referents. Journal of Clinical Psychology/In Session, 57, 1053-1063. doi: 10.1002/jclp.1073 Kivlighan, D. M., Patton, M. J., & Foote, D. (1998). Moderating effects of client attachment on the counselor experience-working alliance relationship. Journal of Counseling Psychology, 45, 274-278. doi: 10.1037/0022-0167.45.3.274 Latts, M. G. & Gelso, C. J. (1995). Countertransference behavior and management with survivors of sexual assault. Psychotherapy, 32, 405-415. 134 Ligi?ro, D. P. (2000). The counselor?s part: Relationships between countertransference behavior, attachment style, and counselor-client working alliance. (Unpublished master thesis). University of Maryland, College Park. Ligi?ro, D. P., & Gelso, C. J. (2002). Countertransference, attachment and the working alliance: The therapist?s contributions. Psychotherapy: Theory, Research, Practice, Training, 39, 3-11. doi: 10.1037/0033-3204.39.1.3 Lo Coco, G. L., Gullo, S., Prestano, C., & Gelso, C. J. (2011). Relation of the real relationship and the working alliance to the outcome of brief psychotherapy. Psychotherapy, 48(4), 359-367. doi: 10.1037/a0022426 Mahrer, A. R. (2001). An experiential alternative to countertransference. Journal of Clinical Psychology, 57, 1021-1028. doi: 10.1002/jclp.1070 Main, M. M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A move to the level of representation. In I. Bretherton & E. Waters (Eds.), Growing points in attachment theory and research, Monographs of the Society for Research in Child Development, 50, (1-2, Serial No. 209), pages 74-79. Mallinckrodt, B. (2010). The psychotherapy relationship as attachment: Evidence and implications. Journal of Social and Personal Relationships, 27, 262-270. doi: 10.1177/0265407509360905 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, 337-350. doi: 10.1037/a0015169 135 Martin, A., Buchheim, A., Berger, U, & Strauss, B. (2007). The attachment organization on potential countertransference reactions. Psychotherapy Research, 17, 46-58. doi: 10.1080/10503300500485565 Martin, D. J., Garske, J. P., & Davis, M. K. (2000). Relation of the therapeutic alliance with outcome and other variables: A meta analytic review. Journal of Consulting and Clinical Psychology, 68, 438-450. doi: 10.1037/0022-006X.68.3.438 McClure, B. A., & Hodge, R. W. (1987). Measuring countertransference and attitude in therapeutic relationships. Psychotherapy, 24, 325-335. doi: 10.1037/h0085723 Mikulincer, M., & Shaver, P. R. (2007). Attachment in adulthood: Structure, dynamics, and change. New York, NY: The Guilford Press. Mohr, J. J. (2001). Client avoidance in first sessions of counseling: The moderating role of therapist attachment. (Unpublished doctoral dissertation). University of Maryland, College Park. Mohr, J. J, Gelso, C. J., & Hill, C. E. (2005). Client and counselor trainee attachment as predictors of session evaluation and countertransference behavior in first counseling session. Journal of Counseling Psychology, 52, 298-309. doi: 10.1037/0022- 0167.52.3.298 Moore, S. R., & Gelso, C. J. (2011). Recollections of a secure base in psychotherapy: Considerations of the real relationship. Psychotherapy, 48(4), 368-373. doi:10.1037/a0022421 Morgan, A. C., Brunshweiler-Stern, N., Harrison, A. M., Lyons-Ruth, K., Nahum, J.P., Sander, L., Stern, D., N., & Tronick, E. Z. (1998). Moving along to things left undone. Infant Mental Health Journal, 19, 324-332. 136 Myers M. D., & Hayes, J. A. (2006). Effects of therapist general self-disclosure and countertransference disclosure on ratings of the therapist and session. Psychotherapy: Theory, Research, Practice, Training, 43, 173-185. doi: 10.1037/0033-3204.43.2.173 Najavits, L. (2000). Researching therapist emotions and countertransference. Cognitive and Behavioral Practice, 7, 322-328. doi: 10.1016/S1077-7229(00)80090-5 Norcross, J. C. (2002). Empirically supported therapy relationships. In Norcross, J. C. (Ed.), Psychotherapy Relationships that Work (pp. 3-16). New York, NY: Oxford University Press. Norcross, J. C. & Lambert, M. J. (2011). Psychotherapy relations that work II. Psychotherapy, 48, 4-8. doi: 10.1037/a0022180 Ogden, T. H. (1990). The matrix of the mind. Lanham, MD: Aronson. (Original work published 1986). Pallant, J. (2010). SPSS survival manual: A step by step guide to data analysis using SPSS. Berkshire, England: Open University Press. Peabody, S. A. & Gelso, C. J. (1982). Countertransference and empathy: The complex relationship between two divergent concepts in counseling. Journal of Counseling Psychology, 29, 240-245. doi: 10.1037/0022-0167.29.3.240 Racker, H. (1957). The meaning and uses of countertransference. Psychoanalytic Quarterly, 26, 303-357. Romano, V, Fitzpatrick, M, & Janzen, J. (2008). The secure base hypothesis: Global attachment, attachment to counselor, and session exploration in psychotherapy. Journal of Counseling Psychology, 55, 495-504. doi: 10.1037/a0013721 137 Rossberg, J. L., Karterud, S., Pedersen, G., & Friis, S. (2008). Specific personality traits evoke different countertransference reactions: An empirical study. Journal of Nervous and Mental Disease, 196, 702-708. doi: 10.1097/NMD.0b013e318186de80 Safran, J. D. & Muran, J. C. (2006). Has the concept of the therapeutic alliance outlived its usefulness? Psychotherapy: Theory, Research, Practice, Training, 43, 286-291. doi: 10.1037/0033-3204.43.3.286 Sable, P. (2008). What is adult attachment? Clinical Social Work Journal, 36, 21-30. doi: 10.100/s10615-007-0110-8 Slade, A. (2008). The implications of attachment theory and research for adult psychopathology. In J. Cassidy & P. R. Shaver (Eds.), Handbook of Attachment: Theory, Research and Clinical Applications (pp.762-782). New York, NY: The Guilford Press. Strupp, H. H. & Binder, J. L. (1984). Psychotherapy in a new key: A guide to time-limited dynamic psychotherapy. New York, NY: Basic. Thompson, R. (2008). Early attachment and later development: Familiar questions, new answers. In J. Cassidy & P. R. Shaver (Eds.), The Handbook of Attachment: Theory, Research, and Clinical Applications (pp. 348-365). New York: Guilford. Wampold, B. E. (2001). The great psychotherapy debate: Models, methods and findings. Mahwah, NJ: Laurence Erlbaum Associates. Winnicott, D. W. (1949). Hate in the countertransference. International Journal of Psychoanalysis, 30, 69-75. 138 Woodhouse, S. S., Schlosser, L. Z., Crook, R. E, Ligi?ro, D. P., & Gelso, C. J. (2003). Client attachment to therapist: Relations to transference and client recollections of parental caregiving. Journal of Counseling Psychology, 50, 395-408. Yulis, S. & Kiesler, D. J. (1968). Countertransference response as a function of therapist anxiety and content of patient talk. Journal of Consulting and Clinical Psychology, 32, 413-419. doi: 10.1037/h0026107 Zeifman, D. & Hazan, C. (2008). Pair bonds as attachments: Reevaluating the evidence. In J. Cassidy & P. R. Shaver (Eds.), Handbook of Attachment: Theory, Research and Clinical Applications (pp.762-782). New York, NY: The Guilford Press 139