ABSTRACT Title of Dissertation: THERAPIST-CLIENT RACIAL MATCHING VS. NON-MATCHING AND THERAPISTS’ COUNTERTRANSFERENCE: EXPLORING THEIR RELATION AND TESTING MODERATORS. Beatriz Isabel Palma Orellana Doctor of Philosophy, 2016 Dissertation directed by: Professor Charles J. Gelso Department of Psychology The present study assessed the influence of clients’ race (i.e., Black/African American or White/European American), therapists’ universal-diverse orientation (UDO), and therapists’ anger discomfort on countertransference reactions. Countertransference was operationalized as therapists’ self-reported state anxiety, their verbal avoidant responses (as manifestation of behavioral countertransference), and their self-reported countertransference. Data were gathered from 63 White, European American therapists and therapists- in-training. Participants completed online measures pertaining to universal-diverse orientation, anger discomfort, trait anxiety, social desirability, and a demographic questionnaire. A week after completing such measures, the participants completed a Lab session. The therapists and therapists-in-training were randomly assigned to one of two conditions: An angry White/European American client or an angry Black/African American client. Participants watched and verbally responded to a video of the assigned scripted analogue client. Right afterward, the therapists and therapists-in-training completed a measure of state anxiety and three single items assessing the influence of the participant’s countertransference in his or her behaviors, thoughts, and feelings while responding to the videotaped client. Additionally, the participants’ verbal responses were transcribed verbatim and coded as approach or avoidant responses, which ultimately provided an index of behavioral countertransference. Results showed that therapists’ anger discomfort, their universal-diverse orientation, and clients’ race predicted state anxiety. No significant effects were found on the other countertransference measures. Additionally, only anger discomfort significantly and uniquely accounted for variance in state anxiety. Contrary to expectations, neither clients’ race nor universal-diverse orientation uniquely accounted for variance in the dependent variables. Results were not significant for the interaction of clients’ race and UDO on therapists’ countertransference reactions. Results were also non-significant for the interaction of clients’ race and anger discomfort on the participants’ countertransference reactions. Implications of the findings are further discussed. THERAPIST-CLIENT RACIAL MATCHING VS. NON-MATCHING AND THERAPISTS’ COUNTERTRANSFERENCE: EXPLORING THEIR RELATION AND TESTING MODERATORS by Beatriz Isabel Palma Orellana Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2016 Advisory Committee: Professor Charles J. Gelso, Chair Professor Clara E. Hill Professor Mary Ann Hoffman Professor Dennis M. Kivlighan, Jr. Professor Barry D. Smith © Copyright by Beatriz Isabel Palma Orellana 2016 ii Dedication To my parents, Beatriz Isabel Orellana Garafulic and Juan Carlos Palma Irarrázaval. Thank you for teaching me the importance of working hard, persevering, and being passionate about what you do. And most of all, thanks for your constant love and support. iii Acknowledgements First, I want to thank Charlie Gelso, my advisor and mentor throughout my Ph.D. program. Charlie is an outstanding scientist-practitioner, and working with him has allowed me to experience first hand that building a bridge between psychotherapy research and clinical practice is possible. I am really thankful for all his guidance and support during my grad school experience. I am also grateful for our intellectually stimulating conversations, that helped me not only generate this project but also bring it to fruition. Additionally, I really appreciate Charlie’s great sense of humor (and laughing with him), and his capacity to challenge me in a caring way. I am also grateful to my committee members: Clara E. Hill, Mary Ann Hoffman, Dennis M. Kivlighan Jr., and Barry D. Smith. Their inquiries about the subjects addressed and their feedback not only helped me improve this project, but also whet my appetite to continue my empirical exploration. Additionally, I want to express my gratitude to Ann Hummel, who generously provided me guidance when I needed it. I also want to thank Stephen Byrne, Kimberly Elliot, and James Riffle, my “undergraduate” research assistants (all of which graduated during this project). They were involved in almost every stage of the process, and demonstrated a strong work ethic through it. I am also grateful to my coders, raters, and everyone who helped me sharpen this work and make this project possible, including every therapist and therapist-in-training who participated in this study. I also want to thank my parents Beatriz Orellana and Juan Carlos Palma, and my siblings: Magdalena, Juan Cristóbal, and Catalina. Thank you for providing me support and care in many different ways, and for always being there for me. Also, thanks to all iv my friends, my sister in law, and Patricia De Cea, who provided laughter, cheered me up, and somehow knew how to be with me when I needed support. Last but not least, I want to thank Cristian Saez, my husband. I am deeply grateful for your care, support, good sense of humor, patience, and encouragement throughout this process. I believe it takes a village to complete a dissertation study… And I am really grateful to everyone who helped me make this project possible. v Table of Contents Dedication Acknowledgements Chapter 1: Introduction………………………………………………………...........1 Multiculturalism 1 Countertransference 3 Universal-Diverse Orientation 4 Experience with Anger 6 Chapter 2: Literature Review……………………………………………………….9 The Concept of Race 10 Research on Therapists and Clients’ Match on Race 12 Universal-Diverse Orientation 14 Measurement and Research Considering UDO 16 The Therapist’s Countertransference 18 Definition and Relevant Aspects Related to Countertransference 19 Countertransference Measurement and Research 21 Measurement 21 Research 21 The Experience of Anger 24 Research Related to Therapists Dealing with Clients’ Anger 26 The Use of Analogue Procedures in Research 29 Advantages 30 Disadvantages and Challenges 31 Chapter 3: Statement of the Problem and Hypotheses……………………………..33 Statement of the Problem 33 Hypotheses 36 Chapter 4: Method………………………………………………………………….41 Design 41 Independent Variable 41 Dependent Variable 42 Participants 42 Measures 45 Miville-Guzman Universality-Diversity Scale –Short From 45 Anger Discomfort Scale 46 The Balanced Inventory of Desirable Responding 47 State-Trait Anxiety Inventory 48 Countertransference Index 49 Avoidance Index 50 Demographic Questionnaire 54 Open Question 54 vi Stimulus Material 54 Procedure 64 Chapter 5: Results…………………………………………………………………..68 Descriptive Data 68 Preliminary Analyses 74 Test of Normality 75 Hypotheses Testing 76 Chapter 6: Discussion ……………………………………………………………...80 Limitations 85 Future Directions 86 Appendix A: Miville-Guzman Universality-Diversity Scale –Short ………………90 Appendix B: Anger Discomfort Scale……………………………………………...91 Appendix C: The Balanced Inventory of Desirable Responding…………………..92 Appendix D: State-Trait Anxiety Inventory………………………………………..94 Appendix E: Countertransference Index……………………………………………95 Appendix F: Response Mode Categories for Units and Whole Turns……………..96 Appendix G: Demographic Questionnaire …………………………………………104 Appendix H: Client Case Presentation and Script………………………………….106 Appendix I: Emails for Recruitment………………………………………………..108 Appendix J: Informed Consent……………………………………………………..111 Appendix K: Debriefing Form……………………………………………………...113 Appendix L: Tables………………………………………………………………...115 References………………………………………………………………………….120 vii List of Tables Table 1: Means and Standard Deviations of Actors’ Client Portrayal and Their Racial Features………………………………………………………………………………63 Table 2: Intercorrelations for Miville-Guzman Universality-Diversity Scale –Short (MGUDS), Anger Discomfort Scale (ADS), Balanced Inventory of Desirable Responding (BIDR6) - Self Deceptive Enhancement (SDE) and Impression Management (IM) scales, State-Trait Anxiety Inventory for Adults (STAI) - Trait and State scales, Countertransference Index Scale (CT- Scale) – assessing influence of countertransference in behavior (B), thoughts (T), and feelings (F), Avoidance Index, and Racial Non-Match………………………………………………………………69 Table 3: Means and standard deviations of the variables of interest per experimental condition…………………………………………………………………………….71 Table 4: Standard Multiple Regression Analysis of Clients’ Race, Universal-Diverse Orientation (UDO) and Anger Discomfort as Predictors of Behavioral Countertransference (i.e., Avoidance Index)………………………………………115 Table 5: Standard Multiple Regression Analysis of Clients’ Race, Universal-Diverse Orientation (UDO) and Anger Discomfort as Predictors of Self-Reported Countertransference (i.e., Countertransference Index Scale)……………………...115 Table 6: Hierarchical Multiple Regression Analysis of Clients’ Race, Universal- Diverse Orientation (UDO) and Anger Discomfort as Predictors of State Anxiety (Trait anxiety as covariate)………………………………………………………..116 Table 7: Hierarchical Multiple Regression Analysis of Clients’ Race, Universal- Diverse Orientation (UDO), Anger Discomfort, and Interaction Terms (Clients’ Race X UDO and Clients’ Race X Anger Discomfort) as Predictors of Behavioral Countertransference (i.e., Avoidance Index)………………………………………117 Table 8: Hierarchical Multiple Regression Analysis of Clients’ Race, Universal- Diverse Orientation (UDO), Anger Discomfort, and Interaction Terms (Clients’ Race X UDO and Clients’ Race X Anger Discomfort) as Predictors of Self-Reported Countertransference (i.e., Countertransference Index Scale)………………………118 Table 9: Hierarchical Multiple Regression Analysis of Clients’ Race, Universal- Diverse Orientation (UDO), Anger Discomfort, and Interaction Terms (Clients’ Race X UDO and Clients’ Race X Anger Discomfort) as Predictors of State Anxiety (Trait anxiety as covariate)……………………………………………………………….119 1 Chapter 1: Introduction Humans are diverse. From our genetic makeup to the experiences that shape us, the possibility of difference among us is vast. Inevitably, as individuals we try to make sense of such a world, thus creating categories to make the world around us more manageable. Yet as soon as we start dividing the world around us, our lenses turn to what is about us (in-group) versus what is about others (out-group). Numerous authors have discussed human differences. In addition, social psychology has extensively addressed the issue that the experience of in-group and out-group brings with it several attributions, attitudes, stereotypes, prejudices, etc…. which affect the way others relate to us and how we relate to others (e.g., Stangor, 2011). One of the aspects that can be central to understand the in-group and out-group experience is culture. Multiculturalism Culture is a large umbrella, which encompasses a wide spectrum of constructs ranging from more biologically based to more societally created. Our cultural background is mostly related to ethnicity, but also influenced by race, social class, gender, sexual orientation, among others (McGoldrick, Giordano & Garcia-Preto, 2005). As Munley, Lidderdale, Thiagarajan & Null (2004) stated, “individuals may be members of more than one cultural or identity group and some group identities may be more salient than others” (p.284). Thus, different aspects pertaining to culture will give rise to both unique experiences and shared experiences among individuals. 2 One of the constructs within the multicultural umbrella that has been a fertile area of research is race. Race can be thought of as constituted by physical markers, yet it is a socially developed notion. In addition, race carries meaning. As Hall (1997) eloquently shared, “The body is a text, and we are all readers of it. We are readers of race, we are readers of human difference. Race works like a language”. It is not about denying physical differences, but mostly is to look at the meaning that we give to such physical differences (Hiles 2007). The meaning we attribute to race (and ethnicity) might be related to the stereotypes, prejudices, attitudes, among others, that we have towards other groups and our own race. Such lenses will also color the way we react and relate to those who we perceive as different or similar to us. When considering the psychotherapy realm, the way we behave towards a client that belongs to a different racial/ethnic group than our own, might be related to the attitudes we have in connection to such group. Pérez Foster (1998) has highlighted that therapists “…carry their cultural view of being a human into all work with their clients…” (p. 258). Therefore, disentangling the aspects that we as therapists bring to the work with our clients might be relevant in the psychotherapy we conduct with those who are similar and those who are different from us. Taking into account that “psychotherapy is an affective experience” (Gelso and Hayes, 2007, p. 18), during treatment therapists and clients will have reactions towards each other. Such reactions might be based on their own experiences with race and what they attribute to the differences between each other that they encounter. As therapists, we shall inevitably get emotionally touched by our clients. Thus, following Hall (1997) a person’s race might evoke things in us that in turn might generate and influence our 3 reactions towards such a person. Sometimes, these reactions might get in the way of the therapeutic work. When that is the case, we might be dealing with countertransference. Countertransference Countertransference can be defined in many ways, yet for the present work, it refers to the therapist’s reactions to the client that are related to the therapist’s past or current unresolved issues (Gelso & Hayes, 2007). Empirical work has shown that countertransference might interfere with therapy process and outcome. For example, countertransference is negatively related to working alliance (Ligiéro & Gelso, 2002), and real relationship (Palma & Gelso, 2013), variables that have been connected to both the process of psychotherapy and treatment outcome. In addition, Hayes, Gelso & Hummel (2011) conducted meta-analyses considering countertransference research, which revealed that countertransference is negatively related to outcome, and that countertransference management predicts better treatment results. Some authors highlight the importance of considering countertransference reactions in cross-racial dyads (e.g., Gelso and Hayes, 2007; Gelso & Mohr, 2001; Jackson, 1973; Jones, 1985; Mirsky, 2011; Pérez Foster, 1998). A useful concept when looking at such cases is cultural countertransference, which has been defined in different ways. For example, Ridley (2005) states that such concept “...involves the therapist’s projecting race-related emotional reactions onto a client of another race” (p.70). Considering cultural countertransference, particularly between White therapists and African American clients, Jones (1985) mentioned that therapists can have unhelpful emotional responses to any client, yet it seems that “black patients may evoke more 4 complicated countertransference reactions and more frequently” (p. 178). As explanations, the author states that the “social images of blacks” (p. 178) allow more space for projections from the therapist, and that cultural difference might be related to more empathic failures. A small number of studies have looked at therapists’ countertransferential reactions to clients in cross-racial dyads. For example, both Brittan (1984) and Harbin (2004) used analogue procedures to assess European American therapists’ countertransference responses to an angry Black client (each author looked at particular variables which will be discussed later). However, these studies only investigated the reactions towards a Black client. Though such empirical efforts are highly valuable, in order to clarify the racial layers that therapist-client interactions might have, studies need to allow comparison of therapists’ reactions to a White client and to a Black client. Therefore, efforts to increase the understanding of the relations between racial match and non-match between therapist and client on the one hand, and countertransference on the other, might be valuable in connection to therapy research. In addition, it would appear to be useful to study the variables that might influence the relation between therapist-client cross-racial and same-race dyads and countertransference. One of such variables is universal-diverse orientation (UDO). Universal-Diverse Orientation Universal-diverse orientation “…reflects an attitude of awareness and acceptance of both, the similarities and differences among people” (Miville et al., 1999, p. 291). UDO encompasses affective, cognitive and behavioral aspects that are interconnected (Fuertes, Miville, Mohr, Sedlacek & Gretchen, 2000; Miville et al, 1999). Miville et al. 5 (1999) based this construct mainly in Vontress (1988, 1996), who highlighted that the awareness and acceptance of the similarities and differences others have in relation to us are important for effective interactions in general, and particularly central for therapists who work with people from different cultural backgrounds. Research has shown that UDO is positively related to empathy (Miville et al., 1999), openness to experience (Thompson, Brossart, Carlozzi, & Miville, 2002), and perspective taking (Miville, Carlozzi, Gushue, Schara & Ueda, 2006), among others. In addition, research has shown that UDO accounts for significant variance in variables such as multicultural counseling competence expectation, attitudes towards diversity, psychological hardiness, and psychosocial functioning (Ponterotto, 2008). To the author’s knowledge, Harbin (2004) conducted the only unpublished study that has looked at the therapist’s universal-diverse orientation and countertransference in a cross-racial dyad (European American therapist and African American client). He found that UDO was significantly and negatively related to countertransference reactions. As previously stated, in this analogue study the author only used cross-racial dyads (Black actor-client and White therapists); therefore, no possible comparisons where made in connection to the relationship of these variables when looking at cross-racial dyads versus same race dyads. Researchers have highlighted the importance of continuing studying UDO in order to advance the understanding of the construct (e.g., Miville, 2006; Ponterotto, 2008). Therefore, research on UDO and countertransference that allows comparison between therapist-client dyads considering racial match vs. cross-racial might be highly valuable to untangle UDO’s role in the reactions we have to those who are racially similar or different to us. In addition, work that increases the knowledge about 6 variables that are related to UDO and countertransference can further advance the empirical knowledge of these constructs. In both Brittan (1984) and Harbin’s (2004) studies, the client was an angry Back male. It has been stated that there are important communicational style differences between Black and White people, which can lead to misinterpretations in the interactions between these groups (Sue & Sue, 2008). Additionally, Sue and Sue posit, “White counselors often believe that Blacks are nonverbal, paranoid and angry, and that they are most likely to have character disorders… (p. 92). Therefore, it can be speculated that a White/European American therapist’s stereotypes and attitudes pertaining a Black male client might get triggered when the client gets angry in session, which could generate intense countertransference reactions. However, these responses might also be related to other variables, like the way a therapist relates to his/her own anger. One of such variables might be the therapist’s anger discomfort. Looking at this variable could be important, because the way a therapist relates to anger might play an important role in the work with clients who are racially different, as “…(a) cross racial interactions are frequently embedded with angry overtones, (b) anger-laden interactions are frequently avoided or ignored...” (Abernethy, 1995, p. 96). Experience with Anger Though anger is considered a basic and commonly experienced emotion (Kannan et al., 2011), research looking at anger, especially maladaptive anger, is less frequent than the one looking at other troublesome emotions (DiGiuseppe & Tafrate, 2001). Considering critical moments in therapy, dealing with client’s anger can be threatening for trainees (e.g., Bandura, Lipsher & Miller, 1960; Gamsky & Farwell, 1966; Sharkin 7 and Gelso, 1993). When looking at experience in relation to dealing with a client’s anger, results have been mixed: Experience has been related to better responses to an angry client (e.g., Russell & Snyder, 1963, Haccoun and Lavigueur, 1979), but one study showed that experienced therapists had more struggles than inexperienced therapists did in responding to an angry client that was directing the anger towards them (Varble, 1968). Two variables that could be related to the way trainees and more experienced therapists respond to an angry client are anger-proneness (the “…general tendency to experience angry feelings”, Sharkin & Gelso, 2001, p. 483) and anger discomfort (“…the degree to which people feel uncomfortable with their own anger”, Sharkin & Gelso, 2001, p. 483). In fact, research with counselor trainees has shown that both anger discomfort and anger-proneness are positively related to anger toward and discomfort with an angry client (Sharkin & Gelso, 2001). In addition, anger discomfort has been positively related to trait anxiety, anger suppression and anger expression (Sharkin & Gelso, 1991). Considering attachment, secure people score lower on anger proneness than insecure people (Mikulincer, 1998). Therefore, we could think that the way a therapist relates to his/her own anger will play a role in his or her reactions to an angry client. Furthermore, due to stereotypes in the US culture, when facing an angry client, issues related to race might become salient. Thus, the way a therapist relates to anger (e.g., anger discomfort) might play a role in the way he/she approaches an angry client from his/her own race versus from a different race, and the countertransferential reactions he/she has to them. Because of the nature of the therapeutic encounter, one might not expect too much overt expression of anger from the therapist; therefore, if a counselor is 8 anger prone, this might not necessarily get played in session. On the other hand, having anger discomfort might influence a therapist’s work, in ways such as not exploring issues of anger brought up by the client, avoiding the reflection of the client’s anger, and/or having difficulties addressing interactions in the here-and-now that have an angry undertone, among others. The current study was an effort to continue disentangling the therapist’s contribution to the therapeutic relationship. For this, the therapist’s reactions to those who are similar to and different from them, and some variable that might predict such reactions and moderate such relationship were examined. Specifically, this study looked at White/European American therapists’ reactions to an angry Black/African American client and to an angry White/European American client. Though the main focus of this work was the experience of cross-racial therapist-client dyads, conclusions in connection to race and countertransference can only be made if there is a group to which the cross racial dyads can be compared to. Such racial groups were chosen mainly due to the fact that relations between these two groups have been a central and highly complex aspect of the US history, and that there are strong attitudes, stereotypes, etc,.. that still influence the interactions between these two groups. In addition, the present study also looked at the relations of universal-diverse orientation, and therapist anger discomfort, to countertransference, and the role of these variables in the relation between therapist-client racial match or no-match, and countertransference reactions. 9 Chapter 2: Literature Review As was previously presented, the aim of this study was to look at therapists’ experience in cross-racial dyads, particularly White/European American therapist and Black/African American client. Specifically, there might be certain critical moments in therapy where a therapist’s stereotypes and attitudes towards in-group and out-group might get triggered, and potentially affect therapeutic work via countertransference reactions. One such moment can be when a client gets angry in session. The therapist’s reactions might be moderated by the way he/she relates to similarities and differences with others (e.g., universal-diverse orientation), and the way he/she relates to anger. The present literature review focuses on the different variables studied (i.e., client race, universal-diverse orientation, countertransference, and discomfort with anger). Due to the breadth and depth of the theoretical and empirical literature in relation to some of the variables of interest (e.g., race), only the work that is central to the current study and provides a framework for it is examined. The review first addresses race, specifically looking at it from the lense of intra- and inter-group attitudes, and presenting empirical findings on therapist-client cross-racial dyads. Then, the review focuses on universal- diverse orientation, defining the construct and reporting research on the topic that is relevant to the current work. Third, information on countertransference is presented, considering its definition, the different ways to measure the construct, and some key findings. Next, the concept of anger is addressed, with a focus on therapists’ response to an angry client. Because the present study was designed as a laboratory investigation, the 10 literature review ends with a brief discussion on analogue research, addressing its advantages and disadvantages. The Concept of Race Race can be defined as “a category of persons who are related by a common heredity or ancestry and who are perceived and responded to in terms of external features or traits”, (Wilkinson, 1993, p. 19). The central aspect of the idea of race might be that, as Yoo and Pituc (2013) highlight, it “…is a sociopolitical construction based on perceived physical differences (e.g., skin color, facial features, hair type) and is often conflated or interchanged with ethnicity (i.e., group membership based on shared values, traditions, behaviors, and language...” (p. 427). Even though empirical findings have extensively and repeatedly shown that there is more difference within-groups than between-groups in terms of biological and physical characteristics, we are inclined to categorize people based on their physical appearances and the group memberships we externally assign to them. Such racial categorization will tend to be pervasive in a person’s life, as “…race shapes group membership, meaning, experiences, and treatment of others” (Yoo & Pituc, 2013, p.427). Therefore, the race(s) people belong to (or are categorized in) will be directly related to the stereotypes, prejudices, attitudes, beliefs, etc... that they have in relation to their in-group and the ones they have about their outer-groups. Social psychology has extensively explored and explained intra and intergroup dynamics (Stangor, 2011). One of the constructs that social psychologists have studied and found central in connection to intra/intergroup experiences is attitude. Attitudes can be defined as “our positive or negative evaluations of an attitude target” (Stangor, 2011, p. 5.2). Attitudes have an affective, cognitive and 11 behavioral component, with the affective one usually being the strongest and most central. The strongest attitudes are the most important, as they easily come to mind. According to Stangor, our attitudes likely guide our behavior (principle of consistency), and meta-analyses have supported the idea that attitudes as reported in self-report measures predict behavior (Stangor, 2011). Thus, the attitudes we have towards various groups will influence the way we relate to others who are similar or different. For example, Jones (1997) addresses issues of prejudice and racism in the relations between White and Black people. According to this author, such relations are “…paradigmatic of intergroup conflict in the United States” (Jones, 1997, p. 16). This author discusses a study that looked at White people’s racial attitudes towards different groups (e.g., Hispanics, illegal immigrants, Whites). Results showed that the most negative attitudes where towards Blacks, and they were perceived as “…lazier, more welfare dependent, more prone to violence, and less intelligent than the other groups (Sears, Citrin and Van Laar, 1995, in Jones, 1997, p. 17). In addition, such attitudes towards Blacks as a group were more highly crystalized and organized than for other groups. When considering such ideas in the psychotherapy realm, one can wonder how these attitudes might influence the work of White/European American therapists with those who are different from them, such as Black/African American clients. As Fuertes and Gelso (2000) have stated, “in counseling, physical features, such as race, have been found to be salient ‘markers’ that are easily encoded and powerful in shaping initial impressions, and to affect the process and outcome in psychotherapy” (p. 212). Therefore, the attitudes we have towards people from our own racial/ethnic group versus those from other groups might have a central role in psychotherapy, as such attitudes will 12 affect the way we work with both clients who we perceive to belong to our racial groups and those who are not. Additionally, such attitudes might be more covert than overt. The interest in race in psychotherapy is not new. Research on race has increased in the past decades (Yoo & Pituc, 2013), and theoreticians and clinicians have been addressing the idea of race in therapy for quite some time (e.g., Sattler, 1970; Vontress, 1971; Jackson, 1973; Proctor & Rossen, 1981; Holmes, 1999). In the following subsection, there is a brief review of the empirical studies on client-therapist matching that are most relevant for the current work. Research on Therapists and Clients’ Match on Race Literature reviews have generally shown that even though people have a preference for therapists of their own race/ethnicity, client outcome does not improve by matching therapists and clients based on race/ethnicity (Coleman, Wampold & Casali, 1995; Maramba & Nagayama, 2002; Shin et al., 2005; Smith & Trimble, 2016). Cabral and Smith’s (2011) meta-analysis revealed that treatment outcomes of therapist-client dyads that are matched by race/ethnicity are not substantially different from those that are not matched. Yet when looking at the different groups, results showed that of all the racial/ethnic groups studied, “the relevance of ethnic matching was greatest amongst African American participants” (p. 543). These clients presented a strong preference for therapists of their own group, the perception of therapists differed as a function of racial/ethnic matching (i.e., evaluated African American therapists more positively than other therapists), and these clients had better outcomes when the therapist was from their own racial group. However, as the authors note, few of the studies used an experimental 13 design, and many investigations looking at preference were based on hypothetical scenarios instead of real therapy. Fewer studies have examined the therapist’s experience in cross-racial psychotherapy dyads. Some studies have looked at the therapist’s reactions to addressing ethnicity and race in cross-cultural therapy. Though a survey revealed that therapists reported being comfortable with and being skilled at discussing racial differences in therapy (Maxie, Arnold, & Stephenson, 2006), a qualitative study showed that when looking at African American and European American psychologists’ experience, the European American therapists reported feeling uncomfortable when addressing race (Knox, Burkard, Johnson, Suzuki & Ponterotto 2003). Other studies have examined therapists’ countertransference reactions when working in cross-racial dyads, and the relation of countertransference to other variables. (Theory and research on countertransference will be more extensively reviewed later.) For example, Brittan (1984) studied White therapists’ countertransference reactions to Black male clients. Participants watched a video analogue, where the client (an actor) portrayed one of two conditions: (1) Presence of racial verbal material – client angrily denigrates White colleagues, or (2) absence – denigrates colleagues without mention of race. In addition, this author wanted to see the moderating role of counselors’ racial identity level on countertransference reactions. Countertransference was assessed by self-reported state anxiety and therapists’ avoidant verbal responses (thus, two dependent variables). Brittan hypothesized that more countertransference in the therapist would be elicited by the Black client who angrily credited his problems to racism from White colleagues. This 14 hypothesis was not supported for either dependent variable, and in fact the mean for avoidance was higher in the neutral condition. Therefore, it seems that the therapist’s experience in cross- racial/ethnic dyads in therapy (particularly for a White/European American therapist with a Black/African American client) is complex, might be influenced by different variables, and may generate different reactions. Such differences seem relevant to explore. For example, clients and therapists might have stereotypes, preferences, biases, etc… that can potentially influence treatment especially when working with clients or therapists who are different from them. One of the variables that might enhance (or hinder) the interactions we have with those who are different from us is Universal-Diverse Orientation, or UDO. Universal-Diverse Orientation Universal-Diverse Orientation (UDO) has been defined as “an attitude toward all other persons that is inclusive yet differentiating in that similarities and differences are both recognized and accepted; the shared experience of being human results in a sense of connectedness with people and is associated with a plurality or diversity of interactions with others” (Miville et al., 1999, p. 292; Ponterotto, 2008). Like all attitudes (Stangor, 2011), UDO includes affective, cognitive and behavioral elements, which are interconnected (Miville et al., 1999). As Singley and Sedlacek (2009) stated, “rather than simply the presence or absence of prejudice, UDO is the movement toward or away from diversity” (p. 405). This construct was first introduced by Miville et al. (1999), and it was based mainly on Vontress’ (1979, 1988, 1996) work on likenesses and differences among 15 people (Harbin, 2004; Miville et al., 1999). Vontress highlighted that people are both similar to and different from each other, and that the awareness of such similarities and differences facilitates effective human relations. Miville et al. underscore that for Vontress, an individual is the result of the interaction of different cultures: Universal, ecological, national, regional, and racioethnic. Behind the universal culture is a “common biological makeup of human beings” (Miville et al, 1999, p. 291), and it encompasses what makes us human beyond particular groups. In regards to the idea of universal aspects that connect human beings, Miville et al. also considered Jung’s perspective of universal images (i.e., archetypes) and collective unconscious, and Yalom’s idea of universality in group counseling (i.e., members recognizing that at a certain level, they are similar). Even though Miville et al. did not include Harry Stack Sullivan’s work as part of the theoretical background, their views about universality remind us of Sullivan’s idea that we are more human than otherwise. The other cultures presented by Vontress (e.g., national, racioethnic) refer to “…group experiences at increasingly local—and thus more diverse—levels” (Miville et al. 1999, p. 292). Therefore, the construct of UDO relates to awareness and acceptance of the fact that commonalities and differences exist simultaneously among people. Miville et al. highlight that such consideration of people as both similar and different is central for therapists, as it would allow joining clients based on the commonalities of humanness, yet respecting and valuing the client’s qualities that are different from their own. It can also be speculated that, on the contrary, therapists with low UDO might be more prone to present negative reactions to clients who might be different from them. In the following 16 subsection, some research findings related to UDO that can illuminate UDO’s role in the therapy process are presented. Measurement and Research Considering UDO Empirical work related to UDO was possible thanks to the development of a scale that allowed its assessment: The Miville-Guzman Universality-Diversity Scale (M- GUDS; Miville et al., 1999). Though the authors hypothesized a three-factor structure, only a single factor emerged from their factor analysis. Thus, the 45-item measure yields a total score of universal-diverse orientation (details on this measure will be presented in the methods section). Further advancements on UDO’s measurement were possible due to Fuertes, Miville, Mohr, Sedlacek, and Gretchen (2000). These authors conducted a series of studies, where they looked at the factor structure of the M-GUDS, and developed a shorter version of the measure. Thus, Fuertes et al. came up with the M- GUDS-S, which yields a total score and a score for three subscales (which correspond with the originally hypothesized three factor structure): Diversity of Contact (a behavioral component of UDO), Relativistic Appreciation (cognitive component of UDO), and Comfort with Differences (more related to emotional aspects). According to Ponterotto (2008), empirical studies have shown that higher UDO levels relate to healthy narcissism, feminism, androgyny, positive racial identity (Miville et al., 1999), academic self-confidence (Fuertes, Sedlacek, Roger, & Mohr, 2000), empathy (Miville et al., 1999) and help-seeking behaviors (Fuertes, et al., 2000), among others. Therapists’ UDO is also positively associated with gay/bisexual clients’ ratings of the working alliance, session depth, and session smoothness (Stracuzzi, Mohr & Fuertes, 2011). On the other hand, UDO is negatively related to prejudice, as measured by 17 homophobia and dogmatism (Miville et al., 1999). In addition, UDO has been shown to share variance with self-identity in multicultural counseling knowledge and awareness (Munley, et al., 2004). Also, several studies have shown a gender difference in UDO ratings, with women scoring higher on average than men (Fuertes & Gelso, 2000; Fuertes, Miville, Mohr, Sedlacek & Gretchen, 2000; Singley & Sedlacek, 2009). Due to the focus of interest of the present work, studies related to race/ethnicity are further addressed. Some results related to UDO and race/ethnicity have been inconclusive. Fuertes et al. (2000) did not find significant differences in UDO ratings by European American and other racial/ethnic groups. On the contrary, other studies have shown significant differences in UDO pertaining to race/ethnicity. For example, in a sample of 2,228 College Freshmen (measures completed during orientation), Singley and Sedlacek (2009) found that Anglo-American students’ overall ratings of UDO were significantly lower than those from people who identified as Asian/Asian American, African American, and Latino. These authors did not find differences among the UDO scores of members from differing minority groups. It seems relevant to highlight that the participants in some of the previously presented studies were college students and not therapy clients or therapists. To the author’s knowledge, Harbin (2004) conducted the first study that looked at “…how therapist UDO may directly relate to counseling process in a cross racial situation” (p. 68). Specifically, this author was interested in the role that information on the client’s strengths and therapists’ UDO had in the countertransference that European American therapists had to an angry African American client (who was videotaped). Harbin found a significant and negative relation between UDO and countertransference. 18 To further explore the role of UDO in the dependent variables (i.e., affective, cognitive and behavioral countertransference), the author divided UDO in low and high, and found that the higher the participants’ UDO scores were, the lower their anxiety when reacting to an angry Black male client. Taking into account the different studies previously reviewed, it seems that UDO might have a relevant role in treatment, and further exploration of therapists’ UDO in cross racial dyads could be valuable. In addition, it seems that looking at UDO only as a main effect might overlook important differences in connection to the construct. An area that seems relevant and fertile to continue exploring is if there is an interaction between therapist-client racial/ethnic match/non-match and UDO on the therapists’ countertransference reactions. Thus, countertransference will be reviewed in the following section. The Therapist’s Countertransference For some time it has been theorized that countertransference has a negative impact on treatment (e.g., Freud 1910/1959; Gelso and Carter, 1994: Gelso and Hayes, 2007). This assertion was recently supported by a meta-analysis that showed that there is a negative and modest relation between countertransference and treatment outcome (Hayes, Gelso & Hummel, 2011). Therefore, efforts directed at uncovering what influences a therapist’s countertransference reactions and what might help a therapist better manage such reactions might be valuable for psychotherapy work. In the present section, the definition of countertransference used in this study and some aspects of the construct that are pertinent to the proposed work are reviewed. Then issues related to the 19 assessment of countertransference, and some central empirical findings pertaining to this variable are addressed. Definition and Relevant Aspects Related to Countertransference Over the years, countertransference has been conceptualized in different ways. The earliest perspective on countertransference can be traced back to Freud (1910/1959), who considered it to be a conflict-based response to the client’s transference, mainly unconscious, and thus, a problem that analysts should get rid of. For the current study, countertransference will be considered from what is termed the integrative view. From this perspective, countertransference is defined as “the therapist’s internal or external reactions that are shaped by the therapist’s past or present emotional conflicts and vulnerabilities” (Gelso and Hayes, 2007, p. 25). This definition captures a key issue pertaining to countertransference: Not all the responses that the therapist has in psychotherapy are countertransferential (Beitman, 1983; Gelso & Hayes, 2007). As Beitman (1983) has highlighted, “Among the non-countertransference therapist responses which are to be distinguished from countertransference are empathy, intuition, clinical and theoretical knowledge, technique and therapeutic caring” (p. 84). In addition, at least to some degree, countertransference is always a joint creation of therapist and patient (Gelso and Hayes, 2007). Consequently, countertransference can inform the therapist not only about his/her own internal world, but also about the client’s dynamics. Also, one can experience countertransference at an affective and cognitive level (i.e., internal reactions), and/or act out such internal responses (i.e., countertransference behaviors). As previously stated, internal reactions can be useful (i.e., they can inform us about the patient). The problem is when countertransference is acted out because “The therapist is 20 no longer attending to the client’s needs, but instead, is playing out his/her own conflicts with the client” (Latts and Gelso, 1995, p. 405). Additionally, the present study focused on cross-racial client-therapist dyads, specifically between a White/European American therapist and a Black/African American client. Thus, it is pertinent to consider the role of culture in countertransference. As Gelso and Hayes (2007) state, “although countertransference is largely based on intrapsychic and interpersonal phenomena, cultural factors are often centrally involved as origins and triggers” (p.134). Furthermore, some authors have even developed a framework for looking at countertransference from a cultural lense. For example, Gelso and Mohr (2001) mentioned that the general definition of countertransference is not enough when cultural aspects are part of the countertransferential experience. These authors introduced two concepts that are central to consider in such cases. The first was cultural countertransference, which they defined as the “therapist's culture-related distortions of the patient or rigid interpersonal behaviors rooted in his or her direct or vicarious experiences with members of the patient's RSM [racial/ethnic or sexual orientation minority status] group.” (p. 58). The second concept they presented was culturally reinforced countertransference, where the only difference with cultural countertransference is that in this case, the root of the transference is early childhood experiences. Several studies have looked at the therapist’s countertransferential responses, however, not much has been written in relation to the influence of cultural factors (such as race) on countertransference and vice versa (Gelso & Hayes, 2007). The following subsection reviews some empirical findings related to countertransference. 21 Countertransference Measurement and Research As can be inferred from the previous examination, the construct of countertransference is complex, and thus is difficult to measure (Hofsess and Tracey, 2010; Fauth, 2006; Gelso & Hayes, 1998). For example, a therapist could self-report some aspects related to internal reactions, yet not be aware that he/she is acting them out. An external observer (e.g. supervisor) could report countertransference behaviors, without recognizing the different layers (e.g., affective and cognitive reactions) that the phenomena might have for the therapist. Measurement. To measure countertransference, empirical studies have focused on its internal or external manifestation (See Hayes, Gelso & Hummel, 2011 for a more detailed review). To assess countertransference at an affective level, researchers have used quantitative measures that evaluate emotional states, as well as qualitative methods. To measure countertransference at a cognitive level, researchers have followed a two-step process. First they have established the number of words of a certain kind that appear in a session (e.g., related to sexual content), and then they have looked at therapist recollecting more or less of these words than actually appear in the session. At a behavioral level, therapists’ countertransference has been measured in different ways. One is based on quantitative measures (either self-report or completed by external judges, such as supervisors). Another approach to behavioral countertransference has been by having external raters trained to assess therapists’ approach or avoidance responses in connection to a patient and the material he/she brings. Research. Though some studies in countertransference have used qualitative methodology (e.g., Hayes et al, 1998), most investigations have sought to quantify the 22 phenomenon. Among the investigations using a quantitative methodology, we find a few that look at actual real-life treatment (e.g., Markin, McCarthy & Barber, 2013; Palma & Gelso, 2013); however, most studies were experimental analogue and quasi-analogue research (e.g., Peabody & Gelso, 1982; Mohr, Gelso & Hill, 2005; Fauth & Hayes, 2006; Hummel, 2013). The following paragraphs will present research related to countertransference. Following both Gelso and Hayes’ (2007), and Harbin’s (2004) presentation of empirical findings, results will be organized based on Hayes’ (1995) model of countertransference (i.e., distinction of origins, triggers, manifestations, effects, and management factors). Though successful management has been shown to positively relate to therapy outcomes (Hayes, Gelso & Hummel, 2011), it will not be addressed in the current review, as it pertains to an aspect of countertransference that goes beyond the proposed study. Considering origins (i.e., the areas of unresolved conflict a therapist has), research has shown that countertransference can stem from a variety of issues, such as unmet needs, issues related to family of origin, different roles (e.g., as a partner, as a parent, professional), narcissism, etc… (Gelso & Hayes, 2007). In relation to triggers (i.e., events in treatment that elicit a therapist’s conflictual issues), it is central to look at the interchange between a therapist’s and a client’s characteristics and/or content presented (Gelso & Hayes, 2007). Furthermore, it is posited that to really gain understanding of causes of countertransference that are related to cultural factors, we need to consider both client and therapist factors (Gelso & Mohr, 2001). Thus, studies have looked at factors that might predispose therapists to present countertransference in conjunction with stimuli that might elicit such reactions. For example, research has shown that a therapist’s 23 homophobia plays a role in triggering countertransference (manifested as avoidance) when responding to lesbian (Gelso, Fassinger, Gomez, & Latts, 1995) and gay clients (Hayes and Gelso, 1993). As can be deduced from the previous information, there can be myriad origins and triggers for a therapist (Hayes et al., 1998; Gelso & Hayes, 2007). As previously presented, the manifestation of countertransference can be at an internal and/or external level. Regarding the internal level, research has consistently shown that a common affective response therapists have when their conflicts gets stirred up is anxiety (e.g., Fauth & Hayes, 2006; Gelso et al, 1995; Hayes et al, 1998). In relation to cognitive manifestations, it seems that these come in the form of therapists’ misperceptions and distortions. Some studies have shown issues with recollection of specific content (e.g., number of sexual words used by a lesbian client) related to countertransference (e.g., Cutler, 1958; Gelso et. al., 1995), yet others have not found issues with recollection (e.g., Hayes and Gelso, 1993). Countertransference may also influence the therapist’s perceptions, as therapists have been found to perceive clients as overly similar or overly dissimilar when they have countertransference (e.g., Fiedler, 1951; McClure and Hodge, 1987). At a behavioral level, countertransference can present itself as difficulties maintaining appropriate therapeutic distance with the client (Gelso and Hayes, 2007). Empirical work has shown that the most common display of a therapist’s overt countertransference is avoidance behavior, such as changing topic, providing less involving statements when involvement is warranted, among others (e.g., Bandura et al., 1960; Hayes & Gelso, 1991, 1993; Hayes et al, 1998; Latts and Gelso, 1995; Mohr et al., 2005). 24 Considering effects, as previously stated, Hayes, Gelso & Hummel (2011) conducted a meta-analysis and found a negative yet modest relation between countertransference and treatment outcome. In addition, some studies have looked at the relation between countertransference and other variables related to the therapy process and outcome. Among others, results have shown a negative relation between countertransference and the working alliance (Ligiéro and Gelso, 2002), the real relationship (Palma & Gelso, 2013), and the level of superficiality in sessions (Markin, McCarthy & Barber, 2013). When examining the previous research, it becomes clear that to explore countertransference phenomena, is central to consider the interplay between therapists and clients. In addition, it seems that countertransferential reactions in cross-racial dyads are a multilayered phenomenon. So far, attention has been given mainly to a client’s anger as stimulus to trigger a therapist’s countertransference. On the other hand, a therapist’s anger has been seen as a potential manifestation of countertransference (e.g., Hayes et al, 1998), but not much research has looked at its role when the therapist’s issues get triggered. Perhaps the therapist’s reactions to the client’s anger have to do with his/her own experience (e.g., discomfort) of anger (which itself can have its origins in unresolved conflicts). The following section focuses on a brief discussion on the construct of anger, particularly regarding the therapist’s experience. The Experience of Anger Anger is one of the basic human emotions and thus part of common human experience. Though it has been defined in many different ways, most current perspectives highlight that it is a multidimensional construct, composed of several interrelated 25 components, such as cognitive, physiological, affective and behavioral (e.g., Sharkin, 1988; Brondolo, DiGiuseppe & Tafrate, 1997; Kassinove & Tafrate, 2006). Furthermore, the expression of anger can be at a motoric level, a verbal level, or a combined verbal- motoric level. In addition, anger may or may not be overtly expressed (Sharkin, 1988). Though anger can be considered as an “adaptive internal signal that cues self-protective action” (Kannan et al., 2011, p. 169), theoreticians and researchers have mostly addressed the maladaptive and disruptive aspects of this emotion. For both adaptive and maladaptive anger, a central aspect has been to look at its “…intensity, frequency, duration, causes, and functions…” (Sharkin, 1988). All these different aspects have brought up challenges in connection to assessment of the construct, as it is hard to measure all the different levels at once (Sharkin, 1988). Considering psychotherapy, many theoretical and empirical efforts have been directed at addressing and treating the client’s anger (e.g., DiGiuseppe & Tafrate, 2001; Saini, 2009; Deffenbacher, 2011). Fewer studies have looked at how therapists deal with a client’s anger in session, despite the fact that being the target of clients’ anger is stressful for most therapists (Hill, Kellems, Kolchakian, Wonnell, Davis, & Nakayama, 2003). Furthermore, as Sharkin and Gelso (2001) state, is essential for therapists to respond to and explore a client’s anger. Theoreticians and researchers have mentioned the importance of training to increase effectiveness when dealing with a client’s anger (e.g., Hector et al., 1979; Abernethy, 1995; Hess, Knox, & Hill, 2006). The possibility of exploring and responding to a client’s anger might also be associated with the therapist’s own way of relating to and dealing with anger, beyond particular skills used to address this emotion in session. 26 Literature on the therapist’s anger and/or his/her way of relating to anger is rather scant. Little has been written about the subject, and several of the writings that do exist are in the realm of group therapy (e.g., Hahn, 1995; Van Wagoner, 2000). Some authors have written about the therapist’s anger as a countertransferential reaction to a client (e.g., Maroda, 2010; McWilliams, 2004; Winnicott, 1949), yet few studies have looked at the therapist’s own issues with anger (e.g., discomfort) and how these might get in the way of addressing a client’s anger in therapy. Furthermore, if a therapist has anger issues, these might get triggered when a client directs anger towards him or her. For example, when looking at countertransference reactions to an angry Black/African American client, one could assess how comfortable the therapist is in relation to anger. Even though a therapist’s anger discomfort might reflect internal conflicts, this variable could allow us to better understand the therapist’s responses to an angry client (e.g., based on issues with anger in general, or due to the stereotypes and attitudes related to the race of the client, or both). In addition, perhaps a therapist’s reactions differ when anger is directed to others versus towards him/her. In the following subsection, relevant empirical work related to therapists dealing with a client’s anger is presented. Research Related to Therapists Dealing with Clients’ Anger Studies looking at therapist trainees’ responses to an angry client have shown that trainees tend to use more avoidance behavior (e.g., information giving instead of self- involvement responses) when anger is directed at them versus towards others (Bandura, Lipsher & Miller, 1960; Gamsky & Farwell, 1966; Davis et al., 1985). It has also been found that trainees present more anxiety when dealing with angry client behavior than friendly behavior, regardless of a trainee’s experience (Russell & Snyder, 1963). The 27 challenges of dealing with direct anger are also present for seasoned therapists: In a qualitative study of therapists’ experience with being the target of a client’s anger, “therapists had more difficulty with hostile than suspected-unasserted client anger” (were suspected-unasserted anger was “…defined as the client not directly expressing anger unless strongly encouraged to do so by the therapist”, Hill et al., 2003, p. 477). These therapists were all Ph.D. level in private practice, with 15.54 years of post-doctoral experience on average. Some studies have shown that experience matters, as practiced therapists respond more positively and acceptingly to angry clients (Beery, 1970), and have less negative perceptions of angry clients (Haccoun & Lavigueur, 1979) than less experienced therapists. Furthermore, Hess, Knox and Hill (2006) found that trainees reported increased self-efficacy for dealing with client anger in session after receiving training in the subject. In a different vein, Varble (1968) found that experienced therapists presented more struggles in responding to a client’s anger directed towards them than inexperienced therapists did. In addition, some trainees’ characteristics have been associated with dealing with anger more effectively. Bandura, Lipsher & Miller (1960) found that therapists who use direct forms of expressing their hostility are more likely to use approach responses (e.g., self-involvement) to a client’s expression of hostility towards other than therapists that had challenges expressing their hostility. Also, high –anxiety trainees tend to give more defensive responses when dealing with clients expressing anger than those with low- anxiety (Yulis & Kiesler, 1968). Also, therapists with high need of approval tend to avoid 28 more a client’s hostility, making fewer efforts to explore it and a tendency to ignore it, instead (Bandura et al. 1960). In an effort to further understand trainees’ experience with anger and its association to responding to angry clients, Sharkin and Gelso (1993) examined the relations among trainees’ anger-proneness (i.e., their “general tendency to experience angry feelings”, p. 483), anger discomfort (i.e., “degree to which people feel uncomfortable with their own anger”, p. 483), and their response to an angry client. Trainees responded to a video of an actress portraying a female client directing anger towards her therapist. Therapists in training were randomly assigned to one of two videos (same format and content just different actresses). The dependent variables in this study were trainees’ discomfort with and anger toward the angry client. Results showed that both trainee anger proneness and anger discomfort were positively and significantly correlated to discomfort with the client. Trainee anger proneness was positively correlated to anger towards the client. However, contrary to expectations, trainee anger discomfort was also positively related to anger towards the client (the authors expected a negative relation). Hierarchical regression analysis also showed that both anger proneness and anger discomfort accounted for variance in their relations with the dependent variables. Also, contrary to expectations, these authors did not find a difference related to gender; however, age was negatively and significantly correlated with discomfort with the client, anger with the client, and anger-proneness. Koo and Park (1998) conducted a similar study in Korea finding that the counselor's anger proneness did not relate to their reactions to an angry client, but counselors’ anger discomfort was positively related to state anxiety. These authors also found that counselors with high anger discomfort and 29 high anger proneness had more state anxiety than counselors with high anger proneness but low anger discomfort. Also, Harbin’s (2004) previously presented study examined clients’ strengths, UDO, and therapists’ countertransference. This author had included therapists’ discomfort with anger as a covariate, but he did not find a significant correlation between discomfort with anger and the dependent variables (i.e., affective, cognitive, and behavioral countertransference). From the previously reviewed work and the information on countertransference presented earlier, it seems that the way a therapist deals with a client’s anger might be related to the therapist’s own experiences with anger. Finally, most of these studies reviewed have used analogue procedures (e.g., videotaped vignettes, response to a videotaped client) instead of real therapy encounters, with the limitations that it can present for psychotherapy research (i.e., external validity). However, the findings are valuable in illuminating the therapist’s experience in dealing with an angry client. Due to the ongoing debate on the utility of using analogue procedures in psychotherapy research (Heppner, Wampold & Kivlighan, 2008), experimental analogues are addressed in the following section. The Use of Analogue Procedures in Research As it has been previously stated, many of the studies reviewed have used analogue research methods. According to Heppner, Wampold & Kivlighan (2008) “…a counseling analogue is an experimental simulation of some aspect of the counseling process involving manipulation of some aspects of the counselor, the client, and/or the counseling process” (p. 406). Gelso (1979) discusses two types of analogue studies: The experimental analogue and the correlational analogue. As the author states, both 30 analogues are a simulation of therapy and occur in a laboratory, but in the experimental analogue you manipulate the independent variable, which does not happen in the correlational analogue. Thus, in the latter you cannot make causal inferences. According to Murdock (2011), the use of analogue methodology had its peak in the early 1980s, and has been used less since. The current study was an experimental analogue. Due to the complexity of the phenomena of interest (clients’ race, therapists’ universal-diverse orientation, therapists’ anger discomfort, and therapists’ countertransference), this methodology seemed the most appropriate to approach the present investigation. As with any research methodology, analogue procedures have their own advantages and disadvantages. Advantages The main advantage is that an analogue methodology allows the researcher to have more control of the study’s setting or context (Gelso, 1979). Also, analogues facilitate the examination of aspects of the therapy process that otherwise might bring up ethical and practical challenges (Heppner et al, 2008). For example, when looking at therapist anger it might be unethical to deliberately assign therapists who have anger- discomfort to angry clients in order to examine the reactions. In addition, experimental analogues in particular allow having control of the variables and the setting (Gelso, 1979). According to Heppner et al. (2008), analogue studies allow control “…primarily by eliminating extraneous variables, controlling confounding variables, and manipulating specified levels of an independent variable” (p. 410). Thus, it allows specificity in operationalizing a variable (Gelso, 1979; Heppner et al, 2008). In sum, the strengths of 31 analogue research are precise control and internal validity (Gelso, 1979; Heppner et al, 2008) Disadvantages and Challenges On the other hand, the cost of precise control of the stimulus situation is problems with external validity or the generalizability of the findings (Gelso, 1979; Heppner et al, 2008). Gelso (1979) suggests that investigators address this issue by seeing whether the simplifications of the study are really oversimplifications and thus do not relate to real life therapy. In the same vein, Heppner et al. mention that to assess external validity of analogue studies, it would be important to evaluate the similarity of the analogue variables and the ones present in real-life counseling, considering the following areas: Client variables, counselor variables, and counseling process and settings. Furthermore, these authors underscore the importance of and need for seeking ways to closely resemble client, therapist and counseling process and setting variables in order to increase generalizability. A further discussion on the ways in which this was achieved is presented in the method section. Considering the challenges that real-life therapy might present to study the variables of interest, the current study was an audiovisual analogue where counselor behavior (in the case of this study, avoidant verbal responses will be considered as behavioral countertransference) was a dependent variable (Munley, 1974). Many times we cannot predict a client’s angry response in session. In addition, there will be particular moments in therapy related to the client’s characteristics and/or content that will touch the therapist’s hot buttons, and trigger the therapist’s countertransference. However, waiting for the specific events to trigger the vulnerabilities relevant for this study (e.g., 32 issues with diversity, issues with anger) might be quite a challenge. To date, few investigations have looked at the relations among the variables of interest. Thus, an audiovisual analogue could allow further exploration and understanding of the phenomena that was the focus of this work. In the following chapter, there is a presentation of the specific hypotheses that guided the present study. 33 Chapter 3: Statement of the Problem and Hypotheses Statement of the Problem For some time, there has been a supposition that client-therapist racial/ethnic matching “…should result in stronger therapeutic alliance” (Cabral & Smith, 2011, p. 537) and therefore, better treatment results. This general idea has been tested in different ways, and results have been mixed. Cabral and Smith (2011) conducted a meta-analyses that looked at client-therapist matching. The authors found “…almost no benefit to treatment outcomes from racial/ethnic matching of clients with therapists” (Cohen’s d = 0.09; Cabral & Smith, 2011, p. 537). However, results also suggested clients’ preference for therapists of their own race/ethnicity and a tendency to view therapists of one’s own race/ethnicity somewhat more positively than other therapists. Moreover, the authors discovered great heterogeneity in the findings related to the different racial/ethnic groups. The most noticeable differences were for African American clients, who showed a very strong preference for therapists of their own race/ethnicity, evaluated therapists from own race/ethnicity better, and appeared to have mildly better outcomes with therapists who were also African American. All these aspects might play an important role in the way a White/European American therapist approaches the work with a Black/African American client and the way he/she might react to such client. In spite of the well- known idea that intra-group differences might be greater than between-group differences, looking at therapist-client matching in therapy can still illuminate certain areas of therapy work which at times might get in the way of the treatment process. One of the aspects that can 34 play a role in the treatment of clients that are from different race/ethnicity is the countertransferential reactions that therapists can have to such clients. Several authors have highlighted the importance of looking at cultural factors related to a therapist’s reactions, especially when working with those who are culturally different from them (e.g., Comas Díaz, 2012; Gelso & Hayes, 2007; Gorkin, 1987; Ridley, 2005; Vontress, 1996). Furthermore, authors have even coined a term for countertransference reactions related to cultural factors: cultural countertransference (e.g., Gelso & Mohr, 2001; Jones, 1985; Pérez Foster, 1998; Ridley, 2005), and underscore that therapists have to attend to such reactions. In order to really identify if cultural countertransference is at play when working with a client, one would have to explore with the therapist the origins and triggers of such responses, which goes beyond the purpose of this study. However, if we control for different client factors where only the client’s race/ethnicity is what is different (i.e., White vs. Black client) and we find that therapists exhibit countertransference differentially to such clients, we can think of the client’s race/ethnicity as being a trigger of countertransference. In other words, we all have “hot buttons” that incite our unresolved issues. The client’s race per se might not necessarily activate these hot buttons. It might be that at particular moments or when facing specific situations in treatment, a client from a particular racial/ethnic group will trigger the specific stereotypes a therapist has about this group, which in turn will fuel his/her responses to such a client, including countertransferential reactions. In therapy dyads whose members are from different cultures, there are multiple potential dissonance opportunities related to differential worldviews (Pérez Foster, 1998). One such moment can be when the client gets angry in session. As I mentioned 35 previously, our racial attitudes and stereotypes will influence the way we relate to the other. Considering African American males, there is a stereotype of the Angry Black male, where African American men are seen as trouble making, resentful and prone to violence (Gilbert, Carr-Ruffino, Ivanevich, & Lownes-Jackson, 2003). Lombardo (1978) mentions two African American male stereotypes: The brute, where the African American male is seen as temperamental, violent, primitive and sexually powerful, and the sambo, where they are seen more childlike. According to Hall (2001), the brute stereotype has conveyed African American lack of self-control. Considering that according to Jones, (1997), “…studies suggests that White Americans continue to cling to the image of the dangerous, violence-prone, and antisocial image of Black men” (Sue & Sue, 2008, p. 80), we can speculate that a White/European American might have stronger countertransference reactions to an angry African American male client than to an angry European American male client. Among the different therapist-client cross-cultural dyads, one that has been extensively addressed by clinicians is the White/European American therapist and Black/African American client. Theoreticians have highlighted the particular reactions that therapists can have in such cases (e.g., Jackson, 1973), and have considered some variables that can be related to these reactions (e.g., therapists’ and clients’ racial identity, universal-diverse orientation). In spite of such theoretical emphasis, to the author’s knowledge no study to date has empirically looked at White/European American therapists’ countertransference reactions when addressing a Black/African American client versus a White/European American client. Looking at the relations between these variables seems important. First, the United States’ history is intertwined with racial 36 issues, especially between White and Black people. Such history might still be present in the encounters between members of these two racial groups. Also, people tend to be unaware of their own racial biases (see Comas Díaz, 2012), and there have been different empirical efforts that have documented negative attitudes and discrimination from White people to African Americans (e.g., Dovidio & Gaertner, 1998; Whaley, 1998). In addition, it has been suggested that there is often an angry undertone in cross- racial discussions. As previously stated, the client’s anger toward the therapist might be a critical event that might trigger biases, stereotypes and specific attitudes in the therapist, which in turn can influence the therapist’s reactions to the client. Thus, in order to further understand and disentangle therapists’ responses to clients who are different from them, the current study looked at White therapists’ countertransference reactions to an angry client (White or Black), and some variables that might serve as a buffer against such reactions. Specifically, the present study focused on the relations among clients’ race (i.e., Black or White), therapists’ universal diverse orientation, therapists’ anger discomfort, and therapists’ countertransference reactions, operationalized as self-reported state anxiety, verbal avoidant responses, and self-reported countertransference. Hypotheses The hypotheses guiding the current work are presented in this section. Each hypothesis is presented first, followed by the rationale for the proposition. There were several goals guiding the present study. The first purpose was to assess the relationship between a client’s race and therapists’ countertransference. Considering such objective, it was hypothesized that: 37 1: Countertransference will be greater for White/European American therapists responding to an angry Black/African American client than for White/European American therapist responding to an angry White/European American client, such that 1.a. Self-reported state anxiety will be greater for White/European American therapists responding to an angry Black/African American client than for White/European American therapist responding to an angry White/European American client, 1.b. Therapist verbal responses that reflect avoidance will be greater for White/European American therapists responding to an angry Black/African American client than for White/European American therapist responding to an angry White/European American client 1.c. Self-reported countertransference will be greater for White/European American therapists responding to an angry Black/African American client than for White/European American therapist responding to an angry White/European American client To the author’s knowledge, the idea that a Black/African American client might trigger “…more complicated countertransference reactions and more frequently” (Jones, 1985, p. 178) has been suggested theoretically yet has not been tested empirically. Furthermore, to the author’s knowledge, no previous published studies have looked at the countertransference reactions a White/ European American therapist has to an angry Black/African American client versus to an angry White/ European American client. With all other conditions being equal, any differential reactions between these two groups could be related to the client’s race. 38 Additionally, a central objective in the present work was to look at the predictive value that clients’ race, therapists’ UDO, and therapists’ anger discomfort had for each of the dependent variables (i.e., therapists’ state anxiety, therapists’ verbal responses that reflect avoidance, and therapists’ self-reported countertransference) for White therapists responding to the White versus Black clients. Another central aspect of this study was the assessment of the unique variance accounted for by the independent variable and the status variables on each dependent variable. Therefore, the following hypotheses were presented: 2: The independent and status variables will jointly predict each of the dependent variables, such that: 2.a. White/European American therapists’ UDO level, anger discomfort, and the client’s race will jointly predict therapists’ state anxiety as manifestation of affective countertransference to an angry client. 2.b. White/European American therapists’ UDO level, anger discomfort, and the client’s race will jointly predict therapists’ verbally avoidant responses as manifestation of behavioral countertransference to an angry client. 3.c. White/European American therapists’ UDO level, anger discomfort and the client’s race will jointly predict therapists’ self-reported countertransference to an angry client. 3. In addition, each independent variable will uniquely account for variance in each of the countertransference measures, in a way such that: 39 3.a. White/European American therapists’ UDO will uniquely account for variance in the state anxiety measure, in the verbal avoidant responses to the angry client, and the self-reported countertransference measure. 3.b. White/European American therapists’ anger discomfort will uniquely account for variance in the state anxiety measure, in the verbal avoidant responses to the angry client, and the self-reported countertransference measure. 3.c. The clients’ race will uniquely account for variance in the state anxiety measure, in the verbal avoidant responses to the angry client, and the self- reported countertransference measure. Pertaining to universal-diverse orientation, Harbin (2004) found that it was positively related to countertransference. When the author looked at each dependent variable separately (i.e., state anxiety, cognitive recall and behavioral avoidance) UDO contributed significantly to only state anxiety. In spite of such result, in the current study the predictive value of UDO in relation to each separate aspect of countertransference (i.e., each dependent variable) was examined. As previously mentioned, there were some differences between the two studies: The current study did not assess cognitive recall but looked at self-reported countertransference. Also, in the present work there were two different stimulus clients (Black and White angry client), whereas Harbin had only one (Black client), and also the samples might differ (e.g., he included only therapists in training however in this study the sample included both therapist and therapists in training). Sharkin and Gelso (1993) found that anger discomfort accounted for variance in discomfort with the angry client. Following their findings, the present work looked at the 40 predictive value of anger discomfort in each of the dependent variables. Additionally, to the author’s knowledge no study has looked at clients’ race as a predictor of different manifestations of countertransference. Lastly, in order to further explore the relation between the variables of interest, it seemed valuable to look at possible interaction effects considering the status variables as moderators in the relationship between the independent variable and the dependent variables. Thus, it was hypothesized that: 4. There will be an interaction between the angry client’s race and White/European American therapists’ UDO in affecting countertransference, such that in the Black/African American angry client condition, people with lower level of UDO will have more countertransference than those with higher UDO, whereas levels of UDO will not make a difference in therapists’ countertransference when facing a White/European American angry client. 5. There will be an interaction between the angry client’s race and White/European American therapists’ anger discomfort in affecting countertransference, such that in the Black/African American angry client condition, people with higher level of anger discomfort will have more countertransference than those with lower anger discomfort, whereas levels of anger discomfort will not make a difference in therapists’ countertransference when facing a White/European American angry client. 41 Chapter 4: Method Design The present study was a counseling analogue. Therapists and therapists-in- training were asked to respond to one of four videotaped actors portraying an angry client. The actors were either White/European American (2 actors) or Black/African American (2 actors). Right after the videotape, participants completed a measure of state anxiety, and three single items assessing the influence of the participant’s countertransference in his or her behaviors, thoughts, and feelings while responding to the videotaped client. In addition, a week prior to watching the videotape the participants completed online measures pertaining to the therapist’s anger discomfort, the therapist’s universal-diverse orientation, his/her trait anxiety, social desirability, and a demographic questionnaire. In order to disguise the focus of the study, the pre-lab measures also included a brief questionnaire of attachment targeting romantic relationships (the Experience in Close Relationship Scale-Relationship Structures Questionnaire). Independent Variables. There were three independent variables in this study. One of these was manipulated: Client’s Race (White/European American Client or Black/African American Client). There were two status variables: Universal-diverse orientation (assessed by the M-GUDS-S), and anger discomfort (measured with the ADS). In addition, trait-anxiety (assessed with the STAI-T) and social desirability (measured with the BIDR6) were included as covariates (the first one was included to be partialed out from state-anxiety and the second to be partialed out from UDO). 42 Dependent Variable. The main dependent variable was countertransference. As different manifestations of countertransference were assessed and treated separately, three dependent variables can be identified: affective countertransference (measured by the STAI-S), behavioral countertransference-verbal manifestation (measured by the Avoidant Index), and self-reported countertransference (assessed with the CT Index). The self-reported countertransference was assessed at a behavior, thoughts and feelings level. Participants Ninety-eight therapists and therapists-in-training completed all the pre-lab measures online. However, only 86 participants completed the two portions of the study (seven did not provide contact information to set up the Lab session, four were repeatedly contacted but mentioned being too busy to complete the second part, and one came to the Lab session but the undergraduate RA did not attend). It had been established a priori that the sample for the study would consist only of White/Caucasian, European-American, non-international students. Therefore, International students, non-White/Caucasian, and Biracial participants were excluded from the sample. Additionally, seven participants who considered themselves White and European American yet reported additional heritage (e.g., Latino, Native American, Middle Eastern) where also not included. The decision for such exclusion was that the extent of the influence of such heritage seemed unknown. For example, in one case the participant told the researcher that her mother is Latina, yet in other case the participant mentioned “background of Cuban, German, French, Puerto Rican”. Before collecting data, a power analysis was conducted to determine the amount of participants needed for the current study that would allow detecting anticipated 43 medium size effects. According to Cohen (1992), if alpha is set up at .05 and an 80 percent likelihood of detecting effects is wanted, the sample needed to detect a medium size effect for a multiple regression with 5 independent variables (in the case of the current study are one independent variable, two status variables, and two covariates) is N = 91. Additionally, two experimental studies that have looked at countertransference and have employed two conditions had samples of N=67 (Gelso et al., 1995) and N= 47 (Latts & Gelso, 1995), and both detected main and interaction effects. Harbin (2004), who looked at UDO and countertransference and had two experimental conditions (client’s strengths vs. no information of strengths) had a sample of 45 participants. Based on the previously presented decision-making guidelines, the current study had a sample of 63 White/European American therapists and therapists-in-training. Considering the power analysis, and the sample size of previous studies, a group of 63 participants seemed to be large enough to detect the expected effects. The sample consisted of 53 females (84.1%) and 10 males (15.9%), with a mean age of 38 years old (Range=24 to 65). Such age average excluded two participants: One participant did not share her age, and one participant mentioned she was “over 75”. Of these 63 participants, 51 identified as heterosexual (81%), four as Gay or Lesbian (6.3%), four as Bisexual (6.3%), three as Queer (4.8%) and one as Other (1.6%), who self- identified as being “Pansexual”. Regarding their highest academic degree completed (if currently enrolled, participants were asked to write the highest degree already received), 11 participants mentioned having a BA/BS (17.5%), 16 a MA/MS (25.4%), one a M.Ed. (1.6%), 12 a MSW (19%), 17 a Ph.D. (27%), four a PsyD. (6.3%), and two mentioned “other” (3.2%). The two participants’ degrees that were in the “other” category reported 44 having attained a “Master in Divinity”, and a “MSW and JD”. Additionally, 29 participants (46%) mentioned they were currently enrolled in an academic degree (46%). The mental health programs that participants endorsed where: 24 participants in Counseling Psychology (38%), 18 in Clinical Psychology (28.6%), 14 in Social Work (22.2%), one in Counselor Education (1.6%), and six other (9.5%). In terms of clinical work, the average of years providing therapy by the participants was 9.7 years (ranging from less than a year to 41 years). Additionally, 33 participants (52.4%) endorsed being a Licensed Mental Health Provider. Participants also rated different theoretical approaches on a scale from 5 (Strongly Representative) to 1 (Not at all) in terms of how such perspectives represented their work. The average representativeness for Psychodynamic/Psychoanalytic theory was 2.25 (SD = 1.28), the mean for Cognitive/Behavioral theory was 2.40 (SD = 1.12), the mean representativeness for Systemic theory was 3.27 (SD = 1.25), for Humanistic Experiential theory was 2.29 (SD = 1.31), and other was 2.94 (SD = 1.72). Some of the theoretical perspectives that participants endorsed as “other” were: Mindfulness, Multicultural, Rogerian, Relational, Schema Focused Therapy, ACT, Feminist, Energy Psychology, Integrative. Finally, 25 participants completed the Lab portion of the study in a Lab established at the Counseling Psychology Program at the University of Maryland. Thirty- three participants completed the Lab portion in the office were they regularly saw clients and five completed it in other areas (e.g, a general office in their practice, a meeting room at their counseling center). 45 Measures Miville-Guzman Universality-Diversity Scale –Short (M-GUDS-S; Fuertes, Miville, Mohr, Sedlacek, and Gretchen, 2000). The M-GUDS-S is a 15-item, self-report measure, which assesses universal-diverse orientation (see Appendix A). It is based on the Miville-Guzman Universality-Diversity Scale (M-GUDS), a 45-item, self-report measure created by Miville et al. (1999) to assess UDO. Miville et al. defined the construct of UDO “… as an attitude of awareness and acceptance of both the similarities and differences that exist among people.” (p. 292). The instrument has adequate validity and reliability (Miville et al. 1999). Specifically, Miville et al. conducted a series of studies in which they looked at the association of the M-GUDS with theoretically related constructs to determine convergent and discriminant validity. As expected, the M-GUDS was positively related to empathy, healthy narcissism, androgyny, and measures of racial identity. In addition, the M-GUDS was negatively related to dogmatism and homophobia. Also, the M-GUDS showed to be reliable, with internal consistency ratings and test-retest reliability that ranged between .89 and .95 (Miville et al, 1999). The M-GUDS-S is a short form of the Miville-Guzman Universality-Diversity Scale. In this self-report measure, respondents have to rate 15 items based on a 6-point Likert-type scale, which ranges from 1= strongly disagree to 6= strongly agree (Fuertes et al., 2000). This instrument has three subscales: Diversity of Contacts (e.g., “I often listen to music of other cultures”), Relativistic Appreciation (e.g., “Persons with disabilities can teach me things I could not learn elsewhere”), and Comfort with Differences (e.g., “I am only at ease with people of my own race”). Thus, it provides a total score and three subscales score. 46 In terms of reliability, Fuertes et al. found the following Cronbach alpha: Diversity of Contact: .82, Relativistic Appreciation: .59, Comfort with Differences: .92 and Total Score: .77. Thus, overall, the MGUDS-S is sufficiently reliable (Fuertes et al., 2000). In addition, the correlation between the short form (MGUDS-S) and the long form of the scale (M-GUDS) was .77. The Cronbach alpha for the current sample was .82 for the total MGUDS-S scale. Finally, according to Fuertes et al. (2000), the short form has three advantages over the M-GUDS: (1) Strong correlation with the original yet is shorter, (2) Clearer factor structure, and (3), it “…allows for an analysis of UDO using subscale scores” (p.167). Anger Discomfort Scale (ADS; Sharkin and Gelso, 1991). The ADS is a 15-item, self report measure (see Appendix B). The scale seeks to assess the level of a person’s discomfort with his/her own anger. Sharkin and Gelso’s conceptualization of anger discomfort includes intrapersonal (i.e., feeling threatened by one’s own experience of anger) and interpersonal (i.e., concerns about how others will react to own anger) elements. Among the items that comprise the ADS we find: “I feel guilty about being angry at others”, “People don’t seem to like me when I’m angry”, and “I am comfortable with my angry feelings”. Items are rated on a 4-point Likert-type scale, ranging from 1= almost never to 4= almost always, with higher scores reflecting greater discomfort with one’s own anger (Sharkin & Gelso, 1991). The authors established the ADS validity by its positive correlation with anger expression, anger suppression, and trait anxiety. Additionally, Sharking and Gelso assessed reliability calculating internal consistency (alpha = .81) and test-re-test coefficient (.87). Analysis of the measure showed four factors; however, only 47 the total score will be considered in the present study. The internal consistency value for the current sample was .82. The Balanced Inventory of Desirable Responding (BIDR6; Paulhus, 1984, 1986, 1994, 1998). The Balanced Inventory of Desirable Responding is a 40- item questionnaire, which assesses a person’s proclivity to respond in socially desirable ways when completing self-reported measures. It consists of two scales: Self-deceptive enhancement (SDE; e.g., “I am fully in control of my own fate”, “I never regret my decisions”) and impression management (IM; e.g., “I never swear”, I sometimes tell lies if I have to”). The SDE scale assesses an agreeable self-presentation that is based on an inaccurate self-image. The IM scale reflects a tendency to consciously deceive others. Each scale consists of 20 items rated on a 7-point likert scale. According to Paulhus, the alphas typically range from .66 to .77, for the SDE scale, and .77 to .85 for the IM scale. The coefficient alpha values obtained in the current study were .73 for the SDE scale and .82 for the IM scale. Validity for the IM scale was established through correlations with MMPI Lie (L) scale and the L scale of the Eysenck Personality Inventory (Paulhus, 1986). The SDE scale has correlated with Block’s ego resiliency scale and Edwards Social Desirability Scale (Paulhus, 1986), among others. The validation sample for the BIDR consisted of College students. There is a newer paid version of the scale, The Paulhus Deception Scales (PDS). The norm groups include prison inmates, military recruits, and general population. The PDS has been used in studies that look at forensic population; however, the BIDR-6 Form 40 (Paulhus, 1991) is the most widely used form in published research (Lanyon and Carle, 2007). (See Appendix C) 48 The State-Trait Anxiety Inventory for Adults (STAI; Spielberger, in collaboration with Gorsuch, Jacobs, Lushene, and Vagg, 1968, 1977). This is a 20-item self-report measure that assesses state anxiety (S-Anxiety; transient state characterized by subjective feelings and thoughts related to tension, nervousness, apprehension and worry) and trait anxiety (T-Anxiety; “…relatively stable individual differences in anxiety proneness…”, Spielberger & Reheiser, 2004, p. 70; Spielberger, et al., 1999, p. 997). In the S-Anxiety scale, individuals are asked to rate the intensity of their present anxiety (“right now, at this moment”) responding to each item on a 4 point-scale (1: Not at all, 2: Somewhat, 3: Moderately or 4: Very much so). Example of items of the S-Anxiety are “I feel calm”, and “I feel tense”. For the T-Anxiety individuals have to report their experience of anxiety based on how they generally feel. In the T-Anxiety Scale, respondents are asked to rate each item on the following 4-point scale: Almost Never, Sometimes, Often, Almost Always. Some of the items of the T-Anxiety are “I am a steady person”, and “I lack self-confidence”. There have been several forms of the STAI. The latest form, Form Y, has shown better structure and better factor differentiation than the previous one (Form X). In terms of reliability, test-retest sores for the STAI T-Anxiety scale ranged from .73 to .86 with a sample of College Students. In terms of internal consistency, the median alpha for independent samples of working adults, students and military recruits was .90 for the T- Anxiety and .93 for the S-Anxiety. For the current sample, the Cronbach alpha for the STAI-Trait scale was .85, and for the STAI-State scale was .93. Spielberger and Reheiser (2004) state that throughout the scale development process, the items for each scale have met strict validity criteria. Pertaining concurrent validity, the STAI T-Anxiety has shown 49 to be highly correlated (i.e., correlations ranging from .73 to .85) to other measures of anxiety (ASQ, MAS). In the present study, the S-Anxiety was used to assess affective countertransference (i.e., therapist’s anxiety). The T-Anxiety Scale was included to partial out Trait Anxiety from the State Anxiety. The State anxiety measure is presented in Appendix D. Countertransference Index Scale (CT Index Scale). This scale was created for the current study, based on the CT Index created by Hayes, Riker, Ingram (1997). The CT Index is a single item, which is rated on a 5-point Likert-type scale, ranging from 1= strongly disagree to 5= strongly agree (see Appendix E). The authors created this measure to assess “the extent to which ‘the counselor’s behavior in session was influenced by countertransference (i.e., areas of unresolved conflicts)’” (p. 147). It is to be completed by counselors or supervisors right after a counseling session. Hayes, Riker and Ingram found a positive and significant correlation between counselors and supervisors scores of the CT index, which provided evidence for interrater reliability. For this study, two modified versions of this question were incorporated, which assessed the extension that thoughts and feelings in session were influenced by countertransference (see Appendix E). Correlations among the different CT Index items (i.e., the three single- item measures, assessing the extension that (1) thoughts, (2) feelings, or (3) behaviors in session were influenced by countertransference), were as follows: CT Index assessing behaviors and CT Index assessing thoughts (r(61) = .57, p=.000), CT Index assessing behaviors and CT Index assessing feelings (r(61) = .60, p=.000), and CT Index assessing feelings and CT Index assessing thoughts (r(61) = .68, p=.000). Due to the high correlation among the different CT Indices, these were combined into one score and thus 50 were considered as only one dependent variable. The Cronbach alpha coefficient for the whole scale was .83. Avoidance Index. The behavioral component of countertransference was considered in terms of therapists’ verbal approach-avoidance responses, following the procedure introduced by Bandura, Lipsher and Miller (1960). In this procedure, participants’ responses are assigned different codes (e.g., approval, exploration, disapproval of the client, topical transitions), which correspond to general categories of approach responses (i.e., responses that generate further expression of feelings and attitudes from the client), or avoidance responses (i.e., responses that discourage, inhibit or divert the client’ s expression). There is also a code of “other” for responses that do not seem to fit in any other category. It is expected that such avoidant responses are a manifestation of a therapist’s countertransference, as they protect the therapist from further expression of the client’s feelings or material that can trigger the therapist’s discomfort (Latts & Gelso, 1995). Then, a cumulative ratio of avoidant responses for each therapist is calculated, based on the number of avoidant responses to the number of approach and avoidance responses. These procedures have been successfully used in previous studies assessing countertransference (e.g., Gelso et al., 1995; Harbin, 2004; Hayes & Gelso, 1993; Hummel, 2013; Latts & Gelso, 1995). The coding system used for the current work followed a slight modification that Hummel (2013) introduced to Bandura et als.’s (1960) coding system. This modified version of the system made more theoretical sense for the present study, as such adjustment introduced the option of “Colluding/Inappropriate approval”, which the previous coding system did not account for. However, it got rid of the code of “silence”. 51 During the coding process for the current research, it became clear that in a few cases there was silence for the whole speaking turn. After coding, these few cases (e.g., one case in which the participant did not respond to any speaking turn; a case in which the participant responded in some turns and not others) were reviewed, and it was decided that the silences in such few cases could be related to the actual phenomenon that the current work was interested in (i.e., countertransference). Gelso et al. (1995) coded silences as avoidance only in the cases in which there was silence for the whole speaking turn. Therefore, it was opted to follow Gelso et al., and the silences that occurred during a whole speaking turn where finally accounted for as avoidance (thus, one more code was added to the system). The specific procedure used in the present study to calculate the Avoidance Index is described in the following two sections. (see Appendix F for a general view of the classification and specifics of the trainings). Transcribing and Unitizing. Three psychology undergraduate students transcribed the participants’ verbal responses. These research assistants were divided in pairs, and each pair was assigned to each transcript; one research assistant transcribed the responses and the other reviewed to ensure nothing was missing. These students were also extensively trained in the unitizing system first introduced by Hill et al. (1981) in which verbal responses are divided into grammatical units (e.g., independent sentences). Once a 90% of agreement in the unitizing of several practice transcripts unrelated to the participants’ responses was reached, the undergraduates started unitizing the participants’ transcripts. All undergraduates unitized half of the transcripts (i.e., 32 transcripts were each unitized by every research assistant), and weekly meetings were held to discuss the given units and to come to agreement on any differences. Once consistent unit agreement 52 was sustained across several participants’ transcripts, pairs of undergraduate students were assigned to unitize the transcripts that were left (thus, each transcript that was left was unitized by two research assistants). Each pair was instructed to discuss any unitizing disagreement until they had reached consensus. Coding. Two advanced Counseling Psychology Ph.D. students and one Counseling Psychologist coded every unitized transcript. The three raters had previous experience with coding verbal responses in counseling/therapy. The coders were extensively trained, which included a whole afternoon meeting to present the coding system, discuss specifics of the work, practice coding of some transcripts and discuss the assigned codes afterwards to reach consensus. After that initial meeting the raters also individually coded several practice transcripts, and we held weekly meetings via Skype to discuss such codings and reach consensus. This process was maintained until the coders had reached 85% consensus in their individual practice codings. Additionally, while coding, raters had a personal “Log” where they could write the reasons behind some of their coding decisions, the questions they had while coding, among others. Every coder rated each participant’s transcript. For each transcript of the participants’ responses, the coding process involved assigning a specific code to every unit in every speaking turn (there were 6 speaking turns and each could have several units). Then, for each speaking turn in the transcript, the coder needed to assign an overall code. For this, coders were asked to consider the unit that might carry more weight (e.g., have a greater impact) for that speaking turn. Ratings at the unit and turn level were based on the specific codes (e.g., approval, ignoring) and categories (i.e., approach or avoidance) previously mentioned. All these ratings were considered later on 53 to create an “Avoidance Index”, which is the ratio of the number of avoidance responses over the number of approach plus avoidance responses. Previous studies have calculated the Avoidance Index based on the unit level rating (i.e., each unit in every speaking turn, e.g., Gelso et al., 1995) or they had chosen analysis at the turn level (i.e., an overall code, one for each speaking turn, e.g., Brittan, 1993; Harbin, 2004; Hummel, 2013). At a theoretical level, it makes sense to analyze the data of the verbal responses that carried more weight. During clinical work, a therapist might speak several sentences during a speaking turn, yet it might take just one of those phrases to really affect the client (for better or for worse). Furthermore, according to Brittan (1993) within a speaking turn, one impactful avoidance response from the therapist could diminish any other approach responses for that turn. In addition, the relation between each coder’s Avoidance Index with ratings at the unit level and at the overall speaking turn was calculated. Results showed that for each coder, the Avoidance Index at the unit level and at the turn level were strongly correlated; r= 76 (for rater 1), .87 (for rater 2), .77 (for rater 3), all at p < .01. Considering the previously presented theoretical basis and the correlational analysis, it was decided that the Avoidance Index to be used for the statistical analysis of the hypotheses would be the one considered at the turn level (i.e., one overall code for each turn). To assess inter- raters’ reliability, a Two-Way Random Consistency Intraclass Correlation was calculated. Results showed ICC (2,3) = .84; Therefore, 84% of the variance in the mean of the three raters does not represent random variation. Additionally, correlations of the Avoidant Index between coder 1 and 2, 1 and 3, and 2 and 3 considering the overall speaking turns were as follows: r=.72, .54, 65. 54 Demographic Questionnaire. Therapists completed a demographic questionnaire, which included questions related to therapist race/ethnicity, gender, sexual orientation, age, international status, most advanced degree, years of clinical experience, and theoretical perspective. As both trainees and therapists who had graduated were invited to participate, the questionnaire also included queries on whether the therapist was currently in training, licensed, etc… (i.e., current professional status). (See Appendix G) Open question. In the present study it was also included an open question that therapists had to complete after watching the video and responding to the client (right after completing the state anxiety measure and counterbalanced with the CT index measures). Therapists were asked to “Please briefly describe any emotional reactions you had to the client”. This question will be utilized in a future study, thus it was not included in the current work. Stimulus Material The stimulus client for the present study was an angry male client. Specifically, therapists watched a tape of either an angry Black/African American male client or an angry White/European American male client. The idea behind having an angry male client was related to several issues. First, as previously stated, dealing with a client’s anger can be a critical moment in session, especially if the anger is directed towards the therapist. Second -and perhaps a more telling reason that is also addressed by Harbin (2004)- is that European Americans might have intense reactions to an angry male Black client. According to Sue and Sue (2008), the way Black people communicate (e.g., often heated, animated communication) generates several emotions and feelings, which is different to the way in which White middle-class people communicate (e.g., are more 55 detached and objective). Furthermore, these authors mention that White people might misinterpret the communication style of Black people. Such misinterpretations might trigger a European American therapist’s “hot-buttons”, generating in the therapist countertransferential reactions related to the stereotypes and attitudes s/he has towards African Americans. In addition, it has been suggested that a Black client might generate more complicated countertransferential reactions than one from other race, due to the societal image of Black people that shapes the therapist’s projections (Harbin, 2004; Jones 1984, 1985). As previously stated there are powerful stereotypes in our society pertaining to the angry male black client (e.g., the brute stereotype as presented by Lombardo, 1978), which might readily surface when the client portrayed in the stimulus video is an angry Black/African American client who directs his anger towards the therapist. Thus, one might expect that the countertransference a White/European American therapist has towards an angry Black/African American male client will be more frequent and/or intense than to an angry White/European American male client. Also, though there are some strong stereotypes in relation to an angry Black/African American woman, these seem less threatening than those of an angry male from the same racial/ethnic group. Therefore, it was decided to only have male clients as stimuli. To portray the stimulus client, actors were recruited via professional acting websites in the DC/MD/VA area. Approximately 80 professional actors contacted the author, and 25 were interviewed either by phone, via Skype or in person. Actors had to perform part of the script in the interview to show how well they could portray the client. Actors’ selection was based on their performance and also on how credibly they represented the intended client (e.g., physically they could be seen as a 36 year old male). 56 To control for actor effects and to increase the generalizability of the results, two actors portrayed each client (thus, there were four actors and videos in total). Because some physical characteristics of a client could be related to a therapist’s reaction to an angry client (e.g., client’s size, certain types of tattoos), the actors were paired regarding these characteristics (e.g., similar weight of one Black/African American client and one White/European American client). The actors followed the same script verbatim, to control for any differences pertaining session content. In addition, the actors wore similar clothes, and were paired-up in terms of non-verbal responses (see below). Also, actors were economically compensated for their participation in the study. Following Brittan’s (1984) and Harbin’s (2004) procedures, the actors were judged on likeability, believability, attractiveness and emotionality on a 5-point Likert- type scale (1= very low, 5=very high). Harbin stipulated a priori that for a tape to be used, the actor in it should receive an average rating of 4 or higher in believability, 3 or more in emotionality, likeability and attractiveness. Preliminary ratings performed by a European American male psychologist and a biracial male psychologist were all above the previously presented cut-off values. However, when assessed by the six officially assigned raters, all the possible videos had average ratings below Brittan’s and Harbin’s cut-off point scores. After reviewing the different ratings and the comments made by the raters, it was decided that the videos would be chosen based on highest ratings and similar ratings of the actors in each condition (e.g., ratings of actor 1-African American- in video 3 similar to ratings of actor 3-European American- in video 2; a few videos were made for each actor). That way, ratings were balanced in each condition. The specific rating values and a detailed discussion of the issue will be presented later on. 57 Harbin’s team of judges also rated their African American actors on Negroid vs. Caucasoid features. They established a rating of 3.0 for tapes to be used. For the Negroid vs. Caucasoid features ratings, the Black/African American client had to receive high ratings (at least 4) in Negroid features and low (2 or less) in Caucasoid features, and vice versa for the White/European American client. This allows client differentiation in terms of race. The same standards were followed in the current work. As in previous studies (e.g., Gelso, et al., 1995; Harbin, 2004; Hayes & Gelso, 1993; Hummel, 2013), the video was presented as individual segments (6 in the case of the current work) with pre-determined stopping points. The actor was facing the camera, and the background was similar to a counseling office. This was to give the therapist a sense of being sitting with a client. In addition, as the session advanced there was a progression of the client being angry in general, then directing it at others and finally directing his anger towards the therapist. One of the main challenges in terms of the script was how to create a simulation that could be believable and reflected the experiences of both a Black/African American angry male client and a White/European American angry male client, yet kept things identical for both clients. First was the issue of representing the communication styles for both clients at verbal and non-verbal levels, as there are important differences between these two groups. Second, there is the matter of providing some brief background that was believable for both clients. It was decided that the client should have at least college education. Matching the clients based on educational level might allow more similarity in the verbal expressions used by both clients. Following Brittan’s (1984) work, it had been determined a priori that the client would work in the helping 58 profession. Brittan’s client was a professional counselor, in order “…to make the stimulus more ego near” (p. 58) for the participants. However, several therapists who reviewed the original script found that the level of anger of the stimulus client might hardly be present in someone working in the helping professions (one of the reasons given was that it would be expected that someone from the helping professions might regulate better). Thus, it was finally decided that the client would work in an Insurance Company. The client’s presentation and the script can be found in Appendix H. These were created based on Brittan’s (1984) work and the literature reviewed. Some of the ideas were presented more vaguely (e.g., attribution of not being promoted), to serve as a blank canvas on to which participants projected their attributions, attitudes and stereotypes (e.g., perhaps attributed not being promoted to problems in the person of the Black/African American versus problems of not having had the opportunity to succeed in the case of White/European American). Several steps were taken to ensure that the script was representative of issues for White/European American and Black/African American males. These steps are described as follows. Development of the script. The main author wrote a script, which was initially reviewed by her advisor, a European American/White male. After that review, the main author sent the written script to an African American/Black male graduate student in the Counseling field, and to a Biracial male psychologist to review it. Then, the author met with three undergraduate students in an ongoing process of re-evaluation and re-writing of the script based on the suggestions. 59 Once it was re-written, the undergraduate students separately rated each segment of the script in terms of intensity of anger (0= No anger to 5= Extreme anger), believability (1= Not believable at all to 5= Extremely believable), and wrote the words that pertained to anger in each segment. The students also gave an overall rating to the script, ranging from 1 to 5 (1=Not believable at all to 5= Extremely believable). The biracial male psychologist and the author’s advisor also completed these ratings. It was predetermined that in order to retain a segment (i.e., to include it in the final version of the script) it needed to attain an average score of at least 3 in believability. Additionally, it was established that to use the script it needed to have an average score of 4 in the overall rating (as previously mentioned, such rating could range from 1=Not believable at all to 5= Extremely believable). Through this revision process, the lengths of the segments were shortened, some words were changed, and a 7th and final segment was eliminated (seemed too “over-the-top and not believable” and/or that such escalation of anger might be more in in-patient work than in private practice). The final version of 6 segments was resent to the author’s advisor, the African American/Black graduate student and the biracial psychologist for a final view. The ongoing revision process was to ensure that the script would be representative of White and Black male’s experiences in a believable way. The revision process ended when the three previously mentioned reviewers agreed that it was a representative script (based on ratings and on general feedback). Ratings of the Videos. Three Black/African American and three White/European American raters watched and rated the videos. Four of these raters were early career 60 counseling psychologists, one was a counseling psychology Ph.D. student, and the other was doing a Ph.D. in counselor education. The raters had the instruction to view six videos. Following Brittan (1993), it had been decided a priori that after each video, the raters would complete ratings in the following categories: Likeability, believability, attractiveness, and emotionality, using a 5-point scale Likert scale (1 = very low, 5 = very high). However, in separate discussions with different raters and with the biracial male psychologist who rated the script, the issue that emotionality encompasses too many different emotions came up. Specifically, a rating of 4 in emotionality for one actor could be related to intense anger, however in another actor such rating might be reflecting intense frustration. Thus, it was agreed that they would rate anger. It was also pointed out that African-American males could show anger with different levels of aggression (e.g., as frustration, as aggressiveness). The means for believability for the four actors given by the raters were as follows: 3.2 (SD = 1.6), 4.7 (SD = 0.5), 4.2 (SD = 0.8), and 3.5 (SD = 1.4). In terms of likeability, the means given by the raters to the actors were 2.3 (SD = 1.0), 3.3 (SD = 1.2), 2.7 (SD = 1.0), and 2.2 (SD = 1.2). Regarding attractiveness, mean ratings given by the judges per actor were 2.5 (SD = 0.5), 4 (SD = 0), 2.3 (SD = 1.2), and 2.3 (SD = 0.8). Finally, the mean ratings given per actor in terms of their portrayal of anger was 3.8 (SD = 0.8), 3.7 (SD = 1.0), 4.2 (SD = 1.0), and 3.2 (SD = 0.8). As previously mentioned, actors were also rated in their Caucasoid and Negroid features. Specifically, in the present study the White/European American actors received a rating of 5 in Caucasoid features, except in one case, in which the rater assigned a value of 4 for one of the actors. In the case of Negroid features for these actors, all but one 61 actor were rated as 1 (the exception was an actor who received a rating of 2 by the same rater that assessed Caucasoid features as 4 for this actor). A similar situation occurred in the ratings of the Black/African American actors, where all but one rating had a value of 5 (the exception was a rater that assigned a rating of 4 to one of the actors) in Negroid features. Regarding Caucasoid features, the two actors were rated as 1 by most raters except one case in which the rater assigned a 2 (which corresponds to this rater’s assignment of 4 in Negroid features). Therefore, based on the ratings it was determined that the actors’ features were representative of their specific Racial/Ethnic groups, and the tapes could be used in the present study. The mean ratings given to the actor’s performance and their features can be found in Table 1. Additionally, to have a better sense of what was behind the different numbers assigned, the raters were asked to write comments regarding their ratings. Thus for example, in a specific video, several raters provided lower ratings. When looking at the comments, it was clear that such values were assigned due to a technical issue (different segments were chosen to create a full video, and when shown it seemed choppy). However, this issue was solved when such video was shown in Qualtrics (an online survey platform), which was segment-by-segment (originally the videos were going to be shown as a full video for participants to stop after each segment. However, due to Qualtrics’ technical aspects, each segment had to be shown separately). As it was previously stated, the ratings were lower than the cutting points we had determined a priori. However, it was difficult to disentangle whether such ratings were due to actors’ performance, raters’ personal reactions to the content presented (e.g., anger towards the therapist), raters’ reactions to the actors (e.g., countertransference related to 62 race), or other issues (e.g., lower ratings in attractiveness because the actor was angry). The main author and her advisor discussed this at length. When looking at the ratings of all the actors, it was seen that the ratings of one Black actor and those of one White actor were closer and higher than the other two actors (i.e., the other White and the other Black actor). For example, the rating for Believability for Actor 2 (Black) was 4.7, and for Actor 3 (White) it was 4.2, therefore these actors seemed to have close ratings in Believability. Additionally, those actors with lower rating also had similar numbers. For example, the rating for Actor 1 (Black) in Believability was 3.2, and for Actor 4 (White) it was 3.5, therefore these actors seemed to have similar ratings in Believability. Thus, each pair was deemed comparable. Also, as previously stated, the raters scored six videos, but only 4 were used. The videos that were dropped were the ones that had lowest ratings for the different categories, and also that were more negatively considered based on the written comments provided by raters. Finally, the written comments related to each of the video ratings were considered. Based on all the considerations, the videos were chosen as usable. Manipulation Check In order to check whether the racial manipulation worked, the experimental material was judged by independent raters (Heppner et al., 2008). As previously reported, the ratings of White/European American actors were high in Caucasoid features and low in Negroid features. The ratings of the Black/European American actors were high in Negroid features and low in Caucasoid features. 63 Table 1 Means and Standard Deviations of Actors’ Client Portrayal and Their Racial Features. Actor 1 2 3 4 Black/African American White/European American Actors Actors Believability 3.2 4.7 4.2 3.5 (1.6) (0.5) (0.8) (1.4) Likeability 2.3 3.3 2.7 2.2 (1.0) (1.2) (1.0) (1.2) Attractiveness 2.5 4.0 2.3 2.3 (0.5) (0) (1.2) (0.8) Anger 3.8 3.7 4.2 3.2 (0.8) (1.0) (1.0) (0.8) Negroid Features 5.0 4.8 1.2 1.0 (0) (0.4) (0.4) (0) Caucasoid Features 1.0 1.2 4.8 5.0 (0) (0.4) (0.4) (0) Note. Standard deviations are provided in parentheses. 64 Procedure Recruitment. Therapists and therapists-in-training were contacted via email, word-of-mouth, and professional listservs. In the cases when contact information was available, potential participants received a personalized message including a brief description of the study, instructions on how to proceed, and a link to complete measures online. Each potential participant received up to three emails inviting him/her to participate (one initial and two follow-ups). (see Appendix I). The first step in the recruitment process was to contact the directors of clinical training in Clinical Psychology, Counseling Psychology, and Counselor Education programs in the DC and MD (programs with close proximity to DC) area. The person in charge of the students of the Family Science Department at UMD was also contacted. Specifically, such key individuals were asked to provide the personal contact information of the students in his/her program (name and email) or to share the recruitment email (in case the first option was not possible) with them. It was specified to the people providing the contact information that only students who had completed at least one practicum course (in one case a student was finishing the first practicum) were illegible to participate. Several strategies to recruit licensed therapists or therapist in the process of licensure were followed. First, an email was sent to the UMD –Counseling Psychology program alumni who live in the DC/MD/VA area. Then, different clinics and health centers at the UMD campus (i.e., Psychology Clinic, MPCRL, Counseling Center, Health Center) were contacted. In the contact email, the therapists were also invited to forward this invitation to any colleague, supervisee, etc.… that might be interested in being part 65 of the study. Finally, the recruitment email was also posted in different listservs related to therapists in the DC/MD area, and in some institutes related to professional training. Pilot test. Before starting to collect data, the procedure was tested with several participants who were not aware of the study’s hypotheses. Participants’ feedback was mostly positive, and supported the chosen procedure. A few minimal adjustments were made based on the given suggestions. Completion of measures pre-laboratory session. The email invitation to participate in the study included a link to complete the measures online. The link directed the participants to a Qualtrics site, where after completing the Informed Consent (see Appendix J), participants had to complete the following measures: Miville-Guzman Universality-Diversity Scale –Short (M-GUDS-S), Anger Discomfort Scale (ADS), Trait Anxiety scale of the State-Trait Anxiety Inventory (STAI - Trait), The Balanced Inventory of Desirable Responding (BIDR), and another measure to distract from the study’s hypothesis (i.e., Experience in Close Relationship Scale-Relationship Structures Questionnaire; ECR-RS). The order of the measures was counterbalanced, to ensure that instrument arrangement did not influence the results. Participants were also asked to send the day/times in which they could complete the lab visit and to choose a preferred location for the Lab (at UMD or at the participant’s office). In a few cases, the lab was completed in a place different than the previous options (e.g., a clinic’s conference room). Extensively trained undergraduate research assistants run the Labs at UMD, whereas the main author run the experiments both in the offices and at UMD. 66 Between the participant’s online data completion and the time when the lab visit was scheduled (approximately a week), the main researcher reviewed the participant’s demographic information. Such facts were relevant for the random assignment of the participants. First, even though therapists from all races/ethnic background were invited to participate, the data used was only from White, European American therapists. Thus, to provide equivalence between experimental groups, White/European American participants were randomly assigned to either condition. Therapist’s gender was also considered in the random assignment, in order to make the two groups equal in the proportion of males and females. Laboratory procedure. Once the participant arrived to the Lab, s/he received a brief explanation of the procedure. Participants were also guided in how to play the video in Qualtrics and how to advance to the following video. Therapists were told to respond as they would usually reply to a client, and that once they had done so to move immediately to the next segment. After the explanation ended, the researcher left the room. The laboratory procedure followed the method used in Gelso et al. (1995). Thus, participants first read a case summary pertaining to the client. Following Brittan (1984), after the case material there was a sentence asking participants to assume that they were about to have the fifth session with the client. Then, the therapists watched the video of a male client (Black or White) who was angry. The 6 segments of the therapy session clip were presented individually, and the therapists needed to advance to the following once they had responded to that segment. After responding to all the video segments, participants completed the state scale of the State-Trait Anxiety Inventory (STAI). Then, 67 in a counterbalanced order, they responded to the following question: “Please briefly describe any emotional reactions you had to the client”, and they completed the three Countertransference Index measures (the original plus two created for this study). Finally, participants were debriefed in relation to the study (Appendix K), with emphasis on the importance of not sharing information about the study until completion of the project. 68 Chapter 5: Results Descriptive Data The descriptive data for all the measures was calculated. The means and standard deviations for each measure were as follows: Miville-Guzman Universality-Diversity Scale –Short (MGUDS) M = 73.5 (SD = 7.64), Anger Discomfort Scale (ADS) M = 26.5 (SD = 6.12), Balanced Inventory of Desirable Responding (BIDR6) - Self Deceptive Enhancement (SDE) scale M = 6.13 (SD = 3.20) and Impression Management (IM) scale M = 6.84 (SD = 4.02), State-Trait Anxiety Inventory for Adults (STAI) - Trait scale M = 1.78 (SD = 0.33) and State scale M = 2.10 (SD = 0.53), Countertransference Index (CT) – assessing influence of countertransference in behavior (B) M = 2.52 (SD = 1.11), thoughts (T) M = 2.9 (SD = 0.98), or feelings (F) M = 3.00 (SD = 1.09), and Avoidance Index M = .35 (SD = .27). The correlations among the variables of interest are presented in Table 2. Additionally, the means and standard deviations per experimental conditions are provided in Table 3. The internal consistency for each measure was also estimated, using Cronbach’s alpha, and these values are reported in the measure description section. The alpha coefficients for all the measures were above .82, except for the Self-deceptive enhancement (SDE) scale of the Balanced Inventory of Desirable Responding (BIDR), which was .73. According to Pallant (2010), acceptable Cronbach Alpha coefficients are above .7. 69 Table 2 Intercorrelations for Miville-Guzman Universality-Diversity Scale –Short (MGUDS), Anger Discomfort Scale (ADS), Balanced Inventory of Desirable Responding (BIDR6) - Self Deceptive Enhancement (SDE) and Impression Management (IM) scales, State- Trait Anxiety Inventory for Adults (STAI) - Trait and State scales, Countertransference Index Scale (CT- Scale) – assessing influence of countertransference in behavior (B), thoughts (T), and feelings (F), Avoidance Index, and Racial Non-Match. Variables 1 2 3 4 5 6 7 8 1. MGUDS 2. ADS .04 3. BIDR6-SDE .00 -.20 4. BIDR6-IM -.02 -.04 .37** 5. STAI-T -.27* .36** -.58** -.14 6. STAI-S -.06 .43** -.23 -.03 .38** 7. CT-Scale .02 .05 -.29* -.28* .32* .37** 8. Avoidance Index .17 -.08 .13 .25 -.20 -.16 -.22 9. Racial Non-Match -.19 -.16 .07 -.10 .06 -.11 .06 -.05 70 Note. MGUDS = Miville-Guzman Universality-Diversity Scale –Short (Pre-Lab); ADS = Anger Discomfort Scale (Pre-Lab); BIDR6- SDE = The Balanced Inventory of Desirable Responding - Self Deceptive Enhancement (SDE) scale (Pre-Lab); BIDR6-IM = The Balanced Inventory of Desirable Responding - Impression Management (IM) scale (Pre-Lab); STAI-T = The State-Trait Anxiety Inventory for Adults – Trait Scale; STAI-S = The State-Trait Anxiety Inventory for Adults – State Scale; CT- Scale = Scale created by combining the Countertransference Index assessing behaviors, the Countertransference Index assessing thoughts, and the Countertransference Index assessing feelings; Racial Non-Match = Refers to whether the client watched was the Black/African American client (coded as 1) or the White/European American Client (coded as 0). *p < 0.05, two-tailed. ** p < 0.01, two-tailed. 71 Table 3 Means and standard deviations of the variables of interest per experimental condition Variable Black/Afr American White/Eur American of Interest Client (N=32) Client (N=31) M SD M SD UDO 72.1 6.8 75 8.2 Anger Discomfort 25.6 5.7 27.5 6.5 BIDR6_SDE 6.3 3.0 5.9 3.5 BIDR6_IM 6.4 3.9 7.2 4.2 Trait Anxiety 1.8 .35 1.8 .32 Countertransference Measures State Anxiety 2.0 .52 2.2 .54 CT-I Behavior 2.6 1.1 2.4 .92 CT-I Thoughts 3.0 1.0 2.8 .95 CT-I Feelings 3.0 1.1 3.0 1.1 Avoidance Index .35 .28 .34 .27 Note. M = Mean, SD = Standard Deviation, UDO = Universal-Diverse orientation; CT-I = Countertransference Index; BIDR6-SDE = The Balanced Inventory of Desirable Responding - Self Deceptive Enhancement (SDE) scale (Pre-Lab); BIDR6- IM = The Balanced Inventory of Desirable Responding - Impression Management (IM) scale (Pre-Lab); 72 Most of the means of the variables of interest for the overall sample were similar to the means found in previous studies. The average score for UDO (assessed by the MGUDS) was 73.5 (SD = 7.6), higher than the one found by Harbin (2004) (UDO M = 70.9, SD = 8.9). Harbin’s sample only included therapist trainees, whereas the current study included trainees and therapists working in the field. Regarding anger discomfort (assessed by the ADS), the sample mean of the present study was 25.5 (SD = 6.2), lower to what Sharkin & Gelso (1993) found in their study: (specifically, men M = 30, SD = 7.3, and women M = 30.5, SD = 4.5). Sharkin and Gelso’s study only included therapist trainees. Considering the dependent variables, the mean for the State Anxiety measure for the current study (i.e., M = 2.1, SD = .53) was similar to the one in Hummel’s (2013) sample (M = 2.0, SD = .49) yet lower than the one found in Harbin’s (2004) sample (M = 4.0, SD = 3.0). Both Hummel’s and Harbin’s sample only included therapists in training. The average verbal avoidance (i.e., Avoidance Index) for this study (i.e., M = .35, SD = .27) was similar to the one in Hummel’s (2013) work (M = .38, SD =. 21) and Harbin’s (2004) study (who reports M = 33.5, SD =36.7; however, transforming these values into a ratio would be M = .34, SD = 0.37), yet lower than the one found by Brittan’s (1984) work (M = .51, SD =. 32). In order to further describe the data, some relations of interest within each of the experimental conditions (i.e., White/European American client v/s Black/African American client) were further investigated. The only significant relationship in the data was between anger discomfort and state anxiety. These two variables were positively correlated buy only in the White/European American client condition (r(29) = .50, p=.004 see table 4). 73 Table 4 Correlations between three different measures of countertransference (i.e., State Anxiety, Avoidance Index and Countertransference Index –CT Index) with Universal-Diverse Orientation (UDO) and Anger Discomfort (AD) for the two experimental conditions (i.e., Black/African American Client and White/European American Client) Relationship Black/Afr American White/Eur American of Interest Client Client (N=32) (N=31) UDO & State Anxiety -.04 -.12 UDO & Avoidance Ratio .06 .31 UDO & CT Index (behavior) -.07 -.06 AD & State Anxiety .32 .50** AD & Avoidance Ratio -.21 -.09 AD & CT Index (behavior) .17 .05 ** p < .01. 74 Additionally, the variables of interest were correlated with some of the quantitative demographic variables. Participants’ age correlated negatively with the countertransference index both at the behavioral level (r(61) = -.31, p = .015) and at the thought level (r(61) = -.26, p = .037). Additionally, years doing therapy was negatively related with anger discomfort (r(61) = -.34, p = .007) and with state anxiety (r(61) = -.29, p = .023). The correlation between age and years providing therapy was r(61) = .74, p = .000). A measure of social desirability had been included to control for social desirability when responding questions about universal-diverse orientation. Correlations between these variables were non-significant, therefore social desirability was not further included in the analyses. Preliminary Analyses For the current experiment, participants were randomly assigned to the one of the two experimental conditions (White angry client vs. Black angry client). Though no initial disparities were expected between the groups, independent sample t-tests where performed to determine if the demographic variables were different between the two experimental conditions. No significant differences were detected between these two groups. Specifically, there was no significant difference in scores for Black client condition (M = 36.44, SD = 13.78) and White client condition (M = 39.52, SD = 14.91; t (61) = -.852, p = .398) regarding age. The independent sample t-test conducted to compare Black client condition (M =1.88, SD =.34) and White client condition regarding gender (M = 1.81, SD = .40; t (61) = .736, p = .465) showed that there were no significant differences regarding gender between these two groups. Also, there was no significant 75 difference in scores for Black client condition (M = 9.69, SD = 11.58) and White client condition (M = 9.64, SD = 9.08; t (61) = .020, p = .984) regarding years providing therapy. Test for Normality The first step in the data analysis was to test for normality. The Normal Q-Q plots and the normality tests showed the violation of the normality assumption (such as skewness and kurtosis) in several scales. Therefore, based on Tabachnick and Fidell’s (2007) recommendations, square root or logarithmic transformations were applied to the scales that presented a skewness coefficient greater than one (in terms of absolute value). After transformation all but the countertransference measures presented non-significant values (i.e., did not violate normality assumptions) in the Shapiro-Wilks test. Tabachnick and Fidell mention that “non-normal kurtosis produces an underestimate of the variance of a variable” (p. 79), however, “...if all the variables are skewed to about the same moderate extent, improvement of analysis with transformation are often marginal” (Tabachnick and Fidell, 2007, p. 87). Because previous studies that have assessed countertransference in the same way (e.g., avoidance ratio, countertransference index) have not reported such transformations, and the countertransference variables were skewed in a similar way, we decided to run the analysis with the original variables. Additionally, even though there were outliers present in the data for some variables, the decision was made to keep them in the dataset. The rationale for such choice was that regarding countertransference, outliers embody the exact phenomenon that this study sought to assess (e.g., cases in which countertransference is too high or when countertransference management fails). Furthermore, histograms revealed different 76 skewness patterns for the countertransference index measures. For example, the countertransference index assessing behavior was positively skewed; however, the one related to feelings was negatively skewed (more participants identified/recognized having feelings of countertransference towards the portrayed client than having behavioral countertransference). Hypotheses Testing The first set of hypotheses stated that countertransference (i.e., a) state anxiety, b) avoidance index, and c) three countertransference indices - extent to which participants’ behavior, thoughts, and feelings were influenced by countertransference) would be greater for European American therapists responding to an angry Black/African American client than to an angry White/European American client. According to the second group of hypotheses, participants’ UDO, anger discomfort and experimental condition (i.e., clients’ race) would jointly predict participants’ countertransference (as a) state anxiety, b) avoidance index, and c) three countertransference indices - extent to which participants’ behavior, thoughts, and feelings were influenced by countertransference). A third set of hypotheses stated that each of the independent variables (i.e, participants’ UDO, participants’ anger discomfort, and experimental condition – clients’ race) would uniquely account for variance in each of the different countertransference measures. A final group of hypothesis posited interactions 1) between UDO and clients’ race in predicting countertransference, and 2) between anger discomfort and clients’ race in predicting countertransference. Standard multiple regressions were conducted to assess the first set of hypotheses, the second group of hypotheses, and the third group of hypotheses. When the dependent 77 variable was state anxiety, a hierarchical regression was conducted (which allowed to covariate trait anxiety). To ensure that there was no violation of the regression test’s assumptions, preliminary analyses were run. As it has been mentioned before, because of the nature of the countertransference phenomenon a non-normal distribution of the dependent variables was expected. The regression tables can be found in Appendix L. To test the hypotheses regarding interaction effects (i.e., 4 and 5), hierarchical multiple regression analyses were conducted. In total, three hierarchical multiple regression analyses were performed, one for each dependent variable (i.e., state anxiety, avoidance behaviors, self-reported countertransference). Hierarchical regression was chosen because it allowed the examination of moderation hypotheses. Additionally, with this procedure, trait anxiety could also be entered in the first block of the hierarchical regression when predicting state anxiety, and therefore statistically control for it (Pallant, 2010). Preliminary analyses were run to ensure that there was no violation of the test’s assumptions. For the hierarchical regression analyses including interaction terms, the guidelines to test moderator effects presented by Frazier, Tix, and Barron (2004) were followed. For this, the categorical independent variable was coded based on a dummy coding system (Racial match –White/European American client=0, Non-match – Black/African American client=1). Also, the status variables (i.e., UDO and anger discomfort) were centered, and two product terms were created: Clients’ race X UDO, and clients’ race X anger discomfort. The hierarchical multiple regression analyses were as follows: Clients’ race was entered in the first block, UDO was entered in the second block, anger discomfort was 78 entered in the third block, and the forth block contained the two product terms. Additionally, when the dependent variable was state anxiety, trait anxiety (also centered) was entered alone as a covariate in the first step, and then the rest of the equation followed the same steps as when the analyses considered the other dependent variables. The regression tables can be found in Appendix L. Considering the result of all the hierarchical regression analyses, it can be stated that the first set of hypotheses (i.e., therapists’ countertransference would be higher when responding to a Black/African American client than to a White/European American client) were not supported, as clients’ race did not significantly predict any of the countertransference measures. The second set of hypotheses stated that the independent and status variables would jointly predict each of the dependent variables. As previously noted, standard multiple regression analysis (hierarchical when considering state anxiety as the dependent variable) was used to assess the ability of therapists’ universal-diverse orientation, their anger discomfort, and clients’ race to predict therapists’ state anxiety, after controlling for the influence of trait anxiety. The total variance explained by the model as a whole was 25.5%, (F (4,58) = 4.945, p = .002). On the contrary, universal-diverse orientation of the therapists, their anger discomfort, and clients’ race did not predict behavioral countertransference (i.e., assessed by the avoidance index) to an angry client (F (3,59) = .768, p = .52). Similarly, therapists’ universal-diverse orientation, therapists’ anger proneness, and clients’ race did not predict self-reported countertransference index scale (at the behavioral, thought, and feeling level combined) (F(3,59) = .182, p = .908). 79 According to the third group of hypotheses, each of the independent variables would uniquely account for variance in each of the dependent variables. White/European American therapists’ UDO did not uniquely account for variance in any of the dependent variables (i.e., state anxiety, avoidance index, nor the combination of the three one-item self-reported countertransference). On the contrary, White/European American therapists’ anger discomfort made a significant unique contribution, but only to the prediction of therapists’ self-reported state anxiety. Specifically, as previously stated, when analyzing the predictive value of the three independent variables (i.e., UDO, anger discomfort and clients’ race) on state anxiety, results showed that the overall model was significant (F (4,58) = 4.945, p = .002), yet only anger discomfort uniquely accounted for variance in state anxiety (B = .31, p = .010). Finally, clients’ race did not uniquely account for variance in any of the dependent variables (i.e., state anxiety, avoidance index, nor the combination of the three one-item self-reported countertransference). The last two hypotheses presented were related to interaction. The first one posited an interaction between clients' race and UDO in the dependent variables, and the second theorized an interaction between clients' race and anger discomfort in the dependent variables. Hierarchical regression analyses showed that none of the interaction terms were significant; therefore, neither of these two hypotheses were supported (see Appendix L). 80 Chapter 6: Discussion The present counseling analogue study was directed at further exploring what affects countertransference in therapy work. Specifically, the investigation focused on the effect that an angry client’s race, therapists’ self-reported universal-diverse orientation, and their self-reported anger discomfort had on several measures of countertransference (both self-reported and assessed by coders). Descriptive correlations among the variables of interests showed that neither the experimental condition (i.e., White/European American client – racial match, Black/African American client – non-match) nor universal-diverse orientation was significantly related to the countertransference measures (i.e., state anxiety, avoidance ratio, and countertransference index scale assessing the level which countertransference reactions to the client affected the therapist’s behaviors, thoughts or feelings). Participant’s anger discomfort was only related to the state anxiety measure, which is consistent with Sharkin and Gelso’s (1993) finding, yet different to Harbin’s (2004) result (i.e., anger discomfort was not correlated to state anxiety). Additionally, hierarchical regression analyses showed that of all the hypothesized relations, only one was statistically significant: White/European American therapists’ UDO level, their anger discomfort, and the client’s race (plus trait anxiety as covariate) jointly predicted therapists’ state anxiety as manifestation of affective countertransference to an angry client. From all the independent variables, only anger discomfort uniquely accounted for variance in state anxiety. 81 One might look at these results from a hopeful perspective, and consider that perhaps the lack of effects is related to therapists’ lack of racial biases. One might even think that the helping professions have evolved, due to increased multicultural and diversity knowledge and training, and that this has positively affected therapists’ stereotypes and prejudices; therefore, therapists currently are not biased, and their emotional reactions are either not triggered or if they are, they are well managed (either at an unconscious level or consciously, thus low report). Furthermore, Smith and Trimble (2016) highlight that therapist-client match might have been more of a concern in previous years than now, as their meta-analysis showed clients-therapists’ racial match impact on clients’ therapy participation has decrease over time. However, a quick review of some events in the US in recent years (e.g., cases of Freddy Gray, Philando Castile, among many others) shows us that racial issues have not changed. Furthermore, “… although self-reported prejudice has reduced dramatically in the past 60 years, discrimination evidence has not decreased accordingly (Dovidio, Brigham, Johnson, & Gaertner, 1996, in Katz and Hoyt, 2014, p. 300). And the helping professions are not an exception. Several studies have shown physicians’ discrimination in the diagnosis and/or treatment of clients based on the physicians’ racial prejudice (e.g., Blair et al., 2013; Stepanikova, 2012). Furthermore, there is some empirical evidence that as therapists we are a part of the broader society, and although explicit racial bias seems to be less frequent, implicit bias is still very present. For example, Kugelmass (2016) conducted a phone based field experiment that looked at the relations between some clients’ demographics (i.e., gender, class and race), and therapists’ response in making an appointment with such clients. Over 300 licensed Ph.D. or PsyD therapists practicing in 82 New York City were contacted through voicemail messages with a client requesting an appointment. The voicemail calls received were from “…one black middle-class and one white middle-class help seeker, or from one black working-class and one white working- class help seeker” (Kugelmass, 2016, p. 168). Among the findings, the author reported a class main effect and a race X class interaction: Potential clients that seemed to be middle class received almost three times more appointment offers than those who seemed from a working-class, and in the middle-class condition, participants that seemed Black received less offers of appointments than those who seemed White. With evidence of racial biases in other studies, what might have influenced the statistical non-significance in the present work? First, perhaps therapists’ prejudices and biases play a role in determining providing appointments (Kugelmass, 2016), and in the expectations of treatment or bond ratings (Katz and Hoyt, 2014), yet in the moment-to-moment communication that happens in therapy work, these might not get necessarily activated. Also, in Kugelmass’s (2016) study, “(v)oice-over artists recorded scripted messages using racially distinctive names and adopting specified race- and class-based speech patterns” (p. 172). Perhaps the effects related to clients’ race might be better captured when a client’s race is represented by characteristics that are socially associated to such race. On the contrary, in the current study, in order to make the two experimental conditions identical, several characteristics of the clients (e.g., names, context, expression, clothing) were kept equivalent. As previously noted, communication styles might be very different between Black and White people (Sue & Sue, 2008). Therefore, by trying to keep both experimental conditions comparable, the present study might have somehow washed out any effects by trying to 83 make communication styles comparable instead of distinguishable and thus, perhaps more triggering for participants. The non-significant results related to race could have also been associated to the sample. More specifically, perhaps there are characteristics of the participants that influenced the results. Participants completed online measures, and approximately one week after, attended a Lab session. Thus, these are therapists who knowingly chose to participate and be exposed to a lab experience that would “examine(d) therapists’ characteristics and feelings that influence the reactions to clients”. Perhaps those who volunteered and completed the two parts of the study were therapists who are open to experience, which is a variable that has been positively related to UDO (e.g., Thompson, Brossart, Carlozzi, & Miville, 2002). Maybe therapists and trainees who tend to get anxious in new experiences refrained from participating. That might have created a ceiling effect and thus not much variance. Anecdotally, several therapists in the field made post-lab comments such as “that was so interesting”, “that was fun”. Additionally, the fact that we could not empirically demonstrate a relation between clients’ race and countertransference issues does not mean the phenomenon does not exist. This non-significance might be related to the way countertransference was assessed. Participants were asked to self-report countertransference in the form of completing a state anxiety measure and three one-item questions. The nature of the countertransference experience might make it very difficult for participants to accurately report it. Additionally, assessment of countertransference also was in the form of calculating an avoidance index based on the verbal responses of the participants. Current clinical training in the helping fields might involve specific skills training. Additionally, 84 supervisors might help trainees hone in what to say at specific moments (e.g., when a client gets angry). However, knowing what to say might be “masking” the fact that the participant might have intense reactions to the client, and such reactions will not be captured by the countertransference measure. Finally, countertransference regarding racial issues might manifest in subtle behaviors and reactions (e.g., a therapist slightly moving in his chair, or looking away from the client) that the measures used in this study might have not captured. In a different vein, the only model tested by the hierarchical regressions that was significant was the one that included White/European American therapists’ anger discomfort, the therapists’ UDO level, and the client’s race (plus trait anxiety as covariate) predicting therapists’ state anxiety. Only anger discomfort uniquely accounted for variance in state anxiety. Previous empirical work had shown mixed results regarding the relationship between anger discomfort and state anxiety: Sharkin and Gelso (1993) found a significant result, yet Harbin (2004) did not (and therefore did not include anger discomfort as a covariate in his study). Additionally, this brings to the table the countertransference interaction hypothesis, introduced by Gelso and Hayes (2007). According to these authors, “… we must nearly always look at how client attributes and behaviors interact with the therapist’s vulnerabilities and issues if we are to understand the triggers for countertransference” (p. 44). It seems that therapists’ anger discomfort might play an important role regarding experiencing state anxiety when facing an angry client. Thus, it would be valuable to include anger discomfort in further studies that look at therapists’ countertransference (particularly state anxiety) responses to an angry client, to further understand the relationship between these variables. 85 Finally, universal-diverse orientation did not significantly relate to any of the variables of interest. This could be connected to the nature of the variable, which according to Miville et al.’s (1999) is “…an attitude of awareness and acceptance of both the similarities and differences that exist among people.” (p. 292). Perhaps if the therapist is not purposely focused on appreciating the similarities and valuing the differences among people (and thus possibly activating in-group and out-group biases), his or her own struggles related to the differences will not get activated in session. Limitations This study has several limitations. First, analogue procedures were used for this investigation. As previously presented, using an analogue client creates limitations in terms of generalization of the results (i.e., external validity issues). Heppner et al. (2008) suggested that in analogue research, looking at how similar are the analogue variables to those in real-life can assess external validity. Thus, in line with previous analogue studies, resemblance with a counseling situation was attempted by: 1) Using analogue clients that were believable in their client role, 2) Recording the videotape in a space that closely resembled a therapist’s office, 3) The client looked directly at the camera, and in a distance with the camera that closely resembled the distance in a therapy session. Furthermore and contrary to other analogue studies, there was no researcher in the room with the therapist while she or he watched the video. Thus, the previous considerations were a way of simulating a counseling session within the limitations of an analogue procedure. Additionally, some of the ratings of the actors (e.g., measures of attractiveness of the actors that played the clients) were lower than those found in other studies, which 86 raises the question of whether that might have influenced the results. First, even though the actors’ ratings were lower than other studies, these ratings were balanced by conditions. Also, if actors in other analogue studies were deemed as highly attractive (e.g., 4 or more in a 5 point scale) it raises the issue of a possible confound, where participants might have been reacting to attractiveness and not only the assessed variable. Also, as previously mentioned the sample recruited might not be representative of therapists in general, and might have affected the results. First, the sample included therapists in training. Second, therapists came from a restricted geographic area (mostly Mid-Atlantic part of the US). Finally, some of the measures (e.g., M-GUDS-S, State Anxiety) were validated mainly in samples consisting of College students, which was not the targeted sample. To address this limitation, the obtained means were compared to those found in other studies pertaining the variables of interest but that had a sample consisting of therapists or therapists in training (e.g., Harbin, 2004). The means found in this study were mostly similar to the means from those other studies. Future Directions Analogue studies can be a valuable way to address and assess clinical issues that might be very difficult to look at otherwise (such as clients’ anger). One of the relevant characteristics of analogue studies is to keep variables controlled. Thus, for the present study the actors followed the same script and similar expressions. However, as previously stated, Black and White people communicate in a different way (Sue and Sue, 2008), which might be highly relevant to capture in future studies. Therefore, a next step related to the current work could be an analogue study in which White/European American clients and Black/African American clients use the language patterns and non-verbal 87 communication of their specific groups (and following Kugelmass’s results, perhaps social class should also be considered). Another way in which researchers could continue studying the variables of interest and still respect the communication difference between these two racial groups is by conducting field studies. For example, therapists working in the field with a racially diverse client population could complete measures of universal-diverse orientation and anger discomfort at the beginning of treatment, and then assess countertransference after every session. This could allow comparison among different client racial groups in a real- life setting. Such studies could also incorporate the clients’ perspective, by rating the therapists’ impact in them. Such addition would provide an interactive process in which we could have repeated measures over time that capture the ebbs and flows of the phenomena of interest and their effect in the therapy work (e.g., are there moments in therapy in which therapists’ countertransference reactions to a Black/African American client are more at a feelings level versus others that are more at a behavioral level and what is the impact these have on the client, if at all?). Additionally, field studies could also incorporate different components of the therapeutic relationship (as addressed by Gelso and his collaborators), and see how they manifest for each racial/ethnic group. This becomes particularly relevant when thinking of therapists’ reactions to Black/African American clients or to White/European American clients, as it would allow us to see whether there are any patterns related to the variables of interest, the different aspects of the therapeutic relationship, and group membership. Efforts to look at the therapists’ reactions to Black/African American clients vs. White/European American clients are still needed, and are becoming increasingly 88 relevant in the US society. Studies with a larger sample and that include participants from a wider geographical area (as people in some areas could present a more favorable or less favorable view towards and angry Black client, for example) might provide a wider range of responses. Additionally, future studies could include different ways of assessing therapists’ biases that might influence the therapists’ work. In a different vein, future studies could also include other client variables, such as age and gender. This might allow us to further explore the relationship between clients’ race and therapists’ countertransference. As was previously presented, there can be a wide range of countertransference triggers. Additionally, these triggers can vary from person to person, and even within a person (e.g., if a person is feeling particularly vulnerable one day she/he might react to something that perhaps he/she would have not at another time). Therefore, future studies could also look at the specific situations that might trigger racial countertransference reactions. Furthermore, as it has been stated, “… researchers need to examine “who, what, when, and where” questions” (Gelso & Palma, 2011, p. 342). For example, the current study asked therapists to assume they were going to observe a 5th session. Perhaps when racial countertransference issues get triggered, their impact might cause clients to not want to return to therapy. And such pull to drop-out might be stronger for a client when it is the first session versus the 5th session. Therefore, future studies could really consider assessing moment-to-moment client-therapist interaction and different specific race-related situations that might trigger therapists’ countertransference. Additionally, such assessment could include behavioral observation from raters, which might be able to detect specific moment-to-moment changes that might be indicating countertransference. 89 Finally, two new versions of the Countertransference Index were created for this study (one in relation to therapists’ thoughts and one related to therapists’ feelings), which were combined with the original one-item measures and used as a scale. Future research work could include the different versions of the Countertransference Index, to further detect any patterns regarding countertransference manifestation. Furthermore, a fourth CT-Index could be created, directed at assessing the “somatic” level (physiological reactions that at times therapists can be aware of, such as increased heart rate, hand perspiration, etc…) of countertransference. Based on the previous information, it can be stated that perhaps the non- significant effects regarding clients’ race and UDO might be related to the nature of the observed phenomenon (e.g., UDO as accepting and acknowledging similarities and differences), measurement of the phenomenon (e.g., countertransference might have not been captured), and/or limitations inherent to this study’s method (e.g. analogue). Additionally, as only anger discomfort uniquely accounted for variance in state anxiety, it makes us wonder whether the therapists’ reaction to clients’ anger is a variable that seems to be more intense than therapists’ reactions to clients’ race. The current study sought to further understand the influence of Universal-Diverse Orientation and anger discomfort in the countertransference reactions that therapists can have when responding to an angry client. Even though none of the hypothesized relationships among the variables was significant, this study is one added step in the effort of increasing the knowledge of what might and might not affect the therapist’s countertransferential experience. 90 Appendix A Miville-Guzman Universality Diversity Scale – Short Form This test is copyrighted material, and the author does not authorize its publication in a Dissertation. Please contact Marie M. Miville, PhD.(mlm2106@columbia.edu ) for the use of this measure. 91 Appendix B Anger Discomfort Scale Use the scale below to respond to each statement. There are no right and wrong answers. Write in the number that corresponds to your answer for each item. _______________________________________________________________________ 1 2 3 4 1. ____ I do not like it when I get angry. 2. ____ I feel guilty about being angry with others. 3. ____ I fear that my anger will hurt other people. 4. ____ I would prefer that people not see me when I am angry. 5. ____ I believe it is natural and healthy to feel angry. 6. ____ I am troubled by my anger. 7. ____ People do not seem to like me when I am angry. 8. ____ I create more problems for myself when I am angry. 9. ____ I should not be as angry as I often am. 10. ____ I believe it is acceptable for people to feel angry. 11. ____ I feel comfortable with my angry feelings. 12. ____ When I get angry, I also get nervous. 13. ____ My anger scares me. 14.____ I am embarrassed when I get angry. 15.____ I fear losing control because of my anger. 92 Appendix C The Balanced Inventory of Desirable Responding (BIDR Version 6 - Form 40A) Using the scale below as a guide, write a number beside each statement to indicate how true it is. ________________________________________________________________________ _____ 1 2 3 4 5 6 7 not true somewhat very true ____ 1. My first impressions of people usually turn out to be right. ____ 2. It would be hard for me to break any of my bad habits. ____ 3. I don't care to know what other people really think of me. ____ 4. I have not always been honest with myself. ____ 5. I always know why I like things. ____ 6. When my emotions are aroused, it biases my thinking. ____ 7. Once I've made up my mind, other people can seldom change my opinion. ____ 8. I am not a safe driver when I exceed the speed limit. ____ 9. I am fully in control of my own fate. ____ 10. It's hard for me to shut off a disturbing thought. ____ 11. I never regret my decisions. ____ 12. I sometimes lose out on things because I can't make up my mind soon enough. ____ 13. The reason I vote is because my vote can make a difference. ____ 14. My parents were not always fair when they punished me. ____ 15. I am a completely rational person. ____ 16. I rarely appreciate criticism. ____ 17. I am very confident of my judgments 93 ____ 18. I have sometimes doubted my ability as a lover. ____ 19. It's all right with me if some people happen to dislike me. ____ 20. I don't always know the reasons why I do the things I do. ____ 21. I sometimes tell lies if I have to. ____ 22. I never cover up my mistakes. ____ 23. There have been occasions when I have taken advantage of someone. ____ 24. I never swear. ____ 25. I sometimes try to get even rather than forgive and forget. ____ 26. I always obey laws, even if I'm unlikely to get caught. ____ 27. I have said something bad about a friend behind his/her back. ____ 28. When I hear people talking privately, I avoid listening. ____ 29. I have received too much change from a salesperson without telling him or her. ____ 30. I always declare everything at customs. ____ 31. When I was young I sometimes stole things. ____ 32. I have never dropped litter on the street. ____ 33. I sometimes drive faster than the speed limit. ____ 34. I never read sexy books or magazines. ____ 35. I have done things that I don't tell other people about. ____ 36. I never take things that don't belong to me. ____ 37. I have taken sick-leave from work or school even though I wasn't really sick. ____ 38. I have never damaged a library book or store merchandise without reporting it. ____ 39. I have some pretty awful habits. ____ 40. I don't gossip about other people's business. 94 Appendix D State-Trait Anxiety Inventory for Adults A number of statements which people have used to describe themselves are given below. Read each statement and then circle the appropriate number to the right of the statement to indicate how you feel right now, that is, at this moment. There are no right or wrong answers. Do not spend too much time on any one statement but give the answer which seems to describe your present feelings best. 1= Not at all 2= Somewhat 3= Moderately so 4= Very much so 1. I feel calm .................................................................................................. 1 2 3 4 2. I feel secure ................................................................................................ 1 2 3 4 3. I am tense ................................................................................................... 1 2 3 4 © 1968, 1977 Charles D. Spielberger. All rights reserved in all media. Published by Mind Garden, Inc., www.mindgarden.com This test is copyrighted material. The license owner allows the reproduction of a maximum of 5 items in a dissertation. The instructions and items you see above correspond to the State Anxiety Scale of the State-Trait Anxiety Inventory for Adults. For any use of this scale, please contact Mind Garden. 95 Appendix E Countertransference Index (CT Index) Please indicate the extent to which your behavior in session was influenced by countertransference (i.e., areas of unresolved conflicts): 1 2 3 4 5 Strongly Disagree Not Sure Agree Strongly Disagree Agree Please indicate the extent to which your thoughts in session were influenced by countertransference (i.e., areas of unresolved conflicts): 1 2 3 4 5 Strongly Disagree Not Sure Agree Strongly Disagree Agree Please indicate the extent to which your feelings in session were influenced by countertransference (i.e., areas of unresolved conflicts): 1 2 3 4 5 Strongly Disagree Not Sure Agree Strongly Disagree Agree 96 Appendix F Response Mode Categories for Units and Whole Turns I. Approach Responses 1) Approval - Therapist appropriately sanctions, accepts, or supports (including minimal encourages) the client’s feelings or behaviors, and/or - Therapist expresses explicit agreement with the client’s feelings or behaviors. 2) Exploration - Therapist asks for further clarification, elaboration, and detailing of the client’s feelings or behaviors; and/or - Therapist makes suggestions that seem to fit well with the client’s material. 3) Reflection - Therapist repeats or restates the client’s feelings; - Therapist accurately re-labels the client’s feelings, attitudes, or behaviors; and/or - Therapist reflects content when only content is given. 4) Labeling/Interpretation - Therapist points out patterns in the client’s feelings or behaviors. - Therapist suggests relationships between present feelings or behavior and past experiences; and/or - Therapist suggests underlying causes of feelings or behavior. II. Avoidance Responses 5) Disapproval - Therapist is critical of the client’s feelings or behaviors. - Even if the statement is phrased supportively, anything that negates or opposes the client’s feelings is disapproval. 6) Ignoring - Therapist responds to the content of the client’s material but ignores the affect; and/or - Therapist seems to miss the point the client is expressing and instead comes from the therapist’s own agenda or needs. 97 7) Mislabeling - Therapist inaccurately identifies the client’s feelings, attitudes, or behaviors; and/or - Therapist inaccurately identifies the degree of feelings. 8) Topic Transition - Therapist changes the focus of discussion to an irrelevant topic or simply to a different topic. 9) Colluding/Inappropriate approval - Therapist expresses excessive approval of client. - Therapist sides or expresses agreement with client when there is not sufficient warrant for such siding with or agreement. - Therapist shares or possesses client’s feelings without warrant, e.g., becomes too easily upset or sad when the client is upset or sad. 10) Other - Therapist’s response does not fit any of the other categories. - Try to absolutely rule out the other possibilities before choosing this category. (From Hummel, 2013. This coding system includes a few modifications to the original coding system. Such modifications are in italics. ) 98 UNITIZING 1. We will be dividing each participant’s responses into “units” of meaning – that is complete sentences or independent clauses. 2. Put slashes (/) after each “unit” of meaning, or each complete thought. Type the unit # in parenthesis after the slash. a. e.g., I hear that you are angry,/(1) and I would really like to help you with this/.(2) 3. A complete thought has a subject and a verb and can stand on its own. a. For example, “She is going to the beach” is a unit because it has a subject (she), a verb (is going), and can stand on its own. In contrast “that she’s going to the beach” is NOT a unit because although it has a subject and a verb, it can’t stand on it’s own as a complete thought. b. If a participant had 3 different thoughts in one speaking turn, you would have 3 units for the 3 different thoughts (If they rambled you could have then several units in that response!). 4. Independent/Dependent Clauses a. A unit is an independent clause i. An independent clause contains a subject, a verb, and is a complete thought. e.g., I hear that you are angry./ b. A dependent clause contains a subject and a verb, but is not a complete thought. i. Dependent clauses start with subordinating conjunctions (e.g., while, when, because, although) or relative pronouns (who, whose, which, that) e.g.,: that you feel angry 5. Simple sentence a. Has one independent clause, and therefore 1 unit. i. e.g., You are angry at me./ ii. In this example, there is a subject (you), a verb (are), an adjective (angry), and a prepositional phrase (at me). b. Simple sentences can still be long and complicated, even with just one clause. i. You and your boss need to sit down and talk about this with each other./ ii. In the example there is a compound subject (you and your boss), a compound verb (sit down and talk) and prepositional phrases (about this; with each other) 6. Compound Sentence a. Has two independent clauses (and therefore two units: one unit per independent clause). Some examples: i. You are angry,/but you don’t like conflict./ ii. You are frustrated,/ and you want a change./ iii. I hear you,/ and I can help./ iv. You want something;/ your bosses want something different./ 7. Complex Sentence 99 a. Has one (or more) dependent clause(s) (headed by a subordinating conjunction or a relative pronoun) joined to an independent clause. i. It has one unit per independent clause. ii. A dependent clause is not a complete thought, and therefore is not a unit. iii. E.g., When you are angry, you have trouble seeing things clearly./ iv. Dependent clause: When you are angry, v. Independent clause: you have trouble seeing things clearly./ 8. Compound-Complex Sentence a. Has two (or more) independent clauses joined to one or more dependent clauses. i. There is one unit per independent clause. ii. A dependent clause is not a complete thought, and therefore is not a unit. iii. E.g., You feel a lot of anger that is affecting you/, and you want to focus on your relationship problems. / iv. Independent clauses: You feel a lot of anger/, and you want to focus on your relationship problems. / v. Dependent clause: that is affecting you 9. Other unit rules: a. False starts (e.g., Well it sounds like…..What was that like for you?) do not count as separate units. The example cited would count as one unit. b. Minimal enouragers (e.g., “Mmm hmm…”) do not count as a separate unit. c. Single word sentences like “Good” or “What?” do count as a separate unit. 10. For more information, you can check: a. http://www.cws.illinois.edu/workshop/writers/clauses/ b. https://owl.english.purdue.edu/owl/resource/598/01/ 11. Add the unitized transcript to the folder in the UMD Box called “Unitizing” 12. Be sure to check the “Unitizing Summary” in that same folder to check the due dates for the unitizing. The Information presented has been taken from UMD’s PSYC433 (Helping Skills course) Syllabus/Packet used in previous years and Ann Hummel’s dissertation. Some examples have been modified to be better suited for the present study. 100 Coding Training Session Overview o We have a lot of therapist responses, and we need to turn this information into meaningful and workable data. Therefore, we will be coding the participant’s responses in terms of approach (i.e., “…verbal responses that were primarily designed to elicit from the patient further expressions of (hostile) feelings, attitudes, and behavior.” Bandura, Lipsher, & Miller, 1960, p. 2) and avoidance (i.e., “…verbal responses designed to inhibit, discourage, or divert the patients’ (hostile) expressions.” Bandura, Lipsher, & Miller, 1960, p. 3) o We will code at a unit level (i.e., a specific sentence) and at a turn level (i.e., each predetermined speaking turn. There were 6 turns). How will we do this? 1. All coders will rate all the transcripts. 2. Data a. Transcripts • To access the revised and unitized transcripts go to a folder called “Coding” in the UMD Box. There, go to a subfolder called “Unitized Transcripts”. • You can access the box at: https://umd.app.box.com/login b. Coder’s File • Each coder has a folder with his/her name inside the “Coding folder”. • Inside the folder, there is a Spreadsheet where you will enter the data. There is also a Coding Log. Coders should write there about any concern, challenge, difficulties that they encounter in the coding process and/or any questions they might have. When entering a comment in the log, please write the date, participant’s code, turn and unit number (if applicable). c. Spreadsheet • The coders will enter the codes they assigned (per unit and per turn - more later) in an excel spreadsheet. • Each coder will have his/her own spreadsheet, located in their specific folder in the UMD Box. The file name will be “CodingYourInitialsMMDDYY”. Each spreadsheet can be found in the specific coder’s folder. • To enter data, first download the file to your computer. Next, update the date, and enter the data to the updated file. Finally, upload the latest version of the file to your folder in UMD Box. Please check that you upload the most recent file. • Keep a copy of the files in your computer. • If the speaking turn has less units than the ones that appear in the spreadsheet, please enter a “0” for those units. • Once you enter information in the spreadsheet be sure to save a copy in your computer and to upload the file to your folder in the UMD Box. Be sure you saved the last version of the file and that it was successfully loaded to the UMD Box. 3. Coding Procedure a. Two important things before starting • Be sure to code a transcript in “one-sit”, that is, once you start it you continue until you finish it. This, because there might be responses and 101 references to a previous speaking turn that you might miss if you look at things separately. The idea is to consider the transcript as a whole. • Participants first read some background information about the client, and were told that they have met with the client for 4 sessions (going to the fifth). Therefore, some participants might reference information beyond what was given in that specific turn, or might assume things as if they have had previous sessions with the client (e.g., say something like “in our last session you talked about haw difficult it was for you to face your boss”). The steps for coding are developed having this in mind, and must be followed for each script. b. First, download your Coding Spreadsheet to your computer. Update the date (MMDDYY). c. Coding goals i. We are ultimately interested in splitting up the responses into either approach or avoidance. Look at your Response Mode Categories sheet. Categories 1- 4 are approach, and categories 5-9 are avoidance. ii. When therapists avoid client material, it is indicative of countertransference issues going on with the therapists. Read over the different categories that we’re talking about to get more familiar with them. iii. When we talk about approach responses, we are looking for responses that are mostly accurate. These responses should be appropriate from a particular theory (e.g., behavioral, humanistic, psychodynamic). Please try to partial out your own theoretical bias since the responses will be coming from a wide range of backgrounds, programs, and theories. d. Coding Procedure i. Read over the client’s background information that participants reviewed. Remember participants might reference or suggest things based on the client’s background and/or the four previous sessions (as the clips they are watching would be from the “fifth session”). § If there is evidence that the reference or suggestion is related to the assumed previous sessions or background information, it would probably be coded as 2 (exploration). § If a suggestion or reference seems to come out of left field and doesn’t seem to fit with the flow of material, it might be coded as 5 (disapproval) or 6 (ignoring). ii. Read the client’s speaking turn that corresponds to the therapist’s speaking turn prior to coding. § For example, in the first speaking turn, the client is talking about being “pissed” about not being promoted for a job, job for which they are looking for someone to take over, while he has been doing all the work. If the therapist doesn’t acknowledge the client's feelings about not being promoted and/or doing the job for which he is not being recognized, at least one therapist response would be coded as a 7 (ignoring) since they are ignoring affect and a major issue that the client has brought up. Additionally, it's important for the feeling that the therapist reflects matches the feeling(s) that the client is experiencing (e.g., reflection (3) vs. mislabeling (7)). iii. For each speaking turn, we will code (a) all of the units within the speaking turn, and (b) the entire speaking turn (i.e., one overall code). 102 iv. Unit Coding § First read the entire speaking turn (e.g., read all the units in T1). For instance, if there are 3 units within the first speaking turn (T1), make sure to read all 3 units before beginning to rate the first one. The idea is to listen to the music rather than the individual notes. Also, most ambiguous responses are much easier to code in the context of the entire unit. § Also, listen to the audio recording, which is available in the UMD Box in a folder called “Audio Responses”. You will see there is a folder created for each participant. § Assign each unit to a response mode category. Use the number assigned the response categories, e.g., approval would be entered as "1." § Please do not spend time debating between categories within approach or avoidance. In other words, if you are torn between choosing one of two approach responses for an individual unit, just choose one without much debate. It is very tempting to think too much about this! However, the distinction between the two general categories of approach vs. avoidance is all that matters in the end. § If a therapist has fewer units in a speaking turn than are listed on the spreadsheet (which is likely) type 0. If a therapist has more units, add another column(s) to the spreadsheet where the additional unit(s) should be. v. Turn Coding § Assign a response category to the entire speaking turn, using the number assigned to the response categories. If needed, read again the entire speaking turn. § The response category for the overall speaking turn might reflect the categories of the units, but sometimes the speaking turn might have a different feel to it when the entire turn is taken as a whole. § A speaking turn might have several units that are avoidant, yet the overall impression of the speaking turn is not avoidant. § Or, few or none of the units might be avoidant, but the speaking turn as a whole may come across as avoidant, for example disapproval or ignoring client affect. vi. To consider when assigning codes § It’s important to note that 5 (disapproval) can be very subtle. Wording/phrasing of the therapist’s response can make the difference between an approach response vs. an avoidance response. For example, if the therapist is dismissive of the clients concerns, this could be subtle disapproval. § Also, remember that the analogue situation was artificial and the therapists may have been nervous about being audio-taped, so remember that “bad responses” don’t necessary equal avoidance or counter-transference. The responses might not be great, but are not 103 necessarily avoidant. Approach doesn’t have to mean good – just somewhat accurate. vii. If you had any questions, doubts, concerns or comments regarding the specific codes you assigned, please write this is the coding Log that you will find in your folder in the UMD Box. viii. After you completed all the coding (i.e., all units and overall speaking turns) for a participant, answer these three questions in your database, with “Yes” or “No”: § Where there any silences? (If yes, write turn that it occurred.) § Did the therapist acknowledge that the anger was directed towards him/her? (If yes, write turn that it occurred.) § Did the therapist acknowledge race differences? (If yes, write turn that it occurred.) ix. Once you are done coding the transcript of a participant, upload the Coding file to your folder in the UMD Box. Remember to check that you changed the date to the spreadsheet, and that you are uploading the most recent file! The Information presented was based on Harbin’s (2004) and Hummel’s (2013) work. 104 Appendix G Demographic Questionnaire for Therapist 1. What is your age?______ 2. What gender do you self-identify as? _____ Female _____ Male _____Transgender _____Other (Please specify) 3. What is your current level of education? ___ BA/BS ___ MA/MS ___ M.Ed. ___ MSW ___ Ed.D. ___ Ph.D. ___ PsyD. ___ Other (specify) ____________________________ 4. Are you currently in school to obtain such a degree? _____ If yes, please specify which year are you in your program ____ 5. What type of program are you in or did you graduate from? ____________ ___ Counseling Psychology ___ Clinical Psychology ___ School Psychology ___ Social Work ___ Psychiatry ___ Counselor Education ___ Other (please specify) _____________ 6. What is your race? (please indicate one or more to which you self-identify): ___ White/Caucasian ___ American Indian or Alaska Native ___ Black or African American ___ Native Hawaiian or Other Pacific Islander ___ Asian ___ Multiethnic (please specify) ___________ ___ Other (please specify) ________ 7. Are you of Hispanic or Latino origin? ____ Yes ____ No 8. Please specify your ethnic background(s) and/or national origin(s) (e.g., Italian American, Jewish, Persian, Italian American, Indian, Korean American, etc…) _____________________________ 9. Were you born in the US? ____ Yes ____ No 105 10. Are you a US citizen ____ Yes ____ No 11. Are you an International Student ____ Yes ____ No 12. In terms of sexual orientation labeling, you consider yourself: ____ Heterosexual or Straight ____ Gay or Lesbian ____ Bisexual ____ Questioning ____ Queer 13. 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 ___ Cognitive/Behavioral ___ Systems ___ Other 10. If you answered “other” in relation to theoretical approach, please specify ______ 11. Years providing therapy (your best estimate) _______________________ 12. Are you a licensed Mental Health Professional? ____ Yes ____ No 106 Appendix H Client Case Presentation Client Name: Jason Williams Age: 37 Gender: Male Presenting Problem: Increased stress and frustration related to work. Background information Jason is a 37-year-old -- European American/African American (changed based on the video) --heterosexual male. He works at an Insurance Company in the DC area. Jason sought your counseling services due to the increased stress related to his current job. This is the first time Jason has been in therapy. Jason has been working for the past three years as an assistant for the insurance claim project manager at an insurance company. A year ago, his boss resigned, and Jason had to take over the leadership of the Insurance Claim division. The company has been looking for a replacement for his boss’s position, to no avail. Jason’s current work focuses mostly on overseeing and managing the restoration of a client’s home/office after a flood, fire, or other disasters (e.g., earthquake). Jason mentions feeling increasingly “stuck”, frustrated, and disappointed about his job. Jason has been married for the past 5 years to Angela (33). Angela is an English teacher, and is currently working at the high school close to their home. Jason describes their marriage as ”good”, sharing that they are supportive of each other. He also mentioned that in the past year there has been more tension between him and his wife, and the two of them have been fighting more often. Jason attributes their increased fighting to his challenges at work. He also wonders if “things change when you have been married for 5 years”. Angela is currently 5 months pregnant with their first child. Jason is the eldest of three children: He has a brother who is two years younger, and a sister who is five years younger. Jason’s parents have been married for almost 40 years. Jason describes his relationship with his father as “cordial”, and with his mother as “close”. Jason’s parents and siblings also live in the DC area, and they all see each other regularly. Jason’s in-laws live on the West coast. Jason volunteers once a month in a tutoring service at the local high school. Additionally, he identifies running as a good way for him to “blow off some steam”. Assume you have seen Jason for four sessions prior to the one you are about to have. Assume you are alone with Jason. 107 Script Segment 1 Yeah… I’m so frustrated with my job… I work hard, and I get nothing! … It’s almost as if it’s not worth trying…. They still haven’t promoted me…. (Frustrated) And it’s irritating that they are searching for someone to be the Project Manager when I’ve been doing the job all this time, only to get passed over like this! (Raising voice slightly) I don’t think my boss sees how much work my team and I do to keep the division going and I’m so pissed! Segment 2 I hate having to work for these people! I’ve been leading this program for a year, why don’t they trust me (emphasis) to REALLY be in charge? I know it is not an age issue… Mike was younger than me when he became Project Manager for the company... I’m not going to fight with them anymore. (Intense) The hell with them all! Segment 3 This is so frustrating!… What will it take for others to see what I’m capable of? … Damn, at times I think that if you don’t fit the mold… What would it be like if all the players in the Football field play the same position? (Angry) They are so narrow-minded! (Angry) I am so god damn tired of this! Segment 4 Dammit… I’ve had it!!… You know, the two times I’ve made a small mistake, I immediately heard about it… …But when I do a great job, no one gives a damn! (Intense) WHAT THE HELL do they expect? It sucks!!… You hear me? What do you think? Segment 5 I… I don’t know if you get what I am saying… Do you get it? … ‘Cause you just stare at me, and don’t say anything… and I need some guidance!! … I don’t know how to change things… Damn, I mean, even here, I come and try my best to do what I have to do, to tell you what’s wrong (frustrated) and you’re not telling me shit! Segment 6 You know, it’s fucking frustrating! You say something here or there, (intense) but you never really give me anything! What the hell? I know I am the one that has to do the work, but, come on, (Emphasize) aren’t you going to say something?? 108 Appendix I First Email for Therapists and Trainees That I Don’t Know Dear ___________, _____ referred me to you as someone who might be interested in participating in my dissertation study. My name is Beatriz (Bea) Palma, and I am a doctoral candidate in the Counseling Psychology PhD program at the University of Maryland, College Park, working under the supervision of my advisor, Dr. Charles J. 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 therapists’ characteristics and feelings that influence the reactions to clients. Given your clinical experience and training, we would really appreciate if you would contribute your time and perspective by participating in my study. Also, we think this is likely to be an interesting experience that might facilitate your reflection on clinical work. The study involves two parts. If you agree to participate, you will first need to complete some brief questionnaires online (the link is provided below). This should take about 10 – 15 minutes. Then, we will ask you to come by a lab at the University of Maryland for thirty minutes to engage in a therapy related experience. If the location does not work for you, we can also meet you at a more convenient location (e.g., your office) and run the computer tasks on our laptop. Participants who complete both portions of the study will automatically be entered in a raffle of two Amazon gift cards - $100 each. We would very much appreciate your participation in this study. We are aware that your time is extremely important, but believe that the nature of this research will make your participation worthwhile. Previous participants have stated that partaking in this study has been very interesting. Although we do not provide individual feedback, participants have described their experience as an opportunity to reflect about therapy and to ponder about one’s work. Additionally, Dr. Gelso and I hope this study will further illuminate therapists’ reactions to clients and also help us for future therapists’ training. That is why we are hoping to get a sample as large as possible. Finally, if you are interested, we can send you a summary of our findings and can notify you of any publications that result from this study. Also, feel free to pass this email along to anyone in the DC/MD area who you think might be interested in participating (I am looking for therapists -e.g., Clinical and Counseling Psychologists, Family Therapists, Couples Therapists, Social Workers, Psychiatrists, Counselor Educators- and therapists-in-training). 109 If you are agreeable to participating in the study please click on the link provided, to complete the first part of the study. ---- STUDY LINK ---- 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., M.S. Doctoral Candidate Counseling Psychology, University of Maryland Charles J. Gelso, PhD Professor Emeritus & Senior Lecturer Department of Psychology, University of Maryland First Email for Therapists and Trainees That I Know Dear XXXXX, Would you be willing to participate in my dissertation study? I am interested in examining therapists’ characteristics and feelings that influence the reactions to clients. Given your clinical experience and training, I’d really appreciate if you would contribute your time and perspective by participating in my study. Also, I think this is likely to be an interesting experience that might facilitate your reflection on clinical work. 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. Under the supervision of my advisor, Dr. Charles J. Gelso, I am running a study that involves two parts. If you agree to participate, you will first need to complete some brief questionnaires online (the link is provided below). This should take about 10 – 15 minutes. Then, I will ask you to come by a lab at the University of Maryland to engage in a therapy related experience (around 30 minutes). If the location does not work for you, I can also meet you at a more convenient location (e.g., your office) and run the computer tasks on my laptop. Participants who complete both portions of the study will automatically be entered in a raffle of two Amazon gift cards - $100 each. I would very much appreciate your participation in this study. I know it’s a busy time, and I understand that you have many demands on your schedule. So, I would be 110 extremely grateful if you would participate. Previous participants have stated that partaking in this study has been very interesting. Although we do not provide individual feedback, participants have described their experience as an opportunity to reflect about therapy and to ponder about one’s work. Additionally, Dr. Gelso and I hope this study will further illuminate therapists’ reactions to clients and also help us for future therapists’ training. That is why we are hoping to get a sample as large as possible. Finally, if you are interested, I can send you a summary of our findings and can notify you of any publications that result from this study. Also, feel free to pass this email along to anyone in the DC/MD area who you think might be interested in participating (I am looking for therapists and therapists-in-training). If you are agreeable to participating in the study, please click on the link provided, to complete the first part of the study. ---- STUDY LINK ---- 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. Beatriz Palma, M.Ed., M.S. Doctoral Candidate Counseling Psychology, University of Maryland Charles J. Gelso, PhD Professor Emeritus & Senior Lecturer Department of Psychology, University of Maryland (Emails are based on the ones used in Beatriz Palma’s Master’s Thesis and Ann Hummel’s Dissertation) 111 Appendix J Informed Consent This project is conducted by Dr. Charles Gelso and Beatriz Palma, M.Ed., M.S., at the University of Maryland, College Park. We are inviting you to participate in this research because of your clinical and training experience. The purpose of this research project is to examine therapists’ characteristics and feelings that influence the reactions to clients. This knowledge could be helpful in understanding and improving the practice of psychotherapy. The procedure involves two parts. If you agree to participate, you will first need to complete some brief questionnaires online. The first one asks for some background information. The other questionnaires address your and your experiences with some emotions. Completing all the questionnaires will take approximately 15 – 20 minutes. Then, I will ask you to come by a lab at the University of Maryland for half hour to complete some questionnaires and do some tasks on the computer. If the location does not work for you, we can meet at a more convenient location and run the computer tasks on my laptop. We will do our best to keep your personal information and responses strictly confidential. Once you complete the measures from the first part of the study, we will assign a code to your responses, and 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. 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 you directly (i.e., no individual, specific feedback will be given), but the results might help the investigators learn more about the therapist characteristics and behaviors that can affect the quality of the work with clients. 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 Mrs. 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 Mrs. Palma at bpalma@umd.edu. 112 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. 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 : _________________________ 113 Appendix K Debriefing Form We want to thank you for participating in this study. The purpose of our work is to examine psychotherapists’ reactions to an angry client. Specifically, we are looking at whether client’s race has an effect on therapists’ reactions to the angry client. In the first part of this investigation, you completed different measures online, including a survey about anger, and a questionnaire that assesses an attitude of awareness and acceptance that people are both the same and different to us. In the lab portion of this research, you participated in one of two conditions. In one condition, participants responded to a client who was White/European American. In the other condition, the client that participants watched was Black/African American. After seeing and responding to the client, you completed different items and questions regarding your reactions to such client. The different measures and the verbal responses to the clients will help us determine if therapists’ personal characteristics and attitudes relate to their reactions towards an angry client. Our study is not designed to provide you with information about your clinical work. We hope that, in the future, it will help other people through increased understanding of relevant aspects in working with angry clients. Your verbal responses to the videos, your responses to the questionnaires, and your answer to the open question will be held in strict confidentiality. Under no circumstances will this be violated. Rather, your responses will only be seen as anonymous. Furthermore, the data that you provided will not be linked to your name. You will be assigned a code, and all information will be saved under such number. For increased security, data will be encrypted and with restricted access. Additionally, any publication of data will be presented without identifiable information. A central aspect of our work is that therapists respond as they would in a natural setting. Being unaware of the specific purpose of our study or the nature of the data that will be collected is key to get such responses from participants. Due to the fact that many therapists have not yet participated in this study, we must ask you not to discuss this study in detail with anyone. This is crucial to maintaining the study’s validity. If you have any questions or concerns about your participation or the data you provided, please discuss this with us. We will gladly provide any information we can to address your questions or concerns. You can contact Beatriz Palma at 240-393-6973, or bpalma@umd.edu, or Dr. Charles J. Gelso at gelso@umd.edu. If your concerns are such that you would now like to have your data withdrawn we will do so. Also, we will be happy to provide a summary of the findings from the present study. If you would like to receive a summary of the findings, please send your request to Beatriz Palma (bpalma@ umd.edu). 114 If you have questions about your rights as a research participant, you may contact the University of Maryland’s Institutional Review Board at 301-405-4212 (irb@umd.edu). Again, our deepest appreciation for your participation in this study. Sincerely, Beatriz Palma, M.S., M.Ed. Doctoral Candidate, Counseling Psychology Program University of Maryland, College Park Charles J. Gelso, PhD. Professor Emeritus & Senior Lecturer, Counseling Psychology Program University of Maryland, College Park 115 Appendix L Table 4 Standard Multiple Regression Analysis of Clients’ Race, Universal-Diverse Orientation (UDO) and Anger Discomfort as Predictors of Behavioral Countertransference (i.e., Avoidance Index) Variable β t Sig Clients’ race -.029 -.217 .829 UDO .169 1.300 .199 Anger Discomfort -.088 -.684 .497 Table 5 Standard Multiple Regression Analysis of Clients’ Race, Universal-Diverse Orientation (UDO) and Anger Discomfort as Predictors of Self-Reported Countertransference (i.e., Countertransference Index Scale) Variable β t Sig Clients’ race .080 .600 .551 UDO .035 .266 .791 Anger Discomfort .064 .487 .628 116 Table 6 Hierarchical Multiple Regression Analysis of Clients’ Race, Universal-Diverse Orientation (UDO) and Anger Discomfort as Predictors of State Anxiety (Trait anxiety as covariate) Variable R2 R2-Change F F-Change Step 1 Trait Anxiety .145 .145 10.361 10.361** Step 2 Clients’ race .255 .110 4.959 2.844 * UDO Anger Discomfort Note. df for step 1 = 1, 61; df for step 2 = 3, 58; n = 63. *p < .05; **p < .01. 117 Table 7 Hierarchical Multiple Regression Analysis of Clients’ Race, Universal-Diverse Orientation (UDO), Anger Discomfort, and Interaction Terms (Clients’ Race X UDO and Clients’ Race X Anger Discomfort) as Predictors of Behavioral Countertransference (i.e., Avoidance Index) Variable R2 R2-Change F F- Change Step 1 Clients’ race .002 .002 .136 .136 Step 2 UDO .030 .028 .926 1.714 Step 3 Anger Discomfort .038 .008 .768 .468 Step 4 Clients’ race X UDO .042 .004 .500 .132 Clients’ race X Anger discomfort Note. df for step 1 = 1, 61; df for step 2 = 1, 60; df for step 3 = 1, 59; df for step 4 = 2, 57; n = 63. *p < .05; **p < .01. 118 Table 8 Hierarchical Multiple Regression Analysis of Clients’ Race, Universal-Diverse Orientation (UDO), Anger Discomfort, and Interaction Terms (Clients’ Race X UDO and Clients’ Race X Anger Discomfort) as Predictors of Self-Reported Countertransference (i.e., Countertransference Index Scale) Variable R2 R2-Change F F- Change Step 1 Clients’ race .004 .004 .247 .247 Step 2 UDO .005 .001 .157 .070 Step 3 Anger Discomfort .009 .004 .182 .628 Step 4 Clients’ race X UDO .037 .028 .442 .834 Clients’ race X Anger discomfort Note. df for step 1 = 1, 61; df for step 2 = 1, 60; df for step 3 = 1, 59; df for step 4 = 2, 57; n = 63. *p < .05; **p < .01. 119 Table 9 Hierarchical Multiple Regression Analysis of Clients’ Race, Universal-Diverse Orientation (UDO), Anger Discomfort, and Interaction Terms (Clients’ Race X UDO and Clients’ Race X Anger Discomfort) as Predictors of State Anxiety (Trait anxiety as covariate) Variable R2 R2-Change F F-Change Step 1 Trait Anxiety .145 .145 10.361 10.361** Step 2 Clients’ race .163 .018 5.849 1.288 ** Step 3 UDO .163 .000 3.843 .022 * Step 4 Anger Discomfort .255 .091 4.959 7.111 ** Step 5 Clients’ race X UDO .266 .011 3.375 .409 ** Clients’ race X Anger discomfort Note. df for step 1 = 1, 61; df for step 2 = 1, 60; df for step 3 = 1, 59; df for step 4 = 1, 58; df for step 4 = 2, 56; n = 63. *p < .05; **p < .01. 120 References Abernethy, A. D. (1995).Managing racial anger: A critical skill in cultural competence. Journal of Multicultural Counseling and Development, 23(2), 96-102. Doi: 10.1002/j.2161-1912.1995.tb00603.x Bandura, A., Lipsher, D., & Miller, P. (1960). Psychotherapists approach- avoidance reactions to patients’ expressions of hostility. Journal of Consulting Psychology, 24, 1–8. Beery, J. W. (1970). Therapist’s responses as a function of level of therapist experience and attitude of the patient. Journal of Consulting and Clinical Psychology, 34, 239-243. Beitman (1983). Categories of countertransference. Journal of Operational Psychiatry, 14(2), 82-90. Blair, I. V., Steiner, J. F., Fairclough, D. L., Hanratty, R., Price, D. W., Hirsh, H. K., Havranek, E. P. (2013). Clinicians’ implicit ethnic/racial bias and perceptions of care among Black and Latino patients. Annals of Family Medicine, 11(1), 43–52. doi:10.1370/afm.1442 Brondolo, E., DiGiuseppe, R., & Tafrate, R. C. (1997). Exposure-based treatment for anger problems: Focus on the feeling. Cognitive and Behavioral Practice, 4, 75- 98. doi: Brittan, C (1984). White therapists’ racial identity and countertransference reactions to black male clients (Unpublished master’s thesis). University of Maryland, College Park, MD. 121 Cabral, R. R. & Smith, T. B. (2011). Racial/ethnic matching of clients and therapists in mental health services: A meta-analytic review of preferences, perceptions and outcomes. Journal of Counseling Psychology, 58(4), 537-554. doi: 10.1037/a0025266 Chang, D. F., & Yoon, P. (2011). Ethnic minority cents’ perceptions of the significance of race in cross-racial therapy relationships. Psychotherapy Research, 21(5), 567- 582. doi: 10.1080/10503307.2011.592549 Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155-159. doi: 10.1037/0033-2909.112.1.155 Coleman, H. L. K., Wampold, B. E., & Casali, S. L. (1995). Ethnic minorities’ ratings of ethnically similar and European American counselors: A meta-analysis. Journal of Counseling Psychology, 42, 55-64. doi: 10.1037/0022-0167.42.1.55 Comas Díaz, L. (2012). Multicultural care: A clinician's guide to cultural competence. Washington, D.C: American Psychological Association. Cutler, R. L. (1958). Countertransference effects in psychotherapy. Journal of Consulting Psychology, 22, 349-356. doi: Davis, K. L., Hector, M. A., Meara, N. M., King, J. W., Tracy, D. C., & Wycoff, J. P. (1985). Teaching counselor trainees to respond consistently to different aspects of anger. Journal of Counseling Psychology, 32, 580-588. doi: Deffenbacher, J. L. (2011). Cognitive-behavioral conceptualization and treatment of anger. Cognitive and Behavioral Practice, 18, 212-221. doi: DiGiuseppe, R. & Tafrate, R. C. (2001). A comprehensive treatment model for anger disorders. Psychotherapy, 38(3), 262-271. 122 Dovidio, J.F., & Gaertner, S. L. (1998). 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. 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. Fiedler, F. E. (1951). A method of objective quantification of certain countertransference attitudes. Journal of Clinical Psychology, 7, 101-107. Frazier, P. A, Tix, A. P., Barron, K. E. (2004). Testing Moderator and Mediator Effects in Counseling Psychology Research. Journal of Counseling Psychology, 51(1), 115- 134. 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). Fuertes, J. N., & Gelso, C. J. (2000). Hispanic counselors’ race and accent and Euro Americans’ universal-diverse orientation: A study of initial perceptions. Cultural Diversity and Ethnic Minority Psychology, 6(2), 211-219. doi: 10.1037/1099- 9809.6.2.211 123 Fuertes, J. N., Miville, M. L., Mohr, J. J., Sedlacek, W. E., & Gretchen, D. (2000). Factor structure and short form of the Miville-Guzman Universality-Diversity Scale. Measurement and Evaluation in Counseling Psychology, 33, 157-169. Fuertes, J. N., Sedlacek, W. E., Roger, P. R., & Mohr, J. J. (2000). Correlates of universal-diverse orientation among first year university students. Journal of the First-Year Experience, 12, 45-59. Gamsky, N. R., & Farwell, C. F. (1966). Counselor verbal behavior as a function of client hostility. Journal of Counseling Psychology, 13, 184-190. Gelso, C. J. (1979). Research in counseling: Methodological and professional issues. The Counseling Psychologist, 8(3), 7-35. 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, NY: Wiley. Gelso, C. J., & Hayes, J. A. (2007). Countertransference and the therapist’s inner experience: Perils and possibilities. Mahwah, NJ: Lawrence Erlbaum. Gelso, C. J., & Mohr, J. J. (2001). The working alliance and the transference/countertransference relationship: Their manifestation with 124 racial/ethnic and sexual orientation minority clients and therapists. Applied & Preventive Psychology, 10, 51-68. Gelso, C. J. & Palma, B. (2011). Directions for Research on Self-Disclosure and Immediacy: Moderation, Mediation, and the Inverted U. Psychotherapy, 48, 342- 348. Gilbert, J., Carr-Ruffino, N., Ivanevich, J. M., & Lownes-Jackson, M. (2003) An empirical examination of inter-ethnic stereotypes: Comparing Asian American and African American employees. Public Personnel Management, 32(2), 251- 266. Gorkin, M. (1987). The uses of countertransference. Northvale, NJ: Jason Aronson. Haccoun D. M. & Lavigueur, H. (1979). Effects of clinical experience and client emotion on therapist’s response. Journal of Consulting and Clinical Psychology, 47, 416- 418. Hall, S. (1997) Race, the Floating Signifier. Media Education Foundation. Hall, R. E. (2001), The Ball Curve: Calculated Racism and the Stereotype of African American Men. Journal of Black Studies 2001 32(1), 104-119. doi: 10.1177/002193470103200106 Hahn, W. (1995) Therapist anger in group psychotherapy. International Journal of Groups Psychotherapy, 45(3), 339-347. Harbin, J. M. (2004). Countertransference reactions in a cross-racial dyad: The role of therapist universal-diverse orientation and presentation of client strengths (Unpublished master’s thesis). University of Maryland, College Park, MD. 125 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: Hayes, J. A. & Gelso, C. J. (1993). Male counselor’s discomfort with gay and HIV- infected clients. Journal of Counseling Psychology, 40, 86-93. Hayes, J. A., Gelso, C. J., & Hummel, A. M. (2011). Managing countertransference. Psychotherapy, 48, 88-97. doi: 10.1037/a0022182 Hayes, J. A., McCracken, J. E., McClanahan, M. K., Hill. C. E., Harp, J. S., & Carozzoni, P. (1998). Therapist perspectives on countertransference: Qualitative data in search of a theory. Journal of Counseling Psychology, 45(4), 468-482. Hayes, J. A., Riker, J. R. & Ingram, K. M. (1997). Countertransference behavior and management in brief counseling: A field study. Psychotherapy Research, 7, 145- 153. Hector, M. A., Davis, K. L., Denton, E. A., Hayes, T. W., & Hector, J. H. (1979). Teaching counselor trainees how to respond consistently to client negative affect. Journal of Counseling Psychology, 26, 146-151. doi: Helms, J. E. (Ed.) (1990). Black and white racial identity: Theory, research and practice. New York, NY: Greenwood Press. Heppner, P. P., Wampold, B. E., & Kivlighan, D. M. (2008). Research design in counseling. Belmont, CA: Thomson Brooks/Cole. 126 Hess, S. A., Knox, S., & Hill, C. E. (2006). Teaching graduate trainees how to manage client anger: A comparison of three types of training. Psychotherapy Research, 16(3), 282-292, doi: 10.1080/10503300500264838 Hiles, D. (2007). Human diversity and the meaning of difference. Paper presented at 10th European Congress of Psychology, Prague, CZ. Hill, C. E., Greenwald, C, Reed, K. G., Charles, D., O'Farrell, M., & Carter, J. (1981). Manual for counselor and client verbal response modes category systems. Columbus, OH: Marathon Consulting and Press. Hill, C. E., Kellems, I. S., Kolchakian, M. R., Wonnell, T. L., Davis, T. L., & Nakayama, E. Y. (2003). The therapist experience of being the target of hostile versus suspected-unasserted client anger: Factors associated with resolution. Psychotherapy Research, 13, 475-491. doi: 10.1093/ptr/kpg040 Hofsess, C. D. & Tracey, T. J. G. (2010). Countertransference as a prototype: development of a measure. Journal of Counseling Psychology, 57(1), 52-67. doi: 10.1037/a0018111 Holmes, D. E. (1999). Race and countertransference: Two “blind spots” in psychoanalytic perception. Journal of Applied Psychoanalytic Studies, 1(4), 319- 332. Hummel, A. M. (2013). Therapist reactions to a client facing terminal illness: A test of ego and countertransference. (Unpublished doctoral dissertation). University of Maryland, College Park, MD. Jackson, A. M. (1973). Psychotherapy: Factors associated with race of the therapist. Psychotherapy: Theory, Research, Practice, 10(3), 271-277. 127 Jones, E. E. (1985). Psychotherapy and counseling with Black clients. En P. B. Pedersen (ed.), Handbook of cross-cultural counseling and therapy (pp.173-179). Westport, CT: Greenwood. Jones, J. M. (1997). Prejudism and Racism. Mc Graw-Hill. Kannan, D., Henretty, J. R., Piazza-Bonin, E., Levitt, H. M., Coleman, R. A., Bickerest- Townsend, M, & Mathews, S. S. (2011). The resolution of anger in therapy: A task analysis. The Humanistic Psychologist, 39, 169-181. doi: 10.1080/08873267.2011.563724 Kassinove, H., & Tafrate, R. (2006). Anger related disorders: Basic issues, models, and diagnostic considerations. In E. L. Feindler. (Ed.), Anger related disorders: A practitioner’s guide to comparative treatments (pp. 1-27). New York, NY: Springer. Katz, A. D., & Hoyt, W. T. (2014). The influence of multicultural counseling competence and anti-Black prejudice on therapists’ outcome expectancies. Journal of Counseling Psychology, 61(2), 299-305. doi: 10.1037/a0036134 Knox, S., Burkard, A. W., Johnson, A. J., Suzuki, L. A., & Ponterotto, J. G. (2003). African American and European American therapists’ experience of addressing race in cross-racial psychotherapy dyads. Journal of Counseling Psychology, 50(4), 466-481. doi: 10.1037/0022-0167.50.4.466 Kugelmas, H. (2016). “Sorry, I’m not accepting new patients”: An audit study of access to mental health care. Journal of Health and Social Behavior, 57(2), 168-183. Latts, M. G. & Gelso, C. J. (1995). Countertransference behavior and management with survivors of sexual assault. Psychotherapy, 32, 405-415. 128 Ligiéro, D., & Gelso, C. J. (2002). Countertransference, attachment, and the working alliance: The therapist’s contribution. Psychotherapy, 39, 3–11. Lombardo, B. (1978). The Harlem Globetrotters and the perception of the Black stereotype.The Physical Educator, 35(2), 60-63. Maramba, G. G. & Nagayama Hall, G. C. (2002). Meta-analysis of ethnic match as predictor of dropout, utilization and level of functioning. Cultural Diversity and Ethnic Minority Psychology, 8, 290-297. doi: 10.1037/1099-9809.8.3.290 Markin, R. D., McCarthy, K. S., & Barber, J. P. (2013). Transference, countertransference, emotional expression, and session quality over the course of supportive expressive therapy: the rater’s perspective. Psychotherapy Research, 23 (2), 152-168. doi: 10.1080/10503307.2012.747013 Maroda, K. J. (2010). Psychodynamic techniques: Working with emotion in the therapeutic relationship. New York: The Guilford Press Maxie, A. C., Arnold, D. H., & Stephenson, M. (2006). Do therapists address ethnic and racial differences in cross-cultural psychotherapy? Psychotherapy: Theory, Research, Practice, Training, 43(1), 85-98, doi: 10.1037/0033-3204.43.1.85 McClure and Hodge, (1987). Measuring countertransference and attitude in therapeutic relationships. Psychotherapy, 24, 325-335. McGoldrick, M., Giordano, G., & Garcia-Preto, N. (2005). Ethnicity and family therapy. New York, NY: Guilford Press. McWilliams, N. (2004). Psychoanalytic psychotherapy: A practitioner’s guide. New York, NY: The Guilford Press. 129 Mikulincer, M. (1998). Adult attachment style and individual differences in functional versus dysfunctional experiences of anger. Journal of Personality and Social Psychology, 74(2), 513-524. Mirsky, J. (2011). Working through countertransference blocks in cultural-competence training. Psychoanalytic Social Work, 18, 136-148. doi: 10.1080/15228878.2011.611788. Miville, M. L., Gelso, C., J., Pannu, R., Liu W., Touradji, P., Holloway, P., & Fuertes, J. (1999). Appreciating similarities and valuing differences: The Miville-Guzman Universality-Diversity Scale. Journal of Counseling Psychology, 46, 291-307. Miville, M. L., Carlozzi, A. F., Gushue, G. V., Schara, S. L., & Masafumi, U. (2006). Mental health counselor qualities for a diverse clientele: Linking empathy, universal-diverse orientation, and emotional intelligence. Journal of Mental Health Counseling, 28(2), 151-165. 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 Munley, P. H. (1974) A review of counseling analogue research methods. Journal of Counseling Psychology, 21, 320-330 Munley, P. H., Lidderdale, M. A., Thiagarajan, M, & Null, U. (2004). Identity development and multicultural competency. Journal of Multicultural Counseling & Development, 32, 283-295. doi: 130 Murdock, N.L. (2011). The ghosts of Counseling Psychology: Is counseling research really dead? The Counseling Psychologist, 39, 715-718. Pallant, J. (2010). SPSS survival manual: A step by step guide to data analysis using SPSS. Berkshire, England: Open University Press. Palma Orellana, B., I. & Gelso, C. J. (2013). What Does the Therapist Bring to the Relationship? The Connections Among Real Relation, countertransference and attachment. Psychotherapy Bulletin, 48(2), 12-16 Paulhus, D. L. (1984). Two-component models of socially desirable responding. Journal of Personality and Social Psychology, 46, 598–609. Paulhus, D. L. (1986). Self-deception and impression management in test responses. In A. Angleitner & J. S. Wiggins (Eds.), Personality assessment via questionnaires (pp. 143–165). Berlin: Springer. Paulhus, D. L. (1994). Balanced Inventory of Desirable Responding: Reference manual for BIDR Version 6. Unpublished manuscript, University of British Columbia, Vancouver, Canada. Paulhus, D. L. (1998). The Balanced Inventory of Desirable Responding. Toronto, Ontario, Canada: Multi-Health Systems. 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 Pérez Foster, R.M. (1998). The clinician’s cultural countertransference: The psychodynamics of cultural competent practice. Clinical Social Work Journal, 26(3), 253-270. 131 Ponterotto, J. G. (2008). Theoretical and empirical advances in multicultural counseling and psychology. In S. D. Brown & R. W. Lent (Eds.) Handbook of Counseling Psychology. (4th Ed) Hoboken, N.J.: Wiley. Proctor, E. K. & Rossen, A. (1981). Expectations and preferences for counselor race and their relation to intermediate treatment outcomes. Journal of Consulting Psychology, 28(1), 40-46. Ridley, C.R. (2005). Overcoming unintentional racism in counseling and therapy: A practitioner’s guide to intentional intervention. Thousand Oaks, CA: Sage. Russell, P. D., & Snyder, W. U. (1963). Counselor anxiety in relation to amount of clinical experience and quality of affect demonstrated by clients. Journal of Consulting Psychology, 27, 358-363. doi: Saini, M. (2009). A meta-analysis of the psychological treatment of anger: Developing guidelines for evidence-based practice. The Journal of the American Academy of Psychiatry and the Law, 37(4), 473-488. doi: Sattler, J. M. (1970). Racial “experimenter effects” in experimentation, testing, interviewing and psychotherapy. Psychological Bulletin, 73(2), 137-160. Sears, D. O., Citrin, J., & Van Laar, C. (1995). Black exceptionalism in a multicultural society. Paper presented at the society of experimental social psychology, Annual Meeting, Washington, D.C. Sharkin, B. S. (1988). The measurement and treatment of client anger in counseling. Journal of Counseling and Development, 66, 361-365. Sharkin, B. S. (1989). How counselor trainees respond to client anger: A review. Journal of Counseling and Development, 67, 561-564. 132 Sharkin, B. S. & Gelso, C. J. (1991). The Anger Discomfort Scale: Beginning reliability and validity data. Measurement and Evaluation in Counseling and Development, 24(2), 61-68. Sharkin, B. S., & Gelso, C. J. (1993). The influence of counselor trainee anger-proneness and anger discomfort on reactions to an angry client. Journal of Counseling and Development, 71, 483-487. doi: Shin, S., Chow, C., Camacho-Gonsalves, T., Levy, R. J., Allen, I. E., & Leff, H. S. (2005).A meta-analytic review of racial-ethnic matching of African American and Caucasian American clients and clinicians. Journal of Counseling Psychology, 52, 45-56. doi: 10.1037/022-0167.52.1.45 Singley, D. B., & Sedlacek, W. E. (2009). Differences in universal-diverse orientation by race-ethnicity and gender. Journal of Counseling & Development, 87, 404-411. Smith T. B., & Trimble, J. E. (2016). Foundations of multicultural psychology: Research to inform effective practice. Washington, DC, US: American Psychological Association. Spielberger, C.D., Gorsuch, R.L., & Lushene, R.E. (1970). The State-Trait Anxiety Inventory: Test manual. Palo Alto, CA: Consulting Psychologist Press. Spielberger, C. D., Jacobs, G., Russell, S., & Crane, R. (1983). Assessment of anger: The State-Trait Anger Scale. In J.J. Butcher & C.D Spielberger (Ed.), Advance in personality assessment (Vol 2, pp.159-187). Hillsdale, NJ: Erlbaum. Spielberger, C. D., Sydeman, S. J., Owen, A. E., & Marsh, B. J.(1999). Measuring anxiety and anger with the State-Trait Anxiety Inventory (STAI) and the State- Trait Anger Expression Inventory (STAXI). In M. E. Maruish (Ed.), The use of 133 psychological testing for treatment planning and outcomes assessment (2nd ed.) (p. 993-1021). Mahwah, NJ: Lawrence Erlbaum Associates Publishers. Spielberger, C. D., & Reheiser, E. C. (2004). Measuring anxiety, anger, depression, and curiosity as emotional states and personality traits with the STAI, STACI, and STPI. In Hilsenroth, M. J., Segal, D. L., & Hersen, M. (Ed.), Comprehensive Handbook of Psychological Assessment (Vol 2, pp. 70-86). Whiley. Stangor, C. (2011). Social Psychology (Preliminar Edition, Special edition for University of Maryland students). Upper Saddle River, NJ: Pearson Custom Publishing. Stepanikova, I. (2012). Racial-ethnic biases, time pressure, and medical decisions. Journal of Health and Social Behavior, 53(3), 329-43. Stracuzzi, T. I., Mohr, J. J., & Fuertes, J. N. (2011). Gay and bisexual male clients’ perceptions of counseling: The role of perceived sexual orientation similarity and counselor universal-diverse orientation. Journal of Counseling Psychology, 58(3), 299-309, doi: 10.1037/a0023603 Sue, D. W. & Sue, D. (2008). Counseling the multicultural diverse: Theory and practice. Hoboken, NJ: Wiley. Tabachnick, B. C., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Boston: Pearson Publication. Thompson, R.L., Brossart, D. F., Carlozzi, A. F., & Miville, M. L. (2002). Five-factor model (big five) personality traits and universal-diverse orientation in counselor trainees. Journal of Psychology, 136, 561-572. 134 Varble, D. L. (1968). Relationship between the therapist’s approach-avoidance reactions to hostility and client behavior in therapy. Journal of Consulting and Clinical Psychology, 32, 237-242. doi: Van Wagoner, S. L. (2000). Anger in Group Therapy, Countertransference and the Novice Group Therapist. In S.S. Fehr, Group Therapy in Independent Practice. New York, NY: Routledge. Vontress, C. E. (1971). Racial differences: Impediments to rapport. Journal of Counseling Psychology, 18(1), 7-13. Vontress, C. E. (1979). Cross-cultural counseling: An existential approach. Personnel and Guidance Journal, 58, 117-122. Vontress, C. E. (1988). An existential approach to cross-cultural counseling. Journal of Multicultural Counseling and Development, 16, 78-83. Vontress, C. E. (1996). A personal retrospective on cross-cultural counseling. Journal of Multicultural Counseling and Development, 24, 156-166. Wilkinson, D. (1993). Family ethnicity in America. In H. P. McAdoo (Ed.). Family ethnicity: Strength in diversity. (pp.15-59). Newbury Park, CA: Sage Winnicott, D. W. (1949). Hate in the countertransference. The International Journal of Psycho-analysis, 30, 69-74. Yoo, H. C., & Pituc, S. T. (2013). Assessment of perceived racial stereotypes, discrimination, and racism. In K. F. Geinsinger (Ed.-In-Chief) APA handbook of testing and assessment in psychology. Washington, DC: American Psychological 135 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.