ABSTRACT Title of Dissertation: SUPERVISEE PERCEPTIONS OF SUPERVISORY FOCUS ON STRENGTHS AND CONSTUCTIVE FOCUS ON DEFICITS: DEVELOPMENT AND VALIDATION OF A MEASURE Melissa Stephanie Roffman, Doctor of Philosophy, 2007 Directed By: Professor Clara E. Hill, Department of Psychology The purpose of the present study was to develop and validate a measure of the constructs of focus on supervisee strengths and constructive focus on deficits from the supervisee perspective, the Supervisory Focus on Strengths and Deficits Inventory-Supervisee Form (SUPSAD-S). Participants (N = 204 for exploratory factor analysis, N = 201 for all other analyses) were a national sample of masters? and doctoral level graduate students in various mental health professions (e.g., counseling psychology; clinical psychology, social work). Data was collected through the use of an internet survey containing the SUPSAD-S as well as measures used to assess its convergent and discriminant validity. The overall return rate ranged from 44-46%. Exploratory factor analysis was used to assess the construct validity and explore the underlying factor structure of the SUPSAD- S. The final two factor solution retaining 24 items accounted for 63% percent of the variance. The first subscale, Focus on Strengths, consisted of 12 items (45.41% of variance) reflecting interventions used by supervisors to identify and enhance supervisees? strengths. The second subscale, Constructive Focus on Deficits, consisted of 12 items (accounting for 17.68% of the variance) reflecting interventions used by supervisors to address supervisees? deficits in a non-critical or non-punitive manner intended to help supervisees grow and improve. Higher subscale scores indicate greater supervisory focus on strengths and constructive focus on deficits, respectively, from the perspective of supervisees rating their supervisors. Evidence of good reliability (i.e., high internal consistency and two week test-retest reliability estimates) for each of the subscales was found. In addition, initial support was found for the convergent and discriminant validity of both subscales. Specifically, the convergent validity of the focus on strengths subscale was supported by its correlations with satisfaction with supervision (r = .71), the supervisory working alliance (r = .69), and counseling self-efficacy (r = .27) as expected. The convergent validity of the constructive focus on strengths subscale was supported by its correlations with satisfaction with supervision (r = .57) and the supervisory working alliance (r = .50) as expected. Discriminant validity of both subscales was supported by their lack of correlation with public self-consciousness. Focus on strengths was also found to be a stronger predictor of positive supervision outcomes than constructive focus on deficits. Implications for supervision practice and future research are discussed. SUPERVISEE PERCEPTIONS OF SUPERVISORY FOCUS ON STRENGTHS AND CONSTUCTIVE FOCUS ON DEFICITS: DEVELOPMENT AND VALIDATION OF A MEASURE By Melissa Stephanie Roffman 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 2007 Advisory Committee: Professor Clara E. Hill, Chair Professor Charles J. Gelso Professor Mary Ann Hoffman Associate Professor Carl Lejeuz Assistant Professor Ty Tashiro ? Copyright by Melissa Stephanie Roffman 2007 ii ACKNOWLEDGMENTS First, I cannot express how grateful I am to my advisor Dr. Clara Hill for her mentoring, enthusiasm for the study, encouragement and belief in me as a researcher, and going above and beyond to help me complete the study (including very fast turn-around time on drafts). I?ve learned so much from working with her and genuinely enjoyed the experience, especially our walking meetings! Second, I would like to express my gratitude to several groups of individuals who made invaluable contributions to various aspects of this study. My committee members, Dr. Charles Gelso, Dr. Mary Ann Hoffman, Dr. Ty Tashiro, and Dr. Carl Lejeuz, provided insightful feedback and suggestions that not only improved the study but deepened my thinking about my research. Many people helped me develop the SUPSAD- S measure, including my advisee team (Jessica Stahl, Patricia Spangler, Wonjin Sim, Jingqing Liu, and Robyn Zakalik), research assistants David Schulzinger and Anna Sheveland, students in the UMD counseling psychology program, and especially Ty Tashiro, whose consultations taught me a great deal about psychometrics. In addition, I could not have completed this dissertation without the help of Shirley Hess, Sarah Knox, Jim Mahalik, Rachel Crook-Lyon, Nicholas Ladany, Daniel Isenberg, Louis Castonguay, James Boswell, Heather Walton, Penny Asay, Jimmy Harbin, and others, who went out of their way to help me recruit participants for my study; I am likewise grateful to all of my participants for taking the time out to complete my survey. Will Dardick, Alan Levy, Ty Tashiro, and Matthew Patton also helped me tremendously by volunteering their time to consult on data analysis, and I learned so much about statistics and data analysis in the process. Jessica Stahl helped with many aspects of the study, including edits to the final iii draft. Third, I want to thank my family and friends for their support and encouragement throughout this entire process: Dr. Jeffrey and Arlene Roffman, Eugene Roffman, Michael Roffman, Sheetal Patel, Nazish Salahuddin, Jessica Stahl, Rachel Freidus, Thomas White, Heather Walton, Kelly Hennessy, Nicole Taylor, Rayna Markin, Matthew Patton, Livia Taravella, Jordan Buffaloe, Nicholas Eliopulos, Patrick Fleming, Greg Wrenn, Andrew Hagan, Emily Hamm, David Schulzinger, and the College Perk coffeehouse. Finally, I would like to dedicate this dissertation to my grandparents, Eleanor and Eugene Roffman, who have always inspired and encouraged me in my academic and writing pursuits. iv TABLE OF CONTENTS Acknowledgments ??????????????????????????.ii Table of Contents ??????????????????????????.iv List of Tables ??????????????????????????? .. vi Chapter 1: Introduction................................................................................................. 1 Chapter 2: Review of the Literature.............................................................................. 4 Definition of Strengths and Deficits ......................................................................... 4 Positive Psychology.................................................................................................. 6 Focus on Strengths in Counseling Psychology ??????????????.9 Theoretical Approaches to Supervision that Incorporate a Focus on Strengths? . 12 Rationale for Focus on Strengths and Constructive Focus on Deficits ................. 26 Empirical Research Related to Strengths Focus and Deficits Focus ...................... 39 Conclusion .............................................................................................................. 59 Chapter 3: Statement of the Problem.......................................................................... 61 Chapter 4: Methods..................................................................................................... 77 Participants.............................................................................................................. 77 Measures ................................................................................................................. 79 Procedures............................................................................................................... 84 Test-Restest Reliability ??????????????????????...91 Chapter 5: Results....................................................................................................... 93 Development of the SUPSAD-S............................................................................. 93 Initial Content Validity of the SUPSAD-S ............................................................. 95 Exploratory Factor Analysis on the SUPSAD-S .................................................... 98 Reliability Estimates of SUPSAD-S subscales..................................................... 116 Convergent and Discriminant Validity of the SUPSAD-S................................... 117 Exploratory Research Questions........................................................................... 125 Chapter 6: Discussion ............................................................................................... 132 Factor Structure of the SUPSAD-S....................................................................... 132 Reliability of the SUPSAD-S................................................................................ 135 Convergent and Discriminant Validity of the SUPSAD-S................................... 136 Exploratory Research Questions........................................................................... 142 Limitations ............................................................................................................ 150 Implications for Supervision Practice and Training ............................................. 154 Implications for Future Research??????????????????...155 Appendix A: Web Survey Introductory Page and Eligibility Requirements???.160 Appendix B: Informed Consent Form??????????????????162 Appendix C: Supervisory Focus on Strengths and Deficits Inventory-Supervisee Form (SUPSAD-S) Measure????????????????????????166 Appendix D: Satisfaction with Supervision Questionnaire??????????169 Appendix E: Supervisory Working Alliance Inventory (Trainee Form)?????171 Appendix F: Counseling Activity Self-Efficacy Scales ???????????172 Appendix G: Public Self-Consciousness Scale ??????????????.174 Appendix H: Demographic Questionnaire????????????????.175 Appendix I: Initial Recruitment Email?????????????????.. 178 Appendix J: Two week/one month follow-up recruitment email???????.. 179 v Appendix K: Follow-up recruitment letter with incentive??????????.180 Appendix L: Test-Retest Recruitment Letter???????????????.181 Appendix M: Structure Coefficients (Factor Loadings) and Pattern Coefficients of Deleted Items???????????????????????????.. 182 References ????????????????????????????.184 vi LIST OF TABLES Table 1. Structure Coefficients, Pattern Coefficients, Item-Total Correlations, and Alphas if Item Deleted for SUPSAD-S Items?????????????.108 Table 2. Item Means, Standard Deviations, and Skewness and Kurtosis Statistics of SUPSAD-S Items????????????????????? 112 Table 3. Correlations, Means, Standard Deviations, Range, Skewness, Kurtosis, Internal Consistency Estimates, and Test-Retest Reliabilities for the SUPSAD-S Subscales????????????????????????115 Table 4. Correlation Matrix, Means, Standard Deviations, Skewness and Kurtosis Statistics for SUPSAD-S, SSQ, SWAI-T, CASES, and SCS?????????120 Table 5. Simultaneous Multiple Regressions with SUPSAD-S Subscales as Predictors of SSQ, SWAI-T, and CASES Scores?????????????..127 Table 6. Correlation Matrix for SUPSAD-S Strengths and Deficits Subscales and Supervisor Theoretical Orientations??????????????????.. 130 1 Chapter 1 INTRODUCTION A focus on strengths in supervision seems called for given the recent emphasis on positive psychology, which has been defined as the scientific pursuit of optimal human functioning and the illumination of the role of human strengths in leading a better life through sound scientific research (e.g., Lopez et al., 2006; Seligman & Czikszentmihalyi, 2000; Aspinwall & Staudinger, 2003). In contrast to mainstream perspectives in psychology which have focused exclusively or disproportionately on the negative while ignoring the positive, positive psychologists advocate for focusing on both the positive and the negative, strengths and weaknesses (e.g., Wright & Lopez, 2002; Lampropoulos, 2001). Although renewed interest in positive psychology in the broader field of psychology has led to an increase in research on human strengths and assets in the past several years, especially in applied areas (e.g., therapy, the workplace; Linley & Joseph, 2004; Joseph & Linley, 2004; Kauffman & Scoular, 2004), how positive psychology may inform and enhance supervision practices remains largely unstudied. In addition, a focus on strengths in supervision seems to be called for in counseling psychology, given the historical emphasis on assets and optimal functioning (Gelso & Fretz, 2001; Lopez et al., 2006; Super, 1955). At the same time, this focus has also been called counseling psychology?s ?unfulfilled promise,? referring to how relatively little has been done empirically and theoretically to advance knowledge of human strengths and positive development (Gelso & Fassinger, 1992; Lopez et al, 2006; Lopez, Edwards, Magyar-Moe, Pedrotti, & Ryder, 2003). Accordingly, counseling psychology training may not be sufficiently grounded in the strengths model, such that 2 counseling psychologists may espouse the philosophy of focusing on strengths but not utilize strength-focused interventions in their therapy and supervision roles (Gerstein, 2006). In order for strength-focused training practices to be implemented into the training of the next generation of counseling psychologists, more empirical research on the effects of strength-focused supervision is needed (Lopez et al., 2006; Gerstein, 2006; Lopez et al., 2003). Finally, clinical and theoretical supervision literature (e.g., Briggs & Miller, 2005; Timm & Blow, 1999; Larson, 1998) suggests that focusing on identifying and enhancing supervisee strengths and focusing constructively on supervisee deficits may each benefit supervisees? professional functioning (e.g., counseling self-efficacy) as well as supervision process and outcome (e.g., the supervisory working alliance, satisfaction with supervision ). Unfortunately, very little empirical literature exists to support these theorized benefits, or to operationalize the processes and interventions used by supervisors to focus on supervisees? strengths or on deficits in a constructive way. Two small bodies of empirical research closely related to the constructs of supervisory focus on strengths and constructive focus on deficits have yielded preliminary support for theorized benefits. First, a small and largely exploratory body of empirical research on feedback in supervision suggests that supervisees value and consider helpful both positive and constructive negative feedback from their supervisors, and that the provision of each may benefit supervisees? short- and long-term professional development (e.g., Heckman-Stone, 2003; Daniels & Larson, 2001; Wulf & Nelson, 2000; Abbott & Lyter, 1998). More research is needed to understand how positive feedback and constructive negative feedback relate to supervision process and outcome. 3 Also, some preliminary empirical research on the impact of solution-focused supervision, an explicitly strength-oriented theoretical approach to supervision, suggests that incorporating solution-focused techniques into supervision may predict enhanced supervisee counseling self-efficacy and other positive aspects of supervision and therapy process and outcome (Triantafillou, 1997; Koob, 2002). These solution-focused supervision studies were limited by a lack of a psychometrically valid measure of strength-focused supervision techniques and because they did not investigate techniques used to by supervisors to constructively address supervisees? deficits. The purpose of the present study, then, is to develop and validate a measure of the constructs of supervisory focus on strengths and constructive focus on deficits. I sought to assess these constructs from the perspective of supervisees because supervisees? perceptions of supervisory focus on strengths and deficits may be more important than what supervisors actually do or report doing, per se. I also sought to explore how supervisory focus on strengths and constructive focus on deficits each impact supervisees? satisfaction with supervision, the supervisory working alliance, and counseling self-efficacy. The development of this measure may allow for further empirical investigation of a positive psychology approach to supervision. By examining and researching these processes more closely, perhaps supervisors may eventually be trained to work more effectively with the strengths and deficits of their supervisees. 4 Chapter 2 REVIEW OF THE LITERATURE In this literature review, I first provide the definitions of supervisee strengths and deficits that guided the current investigation. Next, I provide a summary of the philosophy, historical background, and current research and assessment trends in positive psychology. Then, I discuss focus on strengths or hygiology as a unifying theme throughout the history of the counseling psychology profession. Next, I discuss theoretical approaches to supervision that incorporate a focus on strengths as well as deficits. After that, I describe how focus on strengths and constructive focus on deficits in supervision may each benefit supervisees? professional development and supervision process and outcome. Finally, I review empirical literature related to working with strengths and deficits in supervision. Definitions of Strengths and Deficits Human strengths have been defined in the positive psychology literature as the ?psychological ingredients? or routes to displaying core human virtues such as wisdom, courage, and justice thought to be universally linked to good character (Peterson & Seligman, 2004). Examples of strengths that have been operationalized and studied include hope (Snyder, Rand, & Sigmon, 2002), optimism (Carver & Scheier, 2002), resilience (Masten & Reed, 2002), self-efficacy (Bandura, 1977; Maddux, 2002), and spirituality (Pargament & Mahoney, 2002). Strengths have also been conceptualized as traits which lead to healthy processes (e.g., adaptive coping), which then lead to growth and fulfillment (e.g., well-being, satisfaction, and achievement of highest potential in relationships, work, etc.) (Lopez & Snyder 2003; Lopez, Snyder, & Rasmussen, 2003; 5 Aspinwall & Staudinger, 2003). Based on these definitions as well as strength-oriented theoretical approaches to supervision discussed later in this review, supervisee strengths are defined in the current investigation as proficiencies and inner resources that supervisees bring to the enterprise of conducting therapy (e.g., Timm & Blow, 1999; Carlson & Erickson, 2001). These strengths may be related to the personality, life experiences, interpersonal skills, talents, or abilities (traits or ?psychological ingredients?) of supervisees, as well as aspects of clinical work in which supervisees perform well, correctly, adaptively, or appropriately, either consistently or in an emerging/nascent way (healthy processes) (e.g., Briggs & Miller, 2005; Wetchler, 1990; Carlson & Erickson, 2001; Timm & Blow, 1999; Gelso & Woodhouse, 2003). By utilizing their positive traits to contribute to healthy processes in terms of their performance as therapists, supervisees can move toward competence, growth, and fulfillment of their highest potential as therapists. By virtue of their being human, in training, and involved in the inherently complex nature of the therapy enterprise, supervisees also likely have deficits, defined in the current investigation as ?psychological ingredients? that detract from the ability to learn and perform therapy (i.e., liabilities), as well as aspects of clinical work in which supervisees lack competence, need improvement, make mistakes, use poor judgment, have personal issues that get in the way of their work, or generally fail to work effectively with clients (Abbott & Lyter, 1998; Lopez, Snyder, & Rasmussen, 2003). Therefore, it seems optimal for supervisors to focus on supervisees? strengths as well as deficits, and both of these constructs (focus on strengths and constructive focus on deficits) are central to the current investigation. 6 In the current investigation, focus on strengths and constructive focus on deficits are treated as two separate constructs (rather than two parts of a single construct or two ends of a continuum), given the possibility that supervisors may focus on both, on one and not the other, or on neither. However, each seems important for effective supervision to occur, as will be discussed further in this review of literature. Given that psychology research has traditionally focused disproportionately on deficits while neglecting strengths (e.g., Seligman & Czikszentmihalyi, 2000; Sheldon & King, 2001), the remainder of this review of literature will focus primarily on theory and research related to working with supervisees? strengths, while also highlighting theory and research relevant to constructive focus on deficits. Positive Psychology Philosophy of Positive Psychology Positive psychology has been defined as the pursuit of optimal human functioning and the illumination of the role of human strengths in leading a better life through sound scientific research (Lopez et al., 2006; Seligman, 2002). In contrast to mainstream psychology?s traditional adherence to the medical model and the fundamental assumption that human nature is predominantly negative (Maddux, Snyder, & Lopez, 2004; Seligman & Peterson, 2003), positive psychologists maintain that human nature is motivated toward developing its potential (Seligman & Csikszentmihalyi, 2000; Joseph & Linley, 2004). In fact, positive psychologists assert that mainstream psychology?s focus on pathology detracts from a fundamental mission of psychology to further enhance the lives of all people (Seligman & Csikszentmihalyi, 2000). Therefore, they contend that psychologists should shift their focus away from treating deficits and towards studying 7 how to make people?s lives even better (e.g., Sheldon & King, 2001; Snyder & Lopez, 2002). Positive psychology has been theorized to not only build on people?s strengths but also to increase resilience and improve quality of life (Keyes & Lopez, 2002; Lampropoulos, 2001; Seligman, 2002; Seligman & Csikszentmihalyi, 2000; Linley & Joseph, 2004). For example, Lopez and Snyder (2003) proposed an omnibus hypothesis that measuring, identifying, and enhancing human strengths make a difference in people?s lives. More specifically, they hypothesized that working with strengths should improve achievement in students, productivity in the workplace, mental health, and clinical training. Unfortunately, such hypotheses on the impact of human strengths have largely remained untested. The term ?positive psychology? may be misinterpreted to mean an exclusive focus on the positive. In contrast to mainstream perspectives in psychology such as the ?medical model? which have focused exclusively or disproportionately on the negative while ignoring the positive, positive psychologists acknowledge the importance of focusing on both the positive and the negative, strengths and weaknesses (Wright & Lopez, 2002; Lampropoulos, 2001; Keyes & Lopez, 2002; Lopez & Snyder, 2003; Lopez, Snyder, & Rasmussen, 2003; Snyder et al., 2003). Ignoring weaknesses or focusing too narrowly on the positive is considered by positive psychologists to be overly optimistic or even ?Pollyanna? (Seligman, 2002; Gelso & Woodhouse, 2003; Lopez, Snyder, & Rasmussen, 2003; Sheldon & King, 2001). In contexts such as psychotherapy, focusing only on the positive could invalidate the real suffering of clients (i.e., by not reducing their distressing symptoms), ignore their goal of healing what is wrong through 8 therapy, or result in only superficial gains; it could also limit the variance accounted for in positive psychology research (Seligman, 2002; Lopez & Magyar-Moe, 2006). Hence, the mantra of positive psychology can be summed up in the phrase: ?Develop the strengths, manage the weaknesses? (Lopez & Snyder, 2003, p. xiii). History and Current Trends in Positive Psychology During his tenure as the president of the American Psychological Association, Martin Seligman helped to shift psychology?s focus toward positive psychology, which was historically missing from most research, theory, and clinical practice in psychology (Seligman & Csikszentmihalyi, 2000; Lopez et al., 2006). Although previous theories and movements had addressed positive constructs (e.g., the mental illness reform movement of the early 1900?s, Terman?s studies of giftedness in the early 20 th century, the human potential movement of the 1960?s including Maslow?s concept of the self-actualized person and Rogers? client-centered theory), Seligman?s major contribution was uniting these concepts under the common theme of positive psychology, with an explicit focus on advancing knowledge through stringent empirical investigation and research methodology (e.g., Lopez et al., 2006; Linley & Joseph, 2004). Seligman called for greater theory and research in the area of positive psychology and optimal functioning in his 1998 APA presidential address, and his call has begun to be answered as the movement has achieved momentum (Seligman & Csikszentmihalyi, 2000; Lopez et al., 2006; Aspinwall & Staudinger, 2003). For example, the American Psychologist had two full issues dedicated to the topic (i.e., ?Special Issue on Happiness, Excellence, and Optimal Functioning,? Volume 55, 2000; ?Positive Psychology,? Volume 56, 2001); the Journal of Social and Clinical Psychology (Vol. 19, 2000) had an 9 entire issue dedicated human strengths and virtues; and The Counseling Psychologist (Vol. 34, 2006) had an entire issue dedicated to positive aspects of human functioning. Furthermore, several handbooks have been published on positive psychology such as Positive Psychology Assessment (Lopez & Snyder, 2003), the Handbook of Positive Psychology (Snyder & Lopez, 2002), A Psychology of Human Strengths (Aspinwall & Staudinger, 2003), and Positive Psychology in Practice (Linley & Joseph, 2004), as well as a comprehensive classification of human strengths and virtues (?a manual of the sanities;? Peterson & Seligman, 2004) and a positive psychology approach to ethics encouraging psychologists to fulfill the highest ideals of profession and promoting exemplary ethical behavior rather than punishing unethical behavior (Knapp & VandeCreek, 2006). These recent contributions highlight the renewed interest in positive psychology in the broader field of psychology, and Lopez et al. (2006) noted that positive psychology has also begun to make its way into the mainstream media and the public consciousness. Although there has been an increase in research on human strengths and assets in the past several years, and an increased interest in bringing positive psychology research to applied areas (e.g., therapy, the workplace; Linley & Joseph, 2004; Joseph & Linley, 2004; Kauffman & Scoular, 2004), research on how strengths are used within clinical supervision and training has lagged far behind. For example, supervision and training were not addressed in the recently published handbook on positive psychology in practice. Focus on Strengths in Counseling Psychology A focus on human strengths, assets, and optimal functioning has historically been 10 one of the five unifying themes of counseling psychology (Gelso & Fretz, 2001). Moreover, counseling psychology?s focus on strengths has been central to its identification as a unique applied specialty within professional psychology, distinct from closely related areas such as clinical psychology (Gelso & Fretz, 2001; Lopez et al., 2006; Gerstein, 2006). Super (1955, p.19) attributed counseling psychology?s distinctiveness to its focus on ?hygiology? or ?the normalities even of abnormal persons,? as opposed to a more disease or pathology oriented approach. Consistent with this focus, one of counseling psychologists? central roles is the ?educative/developmental? role, which entails enhancing or maximizing individuals? potential to direct themselves effectively and to discover ways to identify, develop, and use their personal and social resources (Gelso & Fretz, 2001; Lopez, Edwards, et al., 2003). At the same time as a focus on strengths has helped to maintain and ensure the integrity and identification of as a specialty within professional psychology, this focus has also been called counseling psychology?s ?unfulfilled promise,? referring to how relatively little has been done empirically and theoretically to advance knowledge of human strengths, positive development, and optimal functioning (Gelso & Fassinger, 1992). More than a decade later, Gerstein (2006) noted the gap between counseling psychology?s ?rhetoric? of claiming to be grounded in a strength- or hygiology-based model of human functioning, and the reality of counseling psychology training, which largely teaches deficit or pathology models of human behavior and therapy intervention. Consequently, counseling psychologists may espouse the philosophy of focusing on strengths, but lack the skills needed to conduct strength-based therapy and supervision (Gerstein, 2006). Counseling psychology?s relative lack of strength-focused research may 11 be changing, however, as reflected by a recent volume of The Counseling Psychologist (2006, volume 34) devoted entirely to positive psychology; a collection of chapters on positive psychology written by counseling psychologists called Counseling Psychology and Optimal Human Functioning (Walsh, 2003); and the results of a content analysis of counseling psychology peer-reviewed journals suggesting that 29% (N = 1135) of published articles have a positive focus (Lopez et al., 2006). Lopez et al. (2006) called for continued work toward reaffirming the professional identity of counseling psychology through an increased focus on strengths in practice and research, and also offered recommendations for strength-focused practice and research to be implemented into the training of the next generation of counseling psychologists. In order for strength-focused training practices to be implemented into the training of the next generation of counseling psychologists, it seems that more strength- focused scholarship in counseling psychology in general and more research on strength- focused supervision in particular is needed (Lopez et al., 2006). Notably, none of the articles in the recent positive psychology issue of The Counseling Psychologist (2006) or chapters in the Walsh (2003) book on optimal human functioning addressed strengths in supervision or training contexts. Also, the supervision theories that most directly focus on supervisee strengths (i.e., solution-focused supervision, narrativist supervision; e.g., Wetchler 1990; Carlson & Erickson, 2001) were developed not by counseling psychologists but by marriage and family therapists. The development of a measure that assesses supervisory focus on supervisee strengths as well as constructive focus on deficits could encourage more empirical research in this area that could eventually be 12 used to inform training practices in counseling psychology programs, contributing over time to changing the rhetoric of strength-based focus to reality. Theoretical Approaches to Supervision That Focus on Strengths In this section, I will describe several theoretical approaches to supervision that incorporate a focus on supervisees? strengths, while also noting how these theories address deficits. It is notable that most of these theories have not been studied empirically and thus we lack data on their impact on supervisee development and supervision process and outcome. Client-centered/Person-centered/Humanistic Approaches There are several ways in which client-centered approaches to supervision incorporate a focus on supervisee strengths. First, client-centered supervision, like client- centered therapy, is based on a positive view of the individual (supervisee or client), who is seen as motivated toward fulfilling his or her potential (Rogers, 1963; Frankland, 2001; Patterson, 1997). The positive nature of client-centered supervision approaches is reflected within its deep non-directiveness (Patterson, 1997; Gelso & Woodhouse, 2003). Specifically, supervisors follow the supervisees? lead, trusting in their strengths, potential, motivation for growth, and natural tendency to move in an actualizing direction (e.g., toward greater counseling competency) if provided with a nurturing environment that includes the facilitative conditions (i.e., empathy, genuineness, unconditional positive regard) (Rogers, 1963; Frankland, 2001; Gelso & Woodhouse, 2003). Client- centered supervision may be somewhat less nondirective than client-centered therapy given that supervisors are responsible for making sure that supervisees achieve a certain level of competence and provide an acceptable standard of care for clients; therefore, 13 client-centered supervisors must still correct supervisee deficits when necessary (Frankland, 2001). In general, however, client-centered supervision is similar to client- centered therapy in that it is not so much what supervisors do in supervision (i.e., strength-based interventions and techniques) that is relevant to its positive nature, but rather supervisors? attitudes and assumptions about their supervisees? strengths that is of key importance (Joseph & Linley, 2004; Gelso & Woodhouse, 2003). Accordingly, client-centered supervision models do not clearly delineate specific techniques or interventions to foster or build on supervisee strengths, but rather an overall approach that facilitates the emergence of supervisee strengths (Gelso & Woodhouse, 2003; Patterson, 1997). A second way that client-centered supervision addresses supervisee strengths is through valuing or affirmation of the supervisee?s work, similar to Rogers? belief in ?prizing? clients and affirmatively viewing their experience without questioning or diagnosing (Frankland, 2001; Joseph & Linley, 2004; Rogers, 1963). Valuing supervisees? work may allow them to return to their clinical work with greater positivity, self-confidence, and self-awareness (Frankland, 2001; Bernard & Goodyear, 2004; Patterson, 1997). Solution-Focused Supervision An increasing number of authors have begun to discuss solution-focused supervision (Marek, Sandifer, Beach, Coward, & Protinsky, 1994; Selekman & Todd, 1995; Presbury, Echterling, & McKee, 1999; Wetchler, 1990; Rita, 1998; Thomas, 1996; Juhnke, 1996; Koob, 2002; Triantafillou, 1997; Briggs & Miller, 2005), an approach to supervision derived from the solution-focused brief model of therapy introduced by 14 deShazer (e.g., 1988). Solution-focused models have been used primarily in the context of marriage and family therapy and its supervision (e.g., Marek et al., 1994). In general, solution-focused supervision models aim to identify and amplify supervisee strengths, successes, and positives, rather than focusing on weaknesses or mistakes (e.g., Briggs & Miller, 2005; Juhnke, 1996; Presbury et al 1999). Consistent with its name, solution-focused supervision incorporates a focus on generating solutions to problems based on supervisees? past successful experiences and strategies, with less focus on the problems themselves (e.g., Selekman & Todd, 1995; Rita, 1998). For example, instead of focusing on why a supervisee is struggling to be appropriately challenging with a particular client, a solution-focused supervisor might help the supervisee figure out what allowed her to be appropriately challenging with clients on previous occasions that may be applied to the present case (Presbury et al., 1999). Solution-focused theories assume that there is no one universal solution or correct way to intervene, instead defining ?correct? as whatever works for a particular supervisee in a particular case (Wetchler, 1990; Marek et al., 1994; Koob, 2002). Solution-focused supervisors assume that supervisees are competent and possess the strengths and resources needed to solve problems and achieve training goals (Briggs & Miller, 2005). Accordingly, supervisors act in the consultant role, taking a collaborative, goal-oriented approach to supervision that affirms, empowers, and sets up positive expectations for the supervisee (Rita, 1998). Solution-focused supervision theorists argue that a disproportionate focus on supervisee problems and mistakes can reinforce supervisees? low confidence, self- criticism, and feelings of inadequacy as clinicians (e.g., Wetchler, 1990; Briggs & Miller, 15 2005). Pointing out that in the complex enterprise of therapy, there will always be more things clinicians have not mastered than things they have mastered, they argue that a focus on strengths can contribute to the supervisees? development of an identity around success and competence rather than incompetence (Wetchler, 1990; Marek et al., 1994). Importantly, solution-focused supervision theorists do not advocate ignoring supervisee problems and mistakes (Briggs & Miller, 2005), but argue instead that there are always exceptions to problems and times when problems are not present, or are less frequent, less intense, or shorter in duration; therefore, in most cases, supervisees have already displayed or enacted solutions to either resolve problems or lessen their frequency, intensity, or duration (Rita, 1998; Selekman & Todd, 1995). This emphasis on exceptions communicates the supervisor?s confidence that the supervisee can solve the problem, and gives the supervisee confidence that she or he can recognize and overcome the inevitable pitfalls and difficulties that occur in therapy (e.g., Juhnke, 1996; Wetchler, 1990; Marek et al., 1994; Briggs & Miller, 2005). In contrast to other supervision models, solution-focused supervision focuses centrally on the supervisee?s development as a counselor (e.g., counseling self-efficacy, professional identity and autonomy), rather than on the supervisee?s clients and their problems (e.g., Koob, 2002; Briggs & Miller, 2005; Triantafillou, 1997). Reasons for this focus include the greater generalizablity of what is learned in supervision (i.e., the supervisee?s strengths as a therapist) to a variety of clinical situations (Briggs & Miller, 2005); improved clinical work/treatment outcome as a result of increased supervisee counseling self-efficacy and utilization of strengths (Wetchler, 1990; Marek et al., 1994); and the explicit assignment of a more active role to the supervisee, rather than treating the 16 supervisee as a ?conduit? through which the supervisor treats the client (Koob, 2002, p. 166). Solution-focused supervision theorists have done a thorough job identifying specific techniques and interventions used to address supervisee strengths, many of which are derived from solution-focused therapy (e.g., Marek et al., 1994; Selekman & Todd, 1995; deShazer, 1988). For example, solution-focused supervisors may use different types of questions to focus discussion on supervisees? strengths, achievements and competencies, for example: ?What aspect of your counseling have you noticed getting better since we last met?,? ?Tell me best thing you did with client this week,? or ?What strengths did you demonstrate in this particular session?? (e.g., Presbury et al., 1999; Briggs & Miller, 2005). When supervisees feel stuck or ineffective with clients, supervisors may ask the ?miracle question? to elicit supervisee expertise, strengths, and competencies in finding a solution to the problem: ?Pretend the miracle happened overnight and your problem (with the client) is solved. How will you be able to tell, what will you be doing differently, and how did you make it happen?? (e.g., Rita, 1998; Koob, 2002). The miracle question can be used to help therapists define their vision of a successful professional (Koob, 2002). Supervisors may also use scaling questions, which generally take the form of ?On a scale of 1-10, where 1 is the problem at its worst and 10 is the problem is solved, where are you today, where do you want to be in a week/month/year etc.?,? to help supervisees focus and gauge their quantitative progress toward solving specific problems and toward becoming a successful professional in terms of competence, confidence, and satisfaction with their work (e.g., Koob, 2002; Juhnke, 1996). 17 Another solution-focused supervision technique is the use of presuppositional language (e.g., Rita, 1998; Selekman & Todd, 1995), which assumes a supervisee strength is an actuality (e.g., ?Tell me a time when you were assertive with a client? or ?What will you do to be assertive this week with the client??), in contrast to subjunctive language, which assumes that a supervisee positive event is a possibility that may or may not have occurred (e.g., ?Can you think of a time when you were assertive with a client??). Presuppositional language or ?the language of change? (Rita, 1998) is said to be preferable because it makes it more difficult for supervisees to dismiss their strengths, conveys supervisors? belief in supervisees? competency, communicates supervisors? expectation that supervisees will continue to make positive progress, and helps supervisees draw on past problem-solving strategies to address current problems and goals (Presbury et al., 1999; Rita, 1998). Thomas (1996) found that supervisees considered presuppositional language the most influential technique of solution-focused supervision. Solution-focused supervision, like solution-focused therapy, uses the approach of finding exceptions or pattern interruption, which entails focusing on minor alternations that disrupt maladaptive supervisee behavior and lead to more adaptive behavior (Selekman & Todd, 1995; Gelso & Woodhouse, 2003). Under the assumption that there are no complete failures (Presbury et al., 1999), supervisors seek to identify and amplify successes and strengths even within failed interventions (Rita, 1998), help supervisees understand how they made the successful change occur (Briggs & Miller, 2005; Rita, 1998), challenge supervisees to own the success rather than seeing it as an accident or fluke (Presbury et al., 1999), and reinforce the success or underlying strength through 18 compliments (?I?m impressed with your growth?) and ?cheerleading? (e.g., ?Wow! That was great!?) (e.g., Wetchler, 1990). This explicit focus on and reinforcement of supervisee strengths and successes may be important because supervisees might not always realize the strengths they possess in their repertoire, or might not utilize their strengths in a particular intervention because of low self-confidence or countertransference (Gelso & Woodhouse, 2003). With continued focus on and reinforcement of small successes that may be exceptions to failures, supervisee strengths and successes should become more the rule than the exception over time (Wetchler, 1990; Marek, et al., 1994; Briggs & Miller, 2005). Solution-focused supervision literature has not addressed as clearly or thoroughly what supervisors should do when supervisees lack competence or the necessary resources and strengths to address client problems. Several authors suggest taking a more expert, didactic stance to educate supervisees when their problems with clients stem from a lack of knowledge, rather than from a failure to recognize solutions based on their own resources and strengths (e.g., Briggs & Miller, 2005). At the same time, they highlight the importance of encouraging supervisees to find their own answers rather than providing them with the answers, and limiting the educative/didactic role to times when it is specifically requested by supervisees out of their own developmental/training needs, within the limits of ensuring client welfare (Wetchler, 1990; Marek et al., 1994). Narrativist/Constructivist Approaches Narrativist approaches (e.g., Bob, 1999; Timm & Blow, 1999; Carlson & Erickson, 2001) to therapy assume that people are inherently story-tellers who develop stories about themselves that organize past experience and influence future behavior. 19 Applied to supervision, narrativist approaches view supervisees as beginning to develop their stories of themselves as professionals and therapists, in collaboration with supervisors who help them ?revise? or ?edit? their stories (Bernard & Goodyear, 2004). Similar to client-centered supervision, the supervisee is seen as the expert or ?author? of the story, with the supervisor facilitating the process of telling the story through a stance of curiosity and interventions that assist the supervisee?s exploration (Bernard & Goodyear, 2004; Carlson & Erickson, 2001; Timm & Blow, 1999). Constructivist supervision approaches are based on the position that reality and truth are contextual and socially constructed, with a strong emphasis on the supervisee?s subjective experience, collaboration between supervisor and supervisee to construct and reconstruct meaning, and a relatively egalitarian supervision relationship with the supervisor in a consultant role (Bernard & Goodyear, 2004). Narrativist and constructivist supervision approaches, like client-centered supervision, may indirectly communicate a belief in the supervisee?s strengths and competence through their supervisee-centered, collaborative approach. However, several narrativist and constructivist supervision theorists have described approaches that focus more explicitly or directly on supervisee strengths. For example, Carlson and Erickson (2001) described an approach whereby supervisors ?honor? and ?privilege? supervisees? life experiences, and help supervisees identify and utilize strengths, resources, knowledge, and skills derived from these life experiences that are relevant to their work as helpers (e.g., natural helping ability). Also, in facilitating supervisees? narration of their story of how they became helpers, supervisors help supervisees focus on their healthy motivations for pursuing helping careers, rather than pathological reasons such as 20 family-of-origin conflicts and unresolved personal issues. Similarly, Timm and Blow (1999) suggested that supervisors approach supervisees? family of origin issues and historical life events in a balanced way, helping supervisees overcome difficulties or interferences in their clinical work resulting from these experiences (i.e., countertransference, blind spots, pathology or impairment; Gelso & Hayes, 2001), while also helping supervisees access and utilize the strengths, inner resources, and intuitions also developed as a result of the experiences. In the latter approach, supervisors might focus on how supervisees? life experiences make them especially well-suited for a particular case in terms of deeper understanding of the issue and more compassion for the client?s struggle. Timm and Blow (1999) cautioned that supervisors still have the responsibility to be aware of when supervisees? personal issues and conflicts interfere with therapy, for example through manifestation as countertransference behavior, and advocate for attention to both supervisee strengths and deficits stemming from their life experiences. Behavioral and Cognitive-Behavioral Approaches Behavioral and cognitive-behavioral approaches to supervision are characterized by a focus on behaviorally specific supervisee goals, which are achieved through rigorous, ongoing assessment of supervisee skills and progress, and strategies such as practicing or role playing techniques in counseling (e.g., Rosenbaum & Ronen, 1998; Follette & Callaghan, 1995; Bradley & Gould, 2001; Bernard & Goodyear, 2004). Underlying assumptions of these supervision approaches include: (a) both adaptive and maladaptive behaviors are learned and maintained through consequences; (b) proficient therapist performance is a function of learned skills more than ?personality fit;? (c) the 21 purpose of supervision is to teach appropriate behaviors and extinguish inappropriate behaviors; and (d) supervision should employ principles of learning theory within its procedures (Bernard & Goodyear, 2004; Bradley & Gould, 2001). Behavioral and cognitive-behavioral supervision incorporate a focus on supervisee strengths in several ways. First, they assume the basic potential of any supervisee to learn given the right training experiences. Since proficient therapist performance is a function of learned skills rather than personality fit, any supervisee can be trained, and thus no supervisee is inherently unfit to be a therapist (Bradley & Gould, 2001; Gelso & Woodhouse, 2003). Also, these approaches assume the supervisee?s self- directedness and personal resources, with the supervisor acting as a consultant to supervisee in constructing and carrying out strategies (Bradley & Gould, 2001; Bradley & Kottler, 2001). Additionally, like behavioral therapy, behavioral supervision may include techniques such as reinforcement of supervisee strengths, successes, and changes in the desired direction, and helping the supervisee learn to reinforce themselves for what they do right (e.g., Gelso & Woodhouse, 2003; Follette & Callaghan, 1995; Pierce & Epling, 1999). However, although strengths are reinforced in behavioral approaches, practitioners of these approaches may not necessarily be conceptualizing supervisees in terms of strengths and may be focused just as much on deficits in terms of helping supervisees eliminate problematic behaviors and learn from mistakes (Gelso & Woodhouse, 2003; Bradley & Gould, 2001). Social-Cognitive Models of Supervision In the Social-Cognitive Model of Counselor Training (SCMCT model; Larson, 1998), the supervisor?s three functions are to provide (a) modeling experiences, (b) social 22 persuasion (i.e., realistic, supportive encouragement and structured learning situations that increase the chance of counseling successes for the counselor), and (c) feedback on the counselor?s performance (specific, constructive, both positive and negative, focused on changeable aspects of supervisee performance). These three supervisor functions are posited to provide a safe, positive learning environment for supervisees that positively impacts their counseling self-efficacy and performance. According to the SCMCT model, supervisors should monitor supervisees? self-efficacy, which is considered optimal when it slightly optimistic relative to their performance (Larson, 1998). Moreover, supervisors should explicitly intervene when necessary to raise counseling self-efficacy or lower supervisee anxiety, for example through positive performance feedback on what the supervisee did well. Building on the SCMCT model, Lent and colleagues (1998) theorized that a combination of knowledge, skills, and self-efficacy for important counseling tasks, paired with challenging, proximal, specific goals, leads to counseling effectiveness. Critical aspects of supervision that would lead to the acquisition of knowledge, skills, and self- efficacy would include: vicarious learning experiences in important counseling tasks, structured practice where chances of success are purposely maximized (mastery experiences), assistance with anxiety and affect management (e.g., helping supervisee replace non-adaptive attributions for successful and unsuccessful counseling experiences with more adaptive attributions), a focus on progress rather than ultimate goal attainment, reinforcement of successful experiences, ample support and encouragement, and challenge to stretch existing skills. Thus, social-cognitive models of supervision incorporate a focus on supervisee 23 strengths in several ways. First, the emergence of supervisee strengths is facilitated through structuring learning experiences in a way that maximizes the likelihood of supervisee success (Lent, Hackett, & Brown, 1998). Second, supervisors intervene directly at times to build on supervisee strengths, for example through positive feedback, helping supervisees develop adaptive attributions for their successes and failures (e.g., stable, internal attributions for successes), and encouragement/support (Larson, 1998; Lent et al., 1998). In general, social-cognitive supervisors aim to enhance supervisee strengths both in terms of the actual manifestation of strengths (i.e., supervisee?s counseling performance/effectiveness) as well as the supervisee?s (inner) sense of ownership of their strengths (i.e., counseling self-efficacy). Social-cognitive models of supervision also incorporate a focus on supervisees? deficits in terms of the importance of providing constructive negative feedback on the supervisees? performance and challenging supervisees to stretch existing skills and abilities to reach specific goals. Feminist Approaches Characteristics of feminist supervision generally include reduced hierarchy with a collaborative and egalitarian supervision relationship and maximum empowerment of the supervisee in generating of supervision goals, as well as attention to larger themes of gender, power, diversity, social/contextual factors, and responsible action (Syzmanski, 2003; Bernard & Goodyear, 2004). Apart from occasions in which supervisors must exert expert power to ensure client welfare, feminist supervisors generally validate the supervisee?s strengths, trust the supervisee?s experiences and intuitions, and encourage the supervisee?s autonomy (e.g., Syzmanski, 2003; Porter & Vasquez, 1997). Feminist approaches may also be considered deficit focused to the extent that they raise 24 supervisees? awareness of larger themes of gender, power, and social/contextual factors related to their clinical work where awareness had been lacking (Syzmanski, 2003). Summary of Strength-Focused Theoretical Approaches to Supervision The theoretical approaches to supervision discussed here have several elements in common. First, all of the approaches assume, explicitly or implicitly, that supervisees possess strengths, resources, competencies (e.g., things supervisees do well, correctly, adaptively, or appropriately, either in consistent or emerging/nascent way), or at the very least, the potential to learn to be competent with training (i.e., behavioral supervision). These strengths and resources stem from sources such as supervisees? personality, life experiences, interpersonal skills, and talents. Though varying in techniques, supervisors in all of the approaches communicate directly or indirectly their belief in supervisee strengths, and intervene to identify, nurture, develop, and amplify these strengths. Examples of these interventions include positive feedback; reinforcement; a supportive, empowering, and/or non-directive environment that allows strengths to naturally emerge; structuring training experiences to facilitate the emergence of strengths and successes; shaping adaptive attributions; focusing on successful exceptions to problems; and asking questions directed at strengths and competencies. Second, these approaches emphasize a collaborative, egalitarian, non-hierarchical approach to supervision, in which the supervisor is in the consultant role, defined as a more collegial role in which supervisor and supervisee cooperate to plan interventions and solve problems, with the assumption of supervisee competence. It has been suggested that the consultant supervisory role (corresponding the attractive supervisory style) is probably least threatening to supervisees, maximizing their sense of control and 25 empowerment, as compared to the other two supervisory roles/styles (i.e., counselor/interpersonally sensitive; teacher/task-oriented; e.g., Bradley and Kottler, 2001; Bernard & Goodyear, 2004; Friedlander & Ward, 1984; Ladany, Walker, & Melincoff, 2001). Third, these approaches tend to focus on supervisee development, rather than supervisees? clients and case management. This supervisee focus may be helpful to supervisees given that their strengths and assets may generalize across their clinical work, as opposed to the specifics of individual cases (Briggs & Miller, 2005). Finally, these approaches, while strengths-focused, also acknowledge that supervisees will naturally also possess weaknesses, deficits, and problems that must still be addressed. Yet, in addressing weaknesses, strengths (a) should never be neglected, and (b) may in fact be utilized in various ways to overcome deficits (e.g., using strengths to find solutions to problems; replacing inappropriate or maladaptive behaviors and attributions with more positive ones; focusing on how life experiences that contribute to countertransference can also be important sources of empathy and compassion for clients). Although these theoretical approaches suggest that focusing on supervisee strengths is important or would help supervision, researchers have rarely empirically investigated how these approaches impact supervisee development and supervision process and outcome (Bernard & Goodyear; Bradley & Gould, 2001). It is important that supervision theories incorporating a focus on supervisee strengths in addition to deficits gain further empirical validation. 26 Rationales for Focus on Strengths and Constructive Focus on Deficits In the following section, I discuss the potential benefits of supervisory focus on strengths and constructive focus on deficits in terms of: (a) supervisees? growth and development as a clinicians and (b) supervision process and outcome. This division of sections follows from the categorization of major variables addressed in the supervision literature (Ladany & Muse-Burke, 2001). Supervisee Benefits. One of the primary goals of supervision is to enhance the professional functioning of the supervisee, through building competency and promoting the development of professional identity, theoretical orientation, and counseling self-efficacy (Bernard & Goodyear, 2004; Lent, Hill, & Hoffman, 2003). Focus on strengths and constructive focus on deficits in supervision may each contribute to the enhancement of supervisee professional functioning in a number of ways, described next. Enhanced counseling self-efficacy. Counseling self-efficacy refers to counselors' beliefs about their ability to effectively perform counseling-related behaviors or to negotiate particular clinical situations (Larson & Daniels, 1998; Lent et al., 2003; Bandura, 1977; Maddux, 2002) and has been found to relate to reduced supervisee anxiety and increased comfort in the counseling role (Lent et al., 2003; Larson & Daniels, 1998), career development variables (e.g., degree of interest in, and goals regarding, counseling as a central activity in their occupational lives; Heppner, O'Brien, Hinkelman, & Flores, 1996; Lent et al., 2003), performance in counseling situations (Larson & Daniels, 1998), persistence and motivation to learn the complexities of therapy especially when faced with challenges/failure, etc. (Larson & Daniels, 1998), and 27 satisfaction the with counseling role (Larson & Daniels, 1998). Therefore, supervision interventions aimed at increasing supervisee counseling self-efficacy should make these positive outcomes more likely to occur. Theoretical and empirical literature suggest that focusing on strengths may enhance supervisee self-efficacy, and therefore, may indirectly influence these other positive outcomes through their influence on counseling self-efficacy. It has been speculated that trainees, particularly those in early stages of development, have a tendency to be self-deprecating, critical, and insecure of their own abilities, and that this tendency may have a negative impact on their competence as therapists (e.g., Gelso & Woodhouse, 2003; Briggs & Miller, 2005; Stoltenberg, McNeill, & Delworth, 1998). For example, supervisees may limit their range of responses to clients, forgetting the full range of interventions of which they are capable; may be preoccupied during sessions with clients about forthcoming criticism in supervision; and may generally overlook or be unaware of strengths and resources they could utilize to help their work with clients (Gelso & Woodhouse, 2003; Briggs & Miller, 2005; Wetchler, 1990). Supervisors typically intervene in such cases to improve supervisees? counseling self-efficacy, for example with positive feedback, emotional support, a focus on what the supervisee does well, a reminder of similar situations in which the supervisee has effectively handled the problem, teaching the supervisee to self-monitor their positive behaviors, or communicating confidence in the supervisee (Daniels & Larson, 2001; Larson, 1998; Briggs & Miller, 2005; Gelso & Woodhouse, 2003; Barnett et al., 2001). Supervisors? provision of positive feedback has been found to increase supervisees? counseling self- efficacy (Daniels & Larson, 2001). 28 Theoretical literature also suggests that supervisors can enhance supervisees? counseling self-efficacy by constructively focusing on skill deficits through challenge. Specifically, supervisors should challenge supervisees to stretch existing skills in attempting increasingly difficult tasks and behaviors that will optimally result in mastery experiences (e.g., Larson & Daniels, 1998; Larson, 1998; Lent et al., 1998). It seems to be crucial for supervisors to challenge supervisees; without challenge, supervisees may lose motivation to push themselves to improve, which can result in stagnation and diminished self-efficacy and sense of accomplishment. Also, not challenging supervisees may be experienced as infantilizing or convey the belief that they are too fragile or inept to be pushed (e.g., Bernard & Goodyear, 2004; Gould & Bradley, 2001). Supervisors must, however, challenge supervisees in constructive (e.g., gentle, respectful of supervisees? anxiety and limits of competence) ways, so as not to overwhelm, paralyze, discourage, magnify self-doubt and feelings of incompetence, or elicit defenses from supervisees (Blocher, 1983; Bernard & Goodyear, 2004; Barnett, Youngstrom, & Smook, 2001; Daniels & Larson, 2001). Overall, focus on strengths (in terms of positive feedback, support, encouragement) and constructive focus on deficits (in terms of challenge) each may contribute positively to supervisee counseling self-efficacy, helping supervisees develop identities as competent therapists capable of achieving successful outcomes (Briggs & Miller, 2005). Enhanced supervisee learning in supervision. Another fundamental goal of supervision is to help supervisees develop competency in a variety of domains of professional functioning, including counseling skills and techniques, case 29 conceptualization, diagnosis, assessment, treatment planning, and multicultural counseling (e.g., Bernard & Goodyear, 2004; Ladany, Inman, Constantine, & Hofheinz, 1997). Supervision is intended to be a place of learning for supervisees, and accordingly, one of the supervisor?s major roles is as a teacher who imparts new skills and knowledge on the supervisee (Bradley & Kottler, 2001; Bernard & Goodyear 2004). Given that supervision is a place of learning, learning principles such as reinforcement (defined as a consequence that increases the frequency of a preceding befhavior) and punishment (defined as an aversive consequence that decreases the frequency of a preceding behavior) can likely be applied to enhance supervisee learning (Pierce & Epling, 1999; Myers, 1998; Catania, 2001; Follette & Callaghan, 1995). For example, supervisors might build on supervisee strengths by reinforcing what supervisees do effectively or appropriately, as well as provide direction for improvement or alternative behaviors when supervisees inevitably have problems or make mistakes (e.g., Abbott & Lyter, 1998; Bernard & Goodyear, 2004; Gould & Bradley, 2001). Reinforcing supervisee strengths is likely a more effective teaching method than punishing their mistakes and deficits, for all of the reasons that reinforcement is considered a more powerful technique than punishment to teach correct or desired behaviors. First, unlike punishment which merely suppresses unwanted behavior, reinforcement provides information to the learner to guide him or her in the direction of more positive, desired, or correct behaviors. In other words, punishment only tells people what not to do, while reinforcement provides tells them what to do (Pierce & Epling, 1999; Myers, 1998). Applied to supervision, reinforcement of supervisee strengths and behaviors that approximate desired standards provides supervisees with information 30 about what clinical behaviors are appropriate and should therefore be continued or further developed (e.g., Follette & Callaghan, 1995). Focusing on supervisees? mistakes and deficits without calling attention to their strengths and appropriate behaviors only informs them of what not to do, but does not replace eliminated inappropriate behaviors with more appropriate alternative behaviors or suggest a direction for improvement (e.g., Abbott & Lyter, 1998). Second, punishment has been found to have several negative side effects, including fear or hostility toward the punisher, the modeling of aggressive responses, learned helplessness and depression in the learner, and the mere suppression (but not extinction) of the unwanted behavior, which may reappear in other situations where the punishment is avoidable (e.g., untaped sessions that the supervisor will not hear) (Pierce & Epling, 1999; Myers, 1998; Catania, 2001). Applied to supervision, a non-constructive focus on supervisees? deficits may feel punishing to the supervisee, therefore possibly eliciting fear or hostility toward the supervisor which would likely harm the supervisory relationship (e.g., Bernard & Goodyear, 2004; Ladany, Hill, Corbett, & Nutt, 1996; Abbott & Lyter, 1998). For example, Ramos-Sanchez et al. (2002) found that negative experiences in supervision can have observable negative effects on supervisees' clinical work, satisfaction with training, and future career decisions. Similarly, Abbott and Lyter (1998) found that harsh or punitive criticism in supervision can harm supervisees? self- confidence, motivation, and learning. Therefore, it would seem that supervisors may enhance supervisees? learning by reinforcing their strengths and guiding their therapy skills and techniques toward closer and closer approximations of expected levels of competency (e.g., shaping, Myers, 1998; 31 Follette & Callaghan, 1995; Pierce & Epling, 1999), and by helping the supervisee learn to reinforce themselves for what they do right (Gelso & Woodhouse, 2003). For example, supervisors might reinforce supervisees by praising what they do well. Barnett et al. (2001; p. 223) noted that ?Words of encouragement and pointing out what supervisee does well go a long way. Praise not only feels good to the trainee, but also provides useful feedback about strengths and what styles or skills are working well.? It has also been suggested that focusing on supervisees? strengths and personal/professional growth in supervision, rather than focusing on specific clients and their problems, would enhance supervisees? learning and application of learning to clinical work, because supervisees? strengths are likely global and generalizable to a variety of clients and clinical situations (Briggs & Miller, 2005). In addition, a focus on strengths might increase supervisors? interpersonal influence (e.g., perceived expertise, attractiveness, etc.) such that their feedback to supervisees would be more persuasive and therefore more likely to be taken in and implemented (e.g., Stoltenberg, McNeill, & Crethar, 1995). Moreover, it is recommended that supervisors make trainee errors, which are inevitable, into ?teachable moments,? (Knapp & VandeCreek, 2006; p. 224), and address the errors in a constructive, non-punitive way that promotes supervisees? personal growth (Abbott & Lyter, 1998). Specifically, when offering criticism, supervisors should focus only on behaviors that can be changed; be specific; offer criticism as an opinion rather than fact; separate personal feelings about supervisee from the need to criticize; steer away from accusatory comments or ultimatums; model self-critique; create a positive learning environment where it is safe and normal to make mistakes, and critical feedback is to be expected; convey appreciation, encouragement, warmth, and a non-punitive, non- 32 judgmental attitude; sandwich negative feedback between positive feedback by accentuating supervisee strengths/competencies in light of areas needing improvement; and follow criticism with brainstorming to identify methods and means for removing deficiencies and achieving competencies (Abbott & Lyter; 1998; Weisinger & Lobsenz, 1981; Bernard & Goodyear, 1998). Positive expectations for supervisee competence and self-fulfilling prophecy. Seligman (2002; p. 6) referred to the ?birthright? of psychologists as both healing what is weak and nurturing what is strong. Accordingly, he called for psychologists to seek more desirable self-fulfilling prophecies through adherence to a strength-based approach, moving away from an exclusive focus on weakness/deficits (Sandron, 1970; Snyder et al., 2003; Lopez, Snyder, & Rasmussen, 2003). Applied to supervision, focusing on supervisees? strengths may communicate supervisors? confidence in their potential to achieve competence and successful outcomes in their clinical work and professional development (e.g., Briggs & Miller, 2005; Wetchler, 1990). A positive self-fulfilling prophecy may then be set into motion as supervisees internalize their supervisors? positive beliefs and expectations and gain motivation to live up to, or not to disappoint, their supervisors? expectations (e.g., Sandron, 1970; Barnett et al., 2001). One way supervisors might encourage the development of a positive self-fulfilling prophecy is through setting high but achievable goals and expectations in the form of scaffolding. Scaffolding is defined as structuring learning tasks and goals so that the difficulty level is slightly beyond what learners are currently capable of, but not so far above their level that it will be unachievable and frustrating (Vygotsky, 1978). Scaffolding may communicate the expectation that supervisees can reach the desired 33 level, and may set the stage for supervisees to have mastery experiences that will improve their counseling self-efficacy (e.g., Lent et al., 2003; Lent et al., 1998; Larson, 1998; Bandura, 1977). Both constructive focus on deficits (in terms of challenging supervisees in a constructive way to improve existing counseling skills and behaviors through scaffolding) and focus on strengths (in terms of supporting and encouraging supervisees to ensure they reach their goals) may help create a positive self-fulfilling prophecy for supervisees. Building resilience against the stresses of professional development and the therapy enterprise. In their discussion of ways to enhance supervisee excellence, Knapp and VandeCreek (2006) highlighted how supervisors can help inoculate supervisees against the stresses of professional training and the therapist role through attention to enhancing supervisees? resilience, well-functioning, and emotional health (Schwebel & Coster, 1998). For example, when addressing supervisees? countertransference reactions that are interfering with their clinical work, supervisors might reinforce supervisees? strengths related to countertransference management ability, such as appropriate engagement in self-care activities (e.g., a balanced lifestyle), boundary setting, and self- awareness (Gelso & Hayes, 2001; Schwebel & Coster, 1998). The enhancement of supervisee resilience through focusing on strengths may be informed by Keyes and Lopez?s (2002) conception of mental health as both the absence of mental illness (e.g., symptoms) and the presence of well-being. In this model, clients can have either high or low mental illness symptoms while at the same time (independently) having high or low subjective well-being, therefore yielding four ?types? of clients: (a) struggling clients (i.e., those with high mental illness symptoms and high 34 subjective well-being), (b) flourishing clients (i.e., those with low mental illness symptoms and high subjective well-being), (c) floundering clients (i.e., those with high mental illness symptoms and low subjective well-being), and finally (d) languishing clients (i.e., those with low mental illness symptoms and low subjective well-being). Keyes and Lopez (2002) highlighted the importance of building on client strengths through the example of languishing clients, who have resolved their deficits according to disease-oriented models, yet have not received the positive interventions that might increase their well-being and life satisfaction and prevent relapse (i.e., turning them into ?flourishing? clients). Analogously, supervisees could also be described as ?languishing? when they, for example, meet minimum standards of competence and function at an acceptable professional level, yet fail to live up to their highest potential as therapists in terms of competence, satisfaction and enjoyment of their work, and resilience to the stresses of training and the therapist role. Along the same lines, positive interventions such as building on supervisee strengths might produce ?flourishing? supervisees. Focusing constructively on supervisee deficits, in addition to building on strengths, could also be particularly beneficial for ?floundering? or ?struggling? supervisees, who need help meeting the basic standards of competence, but could also benefit from increasing or maintaining their satisfaction with their clinical work and resilience to professional stress, and from maximizing their highest potential (Keyes & Lopez, 2002; Abbott & Lyter, 1998). Supervision Process and Outcome Benefits Supervisory focus on strengths and constructive focus on deficits may each benefit important aspects of supervision process and outcome. Specifically, these foci 35 may improve the supervisory working alliance; decrease supervisee evaluation anxiety and facilitate the evaluative component of supervision/evaluation; and increase supervisee satisfaction with supervision. Improved supervisory relationship/working alliance. A strong, positive supervisory relationship is critical to successful supervision, much like a strong, positive therapy relationship is a major predictor of successful therapy outcome (e.g., Bernard & Goodyear, 2004; Muse-Burke, Ladany, & Deck, 2001; Patton & Kivlighan, 1997). The supervisory relationship has been described as the primary means through which the supervisee becomes involved in supervision and the goals of supervision are achieved (Muse-Burke et al., 2001). Although the supervisory relationship has been variously defined in the literature, one of the most frequently utilized definitions has been in terms of the supervisory working alliance, which involves three aspects: (a) mutual agreement and understanding between the supervisor and supervisee of the goals of supervision; (b) mutual agreement and understanding of the tasks of tasks of supervision; and (c) the emotional bond between the supervisor and supervisee (i.e., mutual liking, caring, and trusting) (Bordin, 1983; Muse-Burke et al, 2001). There are many reasons for the centrality of the supervisory working alliance for effective supervision to occur. For example, a strong supervisory alliance provides supervisees with the safety to discuss issues that that are sensitive, personal, embarrassing, threatening, confusing, or reflect poorly on their work as well as their struggles, doubts, and fears about competency (e.g., Ladany et al., 1996; Muse-Burke et al., 2001). In these areas of struggle and self-doubt, a strong supervisory alliance can provide the supervisee with a safe holding environment where the supervisee can work 36 through the issues, take risks, and increase their confidence (Muse-Burke et al., 2001; Bernard & Goodyear, 2004; Ladany, Ellis, & Friedlander, 1999). A strong supervisory working alliance may allow for the times when supervisors must necessarily be challenging or provide corrective feedback about supervisee deficits (e.g., Hoffman, Hill, & Freitas, 2005; Ladany & Melincoff, 1999; Muse-Burke et al., 2001; Abbott & Lyter, 1998). Overall, the supervisory working alliance may have a long?term impact on the supervisee in terms of professional identity and the extent to which they internalize supervisors? teaching and modeling (Knapp and VandeCreek, 2006; Wulf & Nelson, 2000), and is positively related to supervisee satisfaction with supervision (Ladany et al., 1999; Worthen & McNeill, 1996) and counseling self-efficacy (Efstation et al., 1990). Supervisory focus on strengths may strengthen the supervisory working alliance, in particular by enhancing the emotional bond between the supervisor and supervisee by making the supervisee feel more supported by and trusting of the supervisor (e.g., Briggs & Miller, 2005). Particularly for beginning trainees, focusing on strengths may contribute to factors associated with good supervisory relationships such as warmth, acceptance, respect, understanding, trust, and creation of an atmosphere of experimentation and allowance for mistakes (e.g., Ladany, Hill, Corbett, & Nutt, 1996; Allen, Szollos, & Williams, 1986; Worthen & McNeill, 1996). In addition, the use of the attractive supervision style/consultant supervisory role (Friedlander & Ward, 1984; Bernard & Goodyear, 2004), which characterizes some of the strength-focused supervision theories such as solution-focused supervision (e.g., Wetchler, 1990), has been shown to predict all three components of the supervisory working alliance: task, bond, and goals (Ladany, Walker, & Melincoff, 2001; Syzmanski, 2003). In addition, focusing on supervisees? 37 deficits in a constructive way may strengthen the working alliance, particularly the task and goal components. Supervisees are likely aware that they have deficits and areas for improvement in their clinical work, and look to their supervisors as experts who can help them overcome these problems (Abbott & Lyter, 1998; Barnett et al., 2001). To the extent that supervisors help supervisees work on their deficits in supervision with the goal of eventually overcoming the deficits and becoming better therapists, supervisors may be upholding their mutual agreement with supervisees about the tasks and goals of supervision. In addition, supervisees may feel a stronger bond to supervisors who make an active effort to help them improve. Decreases supervisee evaluation anxiety and facilitates evaluation. The supervisory relationship is inherently power-imbalanced because of the ethical responsibility of the supervisor to evaluate the supervisee as well as the greater experience and status of the supervisor (e.g., Bernard & Goodyear, 2004; Vasquez, 1992; Harrar, VandeCreek, & Knapp, 1990). Consequently, supervisees may feel anxiety over how they are being evaluated and shame over being required to expose themselves and their work to scrutiny and evaluation (Ladany et al., 1996; Hahn, 2002; Gould & Bradley, 2001). Both anxiety and shame can detract from supervisees? ability to learn and perform; for example, supervisees might be preoccupied during counseling sessions by fear of criticism and negative evaluation in supervision (Bernard & Goodyear, 2004; Briggs & Miller, 2005), or may use impression management strategies (e.g., selective disclosure or non-disclosure of information about self and clinical work; Ladany et al., 1996) which serve a protective function, yet may detract from the overall goals of supervision (Ward, Friedlander, Schoen, & Klein, 1985). Also, in response to supervisee evaluation anxiety, 38 supervisors may hold back on providing important and necessary critical feedback out of concern for the supervisee?s self-esteem and confidence, fear of damaging the supervisory relationship, or desire to avoid unpleasant conflict (Robiner, Fuhrman, & Ristvedt, 1993; Ladany & Melincoff, 1999; Abbott & Lyter, 1998; Hoffman et al., 2005). However, it is an ethical imperative for supervisee incompetence to be corrected so that supervisees do not harm or provide substandard treatment to clients (Vasquez, 1992; Harrar et al., 1990; Timm & Blow, 1999). A focus on supervisee strengths may ease supervisee evaluation anxiety and shame, and thereby facilitate the provision of constructive feedback and evaluation. Communicating belief in supervisees? strengths, especially when giving negative feedback, may help supervisees feel supported, validated, and secure in the fact that their supervisors believe they are capable overall of successful outcomes (Larson, 1998; Lent et al., 1998; Briggs & Miller, 2005). By making the supervisory atmosphere seem less risky and decreasing fear of appearing incompetent, supervisees may feel greater permission to be honest about their struggles and be able to take in evaluative feedback with less defensiveness, assuming the feedback is meant to help (Briggs & Miller, 2005; Ladany et al., 1996). In addition, supervisees expect and want constructive negative feedback from their supervisors, and may experience a lack of such feedback from their supervisors as insincere, unhelpful, inaccurate, and perhaps indicative of incompetence or a distortion in the supervisor?s perspective (Bernard & Goodyear, 2004; Gould & Bradley, 2001; Abbott & Lyter, 1998; Kadushin, 1992; Chur-Hansen & McLean, 2006). Supervisees generally recognize that they possess deficits, and look to their supervisors as experts to help them 39 address these deficits (Abbott & Lyter, 1998; Barnett et al., 2001). Also, without corrective feedback, supervisees can make inaccurate assumptions about their competency, fail to make necessary improvements, and be surprised when negatively evaluated at the end of supervision (Abbott & Lyter, 1998; Harrar et al., 1990; Vasquez, 1992). Summary of Rationale for Focusing on Supervisee Strengths and Deficits In conclusion, supervisory focus on supervisee strengths and constructive focus on deficits may each benefit supervisees? professional functioning and development both directly and indirectly. In terms of direct benefits to supervisees, each may enhance supervisees? sense of themselves as therapists (i.e., counseling self-efficacy and professional identity), increase competence through improved learning and positive self- fulfilling prophecy, build resilience against the stresses of professional training, and help supervisees work toward achieving their highest potential as therapists. Supervisees may also benefit indirectly from their supervisors focusing on their strengths as well as deficits through the possible benefits of each to supervision process and outcome (e.g., strengthened working alliance, decreased evaluation anxiety and facilitation of the supervisor evaluation, improved satisfaction with supervision), resulting in an improved supervisory learning environment. Unfortunately, very little empirical literature exists to support these theorized benefits. The next section of this review discusses empirical literature related to supervisory focus on strengths and constructive focus on deficits. Empirical Research Related to Strengths Focus and Deficits Focus Empirical research investigating constructs related to supervisory focus on strengths and deficits was found within two bodies of supervision literature: the feedback 40 literature and the solution-focused supervision literature. The supervision feedback literature may be informative to review given that positive and negative feedback are methods supervisors use to focus on supervisee strengths and deficits, respectively (e.g., Heckman-Stone, 2003; Abbott & Lyer, 1998). The solution-focused supervision literature, consisting of only two empirical studies that could be located (Triantafillou, 1997; Koob, 2002), is reviewed because of the explicit focus on supervisee strengths in the solution-focused model. Although the current investigation is framed within the context of positive psychology, no prior empirical studies on supervision could be located within the positive psychology literature. A small amount of empirical research on positive psychology in therapy has been conducted; one recent study that informed the conceptualization of the current study is reviewed here as an example of this body of research (Harbin, 2006). Feedback Research Research on feedback in supervision has investigated how supervisees value positive, constructive negative, and balanced positive and negative feedback; the positive impact of these types of feedback on supervisees (e.g., increased counseling self- efficacy); and the difficulties experienced by supervisors especially in providing constructive negative feedback. In the following discussion of research on feedback in supervision, I first discuss studies investigating how supervisees value and perceive benefit from positive feedback. Next, I discuss research investigate how supervisees value and perceive benefit from constructive negative feedback; supervisees? perceptions of and problems associated with insufficient constructive negative feedback; supervisors? difficulties with providing 41 constructive negative feedback; characteristics and methods of providing constructive negative feedback; and problems associated with critical (as opposed to constructive) negative feedback. Finally, I discuss studies documenting how supervisees value and perceive benefit from balanced positive and negative feedback. Overall, the empirical supervision feedback literature is sparse; thus, only a few studies are reviewed in each section. Positive Feedback. A few studies have documented how supervisees value and perceive benefit from positive feedback. For example, in an exploratory pilot study of trainee preferences for feedback and evaluation in clinical supervision in a small convenience sample (N = 40) of counseling, clinical psychology, and masters counseling graduate student supervisees, Heckman-Stone (2003) asked trainees four open-ended questions about specific experiences with feedback and evaluation and general characteristics of good and poor use of feedback and evaluation by supervisors. Open- ended responses were content analyzed by only one rater, a methodological limitation of the study. Notwithstanding this limitation, analysis of the open-ended supervisee responses found that supervisees wanted their supervisors to provide more positive feedback and communication of confidence in their abilities. Similarly, in a very exploratory qualitative investigation of eight second-year doctoral students in clinical psychology, Talen and Schindler (1993) found that supervisees considered supervisors? positive regard and validation of their strengths to be the most helpful supervision strategies for helping them overcome anxiety, feel comfortable and trusting in supervision, and meet their supervision goals. The comfort, trust, and meeting of goals described by these supervisees seem similar to the supervisory 42 working alliance, although given the qualitative nature of this study, no validated measures of supervisory working alliance were used. Other research has also demonstrated positive outcomes associated with positive feedback. Daniels and Larson (2001) used an experimental analogue design to explore the effects of bogus performance feedback on counseling self-efficacy and counselor anxiety. Participants (N = 45; 87% female; 83% White) were masters? level trainees in counseling and clinical psychology, counselor education, and MFT training programs. Following a 10 minute mock counseling session with a client actor, trainees randomly received either positive or negative bogus feedback on their performance. In the positive feedback condition, trainees were told that they had performed very well compared to others, while in the negative feedback condition, they were told they had performed poorly compared to others. After receiving the feedback, trainees? counseling self- efficacy and state anxiety were assessed, and compared to their pre-session levels of counseling self-efficacy and anxiety. Supporting the researchers? hypotheses, positive feedback enhanced trainees? counseling self-efficacy from pre- to post-test, while negative feedback diminished their counseling self-efficacy. In addition, as predicted, positive feedback decreased trainees? state anxiety from pre- to post-test, while negative feedback increased anxiety. The major strength of the Daniels and Larson (2001) study included the experimental design which allows for the inference of a causal relationship between type of feedback (positive or negative), and counseling self-efficacy and anxiety. In addition, the use of a validated measure of counseling self-efficacy (COSE; Larson et al., 1992) to assess the impact of each type of feedback was a strength of the study. However, the 43 extent to which the results generalize to real-life supervision settings may be limited given the analogue nature of the design and lack of a supervision relationship in which the feedback was delivered. In addition, the positive and negative feedback given to supervisees was purposely exaggerated to make the conditions as distinct from each other as possible for the sake of the experimental manipulation; thus, the feedback may not have been realistic (particularly the potent critical negative feedback, discussed further in the following section on negative feedback). The Daniels and Larson (2001) study showed how positive feedback may improve counseling self-efficacy in the short-term (immediate) sense and was limited in its external validity. In a more externally valid, yet very exploratory, investigation of the long-term impact of positive feedback, six licensed psychologists were retrospectively interviewed about their internship supervisors? contributions to their post-internship growth and professional development (Wulf & Nelson, 2000). Experiences with invested, affirming supervisors who offered positive feedback contributed to positive trainee long- term growth. Although it is unclear the extent to which these results are generalizable (given that they represent the experiences of a few supervisees and given the lack of rigorous qualitative research methodology), the researchers concluded that supervisory affirmation may encourage supervisees? growth even once the supervisory relationship has ended. The above studies may be limited in generalizability in several respects, including the use of small, non-diverse, or non-random samples and the use of analogue and qualitative methodologies. Also, with the exception of the Daniels and Larson (2001) study, these studies did not use validated measures of supervisee development or 44 supervision process and outcome to assess the impact of positive feedback. Most of the research is best viewed as exploratory. Nevertheless, all the studies seem to support the conclusion that supervisees desire and perceive benefit from positive feedback from their supervisors in terms of their short-term and long-term professional functioning (e.g., increased counseling self-efficacy, decreased anxiety, and continued growth). The research does not assess as clearly whether the positive feedback benefits supervision process and outcome. Constructive Negative Feedback. Other research has documented the perceived benefits of constructive negative feedback, supervisees? desire for such feedback, and supervisees? perception that they do not receive enough of it. In an exploratory pilot study about the impact of constructive critical feedback on supervisees? learning, the vast majority of both social work supervisees? (n = 38) and supervisors? (n = 43) agreed that such feedback can have a positive effect on supervisees? learning, specifically in terms of personal growth (insight, self-confidence, self-awareness); correction of deficits (learning from mistakes and turning deficits into strengths); and enhanced development of skills (Abbott & Lyter, 1998). In other words, constructive critical feedback also seems to contribute to supervisees? professional functioning and growth in terms of learning and skill enhancement. In an anecdotal survey of supervisees? perceived strengths and weaknesses of clinical supervisors, Kadushin (1992) found that supervisees considered their supervisors? major weaknesses to be overly positive feedback, a lack of criticism, and constructive criticism not delivered until the end of the supervisory relationship, when it is too late to correct skills deficits. Similarly, in a very exploratory pilot investigation in which 15 trainee psychiatrists and 21 supervisors were interviewed about 45 characteristics of helpful feedback and good versus poor supervision, supervisees wanted more constructive negative feedback and reported problems with a lack of such feedback; for example, it made them skeptical of their supervisors and did not help correct their deficits (Chur-Hansen & McLean, 2006). While the latter two studies do not use a validated measure of satisfaction with supervision to assess how it is impacted by constructive negative feedback, both studies seem to suggest that a lack of constructive negative feedback may decrease satisfaction with supervision. Consistent with supervisees? perceptions that they do not receive enough constructive negative feedback are several findings that supervisors tend to offer such feedback infrequently. For example, in a case study of a single supervision dyad, Friedlander, Siegel, and Brenock (1989) found that supervisors? feedback was primarily interpersonal, global, and positive, and mainly focused on counselor?s behavior with client, with only a few negative statements. McCarthy, Kulakowski, and Kenfield (1994) surveyed licensed psychologists (N = 232; response rate 45%) about their current experiences in supervision, including the frequency of different techniques they perceived their supervisors to use. Like Friedlander et al. (1989), McCarthy et al. (1994) found that the most frequently used supervision technique was support/encouragement, while one of the least frequently used techniques was confrontation. However, the supervisees in the McCarthy et al. (1994) study were licensed psychologists rather than trainees, so it is unclear whether these results generalize to the supervision of trainees. In addition, it is unclear whether supervisors actually use support/encouragement so frequently and confrontation so infrequently, given that self-report methods (from the supervisee?s perspective), rather than behavioral observation methods, were used. 46 One reason for the relative lack of constructive negative feedback in supervision may be that supervisors have difficulties providing such feedback. For example, Ladany and Melincoff (1999) found that 98% of supervisors of graduate student counselors admitted to withholding some feedback from their supervisees, most often pertaining to negative reactions to their trainees? professional performance in counseling and in supervision. Supervisors? rationale for not disclosing this feedback was the anticipation of a negative reaction from their supervisees. In the small, exploratory pilot study of trainee psychiatrists and their supervisors described previously, supervisors described their reluctance to give and lack of skill in providing constructive negative feedback; confusion about the role of supervisor in providing negative feedback; concerns about damaging the supervision relationship; and fear of legal action, especially when providing difficult providing negative feedback (Chur-Hansen & McClean, 2006). A consensual qualitative research (CQR) study of 15 counseling center supervisors of predoctoral interns in psychology found that supervisors withheld or only indirectly gave difficult feedback when they perceived that the feedback might potentially be injurious to supervisee, that the supervisee was defensive or closed to the feedback, or that the feedback might seriously strain the supervisory relationship. These supervisors seemed to prioritize minimizing negative outcomes over maximizing positive outcomes (Hoffman et al., 2005). Given supervisees? desire for constructive negative feedback, and supervisors? difficulty in providing it, it is helpful to examine research investigating how to provide negative feedback in ways that promote supervisees? growth and minimize harm to supervisees and the supervisory relationship. Strategies defined by social work 47 supervisees and supervisors in the Abbott and Lyter (1998) study for the effective and constructive use of criticism included: criticizing only in ways that promote personal growth and allow trainees to use criticism to their own benefit; creating a positive learning environment where it is safe and normal to make mistakes, and critical feedback is to be expected; conveying appreciation, encouragement, warmth, and a non-punitive, non-judgmental attitude; sandwiching negative feedback between positive feedback (i.e., discussing areas needing improvement in the context of supervisee strengths/competencies); avoiding accusatory comments or ultimatums; only using criticism supported by the professional literature; offering criticism tentatively (i.e., as an opinion, not a fact); focusing on specific behaviors that can be changed; modeling self- critique; and following criticism with brainstorming to identify methods and means for removing deficiencies and achieving competencies (Abbott & Lyter, 1998). Abbott and Lyter (1998) also discussed previous literature suggesting that constructive criticism may be distinguished from ?flaw finding? in that it is improvement-oriented, protects the supervisee?s self-esteem, and communicates a spirit of helping (Weisinger & Lobsenz, 1989). While constructive negative feedback is desired by supervisees and seems to have a positive impact on supervisees and supervision, critical negative feedback seems to have a detrimental impact. For example, the Daniels and Larson (2001) experimental analogue study described previously found that negative feedback decreased trainee counseling self-efficacy and increased anxiety. This negative feedback (telling the trainee they had performed poorly compared to others) was purposely exaggerated for the experimental manipulation comparing it to positive feedback, and it is possible that real- 48 life supervisors might hesitate to give such potent critical negative feedback to their supervisees for fear of damaging the supervisee or the relationship (e.g., Abbott & Lyter, 1998). Although these conditions may not have been realistic of supervision, a strength of the Daniels and Larson (2001) experimental study is the ability to conclude that the critical negative feedback caused the negative effects observed in trainees. Two studies conducted in more realistic supervision settings corroborated the results of the Daniels and Larson (2001) study. Social work supervisors and supervisees speculated that that non-constructive criticism, characterized by participants as harsh and not having the goal of growth promotion could result in damage to supervisees? self- esteem and self-confidence; decreased motivation and discouragement with clinical work; and the impediment of learning and growth (Abbott & Lyter, 1998). Similarly, in a survey of psychology doctoral students (N = 126), Ramos-Sanchez et al. (2002) found that supervisees? perceptions of their supervisors being too critical or harsh with feedback, as well as unsupportive, were related to weaker supervisory alliances and lower satisfaction with supervision. However, this survey suffered from a low response rate (28%), thus introducing the possibility of selection bias in the sample and calling into question the generalizability of the results to psychology trainees/supervisees. Also, this survey did not use a validated measure of satisfaction with supervision, although it did use a validated measure of the supervisory working alliance (Baker, 1990). The negative impact of excessive criticism may be long-term, according to the study described previously in which six licensed psychologists were retrospectively interviewed about their internship supervision experience (Wulf & Nelson, 2000). These former supervisees described how experiences with critical, non-affirming supervisors 49 who failed to validate their strengths exerted a long-term negative impact on their professional development and contributed to significant emotional and developmental difficulties. In sum, a small body of mostly exploratory empirical literature on constructive negative feedback in supervision suggests that it benefits supervisees? professional functioning in terms of learning, skill enhancement, and correction of deficits (e.g., Abbott & Lyter, 1998). Supervisees also seek more constructive negative feedback than they often receive (e.g., Chur-Hansen & McLean, 2006), perhaps because supervisors often avoid giving constructive negative feedback for reasons such as fear of negative consequences or a lack of skills or awareness of methods of providing constructive negative feedback that can have positive consequences for supervisee growth and development (e.g., Ladany & Melincoff, 1999; Hoffman et al., 2005). By contrast, non- constructive (i.e., harsh or critical) negative feedback seems to have a detrimental impact on supervisee professional functioning/development and supervision process and outcome (e.g., the supervisory relationship, satisfaction with supervision) (e.g., Daniels & Larson, 2001; Wulf & Nelson, 2000). Like the research on positive feedback, these studies may be limited in generalizability in several respects, including the use of small, non-diverse, or non-random samples (several very exploratory pilot studies); low response rates in survey studies; and the use of analogue methodology. Also, like the research studies on positive feedback, some of the studies did not use validated measures of supervisee development or supervision process and outcome to assess the impact of constructive negative feedback. 50 Balanced Positive and Negative Feedback. Given research documenting how supervisees value both positive and constructive negative feedback and reasons why each is important and helpful, it is not surprising that the empirical literature also suggests that supervisees value and perceive benefit from balanced positive and negative feedback. As described previously, Heckman-Stone (2003) asked trainees (N = 40) from a convenience sample of graduate students from three training programs (counseling doctoral and masters, clinical doctoral) four open-ended questions about experiences with feedback and evaluation and characteristics of good and poor use of feedback and evaluation by supervisors. Analysis of the open-ended supervisee responses found that the most frequently endorsed concern (22% of the sample) was whether their supervisors provided balanced positive and negative feedback. The study did not define what supervisees meant by balanced (e.g., high levels of both positive and negative, low levels of both, etc.). Similarly, Chur-Hansen and McLean (2006), in a small, exploratory pilot study in which 15 trainee psychiatrists and their 21 supervisors were interviewed about characteristics of helpful feedback and good versus poor supervision, a recurrent theme of interviews was trainees? desire for balanced positive and negative feedback (again, not specifically defined) and supervisors? lack of skill in giving it (i.e., supervisors gave too much positive and not enough constructive feedback). The authors recommended the ?compliment sandwich? or ?positive-negative-positive? approach, in which supervisors comment on supervisees? strengths, identify a specific problem, and finish with a motivating or esteem enhancing statement. Likewise, a majority of both social work 51 supervisees and supervisors emphasized the need for an approach to supervision that identifies both competencies and deficiencies (Abbott & Lyter, 1998). Lehrman-Waterman and Ladany (2001) assessed supervisees? perception of supervisors? provision of effective feedback (including individual items about feedback being balanced positive and negative, systematic, clear, timely, etc.) using a seven-item subscale from an evaluation practices measure (Evaluation Practices in Supervision, Inventory or ESPI) developed and validated in the study. In a national sample of counseling and clinical psychology masters? and doctoral students of all training levels (N = 274; response rate = 35%), supervisee perceptions of effective feedback (which included a single item about balanced positive and negative feedback) were found to predict a positive supervisory working alliance, satisfaction with supervision, and supervisee perceptions of supervisor contributions to counseling self-efficacy. Thus, trainees seemed to feel more connected to, satisfied with, and influenced positively by supervisors who provided them with effective feedback, which includes some balanced positive and negative feedback, among other components. The Lehrman-Waterman and Ladany (2001) study must be interpreted within the context of several limitations. First, the construct of balanced positive and negative feedback, was assessed via only one item on a seven item effective feedback scale. Thus, it may have been operationalized too narrowly, and it is unclear to what extent balanced positive and negative feedback (in isolation of the other components of effective feedback comprising the subscale) related to supervision process and outcome variables. Second, the ex post facto design of the study does not allow for teasing out the directionality of the relationships between variables. In other words, it cannot be 52 determined whether balanced positive and negative feedback (as a component of effective feedback) improved the working alliance, or whether such feedback was given in supervision relationships that already had strong working alliances. Finally, the low return rate of 35% introduces the possibility of a selection bias among participants. Summary of Feedback Literature. The small but growing empirical supervision feedback literature corroborates the positive psychology movement?s position regarding the need for a focus on strengths as well as a focus on deficits. The feedback literature suggests that supervisees value and perceive benefit from positive feedback (Talen & Schindler, 1998), constructive (but not critical) negative feedback (Abbott & Lyter, 1998), and a balance of positive and negative feedback (Chur-Hansen & McLean, 2006), particularly in terms of their professional development and functioning. The feedback literature has less empirical evidence for how positive or constructive negative feedback impact supervision process and outcome (e.g., the supervisory working alliance, satisfaction with supervision). Given the exploratory nature of many of these studies and methodological problems with some of the feedback research, more empirical research is needed to better understand the impact of positive feedback and constructive negative feedback on supervisee development and supervision process and outcome. Solution-Focused Supervision Two published empirical studies have investigated the impact of solution-focused supervision, with its explicit focus on supervisee strengths, on supervision and therapy process and outcome (Koob, 2002; Triantafillou, 1997). Triantafillou (1997) conducted a small, exploratory pilot study of the effects of solution-focused supervision on mental health counselors and their clients at a children?s 53 mental health agency. The solution-focused model was defined in the Triantafillou study as including techniques described by solution-focused theorists (e.g., Wetchler, 1990), such as focusing on supervisees? (and clients?) strengths and successes through compliments and focusing on exceptions to problems, among others. The solution- focused supervision training program involved four three-hour weekly sessions for both supervisors and staff counselors on the solution-focused model and the application of solution-focused techniques in both supervision and therapy. Between training sessions, supervisors (n = 14) were instructed to apply solution-focused supervision techniques to their work with their supervisees (i.e., staff counselors; n = 10) in individual and team supervision, while staff counselors were instructed to apply solution-focused techniques to their work with their clients. All supervisors and supervisees in the sample received training and were told to implement the solution-focused model; in other words, there was no random assignment to treatment conditions comparing the solution-focused model to the model already in place at this agency. The authors do not specify what model was already in place the agency. The effectiveness of implementing the solution-focused model was measured in terms of supervision variables (i.e., supervisees? job satisfaction, supervisors? use of solution-focused techniques such as compliments to supervisees, satisfaction with and desire for further training in solution-focused model among both supervisors and supervisees) and therapy variables (i.e., staff counselors?/supervisees? use of solution- focused techniques, clients? satisfaction with treatment, and reduction in client symptomatology). The researchers did not use validated measures to assess these supervision and therapy variables, relying instead on anecdotal reports (with the 54 exception of the measure of client symptomatology, which was measured in terms of average number of serious incidents per resident and utilization of medications to control behavioral outbursts). The majority of supervisors (70%) anecdotally reported implementing strength- focused techniques in supervision (e.g., compliments to supervisee) as a result of the training in the solution-focused model. Supervisors reported that they preferred the solution-focused model to the model of supervision (not defined in the study) typically used at their agency (31% rated it ?clearly superior? to the typically used model). Anecdotally, both supervisors and supervisees indicated that the model created a more positive atmosphere and improved morale in the agency, increased counseling staff confidence and enthusiasm, decreased staff anxiety, made supervision time more productive, and allowed supervisors to identify and build more on supervisee strengths. In addition, the majority of counselors (70%) reported an increased implementation of strength-building techniques with their clients, including compliments and focus on resources, as a result of their training in a solution-focused model. The solution-focused treatment model was found to have largely positive effects on client treatment (e.g., improved interactions with clients) in terms of anecdotal reports by counselors, as well as more objective measures of client improvement such as reduced symptomatology. Triantafillou (1997) concluded that solution-focused supervision had a positive impact on both supervision and therapy process and outcome, although these results should be regarded as preliminary given the exploratory nature of the study. Given its small sample size, non-standardized implementation of solution-focused techniques, failure to use validated measures of solution-focused techniques or supervision and 55 therapy process and outcome variables, and lack of comparison of the solution-focused model to the ?traditional? model via random assignment with a control group, it is difficult to draw definitive conclusions about the impact of the solution-focused model. In contrast to Triantafillou (1997), who examined the impact of solution-focused supervision on both supervisees and their clients, Koob (2002) restricted the focus of his investigation to supervisee variables. Specifically, Koob (2002) investigated the relationship between solution-focused supervision and the perceived self-efficacy of therapists-in-training. Koob (2002) distinguished solution-focused supervision from ?traditional? supervision models (e.g., psychoanalytic, cognitive-behavioral, developmental). For the purposes of the study, the two supervision approaches were dichotomized: solution-focused supervision was defined has having (a) a deliberate focus on supervisee strengths and successes rather than mistakes, (b) a supervisee development focus rather than a client focus; and (c) the assumption that there is more than one right way to conduct therapy, with ?traditional? approaches defined as having the opposite assumptions, namely (a) a focus on supervisee mistakes, (b) a focus on clients/case management, and (c) a focus on one correct way to conduct therapy. Based on the assumption (which may be erroneous) that ?traditional? supervision contributes to low supervisee counseling self-efficacy, and given that the two approaches have opposing assumptions, Koob (2002) made the following hypotheses: (a) a positive relationship between solution-focused supervision (in terms of belief in its three assumptions, as rated by supervisors) and supervisee counseling self-efficacy (as rated by their supervisees), and (b) a negative relationship between traditional supervision (in terms of belief in its three assumptions, as rated by supervisors) and supervisee-rated counseling self-efficacy. 56 It is important to note that this investigation did not use an experimental design to compare the two supervision approaches, but rather examined correlations between type of supervision (dichotomized as solution-focused versus traditional) and level of supervisee counseling self-efficacy. Participants were a convenience sample of 55 supervisor-therapist dyads. Supervisors were masters? level social workers (95% White, 78% female) with a mean of 7.9 years supervision experience. Supervisees were graduate or undergraduate social work students (91% White, 93% female) with a mean of 8.2 months of field experience. Supervisors completed a measure developed for the study called the Supervisor Opinion Scale containing 30 items (developed on the basis of interviews and feedback from supervisors) and 11 factors consisting of elements posited to characterize the solution- focused (e.g., focus on supervisee strengths and successes, supervisee empowerment, focus on supervisee development) and traditional supervision (e.g., focus on supervisee problems and weaknesses) models. Supervisors rated their agreement with the assumptions of both types of supervision. The internal consistency reliability of the measure was low (.68). Supervisees completed a 30-item measure of their counseling self-efficacy, also developed specifically for this study using field supervisor manuals that outlines the skills and tasks to be mastered by students during their field placement experiences. The internal consistency of this counseling self-efficacy measure was .84. Results supported the study?s hypotheses. Specifically, supervisors with stronger belief in aspects of the solution-focused supervision models (e.g., focus on successes and strengths, supervisee empowerment) tended to have supervisees with more positive counseling self-efficacy. Use of the solution-focused model accounted for 18% of the 57 variance in supervisee counseling self-efficacy. In contrast, none of the aspects of traditional supervision related to supervise counseling self-efficacy. Overall, Koob (2002) concluded that solution-focused supervision enhanced supervisee counseling self- efficacy. The results of the Koob (2002) study must be interpreted within the context of several important limitations. First, their dichotomous conceptualization and operationalization of solution-focused supervision versus ?traditional? supervision may have been flawed, as traditional approaches probably do not necessarily focus exclusively on supervisee problems and mistakes. Additionally, this study was correlational, such that it cannot conclude that strength-focused aspects of solution-focused supervision caused increases in supervisee counseling self-efficacy. Furthermore, the researcher used a non- random (i.e., convenience) sample with limited demographic diversity, thus possibly limiting the generalizability of the findings. This study also used unvalidated measures, one of which lacked adequate psychometric properties (i.e., the measure of supervisors? agreement with assumptions of solution-focused and traditional supervision had low reliability). The measure may have also pulled for socially desirable responses from supervisors, given that items characterizing the ?traditional? supervision approach may have sounded inherently negative (i.e., focusing on supervisee mistakes, only one right way to conduct therapy, etc.). In addition, Koob (2002) did not assess the extent to which supervisors implemented the strength- or deficit- approaches respectively assumed to characterize the two supervision models, only the extent to which they agreed with the assumptions. Therefore, it is unclear what interventions were actually used, which draws the independent variable (supervision approach) into question. 58 In conclusion, the empirical research findings supporting the strength-oriented interventions characteristic of solution-focused supervision (Triantafillou, 1997; Koob, 2002) should be regarded as tentative, given the exploratory and methodologically limited nature of the only two studies testing the theory. Research on Positive Psychology in Therapy Although there is no empirical research thus far on positive psychology in supervision, there is a growing, though still limited, body of empirical literature on positive psychology in therapy. I focus here on only one study of therapists? work with client strengths as an example of this research because it was a stimulus for the present study (Harbin, 2006). Harbin (2006) developed and validated a new measure, the Inventory of Therapist Work with Client Assets and Strengths (IT-WAS), a self-report measure which examines the degree to which therapists possess a tendency to use strength-based approaches in their therapeutic work. The measure, validated on a large sample (N = 225) of both professional psychologists (51% return rate) and trainees (62% return rate) in psychology doctoral programs, was found to have 3 subscales: Theory of Intervention (16 items reflecting therapists? use of theory, such as positive psychology theory, to explain how and why they utilize client strengths in their therapeutic work; alpha = .93); Assessment of Strengths (11 items reflecting therapists explicit and implicit evaluation of client strengths, e.g., asking clients direct questions about their strengths, giving equal attention to strengths as well as weaknesses when writing reports, and interpreting psychological tests in the context of strengths; alpha = .90); and Supporting Progress (10 items reflecting the degree to which therapists openly focus on the gains clients make in 59 therapy, alpha = .90). Harbin (2006) found that most therapists generally indicated that they conducted strength-based clinical work to a high degree, although several differences were found among therapists with different theoretical orientations. Specifically, humanistic, cognitive-behavioral, and multicultural/feminist theoretical orientations were positively related to therapists? work with client strengths (r = .18, .28, .29, respectively), whereas the psychodynamic theoretical orientation was negatively related to work with client strengths (r = -.22 for Supporting Progress subscale). Harbin (2006) discussed how supervisors can use the measure to help incorporate positive psychotherapy into training, for example by helping supervisees learn to incorporate client strengths into their conceptualizations and interventions. Several limitations of the Harbin (2006) study should be noted. First, the IT-WAS measure focused primarily on working with client strengths, despite its grounding in positive psychology literature which advocates attention to strengths as well as weaknesses (e.g., Snyder et al., 2003). In addition, therapists may have responded to IT- WAS items in a socially desirable manner, reporting the use of asset-based therapy approaches to a higher degree than they actually used in their clinical work, although favorable self-presentation is viewed as an inevitable and acceptable aspect of positive psychology research (Lopez, Snyder, & Rasmussen, 2003). Also, the return rates for participants, low by some statistical standards, introduce the possibility of selection-bias in the sample. Finally, the sample was primarily Caucasian, thus possibly limiting the generalizability of findings to other populations. Conclusions Although positive psychology research has expanded overall, how positive 60 psychology may be applied to supervision and clinical training remains largely unstudied. The clinical and theoretical literature (e.g., Gelso & Woodhouse; Frankland, 2001; Briggs & Miller, 2005; Carlson & Erickson, 2001; Larson, 1998) suggests that supervisory focus on strengths and constructive focus on deficits may each benefit aspects of supervisee development and supervision process and outcome, yet little empirical research has been conducted on supervisors? work with supervisee strengths and deficits. A small body of mostly exploratory empirical research on feedback in supervision suggests that supervisees value both positive feedback (Heckman-Stone, 2003; Talen & Schindler, 1993), constructive negative feedback (Abbott & Lyter, 1998; Chur-Hansen & McLean, 2006; Kadushin, 1992, and balanced positive and negative feedback (Chur-Hansen & McLean, 2006; Lehrman-Waterman & Ladany, 2001) from their supervisors, and that each is associated with benefits to supervisee professional functioning/development (e.g., counseling self-efficacy). More empirical research is needed on how positive feedback and constructive negative feedback relate to supervision process and outcome (e.g., the supervisory working alliance, satisfaction with supervision). Also, some preliminary empirical research on the impact of solution-focused supervision suggests that incorporating a focus on supervisee strengths (via solution-focused techniques) into supervision may predict enhanced supervisee counseling self-efficacy and other aspects of supervision and therapy process and outcome (Triantafillou, 1997; Koob, 2002). To facilitate more empirical research on supervisory focus on strengths and constructive focus, a psychometrically valid measure of these constructs is needed. 61 Chapter 3 STATEMENT OF THE PROBLEM The clinical and theoretical supervision literature suggests that supervisory focus on strengths and constructive focus on deficits would each benefit aspects of supervisee development and supervision process and outcome. Unfortunately, however, only a small amount of empirical research, primarily within the supervision feedback and solution- focused supervision literatures, has been conducted to support these predictions (e.g., Heckman-Stone, 2003; Abbott & Lyter, 1998; Koob, 2002; Triantafillou, 1997). Moreover, despite the overall expansion of the positive psychology movement and its increasing focus on applied domains, positive psychology has not yet been applied to the domain of supervision, thus leaving untapped its potential to enhance the effectiveness of supervision practices (Linley & Joseph, 2004). Finally, the field of counseling psychology, historically committed to the identification and enhancement of human strengths, has also largely neglected to apply a focus on strengths to empirical research aimed at improving supervision (e.g., Gerstein, 2006; Gelso & Fassinger, 1992; Gelso and Fretz, 2001). For all of these reasons, it seemed like an optimal time to conduct empirical research examining supervisory focus on strengths and deficits in supervision. To facilitate the empirical investigation of these two constructs, psychologists must first be able to effectively assess the processes and interventions used by supervisors to work with supervisees? strengths and deficits with a psychometrically sound measure (Lopez, Snyder, & Rasmussen, 2003). Hence, the overarching purpose of the present study was to take a step toward generating more research on positive psychology within the context of supervision by developing and validating a measure of 62 the constructs of focus on supervisee strengths and constructive focus on deficits from the supervisee perspective, the Supervisory Focus on Strengths and Deficits Inventory- Supervisee Form (SUPSAD-S). The SUPSAD-S incorporated two constructs, focus on strengths and constructive focus on deficits, stemming from positive psychology theory as well as clinical and empirical supervision literature. The positive psychology movement has explicitly noted how its focus on strengths does not mean ignoring weaknesses (e.g., Lopez et al., 2003; Lopez & Snyder, 2003). Rather, positive psychologists claim that the focus in psychology has typically been exclusively on weaknesses, and therefore a focus on strengths should be added to the typical focus on weaknesses (Wright & Lopez, 2002; Seligman & Csikszentmihalyi, 2000; Sheldon & King, 2001). In fact, positive psychology theorists posit that focusing exclusively on strengths while ignoring weaknesses would be Pollyanna-ish, failing to address the very real problems that people have or resulting in only superficial, short-lived gains (Lampropoulos, 2001; Seligman, 2002; Gelso & Woodhouse, 2003; Keyes & Lopez, 2002). The supervision literature likewise highlights the importance of focusing on supervisee strengths as well as deficits. Focusing on strengths may empower supervisees toward growth, competence, greater confidence, and achievement of their potential as clinicians (e.g., Briggs & Miller, 2005), while focusing constructively on deficits may reduce or eliminate deficits and other problematic aspects of their professional functioning (e.g., Abbott & Lyter, 1998). It is important to emphasize the constructive nature of supervisory focus on deficits as defined and operationalized in the current study. I was most interested in the processes and interventions supervisors use to attend to supervisee deficits in positive, 63 empowering, self-esteem preserving/enhancing ways, as opposed to harsh, critical, or punitive ways of correcting supervisee mistakes and weaknesses (Abbott & Lyter, 1998; Weisinger & Lobsenz, 1981). In other words, the measure developed for the current study was intended to be a positive psychology measure overall, based on the idea that supervisors can focus on both supervisee strengths and deficits in positive, growth- enhancing ways. Supervisory focus on strengths and constructive focus on deficits may be assessed from various perspectives, including the supervisee, the supervisor, or outside raters. While all of these perspectives are important and would likely shed light on different aspects of the processes of strengths and deficits focus, the aim of the present study was to develop a measure from the supervisee?s perspective, in which supervisees were asked to rate the extent to which their supervisors focused on their strengths and deficits. The rationale was that supervisees? perceptions of supervisory focus may be more important than what the supervisors actually do or report doing, per se. The important role of supervisees? perceptions of their supervisors and the supervision process are supported by well-established theories from social psychology, as well clinical and empirical supervision literature. For example, social psychology theorists posit that self-concept is influenced not by what others people actually think of an individual, but what the individual perceives them as thinking (Mead, 1934). Applied to the supervision context, supervisees? self-concept as therapists might be more influenced by their perceptions of how their supervisors see them, more so than any ?reality? of how their supervisors actually see them. 64 The supervision literature points to a similar conclusion. For example, some supervisees might project their own self-criticism onto their supervisors, therefore viewing them as more critical and problem- or deficit-focused than they actually are (e.g., supervisee transference; Bernard & Goodyear, 2004; Lewis, 2001). Moreover, the feedback given by supervisors regarding positive and negative aspects of supervisees? performance may not be the same as what is actually taken in by supervisees, who may selectively attend to and differentially weight aspects of the feedback as a function of their personalities and counseling self-efficacy (Larson, 1998). Larson (1998) distinguished between the objective supervisory environment (what actually happened in supervision or what the supervisor actually said) and the perceived supervisory environment (supervisees? and supervisors? perceptions of what happened), noting that supervisees? and supervisors? perceptions often differ greatly. Given the importance of supervisee perceptions, the measure developed in the current study focused on supervisees? perceptions of supervisory focus on strengths and constructive focus on deficits. Another consideration here is that supervision research has been hindered by measurement problems, specifically the preponderance of measures lacking acceptable psychometric properties, as well as measures designed for other contexts (e.g., therapy) and adapted to supervision with only minimal attention to possible changes in their meanings (e.g. substituting the word ?supervisee? for ?client?, ?supervisor? for ?therapist?, etc.). These adapted instruments then have unknown psychometric properties, and include roles and concepts that may or may not be relevant to supervision. Consequently, one of the current trends in supervision research is the development of 65 measures of supervision process and outcome that are (a) psychometrically valid/sound, and (b) written specifically for the supervision context (Ladany & Muse-Burke, 2001). In the current investigation, attempts were made to avoid these common problems in supervision measurement research by generating items specifically for the supervision context and through the use of established methods for developing psychometrically sound instruments (Dawis, 1987; Walsh & Betz, 2001). Given that the primary focus of the current study was instrument development, and given the lack of previous research on working with strengths and deficits in supervision, exploratory factor analysis was used (Kahn, 2006; Floyd & Widaman, 1995). Thus, the first purpose of this study was to explore the factor structure of the Supervisory Focus on Strengths and Deficits Inventory-Supervisee Form (SUPSAD-S). However, it is important to note that item generation for this new measure was guided by the goal of attaining a two factor solution: focus on strengths and constructive focus on deficits, as informed by the positive psychology and supervision literatures. Although I have conceptualized both factors as positive psychology constructs and theorized that they may both relate to aspects of supervision process and outcome in positive ways, it is possible that they may not relate in identical ways to other supervision variables (e.g., there may be differences in the magnitude or even direction of the relationships, particularly if the strengths and deficits constructs do not correlate strongly with each other). Hence, in all of the following hypotheses, the focus on strengths and constructive focus on deficits constructs are examined separately. The potential utility of the SUPSAD-S was based on its having adequate reliability and validity estimates. Initial construct validity was assessed through 66 exploratory factor analysis, with a two factor solution reflecting supervisory focus on strengths and constructive focus on deficits being predicted. Reliability was determined by both internal consistency and two week test-retest reliability of predicted focus on strengths and constructive focus on deficits subscales; these methods have been shown to be an appropriate means of estimating reliability (Dawis, 1987). At least moderate test- retest reliability estimates were expected for the predicted SUPSAD-S subscales, given that the measure assessed supervisees? perceptions of the same supervisor (over the entire course of their work with that supervisor), once and again two weeks after initial administration. Hypothesis #1: A measure (SUPSAD-S) of supervisory focus on strengths and constructive focus on deficits having adequate construct validity and reliability can be created. Hypothesis #1a: The SUPSAD-S will have construct validity, as demonstrated by a factor structure containing 2 factors, focus on strengths and a constructive focus on deficits. Hypothesis #1b: The focus on strengths subscale of the SUPSAD-S will have adequate internal consistency reliability. Hypothesis #1c: The constructive focus on deficits subscale of the SUPSAD-S will have adequate internal consistency reliability. Hypothesis #1d: The focus on strengths subscale of the SUPSAD-S will have at least moderate two-week test-retest reliability. 67 Hypothesis #1e: The constructive focus on deficits subscale of the SUPSAD-S will have at least moderate two-week test-retest reliability. Assuming that the SUPSAD-S measure yielded subscales for focus on strengths and constructive focus on deficits factors as predicted, the remaining validity hypotheses relate to the two subscales separately. For determining concurrent validity of a new instrument, one should compare the scale with the closest matching scale designed to assess the same construct (Dawis, 1987). After a thorough review of the supervision literature, I could not locate a scale related closely enough to the SUPSAD-S to correspond with this recommendation. The closest measure I could locate, a measure of effective feedback and evaluation practices in supervision (EPSI; Lehrman-Waterman & Ladany, 2001), was not a similar enough construct to use to assess concurrent validity and would not have justified the additional survey length and possible participant fatigue. Hence, the concurrent validity of the SUPSAD-S was not assessed. Dawis (1987) indicated that the utility of a measure can be seen as its ability to predict some practical criterion, or convergent validity. One means of assessing the convergent validity of this measure of supervisory focus on strengths and constructive focus on deficits was to test whether each subscale predicted supervisees? satisfaction with supervision. Research suggests that satisfaction with supervision is related to the provision of positive feedback as well as supervisor support, affirmation, validation, and interpersonal characteristics such as warmth and sensitivity (e.g., Worthen & McNeill, 1996; Allen et al., 1986). In addition, reinforcement of positive, desired, or correct 68 clinical behaviors may enhance supervisees? learning, shaping their skills toward closer and closer approximations of competence (e.g., Bradley & Gould, 2001; Catania, 2001; Follette & Callaghan, 1995). To the extent that learning to conduct therapy is one of the major purposes of supervision (Bernard & Goodyear, 2004), supervisees may feel more satisfied with supervision when such learning is facilitated. Therefore, Hypothesis #2a: Evidence will be found for the convergent validity of the focus on strengths subscale of the SUPSAD-S as demonstrated by a positive correlation with supervisee ratings of satisfaction with supervision. Supervisees are also likely aware that they have deficits and areas for improvement in their clinical work, and look to their supervisors as experts who can help them reduce or overcome these problems, become more competent, and provide better care to their clients (e.g., Abbott & Lyter, 1998; Barnett et al., 2001; Bernard & Goodyear, 2004). Therefore, supervisees may feel more satisfied with supervision when they receive constructive negative feedback that helps them address their problems, and cite problems with supervision experiences in which such feedback was lacking or delivered too late (Gould & Bradley, 2001; Kadushin, 1992; Chur-Hansen & McLean, 2006). Therefore, Hypothesis #2b: Evidence will be found for the convergent validity of the constructive focus on deficits subscale of the SUPSAD-S as demonstrated by a positive correlation with supervisee ratings of satisfaction with supervision. 69 Relationships between supervisory focus on strengths and constructive focus on deficits and the quality of the supervisory working alliance were also used to assess the convergent validity of the SUPSAD-S. By providing supervisees with a safe holding environment where they can struggle, take risks, and increase confidence and self- efficacy, the supervisory working alliance is the primary means through which the supervisee becomes involved in supervision and achieves the learning goals of supervision (e.g., Muse-Burke et al., 2001; Bordin, 1983; Efstation et al., 1990). Supervisory focus on strengths may strengthen the supervisory working alliance, in particular by enhancing the emotional bond between the supervisor and supervisee by making the supervisee feel more supported by and trusting of the supervisor (e.g., Briggs & Miller, 2005; Talen & Schindler, 1993). Factors found to contribute to a strong supervisory working alliance include warmth, acceptance, respect, creation of an atmosphere allowing for experimentation and mistakes, and the consultant supervisory role, which characterizes some of the strength-focused supervision theories (e.g., Bernard & Goodyear, 2004; Allen et al, 1986; Ladany et al., 1996; Ladany et al., 2001; Wetchler, 1990). These predictors of the supervisory working alliance are theoretically similar or related to a supervision approach characterized by a focus on strengths. Hence, Hypothesis #3a: Evidence will be found for the convergent validity of the focus on strengths subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee ratings of the supervisory working alliance. Supervisees seek constructive negative feedback from their supervisors that will help them address their deficits, learn from their mistakes, and become better therapists 70 (Abbott & Lyter, 1998). To the extent that supervisors help supervisees work on their deficits in supervision with the goal of eventually overcoming the deficits and becoming better therapists, supervisors may be upholding their mutual agreement with supervisees about the tasks and goals of supervision (Gould & Bradley, 2001). In addition, supervisees may feel a stronger bond to supervisors who make an active effort to help them improve (Barnett et al., 2001). Hence, Hypothesis #3b: Evidence will be found for the convergent validity of the constructive focus on deficits subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee ratings of the supervisory working alliance. The construct of counseling self-efficacy was also used to assess the convergent validity of the focus on strengths and constructive focus on deficits subscales of the SUPSAD-S. Previous research has found that supervisors? provision of positive feedback increases supervisees? counseling self-efficacy (Daniels & Larson, 2001). Further, some theorists recommend providing support, encouragement, experiences that maximize chances of success and mastery, reminders of resources and previous successes, and communication of confidence in supervisees as ways of increasing counseling self- efficacy, particularly as interventions when supervisees lack confidence and security in their abilities as they learn the complex enterprise of therapy (e.g. Larson, 1998; Larson & Daniels, 1998; Lent et al., 1998; Briggs & Miller, 2005; Gelso & Woodhouse, 2003). Such research lent support to the next hypothesis: 71 Hypothesis #4a: Evidence will be found for the convergent validity of the focus on strengths subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee counseling self-efficacy. In addition, challenging supervisees constructively (e.g., gentle and respectful of limits) to improve on existing skills may motivate supervisees to push themselves in ways that lead to growth, achievement of goals, and higher counseling self-efficacy when goals are met (e.g., Gould & Bradley; Blocher, 1983; Larson, 1998; Abbott & Lyter, 1998). Therefore, Hypotheses #4b: Evidence will be found for the convergent validity of the constructive focus on deficits subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee counseling self-efficacy. A valid measure of supervisees? perceptions of supervisory focus on strengths and deficits should also be unrelated to certain constructs, thus reflecting discriminant validity (Dawis, 1987; Walsh & Betz, 2001). The construct chosen to assess discriminant validity in the present study was public self-consciousness (Fenigstein, Scheier, & Buss, 1975). Public self-consciousness is the degree to which people tend to be concerned about the way in which they present themselves to others. In a previous scale development study, public self-consciousness was found to be unrelated, as hypothesized, to the positive psychology construct of hope (Snyder, Harris, et al., 1991). It seems that supervisees? tendency to be concerned about the way they present themselves to others should also be unrelated to the degree to which they perceive their supervisors as focusing on their 72 strengths or on their deficits. In other words, the measure developed in the current study ought to be independent of any possible supervisee tendency to report only positive things about their supervisors, stemming from their perception that the researchers are looking for this type of positive response or because the items seem to pull for socially desirable responses. Even though items were written to minimize the pull for impression management or socially desirable responses, the nature of the constructs of supervisory focus on strengths and constructive focus on deficits may be inherently value-laden, like most positive psychology constructs (Lopez, Snyder, & Rasmussen, 2003); therefore, relation to supervisee public self-consciousness must be assessed. Hence, Hypothesis #5a: Evidence will be found for the discriminant validity of the focus on strengths subscale of the SUPSAD-S, as demonstrated by a non-significant correlation with public self-consciousness. Hypothesis #5b: Evidence will be found for the discriminant validity of the constructive focus on deficits subscale of the SUPSAD-S, as demonstrated by a non-significant correlation with public self-consciousness. Exploratory Research Questions Several exploratory research questions were investigated. The first question addresses the notion of a balanced focus on strengths and deficits. Originally, I had conceptualized the SUPSAD-S as a measure of balanced focus on strengths and deficits, given that it was written to include both items reflecting a focus on strengths and items reflecting a constructive focus on deficits. Upon further consideration and consultation with the dissertation committee, it seemed to me that focusing on both strengths and 73 deficits in supervision may not be equivalent to facilitating a balanced focus on strengths and deficits in supervision. The concept of balanced focus may imply some optimal amount of each, for example an equal focus on strengths and deficits or a certain optimal ratio of strengths to deficits (e.g., 2:1 strengths to deficits as in the notion of the ?compliment sandwich?), that is associated with desirable supervision process and outcomes (e.g., Lehrman-Waterman & Ladany, 2001; Chur-Hansen & McLean, 2006; Abbott & Lyter, 1998). Due to the difficulty of operationalizing and measuring the construct of balanced focus on strengths and deficits, this idea was not considered any further as a primary focus of the present study and no formal hypotheses regarding balanced focus were made. However, I attempted some exploratory investigation of balanced focus using the SUPSAD-S measure, by examining the amount of unique variance in satisfaction with supervision, the supervisory working alliance, and supervisee counseling self-efficacy accounted for by the unique contributions of focus on strengths and constructive focus on deficits. I wondered whether the two factors would contribute approximately equal amounts of unique variance, or whether (and by how much) one factor might be a stronger contributor than the other factor to overall variance in supervision process and outcome (e.g., given the notion of ?compliment sandwich,? strengths might be a stronger predictor than deficits). Such estimates of unique contribution to variance would offer an (albeit crude) estimation of the balance or ratio of strengths focus to deficits focus that may be associated with the most positive supervision outcomes. Hence, Research Question #1a: How much unique variance in satisfaction with supervision is accounted for by focus on strengths and constructive focus on deficits, respectively? 74 Research Question #1b: How much unique variance in the overall supervisory working alliance is accounted for by focus on strengths and constructive focus on deficits, respectively? Research Question #1c: How much unique variance in overall supervisee counseling self- efficacy is accounted for by focus on strengths and constructive focus on deficits, respectively? I considered a second question related to the issue of amount of supervisory focus on strengths relative to focus on deficits; specifically, I wondered whether, on the whole, supervisees would perceive more strengths than deficits, more deficits than strengths, or equal amounts of both. The positive psychology literature suggests a disproportionate focus on deficits and neglect of strengths in psychology (e.g., Seligman & Czikszentmihalyi, 2000), suggesting that we might find a greater supervisory focus on deficits as rated by supervisees in the sample. However, positive psychology literature also offers the idea that in clinical settings therapists likely focus extensively on client strengths (without necessarily having been formally trained to do so) as a ?deep strategy? of effective therapy (Seligman, 2002); supervisors might likewise focus on strengths as a ?deep strategy? of effective supervision, particularly with therapists-in-training who are still developing their skills and confidence as therapists (e.g., Briggs & Miller, 2005; Stoltenberg et al., 1998). Therefore, given that the entire sample was trainees (of different training levels, but still novices by virtue of being in training), we might expect to find a greater supervisory focus on strengths than deficits. 75 Research Question #2: How much do supervisees perceive that their supervisors focus on strengths as compared to deficits? In addition, I conducted three sets of exploratory analyses examining relationships between SUPSAD-S subscales and demographic variables of supervisees and their supervisors. Given the lack of previous research related, no specific hypotheses were formed. I was first interested in how focus on strengths and constructive focus on deficits would relate to supervisee training level (e.g., year in program, number of clinical hours). A greater focus on strengths might be expected for supervisees in the very early stages of training, as supervisors might perceive these very novice supervisees as needing validation of their strengths in order to enhance confidence and reduce anxiety (e.g., Briggs & Miller, 2005); hence, a negative correlation between focus on strengths and training level might be expected. Alternately, a positive correlation between focus on strengths and training level might be expected, given that supervisees likely develop and manifest more strengths (which can then be pointed out by supervisors) as they gain more experience (e.g., Stoltenberg et al., 1998). In addition, a greater constructive focus on deficits might be expected for more advanced trainees who are secure in their basic skills but looking to learn more and address specific competence issues (e.g., their multicultural competence with specific populations, Ladany et al., 1997); hence, a positive correlation between constructive focus on deficits and training level might be expected. Alternately, a negative correlation between constructive focus on deficits and training level might be expected, as supervisees may have fewer and fewer deficits to focus on as they gain 76 experience, having resolved their deficits through training and experience (e.g., Bernard & Goodyear, 2004; Bradley & Kottler, 2001). Research Question #3a: How do the subscales of focus on strengths and constructive focus on deficits relate to supervisee training level? In addition, given philosophical differences among theoretical orientations regarding the role of strengths and deficits, I was interested in how focus on strengths and constructive focus on deficits would relate to the theoretical orientations of supervisees and their supervisors. Research Question #3b: How do the subscales of focus on strengths and constructive focus on deficits relate to supervisee and supervisor theoretical orientations? Finally, I wondered how focus on strengths and constructive focus on deficits would relate to professional specialization, specifically counseling versus clinical psychology, given philosophical differences between the fields about the role of strengths and deficits (e.g., Gelso & Fretz, 2001). Research Question #3c: Are there differences in mean levels of focus on strengths and constructive focus on deficits for clinical versus counseling psychology trainees, or for trainees whose supervisors hold a degree in counseling psychology versus those whose supervisors hold a degree in clinical psychology? 77 Chapter 4 METHODS Participants Participants (N = 204 for factor analysis, N = 201 for all other analyses; all percentages of sample that follow refer to the sample of 201 who completed the demographic questionnaire) were a national sample of masters? and doctoral level graduate students in a variety of mental health professions, including counseling psychology Ph.D. (n = 78; 38%) and masters? (n = 17); clinical psychology Ph.D. (n = 45; 22%) and Psy.D. (n = 7); counselor education Ph.D. (n = 3) and masters? (n = 10); social work (n = 10); college student personnel (n = 7); and mental health counseling (including rehabilitation counseling and school counseling, n = 24). Participants were recruited from graduate programs and counseling center internships (pre-doctoral psychology and social work) at several universities and a VA hospital in diverse regions of the country. In order to qualify for the present study, participants had to either be currently receiving supervision, or have received supervision within the last semester, of their clinical work. One hundred sixty-four (81.6%) participants were female and thirty-seven (18.4%) were male. With respect to race/ethnicity, 150 (74.6%) were European American/Caucasian, 17 were Asian or Pacific Islander, 16 were African-American, 14 were Hispanic/Latino/a, 2 were Middle Eastern, 1 was Native American, and 1 was biracial or multiracial. The gender and racial composition of this sample is comparable to the gender and racial composition of APA accredited psychology doctoral programs in counseling and clinical psychology (76% female; 75% White) as well as APA accredited 78 pre-doctoral internships (73% female; 76% White), suggesting that the sample is representative of the population of graduate trainees in psychology (APA, 2005). The mean age was 28.5 (SD = 6.05) and ages ranged from 22-57. Additionally, participants were asked to rate their belief in and adherence to eight theoretical orientation clusters on a 5 point scale (1 = low and 5 = high). The following mean ratings emerged: integrative 4.21 (SD = 1.04), humanistic/client-centered 3.96 (SD = .91), cognitive-behavioral 3.58 (SD = 1.11), multicultural 3.54 (SD = 1.06), solution-focused 2.87 (SD = 1.22), psychodynamic-psychoanalytic 2.86 (SD = 1.35), feminist 2.76 (SD = 1.23), and narrativist/constructivist 2.13 (SD = 1.06). Thus, the sample was sample diverse theoretically but favored integrative and client-centered theoretical orientations. Participants had a median of 250 direct clinical hours and ranged from first through final (i.e., internship) years of their graduate programs, with approximately 50% of the sample in the second and third years of their graduate programs. Regarding their work with the supervisors they rated in the survey, participants had met with these supervisors for a median of 18 sessions or 5 months. Several different supervision modalities (individual, n = 192; group, n = 87 group; other e.g., dyadic, n = 10) were represented. Participants worked with their supervisors as part of a variety of different training experiences (practicum, n = 113; externship, n = 25; pre-doctoral psychology internship, n = 29; social work internship, n = 10; other e.g., field placement or graduate assistantship, n = 31) and received supervision for diverse types of clinical experiences (individual emotional-social counseling, n = 161; career counseling, n = 55; group therapy, n = 60; couples counseling, n = 18; family therapy, n = 16; child/adolescent, n = 79 55; supervision-of-supervision, n = 14; supervision of consultation/outreach, n = 19; other e.g., assessment, n = 21). According to information supplied by participants, participants? supervisors held the following degrees (Ph.D., n = 123; Psy.D., n = 12; masters?, n = 67) in a variety of mental health professions (counseling psychology, n = 64; clinical psychology, n = 74; social work, n = 20; counselor education, n = 20; and others such as marriage & family therapy, mental health counseling, college student personnel, n = 24). Participants also estimated their supervisors? belief in and adherence to eight theoretical orientation clusters on a 5-point scale (1 = low and 5 = high). The following mean ratings for supervisors? theoretical orientation emerged: humanistic/client-centered 3.55 (SD = 1.22), integrative 3.51 (SD = 1.27), cognitive-behavioral 3.40 (SD = 1.36), multicultural 3.06 (SD = 1.28), solution-focused 2.91 (SD = 1.38), psychodynamic-psychoanalytic 2.77 (SD = 1.43), feminist 2.41 (SD = 1.22), and narrativist/constructivist 1.97 (SD = 1.05).Thus, participants? supervisors were diverse theoretically but favored client-centered, integrative, and cognitive-behavioral theoretical orientations. Measures In addition to the measure being developed in the current study, the Supervisory Focus on Strengths and Deficits Inventory-Supervisee Form (SUPSAD-S), the following measures were used: the Supervisee Satisfaction Questionnaire (SSQ; Ladany, Hill, Corbett, & Nutt, 1996); the Supervisory Working Alliance Inventory-Trainee Form (SWAI-T; Efstation, Patton, & Kardash, 1990); the Counselor Activity Self-Efficacy Scales (CASES; Lent, Hoffman, & Hill, 2003); the Self-Consciousness Scale (Fenigstein, 80 Scheier, & Buss, 1975); and a demographic questionnaire. Measures are presented in the order in which they appeared in the survey taken by participants. The Supervisory Focus on Strengths and Deficits Inventory-Supervisee Form (SUPSAD-S; see Appendix C) is a self-report measure created for the present study to assess the constructs of supervisor focus on strengths and constructive focus on deficits, from the perspective of supervisees. The original version of the SUPSAD-S used for data collection consisted of 55 items comprising theoretically posited processes and interventions used by supervisors to focus on supervisees? strengths and deficits. Exploratory factor analysis was used to reduce the SUPSAD-S to its final version consisting of 24 items with two subscales, Focus on Strengths (12 items, e.g. ?My supervisor identified areas where I excel as a therapist?) and Constructive Focus on Deficits (12 items; e.g., ?My supervisor attended to my deficits because she/he wanted to see me improve?). Participants (supervisees) indicate the extent to which they agree or disagree that their supervisor focuses on their strengths and deficits, on a scale from 1 to 7 (1 = strongly disagree and 7 = strongly agree), with respect to the entire course of their work with the supervisor. Subscale scores are generated by averaging responses to items after reversing negatively phrased items. Higher subscale scores indicate greater supervisory focus on strengths or constructive focus on deficits, respectively, from the supervisee?s perspective. Description of instrument development procedures, validation, and psychometric properties of the SUPSAD-S can be found in the Results Chapter. The Supervisee Satisfaction Questionnaire (SSQ; Ladany et al., 1996; see Appendix D) is an eight-item self-report measure in which supervisees rate, on a 4-point scale ranging from low (1) to high (4), their satisfaction with various aspects of 81 supervision. Scores range from 8 to 32, with higher scores indicating greater satisfaction. The SSQ has been found to be negatively related to supervisee perceptions of frequency of supervisor ethical violations (Ladany, Lehrman-Waterman, Molinaro, & Wolgast, 1999) and to supervisee nondisclosures involving negative reactions to the supervisor (Ladany et al., 1996). Previous supervision research has shown the internal consistency of the SSQ to range from .96 to .97 (Ladany & Lehrman-Waterman, 1999; Ladany, Lehrman-Waterman, et al., 1999). Internal consistency coefficient alpha in the current sample was .95. The SSQ was used in the current study to evaluate the convergent validity of the SUPSAD-S. The Supervisor Working Alliance Inventory-Trainee Form (SWAI-T; Efstation, Patton, & Kardash, 1990; see Appendix E) is a 19-item self-report measure of supervisees? perceptions of the supervisory working alliance, with two subscales: Rapport (12 items reflecting the rapport or bond between supervisor and supervisee; e.g., ?I feel comfortable working with my supervisor?) and Client Focus (6 items reflecting the process of supervisor and supervisee working together to enhance the trainees? understanding of the client; e.g., ?When correcting my errors with a client, my supervisor offers alternative ways of intervening with that client?). The items are rated on a 7-point Likert scale ranging from almost never (1) to almost always (7). Subscale scores are generated by summing the items and dividing by the number of items on the subscale, whereas total scores are generated by summing all of the items and dividing by 19; therefore, subscale scores as well as the total score can range from 1-7. Higher scores indicate a stronger supervisory working alliance as perceived by the supervisee. Alpha coefficients for the Rapport and Client Focus scales were .90 and .77, respectively 82 (Efstation et al., 1990). Internal consistency coefficient alpha in the current sample was .94 for Rapport subscale, .88 for Client Focus subscale, and .95 for total score. Although Rapport and Client Focus subscales were correlated r = .33 in the measure development sample (Efstation et al., 1990), suggesting the use of subscale rather than total scores of the SWAI-T, previous measure development studies have examined how both SWAI-T subscale and total scores correlate with constructs of interest (e.g., Syzmanski, 2003). Like previous studies, the current study used both subscale and total scores for the SWAI-T, particularly given the high internal consistency for both subscale and total scores and correlation of .72 between subscales in the current sample. Evidence of the validity of the SWAI-T was shown through its relationships to supervisory style and supervisee counseling self-efficacy (Efstation et al., 1990); additionally, the rapport subscale was positively related to satisfaction with supervision (Jackson, 1993). The SWAI-T was used in the current study to evaluate the convergent validity of the SUPSAD-S. The Counselor Activity Self-Efficacy Scales (CASES; Lent, Hill, & Hoffman, 2003; see Appendix F) is a 41-item self-report measure that assesses trainees? self- efficacy over the next week for ?Performing helping skills? (15 items; e.g., ?Open questions--ask questions that help clients to clarify or explore their thoughts or feelings?), ?Managing the counseling process? (10 items; e.g., ?Keep sessions ?on track? and focused?), and ?Handling challenging counseling situations? (16 items; e.g., ?Working effectively with client who is clinically depressed?). Only the first two subscales were used for the present study; the third was dropped in the interest of shortening the length of the entire survey to prevent participant fatigue and attrition. For the total of 25 83 counseling self-efficacy items retained, trainees were asked to rate their confidence in their ability to perform specific tasks or to manage specific scenarios on a 10-point scale ranging from no confidence (0) to complete confidence (9). Scores are generated by averaging items (within subscales for subscale scores or across subscales for total score), therefore yielding scores between 0 and 9 for each of the subscales and the total score, with higher scores reflecting greater confidence. The benefit of the CASES over previous counseling self-efficacy measures is its applicability across a range of developmental levels in counselors. CASES scales were related to positive outcome expectations regarding the counselor role, students' interest in therapy activities, occupational goals in therapy, and negative and positive affect experienced while enacting the counselor role (Lent et al., 2003). Reliability of the CASES is adequate, as internal reliability estimates for the individual subscales ranged from .79 to .94, with an overall alpha of .97. Additionally, scale scores were stable over a 2-week interval, providing evidence of test- retest reliability. Internal consistency coefficient alpha in the current sample was .90 for the Helping Skills subscale, .94 for Session Management subscale, and .95 for total score. The CASES was used in the current study to evaluate the convergent validity of the SUPSAD-S. The Self-Consciousness Scale (SCS; Fenigstein et al., 1975; see Appendix G) is a 23-item measure that uses a 4-point Likert scale, where 1 = extremely uncharacteristic and 4 = extremely characteristic. The SCS contains three subscales that measure different kinds of self-consciousness: Private Self-Consciousness, Public Self-Consciousness, and Social Anxiety. For the present study, only the 7 item Public Self-Consciousness scale was used. The Public Self-Consciousness scale assesses people?s awareness and concern 84 about aspects of the self that others can perceive, and has demonstrated good reliability with a two week test-retest reliability of .84. Internal consistency coefficient alpha in the current sample was .79. The Public Self-Consciousness Scale was used in the current study to assess the discriminant validity of the SUPSAD-S. Demographic Questionnaire. Participants were asked to provide demographic information including their age, gender, race, type of graduate program, year in their program, approximate number of clinical hours, theoretical orientation, approximate number of sessions and length of time working with current supervisor, supervision modality (e.g., individual, group, etc.), supervisor?s highest degree, supervisor?s theoretical orientation (if known), type of training experience they were undergoing in working with the supervisor (e.g., practicum, externship, internship, etc.), and type of clinical work being supervised by the supervisor (e.g., individual, group, career, couples counseling, etc.).See Appendix H. Procedures Participants were recruited for the current study using several methods. First, faculty and program directors of seven graduate programs with whom the author had personal contacts (e.g., graduates of the author?s doctoral program) were contacted by email and asked for assistance with recruiting students from their programs. For five of these programs, contact persons informed students in their programs (n = 249) of the study and forwarded an email containing a cover letter and link to the online survey. For the other two programs, contact persons provided lists of students? emails to the researcher, who then contacted these students (n = 76) individually with an email containing a cover letter and link to the online survey. Second, counseling and clinical 85 psychology doctoral students, counselor education doctoral and masters? students, and pre-doctoral psychology and social work interns at the author?s university (n = 75) were recruited by individual email and provided the link to the online survey. Third, graduate trainees and pre-doctoral psychology interns at sites where the author had a personal contact (e.g., a classmate or graduate of the author?s doctoral program working at the site) were recruited by individual email provided by the personal contact, and provided the link to the online survey (n = 57). While none of these recruiting methods ensure a random sample, they were chosen in order to increase sample size, maximize return rate, and generate a diverse sample in terms of training level, mental health discipline, and geographic region. However, response rate differed by recruitment method. For those participants contacted individually by email, the response rate for completed surveys was 63% (131 out of 208). For participants who were forwarded the recruitment email by the contact person (i.e., training director or faculty), response rate was 29% (73 out of 249). Also, response rate within the local sample was 73% (55 out of 75) compared to 39% for the non-local sample (149 out of 382). (I was able to track response rate by recruitment method and locality by checking participants? electronic signatures against my database of participant names, email addresses, and sites. Participants who were forwarded the email by their training directors were not in my database since I was not provided with their names and e-mail addresses). Data was collected in this study through the use of an online survey on surveymonkey.com (see Appendices A-H; appendix pages are in the order they appear in the web survey). The choice of internet data collection was made given the many 86 advantages of internet research, including the ease of obtaining large and nationwide samples, lower costs, security features, design options and ease of administration, and the fact that results tend to be equivalent to paper-and-pencil survey methods, including the factor structure and psychometric properties of instruments in measure development research (e.g., Gosling, Vazire, Srivastava, & John, 2004; Herrero & Meneses, 2006). One common limitation of internet research includes problems in obtaining accurate response rates, as unknown numbers of individuals could potentially receive emails that link them to the study (Gosling et al., 2004). The recruitment methods used in the current study (i.e., contacting prospective participants individually and keeping track of the number of prospective participants who were forwarded the email by contact persons at various graduate programs) were aimed at preventing this problem with calculating response rate by allowing the researcher to keep an accurate count of how many individuals were recruited to participate. Another limitation of online research is that confidentiality cannot be completely guaranteed; in electronic submissions, there is always a small chance that information could be intercepted and read by a third party (Gosling et al., 2004). In the current study, informed consent included the acknowledgment that confidentiality could not be completely guaranteed if participants chose to complete the survey online. Also, given the focused nature of participant recruitment (i.e., the study was not widely advertised) and the probably limited value of the data to a third party, it seems unlikely that the data was a target for interception. The email to potential participants, either sent directly or forwarded by contact persons (e.g., training directors or faculty of graduate programs), contained a cover letter emphasizing the significance of the research, the relatively short amount of time required 87 to complete the measures (i.e., approximately 15 minutes), and the URL link to the web survey. Also, in two week and one month follow-up recruitment emails, participants were encouraged to fill out the entire survey if they decided to participate (this was added to the follow-up recruitment emails after noting occasional survey drop-out among early participants). Participants were told that the current study was investigating supervisees? perceptions of the processes used by their supervisors to work with their strengths and weaknesses as clinicians. See Appendices I and J. The URL link to the study directed potential participants to a page asking them to indicate whether they met the two participation criteria for the study: (a) trainee or intern in a mental health professional program, and (b) currently receiving supervision of their clinical work, or have received supervision of their clinical work within the last semester (see Appendix A). If participants checked the box indicating ?yes? to meeting both participation criteria, they were directed to the Informed Consent page (Appendix B). If participants checked the box indicating ?no? (i.e., they did not meet BOTH criteria), they were directed to a page asking them to indicate which of the two criteria they did not meet, and subsequently directed to a page explaining why they were not eligible to participate and exiting them from the survey (Appendix A). Having participants enter the survey and then designate whether or not they meet the participation criteria was done to provide information about reasons for non-participation that could inform interpretation of participant response rate. Participants directed to the Informed Consent page were instructed to read the information on the page, provide their electronic signature, and click on a box indicating whether or not they agreed to participate in this research. Indicating agreement to 88 participate directed participants to the survey measures, while indicating non-agreement to participate exited participants from the survey. See Appendix B. Participants who were directed to the survey were then told to select the supervisor they would use in responding to all the questions throughout the survey. For data collected during the break between Fall 2006 and Spring 2007 academic semesters and at the beginning of the Spring 2007 semester, instructions told participants to select a supervisor they had worked with for at least half a semester during the most recent semester (i.e., their Fall 2006 supervisor). This was done to ensure that participants had enough experience with their supervisors (i.e., preferably the entire Fall semester) to be able to evaluate important aspects of supervision process and outcome investigated in the study. Thus, participants who completed the survey during this time period (n = 162, 79.4% of the sample of 204), may have been evaluating supervisors they had already terminated with at the end of the Fall semester, unless they were working with supervisors for the entire academic year or longer. Also, given the winter break, some may not have received any supervision during the week when they completed the survey. Data collection continued into Spring semester, and after approximately a month into the semester, prospective participants were instructed to respond to the survey with respect to their current (Spring) supervisors. Although many of these participants would not have worked with current supervisors for as long as they had worked with their Fall supervisors (unless it was the same supervisor) by that point in Spring semester, the possibility of forgetting aspects of Fall supervision as they became more immersed in Spring supervision might have contaminated the data. There were 42 participants in this 89 latter group who completed the survey with respect to current/Spring supervisors (20.6% of the sample of 204). All participants were instructed to select the supervisor they had met with most recently (whether Fall or Spring), if they had more than one supervisor. This was done to eliminate the possibility of bias introduced by supervisees selecting especially good or especially poor supervisors/supervision experiences. Reminders to participants to answer all measures with respect to the same supervisor were included throughout the survey. Participants completed measures in the following order: SUPSAD-S (55 items divided into two separate web pages), SSQ, SWAIT-T, CASES Part I (Helping Skills), CASES Part II (Session Management), SCS, and demographic questionnaire. A note at the bottom of each web page told participants how many more pages of the survey remained. To increase response rate, two-week and one month follow-up reminder emails containing the URL link to the study were sent to participants whose email addresses were known, and also to contact persons (i.e., faculty and training directors of graduate programs) who were asked to forward the reminder emails to students in their programs (Appendix J). After an initial recruitment effort via the methods described previously along with two-week and one month follow-up reminders, 125 complete/usable surveys had been returned, yielding a response rate of 27% (125 out of 457 total recruited). The target sample size for this study was between 200-250 participants, per the rule of thumb of a 5:1 ratio of participants per item for exploratory factor analysis (e.g., Floyd & Widaman, 1995; Gorsuch, 1983). Therefore, to increase sample size as well as response rate, the procedure for recruiting participants was amended to include small incentives to encourage participation. Paper copies of a follow-up recruitment letter were distributed to 90 prospective participants (n = 219) initially recruited from five graduate programs via faculty or training director contact persons (described previously as first recruitment method). These follow-up paper recruitment letters, nearly identical to the follow-up reminder emails with an additional mention of the incentive enclosed, were printed on Psychology departmental letterhead, personally signed by both the researcher and her dissertation advisor, and contained a $1 incentive (see Appendix K). Additionally, a sub- sample of local participants (n = 41) received similar paper recruitment letters containing candy as an incentive to participate. Thus, 260 prospective participants received incentives to participate, and 197 did not receive incentives to participate. A response rate within each group cannot be calculated because participants did not indicate anywhere in the survey whether or not they had received an incentive to participate. However, the provision of incentives resulted in 79 more complete surveys and increased overall response rate from 27% to between 44-46%. Further explanation of the overall response rate follows. In total, out of 457 participants recruited, 255 entered the survey (56%). Out of these 255 participants, 18 participants indicated that they did not meet one or both of the participation criteria (3 were not currently trainees in mental health programs, 13 were not currently in supervision, and 2 did not specify which of the 2 criteria they did not meet). This left 237 participants who were eligible to participate (52%). Out of these 237 participants, 36 participants provided incomplete data sets (i.e., dropped out of the survey), although 3 of these 36 participants completed the entire SUPSAD-S measure before dropping out of the survey. Thus, there were 204 usable surveys for factor analysis on the SUPSAD-S but only 201 usable surveys available for subsequent analyses (i.e., 91 examining convergent and discriminant validity of the SUPSAD-S). Hence, the overall return rate for this study ranges from 44% (201 out of total 457 recruited) to 45% (204 out of the total 457 recruited) to 46% (201 or 204 out of 439 participants total, which excludes the 18 who were ineligible to participate). Test-Retest Reliability To gather evidence for the test-retest reliability of the SUPSAD-S, a sample of 29 graduate trainees in mental health programs was gathered. Approximately two weeks after completing the initial survey, a sub-sample of 34 participants (primarily local) were sent an email asking them to complete a second online version of the survey that would take approximately 5-10 minutes to complete (see Appendix L). Following the URL link contained in this email brought participants to the test-retest survey, which contained the SUPSAD-S as well as a few demographic items. Instructions in the test-retest survey told participants to respond to SUPSAD-S items with respect to the same supervisor they had in mind when they completed the previous version of the survey two weeks ago. Thirty participants filled out the test-retest survey, although one survey was incomplete. Therefore, the usable return was 85% (29 of 34). Twenty-two participants (76%) were female and 7 (24%) were male. In regard to race/ethnicity, 20 were European- American/Caucasian (69%), 5 were Asian American (17%), 2 were African American (7%), and 2 were Hispanic Latino/a (7%). Twenty-one (72%) were in a counseling psychology doctoral program, 5 (17%) were in a clinical psychology doctoral program, 2 were in a MSW program, and 1 was in a counselor education doctoral program. The mean age of participants was 28 years old (SD = 4.8). 92 Due to the timing of the first portion of data collection during the winter break between academic semesters, some test-retest participants completed the initial survey retrospectively (with respect to Fall supervisors), and then retrospectively again two- weeks later. In these cases, the participants probably did not meet with the supervisor during the two-week period. 93 Chapter 5 RESULTS In this section, I first describe instrument development procedures and the results of the exploratory factor analysis on the SUPSAD-S. Reliability of the SUPSAD-S was obtained by using estimates of internal consistency and test-retest reliability over a two week period. Additionally, initial validity of the SUPSAD-S was assessed by correlations of its subscales with satisfaction with supervision, the supervisory working alliance, supervisees? counseling self-efficacy, and public self-consciousness. Finally, results of several exploratory research questions are reported. Hypothesis 1: A measure (SUPSAD-S) of supervisory focus on strengths and constructive focus on deficits having adequate construct validity and reliability can be created. Development of the SUPSAD-S The method of item generation and scale construction used for creating the SUPSAD-S measure in this study was based on examples of recent scale development (e.g., Gelso et al., 2005) as well as Dawis? (1987) recommendations. Development and validation of the SUPSAD-S encompassed three phases. This first phase involved generating an initial pool of items on the basis of a comprehensive review of existing empirical, theoretical, and clinical literature related to supervisors? work with supervisee strengths and deficits. In addition, two focus groups were conducted with members of the target population (i.e., graduate student trainees/supervisees in mental health professions): one with 4 doctoral students in 94 counseling psychology, and a second with 3 doctoral students in clinical psychology. The perspectives of both counseling and clinical psychology graduate students were sought purposely, given philosophical differences between the two applied fields regarding the role of human strengths and deficits (e.g., Gelso & Fretz, 2001; Super, 1955). In these focus groups, participants were asked open-ended questions related to their personal experiences in supervision and how their supervisors worked with their strengths and deficits. For example, questions were posed regarding the processes and interventions their supervisors used to focus on their strengths and on their deficits, the impact of each focus on their development as counselors, and which strength- and deficit-based interventions they found most helpful and most hindering in their work. On the basis of these interviews and literature review, an initial pool of 220 items was created to measure two theorized constructs: (a) focus on strengths (e.g., ?My supervisor identified areas where I excel as a therapist?) and (b) constructive focus on deficits (e.g., ?My supervisor attended to my deficits because she/he wanted to see me improve?). Items reflecting a focus on strengths were based on the processes and interventions used by supervisors to identify and enhance supervisees? strengths, such as positive feedback, reinforcement of progress and appropriate clinical behaviors, emotional support, and encouragement. Items reflecting a constructive focus on deficits were based on the processes and interventions used by supervisors to address supervisees? deficits in an empowering, growth-enhancing manner, such as challenge, constructive negative feedback, normalizing and helping the supervisee learn from mistakes and difficulties, and highlighting areas for improvement. Both positively worded (e.g., ?My supervisor focused on my strengths?) and reverse worded (e.g., ?My 95 supervisor did NOT reinforce my strengths?) items were generated for both strengths and deficits constructs. Initial Content Validity of the SUPSAD-S The second phase of the development of the SUPSAD-S involved establishing the initial content validity of the measure. Five counseling psychology doctoral students as well as the researcher?s advisor (a professor of counseling psychology) were consulted to reduce the initial pool of items by selecting items that best captured the constructs of interest (i.e., focus on strengths and constructive focus on deficits). Each person was given definitions of these constructs and asked to select the most representative 30 strengths items (specifically, to choose 15 regular and 15 reverse scored) and the most representative 30 deficits items (15 regular, 15 reverse scored). Compiling the items selected by members of this group yielded 83 strengths items and 57 deficits items that were selected by at least one person. The researcher and her advisor reviewed these selected items, eliminating items judged to be poorly written, confusing, redundant, or lacking clear relevance to the concepts of focus on strengths and constructive focus on deficits. Also, items were reworded (based on feedback and edits from the group) to enhance clarity and reduce overlap between hypothesized constructs. This process reduced the pool of items to 110. Next, the process of back translation (Dawis, 1987) was used to further refine the pool of items and to select the items most related to the two hypothesized constructs, focus on strengths and constructive focus on deficits. In this process, two senior level psychology undergraduates who had experience in counselor training issues (i.e., from a helping skills course) were asked to sort the 110 items (in random order) into the two 96 categories, after reading definitions, descriptions, and examples of supervisory focus on strengths and constructive focus on deficits. Items categorized correctly were to be retained; those categorized incorrectly by either of the raters were retained only if they made theoretical sense and were not redundant with other items. The back translation task resulted in the elimination of only 5 items due to incorrect categorization, along with 8 additional items eliminated on the basis of feedback about confusing wording or redundancy. Other edits and wording suggestions from the back translation raters were also incorporated into the item pool, now reduced to 97 items. The 97 item measure was then pilot tested on 16 doctoral students in counseling psychology at the researcher?s university to gather preliminary data on score distribution and to gain overall feedback on its clarity and ease of administration. Pilot participants generally reported finding the measure clear and understandable, while also noting redundancy of many items. After reversing negatively worded items, mean item scores were 5.19 (SD = .82) for strengths items, 5.22 (SD = .76) for deficits items, and 5.24 (SD = .81) overall for all 97 items, on a 7-point scale (where 1 = strongly disagree and 7 = strongly agree). Thus, overall participants? responses were slightly negatively skewed (i.e., skewed toward the higher end of the scale) on both strengths and deficits subscales. Although from a psychometric perspective, items or scales yielding lower mean item scores would have been preferable, these mean item scores were considered acceptable because of the view that on average, strength and deficit focus in actual supervision practice should be on the positive or stronger side. In addition, items with the highest scores contained relevant content that otherwise had satisfactory psychometric properties (e.g., standard deviations and no increase in alpha when deleted). Internal consistency 97 reliability of strengths items (.98) and deficits items (.90) were extremely high, and eliminating items did not increase the alpha. Several criteria were used to eliminate items, including item means that were extremely high or negatively skewed (6.0 or above on a 7-point scale), items with very low standard deviations (less than .2) suggesting little variability among responses, and items with low item-total correlations. Also, a counseling psychologist with expertise in measure development and psychometrics was asked for feedback on the 97 item measure, in particular for suggestions on making the items designed to represent strengths and deficits as distinct from each other as possible. Items judged by this counseling psychologist as likely to load on both strengths and deficits factors due to wording issues, or as not achieving their intended function, were also eliminated. Taken together, these methods resulted in the elimination of 41 items, leaving a pool of 55 items, including 30 strengths items and 25 deficits items. Internal consistency reliability was still high on these 55 items in the sample of 16 pilot participants (.95 for strengths items, .88 for deficits items, .95 for all 55 items), with similar (but slightly less skewed than the 97-item version) means and standard deviations (M = 5.14, SD = .80 for strengths items; M = 5.04, SD = .74 for deficits items; M = 5.13, SD = .79 for all items). This 55-item version of the SUPSAD-S measure was used for the third phase of measure development, which involved data collection from a local and national sample that was used to explore its factor structure and to gather evidence for its reliability and validity. The number of items was purposely kept rather large given that a sufficient number of variables per factor (at least 6) are needed to generate stable factors (Kahn, 2006). Also, 20 reverse scored items were included in the SUPSAD-S to reduce the 98 possibility of participant response bias. Instructions developed for the SUPSAD-S (Appendix C) specified that supervisees should fill out the measure with respect to the supervisor they met with most recently (if they had more than one); this was done to eliminate the potential confound of supervisees selecting only especially good or especially poor supervisors/supervision experiences. Also, participants were told to fill out the measure with respect to the entire supervision experience with that particular supervisor. I had initially considered asking participants to refer to the most recent session as this would provide more concrete, specific feedback for supervisors and would be less subject to memory distortions of the supervisee. However, I speculated that most supervisors could not use all of the strengths and deficits focused interventions in any given session; that the strength or deficit focus of supervision might vary from session to session, depending on the supervision needs for that session; and that some supervision sessions might not be representative of the entire supervision experience. Therefore, I was interested in supervisees? global perceptions of their supervisors? strengths and deficits foci. Exploratory Factor Analysis of SUPSAD-S Hypothesis #1a: The SUPSAD-S will have construct validity, as demonstrated by a factor structure containing 2 factors, focus on strengths and a constructive focus on deficits. To assess the construct validity and explore the underlying factor structure of the SUPSAD-S, exploratory factor analysis was used, even though item generation for this new measure was guided by the goal of a two factor solution (focus on strengths and constructive focus on deficits). We chose to use exploratory factor analysis (EFA) rather than confirmatory factor analysis (CFA) for this purpose for the following reasons. First, 99 EFA seemed more appropriate given the newness of the construct being investigated and the lack of solid theory and empirical research; CFA is most useful in later stages of measure development to refine and improve measures (Kahn, 2006; Floyd & Widaman, 1995). Second, the use of CFA on Likert-scaled items treated as continuous measures, as is the case for the SUPSAD-S, is not recommended due to the likelihood of producing fit statistics that may inaccurately represent the degree of true model fit (Kahn, 2006). Third, lengthy questionnaires (i.e., more than 5-8 variables per proposed factor, as in the SUPSAD-S) often do not result in satisfactory factor solutions when items are submitted to CFA (Kahn, 2006). Finally, estimates based on EFA using the principal-axis method tend to generalize well to those obtained using CFA techniques, and EFA has a long history of being used for both exploratory and confirmatory purposes (Kahn, 2006; Floyd & Widaman, 1995). Hence, EFA guided by a two factor solution was used. While acknowledging the possibility of a different factor structure emerging, we used EFA primarily as a tool to assist in selecting the best and most representative items for the proposed subscales of focus on strengths and constructive focus on deficits. In consultation with several statisticians and psychologists with expertise in measure development, at least two ways of conducting exploratory factor analysis on the SUPSAD-S measure were recommended, each with advantages and disadvantages. The first way involved conducting EFA on the entire sample of 204 participants, while the second way involved splitting the data into two halves, with one half being the replication/confirmation sample used to check the factor stability of the factors identified in the other half. The advantage of using the first method (EFA on the whole sample) is the larger sample size, which decreases the possibility of finding relations in the data by 100 chance; on the other hand, using only one sample for EFA without any confirmation or replication of findings can also result in findings by chance alone and the stability of the factor structure would be unknown. Results would then await confirmation in future research conducted using the SUPSAD-S measure. The advantage of the split half method, therefore, is the increased confidence in findings if they replicate from the first sample to the second (confirmation) sample. However, the smaller sample sizes of each of the split halves (n of 102) could be problematic as some recommendations call for at least 200-250 participants for EFA (Floyd & Widaman, 1995; Gorsuch, 1983). In sum, one could argue for either of the two methods being more appropriate for the present study. We decided on the first method, conducting EFA on the entire sample because of its greater statistical power, while acknowledging the need for confirmation of the factor stability of the SUPSAD-S in future research. Preliminary Analyses. Missing values for the 55 item version of the SUPSAD-S were analyzed using pattern analysis techniques in SPSS 14.0 on the 204 participants who completed the SUPSAD-S. Results suggested no pattern of missing data among the scales. Therefore, data imputation to fill in missing data values was conducted for the 204 participants on the SUPSAD-S using maximum likelihood estimation (EM). The EM technique makes minimal assumptions about the data, and uses an EM algorithm to impute missing data. Separate EM procedures were conducted for SUPSAD-S items written to represent focus on strengths and items written to represent constructive focus on deficits, given the greater similarity of items within these two groups. The possibility of outliers was then investigated using the common criteria of three standard deviations from the mean. No outliers were identified on the 55-item version of the SUPSAD-S 101 measure. Before testing the appropriateness of factor analysis, an independent sample t-test was conducted to determine whether subsamples (i.e., participants recruited by personal email versus email forwarded by training director; participants who completed survey retrospectively for fall supervisors versus current spring supervisors) used in the present study could be combined. Independent samples t-test results found no significant differences on the SUPSAD-S (mean item score for 55 item version) between samples by recruitment method, t (202) = -.507, p > .05, or by timing of completion of survey, t (202) = -.477, p > .05. Hence, subsamples were combined for subsequent analyses. Bartlett?s (1950) test of sphericity and the Kaiser-Meyer-Olkin (KMO) test of sampling adequacy were used to ascertain the appropriateness of factor analysis for the present investigation. More specifically, Bartlett?s (1950) test was used to determine whether the acquired data was a representative sample of the normal population. According to Bartlett?s test, a significant chi-square test indicates that the correlations of the matrix are different from zero and thus, a factor analysis would be appropriate for the data (Tinsley & Tinsley, 1987). This test was significant, ? 2 (df 1485, N = 204) = 9789.81, p < .001, indicating the data is indeed appropriate for a factor analysis. The KMO test of sampling adequacy is an index for comparing the magnitudes of the observed correlation coefficients to the magnitudes of the partial correlation coefficients. For the KMO, a score above .50 is considered to be acceptable and a score above .90 is considered exceptional (Kaiser, 1974). In the present study, the KMO statistic was found to be exceptional at .95. Based on the results of the KMO and Bartlett tests, factor analysis was judged to be appropriate for the present data set. Also, before conducting the 102 factor analysis, all 55 items were examined and deemed appropriate for factor analysis on the basis of means, standard deviations, kurtosis, and skewness. Main Analyses. A principal axis factor (PAF) analysis, as opposed to principal components analysis (PCA), was used to explore the factor structure of the SUPSAD-S. PAF is the best and most commonly used approach for exploratory factor analysis, and is the preferred method when the goal is to understand the relations among a set of measured variables (items) in terms of a smaller number of underlying latent variables (factors) (Kahn, 2006; Floyd & Widaman, 1995). Additionally, PAF produces more accurate final estimates of communality than does PCA, and produces more accurate estimates of factor loadings and factor correlations than PCA when communalities are low (Floyd & Widaman, 1995; Kahn 2006). Notably, the results of PAF and PCA are virtually the same when there are a large number of variables relative to a small number of factors and 50 or so items are being analyzed, as well as when communalities are high (Floyd & Widaman, 1995). Since both criteria were met in the present investigation, PAF was chosen. Also, a promax rotation of extracted principal factors was conducted. Rotated solutions are usually preferable because they create a more even distribution of the variance accounted for among factors, increase the interpretability of factors, and make variables load highly on as few factors as possible (Kahn, 2006). After considering whether to use an orthogonal (e.g., varimax) or oblique (e.g., promax) rotation, a promax rotation was selected because oblique rotations may do a better job than orthogonal rotations of explaining relations among variables when factors are correlated. If factors are uncorrelated, a promax rotation produces results similar to orthogonal (e.g., varimax) 103 rotations. Recent factor development literature suggests using a promax rotation because of its versatility (i.e., ability to produce satisfactory solutions whether or not factors are correlated) and especially if there is a strong possibility that factors may be correlated (Kahn, 2006). Given the newness of the current research, I did not know whether the hypothesized focus on strengths and constructive focus on deficits factors would correlate, and it could be argued on theoretical grounds that they might correlate. Specifically, on the basis of the conceptualization of constructive focus on deficits as a positive psychology construct, one could argue that it would correlate with focus on strengths. Therefore, I selected the promax rotation. (It is important to note, however, that different rotations may provide slightly different results but the differences are not usually dramatic, e.g., Kahn, 2006; such was the case in the present study as I found very similar results when comparing promax to varimax rotations. Only promax rotation results are described further). When determining the number of factors to extract, the use of multiple criteria is recommended because different decision rules sometimes call for the retention of different numbers of factors (Kahn, 2006). The following methods were used. First, Cattell?s scree plots (plots of eigenvalues versus number of factors) for different factor solutions were inspected (Kahn, 2006). Factors falling above the scree line where the eigenvalues flatten out on the scree plot were retained. Second, the proportion of variance accounted for within different factor solution was considered, with the goal of obtaining a factor solution accounting for between 50-80% of the estimated common variance. Third, the percentage of variance explained by each factor was inspected, retaining all factors that explain a large enough percentage and discarding those with small percentages. 104 Fourth, the number of items with significant factor loadings on each factor was considered. Typically, three variables per factor are considered minimum for extraction, and larger numbers of variables per factor help improve factor stability (Kahn, 2006). Fifth, we used the criteria of interpretability of factor solutions. Notably, we did not use the Kaiser-Guttman rule (all factors with eigenvalues > 1.00), the most frequently used criterion for retaining factors, because recent factor analysis literature suggests that it is only appropriate, on statistical grounds, for principal components analysis; additionally, it often results in the extraction of too many factors (Kahn, 2006). Examination of the scree plot showed clear support for a two factor solution, with a sharp drop-off and flattening out of eigenvalues after the second factor. The first and second factors had eigenvalues of 24.24 and 5.64, respectively, while eigenvalues for subsequent factors were 1.8 or less. Also, the first and second factors accounted for a total of 52.764% of the variance, with the first factor accounting for 43.28% and second accounting for 9.49% of the variance, respectively; subsequent factors accounted for substantially less variance (2.69% or less). In addition, both factors met and exceeded the criteria of having at least three items with significant loadings. All items loading on factor 1 loaded above .5, with 12 items loaded above .8, while all items loading on factor 2 also loaded above .5, with 8 items loading in the .7-.8 range. The two factor solution was also deemed interpretable given that the highest loading items on factor 1 related to and were written to represent focus on strengths, while the highest loading items on factor 2 related to and were written to represent constructive focus on deficits. The majority of other items loading onto factors 1 and 2 also related to strengths and deficits, respectively. Therefore, the two factor solution was judged to be the most parsimonious 105 and expedient description of the data set and the two factor solution was retained. The two factors were correlated .57, indicating that the promax rotation was probably appropriate for this data set. Before describing the criteria used to retain items loading onto the two factors, it is first important to note that structure coefficients for promax rotations (analogous to factor loadings) are universally large to the extent that factors are correlated, and items may also appear to cross-load more than they would have if a varimax rotation been performed (Kahn, 2006). The reason is that, by allowing the factors to correlate with each other, promax rotations (unlike varimax rotations) provide estimates of structure coefficients (i.e., factor loadings) that reflect the correlations between factors. Therefore, in addition to examining structure coefficients, it is also helpful to examine pattern coefficients, which reflect the relationships between items and each factor while controlling for the correlation between factors. For example, an item may appear to correlate with more than one factor only because the factors are correlated; after controlling for the correlation, the item may only correlate strongly with one factor and not others (Kahn, 2006). Hence, both structure coefficient and pattern coefficient matrices were examined for making decisions about whether to retain items. In general, our goal in item reduction was to retain the items that most strongly and purely represented each factor (with as little loading on the other factor as possible), and to narrow down the item pool to 24 items, a number that would permit sound reliability while also enhancing ease of usage. Within the 24 items, our goal was to retain 12 items (6 regular and 6 reverse scored to prevent response bias) for each of the two factors. First examining the structure coefficient matrix, there were no items to eliminate 106 on the basis of lacking substantial loading on either factor; all items loaded at least .5 onto one of the two factors, although this is partially related to the use of a promax rotation, as explained previously. After this criteria, items with less than .1 difference between their structure coefficients (factor loadings) on the two factors were eliminated (9 items). I then tried to identify the best/purest 24 items using a combination of the following criteria: (a) largest structure coefficients (highest factor loadings) onto one factor and (b) least cross-loading with the other factor, as reflected by either the difference between structure coefficients (loadings) on the two factors and/or strong correlation with only one factor (and lack of correlation with the other factor, i.e. less than .3) when examining the pattern coefficient matrix. Eliminating items using these criteria leads to the retention of items that are purer measures of the underlying factors (Kahn, 2006). Four iterations with the two factor solution were run, deleting items each time according to the above criteria and conducting factor analysis again on the remaining subset of items (Kahn. 2006; Floyd & Widaman, 1995). These criteria allowed for the identification of the best 12 items on factor 2 (6 regular and 6 reversed) and the best 6 reverse-scored items on factor 1. To select the best 6 regular-scored items on factor 1, several steps were taken. First, application of the above criteria (i.e., examination of structure and pattern coefficients) led to the retention of 3 regular-scored factor 1 items. After retaining these 3 items, it was noted that remaining regular-scored factor 1 items had similar structure and pattern coefficients and were similarly satisfactory (e.g., high loading onto only factor 1), making it difficult to select the best 3 items from this remaining group. To provide some statistical basis for selecting these 3 additional items, 107 a reliability analysis of all factor 1 items was conducted, and the 3 items from this remaining group with the highest item-total correlations and largest decrease in alphas if item deleted were retained. The final two factor solution retaining 24 items accounted for 63% percent of the variance (see Table 1). Kahn (2006) recommends labeling factors on the basis of what items with the largest structure coefficients (highest factor loadings) have in common. Items with the largest structure coefficients on factor 1 included content such as reinforcement of things done well and of strengths, helping increase supervisees? awareness of strengths, positive feedback, and compliments on things done well. Hence, factor 1 of the SUPSAD-S factors appeared to best summarized by the label Focus on Strengths (12 items, accounting for 45.41% of the variance). Higher scores indicate greater supervisory focus on strengths from the perspective of supervisees rating their supervisors. Factor loadings ranged from .79 to .91. Items with the largest structure coefficients on factor 2 included content related to helping the supervisee become more aware of weaknesses; constructive negative feedback; identification of areas needing improvement, ineffective interventions, or what supervisee could do better; and use of specific examples when pointing out weaknesses. Some of these deficits items seemed to entail explicitly constructive, positive, growth- enhancing ways of addressing supervisee deficits (e.g., ?My supervisor attended to my deficits because she/he wanted to see me improve?), while others seemed to entail more neutral ways of focusing on supervisees? deficits (e.g., ?My supervisor helped me identify where I was getting off track with a client?). Thus, it seemed that this subscale could be entitled ?focus on deficits? rather than ?constructive focus on deficits.? 108 Table 1 Structure Coefficients, Pattern Coefficients, Item-Total Correlations, and Alphas if Item Deleted for SUPSAD-S Items Item Structure Coefficients Pattern Coefficients Item-Total Alpha Factor I Factor II Factor I Factor II Correlations if deleted Factor 1: Focus on Strengths My supervisor?. 44rv. did NOT reinforce things I did well in my clinical work. .91 .397 .90 .03 .89 .96 26rv. did NOT give enough positive feedback. .89 .31 .92 -.07 .87 .96 39rv. did NOT compliment me on what I did well. .88 .30 .91 -.08 .87 .96 3. gave me positive feedback about my skills as a therapist .87 .36 .87 -.01 .85 .96 19rv. did NOT reinforce my strengths. .85 .35 .85 .00 .84 .96 46. helped me become more aware of my strengths as a therapist. .84 .49 .77 .17 .82 .96 31rv. did NOT tell me what I did well in my clinical work .83 .39 .81 .05 .82 .96 8rv. did NOT praise my therapeutic skills. .83 .34 .83 -.00 .81 .96 50. praised me for the good work I had done with my clients. .83 .27 .86 -.09 .81 .96 1. focused on my strengths .82 .29 .85 -.07 .81 .96 109 Table 1 (continued) Item Structure Coefficients Pattern Coefficients Item-Total Alpha Factor I Factor II Factor I Factor II Correlations if deleted My supervisor?. 33. identified areas where I was competent. .80 .42 .75 .10 .78 .97 6. identified areas where I excel as a therapist. .79 .34 .78 .02 .78 .97 Factor 2: Constructive Focus on Deficits My supervisor? 49rv. did NOT help me become more aware of my weaknesses .23 .86 -.15 .92 .82 .92 as a therapist. 43. gave me constructive negative feedback .40 .82 .08 .79 .79 .92 47rv. did NOT identify areas where I need improvement. .18 .78 -.17 .85 .74 .92 55rv. did NOT encourage discussion of what I could have done .39 .77 .09 .73 .74 .92 better in my clinical work. 34rv. did NOT use specific examples (e.g., from a tape from my .37 .73 .08 .70 .71 .92 session with a client) when pointing out my weaknesses. 110 Table 1 (continued) Item Structure Coefficients Pattern Coefficients Item-Total Alpha Factor I Factor II Factor I Factor II Correlations if deleted My supervisor?. 23. encouraged discussion of what I could done better .39 .72 .11 .68 .70 .92 in my clinical work 40. helped me identify where I was getting off track with a client. .34 .72 .05 .70 .70 .92 16. used specific examples (e.g., from a tape from my session with .27 .71 -.03 .72 .69 .93 a client) to highlight my areas needing improvement. 21rv. did NOT help me identify where I was getting off track with .35 .69 .08 .66 .67 .93 a client 32rv. did NOT give me constructive criticism. .41 .68 .16 .62 .65 .93 7. attended to my deficits because she/he wanted to see me improve .19 .64 -.09 .67 .61 .93 14. helped me identify ineffective interventions I was using. .21 .61 -.06 .63 .59 .93 Note. N = 204. Kaiser-Meyer-Olkin index = .95. The Focus on Strengths and Constructive Focus on Deficits factors accounted for 45% and 18%, respectively, of the total variance. Factor loadings (structure coefficients) were obtained with the rotated structure matrix of the promax solution. 111 However, it seems that both explicitly positive and more neutral ways of focusing on supervisees? deficits can still considered constructive in the sense that neither is destructive, like harsh, critical, or punitive ways of focusing on supervisees? deficits. In addition, subscale reliability and item-total correlations were high (discussed more below), suggesting that the positive and neutral sounding items hung together well and were tapping into the same construct. Hence, the subscale was named Constructive Focus on Deficits (12 items, accounting for 17.68% of the variance). Higher subscale scores indicate greater supervisory constructive focus on deficits from the perspective of supervisees rating their supervisors. Factor loadings ranged from .61 to .86. The structure coefficients and pattern coefficients for the 24 items are included in Table 1. Item means, standard deviations, kurtosis, and skewness are included in Table 2. In sum, the results of this EFA support Hypothesis #1a in providing evidence for the construct validity of the SUPSAD-S, in terms of a factor structure containing two factors, supervisory focus on strengths and constructive focus on deficits. Other descriptive data. The SUPSAD-S subscale item means were 5.46 (SD = 1.16) for Focus on Strengths and 4.95 (SD = 1.09) for Constructive Focus on Deficits. Despite the elimination of items that cross-loaded or correlated highly with more than one factor, the strengths and deficits factors remained moderately correlated at r = .41. Because the two subscales are not highly correlated, the SUPSAD-S total score may not be meaningful. Thus, in subsequent analyses, only the strengths and deficits subscale scores are used. See Table 3. 112 Table 2 Item Means, Standard Deviations, and Skewness and Kurtosis Statistics of SUPSAD-S Items Item Mean (SD) Skewness Kurtosis___ Factor 1: Focus on Strengths My supervisor?. 44rv. did NOT reinforce things I did well in my clinical work. 5.51 (1.32) -.69 (.17) -.14 (.34) 26rv. did NOT give enough positive feedback. 5.26 (1.75) -.93 (.17) -.13 (.34) 39rv. did NOT compliment me on what I did well. 5.65 (1.37) -.89 (.17) -.11 (.34) 3. gave me positive feedback about my skills as a therapist 5.67 (1.19) -.69 (.17) -.14 (.34) 19rv. did NOT reinforce my strengths. 5.59 (1.37) -.88 (.17) .13 (.34) 46. helped me become more aware of my strengths as a therapist. 4.95 (1.32) -.56 (.17) -.32 (.34) 31rv. did NOT tell me what I did well in my clinical work 5.56 (1.38) -.81 (.17) .12 (.34) 8rv. did NOT praise my therapeutic skills. 5.55 (1.37) -.76 (.17) -.11 (.34) 50. praised me for the good work I had done with my clients. 5.54 (1.31) -.93 (.17) .81 (.34) 1. focused on my strengths 5.40 (1.31) -.62 (.17) -.05 (.34) 113 Table 2 (continued) Item Mean (SD) Skewness Kurtosis___ 33. identified areas where I was competent. 5.47 (1.13) -.50 (.17) -.11 (.34) 6. identified areas where I excel as a therapist. 5.33 (1.32) -.49 (.17) -.46 (.34) Factor 2: Focus on Deficits My supervisor?. 49rv. did NOT help me become more aware of my weaknesses 5.19 (1.46) -1.03 (.17) .86 (.34) as a therapist. 43. gave me constructive negative feedback 4.78 (1.40) -.68 (.17) .13 (.34) 47rv. did NOT identify areas where I need improvement. 5.10 (1.51) -.88 (.17) .40 (.34) 55rv. did NOT encourage discussion of what I could have done 5.34 (1.38) -.87 (.17) .50 (.34) better in my clinical work. 34rv. did NOT use specific examples (e.g., from a tape from my session 4.91 (1.68) -.58 (.17) -.58 (.34) with a client) when pointing out my weaknesses. 23. encouraged discussion of what I could done better in my clinical work 4.97 (1.28) -.70 (.17) .55 (.34) 114 Table 2 (continued) Item Mean (SD) Skewness Kurtosis___ 40. helped me identify where I was getting off track with a client. 4.81 (1.33) -.64 (.17) .50 (.34) 16. used specific examples (e.g., from a tape from my session with a 4.53 (1.74) -.38 (.17) -.68 (.34) client) to highlight my areas needing improvement. 21rv. did NOT help me identify where I was getting off track with 5.23 (1.35) -.70 (.17) .36 (.34) a client. 32rv. did NOT give me constructive criticism. 5.39 (1.45) -.86 (.17) .34 (.34) 7. attended to my deficits because she/he wanted to see me improve 4.79 (1.29) -.39 (.17) .17 (.34) 14. helped me identify ineffective interventions I was using. 4.44 (1.39) -.38 (.17) -.05 (.34) Note. N = 204. SD = Standard deviation, in parentheses. Standard Error of Skewness and Standard Error of Kurtosis in parentheses 115 Table 3 Correlations, Means, Standard Deviations, Range, Skewness, Kurtosis, Internal Consistency Estimates, and Test-Retest Reliabilities for the SUPSAD-S Subscales Scale ` 1___________2___________________________________ 1. Strengths 1 2. Deficits .41** 1 Mean-Item 5.46 4.96 SD-Item 1.16 1.09 Range-Low 2.67 1.50 Range-High 7.00 6.92 Skewness (SE) -.59 (.17) -.75 (.17) Kurtosis (SE) -.53 (.34) .81 (.34) Cronbach?s ? .97 .93 Test-Retest reliability a .91 .85 Note. N = 204. Strengths=Focus on Strengths; Deficits=Constructive Focus on Deficits. Standard Error of Skewness and Standard Error of Kurtosis in parentheses. ** = p < .01 a Two-week test-retest reliability (N = 29). 116 Hypothesis #1b: The focus on strengths subscale of the SUPSAD-S will have adequate internal consistency reliability. Hypothesis #1c: The constructive focus on deficits subscale of the SUPSAD-S will have adequate internal consistency reliability. Internal consistency reliabilities (Cronbach?s ?) of the SUPSAD-S subscale scores were .97 for the Focus on Strengths subscale and .93 for the Constructive Focus on Deficits subscale (see Table 3). Given these very high internal consistency reliabilities, we checked to see if items within subscales were redundant (i.e., correlated > .90). Correlations between items loading onto the Strengths factor ranged from .60 to .84, while correlations between items loading onto the Deficits factor ranged from .40 to .84. These correlations suggest that items within subscales are not redundant or identical. In addition, item-total correlations ranged from .78 to .89 for the Strengths subscale and .59 to .82 for the Deficits subscale (see Table 1), suggesting again that all items were highly related to their respective subscales. Hence, Hypotheses # 1b and 1c were both supported, as both the focus on strengths and constructive focus on deficits subscales have adequate internal consistency reliability. Hypothesis #1d: The focus on strengths subscale of the SUPSAD-S will have at least moderate two-week test-retest reliability. Hypothesis #1e: The constructive focus on deficits subscale of the SUPSAD-S will have at least moderate two-week test-retest reliability. Two week test-retest reliability was determined by Pearson correlation coefficients between participants? SUPSAD-S subscale scores at time 1 and time 2 (for the 29 participants who took the measure twice). Two week test-retest reliabilities of the 117 SUPSAD-S subscale scores were .91 for the Focus on Strengths subscale and .85 for the Constructive Focus on Deficits subscale (see Table 3). Hence, Hypotheses #1d and 1e were both supported, as both the Focus on Strengths and Constructive Focus on Deficits subscales were found to have high two-week test-retest reliability. Overall, results supported Hypothesis #1, with preliminary evidence suggesting that I was able to create a measure (SUPSAD-S) of supervisory focus on strengths and constructive focus on deficits with adequate construct validity and reliability. Given that I found the predicted two subscales, I then went ahead and examined the relationships of each subscale to other variables of interest. Convergent and Discriminant Validity of the SUPSAD-S The convergent validity of the Focus on Strengths and Constructive Focus on Deficits subscales of the SUPSAD-S measure was investigated by assessing the correlations of each subscale with several components of supervision process and outcome (satisfaction with supervision, the supervisory working alliance, and supervisee counseling self-efficacy), while discriminant validity was assessed via correlations with public self-consciousness. For all analyses, an alpha level of .01 was used to provide a more stringent criterion for significant findings and minimize the possibility of finding relations due to chance given the large sample size. Preliminary Analyses. Missing values for the 201 participants completing the SSQ, SWAI-T, CASES, and SCS measures were analyzed using the pattern analysis techniques described previously, and no pattern of missing data among the scales was found. Therefore, data imputation to fill in missing data values was conducted using maximum likelihood estimation (EM) for each individual scale (SSQ, SCS) or subscale 118 (i.e., SWAI-T Rapport and Client Focus; CASES Helping Skills and Session Management). Using the criterion of three standard deviations from the mean, three outliers were identified on the SWAI-T Client Focus subscale. Given that these participants were not outliers on any of the other measures in the survey, they were retained under the rationale that their data was likely valid (i.e., their below average ratings on the SWAI-T probably reflected genuinely poorer working alliances with their supervisors). All measures were then checked for skewness and kurtosis by examining whether the skewness and kurtosis statistics were greater than twice their standard errors, respectively. None of these measures had abnormal kurtosis, while two measures were found to be significantly negatively skewed (i.e., scores concentrated at the higher end of the scale): the SSQ and SWAI-T (both Rapport and Client Focus subscales). However, visual inspection of the distributions suggested that they did not deviate enough from normality to violate assumptions of normality for the statistical tests to used for data analysis (primarily correlations and multiple regression), which are also robust to some deviation from normality. Hence, no data transformations were undertaken. See Table 4 for descriptive statistics of the SSQ, SWAI-T, CASES, and SCS measures. Hypothesis #2a: Evidence will be found for the convergent validity of the focus on strengths subscale of the SUPSAD-S as demonstrated by a positive correlation with supervisee ratings of satisfaction with supervision. Hypothesis #2b: Evidence will be found for the convergent validity of the constructive focus on deficits subscale of the SUPSAD-S as demonstrated by a positive correlation with supervisee ratings of satisfaction with supervision. 119 To test Hypotheses #2a and 2b, bivariate correlations between each of the SUPSAD-S subscales (Focus on Strengths and Constructive Focus on Deficits) and scores on the SSQ (Ladany et al., 1996) were tested for significance (See Table 4). Supervisory focus on strengths was positively correlated with satisfaction with supervision (r = .71, p < .001), suggesting that the more supervisees perceived their supervisors to focus on their strengths, the greater their satisfaction with supervision. Constructive focus on deficits was also positively correlated with satisfaction with supervision (r = .57, p < .001), suggesting that the more supervisees perceived their supervisors to focus constructively on their deficits, the greater their satisfaction with supervision. For determining effect sizes of correlations, the common rule of thumb (that will be applied for all correlations reported in this chapter) is that r values of .5, .3, and .1, correspond to large, medium, and small effect sizes, respectively (Cohen, 1988). Thus, both of the above correlations represented large effect sizes (i.e., r > .5). Each of these relationships was also examined visually via scatterplots for non- linear (e.g., quadratic, cubic) relationships. Both relationships appeared to be linear. Hence, Hypothesis #2a for the convergent validity of the Focus on Strengths subscale and Hypothesis #2b for the convergent validity of the Constructive Focus on Deficits subscales were both supported. 120 Table 4 Correlation Matrix, Means, Standard Deviations, Skewness and Kurtosis Statistics for SUPSAD-S, SSQ, SWAI-T, CASES, and SCS 1 2 3 4 5 6 7 8 9 10 1. Strengths 1 .41 .71** .71** .52** .69** .24** .28** .27** -.07 2. Deficits 1 .57** .36** .64** .50** .07 .15* .11 .00 3. SSQ 1 .77** .73** .81** .20** .22** .22** -.01 4. SWAI-T Rp 1 .71** .96** .29** .29** .30** -.08 5. SWAI-T CF 1 .88** .26** .29** .28** -.03 6. SWAI-T Tot 1 .30** .31** .32** -.07 7. CASES HS 1 .77** .96** -.10 8. CASES SM 1 .91** -.12 9. CASES Tot 1 -.11 10. SCS 1 M 5.47 4.97 3.19 5.56 5.09 5.38 6.65 6.70 6.67 2.60 SD 1.16 1.09 .69 1.09 1.10 1.02 1.11 1.14 1.06 .57 121 Table 4 (continued) 1 2 3 4 5 6 7 8 9 10 Skewness -.60 -.77 -.85 -1.03 -.84 -.92 -.29 -.36 -.32 -.07 Kurtosis -.50 .87 -.00 .50 .72 .33 -.53 -.56 -.52 .24 Note. N = 201. Strengths = Focus on Strengths subscale of SUPSAD-S; Deficits = Constructive Focus on Deficits subscale of SUPSAD-S. SSQ = Supervisee Satisfaction Questionnaire. SWAI-T Rp = Supervisor Working Alliance Inventory-Trainee Form Rapport subscale; SWAI-T CF = Supervisor Working Alliance Inventory-Trainee Form Client Focus Subscale; SWAI-T Tot = Supervisor Working Alliance Inventory-Trainee Form Total Score. CASES HS = Counselor Activity Self-Efficacy Scales Helping Skills subscale; CASES SM= Counselor Activity Self-Efficacy Scales Session Management Subscale; CASES Tot= Counselor Activity Self-Efficacy Scales Total Score. SCS = Public Self-Consciousness Scale. M = Mean. SD = Standard Deviation. Standard Error for Skewness Statistics for all measures =.17, Standard Error for Kurtosis Statistics for all measures = .34. ** p < .01, * p < .05. 122 Hypothesis #3a: Evidence will be found for the convergent validity of the focus on strengths subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee ratings of the supervisory working alliance. Hypothesis #3b: Evidence will be found for the convergent validity of the constructive focus on deficits subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee ratings of the supervisory working alliance. To test Hypotheses #3a and 3b, bivariate correlations between each of the SUPSAD-S subscales (Focus on Strengths and Constructive Focus on Deficits) and the SWAI-T (Efstation et al., 1990) total score as well as subscale (Rapport and Client Focus) scores were tested for significance (see Table 4). Supervisory focus on strengths was positively correlated with both components of the supervisory working alliance, Rapport (r = .71, p < .001) and Client Focus (r = .52, p < .001) as well as the overall supervisory working alliance (r = .69, p < .001), all large effect sizes. These results suggest that the more supervisees perceived their supervisors to focus on their strengths, the stronger they rated the supervisory working alliance (overall as well as both working alliance components). Constructive focus on deficits was also positively correlated with both components of the supervisory working alliance, Rapport (r = .36, p < .001; medium effect size) and Client Focus (r = .64, p < .001; large effect size), as well as the overall working alliance (r = .50, p < .001; large effect size). These results suggest that the more supervisees perceived their supervisors to focus on their strengths, the stronger they rated the supervisory working alliance, particularly the Client Focus component. 123 All of the above relationships were also examined visually via scatterplots for non-linear (e.g., quadratic, cubic) relationships. All relationships appeared to be linear. Hence, Hypotheses #3a for the convergent validity of the Focus on Strengths subscale and Hypothesis #3b for the convergent validity of the Constructive Focus on Deficits subscale were both supported. Hypothesis #4a: Evidence will be found for the convergent validity of the focus on strengths subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee counseling self-efficacy. Hypotheses #4b: Evidence will be found for the convergent validity of the constructive focus on deficits subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee counseling self-efficacy. To test Hypotheses #4a and 4b, bivariate correlations between each of the SUPSAD-S subscales (Focus on Strengths and Constructive Focus on Deficits), and the CASES (Lent et al., 2003) total score as well as subscale (Helping Skills and Session Management) scores were tested for significance (see Table 4). Supervisory focus on strengths was positively correlated with self-efficacy for helping skills (r = .24, p < .001; small effect size), for session management (r = .28, p < .001; small-medium effect size), and overall counseling self-efficacy (r = .27, p < .001; small-medium effect size). These results suggest that the more supervisees perceived their supervisors to focus on their strengths, the higher counseling self-efficacy they had (overall and with regards to helping skills and session management). Examination of the scatterplot of the relationship between supervisory focus on strengths and counseling self-efficacy (total) showed a weak linear relationship. 124 No significant correlations were found between constructive focus on deficits and either self-efficacy for helping skills (r = .07, p >.05), self-efficacy for session management (r = .15, p < .05), or overall counseling self-efficacy (r = .11, p > .05). These results suggest that supervisees? perceptions of their supervisors? constructive focus on their deficits was not related to their counseling self-efficacy. No effect sizes are reported for the relationships between constructive focus on deficits and counseling self- efficacy (total score or subscales), given the non-significant correlations. Hence, Hypothesis #4a for the convergent validity of the Focus on Strengths subscale was supported. Hypothesis #4b for the convergent validity of the Constructive Focus on Deficits subscale was not supported. Hypothesis #5a: Evidence will be found for the discriminant validity of the focus on strengths subscale of the SUPSAD-S, as demonstrated by a non-significant correlation with public self-consciousness. Hypothesis #5b: Evidence will be found for the discriminant validity of the constructive focus on deficits subscale of the SUPSAD-S, as demonstrated by a non-significant correlation with public self-consciousness. To test Hypotheses #5a and 5b, bivariate correlations between each of the SUPSAD-S subscales (Focus on Strengths and Constructive Focus on Deficits) and scores on the SCS (Fenigstein et al., 1975) were tested for significance (see Table 4). Neither supervisory focus on strengths (r = -.07, p > .05) nor constructive focus on deficits (r = .00, p > .05) was significantly correlated with public self-consciousness. Hence, Hypothesis #5a for the discriminant validity of the Focus on Strengths subscale 125 and Hypothesis #5b for the discriminant validity of the Constructive Focus on Deficits subscale were both supported. In sum, results support the convergent validity of the Focus on Strengths subscale through predicted relationships with satisfaction with supervision, the supervisory working alliance, and supervisee counseling self-efficacy. Results also support the convergent validity of the Constructive Focus on Deficits subscale through predicted relationships with the supervisory working alliance and satisfaction with supervision. Evidence for the discriminant validity of both subscales was found through their lack of relationship to public self-consciousness. Exploratory Research Questions Research Question #1a: How much unique variance in satisfaction with supervision is accounted for by focus on strengths and constructive focus on deficits, respectively? A simultaneous multiple regression was run with focus on strengths and constructive focus on deficits entered as predictor variables and satisfaction with supervision as the criterion variable. The overall regression model was significant, F (2, 198) = 149.57, p < .001, and strengths and deficits together accounted for 60.2% of the variance in satisfaction with supervision. The beta weight for focus on strengths was .58 (p <. 001), accounting for 33.64% of the unique variance in satisfaction with supervision (? 2 = .42, large effect size). The beta weight for constructive focus on deficits was .34 (p < .001), accounting for 11.56% of unique variance in satisfaction with supervision (? 2 = .19, medium effect size). (These effect sizes were determined by the same rule of thumb described previously for a correlation r, except that ? 2 is in r 2 units; hence, the previous criteria of r of .1, .3, and .5 as small, medium, and large effects translate to r 2 or ? 2 of 126 .01, .09, and .25 as small, medium, and large effects, respectively; Cohen, 1988). Therefore, focus on strengths accounted for almost three times as much unique variance in satisfaction with supervision as did constructive focus on deficits. See Table 5. Research Question #1b: How much unique variance in the overall supervisory working alliance is accounted for by focus on strengths and constructive focus on deficits, respectively? A simultaneous multiple regression was run with focus on strengths and constructive focus on deficits entered as predictor variables and supervisory working alliance (total score) as the criterion variable. The overall regression model was significant, F (2, 198) = 112.239, p < .001, and strengths and deficits together accounted for 53.1% of the variance in the supervisory working alliance. The beta weight for focus on strengths was .58, (p < .001), accounting for 33.64% of the unique variance in the supervisory working alliance (? 2 = .38, large effect size). The beta weight for constructive focus on deficits was .27 (p < .001), accounting for 7.29% of the unique variance in the supervisory working alliance (? 2 = .12, medium effect size). Therefore, focus on strengths accounted for almost five times as much unique variance in the supervisory working alliance as did constructive focus on deficits. See Table 5. Research Question #1c: How much unique variance in overall supervisee counseling self- efficacy is accounted for by focus on strengths and constructive focus on deficits, respectively? A simultaneous multiple regression was run with focus on strengths and constructive focus on deficits entered as predictor variables and supervisee counseling self-efficacy (total score) as the criterion variable. The overall regression model was 127 Table 5 Simultaneous Multiple Regressions with SUPSAD-S Subscales as Predictors of SSQ, SWAI-T, and CASES Scores _ SSQ SWAI-T______CASES__ Strengths Beta Weight .58* .58* .27* Unique Variance accounted for by subscale 33.6% 33.6% 7.3% Deficits Beta Weight .34* .27* .00 Unique Variance accounted for by subscale 11.6% 7.29% 0% Model Variance R 2 60.2% 53.1`% 7.4% Note. N = 201. Strengths = Focus on Strengths subscale of SUPSAD-S; Deficits = Constructive Focus on Deficits subscale of SUPSAD-S. SSQ = Supervisee Satisfaction Questionnaire. SWAI-T = Supervisor Working Alliance Inventory-Trainee Form Total Score. CASES = Counselor Activity Self-Efficacy Scales Total Score. Model Variance = Total variance accounted for by regression model with strengths and deficits as predictors. * p < .01. 128 significant, F (2, 198) = 7.90, p < .01, and strengths and deficits together accounted for 7.4% of the unique variance in supervisee counseling self-efficacy. The beta weight for focus on strengths was .27 (p < .001), accounting for 7.3% of the unique variance in supervisee counseling self-efficacy (? 2 = .06, small effect size). The beta weight for constructive focus on deficits was .00 (p > .05), accounting for 0% of unique variance in supervisee counseling self-efficacy (? 2 = 0, no effect). Therefore, focus on strengths accounted for all of the model variance in supervisee counseling self-efficacy, although it was still a small amount of variance explained, while constructive focus on deficits did not explain any unique variance in counseling self-efficacy. See Table 5. Research Question #2: How much do supervisees perceive that their supervisors focus on strengths as compared to deficits? A paired samples t-test comparing perception of focus on strengths to constructive focus on deficits (in the sample of 204) was conducted. Results were significant, t (203) = 5.85, p < .001 (? 2 = .14, medium effect size), suggesting that supervisees in the present sample perceived their supervisors as focusing significantly more on their strengths than on their deficits. Research Question #3a: How do the subscales of focus on strengths and constructive focus on deficits relate to supervisee training level? Two indices of supervisee training level assessed via the demographic questionnaire were supervisees? year in their training programs and their number of direct clinical hours. Neither the focus on strengths (r = .09, p > .05) nor the constructive focus on deficits (r = .00, p > .05) subscale was significantly correlated to supervisees? year in their training programs. Nor was either subscale significantly correlated with supervisees? 129 number of clinical hours (r = .09, p > .05 for strengths subscale, r = .08, p > .05 for deficits subscale). Research Question #3b: How do the subscales of focus on strengths and constructive focus on deficits relate to supervisee and supervisor theoretical orientations? Bivariate correlations between the SUPSAD-S focus on strengths and constructive focus on deficits subscales with eight trainee theoretical orientations were also examined; no correlations were significant at the .01 level. Also, bivariate correlations between the SUPSAD-S focus on strengths and focus on deficits subscales with eight supervisor theoretical orientations (as rated by supervisees about their supervisors) were examined (see Table 6). Supervisory focus on strengths was significantly positively correlated with perceived supervisor humanistic/existential/client- centered (r = .23, p < .01), multicultural (r = .22, p < .01), and integrative (r = .21, p < .01) theoretical orientations. Supervisory constructive focus on deficits was significantly positively correlated with perceived supervisor multicultural theoretical orientation (r = .20, p < .01). All of these correlations with supervisor theoretical orientations represent small effect sizes. Research Question #3c: Are there differences in mean levels of focus on strengths or constructive focus on deficits for clinical versus counseling psychology trainees, or for trainees whose supervisors hold a degree in counseling psychology versus those whose supervisors hold a degree in clinical psychology? To examine differences between counseling psychology and clinical psychology trainees on SUPSAD-S subscale scores, independent samples t-tests were conducted. Counseling psychology trainees (N = 95, M = 5.52, SD = 1.14) did not differ significantly 130 Table 6 Correlation Matrix for SUPSAD-S Strengths and Deficits Subscales and Supervisor Theoretical Orientations Psychodyn a Hum b CBT b MC c Fem c Narr/Constr d S-F d Integr b 1. Strengths .12 .23** -.13 .22** .05 .12 -.08 .21** 2. Deficits .00 .07 -.04 .20** .15* .08 .08 .15* Note. Strengths = Focus on Strengths subscale; Deficits = Constructive Focus on Deficits subscale. Psychodyn=Psychodynamic. Hum=Humanistic/Existential/Client-Centered. CBT=Cognitive-Behavioral. MC=Multicultural. Fem=Feminist. Narr/Constr=Narrativist/Constuctivist. S-F=Solution-Focused. Integr=Integrative. a N = 200; b N = 199; c N = 198; d N = 196. ** p < .01. * p < .05. 131 from clinical psychology trainees (N = 52, M = 5.33, SD = 1.23) in their perceptions of supervisory focus on strengths, t (145) =.919, p > .05. Also, counseling psychology trainees (M = 4.89, SD = 1.13) did not differ significantly from clinical psychology trainees (M = 4.99, SD = 1.04) in their perceptions of supervisory constructive focus on deficits, t (145) = .481, p > .05. Also, to examine differences on SUPSAD-S subscale scores between trainees whose supervisors had a degree in counseling psychology versus those whose supervisors had a degree in clinical psychology, independent samples t-tests were conducted. Supervisees did not rate counseling psychology supervisors (n = 77, M = 5.57, SD = 1.18) as significantly different from clinical psychology supervisors (n = 81, M = 5.40, SD = 1.15) in terms of focus on strengths, t (156) = .884, p > .05. Nor did supervisees rate counseling psychology supervisors (M = 5.04, SD = 1.10) as significantly different from clinical psychology supervisors (M = 4.87, SD = 1.09) in terms of constructive focus on deficits, t (156) = .989, p > .05. Finally, no significant differences were found on either SUPSAD-S subscale scores by the gender, race, or age of the supervisee; length of time working with the supervisors; or type/modality of supervision. 132 Chapter 6 DISCUSSION In the present study, the SUPSAD-S measure was developed and used to examine whether supervisory focus on strengths and constructive focus on deficits each relate to aspects of supervision process and outcome (e.g., satisfaction with supervision, the supervisory relationship) and supervisee development (e.g., counseling self-efficacy) in hypothesized ways. In this section, I first discuss the results for each of the hypotheses and exploratory research questions, then present the limitations of the present investigation, and finally discuss implications for supervision practice, training, and future research. Hypothesis #1: A measure (SUPSAD-S) of supervisory focus on strengths and constructive focus on deficits having adequate construct validity and reliability can be created. Factor Structure of the SUPSAD-S Hypothesis #1a: The SUPSAD-S will have construct validity, as demonstrated by a factor structure containing 2 factors, focus on strengths and a constructive focus on deficits. Exploratory factor analyses revealed the existence of two subscales: Focus on Strengths and Constructive Focus on Deficits, supporting Hypothesis #1a. The first subscale, Focus on Strengths, is composed of 12 items that reflect interventions and processes used by supervisors to identify and enhance supervisees? strengths. Higher scores indicate a stronger perception on the part of supervisees that their supervisors give positive feedback; point out, reinforce, or compliment/praise things supervisees did well in their clinical work; focus on, reinforce, and help supervisees become more aware of 133 their strengths; and identify areas where supervisees are competent or excel. Although these interventions might be used in any theoretical approach to supervision, the interventions on the focus on strengths subscale were primarily derived from behavioral and cognitive-behavioral supervision (positive reinforcement of correct or desired behaviors, e.g., Bradley & Gould, 2001), solution-focused supervision (identifying and focusing on successes and strengths, compliments and praise, e.g., Briggs & Miller, 2005), and the social-cognitive model of counselor training (e.g., positive performance feedback, Larson, 1998). In general, strength-focused interventions consisted of ways of calling attention to positive aspects of supervisees? performance or skills, as opposed to inner strengths or resources that a supervisee might bring to the therapy enterprise (e.g., as a result of life experiences) as theorized by strength-focused approaches such as narrativist/constructivist supervision (Timm & Blow, 1999; Carlson & Erickson, 2001). The second subscale, Constructive Focus on Deficits, is composed of 12 items reflecting the processes and interventions used by supervisors to address supervisees? deficits in a non-critical or non-punitive manner intended to help supervisees grow and improve. Higher scores indicate a stronger perception on the part of supervisees that their supervisors give constructive negative feedback; identify and encourage discussion of areas needing improvement, things the supervisee could have done better, places where the supervisee was getting off track with a client, and ineffective interventions; use specific examples (e.g., from a tape from a session with a client) when pointing out supervisees? weaknesses and areas needing improvement; and attend to deficits because they want to see supervisees improve. Like the focus on strengths items, the deficits items are related more to feedback on aspects of the supervisees? performance (e.g., areas 134 needing improvement or areas where performance is ineffective) which are changeable, rather than supervisee traits (e.g., personality or personal issues) that can be considered liabilities. As discussed previously, some of the constructive focus on deficits interventions seem more explicitly positive (e.g., attending to deficits because of wanting to see the supervisee improve), whereas others sound more neutral (e.g., identification of areas needing improvement). The reason for the many neutral (as opposed to explicitly positive) sounding items on the constructive focus on deficits subscale is that many of the explicitly positive sounding deficits items (e.g., ?My supervisor commented on my weaknesses in ways that helped me take in the feedback?) were eliminated during EFA due to cross-loading onto the strengths factor, or weaker loadings onto the deficits factor than the items that were retained. However, all of the interventions on the constructive focus on deficits subscale (both explicitly positive and neutral) were consistent with literature offering recommendations for how to address problematic or ineffective aspects of supervisee performance in a constructive, non-punitive way that promotes supervisees? growth (Abbott & Lyter, 1998). For example, this literature suggests that when offering constructive negative feedback, supervisors should be specific, focus only on supervisee behaviors that can be changed, and ensure the feedback is improvement-oriented and meant to benefit the supervisee, rather than ?flaw finding? (Abbott & Lyter, 1998; Weisinger & Lobsenz, 1989). In addition, neutral items correlated positively with more explicitly constructive sounding deficits items, and the subscale overall correlated positively with satisfaction with supervision and the supervisory working alliance, which suggests that supervisees interpreted the neutral sounding deficits items as constructive, 135 rather than harsh or punitive. The focus on strengths and constructive focus on deficits subscales were moderately correlated (r = .41), although this correlation was not large enough for a total score for the SUPSAD-S to be considered meaningful. Reliability of the SUPSAD-S Hypothesis #1b: The focus on strengths subscale of the SUPSAD-S will have adequate internal consistency reliability. Hypothesis #1c: The constructive focus on deficits subscale of the SUPSAD-S will have adequate internal consistency reliability. Hypothesis #1d: The focus on strengths subscale of the SUPSAD-S will have at least moderate two-week test-retest reliability. Hypothesis #1e: The constructive focus on deficits subscale of the SUPSAD-S will have at least moderate two-week test-retest reliability. Evidence of good reliability for each of the subscales was found, as both estimates of internal consistency and test-retest reliability across a two week period were high. The high internal consistency reliability estimates suggest that items on each subscale were tapping into a similar construct and hung together well. The high test-reliability estimates suggests that supervisees? impressions of their supervisors? focus on their strengths and constructive focus on their deficits remained stable over a two-week period. Hence, Hypotheses #1b-1e were supported and suggest that the SUPSAD-S has good psychometric properties. 136 Overall, results supported Hypothesis #1, with preliminary evidence suggesting that I was able to create a measure (SUPSAD-S) of supervisory focus on strengths and constructive focus on deficits with adequate construct validity and reliability. Convergent and Discriminant Validity of the SUPSAD-S Hypothesis #2a: Evidence will be found for the convergent validity of the focus on strengths subscale of the SUPSAD-S as demonstrated by a positive correlation with supervisee ratings of satisfaction with supervision. Hypothesis #2a was supported in the current study, as the focus on strengths subscale was positively correlated with a measure of satisfaction with supervision. Hence, it appears that supervisees who perceive their supervisors to focus on their strengths tend to be more satisfied with their supervision experience. This finding makes sense given research suggesting that satisfaction with supervision is related to the provision of positive feedback, supervisor support, affirmation, validation, and interpersonal characteristics such as warmth and sensitivity (e.g., e.g., Worthen & McNeill, 1996; Allen et al., 1986). It also makes sense given that reinforcement of positive, desired, or correct clinical behaviors may facilitate one of the major purposes of supervision, namely learning how to be a competent therapist (e.g., Bernard & Goodyear, 2004; Bradley & Gould, 2001). Supervisees may feel more satisfied with supervision when such learning is facilitated. Hypothesis #2b: Evidence will be found for the convergent validity of the constructive focus on deficits subscale of the SUPSAD-S as demonstrated by a positive correlation with supervisee ratings of satisfaction with supervision. 137 Hypothesis #2b was also supported, as the constructive focus on deficits subscale was positively correlated with a measure of satisfaction with supervision. Hence, it appears that supervisees who perceive their supervisors to focus on constructively on their deficits tend to be more satisfied with their supervision experience. This finding makes sense given research suggesting that supervisees expect and want constructive negative feedback from their supervisors, and may experience a lack of such feedback from their supervisors as unhelpful (e.g., Gould & Bradley, 2001; Abbott & Lyter, 1998; Kadushin, 1992; Chur-Hansen & McLean, 2006). Supervisees generally recognize that they possess deficits, and look to their supervisors as experts to help them reduce or overcome these problems, become more competent, and provide better care to their clients (Abbott & Lyter, 1998; Barnett et al., 2001). To the extent that supervisors help their supervisees make necessary improvements, satisfaction with supervision may be increased. Hypothesis #3a: Evidence will be found for the convergent validity of the focus on strengths subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee ratings of the supervisory working alliance. Hypothesis #3a was supported, as the focus on strengths subscale was positively correlated with a measure of the supervisory working alliance, both total score and Rapport and Client Focus subscale scores. Hence, it appears that supervisees who perceive their supervisors to focus on their strengths tend to perceive a stronger working alliance with these supervisors; more specifically, these supervisees perceive a stronger bond with the supervisor as well as stronger collaboration with the supervisor to help understand their clients. This finding makes sense given that supervisory focus on 138 strengths may promote an atmosphere of warmth, acceptance, respect, and acceptance of experimentation and mistakes, all factors theorized to contribute to the supervisory working alliance (Muse-Burke et al., 2001; Allen et al, 1986; Ladany et al., 1996; Ladany et al., 2001). In particular, a focus on strengths may enhance the emotional bond between the supervisor and supervisee by making the supervisee feel more supported by and trusting of the supervisor (e.g., Talen & Schindler, 1993; Briggs & Miller, 2005). Also, through positive reinforcement, a focus on strengths may enhance supervisees? learning about how to work effectively with clients, thus increasing their sense of collaboration with their supervisors in regards to understanding and treating clients (Bradley & Gould, 2001). Hypothesis #3b: Evidence will be found for the convergent validity of the constructive focus on deficits subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee ratings of the supervisory working alliance. Hypothesis #3b was supported, as the constructive focus on deficits subscale was positively correlated with a measure of the supervisory working alliance, both total score and Client Focus and Rapport subscale scores. Hence, it appears that supervisees who perceive their supervisors to focus constructively on their deficits tend to perceive a stronger working alliance with their supervisors; more specifically, these supervisees perceive a stronger collaboration with the supervisor to help understand their clients as well as a stronger bond with the supervisor. This finding makes sense given that supervisees seek constructive negative feedback from their supervisors to help them work more effectively with their clients (e.g., Abbott & Lyter, 1998; Gould & Bradley, 2001). To the extent that supervisors constructively focus on supervisees? deficits with the intent 139 of helping them learn to work more effectively with clients, supervisors may be upholding their mutual agreement with supervisees about the tasks and goals of supervision (e.g., to help the supervisee become a better therapist; Bernard & Goodyear, 2004). Thus, by promoting supervisee learning, a constructive focus on deficits in supervision may enhance the task and goal components of the supervisory working alliance (similar to the Client Focus subscale). In addition, supervisees may feel a stronger bond to supervisors who make an active effort to help them improve by constructively focusing on their deficits (e.g., Barnett et al., 2001). Hypothesis #4a: Evidence will be found for the convergent validity of the focus on strengths subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee counseling self-efficacy. Hypothesis #4a was supported, as the focus on strengths subscale was positively correlated with a measure of supervisee counseling self-efficacy, both total score and Helping Skills and Session Management subscale scores. Hence, it appears that supervisees who perceive their supervisors to focus on their strengths tend to have higher counseling self-efficacy, both overall and specifically with respect to performing helping skills and managing the counseling process. However, the effect sizes for these relationships were small, which lends weak support to research suggesting that focusing on supervisees? strengths (e.g., through positive feedback, support, communication of confidence, etc.) may relate positively to supervisees? confidence and counseling self- efficacy (e.g., Daniels & Larson, 2001; Wulf & Nelson, 2000; Briggs & Miller, 2005; Koob, 2002). One explanation for this finding may be that focus on supervisees? strengths as operationalized in the SUPSAD-S corresponds to only one of several 140 important contributors to counseling self-efficacy as theorized in the Social-Cognitive Model of Counselor Training (SCMCT; Larson, 1998); namely, the focus on strengths subscale corresponds to verbal persuasion (i.e., support, positive feedback) in the SCMCT model. Other contributors to counseling self-efficacy according to the SCMCT model include vicarious learning experiences (e.g., role modeling) in important counseling tasks and structured practice where chances of success are purposely maximized (mastery experiences). In addition, counseling self-efficacy may be influenced by more than just what happens in supervision; it may also relate to the supervisee?s good and bad counseling experiences as well as their feelings about themselves and their counseling abilities, among other things (Larson & Daniels, 1998; Lent et al., 2003). This might account for why focus on strengths related more strongly to satisfaction with supervision and the supervisory working alliance (which are more exclusively tied to the supervision experience) than to counseling self-efficacy, in comparing effect sizes. Hypotheses #4b: Evidence will be found for the convergent validity of the constructive focus on deficits subscale of the SUPSAD-S, as demonstrated by a positive correlation with supervisee counseling self-efficacy. Hypotheses #4b was not supported, as the constructive focus on deficits subscale was not correlated with a measure of supervisee counseling self-efficacy, either total score or either the Helping Skills and Session Management subscale scores. This finding conflicts with theories such as the social-cognitive model of supervision, which suggests that constructive negative feedback and challenge to improve on existing skills may motivate supervisees to push themselves in ways that lead to growth, achievement of 141 goals, and higher counseling self-efficacy when goals are met (e.g., Larson, 1998; Larson & Daniels, 1998; Lent et al., 1998). However, it also makes sense that having one?s deficits as a therapist pointed out would not be positively related to one?s confidence in his/her counseling skills. Perhaps other variables not measured here that relate to what supervisees do with constructive feedback on their deficits (e.g., internalize the deficit as a stable and unchangeable aspect of their therapy style, or view the deficit more adaptively as a challenge to overcome) might shed light on how constructive focus on deficits relates to supervisees? counseling self-efficacy. Hypothesis #5a: Evidence will be found for the discriminant validity of the focus on strengths subscale of the SUPSAD-S, as demonstrated by a non-significant correlation with public self-consciousness. Hypothesis #5b: Evidence will be found for the discriminant validity of the constructive focus on deficits subscale of the SUPSAD-S, as demonstrated by a non-significant correlation with public self-consciousness. Hypotheses #5a and #5b were both supported, providing evidence of the discriminant validity of the SUPSAD-S subscales as neither subscale correlated with the unrelated construct of public self-consciousness (Fenigstein et al., 1975). This suggests that neither perceptions of supervisory focus on strengths nor perceptions of constructive focus on deficits were related to supervisees? general concern about the way they present themselves to others. In other words, the SUPSAD-S subscales seemed to be independent of any possible supervisee tendency to report only positive things about their supervisors, stemming from the perception that the researchers were looking for this type of positive response or because the items seemed to pull for socially desirable responses. 142 Overall, the present investigation demonstrated initial support for both the convergent and discriminant validity of the SUPSAD-S subscales. Specifically, the convergent validity of the focus on strengths subscale was supported by its correlations with measures of satisfaction with supervision, the supervisory working alliance, and counseling self-efficacy as expected. The convergent validity of the constructive focus on strengths subscale was supported by its correlations with measures of satisfaction with supervision and the supervisory working alliance as expected, although it was somewhat surprising that it was not related to counseling self-efficacy. Discriminant validity of both subscales was supported by their lack of correlation with a measure they would not be expected to be related to (i.e., public self-consciousness). Exploratory Research Questions Research Question #1a: How much unique variance in satisfaction with supervision is accounted for by focus on strengths and constructive focus on deficits, respectively? Research Question #1b: How much unique variance in the overall supervisory working alliance is accounted for by focus on strengths and constructive focus on deficits, respectively? Research Question #1c: How much unique variance in overall supervisee counseling self- efficacy is accounted for by focus on strengths and constructive focus on deficits, respectively? Both focus on strengths and constructive focus on deficits were important predictors of satisfaction with supervision. Focus on strengths, however, was a stronger predictor than constructive focus on deficits, accounting for almost three times as much 143 unique variance (34% compared to 12%) in satisfaction with supervision. This result suggests that supervisors should focus more on strengths than on deficits. Similarly, although both the strengths and deficits subscales uniquely predicted the supervisory working alliance, focus on strengths was a stronger predictor than constructive focus on deficits, accounting for almost five times as much unique variance (34% compared to 7%) in the supervisory working alliance. This result also suggests that supervisors should focus more on strengths than on deficits. In regards to why focus on strengths was a stronger predictor of satisfaction with supervision and the supervisory working alliance, two possible explanations are as follows. First, focusing on supervisee strengths may be a more powerful teaching method than focusing on supervisee deficits because reinforcement is a more powerful technique than punishment to teach correct or desired behaviors (although our conceptualization of constructive focus on deficits is not the same as punishment; e.g., Pierce & Epling, 1999). Specifically, reinforcement of strengths and behaviors approximating desired standards provides supervisees with information about what clinical behaviors are working well and should therefore be continued or further developed (e.g., Bradley & Gould, 2001; Follette & Callaghan, 1995). Focusing on mistakes and deficits only informs supervisees of what not to do, but does not necessarily suggest appropriate alternative behaviors or directions for improvement (e.g., Abbott & Lyter, 1998; Catania, 2001; Myers, 1999). To the extent that focusing on strengths may be a better teaching method than focusing on deficits, supervisees may learn more, feel more satisfied with their learning, and feel more connected to their supervisors who facilitate their learning. A second explanation may be that focus on strengths inherently feels better than focus on deficits; it is part of 144 human nature to like hearing praise about oneself more than criticism (no matter how constructively the criticism is offered), and trainees are no exception (e.g., Barnett et al., 2001; Talen & Schindler, 1993). Although supervisees want their supervisors to focus on their deficits somewhat so that they can correct problems in their clinical work (e.g., Abbott & Lyter, 1998), they may enjoy supervision more and feel a stronger emotional bond with their supervisors when they are affirmed and validated in their strengths (e.g., Talen & Schindler, 1993; Briggs & Miller, 2005). Overall, the importance of both strengths and deficits foci as predictors of supervision process and outcome supports positive psychology theory regarding the importance of developing strengths while managing weaknesses (e.g., Lopez & Snyder, 2003). The fact that strengths focus was a bigger predictor also extends positive psychology theory regarding the potential to enhance satisfaction with supervision and the supervisory working alliance by focusing on supervisees? strengths (e.g., Linley & Joseph, 2004). In contrast to the findings that both focus on strengths and constructive focus on deficits uniquely predicted satisfaction with supervision and the supervisory working alliance, only focus on strengths uniquely predicted counseling self-efficacy (although it only accounted for 7% of the variance). As discussed previously, other variables (e.g., practice or counseling experience) may influence counseling self-efficacy more so than focusing on either strengths or deficits. Research Question #2: How much do supervisees perceive that their supervisors focus on strengths as compared to deficits? Supervisees in the present sample perceived a high degree of both focus on 145 strengths and constructive focus on deficits, rating their supervisors as focusing on their strengths an average of 5.47 and on their deficits an average of 4.97 out of 7 points. Moreover, these supervisees perceived their supervisors as focusing significantly more on their strengths than on their deficits, although they focused a lot on each. This finding conflicts with the positive psychology literature, which claims that psychology has traditionally focused disproportionately on weaknesses while neglecting strengths (e.g., Seligman & Czikszentmihalyi, 2000). One possible explanation for why supervisees perceive a greater focus on their strengths than on their deficits may be that it is probably easier for supervisors to praise supervisees or give them positive feedback about what they are doing well; supervisors can be relatively sure the compliments will be well-received (Heckman-Stone, 2003; Talen & Schindler, 1993). It is likely more difficult to tell supervisees what they are not doing well in a way that preserves self-esteem and does not elicit so much defensiveness that the feedback cannot be heard (Gould & Bradley, 2001). In other words, constructively focusing on supervisees? deficits may require more effort, skill and tact than focusing on their strengths. Indeed, previous research suggests that supervisors often hesitate or have difficulty in providing constructive negative feedback (e.g., Ladany & Melincoff, 1999; Hoffman et al., 2005; Chur-Hansen & McLean, 2006). Alternately, the explanation for the greater strengths than deficits focus might be related to the positive psychology notion of focus on strengths as a ?deep strategy? of effective clinical work (Seligman, 2002). Namely, in clinical settings therapists likely focus extensively on client strengths (without necessarily having been formally trained to do so) as a ?deep strategy? of effective therapy. Supervisors might likewise focus on 146 strengths as a ?deep strategy? of effective supervision (without necessarily having been trained in a strengths model), particularly with therapists-in-training who are still developing their skills and confidence as therapists (Briggs & Miller, 2005). Therefore, one reason for the greater supervisory focus on strengths than deficits might be that the supervisees in the sample were all therapists-in-training. Although a diverse range of training levels were represented, all supervisees were still novice therapists by virtue of being in training, and supervisors may, on average, focus more on the strengths than the deficits of trainees, perhaps as a way of bolstering their confidence and professional identity (e.g., Briggs & Miller, 2005; Stoltenberg et al., 1998; Barnett et al., 2001). Research Question #3a: How do the subscales of focus on strengths and constructive focus on deficits relate to supervisee training level? Neither focus on strengths nor constructive focus on deficits was related to supervisee training level (i.e., supervisees? year in their program or number of clinical hours). In other words, across training levels ranging from beginner (e.g., first practicum experience) to more advanced (e.g., predoctoral intern), supervisees perceived similar levels of focus on strengths and constructive focus on deficits from their supervisors. The lack of these relationships is an interesting finding, given that supervisees likely vary across training level in confidence as well as competence/skills (e.g., Stoltenberg et al., 1998), both of which might influence supervisors? use of strength or deficit focused interventions (Briggs & Miller, 2005; Larson, 1998). One explanation for the lack of relationships may be that supervisors do not vary their amount of focus on strengths or on deficits by supervisees? training level, but perhaps vary the content of the strengths or deficits focus. For example, a supervisor might compliment a very beginning trainee on a 147 well-delivered reflection of feelings, whereas the supervisor might compliment a more advanced trainee on making an accurate differential diagnosis of a client. Research Question #3b: How do the subscales of focus on strengths and constructive focus on deficits relate to supervisee and supervisor theoretical orientations? Although supervisee theoretical orientation was not related to either the focus on strengths or constructive focus on deficits subscales, several supervisor theoretical orientations (as rated by supervisees about their supervisors) were related to the subscales of the SUPSAD-S. First, focus on strengths correlated positively to perceived supervisor identification with and adherence to humanistic/existential/client-centered theoretical orientation. This finding makes sense given the primary role of supervisee strengths in client-centered supervision; for example, client-centered supervisors trust in supervisees? strengths, potential, motivation for growth, and natural tendency to move in an actualizing direction (e.g., toward greater counseling competency and self-efficacy) if provided with a nurturing environment that includes the facilitative conditions (i.e., empathy, genuineness, unconditional positive regard) (Rogers, 1963; Frankland, 2001; Patterson, 1997). Perhaps the reason why this correlation was small, however, is because client-centered supervision theory focuses not so much on what supervisors do in supervision (as is the focus on the SUPSAD-S), but more on supervisors? attitudes and assumptions about their supervisees? strengths and creation of an atmosphere where strengths can emerge (Patterson, 1997; Joseph & Linley, 2004; Gelso & Woodhouse, 2003). Both the focus on strengths and constructive focus on deficits subscales were positively related to perceived supervisor multicultural theoretical orientation. The 148 relationship to focus on strengths makes sense because a multiculturally-oriented supervisors might be especially sensitive to issues of power differential in the supervision relationship, given the focus on the lack of power of disadvantaged populations in society (e.g., racial minorities) in multicultural theories (e.g., Ladany et al., 1997; Bernard & Goodyear, 2004). Multicultural supervisors may thus be likely to establish a collaborative supervision relationship that minimizes the power differential, validates the supervisee?s strengths, and encourages the supervisee?s autonomy. The relationship of multicultural theoretical orientation to constructive focus on deficits also makes sense, as multicultural supervisors might try to enhance supervisees? multicultural competence by addressing their biases or lack of knowledge and skills relevant to working with specific culturally diverse groups (e.g., Ladany et al., 1997). Integrative supervisor orientation was also positively related to supervisory focus on strengths. Although it is hard how to interpret this finding given that integrative could entail many different combinations of approaches, perhaps it suggests that supervisors who are open to seeing the strengths in various perspectives are also open to seeing strengths in their supervisees. Neither solution-focused nor narrativist-constructivist supervisor theoretical orientations were related to focus on strengths even though these are explicitly strength- focused supervision approaches (e.g., Marek et al., 1994; Timm & Blow, 1999). One reason might be that these approaches are less common, and participants may not have known whether or not their supervisors adhered to these approaches. Research Question #3c: Are there differences in mean levels of focus on strengths or constructive focus on deficits for clinical versus counseling psychology trainees, or for 149 trainees whose supervisors hold a degree in counseling psychology versus those whose supervisors hold a degree in clinical psychology? No differences were found between supervisees in counseling psychology training programs and supervisees in clinical psychology training programs in their perceptions of their supervisors? focus on strengths or constructive focus on their deficits. Similarly, no differences were found between supervisees whose supervisors had a degree in counseling psychology versus those whose supervisors had a degree in clinical psychology on either subscale of the SUPSAD-S. These findings are interesting given that a focus on strengths or hygiology has been central to counseling psychology?s identification as a unique applied specialty, distinct from closely related areas like clinical psychology, which is traditionally thought to be based more on the medical or pathology-focused model (Gelso & Fretz, 2001; Super, 1955). However, the positive psychology movement initiated by Martin Seligman in 1998 has sparked interest in the wider field of psychology in the concept of focusing on strengths as well as weaknesses (e.g., Seligman & Csikszentmihalyi, 2000; Sheldon & King, 2001). Perhaps as a result, positive psychology theory and research has been incorporated into disciplines other than just counseling psychology, including clinical psychology (Martin Seligman himself is, in fact, a clinical psychologist). Thus, the reality seems to be that trainees in both counseling and clinical psychology have similar supervision experiences in terms of their supervisors? attention to their strengths and deficits. In addition, some recent counseling psychology literature has suggested that counseling psychology training is not sufficiently grounded in the strengths model, such that counseling psychologists may lack the skills to implement strengths-oriented 150 interventions (Gerstein, 2006). In fact, counseling psychology trainees (and their supervisors) were highly represented among the present sample of supervisees who perceived a greater degree of supervisory focus on strengths than on deficits (and a large degree of both). Limitations Although efforts were made to operationalize the constructs of focus on strengths and constructive focus on deficits as thoroughly and broadly as possible (i.e., through a comprehensive review of the literature and focus groups with the population targeted by the measure; Dawis, 1987), it is still possible that the construct was too narrowly defined or included aspects not identified in the current investigation. Cook and Campbell (1979) assert that the best way to understand and measure a construct is to assess it with multiple measures and multiple methods. Therefore, a second limitation of this study is that it relied on a single method approach, specifically the use of all self-report measures from the supervisees? perspective, without corroborating evidence from supervisors? perspectives or third-party observation of supervision sessions. Although I was primarily interested in supervisees? perceptions of their supervisors? focus on strengths and deficits (more than in supervisors? actual focus on strengths and deficits), this reliance on a single-method approach could possibly introduce a mono-method bias. The use of self-report measures is also problematic because of the possibility of biased responses from participants based on social desirability or wanting to help out the researcher by answering in a positive way. Although social desirability may have inflated the item means on the SUPSAD-S, a socially desirable response bias seems less likely 151 given that we asked supervisees, rather than supervisors, about how much supervisors focused on strengths and deficits (i.e., the supervisees were reporting about another person?s behavior, rather than their own). In addition, most positive psychology scholars view favorable self-presentation as an inevitable part of the context of positive psychology that should be acknowledged but not corrected (Lopez et al., 2003). Thus, social desirability may be a limitation of much of positive psychology research, in addition to the present study. A bigger concern than social desirability might be a halo effect (e.g., Blodgett, Schmitt, & Scudder, 1987), whereby supervisees who liked their supervisors may have rated various aspects of the supervisor?s behavior in a (perhaps unrealistically) positive light, regardless of the supervisor?s actual behavior. While social desirability or the halo effect may have contributed to the high mean item scores on the SUPSAD-S, high scores also seem appropriate given the view that, on average, strength and deficit focus in actual supervision practice should be on the positive or stronger side. In addition, self-report measures are limited by the accuracy of participants' memories. Thus, we cannot determine whether supervisees' recall of their supervisors' focus on strengths and deficits reflects accurately their perceptions of what occurred during supervision. Moreover, some participants completed the survey retrospectively with respect to supervisors they had already terminated with and/or had not met with in several weeks. Some test-retest participants completed the survey retrospectively (with respect to Fall supervisors) and then retrospectively two-weeks later, thus heightening the possibility of memory distortion (although test-retest reliability was high, which obviates some of this concern). 152 Another possible limitation is that participants may have self-selected to complete the survey. For example, supervisees with a particular interest in supervision or in supervisory focus on strengths and deficits may have volunteered. I attempted to minimize this possibility by encouraging the participation of as many identified prospective participants as possible (i.e., through several follow-up contacts). Also, to eliminate the possibility of bias introduced by supervisees selecting especially good or especially poor supervision experiences, participants were instructed to respond to the measures with respect to the supervisor with whom they met most recently. The quality of current supervision experiences seems more likely to vary randomly among participants than if participants had been allowed to select the supervision experience of their choice. In addition, by some statistical standards, the return rate for the current study is a potential limitation. However, obtaining a 45% response rate for an internet study is actually quite good. Overall response rates for web surveys are typically somewhat lower than paper and pencil surveys, for reasons such as the ease of discarding email messages and the lesser likelihood of emails getting recipients? attention because they do not physically show up on one?s desk (e.g., Yun & Trumbo, 2000; Kittelson, 1995). Also, the population surveyed in the present study, graduate mental health trainees, is heavily recruited to participate in internet research studies and may be getting selective about volunteering to participate. I followed several procedures suggested by internet research literature for increasing response rate, including sending multiple follow-up emails, sending personalized emails where possible (in the sub-sample of participants who were forwarded the recruitment email by their training directors, it was not possible to obtain 153 individual email addresses for privacy reasons), and the provision of incentives to part of the sample (Yun & Trumbo, 2000; Kittelson, 1995). These procedures increased the response rate. Nevertheless, the low response rate in the present study leaves open the possibility of confounding due to self-selection and a biased or non-representative sample (Heppner, Kivlighan, & Wampold, 1999). A further limitation is the correlational design, which prevents inferences about causality from being made (Heppner et al., 1999). For example, it is not possible to determine whether perceived supervisory focus on strengths or deficits causes improvements to the supervisory working alliance or increased satisfaction with supervision, or whether the situation is reversed (e.g., supervisees with good relationships with their supervisors who feel satisfied with supervision perceive their supervisors as focusing on their strengths and/or deficits). In addition, our use of the technique of exploratory factor analysis (EFA) has several limitations. First, EFA may be thought of as ?garbage in, garbage out,? such that factors are determined by and depend on the quality of the items created (e.g., Floyd & Widaman, 1995). To combat this problem, steps were taken during the measure development phase to ensure that the items subjected to factor analysis were as high quality as possible. Second, the process of identifying (e.g., deciding on the best factor solution) and naming factors is somewhat subjective. Third, the use of a single sample to explore factor structure, without testing whether the factor structure replicates on a confirmation sample, allows for the possibility of findings by chance alone and unknown stability of the factor structure. Consequently, this study should be viewed as exploratory. 154 Also, the concurrent validity of the SUPSAD-S measure could not be assessed because I could not identify a measure of a closely enough related construct. This concern is somewhat alleviated, however, given that the hypotheses regarding convergent validity were supported. In addition, the lack of similar measure to the SUPSAD-S in the literature highlights the newness of the current research and the need for a measure of supervisory strengths and deficits in the supervision literature. Finally, only one order of measures in the survey was used for all participants; therefore, possible order effects could not be assessed. Implications for Supervision Practice and Training The results of the present study suggest that both focus on strengths and constructive focus on deficits predict positive supervision outcomes, with focus on strengths being a stronger predictor than constructive focus on deficits. Hence, supervisors should focus on both, although they should focus more on strengths than on deficits. This is similar to the idea of the ?compliment sandwich,? in which supervisors comment on supervisees? strengths, identify a specific problem, and finish with a motivating or esteem enhancing statement (e.g., Chur-Hansen & McLean, 2006). The results of this study may also inform the training of supervisors. Given that both focus on strength and constructive focus on deficits were related to positive outcomes in terms of the supervisory relationship and satisfaction with supervision, it may be helpful to train new supervisors to implement these approaches into their work. In particular, supervisors may need training in how to focus constructively on deficits, given literature citing supervisors? difficulties with constructive negative feedback (e.g., Ladany & Melincoff, 1999). The SUPSAD-S itself might serve as a training tool in 155 supervision practica or supervision-of-supervision; supervisors-in-training might review its items as examples of strength and constructive deficit focused interventions, and might self-monitor as to how much of each focus they implement in their supervision sessions. In addition, supervisors might have their supervisees complete the SUPSAD-S and use their responses as feedback about the extent to which they focus on their supervisees? strengths and deficits, as perceived by their supervisees. Implications for Future Research First, future researchers should continue to evaluate the psychometric properties of the SUPSAD-S. The initial step in this process would be to perform a confirmatory factor analysis on a new sample to test the stability of the factor structure. It is possible that supervisory focus on strengths and deficits may predict other aspects of supervisee professional functioning, development, and identity that were not investigated in the present study. Some areas of fruitful future research might use the SUPSAD-S to investigate how these foci relate to variables such as anxiety and comfort in the counseling role, use of impression management strategies, persistence and motivation to learn therapy, satisfaction with the counseling role, actual performance in the counseling role, amount or depth of learning about therapy, and career development variables such as degree of interest in and goals regarding, counseling as a central activity in their occupational lives (e.g., Larson & Daniels, 1998; Lent et al., 1998; Lent et al., 2003; Ladany et al., 1996; Heppner et al., 1996). The SUPSAD-S was conceptualized as a global measure of supervisees? perceptions of focus on strengths and constructive focus on deficits over the entire course of supervision with a particular supervisor, and validity and reliability information 156 supports this conceptualization. However, supervisors are also likely to work with supervisees? strengths and deficits to varying degrees in different supervision sessions, depending on the needs of the supervisee for that session. Therefore, the creation of a session-level version of the SUPSAD-S would be useful and might have practical utility for supervisors. For example, supervisors could use the measure to get feedback from their supervisees about the extent to which they are facilitated a focus on the supervisee?s strengths and deficits in a particular session. Also, supervisees might be better able to remember their supervisors? interventions in their most recent supervision session, thus reducing error associated with responding about global impressions of their supervision experience. Such a measure could also be used to track supervisory focus on strengths and deficits from session to session over the course of a semester and see how each focus varies depending on what issues are salient in the supervisory relationship, the supervisee?s development as a counselor, and the supervisee?s caseload. Moreover, researchers could also examine supervisor perspectives by developing a parallel supervisor form of the SUPSAD-S. Having parallel supervisee and supervisor forms of the SUPSAD-S would allow for research investigating how much supervisor and supervisee perspectives coincide and how much ?distortion? is going on in either direction. For example, a supervisor may perceive that s/he is focusing a lot on both supervisee strengths and deficits, whereas the supervisee may perceive that the supervisor is only focusing on deficits. Researchers could also investigate whether convergence of supervisor and supervisee perspectives on strengths and deficits focus is associated with better supervision outcomes, similar to therapy research suggesting that a greater convergence of therapist and client perceptions of events in therapy is associated with 157 better therapy outcomes (e.g, Kivligihan & Arthur, 2000). In addition to investigating supervisor and supervisee perceptions of supervisory focus on strengths and constructive focus on deficits, it may be helpful to investigate these foci from a more ?objective? behavioral perspective, for example by having a trained observer rate supervisor interventions as strength or deficit focused Comparing a behavioral measure to supervisor or supervisee perceptions of supervisory focus on strengths and constructive focus on deficits (on the SUPSAD-S) might yield interesting discrepancies that would have practical utility for supervisors. For example, a supervisor may believe that s/he is focusing on both strengths and deficits in a session, but behavioral observations may suggest that s/he is only focusing on one or on neither. Or, behavioral observations might suggest that a supervisor is focusing on both strengths and deficits, but their supervisee might not be ?taking in? one or both of the supervisors types of focus. Qualitative investigations of the supervisor perspective could also be conducted to gain a richer and more in-depth understanding of how, when, and why they use interventions aimed at focusing on the strengths or deficits of their supervisees. Supervisee perceptions of focus on strengths and constructive focus on deficits could be investigated qualitatively as well to learn more about their experiences of strengths and deficits foci in supervision, their sense of what types of strength and deficit focused interventions are helpful or unhelpful, how they incorporate strengths and deficit focused feedback into their clinical work, and how each type of focus impacts them professionally and personally. Another area for future research would be to determine whether foci on strengths 158 and deficits are differentially helpful for some types of supervisees over others. Supervisee individual difference variables such as personality and pathology factors (e.g., perfectionism, neuroticism, negative affectivity, depression) as well as self-esteem might affect what supervisees take in vis a vis strength and deficit focus and what they do with the feedback. In general, more research is needed to address the issue of balanced focus on strengths and deficits. The SUPSAD-S measure was used in the current study to do some preliminary investigation of the notion of balanced focus on strengths and deficits, but was not itself a measure of balanced focus. More research is needed to understand what proportion of strengths and deficits focus is most helpful for what types of supervisees (e.g., training level, personality, demographic and cultural variables, anxiety level, etc.) under what conditions (training goals, caseload issues, type of training experience, status of supervisory working alliance, etc.). More research is also needed to operationalize more specifically what balanced focus means. It may mean proportion or amount of strengths relative to deficits focus as explored in the present study using the SUPSAD-S. Alternately, balanced strength and deficit focus may refer to specific types or styles of supervision interventions (e.g., compliment sandwich; Chur-Hansen & McLean, 2006), or an overall atmosphere of supervision (e.g., in the context of supervision relationships where the supervisors frequently communicate a belief in supervisees? strengths, supervisees may welcome a focus on deficits as necessary, trusting that their supervisors want to see them succeed; Briggs & Miller, 2005). In addition, experimental research in which focus on strengths and constructive focus on deficits are manipulated (e.g. using an audio-visual analogue design with 159 supervisees rating how satisfied or comfortable they would be working with a fictitious supervisor) might contribute to out knowledge of whether strengths or constructive deficits focus lead to or cause positive supervision outcomes. Finally, and perhaps most importantly, research should be done to investigate how focusing on strengths and deficits in supervision influences the process and outcome of the therapy provided by the supervisee. Given that the ultimate purpose of clinical supervision is to help supervisees provide the best possible treatment to their clients (Freitas, 2002; Holloway & Neufeldt, 1995; Bernard & Goodyear, 2004), we need to know more about how the processes of focusing on strengths and deficits in supervision translate to therapy practices. It is possible that focusing on strengths as well as deficits in supervision might train supervisees to take a similar approach (e.g., via role modeling, experiencing the benefits firsthand, or parallel process; Barnett et al., 2001; Larson, 1998; Friedlander et al., 1989) with their clients, building on client strengths in addition to addressing dysfunction and pathology (e.g., Joseph & Linley, 2004; Triantafillou, 1997). Conclusion In conclusion, the present study offers a new perspective on the complex yet important application of positive psychology theory to the domain of supervision. The subscales that emerged from the factors analysis provide initial insight into the ways in which supervisors may focus constructively on supervisees? strengths and deficits, and how each relates to supervisee development and supervision process and outcome. Continued empirical research in this area will help shed further light on this important topic. 160 Appendix A: Web Survey Introductory Page and Eligibility Requirements Welcome to my Supervision Study I greatly appreciate your consideration of my dissertation research. Please allow 15-20 minutes to complete my survey in one sitting. Before participating, please proceed to the next page to read about eligibility requirements. **Participant clicks ?Next? to proceed and is taken to Eligibility page below Eligibility to Participate in this Study To be eligible to participate in this study, you must meet BOTH of the following criteria: 1) you must be a trainee or intern in a masters' or doctoral program in one of the mental health professions (e.g., counseling psychology, clinical psychology, counselor education, social work, college student personnel, MFT, school counseling, or any other mental health professional program) AND 2) you must be CURRENTLY receiving supervision of your clinical work, or have received supervision of their clinical work within the last semester (e.g., Fall, 2006) Please indicate "YES" if you meet BOTH of these criteria, or "NO" to indicate that do you NOT meet BOTH of these criteria. You may continue participating only if you meet BOTH of these criteria. ___YES?I meet BOTH of these criteria ___NO--I do NOT meet BOTH of these criteria **Clicking ?Yes,? takes participants to Informed Consent page, see Appendix B; Clicking ?No,? takes participants to a page that contains the question below: Please click below to indicate which of the two criteria for participation in this study you do not meet. Check all that apply. ___I am not a trainee or intern in a mental health professional program. ___I am not currently receiving supervision of my clinical work or have not received supervision my clinical work within the last semester **After clicking either or both of these two criteria, participants are taken to the following page explaining why they were not eligible to participate in study: Thank you for your consideration of my study! My study is about supervision experiences for therapists-in-training. Thus, it requires participants to be therapists-in-training in masters' or doctoral programs who are currently receiving or have recently received (i.e., within the last semester) supervision of their clinical work. Your response indicated that you did not meet one or both of these criteria, so you would not be eligible to participate. I appreciate your interest in my study and would be happy to answer any questions. If you have any questions about my study, please contact either Dr. Clara Hill (Department of Psychology, University of Maryland, College Park, 2147G Biology-Psychology Building, College Park, MD 20742; phone: 301-405-5791; email: hill@psyc.umd.edu) or Ms. Melissa Roffman (Department of Psychology, University of Maryland, College Park, 2147H 161 Biology-Psychology Building, College Park, MD 20742; phone: 240-687-6040; email: msroffman@gmail.com). If you have questions about your rights as a research subject or wish to report a research- related injury, please contact: Institutional Review Board Office, University of Maryland, College Park, Maryland, 20742; (e-mail) irb@deans.umd.edu; (telephone) 301-405-0678. 162 Appendix B: Informed Consent Form Please read the Informed Consent below. After reading the Informed Consent, please provide your electronic signature, and then click one of the two boxes at the bottom of the page to indicate whether or not you agree to participate in this research. INFORMED CONSENT FOR PARTICIPATION IN THIS STUDY PROJECT TITLE Supervisee Perceptions of Supervisory Focus on Strengths and Deficits: Development and Validation of a Measure WHY IS THIS RESEARCH BEING DONE? This is a research project being conducted by Dr. Clara Hill and Ms. Melissa Roffman at the University of Maryland, College Park. We are inviting you to participate in this research project because you are a therapist-in-training. The purpose of this research project is to investigate the processes and interventions used by supervisors to facilitate a constructive focus on supervisees' strengths and deficits. By examining and researching these processes more closely, perhaps supervisors may eventually be trained to work more effectively with the strengths and deficits of their supervisees. WHAT WILL I BE ASKED TO DO? The procedures involve completing a series of measures via an online (internet) survey, in which you will respond to questions about yourself as a therapist, as well as your perceptions of your current or previous supervisor and supervision experiences with this supervisor. Questions from the survey will be in Likert scale format (e.g., rating on a scale where 1=strongly disagree and 7=strongly agree), and an example item is "My supervisor focused on my strengths." Another example item is "I feel comfortable working with my supervisor", rated on a scale of 1=almost never to 7=almost always. Participation in this study, involving completion of the entire survey in one sitting in any location where you have internet access, will require a 15-20 minute time commitment. In addition, you may be contacted by email and asked to complete a time-sensitive second copy of the survey that will only take 5 minutes to complete, also via online survey. WHAT ABOUT CONFIDENTIALITY? We will do our best to keep your personal information confidential. To help protect your confidentiality, (1) your name will not be included on the surveys and other collected data; (2) a four-digit code will be placed on the survey and other collected data; (3) through the use of an identification key, the researcher will be able to link your survey to your identity; and (4) only the researcher will have access to the identification key. Data will then be saved in a password-protected file on the student investigator's computer. Only the student investigator will know the password, thus ensuring that other individuals do not have access to data. In addition, when reporting the results of this study, only aggregate data will be reported. If we write a report or article about this research project, 163 your identity will be protected to the maximum extent possible. Your information may be shared with representatives of the University of Maryland, College Park or governmental authorities if you or someone else is in danger or if we are required to do so by law. One limitation inherent in internet research such as this study is that confidentiality CANNOT be completely guaranteed; in electronic submissions, there is always a small chance that information could be intercepted and read by a third party. However, given the focused nature of participant recruitment for this study (i.e., the study will not be widely advertised) and the probably limited value of the data to a third party, it seems unlikely that this data will be a target for interception. WHAT ARE THE RISKS OF THIS RESEARCH? There may be some risks from participating in this research study. It is possible that you may experience slight discomfort when asked to reflect on your experiences in supervision with your current supervisor when filling out the measures included in this study. However, this possible discomfort may be no greater than what you may routinely experience when reflecting on your training experiences as part of your professional development in your graduate program. WHAT ARE THE BENEFITS OF THIS RESEARCH? This research is not designed to help you personally, but the results may help the investigator learn more about the process of supervision so that, in time, supervision as a component of counselor training in graduate programs may be improved. We hope that, in the future, other people might benefit from this study through improved understanding of supervision training practices. Research informing training practices may eventually improve the effectiveness and competence of counselors being trained in graduate programs. DO I HAVE TO BE IN THIS RESEARCH? MAY I STOP PARTICIPATING AT ANY TIME? Your participation in this research is completely 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 any benefits to which you otherwise qualify. IS ANY MEDICAL TREATMENT AVAILABLE IF I AM INJURED? The University of Maryland does not provide any medical, hospitalization or other insurance for participants in this research study, nor will the University of Maryland provide any medical treatment or compensation for any injury sustained as a result of participation in this research study, except as required by law. WHAT IF I HAVE QUESTIONS? This research is being conducted by Dr. Clara Hill and Ms. Melissa Roffman at the University of Maryland, College Park. If you have any questions about the research study itself, please contact Dr. Clara Hill at: Department of Psychology, University of Maryland, College Park, 2147G Biology-Psychology Building, College Park, MD 20742; 164 phone: 301-405-5791; email: hill@psyc.umd.edu, or Ms. Melissa Roffman at: Department of Psychology, University of Maryland, College Park, 2147H Biology- Psychology Building, College Park, MD 20742; phone: 240-687-6040; email: msroffman@gmail.com If you have questions about your rights as a research subject or wish to report a research- related injury, please contact: Institutional Review Board Office, University of Maryland, College Park, Maryland, 20742; (e-mail) irb@deans.umd.edu; (telephone) 301-405-0678 This research has been reviewed according to the University of Maryland, College Park IRB procedures for research involving human subjects. This research has been approved by the University of Maryland, College Park Institutional Review Board (IRB), Approval #06-0566, expiration date 10/30/2007. STATEMENT OF AGE OF SUBJECT AND CONSENT [Please note: Parental consent always needed for minors.] Your electronic signature (typing in your name below) indicates that: you are at least 18 years of age, the research has been explained to you; your questions have been fully answered; and you freely and voluntarily choose to participate in this research project. Your name will be kept separate from the rest of your data. Name: ___________ Date: ____________ Please click below to indicate whether you agree or do not agree to participate in this research. __Yes, I agree to participate __No, I do not agree to participate **Clicking ?Yes? takes participants to measures in survey starting with SUPSAD-S (Appendix C); Clicking ?No? takes participants to a page that exits them from the survey, with the message below: Thank you for your consideration of my study! If you have any questions about this research, please contact either Dr. Clara Hill (Department of Psychology, University of Maryland, College Park, 2147G Biology- Psychology Building, College Park, MD 20742; phone: 301-405-5791; email: hill@psyc.umd.edu) or Ms. Melissa Roffman (Department of Psychology, University of Maryland, College Park, 2147H Biology-Psychology Building, College Park, MD 20742; phone: 240-687-6040; email: msroffman@gmail.com). If you have questions about your rights as a research subject or wish to report a research- 165 related injury, please contact: Institutional Review Board Office, University of Maryland, College Park, Maryland, 20742; (e-mail) irb@deans.umd.edu; (telephone) 301-405-0678. 166 Appendix C: Supervisory Focus on Strengths and Deficits Inventory-Supervisee Form (SUPSAD-S) Measure Before you begin, pick the supervisor you will be using in responding to all the questions throughout this survey. (FOR SUPERVISEES COMPLETING SURVEY WITH RESPECT TO FALL 2006 SUPERVISOR: This supervisor should be someone with whom you worked in the most recent semester for at least half a semester (e.g., Fall 2006). If you currently have or previously had more than one supervisor, please select the one you've met with the most recently)/(FOR SUPERVISEES COMPLETING SURVEY WITH RESPECT TO SPRING 2007 SUPERVISOR: This supervisor should be someone you are working with during the current (Spring 2007) semester. If you currently have more than one supervisor, please select the one you've met with the most recently.) Please indicate the extent to which you agree or disagree with the following statements regarding your experience with this supervisor over the ENTIRE course of your work together. Click the appropriate box next to each item to indicate your response (1-7), where 1=Strongly Disagree 2 3=Disagree 4 5=Agree 6 7=Strongly Agree My supervisor? 1. .... focused on my strengths. 2. .... framed critical feedback as his or her opinion, not a ?fact.? 3. .... gave me positive feedback about my skills as a therapist. 4. .... emphasized that it is normal to make mistakes when in training. 5. .... did NOT identify areas where I had improved. 6. .... identified areas where I excel as a therapist. 7. .... attended to my deficits because she/he wanted to see me improve. 8. .... did NOT praise my therapeutic skills. 9. .... commented on my weaknesses in ways that helped me take in the feedback. 10. .... did NOT communicate confidence in my abilities. 11. .... commented on what I did well as a supervisee. 12. .... helped me learn from my mistakes. 13. .... did NOT identify areas where I was competent. 14. .... helped me identify ineffective interventions I was using. 15. .... did NOT notice the confidence I?ve gained since beginning our work together. 16. .... used specific examples (e.g., from a tape from my session with a client) to highlight my areas needing improvement. 17. .... did NOT point out areas where I was making progress. 167 18. .... provided suggestions about how to resolve problems I was having in my clinical work. 19. ....did NOT reinforce my strengths. 20. .... helped me understand how my life experiences can enhance my clinical work (e.g., increased empathy or compassion for a client?s struggles). 21. .... did NOT help me identify where I was getting off track with a client. 22. .... used specific examples (e.g., from a tape from my session with a client) to provide evidence of my strengths. 23. .... encouraged discussion of what I could have done better in my clinical work. 24. .... did NOT identify areas where I excel as a therapist. 25. .... helped me understand how my personal issues might be interfering with my clinical work. 26. .... did NOT give enough positive feedback. 27. .... gave negative feedback in a tactful way. 28. .... asked me to identify areas where I feel most competent. NEXT PAGE ON WEBSITE: Please indicate the extent to which you agree or disagree with the following statements regarding your experience with your supervisor (the SAME SUPERVISOR you referred to when answering the questions in the previous section) over the ENTIRE course of your work together. Click the appropriate box next to each item to indicate your response (1- 7), where 1=Strongly Disagree 2 3=Disagree 4 5=Agree 6 7=Strongly Agree My supervisor? 29. .... praised my efforts to manage a difficult case. 30. .... framed my difficulties as a therapist as normal. 31. .... did NOT tell me what I did well in my clinical work. 32. .... did NOT give me constructive criticism. 33. .... identified areas where I was competent. 34. .... did NOT use specific examples (e.g., from a tape from my session with a client) when pointing out my weaknesses. 35. .... called my attention to areas where I had improved. 36. .... did NOT help me learn from my mistakes. 37. .... encouraged me to utilize my strengths to enhance my clinical work. 38. .... helped me brainstorm ways to address my problems with a client. 39. .... did NOT compliment me on what I did well. 40. .... helped me identify where I was getting off track with a client. 41. .... helped me understand how my life experiences make me well-suited to be a professional helper. 42. ?. noticed my less obvious strengths as a therapist. 168 43. .... gave me constructive negative feedback. 44. .... did NOT reinforce things I did well in my clinical work. 45. .... buffered negative feedback with positive feedback (i.e., ?compliment sandwich?). 46. .... helped me become more aware of my strengths as a therapist. 47. .... did NOT identify areas where I need improvement. 48. .... encouraged me to talk about cases that were going well. 49. .... did NOT help me become more aware of my weaknesses as a therapist. 50. .... praised me for the good work I had done with my clients. 51. .... checked in with me about my reactions to hearing negative feedback. 52. ?. did NOT try to build on my strengths. 53. .... communicated a belief that I would be able to overcome my deficits as a therapist. 54. .... encouraged discussion of what went well in my clinical work. 55. .... did NOT encourage discussion of what I could have done better in my clinical work. 169 Appendix D: Satisfaction with Supervision Questionnaire (Ladany, Hill, Corbett, & Nutt, 1996) Please respond to each of the following items below regarding your experience with your supervisor (the SAME SUPERVISOR you referred to when answering the questions in the previous sections) over the ENTIRE course of your work together. Indicate your response by clicking on the appropriate box. 1. How would you rate the quality of the supervision you have received? 1 2 3 4 Excellent Good Fair Poor 2. Did you get the kind of supervision you wanted? 1 2 3 4 No, definitely not No, not really Yes, generally Yes, definitely 3. To what extent has this supervision fit your needs? 1 2 3 4 Almost all of my needs Most of my needs Only a few of my None of my needs have been met have been met needs have been met have been met 4. If a friend were in need of supervision, would you recommend this supervision to him or her? 1 2 3 4 No, definitely not No, I don?t think so Yes, I think so Yes, definitely 5. How satisfied are you with the amount of supervision you have received? 1 2 3 4 Quite satisfied Indifferent or mildly Mostly satisfied Very satisfied dissatisfied 6. Has the supervision you received helped you to deal more effectively in your role as a counselor or therapist? 1 2 3 4 Yes, definitely Yes, generally No, not really No, definitely not 7. In an overall, general sense, how satisfied are you with the supervision you have received? 1 2 3 4 Very satisfied Mostly satisfied Indifferent or Quite dissatisfied mildly dissatisfied 170 8. If you were to seek supervision again, would you come back to this supervisor? 1 2 3 4 No, definitely not No, I don?t think so Yes, I think so Yes, definitely Score is sum of items after reverse scoring 1, 3, 6, 7 171 Appendix E: Supervisory Working Alliance Inventory (Trainee Form) (Efstation, Patton, & Kardash, 1990) Please indicate the frequency with which the behavior described in each of the following items seems characteristic of you and your work with your supervisor (the SAME SUPERVISOR you referred to when answering the questions in the previous sections). Click the appropriate box next to each item to indicate your response (1-7), where 1 = "Almost Never" and 7 = "Almost Always." 1 2 3 4 5 6 7 Almost Almost Never Always 1. I feel comfortable working with my supervisor. 2. My supervisor welcomes my explanations about my client?s behavior. 3. My supervisor makes the effort to understand me. 4. My supervisor encourages me to talk about my work with clients in ways that are comfortable for me. 5. My supervisor is tactful when commenting about my performance. 6. My supervisor encourages me to formulate my own interventions with the client. 7. My supervisor helps me to talk freely in our sessions. 8. My supervisor stays in tune with me during supervision. 9. I understand client behavior and treatment technique similarly to the way my supervisor does. 10. I feel free to mention to my supervisor any troublesome feelings I might have about him/her. 11. My supervisor treats me like a colleague in our supervisory sessions. 12. In supervision, I am more curious than anxious when discussing my difficulties with clients. 13. In supervision, my supervisor places a high priority on our understanding the client?s perspective. 14. My supervisor encourages me to take time to understand what the client is saying and doing. 15. My supervisor?s style is to carefully and systematically consider the material I bring to supervision. 16. When correcting my errors with a client, my supervisor offers alternative ways of intervening with that client. 17. My supervisor helps me to work within a specific treatment plan with my clients. 18. My supervisor helps me to stay on track during our meetings. 19. I work with my supervisor on specific goals in the supervisory session. Rapport scale: Sum items 1-12, divide by 12 Client Focus scale: Sum items 13-19, divide by 7 172 Appendix F: Counseling Activity Self-Efficacy Scales (Lent, Hill, & Hoffman, 2003) General Instructions: The following questionnaire consists of two parts. Both parts ask about your beliefs about your ability to perform various counselor behaviors or to deal with particular issues in counseling. I am looking for your honest, candid responses that reflect your beliefs about your current capabilities, rather than how you would like to be seen or how you might look in the future. There are no right or wrong answers to the following questions. CASES Part I Instructions: Please indicate how confident you are in your ability to use each of the following helping skills EFFECTIVELY, over the next week, in counseling MOST clients. Click the appropriate box next to each item to indicate your response (0- 9), where 0="No Confidence" and 9="Complete Confidence." How confident are you that you could use these general skills effectively with MOST clients over the next week? ___________________________________________________________ No confidence Complete confidence 0 1 2 3 4 5 6 7 8 9 ___________________________________________________________ 1. Attending (orient yourself physically toward the client). 2. Listening (capture and understand the messages that clients communicate). 3. Restatements (repeat or rephrase what the client has said, in a way that is succinct, concrete, and clear). 4. Open questions (ask questions that help clients to clarify or explore their thoughts or feelings). 5. Reflection of feelings (repeat or rephrase the client?s statements with an emphasis on his or her feelings). 6. Self-disclosure for exploration (reveal personal information about your history, credentials, or feelings). 7. Intentional silence (use silence to allow clients to get in touch with their thoughts or feelings). 8. Challenges (point out discrepancies, contradictions, defenses, or irrational beliefs of which the client is unaware or that he or she is unwilling or unable to change). 9. Interpretations (make statements that go beyond that the client has overtly stated and that give the client a new way of seeing his or her behavior, thoughts, or feelings). 10. Self-disclosures for insight (disclose past experiences in which you gained some personal insight). 11. Immediacy (disclose immediate feelings you have about the client, the therapeutic relationship, or yourself in relation to the client). 12. Information-giving (teach or provide the client with data, opinions, facts, resources, or answers to questions). 13. Direct guidance (give the client suggestions, directives, or advice that imply actions for the client to take). 173 14. Role-play and behavior rehearsal (assist the client to role-play or rehearse behaviors in-session). 15. Homework (develop and prescribe therapeutic assignments for clients to try out between sessions). CASES Part II Instructions: Please indicate how confident you are in your ability to do each of the following tasks EFFECTIVELY, over the next week, in counseling MOST clients. Click the appropriate box next to each item to indicate your response (0-9), where 0="No Confidence" and 9="Complete Confidence." How confident are you that you could do these specific tasks EFFECTIVELY with MOST clients over the next week? ___________________________________________________________ No confidence Complete confidence 0 1 2 3 4 5 6 7 8 9 ___________________________________________________________ 1. Keep sessions ?on track? and focused. 2. Respond with the best helping skill, depending on what your client needs at a given moment. 3. Help your client to explore his or her thoughts, feelings, and actions. 4. Help your client to talk about his or her concerns at a ?deep? level. 5. Know what to do or say next after your client talks. 6. Help your client to set realistic counseling goals. 7. Help your client to understand his or her thoughts, feelings, and actions. 8. Build a clear conceptualization of your client and his or her counseling issues. 9. Remain aware of your intentions (i.e., the purposes of your interventions) during sessions. 10. Help your client to decide what actions to take regarding his or her problems. 174 Appendix G: Public Self-Consciousness Scale (Fenigstein, Scheier, & Buss, 1975) Instructions: Use the scale below to answer how characteristic you feel the following statements are of yourself. Please click the appropriate box next to each item to indicate your response. 1 2 3 4 EXTREMELY SOMEWHAT SOMEWHAT EXTREMELY UNCHARACTERISTIC UNCHARACTERISTIC CHARACTERISTIC CHARACTERISTIC 1. I?m concerned about my style of doing things. 2. I?m concerned about the way I present myself. 3. I?m self-conscious about the way I look. 4. I usually worry about making a good impression. 5. One of the last things I do before I leave my house is look in the mirror. 6. I?m concerned about what other people think of me. 7. I?m usually aware of my appearance. 175 Appendix H: Demographic Questionnaire INSTRUCTIONS: Please complete the following items, either by clicking on your choice, or by typing in responses where appropriate. 1) Gender: ____Female ____Male ____Other (please specify) 2) Age: _____ 3) Race/Ethnicity: ____African-American ____European American/Caucasian ____Asian/Pacific Islander ____Hispanic/Latino ____Native American ____Middle Eastern ____Biracial/multiracial ____Other (Please specify: ______) 4) Please indicate the type of graduate training program you are currently enrolled in. ____Counseling Psychology Ph.D. program ____Counseling Psychology Masters' program ____Clinical Psychology Ph.D. program ____Clinical Psychology Psy.D. program ____Clinical Psychology Masters' program ____Counselor education Masters' program ____Counselor education Ph.D. program ____College Student Personnel Masters' program ____College Student Personnel Ph.D. program ____Masters' of Social Work program ____Marriage & Family Therapy (Masters') ____Mental Health Counseling (Masters') ____Rehabilitation Counseling (Masters') ____Other (please specify) 5) What year of your program are you in? ___1 st year ___5 th year ___2 nd year ___6 th year ___3 rd year ___7 th + year ___4 th year 6) Approximately how many hours of direct clinical experience (i.e., face-to-face with client) would you estimate you have completed? _____ 7) On a scale of 1 to 5, where 1 is "low" and 5 is "high," please rate how closely you believe in and adhere to each of the following theoretical orientations. Psychodynamic/Interpersonal 1 2 3 4 5 Humanistic/Existential/Client-Centered 1 2 3 4 5 Cognitive/Behavioral 1 2 3 4 5 Multicultural 1 2 3 4 5 176 Feminist 1 2 3 4 5 Narrativist/Constructivist 1 2 3 4 5 Solution-Focused 1 2 3 4 5 Integrative/Eclectic 1 2 3 4 5 8) Approximately how many sessions have you had with your current clinical supervisor (i.e., the supervisor you had in mind when you completed the survey in this study)? _____ sessions 9) Approximately how long have you been working with you current clinical supervisor (i.e., the supervisor you had in mind when you completed the survey in this study)? ______ months ______ semesters 10) What type of supervision do you receive from your current clinical supervisor (i.e., the supervisor you had in mind when you completed the survey in this study)? Check all that apply. _____ Individual supervision _____ Group supervision _____ Other (please specify: ____ ) 11) What is your current supervisor?s highest degree (i.e., the supervisor you had in mind when you completed the survey in this study)? Check all that apply. _____ Counseling Psychology Ph.D. _____ Clinical Psychology Ph.D. _____ Clinical Psychology Psy.D. _____ Counseling Psychology Masters? _____ Clinical Psychology Masters? _____ Counselor Education Masters? _____ Counselor Education Ph.D. _____ Social Work Masters? (LCSW) _____ Marriage & Family Therapy Masters? _____ Psychiatry (M.D.) _____ Other (please specify: _________) 12) On a scale of 1 to 5, where 1 is low and 5 is high, please rate how closely your supervisor (i.e., the supervisor you had in mind when you completed the survey in this study) believes in and adheres to each of the following theoretical orientations. If you are unsure of your supervisor's theoretical orientation, just make an educated guess. Psychodynamic/Interpersonal 1 2 3 4 5 Humanistic/Existential/Client-Centered 1 2 3 4 5 Cognitive/Behavioral 1 2 3 4 5 Multicultural 1 2 3 4 5 Feminist 1 2 3 4 5 Narrativist/Constructivist 1 2 3 4 5 Solution-Focused 1 2 3 4 5 177 Integrative/Eclectic 1 2 3 4 5 13) What type of training experience are you currently undergoing with your current supervisor (i.e., the supervisor you had in mind when you completed the survey in this study)? Please check all that apply in both columns. ___ Practicum ___ Externship ___ Pre-doctoral internship for psychology ___ Internship for social work 14) For what types of clinical work are you receiving from your current supervisor (i.e., the supervisor you had in mind when you completed the survey in this study)? Check all that apply. ___ individual emotional/social counseling ___ individual career/vocational/educational counseling ___ group therapy ___ couples therapy ___ family therapy ___ child/adolescent therapy ___ supervision-of-supervision ___ supervision of consultation or outreach ___ Other (please specify) **At the end of Demographic Questionnaire, clicking ?Next? takes participations to the following last page of the survey: THANK YOU FOR YOUR PARTICIPATION I greatly appreciate the time you took to participate in my study! Your participation will help generate knowledge about the processes used by supervisors to focus on their supervisees' strengths and deficits, which may one day be used to inform more effective supervisory practices. If you would like to be emailed a summary of the results of this research, please provide your email in the space below. Your email address will be kept separate from the rest of your data. ___________________ 178 Appendix I: Initial Recruitment Email Subject: Supervision Dissertation Study Dear (Name) or (Therapist-in-Training at X University), Have you ever noticed what your supervisors do, or don?t do enough of, to focus on your strengths and deficits as a therapist? My name is Missy Roffman, and I?m a 5 th year student in the counseling psychology program at the University of Maryland, College Park. My dissertation research is investigating supervisees? perceptions of the processes used by their supervisors to work with their strengths and weaknesses as clinicians. I am writing to invite you to participate in my study. Participation would involve completing a brief series of measures (total participation time of approximately 15 minutes), which can be accessed online via the URL link below. Your participation would be immensely helpful to me in generating knowledge that can hopefully one day contribute to more effective supervisory practices. Also, I hope you will find this survey will give you an opportunity to reflect on important aspects of your training experience. TO ACCESS THS STUDY, PLEASE CLICK HERE: http://www.surveymonkey.com/s.asp?u=312512980120 Thank you for your consideration. This research has been approved by the University of Maryland, College Park Institutional Review Board (IRB), Approval #06-0566. Please note that by agreeing to participate in this online survey, we are assuming that you are over 18 years of age and have provided your informed consent. If you have questions about your rights as a research subject or wish to report a research-related injury, please contact: IRB Office, University of Maryland, College Park, Maryland, 20742; (e-mail) irb@deans.umd.edu; (telephone) 301-405-0678. Sincerely, Melissa S. Roffman, M.A. Clara E. Hill, Ph.D. Doctoral Student Professor 240-687-6040 301-405-5791 mroffman@psyc.umd.edu hill@psyc.umd.edu 179 Appendix J: Two week/one month follow-up recruitment email Subject: Reminder about Supervision Dissertation Study Dear (Name) or (Therapist-in-Training at X University), My name is Missy Roffman, and I'm a 5th year student in the counseling psychology program at the University of Maryland, College Park. A (couple weeks)/(a month) ago, (I sent you an email)/(Dr. Y forwarded you an email) inviting you to participate in an online study on supervision experiences. If you have already responded--THANK YOU. If you haven't had a chance yet to respond, PLEASE think about doing so--I would be so grateful to have your data. Participation would involve completing a brief series of measures (total participation time of approximately 15 minutes), which can be accessed online via the URL link below. I know this is a commitment of time, but your participation would be immensely helpful to me in generating knowledge that can hopefully one day contribute to more effective supervisory practices. Also, I hope you will find this survey will give you an opportunity to reflect on important aspects of your training experience. The findings of this study will be available to you if you wish. If you do choose to participate, it would be a huge help to me if you could complete the entire survey?the first two pages are longer than the rest, so please stick with me! TO ACCESS THS STUDY, PLEASE VISIT THE FOLLOWING WEBSITE: http://www.surveymonkey.com/s.asp?u=312512980120 Thank you for your consideration. This research has been approved by the University of Maryland, College Park Institutional Review Board (IRB), Approval #06-0566. Please note that by agreeing to participate in this online survey, we are assuming that you are over 18 years of age and have provided your informed consent. If you have questions about your rights as a research subject or wish to report a research-related injury, please contact: IRB Office, University of Maryland, College Park, Maryland, 20742; (e-mail) irb@deans.umd.edu; (telephone) 301-405-0678. Sincerely, Melissa S. Roffman, M.A. Clara E. Hill, Ph.D. Doctoral Student Professor 240-687-6040 301-405-5791 msroffman@gmail.com hill@psyc.umd.edu 180 Appendix K: Follow-up recruitment letter with incentive Dear Therapist-in-Training at X University, My name is Missy Roffman, and I'm a 5th year student in the counseling psychology program at the University of Maryland, College Park. A few weeks ago, you received an email invitation forwarded by Dr. Y, inviting you to participate in my online dissertation study on supervision. If you have already participated, THANK YOU. If you haven't had a chance yet to respond, PLEASE think about doing so--I would be so grateful to have your data. Please accept this small gift as a token of my appreciation. Participation would involve completing a brief series of measures (total participation time of approximately 15 minutes), which can be accessed online by visiting the website listed below. I know this is a commitment of time, but your participation would be immensely helpful to me in generating knowledge that can hopefully one day contribute to more effective supervisory practices. Also, I hope you will find this survey will give you an opportunity to reflect on important aspects of your training experience. The findings of this study will be available to you if you wish. If you do choose to participate, it would be a huge help to me if you could complete the entire survey?the first two pages are longer than the rest, so please stick with me! TO ACCESS THE WEBSITE CONTAINING MY STUDY, YOU CAN EITHER: 1) EMAIL ME AT msroffman@gmail.com AND I WILL EMAIL YOU THE URL LINK TO THE STUDY WEBSITE OR 2) TYPE IN THE FOLLOWING URL LINK TO THE STUDY WEBSITE: http://www.surveymonkey.com/s.asp?u=312512980120 OR 3) FOLLOW THE LINK TO THE STUDY WEBSITE CONTAINED IN THE EMAIL YOU RECEIVED OR WILL SOON RECEIVE FROM DR. Y. Thank you for your consideration. This research has been approved by the University of Maryland, College Park Institutional Review Board (IRB), Approval #06-0566. Please note that by agreeing to participate in this online survey, we are assuming that you are over 18 years of age and have provided your informed consent. If you have questions about your rights as a research subject or wish to report a research-related injury, please contact: IRB Office, University of Maryland, College Park, Maryland, 20742; (e-mail) irb@deans.umd.edu; (telephone) 301-405-0678. Sincerely, Melissa (Missy) Roffman, M.A. Clara E. Hill, Ph.D. Doctoral Student Professor 240-687-6040 301-405-5791 msroffman@gmail.com hill@psyc.umd.edu 181 Appendix L: Test-Retest Recruitment Letter Subject: Supervision Dissertation Study?A Brief Part 2 Dear (Name), Approximately 2 weeks ago, you completed an online survey for my dissertation research investigating supervisees' perceptions of the processes used by their supervisors to work with their strengths and weaknesses as clinicians. It would be incredibly helpful to me if you could complete this brief (5-10 minute) version of my survey that will be used to gather data on the test-retest reliability of my measure. The findings of this study will be available to you if you wish, and hopefully your participation will generate good research karma for your own research endeavors! If you choose to participate, it is important that you fill out this brief survey as soon as possible, given the time sensitive nature of test- retest reliability data. The link to this survey can be found below. TO ACCESS THIS SURVEY, PLEASE CLICK HERE: http://www.surveymonkey.com/s.asp?u=460613080408 Thank you again for your time and interest in my study! This research has been approved by the University of Maryland, College Park Institutional Review Board (IRB), Approval #06-0566. Please note that by agreeing to participate in this online survey, we are assuming that you are over 18 years of age and have provided your informed consent. If you have questions about your rights as a research subject or wish to report a research- related injury, please contact: IRB Office, University of Maryland, College Park, Maryland, 20742; (e-mail) irb@deans.umd.edu ; (telephone) 301-405-0678. Sincerely, Melissa S. Roffman, M.A. Clara E. Hill, Ph.D. Doctoral Student Professor 240-687-6040 301-405-5791 msroffman@gmail.com hill@psyc.umd.edu 182 Appendix M: Structure Coefficients (Factor Loadings) and Pattern Coefficients of Deleted Items from the SUPSAD-S Structure Pattern Coefficients Coefficients Factor Factor Factor Factor My supervisor?. I II I II 2. .... framed critical feedback as his or her opinion, not a ?fact.? .600 .322 .617 -.03 4. .... emphasized that it is normal to make mistakes when in training. .606 .379 .579 .048 5. .... did NOT identify areas where I had improved. .561 .599 .326 .413 9. .... commented on my weaknesses in ways that helped me take in the feedback. .537 .653 .244 .514 10. .... did NOT communicate confidence in my abilities . .734 .380 .767 - .057 11. .... commented on what I did well as a supervisee. .624 .375 .607 .029 12. .... helped me learn from my mistakes. .536 .641 .253 .497 13. .... did NOT identify areas where I was competent. .731 .370 .771 -.07 15. .... did NOT notice the confidence I?ve gained since beginning our work together. .622 .514 .487 .236 17. .... did NOT point out areas where I was making progress. .625 .523 .484 .246 18. .... provided suggestions about how to resolve problems I was having in my clinical work. .429 .517 .198 .404 20. .... helped me understand how my life experiences can enhance my clinical work (e.g., increased empathy or compassion for a client?s struggles). .537 .589 .299 .418 22. .... used specific examples (e.g., from a tape from my session with a client) to provide evidence of my strengths. .642 .584 .458 .323 24. .... did NOT identify areas where I excel as a therapist. .875 .495 .878 -.01 25. .... helped me understand how my personal issues might be interfering with my clinical work. .361 .557 .064 .520 27. .... gave negative feedback in a tactful way. .591 .624 .349 .425 28. .... asked me to identify areas where I feel most competent. .525 .638 .238 .502 29. .... praised my efforts to manage a difficult case. .676 .362 .696 -.0 35. .... called my attention to areas where I had improved. .615 .720 .303 .547 36. .... did NOT help me learn from my mistakes. .707 .662 .489 .383 37. .... encouraged me to utilize my strengths to enhance my clinical work. .757 .536 .670 .154 38. .... helped me brainstorm ways to address my problems with a client. .513 .462 .369 .251 41. .... helped me understand how my life experiences make me well-suited to be a professional helper. .493 .516 .294 .348 42. ?. noticed my less obvious strengths as a therapist. .678 .494 .587 .160 45. .... buffered negative feedback with positive feedback (i.e., ?compliment sandwich?). .509 .541 .298 .371 48. .... encouraged me to talk about cases that were going well. .633 .456 .553 .141 52. .... checked in with me about my reactions to hearing negative feedback. .490 .610 .210 .491 53. ?. did NOT try to build on my strengths. .808 .512 .765 .076 54. .... communicated a belief that I would be able to overcome .527 .594 .279 .435 183 my deficits as a therapist. 55. .... encouraged discussion of what went well in my clinical work. .718 .542 .606 .196 Note. First Iteration of Exploratory Factor Analysis, Promax Rotation. N = 204. 184 References Abbott, A.A., & Lyter, S.C. (1998). The use of constructive criticism in field supervision. Clinical Supervisor, 17, 43-57. Allen, G.J., Szollos, S.J., & Williams, B.E. (1986). Doctoral students' comparative evaluations of best and worst psychotherapy supervision. Professional Psychology: Research and Practice, 17, 91-99. American Psychological Association Office of Program Consultation and Accreditation. (2005). Research Information from Annual Report Submissions, Accredited Doctoral Program Statistics Student Characteristics (Table 2) and Accredited Internship Program Statistics Intern Statistics (Table 7). Retrieved April 25, 2007, from http://www.apa.org/ed/accreditation/2005_research.html. Aspinwall, L.G., & Staudinger, U.M. (2003). A psychology of human strengths: Some central issues of an emerging field. In L.G. Aspinwall & U.M. Staudinger (Eds.), A psychology of human strengths: Fundamental questions and future directions for a positive psychology (pp. 9-22). Washington DC: American Psychological Association. Baker, D. E. (1990). Relationship of the supervisory working alliance to supervisor and supervisee narcissism, gender, and theoretical orientation. Dissertation Abstracts International, 51, (7-B), 3602?3603. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavior change. Psychological Review, 84, 191-215. Barnett, J.E., Youngstrom, J.K., & Smook, R.G. (2001). Clinical supervision, teaching, and mentoring: Personal perspectives and guiding principles. Clinical Supervisor, 20, 217-230. Bartlett, M. S. (1950). Tests of significance in factor analysis. British Journal of Psychology, 3, 77-85. Bernard, J.M., & Goodyear, R.K. (2004). Fundamentals of clinical supervision. 3 rd edition. Boston, MA: Pearson Education. Blocher, D. (1983). Toward a cognitive developmental approach to counseling supervision. The Counseling Psychologist, 11, 27-34. Blodgett, E.G., Schmidt, J.F., & Scudder, R.R. (1987). Clinical session evaluation: The effect of familiarity with the supervisee. Clinical Supervisor, 5, 33-43. Bob, S. (1999). Narrative approaches to supervision and case formulation. Psychotherapy: Theory/Research/Practice/Training, 36, 146-153. 185 Bordin, E.S. (1983). A working alliance based model of supervision. The Counseling Psychologist, 11, 35-41. Bradley, L.J., & Gould, L.J. (2001). Psychotherapy-based models of counselor supervision. In L.J. Bradley & N. Ladany (Eds.), Counselor supervision: Principles, process, and practice 3 rd edition (pp. 147-175). Philadelphia, PA: Taylor & Francis. Bradley, L.J., & Kottler, J.A. (2001). Overview of counselor supervision. In L.J. Bradley & N. Ladany (Eds.), Counselor supervision: Principles, process, and practice 3 rd edition (pp. 3-21). Philadelphia, PA: Taylor & Francis. Briggs, J.R., & Miller, G. (2005). Success enhancing supervision. Journal of Family Psychotherapy, 16, 199-222. Carlson, T.D., & Erickson, M.J. (2001). Honoring and privileging personal experience and knowledge: Ideas for a narrative therapy approach to the training and supervision of new therapists. Contemporary Family Therapy, 23, 199-220. Carver, C.S., & Scheier, M.F. (2002). Optimism. In C.R. Snyder & S.J. Lopez (Eds.), Handbook of positive psychology (pp. 231-243). New York: Oxford University Press. Catania, A.C. (2001). Positive psychology and positive reinforcement. American Psychologist, 56, 86-87. Chur-Hansen, A., & McLean, S. (2006). On being a supervisor: The importance of feedback and how to give it. Australasian Psychiatry, 14, 67-71. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2 nd ed). New Jersey: Lawrence Erlbaum. Cook, T.D., & Campbell, D.T. (1979). Quasi-experimentation: design & analysis issues for Field settings. Houghton Mifflin: New York Daniels, J.A., & Larson, L.M. (2001). The impact of performance feedback on counseling self-efficacy and counselor anxiety. Counselor Education and Supervision, 41, 120-130. Dawis, R. V. (1987). Scale construction. Journal of Counseling Psychology, 34, 481-489. deShazer, S. (1988). Clues: Investigating solutions in brief therapy. New York: Norton. Efstation, J.F., Patton, M.J., & Kardash, C.M. (1990). Measuring the working alliance in counselor supervision. Journal of Counseling Psychology, 37, 322-329. 186 Fenigstein, A., Scheier, M., & Buss, A. H. (1975). Public and private self-consciousness: Assessment and theory. Journal of Consulting and Clinical Psychology, 43, 522- 527. Floyd, F.J., & Widaman, K.F. (1995). Factor analysis in the development and refinement of clinical assessment instruments. Psychological Assessment, 7, 286-299. Follette, W.C., & Callaghan, G.M. (1995). Do as I do, not as I say: A behavior-analytic approach to supervision. Professional Psychology: Research and Practice, 26, 413-421. Frankland, A. (2001). A person-centred model of supervision. Counselling Psychology Review, 16, 26-31. Freitas, G.J. (2002). The impact of psychotherapy supervision on client outcome: A critical examination of 2 decades of research. Psychotherapy: Theory/Research/Practice/Training, 39, 354-367. Friedlander, M.L., Siegel, S.M., & Brenock, K. (1989). Parallel processes in counseling and supervision: A case study. Journal of Counseling Psychology, 36, 149-157. Friedlander, M.L., & Ward, L.G. (1984). Development and validation of the Supervisory Styles Inventory. Journal of Counseling Psychology, 31, 542-558. Gelso, C. J., & Fassinger, R. E. (1992). Personality, development, and counseling psychology: Depth, ambivalence, and actualization. Journal of Counseling Psychology, 39, 275-298. Gelso, C. J., & Fretz, B. (2001). Counseling Psychology (2 nd ed.). Fort Worth, Texas: Harcourt. Gelso, C.J., & Hayes, J.A. (2001). Countertransference management. Psychotherapy, 38, 418-422. Gelso, C. J., & Woodhouse, S. (2003). Toward a Positive Psychotherapy: Focus on Human Strengths. In W. B. Walsh (Ed.), Counseling psychology and human strengths (pp. 171-197). Mahwah: Erlbaum. Gelso, C.J., Kelley, F.A., Fuertes, J.N., Marmarosh, C., Holmes, S.E., Costa, C., & Hancock, G.R. (2005). Measuring the Real Relationship in Psychotherapy: Initial Validation of the Therapist Form. Journal of Counseling Psychology, 52, 640- 649. Gerstein, L.H. (2006). Counseling psychology?s commitment to strengths: Rhetoric or reality? The Counseling Psychologist, 34, 276-292. 187 Gorsuch, R. L. (1983). Factor analysis (2nd ed.). Hillsdale, NJ: Erlbaum. Gosling, S.D., Vazire, S., Srivastava, S., & John, O.P. (2004). Should we trust web-based studies?: A comparative analysis of six preconceptions about internet questionnaires. American Psychologist, 59, 93?104. Gould, L.J., & Bradley, L.J. (2001). Evaluation in supervision. In L.J. Bradley & N. Ladany (Eds.), Counselor supervision: Principles, process, and practice 3 rd edition (pp. 271-295). Philadelphia, PA: Taylor & Francis. Hahn, W.K. (2002). The experience of shame in psychotherapy supervision. Psychotherapy, 38, 272-284. Harbin, J.M. (2006). Therapist work with client strengths: Development and validation of a measure. Unpublished doctoral dissertation, University of Maryland, College Park. Harrar, W.R., VandeCreek, L., & Knapp, S. (1990). Ethical and legal aspects of supervision. Professional Psychology: Research and Practice, 21, 37-41. Heckman-Stone, C. (2003). Trainee preferences for feedback and evaluation in clinical supervision. Clinical Supervisor, 22, 21-33. Heppner, P. P., Kivlighan, D. K., & Wampold, B. E. (1999). Research design in counseling (2 nd Edition). New York: Brooks/Cole. Heppner, M.J., O?Brien, K.M., Hinkelman, J.M., & Flores, L.Y. (1996). Training counseling psychologists in career development: are we our own worst enemies? The Counseling Psychologist, 24, 105-125. Herrero, J., & Meneses, J. (2006). Short web-based versions of the Perceived Stress (PSS) and Center for Epidemiological Studies-Depression (CESD) Scales: A comparison to pencil and paper responses among internet users. Computers in Human Behavior, 22, 830?846. Hoffman, M.A., Hill, C.E., & Freitas, G.F. (2005). Supervisor perspective on the process and outcome of giving easy, difficult, or no feedback to supervisees. Journal of Counseling Psychology, 52, 3-13. Holloway, E.L., & Neufeldt, S.A. (1995). Supervision: Its contributions to treatment efficacy. Journal of Consulting & Clinical Psychology, 63, 207-213. Jackson, A. P. (1993, August). The Supervisory Working Alliance Inventory: An exploration of predictor and outcome variables. In K. D.Multon (Chair), The working alliance in counselor supervision Symposium conducted at the 101st 188 Annual Convention of the American Psychological Association, Toronto, Ontario, Canada. Joseph, S., & Linley, P. A. (2004). Positive therapy: A positive psychological theory of therapeutic practice. In P. A. Linley & S. Joseph (Eds.), Positive psychology in practice (pp. 354-370). Hoboken, NJ: Wiley. Juhnke, G.A. (1996). Solution-focused supervision: Promoting supervisee skills and confidence through successful solutions. Counselor Education and Supervision, 36, 48-57. Kahn, J.H. (2006). Factor analysis in counseling psychology research, training, and practice: Principles, advances, and applications. The Counseling Psychologist, 34, 1-35. Kauffman, C., & Scoular, A. (2004). Toward a positive psychology of executive coaching. In P.A. Linley & S. Joseph (Eds.), Positive psychology in practice (pp. 287-304). Hoboken, NJ: Wiley. Kadushin, A. (1992). What?s wrong, what?s right with social work supervision. Clinical Supervisor, 10, 3-19. Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39, 31-36. Keyes, C. L. M., & Lopez, S., J. (2002). Toward a science of mental health: Positive directions in diagnosis and interventions. In C. R. Snyder & S. J. Lopez (Eds.), Handbook of positive psychology (pp. 45?62). New York: Oxford. Kittleson, M. J. (1995). An assessment of the response rate via the Postal Service and e- mail. Health Values, 19(2), 27-39. Koob, J.L. (2002). The effects of solution-focused supervision on the perceived self- efficacy of therapists in training. Clinical Supervisor, 21, 161-183. Knapp, S.J., & VandeCreek, L.D. (2006). Practical ethics for psychologists: A positive approach. Washington, DC: American Psychological Association. Ladany, N., Ellis, M.V., & Friedlander, M.L. (1999). The supervisory working alliance, trainee self-efficacy, and satisfaction. Journal of Counseling and Development, 77, 447-455. Ladany, N., Hill, C.E., Corbett, M.M., & Nutt, E.A. (1996). Nature, extent, and importance of what psychotherapy trainees do not disclose to their supervisors. Journal of Counseling Psychology, 43, 10-24. Ladany, N., Inman, A.G., Constantine, M.G., & Hofheinz, E.W. (1997). Supervisee 189 multicultural case conceptualization ability and self-reported multicultural competence as functions of supervisee racial identity and supervisor focus. Journal of Counseling Psychology, 44, 284-293. Ladany, N., & Lehrman-Waterman, D. (1999). The content and frequency of supervisor self-disclosures and their relationship to supervisor style and the supervisory working alliance. Counselor Education and Supervision, 38, 143-160. Ladany, N., Lehrman-Waterman, D., Molinaro, M., & Wolgast, B. (1999). Psychotherapy supervisor ethical practices: Adherence to guidelines, the supervisory working alliance, and supervisee satisfaction. The Counseling Psychologist, 27, 443-475. Ladany, N., & Melincoff, D.S. (1999). The nature of counselor supervisor nondisclosure. Counselor Education and Supervision, 38, 161-176. Ladany, N., & Muse-Burke, J.L. (2001). Understanding and conducting supervision research. In L.J. Bradley & N. Ladany (Eds.), Counselor supervision: Principles, process, and practice 3 rd edition (pp. 304-324). Philadelphia, PA: Taylor & Francis. Ladany, N., Walker, J.A., & Melincoff, D.S. (2001). Supervisory style: Its relation to the supervisory working-alliance and supervisor self-disclosure. Counselor Education and Supervision, 40, 263-275. Lampropoulos, G. K. (2001). Integrating psychopathology, positive psychology, and psychotherapy. American Psychologist, 56, 87-88. Larson, L.M. (1998). The social-cognitive model of counselor training. The Counseling Psychologist, 26, 219-273. Larson, L.M., & Daniels, J.A. (1998). Review of the counseling self-efficacy literature. TheCounseling Psychologist, 26, 179-218. Larson, L.M., Suzuki, L.A., Gillespie, K.N., Potenza, M.T., Bechtel, M.A., & Toulouse, A.L. (1992). Development and validation of the Counseling Self-Estimate Inventory. Journal of Counseling Psychology, 39, 105-120. Lehrman-Waterman, D., & Ladany, N. (2001). Development and validation of the Evaluation Process within Supervision Inventory. Journal of Counseling Psychology, 48, 168-177. Lent, R.W., Hackett, G., & Brown, S.D. (1998). Extending social cognitive theory to counselor training: Problems and perspectives. The Counseling Psychologist, 26, 295-306. Lent, R.W., Hill, C.E., & Hoffman, M.A. (2003). Development and validation of the 190 Counselor Activity Self-Efficacy Scales. Journal of Counseling Psychology, 50, 97-108. Lewis, W.C. (2001). Transference in analysis and in supervision. In S. Gill (Ed.). The supervisory alliance: Facilitating the psychotherapist?s learning experience (pp. 75-80). Northvale, NJ: Jason Aronson. Linley, P.A., & Joseph, S. (2004). Positive psychology in practice. Hoboken, NJ: Wiley. Lopez, S. J., Edwards, L. M., Magyar-Moe, J. L., Pedrotti, J. T., & Ryder, J. A. (2003). Fulfilling its promise: Counseling psychology?s efforts to understand and promote optimal human functioning. In W. B. Walsh (Ed.), Counseling psychology and human strength (pp. 297-308). Mahwah: Erlbaum Lopez, S.J., & Magyar-Moe, J.L. (2006). A positive psychology that matters. The Counseling Psychologist, 34, 323-330. Lopez, S.J., Magyar-Moe, J.L., Petersen, S.E., Ryder, J.A., Krieshok, T.S, O?Byrne, K.K., Lichtenberg, J.W., & Fry, N. (2006). Counseling psychology?s focus on positive aspects of human functioning. The Counseling Psychologist, 34, 205-227. Lopez, S. J., & Snyder, C. R. (Eds.). (2003). Positive psychology assessment: A handbook of models and measures. Washington, DC: APA. Lopez, S. J., Snyder, C. R., & Rasmussen, H. N. (2003). Striking a vital balance: Developing a complementary focus on human weakness and strength though positive psychological assessment. In S. J. Lopez & C. R. Snyder (Eds.), Positive psychological assessment: A handbook of models and measures (pp. 3-20). Washington, DC: APA. Maddux, J.E. (2002). Self-efficacy: The power of believing you can. In C. R. Snyder & S. J. Lopez (Eds.), Handbook of positive psychology (pp. 277-287). New York: Oxford. Maddux, J. E., Snyder, C. R., & Lopez, S. J. (2004). Toward a positive clinical psychology: Deconstructing the illness ideology and constructing an ideology of human strengths and potential. In P. A. Linley & S. Joseph (Eds.), Positive psychology in practice (pp. 320-334). Hoboken, NJ: Wiley. Marek, L.L., Sandifer, D.M., Beach, A., Coward, R.L., & Protinsky, H.O. (1994). Supervision without the problem: A model of solution-focused supervision. Journal of Family Psychotherapy, 5, 57-64. Masten, A.S., & Reed, M.J. (2002). Resilience in development. In C. R. Snyder & S. J. Lopez (Eds.), Handbook of positive psychology (pp. 74-88). New York: Oxford. 191 McCarthy, P., Kulakowski, D., & Kenfield, J.A. (1994). Clinical supervision practices of licensed psychologists. Professional Psychology: Research and Practice, 25, 177- 181. Mead, G.H. (1934). Mind, self, and society: From the standpoint of a social behaviorist. Oxford, England: University of Chicago Press. Muse-Burke, J.L., Ladany, N., & Deck, M.D. (2001). The supervisory relationship. In L.J. Bradley & N. Ladany (Eds.), Counselor supervision: Principles, process, and practice 3 rd edition (pp. 28-57). Philadelphia, PA: Taylor & Francis. Myers, D.G. (1998). Psychology 5 th edition. New York: Worth Publishers. Pargament, K.I., & Mahoney, A. (2002). Spirituality: Discovering and conserving the sacred. In C. R. Snyder & S. J. Lopez (Eds.), Handbook of positive psychology (pp. 646-662). New York: Oxford. Patterson, C.H. (1997). Client-centered supervision. In C.E. Watkins, Jr. (Ed.). Handbook of psychotherapy supervision (pp. 134-146). New York: John Wiley. Patton, M.J., & Kivlighan, D.M. (1997). Relevance of the supervisory alliance to the counseling alliance and to treatment adherence in counselor training. Journal of Counseling Psychology, 44, 108-115. Peterson, C. & Seligman, M. E. P. (2004). Character strengths and virtues: A handbook and classification. Washington, DC: APA. Pierce, W.D., & Epling, W.F. (1999). Behavior analysis and learning 2 nd edition. Upper Saddle River, NJ: Prentice Hall. Porter, N., & Vasquez, M. (1997). Covision: Feminist supervision, process, and collaboration. In J. Worell & N.G. Johnson (Eds.). Shaping the future of feminist psychology: Education, research, and practice. American Psychological Association: Washington, DC. Presbury, J., Echterling, L.G., & McKee, J.E. (1999). Supervision for inner-vision: solution-focused strategies. Counselor Education and Supervision, 39, 146-155. Ramos-Sanchez, L., Esnil, E., Goodwin, A., Riggs, S., Touster, L.O., Wright, L.K., Ratanasiripong, R., & Rodolfa, E. (2002). Negative supervisory events: Effects on supervision satisfaction and supervisory alliance. Professional Psychology: Research and Practice, 33, 197-202. Rita, E.S. (1998). Solution-focused supervision. Clinical Supervisor, 17, 127-139. Robiner, W.N., Fuhrman, M., & Ristvedt, S. (1993). Evaluation difficulties in 192 supervising psychology interns. Clinical Psychologist, 46, 3-13. Rogers, C. R. (1963). The concept of the fully functioning person. Psychotherapy: Theory, Research, and Practice, 1, 17-26. Rosenbaum, M., & Ronen, T. (1998). Clinical supervision from the standpoint of cognitive-behavior therapy. Psychotherapy, 35, 220-230. Sandron, L. (1970). Two opposing models for a self-fulfilling prophecy. The Counseling Psychologist, 6, 82-87. Schwebel, M., & Coster, J. (1998). Well-functioning in professional psychologists: As program heads see it. Professional Psychology: Research and Practice, 29, 284- 292. Selekman, M.D., & Todd, T.C. (1995). Co-creating a context for change in the supervisory system: The solution-focused supervision model. Journal of Systemic Therapies, 14, 21-33. Seligman, M. E. P. (2002). Positive psychology, positive prevention, and positive therapy. In C.R. Snyder & S. J. Lopez (Eds.), Handbook of positive psychology (pp. 3?12). New York: Oxford. Seligman, M., E. P., & Csikszentmihalyi, M. (2000). Positive psychology: An introduction. American Psychologist, 19, 5-14. Seligman, M. E. P. & Peterson, C. (2003). Positive clinical psychology. In L. G. Aspinwall & U. M. Staudinger (Eds.), A psychology of human strengths: Fundamental questions and future research directions for a positive psychology (pp. 305?318). Washington, DC, APA. Sheldon, K. M., & King, L. (2001). Why positive psychology is necessary. American Psychologist, 56, 216-217. Snyder, C. R., Harris, C., Anderson, J. R., Holleran, S. A., Irving, L. M., et al. (1991). The will and the ways: Development and validation of an individual-differences measure of hope. Journal of Personality and Social Psychology, 60, 570-585. Snyder, C.R., & Lopez, S.J. (2002). Handbook of positive psychology. New York: Oxford University Press. Snyder, C. R., Lopez, S. J., Edwards, L. M., Pedrotti, J. T., Prosser, E. C., Walton, S. L., et al. (2003). Measuring and labeling the positive and the negative. In S. J. Lopez & C. R. Snyder (Eds.), Positive psychology assessment: A handbook of models and measures (pp. 21?40). Washington, DC: APA. 193 Snyder, C.R., Rand, K.L., & Sigmon, D.R. (2002). Hope theory: A member of the positive psychology family. In C.R. Snyder & S.J. Lopez (Eds.), Handbook of positive psychology (pp. 257-276). New York: Oxford University Press. Stoltenberg, C.D., McNeill, B.W., & Crethar, H.C. (1995). Persuasion and development in counselor supervision. The Counseling Psychologist, 23, 633-648. Stoltenberg, C.D., McNeill, B.W., & Delworth, U. (1998). IDM: An integrated developmental model for supervising counselors and therapists. San Francisco: Jossey-Bass. Super, D.E. (1955). Transition from vocational guidance to counseling psychology. Journal of Counseling Psychology, 2, 3-9. Szymanski, D.M. (2003). The feminist supervision scale : A rational/theoretical approach. Psychology of Women Quarterly, 27, 221-232. Talen, M.R. & Schindler, N. (1993). Goal-directed supervision plans : A model for trainee supervision and evaluation. Clinical Supervisor, 11, 77-88. Thomas, F.N. (1996). Solution-focused supervision : The coaxing of expertise. In S.D. Miller, M.A. Hubble, & B.L. Duncan (Eds.). Handbook of solution-focused therapy (pp. 128-151). San Francisco: Jossey-Bass. Timm, T.M., & Blow, A.J. (1999). Self-of-the-therapist work : A balance between removing restraints and identifying resources. Contemporary Family Therapy, 21, 331-351. Tinsley, H. E. & Tinsley, D. J. (1987). Uses of factor analysis in counseling psychology research. Journal of Counseling Psychology, 34, 414-424. Triantafillou, N. (1997). A solution-focused approach to mental health supervision. Journal of Systemic Therapies, 16, 305-328. Vasquez, M.J. (1992). Psychologist as clinical supervisor: Promoting ethical practice. Professional Psychology: Research and Practice, 23, 196-202. Vygotsky, L.S. (1978). Mind in society: The development of higher mental processes. Cambridge, MA: Harvard University Press. Walsh, W. B. (Ed.). (2003). Counseling psychology and optimal human functioning. Mahwah, NJ: Erlbaum. Walsh, W.B., & Betz, N.E. (2001). Tests and assessment, 4 th edition. Upper Saddle River, NJ: Prentice-Hall. 194 Ward, L.G., Friedlander, M.L., Schoen, L.G., & Klein, J.G. (1985). Strategic self- presentation in supervision. Journal of Counseling Psychology, 32, 111-118. Weisinger, H., & Lobsenz, N.M. (1981). Nobody?s perfect: How to give criticism and get results. Los Angeles: Stratford. Wetchler, J.L. (1990). Solution-focused supervision. Family Therapy, XVII, 129-138. Worthen, V., & McNeill, B.W. (1996). A phenomenological investigation of ?good? supervision events. Journal of Counseling Psychology, 43, 25-34. Wright, B.A. & Lopez, S.J. (2002). Widening the diagnostic focus: a case for including human strengths and environmental resources. In C.R. Snyder & S.J. Lopez (Eds.), Handbook of positive psychology (pp. 26-44). New York: Oxford University Press. Wulf, J., & Nelson, M.L. (2000). Experienced psychologists? recollections of internship supervision and its contribution to their development. Clinical Supervisor, 19, 123-145. Yun, G.W., & Trumbo, C.W. (2000). Comparative Response to a Survey Executed by Post, E mail, & Web Form, Journal of Computer-Mediated Communication, 6, no pagination specified.