On becoming a psychotherapist


Material Information

On becoming a psychotherapist an experience of good supervision
Physical Description:
vii, 150 leaves : ill. ; 29 cm.
Buehner, Timothy M
Publication Date:


Subjects / Keywords:
Psychology thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1997.
Includes bibliographical references (leaves 138-149).
General Note:
General Note:
Statement of Responsibility:
by Timothy Mark Buehner.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 028624630
oclc - 38863619
System ID:

Table of Contents
    Title Page
        Page i
        Page ii
    Table of Contents
        Page iii
        Page iv
    List of Tables
        Page v
        Page vi
        Page vii
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
    Review of literature
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
    Appendix A. Sample protocol with meaning units
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
    Appendix B. Fundamental structure of positive supervision
        Page 135
        Page 136
        Page 137
    List of references
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
    Biographical sketch
        Page 150
        Page 151
Full Text







I would like to acknowledge my committee chairman, academic advisor, and

friend, Franz Epting, Ph.D. His introduction of phenomenological research sparked my interest and his support allowed my enthusiasm to grow. I would also like to acknowledge my other committee members, Drs. Jim Archer, Wayne Francis, David Suchman, and Robert Ziller, for their interest and support of qualitative research in psychology. Finally, I would like to acknowledge a supervisor, Enrique Casero, Ph.D., who re-ignited my passion for qualitative research and who hinted that it is okay to be regarded as a "phenomenologist."

I would also like to acknowledge my wife, Randi, and my children Anja, Linnea, Nathaniel, and Tyra. Without their perpetual love, support, and faith in me, this project and this manuscript would never have become complete.


A CKN O W LED GEM EN TS ................................................................................................ ii

LIST OF TA BLES ............................................................................................................... v

ABSTRA CT ....................................................................................................................... vi

IN TRODU CTION ............................................................................................................... I
Clinical Supervision of Psychotherapy .......................................................................... 2
Research M ethodology in Counseling Psychology ....................................................... 5
Phenom enological Research .......................................................................................... 9
Sum m ary ...................................................................................................................... 11

REV IEW OF LITERATU RE ............................................................................................ 13
Im portance of Supervision ........................................................................................... 13
Conflicts in Supervision Theory .................................................................................. 15
M odels and Theories of Supervision ........................................................................... 18
W hat is Good Supervision? ......................................................................................... 52
Sum m ary ...................................................................................................................... 58

M ETH OD .......................................................................................................................... 59
Subjects ........................................................................................................................ 61
Procedure ..................................................................................................................... 62

RESU LTS .......................................................................................................................... 68
Essential D escription of Good Supervision ................................................................. 76
Structural D escription of Good Supervision ................................................................ 79

DISCU SSION .................................................................................................................... 81
Personal Reflections ..................................................................................................... 82
Orientation of the Subject ............................................................................................ 84
Qualities of Supervision ............................................................................................... 86
H ow D oes This Relate to D evelopm ental M odels? ..................................................... 93
Other Phenom enological Studies of Supervision ........................................................ 96
Training in Supervision .............................................................................................. 105
Conclusions and Recom m endations for Future Research .......................................... 106
Com m ents Regarding the M ethodology .................................................................... Ill


APPENDIX A SAMPLE PROTOCOL WITH MEANING UNITS ............................. 114

APPENDIX B FUNDAMENTAL STRUCTURE OF POSITIVE SUPERVISION ....... 135 LIST O F RE FER EN C E S ................................................................................................. 138

BIO G RA PH ICA L SK ETCH ........................................................................................... 150




1: Characteristics of Beginning and Advanced Supervisors ................................. 34

2: Empirical support of supervisee developmental models .................................. 41

3: Empirical support of supervisor developmental models................................... 43

4: Bernard's 3x3 matrix of supervisor roles and functions................................... 49

5: Littrell, Lee-Borden, and Lorenz four-stage unidimensional. model of supervision ...50 6: Three additional supervisor roles added by Ellis, Dell, and Good ...................... 50

7: Meaning units derived from the five subject protocols.................................... 69

8: Four phases of a Good Supervision Experience for Advanced Supervisees............ 97


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy


Timothy Mark Buehner

December, 1997

Chairman: Franz R. Eating, Ph.D.
Major Department: Psychology

Supervision in the training of professional psychologists is recognized as a

necessary and valuable practice. Despite its clinical relevance, supervision has received relatively little attention in the forin of empirical research. This study uses a qualitative research method, the phenomenological method as developed by Amedeo Giorgi, to explore good clinical supervision as experienced by predoctoral psychology interns. Five interns were interviewed to elicit narratives of their lived experiences of good supervision. Each interview began with the same question, then proceeded to clarify the experience as reported by each subject. No questions were asked to support nor deny preexisting hypotheses regarding the phenomenon. An analysis and eidetic reduction were then performed on the interview transcripts to reveal the essential nature of the lived experience of good supervision. The result of the analysis is reported as the essential description and the result of the reduction is reported as the structural description of good


supervision. This study found that supervisees experience supervision to be a direct extension of the supervisor. Experiences of good supervision were attributed directly to how the supervisor interacted with the supervisees rather than to learned skills or behaviors. Supervisees were able to develop and invest in a supervisory relationship in which they felt respected and supported. Supervisees valued supervision by competent and invested supervisors. When supervision was valued, supervisees performed harder and were more likely to actively participate in the supervision process. Supervisees experienced good supervision as developmental in nature-building upon their existing strengths rather than being corrective or critical. Other necessary components of a good supervisory relationship include, trust, confidence, and likability of the supervisor. The relevance of these results to existing empirical research in supervision is discussed.



Since Freud began delving into the unconscious over one hundred years ago, psychotherapy has become an important treatment modality in the United States and throughout the world. Psychotherapy has been referred to as an art, a craft, and a science. Regardless of how it is classified, psychotherapy is, in essence, a relationship between two (or more) persons striving towards the mutual goal of the restoration of a person's mental health and/or well-being. How one of those persons becomes qualified to assist the other(s) in the pursuit of health is the focus of this paper. We will examine how a person becomes a psychotherapist through good clinical supervision by conducting a phenomenological investigation of good super-vision.

The training of professional psychologists is an area that has stimulated discussion and research in the professional literature for many years. In accordance with the scientist-practitioner model advocated by the American Psychological Association, the psychologist should be trained as a researcher as well as a clinician. The training of the scientist-practitioner is often reported to be composed of four primary elements: academic course work, supervised practical experience, personal psychotherapy, and applied research experience. While the requirement of engaging in personal psychotherapy varies greatly across training programs, the requirements of academic course work, supervised clinical experience, and applied research experience are the essence of all training programs accredited by the American Psychological Association



(APA). Research into how psychology trains its professionals has become an area of increased interest during the past 20 years.

The purpose of this research project is to examine two areas that are essential in the training of professional psychologists, clinical supervision and research methodology. The process of clinical supervision will be empirically studied from the perspective of the psychologist trainee. The examination of research methodology will be experiential in nature-this project intends to utilize a human science methodology, the phenomenological method, to perform the empirical analysis. This research project is an effort to not only contribute to the content of empirically-based psychological knowledge, but to also support the profession's call to embrace alternative methodologies in the collection and analysis of such knowledge.

Clinical SiWervision of Psychotherapy

Supervision is considered by educators, trainers, and professional regulatory

agencies to be a critical component of psychotherapy training (Holloway and Neufeldt, 1995). The importance of supervised training is underscored by the pre- and post-doctoral practice standards of the American Psychological Association and licensing requirements of most states (American Psychological Association, 1986). While supervision of clinical training has long been considered a significant aspect to the training of professional psychologists, it has generated an inadequate body of research compared to research of psychotherapeutic issues in general. Prefacing a special issue of The Counseling Psychologist devoted to supervision, Bartlett and associates noted the inadequacy of


research literature on counseling supervision (Bartlett, Goodyear, and Bradley, 1983). Over a decade later, Holloway and Neufeldt (1995) noted that the body of research on supervision was mediocre considering its importance to the clinical practice of psychology. Robertson (1995) suggests this lack of research is due to the fact that super-vision generates relatively little controversy. He adds that research in the area has increased recently due in part to the increasing pressure of accountability.

Research of the supervision of trainees in clinical and counseling psychology is becoming an area of growing interest. During the past 15 years, there has been an increase in the theoretical discussion of supervision although empirical research testing those theories has been limited (Swanson and O'Saben, 1993; Worthington, 1987). The published empirical research has focused on skills training and supervisor technique (i.e., Russell, Crimmings, and Lent, 1984; Marikas, Russell, and Dell, 1985) with an increasing emphasis on testing developmental models of supervision. The testing of these models builds on the work begun by Stoltenberg (1981) and Littrell, Lee-Borden, and Lorenz (1979). Such developmental models have guided the research and, presumably, the training of counseling psychologists.

As supervision is typically provided by individuals trained within a particular

orientation of psychotherapy, the supervision is often provided following the conceptual framework of that theory, whether it is psychodynamic, cognitive-behavioral, humanistic, or family systems in nature. There are, however, cross-theoretical models of supervision, such as the developmental models which are not bound to any particular theory of psychotherapy and may be employed by supervisors from a variety of theoretical


orientations. The field of counseling psychology appears to have embraced the developmentally based models of supervision and the research has focused on the testing and enhancement of those models (McNeill, Stoltenberg, and Romans, 1992; Stoltenberg and Delworth, 1988; Wiley and Ray, 1986).

Although empirical research has focused on developmental models of supervision, a review of the literature suggests a growing amount of research on the supervisory relationship and process. Recent papers written on the supervisory relationship include discussions of power and control (Salvendy, 1993; Weisgerber, 1992), stagnation in supervision (Yerushalmi, 1993), resistance (Glickauf-Hughes, 1994), mirroring (Manasis and Traub-Werner, 1994), separation-individuation (Watkins, 1992), and parallel process (Adelson, 1994). Various aspects of the supervisor have been discussed, including the supervisor as a mentor (Vallery, 1992), dimensions of the role (Duci, 1992), the supervisor's devotion (Yerushalmi, 1992), and the experience of supervising (Paskiewicz, 1993; Collett, 1995; Clarkson and Aviram, 1995). The role and perspective of the supervisee has also begun generating interest in the professional literature, including the experience of shame in the supervisory experience (Talbot, 1995), the struggle of perfectionism (Arkowitz, 1994), role ambiguity (Olk and Friedlander, 1992), and trainee anxiety (Udis, 1991). While the scope of this list is impressive, it should be noted that much of this literature is accounted for by Dissertation Abstracts International and specialized journals. Furthermore, many of these studies represent research endeavors that apply non-traditional research methodologies. Perhaps the previously mentioned lack of empirical research in counseling supervision is due not to a lack of interest in the


multifaceted process of supervision, but rather to the inadequacy of traditional, i.e., natural science, research methods in handling these meaningful, but experiential, aspects of the supervisory process.

Research Mgthodology in Counseling-Psychology

When psychology was first taught at institutions of higher learning, the teaching professors held appoints in departments of philosophy. As psychology became more applied in focus, however, psychologists strove to become more science-like-adopting research methodologies utilized by the "natural" sciences-biology, physics, chemistryin an effort to become more efficacious in the pursuit of predicting human behavior. While this tact has been effective in promoting the application of psychological principles in the treatment of mental illness, it has also resulted in a departure from studying other areas of psychological interest that are not conducive to being studied by the experimental method. As the strength of psychology's rigor in natural science methodologies has increased, so too has the interest in alternative forms of research methodology. Perhaps psychology is reaching an age where it can critically examine its development, identify its strengths and weaknesses, and return, in part, to an examination of issues originally tossed aside in hopes of becoming accepted by the scientific community. The specialty area of Counseling Psychology is no exception to this developmental progress in psychology. The advancement of counseling psychology's scientific rigor has been hailed, and utilization of natural science methodologies is the


accepted norm. Yet, there is a growing voice calling for methodological diversity that has increased in volume and influence during the past two decades.

In 1979, Gelso called for increased methodological diversity, a call that has been increasingly answered throughout the 1980s and into the 1990s. The three major American publications of counseling psychology, the Journal of Counseling Psychology, The Counseling Psychologist, and the Journal of Counseling and Development, have given much attention to the advancement of methodological diversity (Hoshmand, 1989; Gelso, Betz, Friedlander, Helms, Hill, Patton, Super, and Wampold, 1988; Heppner, Gelso, and Dolliver, 1987; Howard, 1984,1985; Fretz, 1986; Gelso, 1979; Gelso and Johnson, 1982; Patton, 1984; Polkinghorne, 1984). A special subsection of the Journal of Counseling Psychology (Volume 31, 1984) was devoted to alternative methodologies and included 3 major papers that examined the philosophical foundations of such research (Howard, 1984; Patton, 1984; Polkinghorne, 1984). In 1988, the Research Group at the Third National Conference for Counseling Psychology reported on prospects and recommendations for research in counseling psychology (see Gelso et al., 1988). Among the fifteen recommendations offered regarding research, three directly supported the trend toward alternative methodologies. The Counseling Psychologist, in 1989, devoted an issue to alternative methodologies, and in 1994, the Journal of Counseling Psychology included a special section on qualitative research in counseling process and outcome. That section in Journal of Counseling Psychology comprised of eight qualitative studies (Friedlander, Heatherington, Johnson, and Skowron, 1994; Elliot, Shapiro, Firth-Cozens, Stiles, Hardy, Llewelyn, and Margison, 1994: Cummings, Hallberg, and Slemon, 1994;


Rhodes, Hill, Thompson, and Elliott, 1994; Thompson and Jenal, 1994; Rennie, D.L., 1994; Frontman and Kunkel, 1994; Watson and Rennie, 1994) and a reaction by Polkinghorne (1994), in which he noted that this issue was the first significant compilation of qualitative studies produced, despite the call for increased methodological diversity for over a decade. Polkinghome (1984) also suggests that counseling psychology should not only support the development of alternative methodologies, but should assume a leadership role in training researchers in their use.

Counseling psychology's growing interest in alternative methodologies has not been a spontaneous plunge into diversity for diversity's sake. There has been a growing concern that heavy reliance upon natural science methodology has resulted in inadequacies in addressing the full realm of psychological phenomena of interest to counseling psychologists. Within the movement of methodological diversity has been a commitment to go beyond diversity, to embrace alternative methodologies as a way to compensate for the inadequacies of natural science methodologies. It has been argued that the over reliance on natural science methods has failed counseling psychology research in two fundamental ways.

The first problem is related to the types of questions, in terms of meaning and

language, asked by natural science researchers versus the types of questions pondered by clinicians. Human scientists argue that questions regarding the depth and complexity of human experience cannot be adequately addressed within the limitations of natural science methodologies. This is an important distinction, they argue, as such questions are most often the primary concerns of the practitioner. The human scientist attempts to


understand the subject from the perspective of his or her own subjective experience, emphasizing exploration of the whole phenomenon rather than discrete parts. Human science methodologies are offered as a source to address these distinctly different research questions of interest to clinicians.

The second problem, according to the human science perspective, is that the

assumptions of the human scientist regarding human nature are closer to the assumptions of the clinician than are the assumptions of the natural scientist. Specifically, it is argued that most clinicians view their clients as being self-determining. Natural science methodologies, however, conceptualize the individual as an object upon which environmental variables are manipulated which then determine outcome. Human science methodologies, on the other hand, treat the subject from a perspective more familiar to the practitioner, incorporating the assumption of free-will into the data collection and analysis.

Human science methods have developed through the investigations of various researchers who have taken it upon themselves to seek alternative methodologies to the natural science approach in understanding humanistic theoretical positions. Polkinghorne (1982) argues that the development of new methodologies is a result of the humanistic psychologist's commitment to "understanding that human beings exist within an experience of meaning and retain the possibility of acting with self-determined purpose" (p. 48). While humanistic theory does not propose that human action is completely independent of environmental or genetic origins, it does allow that persons can choose to act in ways not predetermined. Polkinghorne describes these alternative methodologies


as structural research designs--those which seek underlying themes or patterns of the human experience. While these structural designs are similar to each other in their focus on human experience, they vary in terms specific research questions addressed, types of data required, and specific analytical procedures employed. The scientific method which will be utilized in this study is the phenomenological method, as conceptualized and developed by Giorgi (1989).

Phenomenological Research

Phenomenology, according to Giorgi, is the study of phenomena as experienced. For the phenomenologist, phenomenon refers to what is given in experience, to be understood in direct relationship with awareness (Giorgi, 1976). Therefore the phenomenological researcher is interested in discovering the structure of the phenomena (experience) as experienced. The phenomenologist refers to the context of the lived world as a point of departure and searches for the meaning of the lived experience. Phenomenology goes beyond the collection and analysis of facts to determine the meanings of those facts-facts are important insofar as they have meaning. The focus of human consciousness, going beyond the quantifiable measurement of the experience, phenomenological research can provides a deeper understanding of human experience than natural science methods.

Phenomenology is a process of discovery and it is essential that the phenomenon be examined prior to drawing on pre-existing knowledge of the phenomenon. As it is unlikely that a researcher can ever study a phenomenon without possessing preconceived


beliefs about it, the researcher must consciously separate and disregard all known experience of the phenomenon of study. This partitioning off, of prior knowledge, is known as bracketing, and it plays a vital role in the phenomenological research process. Bracketing was first described as epochi by Husserl, in his explanation of psychology in terms of pure phenomenology (Husserl, 1913/1983). Using a contemporary definition, bracketing is the "refraining from accepting the actually existing, real world in the customary sense of being" (Kersten, 1989, p. 23). It is the removal of preconceived ideas and beliefs about the phenomenon of study from the experimenter's conscious thought. Thus bracketing is, for the phenomenologist, experimental control-the effort at limiting the amount of extraneous variables and biases into the scientific method and analysis.

A variety of phenomena have been studied utilizing phenomenological

methodologies, and the quantity of published studies has grown steadily, particularly within recent years. Some of the phenomena studied are: being a supervisor (Clarkson and Aviram, 1995), shame (Ablamowicz, 1992), moral sense (Giorgi, 1992), mutual gaze (Angus, Osborne, and Koziey, 1991), maternal mourning (Brice, 1991), shyness (Guglietti-Kelly and Westcott, 1990), anger (Buehner, 1993; Stevick, 1971), aging (Wondolowski and Davis, 1988), decision and choice (Karlsson, 1988), friendship between women (Becker, 1987), courage in chronically ill adolescents (Haase, 1987), the formation of clinical impressions (Churchill, 1984), the experience of failing to learn (Deegan, 1981), the premenstruum (Montgomery, 1982), tragedy (Carrere, 1989), forgetting (Kral, 1989), intimacy (Register and Henley, 1992), and glaucoma (Kugelmann and Bensinger, 1983).


The limited production of supervision research may be due, in part, to the

difficulty in applying natural science methodologies to the complex and dynamic nature of clinical supervision. While behaviors can be monitored and self-report surveys can reflect perceived effectiveness of supervisory styles and interventions, they cannot fully describe the process that promotes the evolution of novices into competent therapists. At a time when counseling psychology is increasing its utilization of human science methodologies to better understand the psychotherapeutic process and to evaluate its efficacy, it is appropriate to utilize such methods to examine the training of its psychotherapists.

This study intends to examine the process of supervision using the

phenomenological method as developed by Giorgi. In particular, this study will attempt to understand the experience of being supervised from the perspective of the trainee. By performing a phenomenological analysis on interview data collected from psychologistsin-training, it is anticipated that a structure of meaning can be discovered which will identify the essences of the supervisory experience for trainees. This structure will then be used as a guidepost to understand the meaning of being supervised relative to existing research of supervision.

There are many general aspects of supervision, including, but not limited to,

didactic teaching, behavioral modeling, mentoring, and self-examination. At times, these events may be experienced as anxiety-provoking, comforting, boring, frustrating, angerprovoking, enlightening, educating, threatening, and so on. One's valuation of the


supervisory experience, however, is more likely to be broken down into less defined categories, simply as helpful or hurtful, positive or negative, good or bad. Given these gross, yet clinically valuable categorizations of the supervisory experience, it seems a study of experiences that are most indicative of an ideal supervisory process would provide the most beneficial contribution to the professional literature. This study, then, will be a study of good supervision, where "good" is defined by the subject, the supervised.


This chapter will discuss several aspects of supervision. After discussing the importance of supervision, there will be a discussion of conflicts regarding supervision theory that have been identified in the literature. There will then be a section on models and theories of supervision which will include developmental models, supervision as a relationship, and the dimensions of super-vision. The discussion of developmental models will include models of supervised development, supervisor development, and a review of the empirical support of developmental models. Finally, there will be a review of what has been identified as "good" supervision.

Importance of Supervisio

The importance of psychotherapy supervision in the training of psychotherapists cannot be understated. It is a fundamental training requirement for accreditation by the American Psychological Association, the American Psychiatric Association, Social Workers Association, as well as a licensing requirement by state regulatory agencies. It is a key element in the process by which psychotherapy is taught and learned (Ekstein and Wallerstein, 1972; Jacobs, David, and Meyer, 1995) regardless of specialty area, level of training, theoretical orientation, or work setting (Watkins, 1995a). Despite this apparent importance, research on the effectiveness of supervision has been limited.



Research into the supervisory process has increased significantly during the past decade. Research of supervision is methodologically diverse, reflecting, perhaps the diverse nature of theoretical beliefs regarding supervision. Supervision is a complex process of teaching, modeling, and preparing the supervised for the professional role. Supervision affects not only behaviors designated as "treatment interventions," but attitudes, beliefs toward the practice of psychotherapy and mental health in general. Supervision provides the framework for an individual as one becomes a psychotherapist. Experiences of supervision will likely affect the supervised's professional demeanor for a significant portion of the supervisee's professional life. In considering the effectiveness of supervision, one must consider what is most important, outcomes of supervised change (and how that may be measured) or outcomes to the psychotherapy being performed by the supervised.

While the importance of supervision is typically accepted as self-evident by those who practice, teach, and regulate psychotherapy, its efficacy toward treatment outcomes is largely unproved. In their introduction to a special section on training in the Journal of Consulting and Clinical Psychology, Beutler and Kendall (1995) reported that professional training may enhance clinical efficacy, but there is insufficient knowledge regarding the specifies aspects of professional training and professional experience which most effectively contribute to such a relationship (p. 180). In their paper within the same special section, Holloway and Neufeldt (1995) concluded that there was little evidence of a direct connection between supervision and therapist efficacy, though it is likely that supervision increases the proficiency with which one uses complex therapy procedures,


enhances knowledge and role definitions, and increases one's ability to facilitate internal therapeutic processes. Holloway and Neufeldt (1995) concluded that the effects of specific supervisory interventions to therapist and client change in behavior remained largely unknown. Rather than accept the whole of supervision as contributing to improved therapeutic technique, Holloway and Neufeldt argue for the identification of specific behaviors and interventions leading to specific changes. Subsequently, Holloway and Carroll (1996) argue that research on supervision must move from a theoretical approach to the pragmatic application of behaviors. Supervision research must, according to Holloway and Carroll (1996) advance from understanding how supervision is performed to identifying how the effectiveness of supervision may be improved

Given the nature of supervision, however, as a complex process requiring an

interpersonal relationship of at least three individuals (supervisor, supervised, and client) engaged in a process of learning, modeling, and relating, it is possible that specific behaviors to specific outcomes may never be proven. Yet given that possibility, it is unlikely that anyone engaged in the practice of psychotherapy would advocate elimination of training through supervision. Supervision is the mechanism wherein individuals not only learn how to do psychotherapy, but how to be a psychotherapist as well.

CQnflicts in SILpervision heory

Despite one hundred years of psychotherapy, there has not developed one

accepted manner of training psychotherapists. The development of supervision theory is


in its infancy when compared to the theoretical development of psychotherapy. Due, in part, to its immature status, there are several issues that must be resolved as theories of supervision develop. One conflict in approaching supervision is the question whether supervision requires a theoretical understanding of its own or whether is it can be sufficiently understood by applying theories adapted from psychotherapy (Hess, 1 987a). Historically, particularly when psychodynamic theories were at their strongest, training has duplicated the processes of the theory and psychotherapy the student was being trained to implement. This traditional course of action may explain, in part, the historical paucity of separate supervision theory. As the theoretical field of psychotherapy has become more eclectic, there has also been an increase in the development of supervision theory unlinked to a specific psychotherapeutic approach. Accepting the need for a separate supervision theory, other conflicts remain.

A second conflict in supervision theory exists in how supervisors should perform supervision. This conflict is whether supervision should be proactive or reactive (Worthington, 1987). In proactive supervision, the supervisor effectively leads the supervisee; in reactive supervision, the supervisor still facilitates growth, but allows the supervisee to determine the course. Some supervision follows an agenda set by the supervisor. Each session has defined goals and the task of the supervisor is to assist the supervisee to attain those goals. In this form of supervision, the trainee is lead by the supervisor toward professional development. Other supervision is lead by the supervisee. Goals are still identified and tracked, but the supervisee is more instrumental in defining those goals as well as the manner they should be attained. The proactive-reactive conflict


has been resolved in large part in the developmental models. Almost uniformly these models advocate a more proactive stance by the supervisor during the supervised's early development, and a more reactive stance during the later stages of development.

A third conflict identified in the literature is the question of whether the

supervised should learn the theory of the supervisor or whether the supervisor should adapt supervision to the theoretical stance of the supervised (Worthington, 1987). In the early years of psychotherapy, this was not much of an issue as supervision was sought out to learn the perspective and methods of a particular supervisor. In the contemporary days of eclecticism, as well as a broad range of theories to choose from, this has become more of an issue. While specialty training stills requires the trainee to learn the methods of the supervisor, the generalist approach used in most training programs must account for theoretical differences between supervised and supervisor. If a supervised wishes to learn from the experience of a given supervisor without necessarily adopting the theoretical orientation, there is a question whether it would be effective for that supervisor to use counseling based super-vision for that supervisor. On the other hand, to expect a super-visor to effectively adopt the theoretical perspective of each trainee does not seem to be reasonable. Again, the developmental models step in to provide a solution. The developmental models are, for the most part, theoretical regarding counseling theory. They can be incorporated into a wide range of supervision situations regardless of the theoretical underpinnings of supervisor or supervised.

Another question of supervision involves the amount of counseling that should

take place within the supervision environment. From one perspective some writers argue


that it is necessary for supervisors to help supervises identify defenses, overcome resistances, and facilitate personal growth. On the hand, other authors contend that those activities are best saved for another environment. It is generally agreed, however, that extensive psychotherapy should not be a part of the supervision process.

Models and Theories of S=ervisio

In reviewing the current literature of supervision, few broad categories of supervision theory and research can be identified. Theories of supervision can be intuitively classified into three categories: 1) developmental models, 2) supervision which parallels a specific counseling theory, and 3) those that emphasize a particular aspect, or dimension, supervision. As the interest of this paper is the unique process of supervision, this review of theories will exclude those which simply apply counseling principles. On the other hand, the importance of the supervisory relationship is typically underscored by most supervision theorists, and a unique theory regarding that relationship will be presented. There will also be a review of the limited research into the dimensions of supervision. As developmental models have received the most attention in the past twenty years, this chapter will begin with a review of supervised and supervisor development.

Developmental Models

The conceptualization of clinical supervision has become increasingly dominated by thinking in developmental terms. In the words of Holloway (19 8 7), the developmental thrust has constituted "the zeitgeist of supervision thinking and research" (p. 209). While


Holloway made her observation ten years ago, the statement remains as true today as it was then (Watkins, 1995a). According to Watkins, these models have proven appealing over the course of the years for three fundamental reasons: 1) they are meta-theoretical; 2) they have direct practice and training implications, and 3) they provide a framework for tracking progress over time (Watkins, 1995a).

Developmental models of supervision conceptualize the supervised and, in some models, the supervisor, as proceeding through a series of developmental stages across time and, presumably, training. These models are often based upon the developmental theories of Piaget, Erikson, and others who have left their mark on human developmental theory. The person becoming a psychotherapist continues through a universal series of stages which predict certain emotional experiences as well as certain needs that require appropriate attention from the trainee's supervisor. The supervises progress from being less developed, i.e., naYve and unskilled, to becoming more developed, i.e., aware and competent. The supervisee's development is largely dependent upon how the supervisor interacts with the supervised at any given point during the trainee's development. It is assumed that the supervised will develop as a natural function of supervision. The supervisor may enhance and encourage growth by providing an appropriate environment and supervisory experience, or conversely, may stunt development by providing an inappropriate environment or engaging in inappropriate behaviors.

There have been many developmental models proposed in the literature. In an 1987 review, Worthington identified sixteen different developmental models in the literature, from 1953 to 1986. In an updated review, Watkins (1995) identified an


additional six models proposed between 1986 and 1994. Although these models may vary in the specifics of their language, number of stages, and developmental focus, both researchers concluded the models were more similar than dissimilar. Watkins noted the similarities began with the assumptions inherent in the models.

Watkins (1995) identified four key, identifiable assumptions shared by the developmental models:

1) therapists in training develop and grow, provided they are not exposed to a
pernicious training environment;
2) their development proceeds through a sequence of stages, from less to more
3) that during those stages, they struggle with various developmental issues and
concerns (e.g. competency, identity);
4) that supervisors would do well to consider the developmental level of their
supervisees and structure supervision accordingly. (p. 647).

Often using the theories of Erikson, Piaget, or other models of human development as their basis, developmental models of supervision and therapist development describe the transformation of an unskilled, nafve novice into a "master psychotherapist." As models of supervisee development have grown in popularity, parallel processes have been noted in supervisor development as well, spawning literature in support of supervisors developing in much the same manner as supervisees. The following sections will describe models of supervisee development, models of supervisor development, and empirical support for those models. Supervisee development

Hogan (1964) proposed four stages in the development of the psychotherapist. His model emphasized the supervisee' s increasing autonomy that culminated in a collaborative relationship with the supervisor. According to Hogan, it is the supervisor's


responsibility to provide a supervision environment that fosters the counselor's growth and development. The proper supervisory role is dependent upon the supervisee's stage of development. In the beginning, the supervisee has feelings of insecurity, lacks insight, and is imitative of the supervisor. The supervisee is dependent upon the supervisor to provide interpretation, support, and predictable outcomes to allow the supervisee to begin developing self-confidence as a psychotherapist. In the second stage, Hogan identifies a conflict centered upon dependency-autonomy issues in which the supervisee struggles between overconfidence and feeling overwhelmed. During this stage of development, Hogan recommends that the supervisor should provide support, exemplification, teaching, and clarification of the ambivalence experienced by the trainee. Hogan also recommends the initiation of personal psychotherapy for the trainee (with someone other than the supervisor). The trainee's successful resolution of the conflict in stage two leads to a stage marked by conditional dependency. In this third stage, the supervisee has increased professional confidence which includes greater insight into the supervisee' s neurotic and healthy motivations. During this stage, the supervisor should continue exemplification, begin sharing as a peer, and incorporate confrontation to ensure the trainee's continued development. Finally, the psychotherapist may now reach the fourth, and Hogan's final stage, that of Master Psychologist. This stage is identified when the psychotherapist has gained personal autonomy, insightfulness with motivation, and experiences a need for confrontation for continued and increased expertise. At this stage, the supervisee and supervisor should interact within a mutually sharing relationship, in consultation as well as confrontation. Hogan's stages are expressed in terms of


developing within the supervisory relationship-all development is experienced and progresses within the context of the supervisory relationship. Stoltenberg (1981) accepted Hogan's speculations about the development of counselors and further elaborated upon how a supervisor might provide the optimal environment for optimal growth by the supervisee (Worthington, 1987).

Stoltenberg's (198 1) Counselor Complexity Model focused on the supervisee's increasing levels of autonomy and indicated the supervisor's behavior should change according. The Counselor Complexity Model has four stages and is so named because Stoltenberg viewed the trainee as becoming more complex as training progressed and the trainee gained experience. Stoltenberg' s model speaks less to the varying levels of dependency within the supervisory relationship and more to the development of identity by the supervisee. Stoltenberg's model was also more elaborate regarding how a supervisor can promote the most growth. According to Stoltenberg, the novice psychotherapist begins by attempting to define boundaries between the "counselor" and the "person" (Worthington, 1987). During this initial stage, the supervisor should provide a structured environment, encouraging autonomy and risk taking within that structure. During the second stage, the supervisee' s professional identity begins to become defined and the trainee begins to experiment with different styles. The supervisee is no longer satisfied imitating the supervisor and will begin expressing disagreement. At this time the supervisor should provide new skills and advice as necessary, encouraging continued autonomy within a low pressure environment. The acquisition of new skills allows the trainee to tolerate a broader variety of clients,


becoming less dependent upon a single doctrine of therapy. In this third stage, Stoltenberg recommends an increased emphasis on sharing, allowing appropriate professional and personal confrontation. In the final stage of Stoltenberg's Counselor Complexity Model, the therapist is capable of practicing independently, integrating the standards of the profession within a personal value system. In 1987, Stoltenberg and Delworth elaborated further on this model, and provided a model of supervisor development within its framework.

In the revised Counselor Complexity Model, Stoltenberg and Delworth (1987) identified three levels of development. The third level, however, has an integrated and non-integrated component to comprise a four-stage model. In Level I, the supervisee is dependent and lacks self-awareness and in-depth understanding of the therapy process. The supervisee experiences anxiety and desires direction and guidance, searching for "the right way" to operate. The supervisor should, at this time, provide structure and direction, acting as a teacher and/or supporter. The dependency-autonomy conflict is a characteristic of Level II. At this time, the supervisee's heightened self-awareness is coupled with confusion as the trainee vacillates between being overly confident and overwhelmed with the increased insight. The supervisor should continue with empathy and encouragement. Teaching should continue, though it plays a smaller role than earlier in the training. As instruction decreases, the supervisor may perform more as a reference source. In Level III, the supervisee experience high, consistent motivation and acts in a more autonomous and confident manner. The psychotherapist possesses increased insight, empathy and sense of professional identity. At this time, the supervisor functions


more as a peer with increased sharing and mutuality. A decrease in structure is matched with an increase in appropriate confrontation. Stoltenberg and Delworth's final stage, Level III Integrated is marked by security, insightfulness, and healthy interdependence. The "master practitioner" is aware of professional and personal strengths and weaknesses and possesses integration of skill across many domains of practice. The supervisor's role is one of a collegial consultant. The difference between the non-integrated and integrated phases of Level III is related to the supervisee-supervisor relationship and the supervisee's professional identity. As the supervisee becomes more integrated, there is more insight regarding one's sense of self as a psychotherapist. The supervisee becomes simultaneously less dependent upon the supervisor for role definition as well as the process of supervision.

Blount (1982) proposed a four-stage model that borrows heavily upon Erikson's model of personality development. The four stages identified by Blount are: adequacy versus inadequacy, independence versus dependence, conditional dependency versus individuation, and professional integrity versus personal autonomy. Supervisor behaviors recommended by Blount are similar to those identified by Hogan (1964), Stoltenberg (1981), and Stoltenberg and Delworth (1987), beginning with a supportive "teaching" role and progressing to collegial consultation with greater supervisee autonomy.

Loganbill, Hardy, and Delworth (1982) identified three stages of counselor development. The stages they identified were stagnation, confusion, and awareness. Each of these stages is similar to those identified by Hogan (1964) and Stoltenberg (1981). The focus, however, of the model proposed by Loganbill and associates (1982) is


that an individual must resolve eight critical issues before becoming a master psychotherapist. The eight issues are: competence, emotional awareness, autonomy, theoretical identity, respect for individual differences, purpose and direction, personal motivation, and professional ethics. According to Loganbill and his colleagues (1982), each of these issues is resolved separately within the structure of the three stages. Therefore, competence develops following the progression from stagnation through confusion to awareness, as does emotional awareness, autonomy, and the other five critical issues. The trainee does not necessarily progress through each issue at the same pace. It is the responsibility of the supervisor to assess the trainee's development on each of the eight issues, then to promote growth to the next level of development of each issue.

Sansbury (1982) proposed a three-level developmental model based upon the specific training levels of professional training programs. The stages are referred to as prepracticum, practicum, and internship and are identified by the level and sophistication of skills acquired. In the first stage, basic listening skills and the assimilation of the counselor role are fostered by the supervisor who provides evaluative feedback, needs assessment, positive modeling, and support. In the second stage, the trainee broadens the therapeutic repertoire, improves conceptualization, refines one's personal theory, and establishes limits of responsibility. The supervisor, to promote this growth, analyzes cases and promotes the supervised's understanding through confrontation, role reversals, interpretation, and feedback. During internship, the trainee should be broadening and refining understanding of clients, examining one's issues, and learning self-reliance. An


intern's supervisor should confront the supervisee in discrepancies between talk and behavior, assist with personal issues, and encourage risk taking.

Friedlander, Dye, Costellos, and Kobos (1984) proposed a model which

emphasizes the trainees' development of their meaning-making and understanding of psychotherapy. The learning and acquisition of skills is dc-emphasized and Friedlander and associates (1984) emphasize the supervisor's responsibility to help the supervisee's meaning-making of psychotherapy as well as the supervisee 's role in it. The first stage is marked by ambiguity in the supervisee. The supervisor's role at this time, is to help the trainee cope with the demands of ambiguity with an emphasis on learning how to learn. During the second stage, the supervisee is able to recognize the limits of the therapeutic condition. During this period the supervisor should help the supervisee see the differences between theory and practice. The supervisor should also help the trainee accept mistakes as well as unanticipated client responses. The supervisor should also assist the supervisee to cope with guilt over failures. Friedlander and associate's (1984) third stage of development is marked by the trainee's discovery of psychotherapy as deep communication. As the communication becomes emphasized, the supervisor should shift the focus from techniques of therapy to the human relationship of therapy. In the final stage, eclecticism relative to the client's need becomes the emphasis as the supervisor helps the trainee develop a repertoire of interventions to apply in the appropriate setting according to the client's need based upon a careful assessment.

Friedman and Kaslow (19 86) identified six developmental stages that the

supervisee must pass through to become a "master" psychotherapist. The first stage is


marked with excitement and anticipatory anxiety as the trainee attempts to become acquainted with the training agency. The supervisor is to provide a secure holding environment, empathic to the trainee's anxieties and vulnerabilities. In the second stage, which begins when the supervisee is assigned a case, the supervisee becomes dependent upon, and identifies with, the supervisor. The supervisee's development at this point is dependent upon the supervisor's help in organizing and anticipating the processes of therapy. It is also important for the supervisor to carefully select cases to avoid major management issues. The third stage identified by Friedman and Kaslow is one of activity and continued dependency. This stage begins when the trainee realizes the client is taking therapy seriously. The trainee becomes more active in the psychotherapy process, assuming more professional responsibility. At this point, it is important that the supervisor acknowledges difficulties, limits criticism, and is affectively predictable. When the fourth developmental stage is reached, the supervisee feels exuberant and "takes charge." Identifying oneself as a psychotherapist, the trainee has an increased understanding of the relationship between theory and practice, is able to explore issues related to countertransference, and is more independent. As the supervisee "breaks away" it is important for the supervisor to relinquish control while remaining supportive. The fifth stage of identity and independence is referred to as "professional adolescence." The supervisee has a firmly established internalized frame of reference. The trainee is also well aware of the supervisee's own weaknesses as well as those of the supervisor. During this stage, the supervisee may devalue the supervisor. Despite the potential for devaluation, the supervisor continues to support the trainee's increasing autonomy and


being readily available to the trainee. The final stage identified by Friedman and Kaslow is described as calm and collegial. When fully developed, the psychotherapist has a firmly established sense of professional identity, is highly integrated, and experiences less intense self-doubt and affective variability. During this final stage, the supervisor should facilitate growth, challenge the supervisee's inconsistencies, and focus on countertransference.

Hess (1986, 1987) proposed a four-stage model of counselor development that he felt was representative of the development models proposed to date. During the Inception phase, the supervisee feels unanchored, insecure, and anxious. The supervisee is rigid, ambivalent, unaware of therapy's impact upon clients, and is dependent upon the supervisor. As the supervisee moves through the skill development phase, the trainee becomes more comfortable in the therapist role. The supervisee also has increased awareness regarding client dynamics and is more accepting of one's own weaknesses. As awareness and competence increase, the supervisee typically fluctuates between being over-confident and overwhelmed with one's new role and skills. The next phase, consolidation, is marked by an internal sense of mastery. Skills continue to develop and the professional identity becomes solidified. Hess' final phase, Mutuality, is marked by the trainee becoming fully autonomous and able to give and receive consultation. At this time the psychotherapist is also able to be creative and utilize intuition. The supervisory relationship is collegial and consultative.

The four-stage model proposed by Rodenhauser (1994) emphasizes the

developing psychotherapist's understanding of the therapeutic process and the role of the


self. The first stage, Restoration, is so named because the trainee attempts to restorer" the client. During the second stage, Interpretation, the trainee adopts a theoretical base, though it is often in a concrete and undifferentiated manner. As the supervisee recognizes the importance of the relational aspects of therapy, the trainee works his way through Rodenhauser's Realization stage. Finally, the psychotherapist becomes fully developed once the self, including one's feelings and reactions, becomes utilized in practicing psychotherapy.

Chazan (1990) proposed a three-stage model based on psychoanalytic theory. At the beginning of training, the trainee feels uncertain and insecure yet is excited and hopeful about becoming a psychotherapist. During the first stage of development, the trainee's effort is toward the creation of a space to provide a grounded psychological environment in which to learn and work. The supervisor's task during this stage is to provide a "home base," a secure holding environment. During the second stage, the supervisee begins building structure within the created space. The supervisee identifies with the supervisor and begins to venture out and explore newly learned techniques and approaches. During this stage, the supervisor must provide empathy, flexibility, tolerance, and tact. The supervisory environment continues to serve as a secure home base while exploration is encouraged. During Chazan's final stage of supervisee development, the supervisee experiences reciprocity and well-being. The supervisee now has a dependable sense of well-being coupled with professional identification. The trainee has become "good enough." Once the supervisee has reached this stage of development, the supervisor acts in a collegial, consultative manner.


Watkins (1990) proposed a six-stage model based upon Mahler's theory of

individuation. During the first stage, symbiosis, the supervised is anxious, dependent, and lacks insight. The trainee desires direction and advice while idealizing the supervisor. A primary task for the supervisor during this period is to provide a secure holding environment providing structure, empathy, and facilitating trust. Watkins refers to the next stage of counselor development as differentiation. During this period, the supervised becomes more attentive and discriminating with a more realistic image of the supervisor. The trainee is still anxious but is also excited about the anticipation of becoming a therapist. It is important that the supervisor remain active in providing a secure learning environment with empathy and support. The supervisor should use modeling, role playing, and discussion about the supervisee-supervisor relationship to stimulate thinking about the psychotherapist-client relationship. In the third stage, early practicing, the trainee is open to experimentation. Despite an increase in confidence, there continues to be a need to check back with the supervisor as the trainee's curiosity is mixed with hesitancy and doubt. The supervisor still maintains security within the structure of supervision, provides encouragement, and encourages the trainee to begin measuring the trainee's own feelings and reactions in response to using different techniques. Watkins' fourth stage is practicing proper. The supervised's confidence continues to grow as does excitement at being a therapist. As the therapist identity begins to crystallize, there is a tendency to overestimate one's competence. During this period the supervisor should function as the trainee's alter ego, providing a "reality check" when needed. During the fifth stage, rapprochement, the trainee struggles to individuate. There is a heightened but ambivalent desire for super-vision and the trainee exhibits


progressive and regressive behaviors. During this critical period, the supervisor must reinforce gains and offer ego-supportive interventions. While helping the supervisee to become fully aware of the developmental progress made, the supervisor validates the difficulties of being independent and autonomous. Watkins' sixth stage finds the trainee on the way to object constancy. Rather than achieving a status such as "master practitioner" the trainee becomes able to accept that professional development is, and will be, in process throughout the professional career. The trainee perceives the supervisor as an integrated whole, with both strengths and weaknesses, and has a strongly developed sense of professional identity. Supervision now operates with less structure in a collegial and collaborative manner. Responsibility for supervision can now be shared by both the supervisee and the supervisor.

Models proposed by Littrell, Lee-Borden, and Lorenz (1979) and Grater (1987) focus on the evolving process of supervision rather than the supervisee's or supervisor's specific needs. Littrell and associates (1979) proposed a model of counselor development where the trainee becomes a psychotherapist through an evolving process of supervision as directed by the supervisor. This model emphasizes how the responsibility of supervision begins with the supervisor and, as the trainee develops, is gradually shifted to the supervisee. During the initial stage of supervision, the supervisor provides goal setting and clarifies the nature of supervision and its components. During the counselortherapeutic stage, the supervisor acts as a counselor to the supervisee. According to Littrell and associates (1979), the teacher stage, the second stage, is when the supervisor teaches the skills of counseling to the supervisee. The supervisee continues developing


and reaches the third stage, the consultation stage, in which the supervisor works together with the supervisee in cooperation. Once fully developed, the psychotherapist selfsupervises, seeking to improve counseling skills through self-observation.

Based upon the supervisee development models presented earlier (Stoltenberg,

1981; Loganbill, et al., 1981; Hogan, 1964) Grater (1985) proposed a four-stage model of super-vision. Rather than proposing specific developmental issues for supervisee or supervisor, Grater proposed stages of tasks of developing complexity that should be accomplished by the supervisee with the direction and guidance of the supervisor. He reports his model is based upon the conclusions of several writers (Marmor, 1976; Parloff, 1979; Paul, 1967) that progress in psychotherapy results from an interaction between the client, the presenting problem, the therapy techniques, and the personal interactions of the psychotherapist. With this model, Grater argues that the goals of supervision should seek to promote therapeutic progress. The first stage of supervision proposed by Grater is the development of basic therapy skills and the adoption of the psychotherapist role. The second stage is composed of an expansion of the range of skills and roles to match the client problems and role expectations. The goal in the third stage of supervision is to develop the supervisee's ability to assess the client's habitual and conflicting behavior patterns, paying particular attention to those that occur within the psychotherapy environment. During the third stage the supervisor should also assist the supervisee in learning to select the appropriate intervention methods. During the fourth stage of supervision, the supervisor should help the trainee learn how to use the self in assessment and intervention. Grater points out that the first two stages are devoted to


skill development, whereas the last two stages emphasize a skill foundation to perform psychotherapy in an integrated manner.

While the above review does not include every model of supervised development that has been proposed, it includes the most significant proposals to date. It also allows the reader to gain an understanding of the highly similar nature of the models, as well as the variations. Models of supervisor development, reviewed in the next section are similar in stage composition and progression. Supervisor development

As developmental models of supervision have continued to be prevalent in the literature, increased attention has be given to the supervisor behaviors necessary to promote growth. Desirable supervisor behaviors have become more defined. Subsequently, there arose a need to account for the supervisor's acquisition of those behaviors. Models of supervised development were soon accompanied by models of supervisor development as it was acknowledged that ideal supervision does not "just happen." Haber (1996) has identified characteristics of beginning and advanced supervisors (see Table 1) which summarizes his perspective of how supervisors interact according to their development. Developmental models of the supervisor follow paths similar to those laid out by the models of counselor development.

Alonso (1983) proposed a theory of supervisor development which suggests

supervisors must resolve three key issues repetitively throughout their career. The three issues she identified were: self and identity, the relationship between psychotherapist and supervisor, and administration. Alonso (1983) proposed that throughout one's career,


TABLE 1: Characteristics of Beginning and Advanced Supervisors Beginning Supervisor Advanced Supervisor

Lacks substantial knowledge of supervision theory Well-informed about supervision theory and
and methodology methodology

Overuses one approach to supervision Implements diverse approaches suitable to specific
Anxious in the role of supervisor; overidentifies Understands role and responsibilities in the
with therapist role supervisory process

Intervenes largely by being the expert clinician More apt to use the resources of the supervisee Engages in power struggles with the supervisee Able to connect and work with the supervisee' s
over the right way world view

Underemphasizes or overemphasizes the person of Discriminantly explores the personal/professional
the therapist in supervision interface when it is relevant for clinical issues

Overreacts to the supervisee More aware of and able to utilize personal reactions
in the supervision process

Experiences diffuse boundaries and inability to Able to use clear boundaries to keep the supervisee
confront supervisee appropriately on task

Ignores cultural differences and learning styles of Respects cultural and learning idiosyncrasies
the supervisee

Works with supervisee in isolation ignoring Considers all of the floors in the professional and
ecological context client houses

Content-oriented Works with process, including the parallel process
between the therapeutic and supervisory

Source: Haber (1996, p.57)

novice to mid-career to late career, the issues remained the same but were perceived and resolved differently due to the perspective change due to experience and demands of life and the profession differing at each life stage. While Alonso's model demonstrates that supervisors do not reach professional stagnation once they assume the supervisory


mantle, it does not relate directly to the supervisor's development in regard to role competency or mastery.

Hess (198 6), on the other hand, proposed a three-stage model of supervisor

development that focuses on the supervisor's performance of supervision. The first stage is marked by anxiety as the new supervisor struggles with the change of role. Hess (1986) predicts that the new Ph.D. is frequently thrown into a supervisory role without much preparation. Only one in three interns receives any training in supervision (Hess and Hess, 1983) and the new supervisor struggles in the role change from supervised to supervisor. Hess (1986) suggests new supervisors cope with the self-doubt and ambiguity of their new role by adopting a concrete approach to supervision-focusing on skill development and client diagnosis. In Hess' second stage of supervisor development, exploration, the supervisor has gained competence and confidence and is often able to amaze and baffle the trainee with feats of psychological "magic." Supervision is valued as a professional activity and the supervisor's enthusiasm often promotes increased interest in learning by the student. Hess suggested the existence of two potential pitfalls during this second stage of development: 1) supervision that is too restrictive; and 2) supervision that is too intrusive. Hess' third stage of supervisor development is marked by a continuing increase in respect-the supervisor's respect for the trainee and trainee's respect for the supervisor. Increased attention is given to the trainee's learning and a deeper relationship develops. Supervisors at this stage are sought out by trainees due to their reputation as excellent teachers of psychotherapy. The fully developed supervisor


receives gratification when the student excels rather than when the supervisor is recognized as a good supervisor.

Rodenhauser's (1994) model of supervisor development has four stages:

emulation, conceptualization, incorporation, and consolidation. During the first stage, the new supervisor draws upon past role models for guidance and information of what to do. The second stage of conceptualization involves an active search by the supervisor for a systematic approach to super-vision. At this time, alliances can be formed with other supervisors to limit over-identification with supervisees. During the incorporation phase, the supervisor becomes aware of how one's personal style impacts supervision. This phase also includes heightened awareness of differences between one's self and the supervisee, such as gender, ethnicity, and lifestyle. In the consolidation phase of supervisor development, the supervisor is able to use theory, experience, the self, and supervisee countertransference within the supervisory process.

When Stoltenberg and Delworth (1987) revised Stoltenberg's (1981) Counselor Complexity Model, they included a model of supervisor development (Stoltenberg and Delworth, 1987). The supervisor model mirrors their supervisee model-it has four stages defined by three distinct levels, the third level having two components. At Level I, the supervisor is highly anxious and relatively nalve. Wanting to "do the right thing," supervisors at this first stage of development typically provide a structured environment for supervision, often performing as "the expert." As the supervisor develops, the supervisor may become confused and conflicted as supervision becomes perceived as a complex and multidimensional process. During this second stage, Level II, the


supervisor may be dependent upon colleagues for guidance. Level III sees the supervisor having resolved earlier conflicts and operating in a more stable manner. The Level III supervisor becomes more autonomous and wishes to improve supervisory skills. This supervisor's comfort level has increased as there is more security in the supervisor role. The integrated phase of Level III is marked by the supervisor becoming fully integrated across many domains. The fully integrated supervisor is able to work with all types of supervises and is willing to seek consultation as needed.

Watkins (1990, 1993, 1994) has proposed a four-stage model of supervisor

development that emphasizes the individual's role development. The four stages of this model are role shock, role recovery and transition, role consolidation, and role mastery. Unlike the other supervisor development models presented here, Watkins also provides us with recommended behaviors by the supervisor's supervisor, recommending that the professional should continue training during development of the supervisor role. During role shock, the new supervisor may feel like an impostor. The new supervisor experiences anxiety, lacks confidence, and feels overwhelmed. The new supervisor has no "supervisor identity" and struggles to provide competent supervision, going through much "soul-searching." It is likely that the new supervisor will be overly structured and concrete in supervision. The supervisor's supervisor performs as a teacher and a guide within a secure holding environment providing encouragement, direction, and teaching.

In role recovery and transition, the supervisor recovers from the shock of being a supervisor. There is an increase in confidence as well as awareness regarding one's strengths and weaknesses although the core of supervisor identity begins to form. It is


likely that one feels ambiguous about one's work as a supervisor. During this period the supervisor's supervisor begins to lessen the hold on the secure training environment and teaches and directs less. There is an increase in effort by the supervisor's supervisor toward stimulating an examination of the new supervisor's strengths and weaknesses and their effect on supervision. During the third stage of role consolidation, the new supervisor has a broader and more informed perspective of supervision. One's perceptions of one's own strengths and weaknesses are more realistic and is more trusting in oneself. The core supervisor identity is established and during supervision the supervisor is more able to address the process of supervision. The development of a personal theory of supervision also takes form by this point in development. The supervisor's motivations are driven more by the trainee's and clients' needs rather than narcissistic fulfillment. The supervisor's supervisor now acts as a supportive guide facilitating a focus on process issues and their impact. There is also an increased focus on the supervisor's personal issues and their impact on supervision. The new supervisor's independent thought and action is encouraged and supported.

Finally, the last stage, role mastery, marks the new supervisor's consolidation of skill and supervisor identity. The new supervisor handles all aspects of supervision well, possesses an integrated style and theory, and deals well with process issues. The supervisor's supervision is now sought and received in a collaborative and collegial envirom-nent. The supervisor's supervisor can be more confrontational while encouraging continued examination of the supervision process, personal issues, and the interaction. Supervision of supervision is now received on an as needed basis.


Supervisor development has not received as much attention as supervised development in the literature. As the influence of supervisor behavior upon the development of trainees becomes better illustrated, there are more questions as to how a supervisor can best promote growth and development. Questions regarding how a supervisor becomes a supervisor has prompted a discussion in the literature which seeks to describe the developmental stages of becoming a supervisor. These models follow the same developmental pathways delineated by the supervised models-the supervisors begin anxious and nalve to the task and progress toward an internalized professional identity. As the supervisor develops, competence and confidence increase, motivations become more attuned toward the learning of the supervised, professional identity becomes formed and the supervisor values the task of supervision. Empirical support of developmental models

The developmental models described above are more similar than dissimilar. Whether the model focuses on the developing counselor, the developing supervisor, or the process of supervision, each model assumes the beginning trainee to be nalve in regard to the psychotherapy (or supervision). They also assume the trainee will develop over time, with supervision, to become an integrated "master" psychotherapist (or supervisor). The models, for the most part, suggest that the supervisor should facilitate the trainee's growth by providing the environment, instruction, and feedback appropriate to the trainee's level of development. In the beginning, that appropriate environment is seen as a secure "holding" environment, safe and structured. The models recommend that the supervisor provide skills training in a proactive manner to the novice psychotherapist.


As the trainee learns and develops, it is suggested that the supervisor allow and encourage increased autonomy while continuing to support the supervised. As the trainee becomes increasingly developed, the supervisee-supervisor relationship is less teacher-student and more consultative and collegial with supervision occurring is a peer-like setting.

To determine the empirical support of these models, Worthington (1987) reviewed the research related to sixteen counselor development models and two supervisor development models proposed through 1985. In a similar review, Watkins (I 995a) reviewed the research related to six counselor development models and four supervisor development models proposed since 1985. Table 2 summarizes the findings of both Worthington and Watkins. According to these summaries, the empirical support has grown significantly during the past 10 years. The findings of these reviews show support of developmental models as a valid way of conceptualizing supervision. Furthermore, the empirical data reviewed by Watkins suggests supervision does indeed change as the supervised gains experience. But while continuing support for developmental models may further entrench them as the dominant way of conceptualizing supervision, Watkins (I 995a) laments the fact that empirical support has not progressed far enough during the last fifteen years. Watkins joins Worthington (1987) and Borders (1989) in suggesting that research needs a more substantial focus on the behaviors of supervised and supervisor to understand what is changing within the development process. They also argue that those behaviors need to be identified to understand how they improve psychotherapist functioning. According to Borders,

There is a need for descriptions of the thoughts, feelings, and behaviors of
supervises at various developmental stages that are more detailed than
these global descriptions, as well as more precise depiction of the


TABLE 2: Empirical support of supervisee developmental models

Conclusions by Worthington Conclusions by Watkins

a) There is some support for developmental a) The majority of studies found some degree of
models as proposed by Hogan (1964) and support for developmental models or a
others. developmental dimension of supervision.

b) In general, perceptions of supervisors and b) Perceptions of both supervisors and supervisees
supervisees have been broadly consistent with were broadly consistent with developmental
developmental theories, model thinking.

c) The behavior of supervisors changes as c) The supervision relationship and supervisor
counselors gain experience, behaviors were perceived at least by
supervisors to change in developmentally appropriate ways as supervisees gain experience.

d) The supervision relationship changes as d) Beginning supervisees were generally found to
counselors gain experience. The relationship be more in need of structure, direction, and
between supervisor and supervisee is guidance and supervisors were found to
influenced by the supervisee's perceptions of provide that more so for them.
his/her supervisor.

e) Advanced supervisees were found to generally need less structure, direction, and guidance and supervisors were found to act accordingly.

f) Advanced as opposed to beginning supervisees were found to be more willing to consider their own personal issues and their effect on therapy.

g) Some research showed that, with experience, supervisees' needs changed in developmentally appropriate ways and that, over time, supervisees showed increase in developmentally relevant structures.

h) The majority of studies tested Stoltenberg's (198 1) complexity model or tested developmental tenets more generally.

i) Stoltenberg's complexity model received the most support overall.

Source: Worthington (1987, p. 195-201); Watkins (1995a, p. 655 and 668).

supervision environment that 'match' each stage or foster movement
toward higher stages. (1989, p. 17).


Worthington (1987) and Watkins (1995a) both found the state of supervisor development theory to be lacking. In his review of supervisor development models, Watkins concluded,

supervisor development models have been fashioned after supervisee
development models; the number of stages, the content of those stages,
and the progression within and across stages are all largely similar for both supervisor and supervised. The main difference seems to be in the foci of the models more than anything else, with the supervisor being the primary
focus in one case and the supervised being the primary focus in the other.
(I 995a, p. 674).

Worthington (1987) reviewed eight research articles that dealt specifically with supervisor development. Watkins reviewed an additional three research articles that dealt with the same issues. A summary of their conclusions is presented in Table 3. After considering these conclusions, Watkins (1 995a) reported that the research in this area has been minimal. In addition, there needs to be more of a focus on behaviors specific to the developing supervisor if much is to be gained. He adds that since the majority of supervisor developmental models simply mirror supervised developmental models, not much more would be gained with the addition of more supervisor developmental models. Watkins also reiterates Worthington's conclusion, "There is little specification of what makes a supervisor effective and thus of how one builds the skills necessary to become effective" (1987, p. 206).

While developmental models continue to be the preferred perspective reported in the supervision literature, little advancement has been made beyond the confirmation that supervision does change as supervises progress through their training. It is likely that supervision is generally more structured for beginning trainees who are typically anxious,


unskilled, and naIve regarding the psychotherapeutic process. As trainees gain experience, competence, awareness, and insight, it appears likely that supervisors act in an increasingly collegial and mutually collaborative manner, enforcing less structure and becoming less directive. Despite the prevalence of developmental models, many reviewers argue that the focus of research should not be directed toward the development of new models, but rather the validation and refinement of the existing models (Borders, 1989; Worthington, 1987; Watkins, 1995a). Furthermore, many argue that specific behaviors and processes should be identified that promote growth within the developmental framework.

TABLE 3: Empirical support of supervisor -developmental models Conclusions by Worthington Conclusions by Watkins

a) There are differences in skillfulness in a) Some of the dimensions that define
supervision across supervisors.) supervisors' perceptions or cognitive map of
supervision have been identified.

b) Supervisors do not become more competent as b) Experience level has not been found to affect
they gain experience. supervisors' cognitive maps of supervision.

c) Supervisors change little in other ways [ways c) Some of the issues, such as competency,
other than competence] as they gain critical to supervisor trainees have been
experience. identified.

d) More experience supervisors have been found to be more highly developed with regard to their views of themselves as supervisors." e) The supervised developmental models of Littrell et al. (1979) and Loganbill et al. (1982) have received minimal to no research support [when applied to supervisors].

f) Sansbury's (1982) supervised developmental model or hierarchy of supervisory issues, when applied to supervisors, has been partially supported.

Source: Worthington (1987, p. 203-205); Watkins (1995a, p. 676)


Theories of supervisor development also lack empirical support. Although levels of skillfulness has been shown to vary across supervisors (Worthington, 1984; Zucker and Worthington, 1986), it has been demonstrated that supervisors do not necessarily improve across time or experience (Marikas, Russell, and Dell, 1985; Worthington, 1984). There is also no evidence that supervisors actually change over time though they may have a more highly developed sense of self as a supervisor (Watkins, Schneider, Haynes, and Nieberding, 1995). The call for research designed to identify effective supervisor behaviors as well as how those behaviors can be developed has been made (Worthington, 1987; Borders, 1989) and essentially ignored (Watkins, 1995a). SLipervision as Relationship

While the developmental models continue to be the zeitgeist of supervision

theory, there is a small body of literature devoted to the relationship of supervisor and supervised. In the words of Hess, "supervision is a relationship in which one person's skills in conducting psychotherapy and his or her identity as a therapist are intentionally and potentially enhanced by the interaction with another person" (1987, p. 255-256). By developing this relationship, Hess argues, one can transcend the need for theory based on developmental models of supervision. In presenting his argument for a theory of relationship, Hess relates the research and experience of four other writers, Markowitz, Grunebaum, Aldrich, and Swain.

Markowitz (19 5 8), though an accomplished psychotherapist, decided to enter into supervision with six supervisor of different theoretical orientations. Markowitz


discovered that the supervisor's ability to relate was the most important factor in being able to provide effective supervision. Grunebaumi (1983) studied the qualities that psychotherapist's look for in their own therapist. Of the four fundamental criteria identified, two were related to personal style and one's ability and/or willingness to interact. The four criteria were: 1) competence; 2) non-overlapping professional and social networks; 3) warmth, caring, and flexibility; and 4) the therapists' willingness to talk rather than remain silent. Aldrich and Swain studied the criteria on which supervisees and supervisors evaluated each other.

Aldrich (198 1; Aldrich and Hess, 1983) found that supervisees appraised

supervisors on: defensiveness (how comfortable is the supervisor with the supervisee's comments); professionalism (modeling); clinical experience (what skills can be provided); theoretical base (how adept conceptually); experience as a teacher; appropriate interest in the student's life; likability; and motivation. While a concern for skills training and what can be learned is evident, so is the quality of the supervisor-supervisee relationship. Swain (1981) found supervisors to judge supervisees on eight criteria: interest in client and client welfare; preparation for supervision; knowledge; selfawareness, self-exploration, self-disclosure and self-esteem; openness to suggestions; clinical and interpersonal skills; boundary management; and decision-making skills. Issues pertaining to the relationship are also present here, though not as prevalent as in Aldrich's (1981; Aldrich and Hess, 1983) findings. Reviewing these findings it appears that the supervisor and supervisee both regard the quality of their relationship an integral component of the supervisory process. The supervisee appears to measure the quality of


supervision directly from the quality of the relationship, regardless of the supervisee' s developmental level. Despite the apparent importance of the relationship, few studies have focused on supervisor-supervisee relationship characteristics (Watkins, 1995 a).

Hess (1987) proposed a theory of supervision relationship based on the work of Martin Buber (1970). Buber proposed that the concept of "I," as in one's self, does not stand alone, but exist only in relation to a "You," a "Thou," or an "It." As we each exist in the world, we ("I") can relate to others as a "You," a "Thou," or an "It." Buber pointed out that relationships exist at varying levels of meaningfulness. The levels vary from superficial wherein a person relates to the other as an object rather than a fellow human being, to deeply personal when a person experiences and acknowledges the humanness of the other person. Designating the relationship to be of "I-It," "I-You," or "I-Thou" reflects the understanding of the depth of the relationship. As we order our hamburger at the drive-up window, we are typically engaged in an "I-It," or even "It-It" relationship as we treat each other as an object at the other side of a goal-one to serve and the other to be served. To engage in small talk while waiting for the order, to acknowledge each person as a person, is a deepening of the relationship to "I-You." To feel genuine empathy with the worker's sore feet and frustration of a seemingly never ending line of customers steps toward an "I-Thou" relationship. To acknowledge the humanness of the other and to have the other respond to your acknowledgment further deepens the relationship as it takes on the flavor of "I-Thou." Within the "I-Thou" relationship occurs a dialogue between two people, each recognizing the humanness of the other. The dialogue takes place within what Buber referred to as the "gaze" (Buber, 1970).


The gaze can be conceptualized as a specific, molecular, behavioral unit (Hess, 1987). The gaze occurs when a person hears the "I" of another person addressing the "Thou" in the self. In our example, the gaze occurs when the fast food clerk hears the customer express her concerns about his fatigue and frustration. The gaze can occur at any time and any place, whenever and wherever two people may interact. It is a purely psychological event with no requirement of space or time. It is for this reason that Hess argues the gaze is a useful way for a supervisory interaction to occur. In addition to being free of space and time constraints, the gaze is super-theoretical, meaning it can be used within the context of any approach to supervision, whether from within a counseling model or developmental model of training. For the gaze to occur, the supervised and supervisor must connect on a meaningful level, communicating within a meaningful relationship. By attaining an "I-Thou" relationship, the supervisee-supervisor communication transcends the constraints of any specific or given technique. The establishment of such a relationship, Hess argues, should be fundamental within any supervisory, or psychotherapeutic, relationship.

Development of the supervisee-supervisor relationship has received little attention. While the importance of the relationship is almost universally accepted, whether a supervisor can develop or learn how to promote a meaningful relationship has not been demonstrated. Furthermore, it is unclear what components of the relationship or aspects of the supervisor and supervised are most salient to a productive supervisory relationship. The next section will review the limited amount of research that has attempted to identify dimensions of supervision.


Dimensions of Supervision

Dimensions of supervision can be thought of as broad-based structures in which supervision occurs. By identifying relevant dimensions it is hypothesized that appropriate behaviors within each dimension can be identified, quantified, and measured. Ellis (Ellis and Dell, 1986; Ellis, Dell, and Good, 1988) has attempted to identify and confirm the dimensions as proposed within a developmental framework. Ellis and Dell (1986) tested the content and dimensionality of Bernard's (1979) dual dimension and Littrell and associate's (1979) singular dimension supervision models from the supervisor's perspective. Using a multidimensional scaling research design, Ellis and Dell found a three-dimensional cognitive map of supervision.

Bernard's (1979) supervision model is a 3x3 matrix defined by two dimensions. The two dimensions identified by Bernard are supervisor roles and supervisor functions. The three supervisor roles identified by Bernard are teacher, counselor, and consultant. The three supervisor functions identified by Bernard relate to process, conceptualization, and personalization skills. The function of supervision is to assist the supervised develop these three types of skills. Process skills focus on the execution of counseling techniques. Conceptualization skills includes those skills that organize and synthesize information about the client. Personalization skills include awareness about one's feelings, values, and beliefs and how they may impact the therapy. The three functions of supervision interact with the three supervisor roles to produce nine categories of supervisor behavior (see Table 4). Each of these nine categories presumably corresponds to supervised needs.


TABLE 4: Bernard's 3x3 matrix of supervisor roles and functions.
Supervisor Supervisor Roles
Functions Teacher Counselor Consultant
Process Demonstrates specific Helps trainee determine what Works with trainee to
interpersonal, treatment, or hinders or facilitates their explore different uses of an intervention skills interventions with a specific intervention and jointly
client; focuses on reducing practice them; mutual inhibitions, increase learning of skills
Conceptualization D-emonstrates one or more Helps trainee understand how Works with traine-e-to ways to classify', organize, their stereotypes, explore mutually issues
and understand client's conceptualizations, and and implications of
behavior, thoughts, and unresolved issues affect the theories, models, and problems counseling session and provide alternative
alternate perspectives conceptualizations to
Personalization Demonstrates or describes Helps trainee work through Works with trainee to
the potential importance of personal issues or feelings explore mutually personal trainee's affect and ways of associated with counseling concerns relevant to recognizing and using one's sessions counseling.
own affect during counseling ,______________ I_________I_Source: Ellis and Dell (1986) p. 284; Bernard (1979)

The four-stage unidimensional model proposed by Littrell and associates (1979) is

unidimensional. The dimension identified by Littrell is the control and responsibility for

supervision. Table 5 summarizes this unidimensional approach. As the counselor

develops, the trainee moves along the developmental stages. At the beginning of

training, the nalve supervisee is directed by the supervisor regarding the development of

training goals as well as what to expect within the supervisory relationship. As training,

and presumably development progresses, the supervisor gradually relinquishes control of

the supervisory processes until the supervisee is finally self-sustaining. The supervisee

progresses from complete reliance upon the supervisor to complete self-reliance.

In testing Bernard's and Littrell's models, Ellis and Dell (1986) found partial

support for Bernard's model, but little support for that of Littrell. Ellis and Dell (1979)

found, however, that Bernard's model was better supported when a fourth role was added

to their dimension of supervisor roles. The addition of self-supervision to the matrix


TABLE 5: Littrell, Lee-Borden, and Lorenz four-stage unidimensional model of supervision

Stage Focus Control/Responsibility of Supervision

1. Establishment of supervisory Supervisor facilitates goal setting and the definition of the
relationship supervisory relationship

2. Counselor-teacher Supervisor teaches the techniques of psychotherapy

3. Consultant Supervisee assumes more responsibility for own leading, using the
supervisor as a consultant is a mutually collaborative role

4. Se If- supervision The supervised is responsible for his/her own leading, operates

Source: adapted from Ellis and Dell (1986).

changed the original 30 matrix (Supervisor Role x Supervisor Function) to a 40 matrix

with twelve identifiable supervisor categories (see Table 6). These new supervisor roles

account for the continued development of the psychotherapist after completing formal


TABLE 6: Three additional supervisor roles added by Ellis, Dell, and Good Self-Supervision Self-supervision Role
Process Leaming new interventions, techniques, or diagnostic strategies or
skills; self-exploration to reduce inhibitions, encourage experimentation
Conceptualization Learn new conceptual approaches or models; determining multiple perspectives of client's behaviors, thoughts, affect, and problems, then developing multiple intervention goals and strategies
Personalization Personal issues and feelings related to counseling sessions or
clients; leading new ways to recognize and use one's feelings and intuitions
Source: Ellis, Dell, and Good (1988, p. 318)

Having identified twelve distinct categories of supervisor behaviors, Ellis, Dell,

and Good (198 8) found that the twelve categories could be evaluated along three

additional dimensions of supervision. The three dimensions identified by Ellis and


associates (19 88) are bipolar in nature and interact with each other to produce a threedimensional model of supervision. The three bipolar dimensions are:

Process and personal foci Conceptual focus
Directive Non-directive
Behavioral Emotional

According to Ellis and his colleagues (1988), each of the twelve supervisory behavior categories (see Table 4 and Table 6) vary uniquely along the continuum of each of the above three dimensions. For example, Teacher-conceptualization behaviors are high on conceptual focus and directive focus, and toward the midpoint on the behavioralemotional axis. On the other hand, Self-supervisor-personalization behaviors are toward the process, non-directive, and emotional poles (Ellis et al., 1988). Ellis and Dell (1988) have commented on the value of these dimensions as a step toward being able to quantify appropriate and facilitative supervisory behaviors as well as to identify when and how these behaviors should occur during the development of the supervisee.

Regardless of the theoretical stance on supervision, the purpose of supervision is to train aspiring psychotherapists and prepare them to operate independently in a professional manner. It has been demonstrated that supervisors vary in their effectiveness-there is, in effect, "good" supervision as well as "bad" supervision. It is presumed that "good" supervision leads to improved training and subsequently better prepared and more effective psychotherapists.


What is GoQd-Supervision?

From a theoretical or intuitive perspective, it is relatively easy to identify the

components of good supervision. Reviewing recommended supervisor behaviors by the authors of counselor development models can provide us with a fairly comprehensive picture of good supervision. On the other hand, research of what makes good supervision is limited to a handful of studies. In attempting to define "good supervision," one must first acknowledge that there is no "ideal formula" for conducting supervision (Bruch, 1974; Fleming and Benedek, 1966). We can acknowledge, however, that, like in psychotherapy, there are general "guideposts" in supervision that can, and should govern its practice (Storm and Heath, 1985; Carifio and Hess, 1987). The search for good supervision leads to the performance of good supervisors.

In their review of the "ideal" supervisor, Cariflo and Hess (1987) divided their quest into three components. The first component is identified by the question, "Who is the ideal supervisor?" This question refers to personal characteristics that make the supervisor a good teacher, educator, mentor, or psychotherapist. The second component and question, "What does the ideal supervisor do?" refers to the activities and techniques such as role modeling and role playing used by the supervisor. The third question, "How does the ideal supervisor perform supervision?" relates to the conditions, styles, and overall environment the supervisor creates and/or uses in conducting supervision. This intuitive categorization will be used in this section of the paper to explore the concept of good supervision.


Who is the Good S"ervisor?

According to Rogers (195 7), supervision should be a model of the psychotherapy process. Therefore, he thought supervisors should exhibit empathy, understanding, and unconditional positive regard. Furthermore, Rogers felt that congruence and genuineness should also be demonstrated by supervisors in their interactions with supervises. Coche (1977) agreed with Rogers about the importance of these fundamental characteristics, and added his belief that the supervisor's willingness for self-disclosure is an important characteristic. Research and theory in the area of supervision have generally supported Rogers' notions about supervision (Carifio and Hess, 1987). To this basic foundation of supervisor "musts," other theoreticians and researchers have described additional factors believed to be important in positive supervisory interactions (Albott, 1984; Aldrich, 198 1; Gitterman and Miller, 1977; Hess, 1980). The supervisor characteristics these writers thought were important include flexibility, concern, attention, investment, curiosity, and openness. Unlike the characteristics described by Rogers, however, few of these characteristics have received empirical support (Carifio and Hess, 1987).

While the facilitative conditions of empathy, respect, genuineness, and

concreteness are deemed important in supervisory interactions, it is unknown how much of these aspects is the right amount, or the appropriate amount. Lambert (1980, 1974) investigated this question by comparing levels of facilitative conditions between therapists and their clients as well as between supervisors and their supervises. He found that supervisors generally exhibited less of these conditions than did therapists, and that both groups varied the amount in varying circumstances. While we do not know


how much is the appropriate amount, it does appear that the appropriate amount may vary according to specific circumstances.

What Does the Good Supervisor Do?

As noted earlier, supervision is a complex task. The good supervisor must be able to perform a variety of tasks during her interactions with supervisees. A fundamental criteria is that the supervisor should be knowledgeable and experienced in therapy (Albott, 1984; Fox, 1983; Hess, 1980; Windholz, 1970). It has been recommended that early meetings between supervisor and supervisee should be structured with the intention of providing information to the supervisee and supervisor. These early meetings lay the foundation for the developing relationship between supervisor and supervisee. That relationship should involve openness, trust, mutual understanding, two-way communication, and collaboration (Bruch, 1974; Finch, 1977).

An important component in the supervisee's and supervisor's collaborative

relationship is goal-setting. It has been suggested that trainees and their supervisors set specific, explicit, and measurable goals (Archer and Peake, 1984; Fox, 1983). Mutually agreed upon goal setting allows both the supervisor and supervisee mark the supervisee's progress during the course of supervision. It has been suggested that an open discussion of progress on pre-established goals will ease the trainee's inherent anxiety about evaluation (Freeman, 1985).

Another task for the supervisor is disentangling the trainee's report of the events in psychotherapy, and to determine what actually happened (Bruch, 1974; Melchiode,


1977; Coche, 1977). This task appears to be a more covert goal of supervision and supervisors appear to be'less explicit in defining it to their trainees (Cariflo and Hess, 1987). In comparing therapists' written summaries with recorded material of psychotherapy sessions, it has been demonstrated that the trainees' recollection of events may not be particularly helpful (Couner, 1944; Chodoff, 1972). The use of audio or video tape to record trainee's psychotherapy sessions has proven to be a more reliable record, though the intrusiveness of such measures may counter-indicate their use, depending upon the goals of supervision (Goldberg, 1985).

It has been suggested that the supervisor should use a variety of teaching methods and modes of data presentation. For example, Brannon (1985) described how a variety of teaching techniques could be used to communicate information and knowledge. Brainstorming, for instance, allows the exploration and exchange of novel ideas in a noncritical environment. Role playing allows the participants to "try out" a variety of techniques therapeutic outcomes, finding new solutions. Modeling behavior by the supervisor allows the demonstration of certain behaviors for the supervisee's later use. Guided reflection allows the replay in supervision of therapist-client interactions that in turn promotes supervisee awareness.

Due to the nature of the supervisory relationship and supervision's often parallel relationship to psychotherapy, the issue of performing psychotherapy as part of supervision is often discussed. Generally, the literature recommends against psychotherapy during supervision. While some authors (i.e. Orzek, 1984; Riorch, 1980) suggest the supervisor should help the supervisee fight inner battles and penetrate the


supervisee's own defenses, it appears that this is better suited for a separate therapist in a separate setting (Ekstein and Wallerstein, 1972; Lambert, 1980). How Do Good Supervisors Perform Supervision?

The good supervisor is able to influence the supervisee's behavior through a variety of ways (Cariflo and Hess, 1987). A supervisee's perspective can be significantly altered by a supervisor's personal style (Markowitz, 1958). It has been suggested that to provide a positive supervisory experience, the supervisor will create a trusting and open environment in which two-way communication and collaboration are valued (Bruch, 1974). In their approach, the supervisor should be supportive (Bruch, 1974), noncritical (Melchiode, 1977), confident, enthusiastic, and open to a supervisee' s questions and input (Albott, 1984). Dorn (1985), using a social influence model, suggested that a supervisor's expertness, trust, and attractiveness improve his/her ability to influence supervisees.

A particularly important factor in how the supervisor conducts supervision

appears to be how he/she gives feedback to the trainee. Barth and Gambrill (1984) and Freeman (19 85) have suggested that feedback should be direct, immediate, and closely tied to the trainee's performance. Freeman also identified four characteristics to ideal feedback. To be most effective, feedback should be: 1) systemic, including objective, accurate, consistent, and reliable; 2) timely; 3) clearly understood and based on specific performance criteria; and 4) reciprocal. In addition to the feedback itself, it has been found that how it is delivered can greatly influence the trainee's acceptance. Cherniss


and Egnatios (1977) found that didactic, insight-oriented, and feeling oriented styles were better received by trainees than were confrontational or authoritarian styles. In terms of the teaching that occurs in supervision, Goin and Kline (1974) found that a supervisory approach that was neither passive nor directive but "middle ground" was most suitable to most supervises. On the other hand, there is no conclusive evidence that teaching styles should be more didactic or experiential in nature (Cariflo and Hess, 1987). What is Good Supervision?

In attempts to define good supervision, the discussion invariably turns to a

discussion of good supervisors. Good supervision is apparently performed by a good supervisor, and good supervisors apparently provide good supervision. It appears that the good supervisor shows respect, empathy, genuineness, concreteness, and self-disclosure in interactions with supervises. It appears that these characteristics vary in the level in which they are presented. The good supervisor is knowledgeable and experienced in both psychotherapy and supervision. The good supervisor works within a collaborative relationship with the trainee, setting mutually agreed upon goals. The supervisor uses an appropriate variety of teaching methods and tools and avoids doing psychotherapy in supervision despite the development of a meaningful relationship. Good supervision is performed in an open and trusting environment and the supervisory relationship is typically viewed as collaborative involving two-way communication. The good supervisor performs supervision in a supportive and thoughtful manner, being neither too directive nor too passive. Feedback is provided in a manner to enhance learning and


acceptance by the trainee. These are all qualities of good supervision that have been identified in the literature.


The primary purpose of the present study is to determine the compatibility of psychological literature research describing good supervision to how the supervised experiences good supervision. The bulk of the literature on supervision is theoretical in nature. Furthermore, the empirical data reflects the measurement of a few characteristics believed to be important in the provision of psychotherapy supervision. The phenomenological method to be used in this project is designed to discover the essential elements of an experience rather than confirm or deny theorized components. Once the essential elements of the supervises' experience have been elicited and identified, they will be reintegrated into a description of good supervision as experienced by the trainees as they develop into professional psychologists. The obtained description will then be compared to the existing research of good supervision.


A phenomenological method was used in this study to explicate supervises'

experiences of good supervision. The particular method used was developed by Giorgi (1989). Using the fundamental precepts defined by Husserl (1983/1913) as guidelines and assumptions, Giorgi has developed a specific and rigorous phenomenological method for psychological research (Giorgi, 1989) based, in part, on the model of Merleau-Ponty (1962). The methodology is comprised of four theoretical processes: 1) empirical observation, reduction of the data, search for essences, and synthesis into the whole.

The first "step" of the method is to empirically obtain a description of the

phenomenon to be studied. The goal of the experimenter at this stage is to obtain an experimentally non-biased description of the phenomenon to be studied. To achieve this goal, it should be a candid concrete description obtained from another person, the "subject," without additional explanation or inquiry. Once the description is obtained, the experimenter performs a reduction of the data into psychological terms.

The second step, the reduction of the data, is the transformation of the descriptive narrative provided by the subject into a psychological description of the subject's experience. For the reduction to remain true to the subject's experience, "the experience must be present to what is given precisely as it is given" (Giorgi, 1989, p. 45). To achieve this "presence," Giorgi emphasizes the importance of intuition on the part of the experimenter. Intuition which in part aids the absorption of the uncontaminated and 59


meaningful description. To exist at the time of the reduction is not enough; one must be fully present to understand the experience. Additionally, for the absorption to remain uncontaminated, bracketing is vital, as prior knowledge and beliefs of the phenomenon must be disregarded in order to be fully present to the description as obtained. Bracketing is the process of actively partitioning preconceived ideas and beliefs about the phenomenon from conscious thought.

The third "step" of the method calls for a search for the essences of the description. The essences of an experience are the most invariant meaning for a phenomenon within the given context (Giorgi, 1989). These essences are what make up the essential structure of the phenomenon studied. This step involves reflection on the part of the researcher to discover the psychological meaning of the obtained description as it comes forth through the transformed narrative. The narrative, however, is more than the sum of its parts, and each part cannot be understood without an understanding of its relationship to the whole. The phenomenologist labels these parts as constituent elements.

Finally, once the constituent elements of the experience are discovered, they are synthesized into a whole description of the meaning. This description constitutes the structure of the experience, the results of the study. The synthesis, or integration involves reflection on the continual influx of consciousness within the context of the situation as the final step of the analysis of the description.

The four "steps" described above provide a general prototype for

phenomenological research, leading to a number of possible methodologies which may be


utilized to answer the type of research questions addressed by phenomenology. The methodology used is this study is the phenomenological method developed by Giorgi (1989).


The design of this study, including the number of subjects, is based on several published studies which have utilized phenomenological methodology (Giorgi, 1992; Anjus, Osburne, and Koziey, 1991; Guglietti-Kelly and Westcott, 1990; Stevick, 1971). Five subjects, all predoctoral psychology interns, were selected on the basis that the experimenter anticipated that they would be able to provide full descriptions of their lived experience of being supervised in a clinical setting. So they were able to communicate freely with the experimenter, an additional selection criterion was that they were professional peers, in good standing, of the experimenter.

The subjects ranged in age from 26 years old to 34 years old. Four of the subjects were women, one was a man. All five subjects were in the final month of their predoctoral internship at the same APA-accredited internship site. All five subjects were enrolled in clinical psychology training programs; three were in university programs, two were in professional school programs. Two of the subjects were enrolled in the same university program, but in different specialty tracks. The five subjects had received training in twelve of the fourteen training rotations available at the internship site.



The phenomenological research process as outlined by Giorgi follows a rigorous process. The qualitative data was collected through the use of interviews in which each subject's experience of supervision was fully explored. The audio recorded interviews were then transcribed verbatim, and these transcripts, hereafter referred to as protocols, comprised the data of the study. The data was then analyzed following Giorgi's method of phenomenological research, a four step method comprised of observation, identification of meaning units, identification of constituent elements, and reintegration of constituent elements into the description of the experience. Prior to engaging in this process, it was necessary for the experimenter to identify his own biases regarding the experience and then bracket them, consciously separating them away from the data collection and analysis procedure.

Self-Reflection and Bracketing

Prior to beginning the data collection, it was necessary for the experimenter to bracket all known experiences and beliefs regarding supervision. To be successful, the experimenter first performed a "self-reflectioif' on the phenomena. This process entailed writing out all thoughts and beliefs regarding supervision, then reducing that written expression to identify previously held beliefs and potential biases. The results of that reduction are presented in the Discussion chapter of this paper so the reader may be informed of potential biases of the researcher. After these biases were identified and bracketed, the experimenter continued with the data collection.


Data Collectio

The data was collected by conducting unstructured interviews. All five interviews were conducted by the experimenter. Each interview was scheduled one week in advance. When the interview was scheduled, each subject was told that the nature of the interview would be a discussion of the subject's positive experiences of supervision received throughout professional training. At that time each subject was asked to reflect upon such experiences as a way to prepare for the interview.

Prior to beginning the interview, each subject was asked to read and sign an

informed consent which included information about the audio taping of the interviews. Each interview began with the same question, "Think back to a time when you experienced what you believe to be a positive experience in supervision. Tell me about the experience, what happened, what was said, what was felt, anything you believe was part of that experience." After the subject answered the initial question, the interview proceeded by asking the subject to elaborate upon what was originally introduced into the discussion by the subject. Except for the first question, the interview did not follow any set plan or seek to confin-n pre-established hypotheses, but reflected and followed the subject's lead. The five interviews ranged in length from 50 to 75 minutes. The five interviews were conducted within a ten day period.

After the fifth interview was conducted, each of the five interviews were

transcribed, in the order in which they were completed. The experimenter completed all five transcriptions. The interviews were transcribed verbatim with the exception of names and other identifying data which was changed to protect the anonymity of the


subjects as well as any individuals they discussed during the interview. After the accuracy of the protocols was confirmed by the experimenter, the audio tapes were destroyed. The written protocols comprise the data set of the study which was then analyzed as follows.

Data Analysis

The method of analysis for each protocol followed the method for psychological phenomenological analysis as outlined by Giorgi (1989). This method is a specific four step procedure which is based on the general phenomenological model described earlier, and is detailed below. Prior to beginning the analysis, it was essential for the experimenter to continue to bracket all he knew concerning the phenomenon of good supervision. The conscious and purposeful act of bracketing drew upon the selfreflections discovered by the experimenter prior to beginning the interviews. Each protocol was analyzed individually following the procedure described below. A sample protocol with corresponding meaning units is presented in Appendix A.

First, the entire description was read through in its entirety two times. The

purpose of these initial readings was to get a sense of the protocol as a whole. Reading it twice allowed the experimenter to better understand the protocol as a description of one person's lived experience. After a general understanding of the protocol was achieved, the next step was taken, the process of understanding the meaning of the experience.

Since it is difficult, if not impossible to analyze the entire protocol at once, the

second step of the procedure consisted of dividing the protocol into parts. As the goal of


the phenomenological method is to discover meaning, the protocol was divided into what Giorgi calls, "meaning units" (Giorgi, 1989). These were delineated by points of transition of meaning within the protocol. In other words, each individual unit contained one individual meaning which revealed something about the subject's experience of supervision. The division into meaning units was accomplished by beginning at the beginning of the protocol, reading until one meaning unit delineated itself, and marking the point of transition on the page. The experimenter then continued reading the protocol from that point, marking each successive meaning unit as it appeared. After completing the protocol, the researcher returned to the beginning of the protocol, testing each meaning unit individually. This "test" determined 1) whether each unit could stand on its own, and 2) if each unit contained only one meaning. If the meaning unit withstood the test, the researcher advanced to the next unit. If not, the meaning unit was corrected. This correction was performed by either adding the unit to an adjacent meaning unit (if the unit failed to stand on its own) or by splitting the meaning unit into two or more units (if the unit contained more than one meaning). This correction procedure was conducted once through the entire protocol, at which time the determination of meaning units was complete.

Once the meaning units were determined, they were transformed to reveal their implicit psychological value to the meaning of good supervision. This step was the transformation of the subject's own words into psychological terms-the transformation of the subject's own narrative description of his/her experience into the generalized psychological meaning of that experience. These psychological meanings were noted in


the margins of the protocol to correspond to the meaning unit from which it came. After the psychological meanings were identified for each meaning unit of the protocol, the constituent elements for the protocol were identified. The constituent elements were comprised of psychological meanings which appeared to be most central to the subject's experience. The constituent elements were identified by the experimenter's reflective presence with the data and subsequent understanding. The constituent elements could be conceptualized as the significant themes that emerged from the subject's narrative. After the constituent elements of the protocol were identified, the analysis of the individual protocol was complete. Each of the remaining four protocols were analyzed following the same procedure.

When the five protocols were analyzed, the constituent elements were listed

together. This list formed the constituent elements of the experience of good supervision as determined by the five individuals interviewed. Using all of the constituent elements for the group, the final step of the analysis was undertaken, following a prototype set by Colaizzi (1978) and described by Stevick (1971). The elements of the group were used to obtain an overall structure of being supervised by synthesizing them into a whole. This process began by determining which, if any, of the elements were common among the subjects. Elements that occurred on a less frequent basis were then considered. Common and uncommon elements were integrated to form a whole, unified, description of good supervision, The aim of the integration is to describe common aspects of the experience while allowing for individual elements to be considered. This initial integration of elements is reported as the essential description of good supervised in the Results chapter.


This description was then further reduced after reflecting upon the themes of the essential description. This final reduction yielded the structural description of good supervision. This description is also reported in the Results chapter of this paper.


The results of this study is presented at three levels of depth and complexity: the original meaning units; the essential description which is comprised of the constituent elements; and the structural description, which is a reduction of the essential description. The data analysis of the five subject protocols yielded meaning units for each narrative. The meaning units are presented in Table 7 beginning on the next page. The meaning units were then reflected upon to determine the most significant, central, aspects to the experience of being supervised. The central elements are identified as the constituent elements. The constituents elements may be comprised of one or more meaning units. In other words, a meaning unit may be significant enough on its own to be deemed a constituent element. Additionally, a constituent element may be composed of two or more meaning units to comprise a unit of higher and more complex meaning. These constituent elements are reintegrated to produce the essential description of the experience of good supervision. Finally, the essential description was reflected upon and ftirther reduced to produce the structural description. The results of this study are comprised of all three of these types of information.



TABLE 7: Meaning units derived from the five subject protocols
Subject I Subject 2 Subject 3 Subject 4 Subject 5

1.SPRa experienced Aware of Depth Organized Experience SPR was close in
as personable and of knowledge depends upon the age, allowed
professional interpersonal identification
process between
SPE b and SPR

2. Committed time Love of teaching SPR explicit re: Seriousness of respected due to
to supervisee expectations, responsibilities, competency and
performance evident through organization
requirements organization

3. Provided specific Relationship-- Provided in vivo SPR gave clear SPE actively
feedback SPR's investment, modeling in expectations of sought SPNc
attributed to session what was wanted because of
personality respect, wanting
factors to get info from

4. Open Genuineness Assessed current Expectations Invested,
communication, level of ability at implied committed to
allowed the onset of SPN confidence, training and
disagreement maturity, fulfilling SPE's
experience needs

5. Felt free to ask Supportive and Gave confidence Kept scheduled Regular
questions validating when @ beginning, appointments communication
teaching after assessing,
gave approval

6. Informal Growth within Monitored Structure and SPE challenged to
structure, relationship performance consistency is push for own
boundaries periodically with important to learning
minimized, tapes learning new
permitted information

7. Personality and Relationship Flexible in Structure and Nonthreatening
behavior reflected moves from process of SPN, consistency not as
openness superficial to depending on important when
deeper level, needs of SPE familiar with
bonding, fostered subject
by SPR's
personality style

8. Explicit regarding Process moves Always made Structure may Nonpunitive
process and from info time and gave change--requires
expectations of gathering to attention to SPE, adaptation, more
SPN consulting sense of priority aggressive
behavior by SPE
to have needs met
Note: a refers to supervisee, 'refers to supervisor, 'refers to supervision


TABLE 7--continued
Subject 1 Subject 2 Subject 3 Subject 4 Subject 5

9. Process attributed Bond allows for Feedback Implies potential Validating of
to personality discussion of perceived as of being feelings,
deeper issues accurate embarrassed if particularly
need to ask important at
questions difficult times
with patients
10. Although Bond dependent SPR got feedback Trainee actively Understanding of
informal, there upon from other team seeks SPN if SPE's
were boundaries interpersonal members, aware necessary perspective,
style/personality of interactions, confirmingrole, effort allowed increased

11. SPR encouraged Personality Reinforced good Course of SPN is Provided
self-disclosure factors: open, work responsibility of instruction for
genuine SPR growth within a
safe environment
12. Friendly, Self-confident, Identified poor Negative Interested in SPE
empathic centered, and work, made experience when at a personal
honest suggestions for not informative level-invoked a
improvement good feeling in
13. Likable Being supportive Honest Seeing potential Accurate
important to of SPR increases attentiveness to
decreasing eagerness to seek SPE's state of
anxiety SPN mind and
confronted SPE
14. Interactive, Relationship Impressive as role Structure not as SPN was at a
encouraged works on model important as new personal level,
discussion reciprocal basis: information, SPR personal growth
take something of had positive long which impacted
value from SPR, termi relationship SPE professional
work harder to behavior
give back/repay
15. Reputation among Interpersonal SPR expressed Respect of SPR SPR must be
peers style factors, confidence in knowledge and competent and
behavioral in SPE, increased willingness to good role model
addition to verbal self-confidence acknowledge what they don't
16. Was not as Style identified as clarity of SPR's Respect Balanced support
available as being due to who expectations supersedes and validation
desired, due to the person is, provided an scheduling with challenging
her popularity including external measure, structure and identifying
confident, non- helped SPE things to improve
punitive organize her goals upon


TABLE 7-coninued
Subject I Subject 2 Subject 3 Subject 4 Subject 5

17. Joint decision SPN as an "art" Organization and SPN form-at Left SPN
making regarding explicitness varied- question knowing I learned
direction of demonstrated and answer, something today
therapy, helped SPR experience review, report
build SPE and knowledge

18. SPE's judgment SPR Assistance with Interaction varied Able to open up
was trusted communicates integration into and receive
needs, changes, the team feedback on
corrections, SPE increased sense of personal aspects
maintain self- security while maintaining
respect appropriate
19. Feedback, Experienced at an Good professional SPR has desire to Genuineness
possibility for affective level reputation with be a SPR, to teach
learning, other disciplines
dependent upon

20. Weakness of SPN In relationship, Training included An intrinsic desire I liked SPRs as
overcome with developed processing of to see another people
SPR's interest in potential through specific person grow,
SPE's growth as a reassurance, interventions and pleasure at
therapist support, suggestions fostering that
understanding growth

21. Relationship SPR's investment Adapted to SPE's Confident and Given opportunity
beyond SPN is shown through professional, secure, wants to to open up,
important, time, energy, interpersonal share knowledge,
whether or enthusiasm; style-built on not fearful of
teacher or socially observed and strengths of SPE competition exchanged with rather than mimic the relationship SPR.

22. Mutual respect SPR appears to Makes specific Format and Trust develops
enjoy what they suggestions for process varies over time, are doing. interventions includes SPR
sharing with SPE

23. Responsiveness Commitment to Provides Positive feedback, Trust required
and support for time appropriate acknowledgment
SPE reinforcement of growth,
including learning
recognition to 3rd


TABLE 7-continued _____Subject 1 Subject 2 Subject 3 Subject 4 Subject 5

24. Self-disclosure Involvement at an Sets professional Identifies Identification
affective level model individual as with SPR
mentor important in
establishing goals
and establishing
SPR as a role
25. Professional SPE experiences Maintains Identified specific Gender aspect
freedom in work at intuitive level professional requirements minimized by
boundaries needed to attain SPE, though
though professional goals gender important
increasingly re: identification
''personable'' over

26. Style of giving Learning in a Respect. Honest SPR competence
feedback is developmental and SPE respect
attributed to building model, is a prerequisite
personality rather than of having +
factors correction. supervisory
27. Important for SPE Time is Professional and Able to overcome No respect equals
to not feel as productive personal personality flaws questionable
though she was reputation. learning equals
wrong, but that why bother with
there are many SPN?
ways to do

28. No gender related SPE is more Conscientious An understanding No respect, more
differences open, less regarding her of SPR guarded
defensive presentation personality led to
mutual respect

29. Confidence SPE seeks Takes time, Expectations were SPR expresses
built-trusted my suggestions, increases effort at rigid confidence in
judgment, corrections times of increased SPE, particularly
provided varying SPE need. when specifics are
experience to identified
push the training


TABLE 7--continued_____Subject 1 Subject 2 Subject 3 Subject 4 Subject 5

30. Beginning at a Teaching can be SPE learned Experiences SPR Poor SPN results
new site, it is reciprocal-- about boundaries/ at an affective in + learning,
necessary for SPR relationship unprofessionalism level through
to help give dependent through a poor experiencing
confidence SPN experience, anger towards
SPR, how it can
3 1. Dedication to SPE's needs Poor boundaries, Despite closeness, Personal and
training-taking change with SPE was scared no experience of professional
some risks, development boundaries disclosure does
providing new crossed not include
challenges emotional
disclosure re:
SPN relationship
32. compromises with As SPE develops Poor boundaries, SPR open to SPE's activity,
SPE to provide as a therapist, SPE lost respect diverse ways of willingness to new experiences becomes more for SPy, thinking participate in
while getting the knowledgeable, important, was
necessary work confident, reinforced
done expectations

33. Own willingness Expectations Poor boundaries, SPR open to Instillation of
to participate, include "I know SPE learned diversity--more confidence
expose more of what I'm importance of important later necessary to
weaknesses, to doing" professional when less need to perform and
learn boundaries, more learn new develop
aware of when material professionally
they are being

34. SPE's openness SPE has needs, Character traits: Rigidity okay Personal aspects
has increased expects to be self-confident, when learning must be addressed
over time, with fulfilled, organized, new material in SPN, goes
confidence, insightful, warm, beyond skill
personal and caring, empathic training
professional with boundaries.

35. Maturity may be SPE expectation- Helpful regarding Modeling role not Technician vs.
related to be able SPR shares other professional primary therapistto handle negative experience issues, i.e. office importance in therapist must be
feedback and politics evaluating willing to
perceptions of supervisory face/work through
SPN experience experience personal issues


TABLE 7-continued
Subject I Subject 2 Subject 3 Subject 4 Subject 5

36. Felt connected SPE belief. it is Honest-straight Behavior is a To be good
important to still forward function of who therapist personal
be learning the person is issues must be
addressed in SPN
37. Interpersonal trust Identifies warm Positive, even SPE has respect Reach a point in affective when giving for SPR's position relationship able
experience as due negative feedback and to disagree,
to personality responsibilities comes later in
variables development
38. Sense of equality, A guide, mentor SPE expectations Respect- SPR shares own
learn from each of needing to universal training
other, process of improve clinical requirement experiences
change, different skills insecurities
39. Relationship Highly Identification of Ethics- Time and
outside SPN trustworthy specifics prerequisite for availability
affects experience illustrated SPR's positive important
and perception, interest and experience
both + and attention

40. Involved as a Expectation of Felt respected by
person, not just a SPR to help- being given SPE give confidence responsibilities
and to teach
41. Roles change, Knowledgeable, Factors required
evolve, but not specific in to be a good SPR:
separable feedback, is interest,
genuine, commitment,
approachable, patience
role model.

42. Flexible within Went out of his
boundaries way to advance

43. Needs of SPE

44. Relationship is
experienced as
"special" creates
an extraordinary
experiencebeyond good.

45, Requires a
willingness to be
close, involved


TABKE 7-coninued
Subject 1 Subject 2 Subject 3 Subject 4 Subject 5

46, Challenged to
develop/grow as therapist 47. Positive
reinforcement 48. Positive SPR is in
tune with SPE, touching all aspects during SPN

49. SPR, modeling
how to do therapy

50. Knowledge and
competence prerequisite (otherwise no learning) 51. SPR's knowledge
is measured by how much SPE learns
52. Trust developsbeing to SPN on time, how they treat you at difficult times 53. Dynamic and
evolving 54. Development of
relationship mirrors
therapeutic relationship 55. One who is
willing to share one's affective experience 56. SPE likes them as

57. Increased trust
allows SPE to expose more of self, develop into a better therapist


Essential Description of Good Supervision

The compilation of constituent elements of the five subjects' descriptions of good supervision generated the following essential description of good supervision.

For the five subjects interviewed in this study, good supervision was experienced as a period of learning and professional growth within the context of a meaningful interpersonal relationship. The relationship was experienced as being collaborative between supervisor and supervisee, teacher and student, mentor and novice. As supervisees grow and develop, they expect their supervisors to change with them, matching their interventions and teaching methods to the supervisee's need. The supervisor's commitment to the supervisee's learning was identified as a fundamental component of good supervision.

Supervisees approach supervision anticipating their own professional

development. There is an implicit power differential in the supervisory relationship in which the supervisee initially concedes control and power to the supervisor. On the other hand, the supervisee expects to be treated with respect in addition to being taught by the supervisor. The supervisor is also expected to relinquish power and control to the supervisee as the supervisee develops skills and proficiency at the psychotherapeutic task. The supervisee's evaluation of supervision is dependent in part on how well these expectations are met. The value attributed to supervision by the supervisee is also dependent upon how much the supervisee can and does learn from the supervisor during the course of supervision.


In order for ideal learning to occur, the supervisee must feel safe enough to expose weaknesses and open oneself to scrutiny by the supervisor. The supervisee's desire and ability to take more risks, both in therapy as well as in supervision, is a result of developing confidence and feeling safe within the supervisory environment. Supervisee confidence is developed by being trusted, given positive feedback, and being provided with challenging clinical cases to further develop clinical skills. Safety is not experienced unless the supervisee is able to trust the supervisor and the supervisory relationship. Good supervision also draws the supervisee into the process, encouraging the supervisee to be more active in the training process.

Good supervision was attributed directly to the actions and personality of the supervisor. Professional competence of the supervisor is measured by how much the supervisee learns from the supervisor. The extent to which the supervisee values supervision is determined by how much the supervisor has to offer. Supervision with supervisors perceived as competent, knowledgeable, and skilled is actively sought out by supervisees. Good supervisors identified in the study were also described as honest, ethical, genuine, supportive, self-confident, warm, caring, insightful, and empathic. Good supervisors encourage, or at least allow, the supervisory relationship to develop from a superficial task-oriented relationship to a deeper relationship wherein the supervisee can share and better understand affective experience. Good supervisors are also explicit in their expectations and evaluation requirements. Good supervisors are in tune with and responsive to their supervisees. Commitment to training in general as well as


commitment to the supervised's training were also identified as important components to good supervision.

Supervisors demonstrate their commitment to training through their commitment of time to their trainees and their attention to the trainees during their allotted time. Supervisors also demonstrate their commitment by being willing to share their own training experiences, particularly difficulties. Supervisors can also demonstrate their interest in the trainees' learning by asking questions about their training expectations, needs, and whether they are being met.

While the supervisor's personality and likability are identified as important, it is more important to maintain a professional relationship; although the relationship may become more informal as the trainee develops. The maintenance of appropriate and professional boundaries was experienced as being the supervisor's responsibility. Boundary concerns were expressed by all five subjects and was experienced as a necessary component to a safe and productive relationship. Mutual respect was also identified as a necessary prerequisite of good supervision.

Feedback was reported to be an important part of good supervision.

Appropriately given feedback was experienced as validating, accurate, and contributing toward the supervised's professional development. Positive feedback contributed to the self-confidence of the trainee and increased the trainee's commitment to work for the supervisor. Negative feedback, when received, was experience by the trainee as an honest effort by the supervisor to contribute to the supervised's development. Helpful


negative feedback was described by the subjects as being timely, noncritical, accurate, specific, and focused on behaviors or interventions rather than personal attributes.

The above essential description is a compilation of constituent elements extracted from the meaning units identified in the five subject protocols. This description is a summary of good supervision as experienced by the five subjects interviewed as part of this study. This description is reduced one more time to yield the structural description.

Structural Description of Good Supervision

Further reflection upon the above essential description of good supervision

yielded the following structural description of good supervision as experienced by these five subjects.

Good supervision was expressed by all the subjects to be a highly personal and individualized process. Good supervision was attributed directly to a good supervisorgood supervision is performed by a person with a commitment to the supervised's training. From the trainee's perspective, supervision is not a skills dependent process but rather a person dependent process, where the supervisor's own personality and interpersonal style is identified as the reason supervision is experiences as good or bad.

Important supervisor attributes which were identified by the subjects include

honesty, genuineness, insightfulness, and self-confidence. Good supervisors also support and validate the trainee's experience, are committed to the profession and to professional training, and respected by students and professional peers. Good supervisors exhibit professionalism, are cognizant and respectful of trainee-supervisor boundary issues, and


are explicit regarding performance expectations. Nonauthoritarian styles of communication and interpersonal warmth were also identified as desirable, though not necessary, characteristics.

Good supervision is experienced as a period of learning when the trainee is able to invest in the training process through the establishment of a trusting interpersonal relationship with the supervisor. The development of a meaningful and safe interpersonal relationship allows the trainees to become more invested in the training process by permitting themselves to ask more questions, admit to difficulties, and be less defensive in the supervisory process. The trainee enters supervision with expectationsexpectations to learn, expectations to be respected, expectations for the supervisor to commit to the training process. How well the supervisor meets these expectations has a direct impact on how beneficial the trainee experiences the training. Good supervision is experienced as a dynamic process with the supervisor able to track and respond to the trainee's changing needs. Good supervision facilitates growth of the trainee by encouraging the trainee to seek additional supervision and training.


The purpose of utilizing the phenomenological method in this study was to

discover a structural account of the experience of good supervision during psychotherapy training. This study strives to understand supervision as experienced in the lived world, related by the subjects in their own words. Supervision was selected as the phenomenon of study due to its nature as a fundamental method in training psychologists. Despite its central role in professional training, clinical supervision has received relatively little empirical research attention (Watkins, 1995a). A discovery oriented approach such as this is particularly useful when applied to phenomena with relatively scarce research findings.

The results of this phenomenological investigation will be discussed in this

chapter. The primary constituent elements of the discovered reductions will be discussed individually and related to appropriate traditional research findings. The reductions in their entirety will then be discussed in a comparative manner to previous phenomenological studies on supervision. Before beginning with the discussion of the results, however, the experimenter's personal reflections regarding supervision are presented. These reflections were written prior to beginning the data collection. They are presented here so the reader may be informed of any potential biases held by the experimenter.



Personal Reflections

An important aspect of this phenomenological method is the bracketing of preconceived knowledge regarding the phenomenon being studied. To effectively bracket, it is necessary to identify those preconceived beliefs. The following is a written representation of the experimenter's beliefs regarding good supervision prior to beginning the data collection. It is presented here so the reader may be informed of the experimenter's biases prior to beginning the project.

As I was preparing the proposal for this project, a then current supervisor of mine suggested to me that good supervision is nothing more than the provision of a safe and secure environment in which the trainee could learn psychotherapy. I was initially drawn to this elegantly simple conclusion and began to believe that yes, perhaps good supervision could be reduced to such a simple but potentially comprehensive thought. As I began to reflect upon my own experiences of being supervised, however, the matter became a great deal more complex. My first recollections of good supervision began with memories of supportive supervisors who encouraged me, expressed confidence in my work (even when I was not confident in it), and, of course, gave me glowing reviews at the end our training period together. As I attempted to identify specific incidents or experiences in supervision which seemed to have a particular impact upon my growth as a therapist those experiences flattened out into average experiences. Such experiences were experienced as good, as in average. These experiences must have contributed to my development as a psychotherapist, but not in a way that prompted me to identify them as critically important to my professional identity.


As I reflected further upon my past experiences, it became more apparent to me that supervision most profoundly affected my professional development when it helped me overcome some form of adversity. The adversity typically took the form of an inner struggle-a crisis of loss of confidence, boredom within a therapeutic context, an interpersonal conflict with another supervisor, peer, or client. In my discovery I found that when I was in a good supervisory relationship, I had assistance in resolving the matter. When I was not, I was left to my own devices to handle the crisis, which at times felt as though I were negotiating a land-mine. So, what was the difference?

When I had "good" supervision I felt I could bring the struggle to the attention of my supervisor who would then share a past experience, insight, or a fresh perspective to help me resolve the issue. I felt, in a word, "safe." When I reflected upon the times when I resisted bringing the issue to supervision, it was, I think, because I was afraid of being "scolded" or penalized for doing something wrong. I determined I would be safer negotiating the land-mine on my own. While my hesitancy clearly falls under the category of personal "issues," of interest here is that there were times when I did take the issue to supervision. It seems, then, that a sense of safety is a fundamental component of good supervision, after all, as supervision may not even take place without it.

As far as other characteristics of good supervision, I expect to learn in supervision and I expect to be treated in a fair manner. While I don't know if likability is a necessary component, I do know that I liked those supervisors from whom I received good supervision. Good supervision has challenged me, supported me, and encouraged me. But most of all, good supervision has taught me new things about psychotherapy-new


skills, new processes, new understandings-and has encouraged me to use that knowledge in the psychotherapeutic setting.

Orientation of the Subject

In reviewing the results of this study it becomes clear that experience of good supervision is understood within the context of the supervisee's world. A significant aspect of their "world of supervision" is their understanding of the supervision process. Each of the subjects reported that they approached supervision with expectations. The expectations of these supervisees involved their development as professionals and how they expected to be treated as human beings. These expectations could be summarized as expectations to learn, be supported, and be treated with respect.

As supervision is a context for training, expectations to learn are reasonable and normal. Supervision has been described as primarily a teaching-learning activity (Holloway and Neufeldt, 1995). The learning process in supervision is also well documented, at least theoretically. Developmental models emphasize the learning processes of the supervisee with the assistance of the supervisor (i.e. Stoltenberg and Delworth, 1987; Watkins, 1993, Ronnestad and Skovholt, 1993) and supervisors themselves experience the supervisor role to include the role of teacher (Paskiewicz, 1993). The quality of learning has also been shown to be related to the quality of the supervision (Carifio and Hess, 1987; Shanfield, Matthes, and Hetherly, 1993)

Expectations to be treated with respect are consistent with APA guidelines for ethical conduct. Writing as a supervisor, Pickering (1987) tells us that,


students have a right to expect that the supervisory process and
relationship will not wound them to the point of bewilderment, will not
create anxieties that they are given no guidance in handling, and will not
be a place where harmful interpersonal games are played (p. 50).

Supervised expectations and value of respect for self-autonomy and integrity has been documented elsewhere (Hutt, Scott, and King, 1983). Additionally, respect as a necessary component in the supervisor-supervisee relationship was a conclusion of Carifio and Hess (198 7) in their search for the ideal supervisor.

Expectations to be supported may be a reflection of the supervises' needs to be guided through the process of becoming a psychotherapist in a thoughtful and human manner. Such expectations are legitimate (Pickering, 1987). Supervised expectations of support are also consistent with the expectations and intent of supervisors when they are engaged in a supervisory relationship (Paskiewicz, 1993; Pickering, 1987). Furthermore, the relationship qualities of warmth, acceptance, trust, and understanding, the "facilitative conditions," are well-documented in the literature as important dimensions in all helping relationships.

One other significant expectation held by the trainees is the expectation to engage in a meaningful interpersonal relationship with the supervisor. In fact, for the subjects in this study, supervision is not experienced as a process of learning with the assistance of a person, but rather as a relationship in which learning occurs. Supervision is not attributed to learned skills, but rather, who the supervisor is and the relationship between the supervisor and trainee. Identification of the supervisor as the instrument of change suggests trainees do not perceive supervision as a professional skill which may be learned


as much as a process of shared learning within the context of an interpersonal relationship.

The perceived quality of the supervisory relationship, and consequently

supervision, is dependent upon the perceived quality of the supervisor, who is evaluated in terms of competence and commitment to training based, in part, on experience and maturity. Experience and maturity of the supervisor was measured by one subject, Karen (all names have been changed), as the supervisor's level of organization and explicitness regarding expectations and evaluation criteria. Karen inferred that since the supervisor was able to verbalize exactly what she expected during the course of training, that the supervisor understood the training process and was able to predict certain outcomes. This could only be accomplished, inferred the supervised, through experience and knowledge. It could not occur, apparently, through training and knowledge about the supervisory process. A significant consequence of this is that trainees do not anticipate the possibility of learning supervision but hold on to the notion that supervision is person dependent and a matter of either "you've got it or you don't." Consequences of this perspective are discussed later in this chapter.

Qualities of Supervisio

As stated earlier, it is difficult, if not impossible, to separate the quality of

supervision from the quality of the supervisor. In fact, in the present study, the benefits of good supervision were attributed directly to the personal characteristics and professional behaviors of the supervisor. For practical purposes, good supervision is a


good supervisor, and visa versa. Studies of supervision invariably include aspects of the supervisor, whether it is what they do, how they do it, or who they are. As it would not be reasonable to attempt to remove the supervisor from the process of supervision, the following discussion of qualities of supervision will include all aspects of supervision, including the supervisor.

While the facilitative conditions have been well-documented in the literature, other proposed important characteristics have received less empirical support. These intuitively derived characteristics include flexibility, concern, attention, investment, curiosity, and openness (Albott, 1984; Aldrich, 198 1; Gitterman and Miller, 1977). Despite the lack of previous empirical support, all of these characteristics were present in the subjects' protocols. Concern, attention, investment, and openness all appear to be significant supervisor characteristics present in good supervisory experiences while curiosity may be reflected in the supervises' perceptions of the supervisor's interest in the supervised. Perhaps the addition of these types of characteristics to the "facilitative conditions" is what delineates the supervisor from other helping roles. Super-visor investment in the supervisee's training seems to be a particularly important characteristic when fulfilling the mentoring aspect of the supervisory role,

The development of supervised confidence and trust was reported by the subjects to be related to self-disclosure by the supervisor. The importance of self-disclosure is consistent with the findings of Hutt and her associates (1983) although the desirability of supervisor self-disclosure has not been emphasized in the literature in general. Hutt and associates (1983) also proposed that self-disclosure by the supervisor is the link between


the supervisory relationship and the content of supervision, moving the process of supervision forward when the supervisee may otherwise be inhibited or reluctant to move forward. Carkhuff (1969) proposed that self-disclosure is an extension of genuineness and a necessary part of any helping relationship. According to Carkhuff (1969), the supervisor's self-disclosure relates to the supervisee's concerns and provides a model for increasingly deeper levels of self-exploration by the supervisee. The present findings suggest that self-disclosure represents an intimate sharing by the supervisor which in turn opens the door for similar sharing by the supervisee. These explanations do not seem mutually exclusive but complement each other nicely. Supervisor self-disclosure can dissolve blockages in supervision by increasing risk taking by the supervisee through the instillation of trust and commitment by the supervisor. This would also be consistent with the findings in this study that trust develops by how the supervisor treats the supervisee at difficult times.

Supervisee trust in the supervisor was described as a necessary component of good supervision.. These findings in the present study support similar findings elsewhere. Allen, Szollos, and Williams (1986) found that perceived expertise and trustworthiness of the supervisor were the best discriminators of quality according to supervisees. Other phenomenological investigations of good supervision found trust to be a central component of a positive experience (Worthen and McNeill, 1996; Hutt et al., 1983). Cariflo and Hess (1987) also concluded that trust is a necessary component of the supervisory didactic relationship.


Like trust, perceived competence of the supervisor is a prerequisite to a positive supervisory experience. It was stated that the supervisor's competence was a driving force in motivating the supervisee to seek supervision. Furthermore, the supervisor's competency was directly related to how much the supervision was valued by the supervisee. In the words of one subject, Manny,

She was very competent, she's really organized, which made me really respect her and want to seek out supervision even more because I knew
she knew a lot and I knew that she was competent so I wanted to learn, get
that information from her.

It is interesting to note that another subject evaluated the supervisor's competence by measuring how much she learned, suggesting the supervisor must always have something to offer the supervisee.

Trust, competence, facilitative qualities-all of these aspects of supervision are components of the supervisory relationship. The relationship between supervisee and supervisor has been proposed as the single most important aspect of effective supervision (Worthen and McNeill, 1996; Cariflo and Hess, 1987; Hess, 1987). In the present study, all positive aspects of supervision are experienced to occur within or as a direct consequence of the supervisory relationship. But is the relationship actually as important as it is perceived? In a recent study by Patton and Kivlighan (1997), the working alliance of the supervisory relationship had a differential impact upon the types of learning that occur in supervision. Patton and Kivlighan (1997) found that the working alliance had no impact upon the technical functioning of the supervisee but did impacted positively upon the trainee's comfort in supervision. More importantly, however, they found that the working alliance between supervisee and supervisor was related to the supervisee' s


performance in counseling sessions, as measured by independent raters. This appears to be the first empirical support that the quality of the supervisory relationship has an impact upon the professional performance of the trainee, validating years of speculation and theorizing.

Although the subjects in this study valued a collaborative supervisory

relationship, they also placed the responsibility of supervision upon the supervisor. This appears contradictory to the idea of establishing a collaborative relationship, but consistent with the supervisee' s implied deference to the supervisor as an authority figure. This suggests a tension between establishing a peer-like relationship with someone who has evaluative power over the supervisee. The supervisee's demand for respect and maintenance of professional boundaries may be a way of negotiating what could become an ambiguous and potentially harmful relationship. By maintaining boundaries, supervisees establish a structure in which they are able to learn, grow, and relate to the supervisor while being protected from arbitrary acts that may occur in a less structured or defined relationship.

While the responsibility of the supervision is perceived to be in the supervisor's hands, the supervisees do acknowledge that their active participation is required to make supervision work. In the words of another subject, Catherine,

part of the positive experience has been because of my willingness to learn
and expose myself ..to learn and receive the training and n ot be afraid to
have things dissected and pointed out.

Pickering (1987) states that it is likely that supervisors expect their supervisees to be active in the supervisory process.


Although this project did not intend to explore negative supervision, experiences of "bad" supervision were related by the subjects as a way of contrasting the positive experiences. The most significant expression of "bad" supervision was related to the breakdown of professional boundaries by the supervisor. One subject, Karen described her experiences with crossed boundaries as scary and unprofessional.

It was scary. It was unnerving for me, because here's this person [crying
to her about her personal problems] who's supposed to be my mentor ... I'd
hope they'd have some kind of professional sense ... I was scared. I was
overwhelmed ... I lost respect for her.

In such experiences, the supervised experiences anxiety, anger, and frustration and the supervisory relationship is burdened with mistrust, disrespect, and a lack of honest disclosure (Hutt et al., 1983). When supervision falls into a category somewhere between good and bad, it is likely that both the supervised and supervisor discount its value. Though it may not be harmful toward the supervised, when supervision fails, the relationship is dulled and an air of tolerance prevails (Paskiewicz, 1993). Hutt and associates (1983) report such an experience is likely to result in anger and anxiety in the supervised as a result of frustrated professional needs.

Boundary issues were identified by the subjects to be an issue which is necessary to a good supervision experience as well as being a primary contributor to a negative experience. Boundaries were experienced as a dynamic process rather than a static wall. Trainees experience the boundaries to become more flexible as the relationship develops. As the supervisory relationship deepened, the supervisor was experienced on a more personal level. All subjects reported, however, that a professional relationship with appropriate boundaries was maintained, even when interacting with the supervisor on a


social level. The boundaries were a desired and welcomed aspect to the relationship. In no instances did the subjects report frustration at wanting to be closer to the supervisor than was permitted. It was as though the trainees experienced the boundaries as a protective mechanism for themselves, and they were able to apply their experience with supervisory boundaries to counseling situations.

Another characteristic of good supervision is the supervisor's ability to track, attend to, and respond to the needs of the supervised. The subjects experienced this aspect as attention and responsiveness-attention to their needs and emotional states followed by appropriate responses and interventions. In a study of psychiatry resident training, supervisors with the highest ratings identified and tracked the residents' central and effectively charged concerns about the client (Shanfield et al., 1993). This is attentiveness in a direct sense, and can be illustrated by Manny's experiences.

The one thing that I've had with her that I haven't had with most of my supervisors, is that she would really key in on my attitude, or what was
going on in the moment. And she would confront me with it.

Another aspect of the attentiveness is how a supervisor tracks unidentified supervised needs and tailors the supervision to match it. The types of needs a supervised can have in supervision can vary. Some needs vary according to developmental level, as in what type of learning is most appropriate for the trainee. Other needs are context sensitive, such as difficulties with a client, spoken or unspoken. Still other needs may be related to the interactions with the supervisor as when the trainee needs more direction or more autonomy. Appropriate interventions at times of difficulty were described as being instrumental in developing a trusting relationship with the supervisor.


It has been suggested that the most significant interventions provided by

supervisors is feedback (Cariflo and Hess, 1987). Feedback is anticipated, expected, and closely tied to a trainee's performance (Barth and Gambrill, 1984). Supervisees in this study experienced feedback as an instrumental aspect of their development as psychotherapists. According to Freeman (1985), feedback should be systematic, timely, clearly understood, and reciprocal. According to the subjects in this study, it is most important for feedback to be accurate (part of being systematic), noncritical, and relevant to what the supervisee is doing. Both of these guideline address the fact the feedback must be received and accepted by the trainee before any behavioral and attitudinal change will take place. In this study, negative feedback given within the context of a good supervisory relationship was experienced as the supervisor's contributions toward improving the skills of the trainee. While the content of the feedback was aimed at changing a supervisee' s behavior, the manner in which it was presented was experienced as supportive, encouraging, and building upon the supervisee's strengths and knowledge. Positive feedback was experienced as a motivating force for continued development by the trainee. Positive feedback let the trainee know that the supervisor was paying attention and interested in the trainee's growth. It also increased trainee confidence and, in some cases, motivated the trainee to work harder for the appreciative supervisor.

How Does This Relate to Developmental Models?

It was not the intent of this project to support or not support hypotheses proposed by the various developmental models of supervisee and supervisor development. Some