Development of the Psychotherapy Supervisor: Review of and Reflections on 30 Years of Theory and Research

Introduction

“I feel like a fraud, a female resident in supervision said.

What do we know about the interior experience of the developing psychotherapy supervisor? Does it sometimes involve the fraudulence of feeling referred to in the preceding quote? Do fear, anxiety, and conflict routinely besiege beginning supervisors as they practice their craft? How might supervisors’ internal life (e.g., cognitions, affect, self-efficacy), skill level, and supervisory identity formation change as they acquire supervision knowledge, training, supervision, and practical experience? Those questions, which have long been matters of concern in the area of psychotherapy supervisor development (Alonso, 1985; Blair & Peake, 1995; Hess, 1987; Cohen & Lim, 2008; Heid, 1997; Pelling & Agostinelli, 2009; Rau, 2002; Watkins, 1995c), continue to be highly pertinent today because: (1) our understanding of the interior experience of the developing psychotherapy supervisor remains quite limited (Majcher & Daniluk, 2009; Pelling, 2008), and (2) if we are to have a more informed, wholistic perspective on supervision, then a more complete understanding of both supervisee and supervisor as participants in the psychotherapy supervision process would seem requisite (Watkins, 1995d).

As a recognized area of theory, research, and practice, psychotherapy supervisor development can be defined as the examination and explication of:

Psychotherapy supervisor development is of considerable importance to better understand because the supervisor plays a central, substantive, and pivotal role in the whole of the supervisory process, affecting all aspects of supervisee learning and growth and, in turn, affecting patient learning and growth.

That appears to be a view that finds some resonance within the field. As Majcher and Daniluk (2009) have recently stated, “The consensus in the literature is that there is a critical need to study supervisor development …” (pp. 63-64).

At this juncture, the matter of psychotherapy supervisor development has been with us now for approximately a generation–with Alonso’s (1983) developmental life-span theory perhaps serving to mark its beginning point. As we look back over that past generation of activity, punctuated by theoretical construction, empirical research, and qualitative inquiry, what can be said about psychotherapy supervisor development now? What do we know? What do we need to know? What needs to happen for this area of study and inquiry to blossom and grow?

In this paper, I would like to consider those questions. Bernard and Goodyear (2009) recently speculated that there is a “lack of any new interest” (p. 297) in supervisor development and declared that research “is virtually nonexistent” (p. 296). But what do the data actually say? I believe a current review about psychotherapy supervisor development could be useful because:

In what follows, I hope to: (1) briefly describe the primary models of psychotherapy supervisor development that have been proposed; (2) provide a summary review of supervisor development theory/research reviews that have been conducted up through 1995; (3) provide a current critique of supervisor development research studies, surveying and scrutinizing all investigations—quantitative and qualitative—that have appeared from 1/1996 up through mid 2011; and (4) provide a contemporary status report about supervisor development and offer some possible directions for its future.

Psychotherapy Supervisor Development Models: 1983-1995

Psychotherapy supervisor development models were products of the 1980s and 1990s. During that time period, five primary models were proposed: Alonso (1983, 1985), Hess (1986, 1987), Rodenhauser (1994, 1997), Stoltenberg and Delworth (1987), and Watkins (1990, 1993, 1994). No new supervisor models have been proposed since then (Bernard & Goodyear, 2009). Although Stoltenberg has continued to include a chapter on supervisor development in each new edition of his book (Stoltenberg & McNeil, 2009; Stoltenberg, McNeil, & Delworth, 1997), no real changes to his original model have been made over the past twenty-five years. Table 1 provides a summary of the primary features of each of the five models.

In surveying those models, the essential orienting points to bear in mind are: (1) all models are generally cast as either a three or four stage linear progression; (2) with Alonso’s life-span model excepted, the models track supervisors from when they begin to supervise through the process of their becoming seasoned, master practitioners; and (3) as supervisors successfully progress through the stages, supervisory identity and skills are theorized to become increasingly enhanced, and supervisory self-doubt, insecurity, and negative affect about oneself as supervisor are theorized to become decreasingly in evidence. Cohen and Lim (2008) nicely capture this process of supervisor developmental unfolding across models:

It is that very process–stage-by-stage progression through and increased mastery of developmental struggles, issues, and crises–that seemingly defines the essence of supervisor developmental models (Bernard & Goodyear, 2009; Cohen & Lim, 2008; Falender & Schafranske, 2004). Supervisor growth is theorized to occur cognitively (e.g., conceptual clarity), affectively (e.g., anxiety giving way to joy in supervising), behaviorally (e.g., expanding one’s arsenal of supervision strategies), and in identity formation and consolidation. Heat becomes light; maturing into the supervisor role increasingly evolves; supervisor development comes to be seen as a welcome, ongoing, lifelong process (Falender & Shafranske, 2004; cf. Grant & Schofield, 2007). As Table 1 shows, all the models are far more similar than dissimilar in their content, structure, and progression.

Considering Findings From the Most Recent Reviews, 1980-1995

As we attempt to look across the last 30 years of supervisor development, I would like to begin by summarizing the primary findings of the three earlier reviews: Worthington (1987), Russell and Petrie (1994), and Watkins (1995b). Over the course of the time period covered in those three reviews, approximately 1980 through 1995, the five previously-mentioned psychotherapy supervisor development models were proposed, and the status of those models–their theoretic specifics and empirical support (or lack thereof)–was addressed by the reviewers.

Worthington (1987)

His theory/research review, appearing at the beginning point of psychotherapy supervisor development theory, identified the then recently proposed models of Alonso (1983) and Hess (1986). Because this area was just emerging, there was very limited material upon which to draw. Worthington indicated that “the field [theoretically] is at a rudimentary level” (p. 191); “theories are still general and imprecise” (p. 206), and are lacking much crucial explanatory material (e.g., identifying the variables that effect change in supervision).

Eight studies were identified that in some way addressed an aspect of supervisor development, such variables as planning statements, supervision emphases, and licensure status were examined for their supervisory impact. Only one study, however, was informed by a (supervisee, not supervisor) developmental model (Miars, Tracey, Ray, Cornfeld, O’Farrell, & Gelso, 1983). Based on Worthington’s analysis, he concluded that the research suggested:

With such a limited body of studies, however, there was little else left to say: “[T]he empirical investigation of how supervisors change with experience is at a rudimentary level” (p. 206). Perhaps the one word that Worthington repeated a few times in his review, which seemed to best capture the state and status of psychotherapy supervisor development at that time, was “rudimentary.”

Russell and Petrie (1994)

In addition to the Alonso and Hess models, Russell and Petrie identified two other proposed models of supervisor development—those of Stoltenberg and Delworth (1987) and Watkins (1990, 1993). Based on their review, they concluded that “there is considerable similarity among the [four] theories of supervisor development” (p. 34). “The models … [1] provide preliminary guidelines for creating effective supervisory dyads in training environments …. and [2] provide guidelines for developing training environments for supervisors” (p. 35). Due to the similarity across models, they also indicated that no further model building was needed.

Russell and Petrie made clear that the supervisor development models lacked any empirical support, but they suggested that that possibility could be remedied: “These models provide directions and hypotheses for research on supervisor development” (pp. 35). They further added that “… these models allow for specific hypotheses concerning supervisors’ behaviors, thinking styles, emotions, and perceptions to be proposed and tested” (p. 35). In their view, the models were eminently testable and ready for study. Model testing, not further model building, was sorely needed.

Watkins (1995b)

In this review, the newly proposed supervisor development model of Rodenhauser (1994) was also recognized. Watkins’ conclusions were highly consistent with the earlier reviews and largely echoed their findings. With regard to theory, he concluded that (1) “supervisor development across stages shows far more similarity than dissimilarity”… (p. 673), (2) “important [model] elements are still missing, for example, the identification of specific impediments to supervisor growth, identification of critical incidents”… (p. 674), (3) “these models are lacking in … a transition theory to guide them”… (674), and (4) “the role of key personality, moderator variables needs to be delineated better …” (p. 675). From Watkins’ perspective, theory needs in 1995 were no different from theory needs identified by Worthington almost a decade earlier.

Not surprisingly, that “sameness” applied for research as well: “[A]ll of what Worthington identified as research needs then remain as research needs now” (p. 677). Watkins identified only three studies of supervisor development that had appeared since Worthington’s review; that limited body of work left little new to say. He called for attention to such neglected variables as supervisor behaviors, cognitions, intentions, and personality traits in the study of supervisor development. Like Russell and Petrie (1994), Watkins also saw empirical possibility here: “[S]ome researchable hypotheses … present themselves for study … [and] await… systematic, rigorous empirical attention …” (p. 675). Again, these models were seen as eminently testable and ready for study. In this review, the need for attention to supervision of supervision was also accentuated.

Summary

As of the mid-1990s, psychotherapy supervisor development had been declared “the richest yet most untapped facet of the clinical supervision endeavor” (Watkins, 1995a, p. 34). Five supervisor development models had been proposed: Alonso (1983, 1985), Hess (1986, 1987), Rodenhauser (1994, 1997), Stoltenberg and Delworth (1987), and Watkins (1990, 1993, 1994). All the models were judged to be quite similar in structure, seemed clinically valid (see Russell & Petrie, 1994), yet remained incomplete in some respects (e.g., lacking a transition theory) because of their newness; research was limited but the models were deemed to be research ready and to hold empirical promise. Attention to and interest in psychotherapy supervisor development appeared to be high, and all indications suggested that this area was poised and primed to be taken to its next level of conceptual, experimental, and practical scrutiny and sophistication.

Considering Research on Psychotherapy Supervisor Development, 1996 to 2011

I would like to complement that 1980 to 1995 summary review with an examination of supervisor development research that has appeared during the last approximate 15 years. What has happened since that beginning, formative period? I thought it best–for a possible body of more recent supervisor development research to be identified–that I cast my review net wide. I defined “psychotherapy supervisor” as a mental health professional who provides psychological treatment supervision. I hoped to include supervision studies across the various disciplines (psychology, psychiatry, social work, psychiatric nursing, and counseling) where psychological treatment supervision has long been considered important. I defined psychotherapy or psychological treatment supervision as: “[A] distinct professional activity” (Falender & Shafranske, 2004, p. 3) or “intervention” (Bernard & Goodyear, 2009, p. 7) in which a senior professional (supervisor) serves as mentor or guide to a junior professional (supervisee) who is in the process of learning and practicing psychological treatment; its primary objective is enhancement of the supervisee’s professional functioning, it involves evaluation of that professional functioning by the supervisor, and it is a hierarchical monitoring process (supervisor to supervisee) that serves a protective function for both patients and profession (Milne, 2007; Thomas, 2010). I defined a “psychotherapy supervisor development study” as one in which supervisor development theory was specifically investigated in some respect, and research findings were then related back to and explained within the context of that body of theory.

Method

To identify articles for review, four steps were taken: (1) PsycInfo, MedLine, Education Research Complete, Digital Dissertations, and Google Scholar database searches were conducted using “supervisor development” as the key search words; (2) reference sections of identified studies were examined to further identify other appropriate articles for inclusion that might have been missed (“ancestry approach”; Cooper, 1989); (3) supervision journals or journals that publish some supervision material were examined for any recent articles that might have appeared; and (4) various supervision texts (e.g., Bernard & Goodyear, 2009; Falender & Shafranske, 2004; Hess, Hess, & Hess, 2008; Watkins, 1997a) were also examined to further find any other possible missed work. Because dissertations have proven to be a valuable source of supervisor development study (e.g., Majcher, 2001; Majcher & Daniluk, 2009; Pelling, 2001, 2008; Stevens, 1994; Stevens, Goodyear, & Robertson, 1997), I chose to include that database as a part of this review. Based on those steps, a total of 18 studies were identified–13 quantitative and mixed method investigations, 3 qualitative investigations, and 2 measurement studies. The review time period spanned from January, 1996 through mid-2011. Each article was reviewed to determine: setting/sample characteristics, measures used, analyses, procedure, findings/conclusions, and limitations/strengths. Tables 2, 3, and 4, respectively, provide a summary of the features of those quantitative/mixed method, qualitative, and measurement studies.

Results

Results are presented that address general study characteristics, quantitative and mixed method studies, qualitative studies, and measurement studies.

General Study Characteristics

A wide range of helping professionals was involved in this body of research: Psychologists, psychiatrists, social workers, and counselors or students being trained for one or more of those professions. The predominant groups to receive research attention, however, were from the disciplines of psychology and counseling. Some primary features of the samples were: (1) approximately two-thirds of the participants were female (1,918 women versus 1, 139 men); (2) 83% of the participants were Caucasian; and (3) ages ranged from twenty-eight to fifty-two years. Some primary features of the studies were:

Quantitative/Mixed Method Studies

Data from these 13 studies suggested that:

Qualitative Studies

Data from these three studies (Majcher & Daniluk, 2009; Nelson, Oliver, & Capps, 2006; Rapisarda, Desmond, & Nelson, 2011) suggested that: (1) the process of supervisor development involved movement—from a place of uncertainty, ambiguity, anxiety, and limited confidence—to a place where greater certainty, clarity, emotional comfort, and confidence were preponderant; (2) issues with which supervisors tended to struggle in the development process surrounded matters of competence, confidence, role definition/clarification/identification, boundary clarification, and personal and professional integration; and (3) these qualitatively-identified process and issues show some similarity to the process and issues incorporated into supervisor development models. Limitations included small sample size and participants being drawn from a single doctoral program.

Measurement Studies

Data from these two studies (Barnes & Moon, 2006; Hillman, McPherson, Swank, & Watkins, 1998 [drawn from Hillman, 1996]) suggested that: The Psychotherapy Supervisor Development Scale (PSDS; Watkins, Schneider, Haynes, & Neiberding, 1995) appears to possess some reliability and validity as a supervisor development measure. Limitations included self-report survey methodology and possible social desirability effect.

Discussion

In their recent research review, Inman and Ladany (2008) indicated that focus on “the supervisory role is lagging” (p. 501) … “supervisory experiences continue to receive little [research] attention” (p. 511). The primary problem with that unfortunate fact, as they noted, is that “the supervisor is integral to counselor [and therapist] development” (p. 511). We empirically know the least about the party who may exert the most substantial impact on supervisees’ therapeutic development and actualization.

Supervisor development models were, perhaps, created in an effort to offset that trend; they were attempts to shine a light on the inner workings of the supervisor “in process” and use that as a fulcrum to consider how supervision was accordingly affected. But as Inman and Ladany’s (2008) conclusions suggest, the promise of supervisor development models has yet to be fulfilled. As we reflect on these 18 studies and the earlier 15-year period of theory/research, what conclusions can we draw about the contemporary status of psychotherapy supervisor development theory, research, and practice? How might we use those conclusions to point us forward in our pursuit to better understand the “integral supervisor”?

Thoughts on theory

Perhaps the strongest, safest theoretical statement that can be made about these five supervisor development models (Table 1) is: They provide a somewhat general, global feel for how supervisor development may unfold over time. The models seemingly have their clearest, most identifiable distinctions at the beginning and end points; it is the middle parts that become a bit more fuzzy and difficult to track. Supervisor development theory is still very much stuck in the ’80s and ’90s; the limitations, imperfections, or deficiencies that existed then still exist now, and no efforts have been made to remedy those. Our most meaningful explications of supervisor development thus far may well have come from studying the clarity that supervisors gain and the growth that they experience with regard to crucial supervision issues and themes over time (e.g., role definition and identification, boundary clarification, professional integration; cf. Majcher & Daniluk, 2009; Kurdt, 2001; Nelson et al., 2006; White, 1997). In my view, that empirical focus on the continuua of supervisor clarity (e.g., expanding self-awareness) and growth (e.g., “growing” skills)—and how those can change across manifold supervisory issues/themes—has yielded some interesting, instructive theoretical insights about supervisor development that we have not had before. While those explanations could be placed within a stage framework, I am not sure what that necessarily adds; they seem quite meaningful without any reference to or even consideration of stages whatsoever.

These five models do appear to provide some useful perspective for thinking about the potential mechanisms by which supervisor development is thought to be actuated. From my reading and reflection, there seem to be at least four particular mechanisms that emerge as critical and pivotal for change and growth across models: (1) sufficient interest in being and desire to improve as a supervisor; (2) openness to one’s supervisory self-experiencing; (3) capacity for and willing embrace of supervisory self-examination and self-reflection; and (4) action, practice, and experimentation (e.g., with regard to supervision skills, strategy deployment, and alliance formation). Whatever the model, those elements seem to be both the prerequisites and lubricant for the initiation and maintenance of any sort of supervisor development process. In my clinical experience, those variables remain every bit as significant today in the “making” of psychotherapy supervisors. While it is vitally important that we better understand the psychotherapy supervisor development process, it also seems well worth taking note of those variables that make that very process possible (or not). These models, however imperfect they may be, still give us some useful ideas about what must be sine qua non for supervisor growth to occur.

What changes to supervisor development theory might be needed? Because supervisor development theory has not evolved, the answer to that question really seems no different now than it would have been 15 to 20 years ago. For example, we still lack attention to critical incidents in supervisor development, the impact of various personality and other moderator variables upon the growth process, and a transition theory that guides our thinking about supervisor movement (forward and backward). If any sort of developmental theory is to productively be our guide for research, such missing elements will need to be incorporated into future theoretical refinements or revisioning. We need to move beyond the “general,” “imprecise,” and “rudimentary” that still seem to characterize far too much of our stage-theory conceptualizations. Even if stage theory itself is ultimately abandoned (e.g., for a more continuous conceptual framework), supervisor development theory requires a level of much greater precision, definition, and specificity than has previously been the case for advancement to occur.

Thoughts on Research and Measurement

These most recent 18 studies (1996 through mid 2011) were all driven in some way by supervisor development theory. Study results have tended to give some support to two fundamental developmental propositions: (1) in the supervisor development experience, supervisors proceed through some type of progressively unfolding supervisor growth process—where they seemingly move from a place characterized by less to more supervisory clarity, definition, identity, and skill; and (2) in the supervisor development experience, supervisors often tend to struggle with some of the same key developmental issues (e.g., gaining a sense of competence) during that growth process. While offering some support for a developmental view, this body of studies, however, remains quite limited in number, and any conclusions drawn from them would perhaps best be considered suggestive. No supervisor development theory has emerged empirically unscathed (e.g., Stevens et al., 1997; White, 1997).

Dissertations have been a ready source of supervisor development study, with some high quality unpublished and publication-bound research being done (e.g., Bencivenne, 1999; Majcher, 2001; Majcher & Daniluk, 2009). Approximately two-thirds of the 18 studies were dissertation projects, with about half of those being published in peer-reviewed journals (e.g., Barnes & Moon, 2006 [drawn from Barnes’ database, 2002]; Pelling, 2008). But research on this topic by more senior professionals—to use the words of Bernard and Goodyear (2009)—“is virtually nonexistent.” The supervisor development theory progenitors contributed almost no research on this subject and their efforts were far more conceptual than empirical. Neither Alonso nor Hess, who sadly are no longer with us, ever conducted any research on supervisor development; their first supervisor development theoretical contributions were essentially their last. Much the same could be said for Rodenhauser, who—after the horrific onslaught and aftermath of Hurricane Katrina—left psychiatry at Tulane University to become a full-time artist in Albuquerque. I only conducted one study on supervisor development (Watkins et al., 1995) and last wrote on the topic about 15 years ago (Watkins, 1997b). While Stoltenberg has conducted research on supervisee development (see Stoltenberg, 2005; Stoltenberg & McNeil, 2009), I am not aware that he has ever turned an empirical eye to supervisor development. None of those theory progenitors took up the research mantle, and no one has since stepped in to fill that research void. Psychotherapy supervisor development, as an area of empirical inquiry, has lacked for and has never really had any research leadership.

While dissertation contributions are helpful and suggest that study of supervisor development is possible, these studies as a group have produced nothing but “one and done” efforts: Not one of these dissertation contributors has ever carried out another investigation of psychotherapy supervisor development. But research beyond dissertations (e.g., Baker et al, 2002; Nelson et al., 2006; Ronnestad et al., 1997) fares no better: None of these contributors has carried out any supervisor development study beyond their original investigation either (more “one and done” efforts). There has been an absolute absence of any systematic or programmatic inquiry into this subject; unfortunately, when the totality of supervision research is considered, that, perhaps, should not come as a big surprise. Systematic programs of research in supervision have been and tend to be quite few in number (Ellis, 2010; Ellis & Ladany, 1997; Inman & Ladany, 2008). Furthermore, supervisor development—because of its focus on change and growth over time—can be a highly challenging and difficult area to study. While this body of “one and done” efforts (dissertation and non-dissertation) has provided some useful contributions, still movement beyond the one-shot study would seem sorely needed if knowledge and understanding of supervisor development is ever to advance most meaningfully; that at least would be a reasonable, though not necessarily easily achieved, desideratum to pursue as we think about trying to push this line of inquiry forward.

In further considering the challenge and difficulty of researching psychotherapy supervisor development, this body of empirical work seemingly brings into focus yet another crucial reality: That the missing element in supervisor development study thus far has largely been development itself. The majority of investigation has been cross-sectional, ex post facto, one-time surveys. Twenty-five years ago, Holloway (1987) stated that “At present, the most obvious problem in supervision research is the absence of longitudinal data to investigate developmental change” (p. 213). Her statement largely appears no less true today. Development, or the examination of change and growth over time, has only been researched in six supervisor development studies—with the time period being but 4 months in 3 studies (Baker et al., 2002; Rapisarda et al., 2011; White, 1997), 8 months in 1 study (Majcher & Daniluk, 2009), and a year and a half in 2 studies (Nelson et al., 2006; Ybrandt & Armelius, 2009). Ybrandt and Armelius (2009) conducted the only developmental study involving seasoned professionals, whereas all other investigations used doctoral students as subjects.

How supervisors might develop and grow beyond that four-month to year-and-a-half time period remains a mystery. Longitudinal study by definition tends to be a time-consuming, often demanding, even daunting process, and psychotherapy supervisors (whether experienced professionals or advanced doctoral students) can be a far less available participant pool for study than beginning supervisees or their psychotherapy patients.

When you combine those two realities, that limited research attention to actual supervisor development over time should come as no great surprise either and, again, only serves to further accentuate the often problematic nature of investigating this subject area.

Nevertheless, each of these studies—qualitative and/or quantitative in design and execution—proves quite instructive and informative in how to approach the project of supervisor development, and all are well worth reading for that reason alone. (White’s [1997] dissertation, which weighs in at a staggering 437 pages, is really a hefty book in and of itself—but a most interesting read, nonetheless.) While both quantitative and qualitative investigation seem needed and eminently possible in psychotherapy supervisor development, Majcher and Daniluk (2009) recently indicated that qualitative methodologies might be most useful for inquiry now because supervisor development investigation is at such an early stage of empirical scrutiny. Perhaps, through building a more substantive base of knowledge by means of qualitative study, we would then be placed in a much better position to proceed quantitatively. As we consider how best to approach study of supervisor development, Majcher and Daniluk’s (2009) thoughts at least seem well worth weighing in that regard.

From my analysis, the matter of measurement may also provide yet another reason to further consider the potential viability and advisability of taking a qualitative approach first at this early juncture. After all, what valid, reliable measures do we actually have to now study the construct of psychotherapy supervisor development? The Psychotherapy Supervisor Development Scale (PSDS; Watkins et al., 1995) has been the primary means by which supervisor development has been measured quantitatively. For the 13 quantitative/mixed method studies in Table 2, and 2 measurement studies in Table 4, 11 of the 15 used the PSDS. Yet in many respects, this measure has yet to be proven. What we know about the PSDS is:

Further validation efforts clearly seem needed if this measure is to be confidently used in supervisor development research; the data about the PSDS present a somewhat mixed picture. Furthermore, if this scale is to be used to assess supervisor development, it seems supremely important to keep in mind what it provides: At its best, the PSDS is a measure of self-perceived supervisor development and nothing more. While that perceptual variable surely has a place in research study, the extent to which (or even if) perceived supervisor development relates to actual supervisor development is a question without empirical answer at the moment. Until we have a more solid supervisor development research measure to use, that question will unfortunately have to wait quantitatively.

Perhaps if there is one preeminent obstacle to progress to be identified here, it would have to be measurement. Our research is only as good as our research measures, and in psychotherapy supervisor development, we are left with little upon which to draw. While measurement has been and continues to be a serious problem in supervision research generally (Ellis, D’Iuso, &Ladany, 2008; Ellis & Ladany, 1997; Watkins, 1998, 2011), it would seem to be particularly acute in supervisor development. Until some remedy is forthcoming, quantitative study may well be stymied. Until that remedy arrives, and it is something on which we can be now working, the recommendation of Majcher and Daniluk (2009) for qualitative study of supervisor development seems all the more apropos.

As a final limitation to mention about this body of research, the vast majority of the studied supervisors was Caucasian (83%). We know little, then, about supervisors of other ethnicities. If study of supervisor development is to advance, efforts to be more inclusive with regard to diversity would also seem critically important.

Thoughts on Practice and Education

In the learning and practice of psychotherapy supervision, data over the past 15 years have seemed to converge with relative consistency on the following point: Training in supervision matters (Baker et al. 2002; Bencivenne, 1999; Lyon et al., 2008; Pelling, 2008; Stevens et al., 1997; Vidlak, 2002; Vieceli, 2006; Ybrandt & Armelius, 2009). Not that long ago, supervision training was far more the rarity than reality. Hoffman (1994) even once referred to the lack of supervision training as the mental health profession’s “dirty little secret.” But considerable change in thinking seems to have occurred about the importance of supervision training since the mid-’90s, and we now see such training being more readily available, required, or at least recommended across the disciplines of psychology, psychiatry, psychiatric nursing, social work, and psychoanalysis (e.g., American Psychological Association, 2009; Cutcliffe, Hyrkas, & Fowler, 2011; Munson, 2001; Pegeron, 2008; Riess & Herman, 2008; Szecsody, 2008; Whitman, Ryan, & Rubenstein, 2001). Data from these supervisor development studies seem to reflect nicely and be supportive of that shift in thinking about supervision.

But why should supervision training really make that much difference? In his now classic review, Worthington (1987) brought to light the possibility that “[s]upervisors do not become more competent as they gain experience” (p. 203). As he stated,

That conclusion, which was based on only four studies at the time, has tended to be confirmed by a number of subsequent research investigations. It indeed seems increasingly likely, as Milne and James (2002) have indicated, that “[c]ompetence in supervision appears to require training” (p. 55), and recent research about the potential benefits of structured supervisor training programs seems to offer some support for their statement as well (see Milne, 2010; Sundin, Ogren, & Boethius, 2008). We seem to have arrived ever more so at the conviction that “… experience, without formal training in supervision, is insufficient as a basis for reliably producing those behaviors and values … that … will enhance the supervisory relationship” (Stevens et al., 1997, p. 87).

As a part of the supervision training experience, the supervision of supervision process has also been brought into focus in some of these reviewed studies (Baker et al., 2002; Lyons et al., 2008; Majcher & Daniluk, 2009; Nelson et al., 2006; Vidlak, 2002; Ybrandt & Armelius, 2009). The results have appeared to offer some support for the potential complementary value of supervisor trainees having their work supervised. For decades now, professional opinion has tended to converge on the importance of both didactic and experiential components in the training of psychotherapy supervisors (see Borders, 2010; Falender et al., 2004; Hoffman, 1994; Kaiser & Kuechler, 2008; Russell & Petrie, 1994; Stoltenberg & Delworth, 1987; Stoltenberg & McNeil, 2009; Watkins, 1992; Whitman et al., 2001). But the supervision of supervision process has often seemed to be shortchanged.

Almost 20 years ago, Ellis and Douce (1994) summed up the state of supervisor supervision as follows: “The scarcity of information regarding the practice of supervising supervisor trainees is particularly apparent … only a handful of published articles address the topic … and none explicitly detail the practices involved” (p. 520). Watkins (1995b) concurred: “… the supervision of supervision process … is something about which virtually nothing is known … (p. 677). Rodenhauser (1997) opined shortly thereafter that “[s]upervision of psychotherapy supervisors is uncommon” (italics in original; p. 539). While those statements still largely ring true today, there at least appears to be increasing recognition of the reality that—just as psychotherapists in training can benefit from and need their psychotherapy supervised—supervisors in training can benefit from and need their supervision supervised as well. Across disciplines, we seem to be increasingly moving toward an expanding professional vision of effective supervision training—characterized by a mixture of organized, systematic coursework and/or seminars and sustained practical supervisor supervision experiences. In my view, that vision is a welcome change in perspective that ultimately can have immensely positive ramifications for supervisor trainees, their supervisees, and the patients that they serve. Perhaps the reality of supervision training opportunities will come to eventually match our rhetoric about need for supervision training and supervisor supervision. That at least seems to be a worthy objective which we can continue to vigorously pursue in the decades that lie ahead.

With the importance of supervisor training/supervision increasingly recognized and embraced, it also seems worth emphasizing that: Supervisor learning, education, growth, and development should not stop after having received some training in the fundamentals of how to supervise (Goin, 2006). Ideally, supervisor development is an active, vital, lifelong educational process and commitment that is joyously and relentlessly pursued with vigor and determination. Ongoing supervisory growth is most apt to happen when ongoing deliberate efforts are made to challenge and cultivate supervision skills and perspectives over time. Supervisory skills and vision, when not subjected to such deliberate efforts, can indeed become mired in a complacency that is antipathetic to the very spirit of supervisor development.

Discussion Summary

In surveying this first generation of psychotherapy supervisor development theory, research/measurement, and practice/education, some of the primary points to emerge were:

Conclusion

The reach of the psychotherapy supervisor is broad, far, and affecting. In the supervision training experience, supervisors strive to enhance the learning and development of their supervisees and, accordingly, their supervisees’ patients. It is well recognized that the potential influence of the supervisor on the development of budding psychotherapists can be both considerable and enduring. But what do we really know about the actual development of psychotherapy supervisors themselves? In this paper, I have examined that question by reviewing the last generation of supervisor development theory, research, and practice. In my opinion, if we are to have the most complete and informed understanding of the supervision process, its outcome, and its participants, then we need a vision of supervision that gives voice to and incorporates the intersecting developmental processes, trajectories, and experiences of the supervisee and supervisor. Unfortunately, the developmental process, trajectory, and experiences of the supervisor have remained more mystery than manifest for far too long. As I have hoped to show here, that can change, and work done thus far nicely identifies needs that require redress and provides some constructive direction for remediation and advancement. From my perspective, the study of supervisor development—though still in its infancy—is an area punctuated by much promise and empirical possibility. The somewhat “barren scape” of supervisor development need not stand. The developmental interior of the psychotherapy supervisor has the potential to be charted and better understood, and our vision of the supervision experience has the possibility of being accordingly transformed.