changes in supervision as counselors and supervisors … supervision as experience is... · changes...

Changes in Supervision as Counselors and Supervisors GainExperience: A Review

Everett L. Worthington, Jr.Virginia Commonwealth University

Evidence bearing on whether and how counselors and supervisors receive or givedifferent types of supervision of psychotherapy as they each gain experience wasinvestigated. Theories describing changes in supervision of counselors as they gainexperience are reviewed. Most are similar to each other. They posit changes in thesupervisee, with supervision environments being matched to the changing needs of thesupervisee. There are three theories concerning how the supervisor changes as he or shegains experience. Findings from empirical studies are consistent with theories ofcounselor development but only weakly supportive of the theory that actual supervisionenvironments are matched to supervisee needs. Findings from empirical research onchanges in supervisors as they gain experience reveal few differences in supervisors atany level beyond the master’s degree.

Presumably, therapists and supervisors agelike wine. In this article, I sample their sounds,sights, bouquets, and tastes as they gain expe-rience. Supervision of prepracticum counselorsis not covered in this review (for a recent re-view, see Kurtz, Marshall, & Banspach, 1985).Rather, this article is a summary of supervisionof counselors from their first practicum andbeyond. First, the issue of what supervisionshould be is discussed. Then research onchanges in supervision as counselors gain expe-rience is reviewed. Last, research on changes insupervision as supervisors gain experience atsupervision is examined.

Theories of Changes in Supervision WithExperience

We assume (and hope) that counselors andsupervisors learn and improve as they gain ex-perience. As professional helpers who have in-vested time, energy, and professional identity in

the wine rack of learning helping skills, wepsychologists have a vested and emotional in-terest in finding that change occurs with expe-rience. Should we find that time and experienceturn our wine to vinegar, we would be incredi-bly threatened. In general, we assume thatchange occurs and ask how counselors and su-pervisors change as they gain experience. Thishas generated several conflictual issues thatmust be understood in order to understand theresearch on changes in supervision resultingfrom changes in experience.

Conflicts in Supervision Theory

One conflict is this: Should supervision beproactive or reactive? Some supervision isdriven by an agenda: Sessions are planned;goals are clearly identified; interventions areusually initiated by the supervisor or evenplanned before a supervision session. Anothertype of supervision is reactive: Goals are iden-tified, but the supervisor awaits critical inci-dents and intervenes when those incidents arise,not initiating his or her agenda. Should thissupervision style change as a counselor gainsexperience? In practice, most training of pre-practicum students and often students in earlypractica is more proactive than reactive. As thecounselor becomes more proficient, the super-visor becomes more reactive. Is this desirable?If we could identify the component skills ofadvanced counseling or even master-level coun-

EVERETT L. WORTHINGTON, JR., PhD, is Associate Pro-fessor of Psychology at Virginia Commonwealth Univer-sity. His current research activity includes supervision ofpsychotherapy, religious counseling, adolescent pregnancy,and other topics relating to marriage, family, and values.

CORRESPONDENCE CONCERNING THIS ARTICLE should beaddressed to Everett L. Worthington, Department of Psy-chology, Virginia Commonwealth University, 800 WestFranklin, Richmond, Virginia 23284-0001.

This article is a reprint from Professional Psychology:Research and Practice, 1987 Vol. 18, No. 3, 189–208.

Training and Education in Professional Psychology Copyright 2006 by the American Psychological Association2006, Vol. S, No. 2, 133–160 1931-3918/06/$12.00 DOI: 10.1037/1931-3918.S.2.133

133

Table 1Supervision of the Developing Counselor

Stage/stage of counselor Supervisor behaviors

Fleming (1953)

1. Imitative learning Anxious supervisees learn by imitating their supervisors, whogive suggestions and demonstrate ways of counseling

2. Corrective learning Supervisor corrects inaccurate interpretations and techniques(less support is necessary because counselor self-confidence is relatively high)

3. Creative learning Supervisor investigates the supervisee’s personal reactions tothe client and how these affect counseling

Grotjahn (1955)

1. Period of preparation Provide technical help and respect and encouragement,restrain the counselor from beginner mistakes (hastyreassurance and support)

2. Elaborate on therapist’s knowledge of client Help the therapist to understand the client’s personalitydynamics and psychopathology

3. Working through Help the therapist to understand and deal with own feelingsand conflicts that are part of the therapeutic process

Hogan (1964)

1. Dependent on supervisor: Neurosis bound, insecure,uninsightful, highly motivated, imitative

Tuition (identifying predictable outcomes), interpretation,support, awareness training, exemplification

2. Dependency-autonomy conflict: Struggle betweenoverconfidence and being overwhelmed, ambivalent,fluctuating motivation; personal therapy withsomeone other than supervisor is recommended

Support, ambivalence clarification, exemplification, sometuition

3. Conditional dependency: Increased professionalself-confidence; greater insight, especially aboutneurotic and healthy motivations; more stablemotivations

Sharing as peers, examplification, confrontation

4. Master psychologist: Personal autonomy,insightfulness with motivation, need forconfrontation

Sharing, confrontation, mutual consultation

Ard (1973)

1. Perceptorship: Student has need of orientation Orients beginning student2. Apprenticeship: Requests specific instruction Gives specific instructions3. Mentorship: Student demonstrates work and

wrestles with personal issuesCritiques the work of the supervisee, helps supervisee in self-

examination4. Sponsorship: Student is largely competent Instills further confidence in an already competent counselor5. Peership: Student has emerged from training to full

professional statusEstablishes coequal relationship after termination of formal

supervision

Gaoni & Neumann (1974)

1. Teacher-student stage2. Apprenticeship3. Developing the therapeutic personality4. Mutual consultation among equals

Littrell, Lee-Borden, & Lorenz (1979)

1. Initial Goal setting, clarification of nature of supervision and itscomponents

2. Counselor-therapeutic Supervisor acts as counselor to supervisee3. Teacher Supervisor teaches supervisee the skills of counseling4. Consultation Cooperation; supervisor and supervisee work together5. Self-supervising Supervisee systematically seeks to improve own counseling

through self-observation

134 WORTHINGTON

Table 1 (continued)


Stoltenberg (1981)a

1. Attempts to define boundaries between the“counselor” and the “person‘

Encourages autonomy within a normative structure;encourages risk taking, acts as teacher, integrates theoryand practice, does not answer all questions correctly,attends to the supervisee’s behavior in supervision as wellas in counseling, gives opportunities to observe thesupervisor

2. Begins to define own identity; not content toinitiate; experiments with different styles; begins todisagree with supervisor

High autonomy, low normative pressure; instructs the traineein new skills and gives advice when necessary

3. Increased empathy, more highly differentiatedinterpersonal orientation, no longer a disciple of anygiven technique, tolerant, can work with a varietyof clients

Increased emphasis on sharing; an exemplification by bothpartners, appropriate professional and personalconfrontation; supervisor might acknowledge ownweaknesses

4. Capable of independent practice, willfulinterdependence with others, integrates the standardsof the profession within a personal value system

Consultation given when deemed appropriate by thesupervisee

Hart (1982)

1. Didactic instruction by the supervisor, acquisition of the caseconceptualization and intervention skills

2. Additional work on supervisee skills but also supervisoryfeedback on supervisee therapeutic work, some work onpersonal awareness

3. Integration of skill development with personal awareness,especially as it applies to clients

Yogev (1982)

1. Role definition: Student acknowledges commitmentto becoming a therapist, demystifies therapy,experiences anxiety, feels inadequate, recognizessome strengths

Helps student to define role, clarifies expectations insupervision, evaluates the supervisee

2. Skill acquisition: Learning the skills of counseling Observation of student, possible cotherapy with thesupervisor

3. Solidification and evaluation of practice Uses both emotional aspects and didactic and skill-practiceaspects

Blount (1982)

1. Adequacy versus inadequacy Supportive relationship, awareness training, modeling,didactic skills instruction

2. Independence versus dependence Exemplification and integration of dynamics, advanced skilldevelopment

3. Conditional dependency versus individuation Greater autonomy and appropriate confrontation,encouragement of peer relationship

4. Professional integrity versus personal autonomy Shaping and collegial consultation, self-supervision,supervision of others, mentoring

Wiley (1982)

Expands on Stoltenberg’s model of counselor complexitybut identifies five critical issues that are behaviorallydefined for each of Stoltenberg’s four stages

Expands on Stoltenberg’s model of supervision environmentsbut identifies five critical issues that are behaviorallydefined for each of Stoltenberg’s four stages

Loganbill, Hardy, & Delworth (1982)

1. Stagnation: Naive unawareness, simplistic dualisticthinking; counseling may lack intensity; linearproblem solving, low self-concept, extremedependence on the supervisor

Supervisors are thought to assess and evaluate supervisee andto have five types of interventions: facilitative, confrontive,conceptual, prescriptive, and catalytic interventions

(table continues)

135CHANGES IN SUPERVISION

Table 1 (continued)


Loganbill, Hardy, & Delworth (1982) (continued)

2. Confusion: Instability, conflict, fluctuation betweenfeelings of incompetence and feelings of greatability; supervisor may be thought of as all-knowing or incompetent

All types of interventions may be appropriate for any stageon any issue

3. Integration: Calm reorganization, a refreezing ofattitudes, basic acceptance of self and of limitations;supervisor is seen as a realistic person with strongand weak areasb

R. Miller (1982)

1. Quiescence Interventions are represented along three continua: Intrusive-reflective, oppositional-supportive, and prescriptive-elicitive

2. Early exploration3. Imitation4. Partial autonomy5. Autonomy

Sansbury (1982)

1. Prepracticum: Basic listening skills and assimilationof the role of counselor

Evaluative feedback, needs assessment; models goodcounseling skills, reinforces and supports students

2. Practicum: Develops new therapeutic techniques;improves conceptualization, refines personal theory,develops competence, establishes limits ofresponsibility for self and client

Analyzes cases; helps resolve counselor-client impasses;promotes counselor understanding through confrontation,role reversals, interpretation, and feedback; teachessupervisees to ask for help in supervision

3. Internship: Broadens and refines understanding ofclients, learns types of clients that are best helped,examines personal issues, learns reliance on self

Confronts supervisee on differences in talk and behavior,supports increased risk taking, helps supervisee withpersonal issues, assists supervisee to self-evaluate

S. R. Friedlander, Dye, Costello, & Kobos (1984)

1. Ambiguity in supervisee Helps supervisee deal with demands for wide-rangingtolerance of ambiguity; emphasizes learning to learn

2. Recognition of limits of therapeutic conditions Helps supervisee to see difference in theory and practice;accepts mistakes and unanticipated client responses; helpsto deal with guilt over failures

3. Discovery of therapy as deep communication Helps to take focus off techniques and onto humanrelationships

4. Eclecticism in light of client’s needsc Helps to develop a repertoire of interventions; helps to applyon the basis of sensitive assessment of clients

Hess (1986)

1. Inception: Confusion, unanchored experience,anxiety, identity formation, adequacy versusinadequacy

Helps to identify experience with cognitive maps forhandling experience; encouragement, support, mutual trust

2. Skill development: Choice of theory, dependenceversus independence

Induces students to try out techniques; rehearses techniqueswith them; gives corrective feedback

3. Consolidation: Skills become “owned,” new skillsdevelop; may supervise less experienced colleagues

Helps students to learn new skills; encouragement

4. Mutuality: Establishes professional identity,conditional dependency versus individualism

Mentoring; collegial supervision; focuses on how therapists’personality affects the case; stages are “spiraled” (e.g.,repeated at different depths)

a Uses Hogan’s (1964) four stages (dependent, dependency-autonomy conflict, conditional dependency, master counselor)but expands on them.b There are thought to be eight critical issues: competence, emotional awareness, autonomy, theoretical identity, respect forindividual differences, purpose and direction, personal motivation, and professional ethics. For each issue, the superviseeis thought to negotiate the three stages listed.c All four issues may arise simultaneously but may be resolved in the order given.

136 WORTHINGTON

seling with the same precision with which wearticulate basic listening skills, would supervi-sion still largely be reactive at advanced coun-selor levels? Or does our inability to preciselyidentify higher order counseling skills shape ourtheory of supervision?

Another conflict in supervision is whether thecounselor should learn the theory of the super-visor or whether the supervisor should adapt hisor her methods to work within the theoreticalframework of the supervisee. A supervisor’sposition on this issue may change as the super-visee gains experience. Beginning traineesmight be taught a doctrinaire way of counseling,whereas interns and postdoctoral superviseesare allowed more freedom. An exception to thisgeneralization is when the supervisee explicitlyasks to learn the counseling style of the super-visor (e.g., learning strategic family therapyfrom Jay Haley, or Paul Wachtel’s learningbehavioral counseling from Gerald Davison).Even in these exceptions, though, what istaught, how fast it is taught, and what is as-sumed to be known differs with the experiencelevel of the counselor.

Whether and how supervision changes ascounselors gain experience also depends on thesupervisor’s beliefs about supervision andcounseling (Bartlett, Goodyear, & Bradley,1983). Some models of supervision are basedon counseling theories, which are adapted andgenerally used consistently with supervisees.The component parts of the counseling theoryare identified and taught to the supervisees, andthe methods by which they are taught are oftensome of the same methods that are used withclients within counseling. Supervision changeswith skill level of the therapist, but changesmight be more content specific than processspecific. A second approach to supervision isnot based on a specific counseling theory.Counselors are thought to change in needs andabilities as they gain counseling experience. It isassumed that supervision should be matched tothe level of the counselor. According to thisdevelopmental model, how a supervisor inter-venes is determined by the supervisor’s view ofthe stages of counseling, by the experience thatthe supervisor has with counselors of varyingexperience, and by the supervisor’s assessmentof the level of both performance and knowledgeof the supervisee. The content of the supervis-

ee’s theory is generally irrelevant in this devel-opmental approach.

Supervision of the Developing Counselor

Much of the research on how supervisionchanges with time has been done by develop-mental theorists who have described supervi-sion apart from the supervisor’s theory of coun-seling. Several such theories have been articu-lated (see Table 1 for a summary). Generally, inthese theories an implicit stage theory of coun-selor development is assumed, and supervisorybehaviors that are thought to be consistent withthe hypothesized level of development of thecounselor are specified. What is surprising isthat few researchers have directly investigatedthe actual development of counselors as theygain supervised and unsupervised experience.One notable exception is Hill, Charles, andReed (1981), who found that counselors in theirdoctoral program in psychology (at the Univer-sity of Maryland) progressed through fourstages. The first stage involved self-conscious-ness of the counselor and attention to their in-ternal experiences, sometimes to the exclusionof understanding the clients’ experiences. Dur-ing a second stage, counselors adopted a coun-selor’s stance in which they used some standardapproach to counseling. Therapeutic failureswere explained in terms of the counselors’ in-ability to execute counseling skills prescribedby the theory. Application of the theory wasgenerally rigid; clients were fit into a procrust-ean bed of prescribed theory. As the counselorgained expertise in the chosen model, however,the occasional therapeutic failures did not stop,though they decreased in numbers. These anom-alies, to use a Kuhnian term, generally intro-duced uncertainty into the counselor’s theoriz-ing. There occurred a third stage, whichspanned late practicum, internship, and early(and sometimes later) professional experience,during which the counselor became confusedover the anomalies and rejected the originaltheory in favor of eclecticism. Sometimes thecounseling theory that guided their practice wasleft unarticulated. At other times theoreticalpropositions were inconsistent and contradic-tory. If counselors progressed beyond this stage,they articulated a reasonably clear and inter-nally consistent personal theory of counselingand behaved consistently with that articulation.


This empirically derived model of counselordevelopment is similar to the theorizing ofHogan (1964). Hogan hypothesized four stagesof counselor development. Beginning counsel-ors were thought to be insecure, neurosis bound,and uninsightful; second-stage counselors werethought to struggle with dependency-autonomyconflicts and to have supervision relationshipscharacterized by ambivalence; in the third stage,the trainee was thought to gain self-confidenceand to evidence more stable motivation; and thefourth stage was termed master psychologistand was hypothesized to be characterized bypersonal autonomy and self-assurance.

Hogan (1964) proposed that ideal supervisionenvironments promote counselor growth ineach stage. If the environment was matched tothe developmental needs of the counselor, thenoptimal growth and improvement of the super-visee should result. Hogan’s operationalizationof supervision behaviors ideal for such stage ofsupervisee development was skimpy in compar-son to the specification of the supervisee’s de-velopment. Nonetheless, this seminal paper pro-voked both subsequent theorizing (Stoltenberg,1981) and research (Reising & Daniels, 1983).

Stoltenberg (1981) expanded Hogan’s (1964)model, terming his elaboration the counselorcomplexity model. As the counselor develops,he or she is thought to become more cognitivelycomplex. Stoltenberg accepted Hogan’s specu-lations about the development of counselors andmore carefully described how supervisors mightcreate growth-producing environments. Duringthe first stage, the supervisee is provided with astructure and encouraged to develop autonomywithin the safety of the structure. In the secondstage, the supervisor deals with identity issuesby offering new skills and advice from whichthe supervisee can choose. In the third stage,increased sharing and collegiality exist, and per-sonal confrontation is sometimes sought andgiven. In the master counselor stage, consulta-tion is given when sought.

Loganbill, Hardy, and Delworth (1982) iden-tified three stages of counselor development:stagnation (naive unawareness), confusion, andintegration. These stages are similar to stagesidentified by Hill et al. (1981) as well as byHogan (1964) and Stoltenberg (1981). How-ever, Loganbill et al.’s (1982) added twist wasthat counselors need to resolve eight criticalissues before becoming master counselors:

competence, emotional awareness, autonomy,theoretical identity, respect for individual dif-ferences, purpose and direction, personal moti-vation, and professional ethics. The counselor isthought to resolve the issues independently ofeach other. Thus the counselor could be in anyof the three stages (stagnation, confusion, orintegration) with any issue. The supervisor’stask is to assess the level of development on theissue that is being dealt with in supervision andto promote growth to the next level of develop-ment on that issue. Loganbill et al. thus pro-posed the most flexible theory of the develop-ment of the counselor. Hess (1980) was simi-larly flexible within a less completelydeveloped theory. He proposed that develop-ment of the counselor occurs in spirals of in-creasing complexity rather than linearly.

Extant stage theories of supervision are anal-ogous to California wines, showing nuances inquality and subtlety but often being quite simi-lar to each other. There are some deficiencies inthe current theories. For example, each theoryof supervision depends on a picture of counselordevelopment that is clear in what it says but ispainted with broad brush strokes. From afar theshapes are noticeable but on closer inspectionthere are no details. The descriptions of coun-selor development rest on scant research. Thereis no specification of what higher order coun-seling skills are or when each level of counsel-ing skill might come into ascendency. For ex-ample, it is generally agreed that basic listeningskills are the building blocks of therapy and thatadvanced empathy, confrontation, focusing,conceptualization, and intervention skills arenecessary to good counseling. However, whenare counselors most ready to learn conceptual-ization? When and how does the counselorshow readiness to learn how to use the concep-tualization arrived at in the supervision sessionor in private thought to help the client adopt thatconceptualization as his or her own? When doesthe counselor learn to deal with defensivenessand resistance? Obviously, all of these skills areneeded with most clients, but counselors in theirfirst practicum cannot learn everything. Con-cepts are postponed until less self-conscious-ness and anxiety inhibit the counselor. A spec-ification is needed for how and when the super-visor can tell that the counselor is receptive tolearning.

138 WORTHINGTON

Another deficiency with current developmen-tal stage theories of supervision is that they areprimarily stage theories rather than theories ofhow transitions take place between stages. Theyspecify—albeit broadly—what the counselorand the supervisor experience and do duringeach stage. But how does the supervisor pro-mote movement from one stage to another? Atransition theory of counselor development andconsequent supervisor behavior is sorelyneeded.

Supervision by the Developing Supervisor

Few theorists have addressed how supervi-sors change as they gain experience. Becausethere has been little attention to this problem,research on the issue has been helter-skelter.

Bernard (1979, 1981, 1982) described howsupervisors can be trained to increase theirawareness of options during supervision. Sheoutlined a training method that is equally appli-cable to beginning and experienced supervisors.Three segments constitute her 16-hr trainingmodule: baseline data gathering, exposure tomodels of supervision, and evaluation plus eth-ical dilemmas. In the first segment, supervisorsidentify the focus of the supervision that theyhave audio- or videotaped. Three areas of focusare defined: process skills, which include howthe counselor behaves during counseling; con-ceptualization skills, which are concerned withhow the counselor thinks about counseling; andpersonal skills, which include how the coun-selor reacts to counseling. Supervisors deter-mine the approximate weighting that they cur-rently give to each focus. Three roles are alsoidentified: teacher, consultant, and counselor.Bernard (1981, 1982) reported that supervisorsoften become aware of a discrepancy betweenintention and performance. The most commondiscrepancy is the case of the supervisor whothinks that he or she uses the counselor rolemost often but finds, upon objective self-observation, that he or she usually uses theteacher role (see also Hess & Hess, 1983). In thesecond segment, Bernard broadened the theo-retical horizons of the supervisor. She explainedfour approaches to supervision and had the su-pervisors use each model briefly: Bernar’s(1979) discrimination model; Interpersonal Pro-cess Recall (Kagan, 1980); microtraining (For-syth & Ivey, 1980); and live supervision (con-

trasting with retroactive supervision rather thandead supervision). In the third part of her pro-gram, Bernard discussed evaluation in supervi-sion, showing supervisors the difference be-tween pinpointing evaluation issues and com-municating the issues to the supervisee (seeTyler & Weaver, 1981, for a discussion). Shediscussed ethical behavior with the use of vi-gnettes that pose issues and require decisionsthat have generally not been thought through bythe supervisors. Bernard’s training is system-atic, and she reported that it was useful both forthe neophyte and the experienced supervisor.She found that the experienced supervisors gen-erally show more responsivity to the materialthan the inexperienced supervisors.

Alonso (1983) proposed a different type oftheory of supervisor development. She consid-ered how a supervisor might change throughoutthe entire professional life cycle, moving fromnovice to midcareer to late-career concerns. Ateach stage in the career of the supervisor, thesupervisor must wrestle with three issues: selfand identity, the relationship between therapistand supervisor, and administration. The super-visor in each of the stages resolves the issuesdifferently because the demands of life and pro-fessions differ at each life stage.

Hess (1986) noted the paucity of theories ofdevelopment of the supervisor and has proposeda three-stage model of supervisor development.In the beginning stage, the new PhD assumesthe mantle of the supervisor by virtue of grad-uation rather than by training or experience. Thenew supervisor must therefore deal with the rolestatus change from trainee to trainer, superviseeto supervisor. Because only about one third ofinterns receive training in supervision (Hess &Hess, 1983), lack of awareness of the structureof supervision and the techniques of supervisionmake the supervisor sensitive to the criticism ofpeers and students and often promote self-consciousness. The new supervisor often copeswith the self-doubts and ambiguity by adoptinga concrete structure for supervision and focus-ing on techniques of counseling or on clientdiagnosis.

In the second stage, exploration, the supervi-sor has gained confidence and competence andis often able to baffle and amaze the superviseewith feats of apparent psychological legerde-main. Supervision is accepted as a professionalactivity of value, and the supervisor’s enthusi-


asm promotes increased student interest incounseling. Two pitfalls are common to thesecond stage of supervisor development: givingsupervision that is too restrictive or toointrusive.

The third stage of development is character-ized by continued and increased respect fromstudents and respect for students. More atten-tion is given to the student’s learning agenda,and more relationship per se occurs (rather thancognitive attention to relationship). Supervisorsare sought because they are perceived to beexcellent teachers of psychotherapy. Gratifica-tion is achieved when the supervisee excelsrather than when the supervisor is recognized asbeing a good supervisor.

Despite the beginning of theorizing aboutthe professional development of the supervi-sor, the field is at a rudimentary level, likegrapes fermenting in the sun in comparison tomaking California wines (i.e., theories aboutthe development of supervisors). Missing isany explication about (a) how supervisorsmight behave at different levels of develop-ment, (b) differentiation between develop-ment of counseling skills and supervisionskills, (c) whether continuing to counsel cli-ents impedes, accelerates, or does not affectsupervision competence (and if it affects it,how), (d) the ways in which supervisors learntheir trade, (e) how a supervisor develops andmodifies his or her theory or model of super-vision with experience, (f) what types of ex-periences and critical incidents help supervi-sors to improve, (g) what might impede thedevelopment of a supervisor, (h) how otherprofessional experiences dovetail with devel-opment of supervision skills, and (i) whethera supervisor can provoke critical incidents insupervision to help the supervisee improve.An explicit, testable theory of the developingsupervisor is needed in order to drive andfocus research concerning the supervisor. Thetheory must transcend the descriptive modelscurrently available and identify theoreticalvariables of import within the larger frame-work of psychology. Whatever the thrust ofthe theory, new thinking is needed in order tofurther the understanding of the supervisorand how he or she can help people to learnpsychotherapy.

Supervision of the Developing Counselor:Empirical Research

Historical Perspective

The types of training that counselors withdifferent levels of experience receive haschanged over time (Leddick & Bernard, 1980).Early in the history of supervision, psychoanal-ysis dominated the field and supervisees under-went training analysis, presumably learningpsychotherapy skills through experiencing therole of client and through observing the traininganalyst at work. Later, it was thought that teach-ing of theories of therapy and personality de-velopment occurred in the classroom, whereastraining in counseling occurred at practicumsites or counseling agencies. As Carl Rogers,Robert Carkhuff, Charles Truax, Allen Ivey,Gerard Egan, Steven Danish, and others devel-oped technologies of training, though, skilltraining began to occur earlier in programs thattrained therapists. Currently, counselors are ex-pected to enter their first counseling practicumalready proficient at beginning counselingskills. Technologies continue to advance. Thereis increasing use of videotaping, bugs in the ear,and even computer simulation (Phillips, 1984).The component skills of advanced psychother-apy are being identified and studied, and pro-cess models are being explicated. With increas-ing sophistication, we researchers might expectthe types of training offered to therapists tocontinue their evolution.

Methodology

In general, studies of supervision as counsel-ors gain experience have been of two types:investigations of one level of counselor (begin-ning counselors or post-master’s counselors)and investigations of several levels of counselorsimultaneously (see Table 2 for a summary ofthe method and findings arranged by thosecategories).

Samples for four of the six studies of begin-ning counselors were drawn from the Universityof Missouri. The other four studies were drawnfrom one practicum site also. This is a seriousrestriction in generalizing the results becauseWorthington (1984a) found that the supervisiongiven at different university counseling centers

(text continues on page 149)

140 WORTHINGTON

Tab

le2

Stud

ies

ofSu

perv

isio

nof

Beg

inni

ngC

ouns

elor

s,P

ost-

MS

Cou

nsel

ors,

and

Cou

nsel

ors

atSe

vera

lL

evel

sof

Exp

erie

nce

Aut

hors

Sam

ple

Supe

rvis

ors

Supe

rvis

ees

Des

ign

Inst

rum

ent

The

ory

Sele

cted

findi

ngs

Beg

inni

ngco

unse

lors

Wor

thin

gton

&R

oehl

ke(1

979)

Uni

vers

ityco

unse

ling

cent

er,

mid

wes

t14

inte

rns,

2po

st-

PhD

s31

begi

nnin

gpr

actic

umst

uden

ts

Cor

rela

tiona

lSu

perv

isio

nQ

uest

ionn

aire

Em

piri

cal

anal

ysis

ofpe

rcei

ved

supe

rvis

orbe

havi

ors

1.Su

perv

isor

sth

ough

tgi

ving

feed

back

abou

tev

alua

ting

and

mon

itori

ngpe

rfor

man

cew

asm

ost

impo

rtan

tsu

perv

isio

nac

tivity

.Sup

port

and

teac

hing

wer

eno

tco

nsid

ered

tobe

impo

rtan

t2.

Supe

rvis

ees

perc

eive

dsu

ppor

tan

dte

achi

ngto

bem

ost

impo

rtan

tin

begi

nnin

gpr

actic

um3.

Beg

inni

ngpr

actic

umst

uden

tsas

ked

for

feed

back

(pos

itive

and

nega

tive)

,bu

tne

gativ

efe

edba

ckw

asno

tre

late

dto

actu

alsa

tisfa

ctio

n

Hep

pner

&H

andl

ey(1

981)

One

mid

wes

tern

univ

ersi

ty,

four

coun

selin

gpr

actic

a

20do

ctor

alst

uden

ts33

begi

nnin

gM

Sst

uden

tsC

orre

latio

nal,

self

-rep

ort

ques

tionn

aire

sC

RF-

SRF;

Bar

rell-

Len

nard

Rel

atio

nshi

pIn

vent

ory.

Five

item

s:2

satis

fact

ion,

3im

pact

Soci

alin

fluen

ce1.

Supe

rvis

eepe

rcei

ved

supe

rvis

oras

trus

t-w

orth

y,at

trac

tive,

and

expe

rt2.

Mos

tsu

perv

isee

ssa

tisfie

dw

ithsu

perv

isio

n,bu

tno

tm

uch

influ

ence

dby

supe

rvis

or3.

Supe

rvis

orpe

rcei

ved

little

impa

ct,

too

4.R

elat

ions

hip

char

acte

rist

ics

ofsu

perv

isor

sre

late

dto

satis

fact

ion

but

not

toim

pact

5.Fi

ndin

gsdo

not

supp

ort

Stro

ng’s

(196

8)tw

o-st

age

mod

elof

influ

ence

Hep

pner

&H

andl

ey(1

982)

One

mid

wes

tern

univ

ersi

ty20

doct

oral

stud

ents

33be

ginn

ing

MS

stud

ents

Cor

rela

tiona

l;tw

ose

lf-

repo

rtqu

estio

nnai

res

CR

F-SR

FSu

perv

isio

nQ

uest

ionn

aire

:17

item

s;Su

perv

isio

nQ

uest

ionn

aire

(sup

ervi

sor)

:17

item

s

Soci

alin

fluen

ce1.

Eva

luat

ion

bysu

perv

isor

sre

late

dto

supe

rvis

ees’

perc

eptio

nof

them

asex

pert

,at

trac

tive,

and

trus

twor

thy

2.T

rust

wor

thy

and

attr

activ

ew

ere

rate

dm

ore

impo

rtan

tth

anex

peri

ence

(tab

leco

ntin

ues)

Tab

le2

(con

tinu

ed)

Aut

hors

Sam

ple

Supe

rvis

ors

Supe

rvis

ees

Des

ign

Inst

rum

ent

The

ory

Sele

cted

findi

ngs

Beg

inni

ngco

unse

lors

(con

tinu

ed)

Han

dley

(198

2)M

idw

este

rnun

iver

sity

coun

selin

gce

nter

20do

ctor

alst

uden

ts33

begi

nnin

gpr

actic

umC

orre

latio

nal

ques

tionn

aire

MB

TI,

BL

RI,

Cou

nsel

ing

Eva

luat

ion

and

Rat

ing

Scal

e(o

ne-i

tem

ratin

gsof

satis

fact

ion

with

supe

rvis

ion)

Pers

onal

ityty

pe1.

Intu

ition

(vs.

sens

ing)

pred

icte

dsa

tisfa

ctio

nw

ithsu

perv

isio

n

2.M

atch

ing

onI–

Sva

riab

lew

asre

late

dto

satis

fact

ion

Hol

low

ay&

Wam

pold

(198

3)

Aw

este

rnun

iver

sity

9do

ctor

alst

uden

ts30

begi

nnin

gM

Sst

uden

tsT

hree

step

wis

em

ultip

lere

gres

sion

s(p

redi

cted

vari

able

s:su

perv

isor

eval

uatio

nof

trai

nee,

supe

rvis

orse

lf-

eval

uatio

n,an

dco

mfo

rtle

vel;

pred

icto

rs:

code

sof

beha

vior

and

sequ

ence

s)

Aud

iota

pes

ofSe

ssio

ns3,

6,an

d9

(Blu

rnbe

rgsy

stem

);SP

RS,

TPR

S

Em

piri

cal

1.Su

perv

isor

eval

uatio

nof

supe

rvis

eepr

edic

ted

byhe

lper

-he

lpee

role

s;su

perv

isor

self

-ev

alua

tion

byin

tera

ctio

n(n

otsi

lenc

e);

leve

lof

com

fort

byre

peat

edqu

estio

ning

bysu

perv

isee

2.Su

perv

isee

eval

uatio

nof

supe

rvis

orpr

edic

ted

bysu

perv

isor

prob

lem

orin

dire

ctne

ss;

supe

rvis

eese

lf-

eval

uatio

nby

repe

ated

ques

tioni

ngby

supe

rvis

or;

leve

lof

com

fort

byde

fens

ive

supe

rvis

or

Ric

kard

s(1

984)

Eas

tern

univ

ersi

tyco

unse

ling

prac

ticum

17pr

actic

umin

stru

ctor

s(5

PhD

,10

MS,

2M

SW)

28be

ginn

ing

MS

stud

ents

Cor

rela

tiona

lbe

twee

non

ese

lf-r

epor

tan

dco

ded

verb

albe

havi

or

SRF;

25-m

inau

diot

ape

(Blu

m-b

erg

syst

emfo

rco

ding

)fr

omSe

ssio

ns9–

12

Soci

alin

fluen

ce1.

Mod

erat

ere

latio

nbe

twee

nsu

perv

isor

verb

albe

havi

oran

dpe

rcep

tions

ofsu

perv

isee

2.B

egin

ning

supe

rvis

eepe

rcei

ves

supe

rvis

oras

posi

tive

unle

sssu

perv

isor

’sbe

havi

oris

nega

tive

3.SR

Fun

itary

fact

or

Post

-MS

coun

selo

rs

Lam

bert

(197

4)U

nive

rsity

coun

selin

gce

nter

4Ph

D,

1M

SW;

staf

fm

embe

rsw

ithat

leas

t5

year

s’co

unse

ling

expe

rien

ce

10(8

coun

selin

gps

ycho

logy

,2

first

-yea

rM

SW;

7m

en,

3w

omen

)

AN

OV

As:

IV�

coun

selin

gor

supe

rvis

ion:

DV

�em

path

y,re

spec

t,ge

nuin

enes

s,an

dsp

ecifi

city

;H

IM(w

ork

inco

unse

ling

orsu

perv

isio

n)

Car

khuf

fra

ting

scal

es;

audi

otap

esra

ted

with

Hill

inte

ract

ion

mat

rix

Rog

eria

n1.

Inco

unse

ling

ther

ew

asm

ore

empa

thy,

spec

ifici

ty,

and

“wor

k”th

anin

supe

rvis

ion.

2.R

espe

ctan

dge

nuin

enes

sw

ere

equa

lin

each

.

Tab

le2

(con

tinu

ed)

Aut

hors

Sam

ple

Supe

rvis

ors

Supe

rvis

ees

Des

ign

Inst

rum

ent

The

ory

Sele

cted

findi

ngs

Post

-MS

coun

selo

rs(c

onti

nued

)

Hut

t,Sc

ott,

&K

ing

(198

3)U

nive

rsity

ofPi

ttsbu

rgh

trai

ning

prog

ram

s

6po

st-M

Ss,

2ea

chfr

omco

unse

lor

educ

atio

n,cl

inic

alps

ycho

logy

,so

cial

wor

k

Phen

omen

olog

ical

(int

ervi

ews,

iden

tifica

tion

ofco

nstr

ucts

,re

inte

rvie

w)

Aud

iota

peof

stru

ctur

edin

terv

iew

Phen

omen

olog

ical

theo

ry1.

Posi

tive

supe

rvis

ion

mar

ked

bybo

thta

skan

dre

latio

nshi

por

ient

atio

n;fo

cus

onre

latio

nshi

pan

dsu

perv

isor

self

-di

sclo

sure

used

tobr

eak

impa

sses

2.N

egat

ive

supe

rvis

ion

mar

ked

bysu

perv

isor

criti

cism

with

out

supp

ort,

infle

xibl

esu

perv

isio

n,un

reso

lved

impa

sses

,su

perv

isee

resi

stan

ce;

supe

rvis

eefo

cuse

son

rela

tions

hip

and

not

task

Cou

nsel

ors

atse

vera

lle

vels

Wile

y(1

982)

9m

ajor

univ

ersi

tyco

unse

ling

cent

ers

(Eas

tC

oast

toPl

ains

;8

stat

es)

71su

perv

isor

s(6

5%Ph

D,

28%

inte

rns,

7%ot

her)

107

(10

atL

evel

I,36

atL

evel

II,

30at

Lev

elII

I,31

atL

evel

IVas

rate

dby

supe

rvis

or(1

–9

sem

este

rs)

AN

OV

As:

IV�

supe

rvis

eele

vel,

DV

�no

.pr

actic

a;IV

�su

perv

isio

nen

viro

nmen

t,D

V�

no.

prac

tica

Psc

ale

(rat

ing

ofsu

perv

isee

leve

l);

Esc

ale

(rat

ing

ofsu

perv

isio

nen

viro

nmen

t);

end

ofse

mes

ter

ratin

gsin

satis

fact

ion

and

impa

ctof

supe

rvis

ion

onsu

perv

isee

(fro

mSQ

)

Stol

tenb

erg’

s(1

981)

Cou

nsel

orC

ompl

exity

Mod

el

1.Su

perv

isee

leve

l(P

scal

e)re

late

dto

supe

rvis

edbu

tno

tun

supe

rvis

edex

peri

ence

2.Su

perv

isio

nen

viro

nmen

t(E

scal

e)re

late

dto

supe

rvis

edbu

tno

tun

supe

rvis

edex

peri

ence

3.C

ongr

uenc

yof

supe

rvis

ion

leve

lan

dsu

perv

isio

nen

viro

nmen

tw

asno

tre

late

dto

satis

fact

ion

orim

pact

ofsu

perv

isor

4.Su

perv

isor

ste

nded

topr

ovid

een

viro

nmen

tsth

atei

ther

wer

em

atch

edto

orw

ere

less

adva

nced

than

supe

rvis

ee’s

leve

l

Rap

hael

(198

2)U

nive

rsity

coun

selin

gce

nter

,U

nive

rsity

ofM

aryl

and

10(M

�9.

5ye

ars

ofex

peri

ence

assu

perv

isor

)

10(4

infir

stor

seco

ndpr

actic

um,

6w

ithfo

uror

mor

ese

mes

ters

ofpr

actic

um,

usua

llyin

tern

)

AN

OV

As

Supe

rvis

ion

Ver

bal

Res

pons

eC

ateg

ory

Syst

em

1.N

odi

ffer

ence

inam

ount

ofta

lk,

but

diff

eren

ces

inty

peof

talk

bysu

perv

isor

ofbe

ginn

ing

and

adva

nced

supe

rvis

ees

2.Su

perv

isor

ofad

vanc

edsu

perv

isee

mad

em

ore

stat

emen

tsfo

cusi

ngon

clie

nt,t

hera

py,a

ndsu

perv

isor

;su

perv

isor

ofbe

ginn

ing

supe

rvis

eem

ade

mor

est

atem

ents

focu

sing

onsu

perv

isee

beha

vior

duri

ngth

erap

y,su

perv

isee

feel

ings

rega

rdin

gth

erap

y,an

dsu

perv

isio

nre

latio

nshi

p.(t

able

cont

inue

s)

Tab

le2

(con

tinu

ed)

Aut

hors

Sam

ple

Supe

rvis

ors

Supe

rvis

ees

Des

ign

Inst

rum

ent

The

ory

Sele

cted

findi

ngs

Cou

nsel

ors

atse

vera

lle

vels

(con

tinu

ed)

Cro

ss&

Bro

wn

(198

3)

Mar

riag

eco

unse

ling

agen

cy9

(7m

ale,

2fe

mal

e)pr

ofes

sion

als

inps

ycho

logy

orso

cial

wor

kw

ithm

ean

of4

year

s’ex

peri

ence

assu

perv

isor

s

19(5

mal

e,14

fem

ale)

trai

ned

but

nopr

actic

um;

22(5

mal

e,17

fem

ale)

trai

ned

plus

1ye

arof

prac

ticum

:9

(1m

ale,

8fe

mal

e)tr

aine

dpl

us2

year

sof

prac

ticum

Fact

oran

alys

is,

then

MA

NO

VA

with

IV�

supe

rvis

eeex

peri

ence

,D

V�

four

fact

orgr

oups

Supe

rvis

ion

Que

stio

nnai

repl

ustw

oite

ms

Em

piri

cal

anal

ysis

ofpe

rcei

ved

supe

rvis

orbe

havi

ors

1.Fa

ctor

anal

ysis

reve

aled

four

fact

ors:

eval

uativ

e,su

ppor

t,tim

e/st

ruct

ure

met

hod

ofsu

perv

isio

n,an

dra

ppor

t2.

Gro

uptr

aine

dpl

us1

year

ofpr

actic

umdi

das

wel

las

grou

ptr

aine

dpl

us2

year

sof

prac

ticum

for

met

hod

ofsu

perv

isio

n3.

Gro

uptr

aine

dpl

us1

year

ofpr

actic

umdi

das

wel

las

grou

ptr

aine

dpl

us2

year

sof

prac

ticum

for

supp

ortiv

ean

din

tens

ive

and

for

effe

ctiv

enes

s

M.

L.

Frie

dlan

der

&Sn

yder

(198

3)

New

Yor

k,O

hio,

Uta

hN

one

82tr

aine

es(2

9be

ginn

ing,

31ad

vanc

edpr

actic

a;22

inte

rns)

Hie

rarc

hica

lm

ultip

lere

gres

sion

Self

-Effi

cacy

Inve

ntor

yT

rain

ing

Exp

erie

nces

Que

stio

nnai

re.

SRF

Supe

rvis

ion

Que

stio

nnai

re:

17

Cou

nsel

orco

mpl

exity

,so

cial

influ

ence

,se

lf-e

ffica

cy

1.Su

perv

isee

expe

cted

supe

rvis

orto

betr

ustw

orth

y,th

enex

pert

,th

enat

trac

tive

and

eval

uativ

e,an

d(l

ast)

supp

ortiv

e2.

Exp

erie

nce

leve

lac

coun

tsfo

rlit

tledi

ffer

ence

inex

pect

atio

ns

Mia

rset

al.

(198

3)U

nive

rsity

facu

ltyor

coun

selin

gce

nter

staf

f

37(1

6m

ale,

21fe

mal

e)Ph

Dco

unse

ling

orcl

inic

alps

ycho

logi

sts

with

supe

rvis

ion

expe

rien

ce(M

�8.

2ye

ars)

Firs

tse

mes

ter,

seco

ndse

mes

ter,

adva

nced

prac

ticum

,in

tern

MA

NO

VA

:IV

�su

perv

isee

expe

rien

ce;

DV

�su

perv

isor

’spe

rcep

tion

ofth

eir

beha

vior

.A

NO

VA

s:IV

�su

perv

isor

’spo

st-P

hDex

peri

ence

,su

perv

isor

’sse

x,su

perv

isor

’sjo

bfu

nctio

n;D

V�

supe

rvis

or’s

perc

eptio

nof

beha

vior

Lev

elof

Supe

rvis

ion

Surv

ey(6

5ite

ms

from

Stol

tenb

erg’

sth

eory

and

Supe

rvis

ion

Que

stio

nnai

re)

Cou

nsel

orco

mpl

exity

1.G

ener

ally

,su

perv

isor

spe

rcei

ved

them

selv

esas

beha

ving

diff

eren

tlyw

ithsu

perv

isee

sat

first

and

seco

ndpr

actic

avs

.ad

vanc

edpr

actic

aan

din

tern

ship

2.G

ener

ally

,su

perv

isor

svi

ewed

them

selv

esas

prov

idin

gm

ore

mon

itori

ng,

inst

ruct

ion,

dire

ctio

n,an

dsu

ppor

tan

dle

sspe

rson

alfo

cus

and

deal

ing

with

clie

ntre

sist

ance

for

less

expe

rien

ced

than

for

mor

eex

peri

ence

dco

unse

lors

3.Su

perv

isor

sof

diff

eren

tco

unse

ling

orsu

perv

isio

nex

peri

ence

did

not

mak

edi

ffer

entia

ldi

scri

min

atio

nac

cord

ing

tole

vel

ofsu

perv

isee

Tab

le2

(con

tinu

ed)

Aut

hors

Sam

ple

Supe

rvis

ors

Supe

rvis

ees

Des

ign

Inst

rum

ent

The

ory

Sele

cted

findi

ngs

Cou

nsel

ors

atse

vera

lle

vels

(con

tinu

ed)

Mia

rset

al.

(198

3)(c

onti

nued

)

4.Su

perv

isor

sw

how

ere

psyc

hoan

alyt

ical

lyor

ient

edm

ade

mor

edi

scri

min

atio

nsby

supe

rvis

eele

vel

than

did

supe

rvis

ors

who

wer

ehu

man

istic

orot

her

orie

nted

5.Su

perv

isor

sdi

dno

tdi

ffer

indi

scri

min

atio

nby

supe

rvis

eele

vel

byei

ther

gend

eror

job

func

tion

Rei

sing

&D

anie

ls(1

983)

20un

iver

sity

coun

selin

gce

nter

sth

roug

hout

U.S

.

Non

e14

1(3

0pr

e-M

S;42

MS

plus

upto

2ye

ars’

expe

rien

ce;

44M

Spl

us3

year

s’ex

peri

ence

orin

tern

ship

;25

PhD

)

Fact

oran

alys

is;

disc

rim

inan

tan

alys

is;

AN

OV

AS

with

plan

ned

com

pari

son

acro

ssex

peri

ence

Cou

nsel

orD

evel

opm

ent

Que

stio

nnai

re,

crea

ted

and

valid

ated

inth

isst

udy.

Tw

osu

b-te

sts:

Tra

inee

and

Supe

rvis

ory

Nee

ds(i

tem

sfr

omSQ

,re

sear

chby

Kir

chne

r[1

975]

and

Rei

sing

&D

anie

ls)

Hog

an’s

deve

lopm

enta

lth

eory

1.St

rong

supp

ort

for

the

cons

truc

tva

lidity

ofH

ogan

’sm

odel

thou

ghno

tfo

rhi

ssu

perv

isio

nre

com

men

datio

ns2.

Supe

rvis

eeat

adva

nced

MS

leve

lor

belo

wis

anxi

ous,

depe

nden

t,te

chni

que

orie

nted

,an

dno

tre

ady

for

conf

ront

atio

nin

rela

tion

tosu

perv

isee

atin

tern

and

PhD

leve

l3.

Inte

rns

and

PhD

sdi

ffer

edin

inde

pend

ence

,w

ork

valid

atio

n,an

dco

mm

itmen

tas

coun

selo

rsM

.L

.Fr

iedl

ande

r&

War

d(1

984)

Stud

y1

API

Csi

tes

(35

stat

es)

202

PhD

s(1

1.5

year

ssu

perv

isor

yex

peri

ence

)

Fact

oran

alys

is,

cano

nica

lco

rrel

atio

nSu

perv

isor

ySt

yles

Inve

ntor

y(p

aral

lel

vers

ions

for

supe

rvis

oran

dsu

perv

isee

)

Supe

rvis

ory

styl

eis

mul

tidim

ensi

onal

(attr

activ

e,in

terp

erso

nally

sens

itive

,an

dta

skor

ient

ed.

Psyc

hody

nam

icsu

perv

isor

diff

eren

tin

styl

efr

omco

gniti

ve-b

ehav

iora

lsu

perv

isor

;m

ore

inte

rper

sona

llyse

nsiti

ve,

less

task

orie

nted

Stud

y2

Gra

duat

epr

ogra

ms

orin

tern

ship

site

sin

9st

ates

183

grad

uate

stud

ent

trai

nees

(36

MS,

147

doct

oral

)

Sam

eas

for

Stud

y1

Sam

eas

for

Stud

y1;

32st

uden

tsal

soco

mpl

eted

the

Cro

wne

-Mar

low

e(1

964)

Soci

alD

esir

abili

tySc

ale

Supe

rvis

ory

styl

eis

mul

tidim

ensi

onal

(as

for

Stud

y1)

(tab

leco

ntin

ues)

Tab

le2

(con

tinu

ed)

Aut

hors

Sam

ple

Supe

rvis

ors

Supe

rvis

ees

Des

ign

Inst

rum

ent

The

ory

Sele

cted

findi

ngs

Cou

nsel

ors

atse

vera

lle

vels

(con

tinu

ed)

Stud

y3

API

Csi

tes

(19

stat

es)

135

PhD

s(8

.8ye

ars’

supe

rvis

ory

expe

rien

ce)

Fact

oran

alys

is;

AN

OV

As

Sam

eas

for

Stud

y1

Supe

rvis

orra

ted

selv

eshi

gher

onin

terp

erso

nal

sens

itivi

tyan

dat

trac

tiven

ess

with

inte

rns

than

with

prac

ticum

stud

ents

but

wer

em

ore

task

orie

nted

with

prac

ticum

stud

ents

than

with

inte

rns

Stud

y4

Gra

duat

epr

ogra

ms

105

grad

uate

stud

ents

52M

S;48

doct

oral

,5

psyc

hiat

ric

resi

dent

s)

Sam

eas

for

Stud

y3

Sam

eas

for

Stud

y1

Supe

rvis

eera

ted

supe

rvis

orhi

gher

onin

terp

erso

nal

sens

itivi

tyan

dlo

wer

onta

skor

ient

atio

nfo

rin

tern

sth

anfo

rpr

actic

umst

uden

ts

Stud

y5

One

nort

heas

tern

univ

ersi

ty’s

trai

ning

prog

ram

s

28pr

e-Ph

Dst

uden

tsR

atin

gof

vide

otap

e;A

NO

VA

with

repe

ated

mea

sure

s

Sam

eas

for

Stud

y1

Exp

ert

supe

rvis

ors

(n�

4)w

ere

rate

das

diff

erin

gfr

omea

chot

her

Hep

pner

&R

oehl

ke(1

984)

Stud

y1

Uni

vers

ityco

unse

ling

cent

erst

uden

tsSu

perv

isor

sin

begi

nnin

gpr

actic

umw

ere

all

inte

rns;

for

adva

nced

prac

ticum

,al

lw

ere

post

-PhD

exce

ptfo

r3

who

wer

ein

tern

s.A

llin

tern

coun

selo

rs’

supe

rvis

ors

wer

epo

st-P

hD

56(2

5be

ginn

ing,

19ad

vanc

edpr

actic

um;

12in

tern

s)

MA

NO

VA

:IV

�ex

peri

ence

leve

ls,

DV

�ex

pect

atio

ns;

LO

C;

SRF

Rot

ter’

sin

tern

al-e

xter

nal

locu

sof

cont

rol

scal

e;Su

perv

isor

yE

xpec

tatio

ns,

expe

ctat

ions

abou

tco

unse

ling,

CR

F-SR

FSu

perv

isio

nPe

rcep

tion

Form

(thi

sst

udy)

Inte

rper

sona

lin

fluen

cea.

No

diff

eren

ces

insu

perv

isor

yex

peri

ence

leve

lfo

rex

pect

atio

ns,

locu

sof

cont

rol,

orpe

rcep

tions

ofsu

perv

isor

asex

pert

,at

trac

tive,

ortr

ustw

orth

yb.

Few

diff

eren

ces

insu

perv

isee

char

acte

rist

ics

and

perc

eptio

nsof

supe

rvis

or

Stud

y2

Sam

eas

for

Stud

y1

Supe

rvis

ors

wer

eal

lpo

st-P

hDex

cept

for

3w

how

ere

inte

rns.

All

inte

rnco

unse

lors

’su

perv

isor

sw

ere

post

-PhD

49(1

8be

ginn

ing,

19ad

vanc

edpr

actic

um;

12in

tern

s)

Cor

rela

tiona

lSu

perv

isio

nQ

uest

ionn

aire

(rev

ised

)E

mpi

rica

la.

Beg

inni

ngst

uden

ts’

ratin

gsof

supe

rvis

orbe

havi

orw

ere

mos

tco

rrel

ated

with

posi

tive

eval

uatio

nof

supe

rvis

or(5

9of

135

sign

ifica

nt);

adva

nced

prac

ticum

stud

ents

(41

of13

5)an

din

tern

s(1

2of

135)

show

edle

ssre

latio

nshi

pb.

Beg

inni

ngsu

perv

isee

sra

ted

rapp

ort,

supp

ort,

and

none

valu

ativ

ete

achi

ngas

mos

tim

port

ant;

adva

nced

supe

rvis

ees

rate

dte

achi

ngm

ost

impo

rtan

t,th

ensu

ppor

t;in

tern

supe

rvis

ees

rate

dsu

ppor

t,pe

rson

alco

nfro

ntat

ion,

and

criti

cal

feed

back

asm

ost

impo

rtan

t

Tab

le2

(con

tinu

ed)

Aut

hors

Sam

ple

Supe

rvis

ors

Supe

rvis

ees

Des

ign

Inst

rum

ent

The

ory

Sele

cted

findi

ngs

Cou

nsel

ors

atse

vera

lle

vels

(con

tinu

ed)

Stud

y3

Sam

eas

for

Stud

y1

Supe

rvis

ors

wer

eal

lpo

st-P

hDex

cept

for

2w

how

ere

inte

rns.

All

inte

rnco

unse

lors

’su

perv

isor

sw

ere

post

-PhD

40(1

5be

ginn

ing,

13ad

vanc

edpr

actic

um;

12in

tern

s)

Freq

uenc

yco

unt

Cri

tical

Inci

dent

sQ

uest

ionn

aire

(thi

sst

udy)

Log

anbi

llet

al.’

sde

velo

pmen

tal

issu

esin

supe

rvis

ion

a.C

ritic

alin

cide

nts

occu

rred

earl

ier

for

inte

rns

than

for

prac

ticum

stud

ents

b.B

egin

ning

and

adva

nced

prac

ticum

stud

ents

had

criti

cal

inci

dent

sre

latin

gto

emot

iona

law

aren

ess/

conf

ront

atio

n;co

mpe

tent

and

supp

ortiv

ein

tern

sha

dcr

itica

lin

cide

nts

rela

ting

tope

rson

alis

sues

Wor

thin

gton

(198

4a)

10un

iver

sity

coun

selin

gce

nter

sfr

om6

geog

raph

ical

regi

ons

with

inco

ntin

enta

lU

.S.

155

post

-PhD

and

82pr

e-Ph

Dsu

perv

isor

s

237

(67

infir

stse

mes

ter,

31in

seco

ndse

mes

ter,

53in

thir

dse

mes

ter,

47in

four

thse

mes

ter;

37in

tern

s)

MA

NO

VA

s:IV

�hi

ghan

dlo

wsa

tisfa

ctio

n,pe

rcei

ved

com

pete

nce,

and

perc

eive

dim

pact

,ea

chcr

osse

dw

ithfiv

ele

vels

ofex

peri

ence

(thr

ee2

�5

desi

gns)

;D

V�

12fa

ctor

grou

ping

sof

Supe

rvis

ion

Que

stio

nnai

re(r

evis

ed)

Supe

rvis

ion

Que

stio

nnai

re(r

evis

ed)

Em

piri

cal

anal

ysis

ofpe

rcei

ved

supe

rvis

orbe

havi

ors

1.Su

perv

isio

ndi

ffer

edac

ross

setti

ngs

2.Su

perv

isio

nat

all

leve

lsin

volv

edm

onito

ring

,pe

rson

alin

volv

emen

t,sk

ills

trai

ning

,an

dra

ppor

t3.

Supe

rvis

ion

diff

ered

acro

ssle

vels

on(a

)in

depe

nden

cew

ithdi

rect

ion,

(b)

infr

eque

ntly

taug

htsk

ills,

(c)

dire

ctm

onito

ring

,an

d(d

)es

tabl

ishi

nggo

als

4.Su

ppor

tst

rong

est

infir

stpr

actic

uman

din

tern

ship

5.C

once

ptua

lizat

ion

and

inte

rven

tion

cuts

acro

ssle

vels

6.E

ffec

tive

supe

rvis

ion

was

rela

ted

toac

cept

ance

and

supp

ort

and

tocl

ear

teac

hing

7.Im

pact

ful

supe

rvis

ion

was

rela

ted

toev

alua

tion;

teac

hing

adva

nced

skill

s,an

dsu

ppor

t8.

Pre-

and

post

-PhD

supe

rvis

ors

did

not

diff

er

Yog

ev&

Pion

(198

4)U

nive

rsity

coun

selin

gce

nter

31(1

7m

ale,

14fe

mal

e),

M�

10ye

ars’

coun

selin

gex

peri

ence

and

5ye

ars’

supe

rvis

ion

expe

rien

ce

10fir

st-y

ear

prac

ticum

,13

seco

nd-y

ear

prac

ticum

,8

inte

rns

One

-way

AN

OV

AS

Que

stio

nnai

re(t

his

stud

y)fo

ras

sess

ing

goal

s,ex

pect

atio

ns,

and

proc

edur

esas

perc

eive

dby

supe

rvis

ors

Dev

elop

men

tal

mat

chin

gN

odi

ffer

ence

sw

ithsu

perv

isee

coun

selin

gex

peri

ence

insu

perv

isor

s’go

als,

expe

ctat

ions

,an

dpr

oced

ures

(tab

leco

ntin

ues)

Tab

le2

(con

tinu

ed)

Aut

hors

Sam

ple

Supe

rvis

ors

Supe

rvis

ees

Des

ign

Inst

rum

ent

The

ory

Sele

cted

findi

ngs

Cou

nsel

ors

atse

vera

lle

vels

(con

tinu

ed)

Wor

thin

gton

&St

ern

(198

5)St

uden

tsin

MS

and

PhD

prog

ram

sat

larg

em

idw

este

rnun

iver

sity

92(1

3po

st-P

hD,

23pr

e-Ph

D;

34m

ale,

14fe

mal

e)

8623

pre-

MS,

23po

st-M

S;12

mal

e,12

fem

ale)

MA

NO

VA

s:IV

�su

perv

isee

degr

eele

vel

and

stat

us,

supe

rvis

eege

nder

,su

perv

isor

gend

er,

gend

erm

atch

ing

(eac

hcr

osse

dw

ithtim

e):

DV

�(f

orsu

perv

isor

)fo

urm

easu

res

ofre

latio

nshi

pst

reng

th,

one

ofsa

tisfa

ctio

n,on

eof

supe

rvis

eeco

mpe

tenc

e;(f

orsu

perv

isee

)th

ree

mea

sure

sof

rela

tions

hip

qual

ity,

thre

eof

eval

uatio

nof

supe

rvis

or

Supe

rvis

ion

Que

stio

nnai

re(r

evis

ed);

6L

iken

scal

eite

ms

onsu

perv

isio

nre

latio

nshi

p

Em

piri

cal

exam

inat

ion

ofsu

perv

isio

nbe

havi

oran

dre

latio

nshi

ps;

gend

erdi

ffer

ence

s

1.Su

perv

isor

and

supe

rvis

eedi

ffer

inth

eir

perc

eptio

nsof

supe

rvis

ion

2.Su

perv

isor

sra

tepr

e-M

Ssu

perv

isio

nre

latio

nshi

pto

chan

geov

ertim

e,po

st-M

Sre

latio

nshi

psno

t3.

Supe

rvis

ees

rate

rela

tions

hips

insu

perv

isio

nto

impr

ove

4.Pr

e-M

Sst

uden

tsra

ted

thei

rre

latio

nshi

psw

ithsu

perv

isor

sbe

tter

than

did

post

-MS

stud

ents

5.Su

perv

isor

stat

usas

facu

ltym

embe

ror

doct

oral

stud

ent

mad

eno

diff

eren

cein

rela

tions

hip

6.M

ale

supe

rvis

ees

thou

ght

that

they

had

bette

rre

latio

nshi

psth

andi

dfe

mal

esu

perv

isee

s;m

ale

supe

rvis

ors

thou

ght

that

they

had

bette

rre

latio

nshi

psth

andi

dfe

mal

esu

perv

isor

s7.

Gen

der

mat

chin

gis

impo

rtan

tto

supe

rvis

ees

but

not

tosu

perv

isor

s8.

Supe

rvis

ees

rate

dre

latio

nshi

psas

impr

oved

mos

tby

supe

rvis

orac

tivity

,go

alor

ient

atio

n,an

dsu

perv

isor

supp

ort

Zuc

ker

&W

orth

ingt

on(1

986)

25un

iver

sity

coun

selin

gce

nter

sth

roug

hout

U.S

.

59po

st-P

hDsu

perv

isor

s,84

%lic

ense

d,67

%w

ith6

orm

ore

year

s’ex

peri

ence

atsu

perv

isio

n

25po

st-P

hDps

ycho

logi

sts

bein

gsu

perv

ised

for

licen

sure

and

34in

tern

s

Five

MA

NO

VA

S:IV

�hi

ghan

dlo

wsa

tisfa

ctio

n,co

mpe

tenc

e,im

pact

;in

tern

svs

.po

st-P

hDan

dlic

ensu

rest

atus

ofsu

perv

isor

;D

V�

12fa

ctor

grou

ping

sof

the

Supe

rvis

ion

Que

stio

nnai

re(r

evis

ed)

Supe

rvis

ion

Que

stio

nnai

re(r

evis

ed),

open

-end

edSu

perv

isio

nQ

uest

ionn

aire

conc

erni

ngtim

ein

supe

rvis

ion

Em

piri

cal

anal

ysis

ofpe

rcei

ved

supe

rvis

orbe

havi

or

1.In

tern

san

dpo

st-P

hDps

ycho

logi

sts

wer

esu

perv

ised

sim

ilarl

yex

cept

for

eval

uatio

n

2.A

dvan

ced

supe

rvis

ion

focu

sed

onpr

ovid

ing

supp

ort,

pers

onal

inte

rsts

,goo

dra

ppor

t,m

aint

enan

ceof

role

dist

inct

ions

,gi

ving

feed

back

,and

conf

ront

atio

nan

dev

alua

tion

3.L

icen

sure

stat

usof

supe

rvis

orm

ade

nodi

ffer

ence

Not

e.C

RF

�C

ouns

elor

Res

pons

eFo

rm;

SRF

�Su

perv

isor

Res

pons

eFo

rm;

MB

TI

�M

yers

-Bri

ggs

Typ

eIn

dica

tor;

BL

RI

�B

arre

tt-L

enna

rdR

elat

ions

hip

Inve

ntor

y;I-

S�

Intu

ition

-Sen

sing

dim

ensi

on;

SPR

S�

Supe

rvis

orPe

rson

alR

eact

ion

Scal

e;T

PRS

�T

rain

eePe

rson

alR

eact

ion

Scal

e;A

NO

VA

�an

alys

isof

vari

ance

;IV

�in

depe

nden

tva

riab

le;

DV

�de

pend

ent

vari

able

;M

AN

OV

A�

mul

tivar

iate

anal

ysis

ofva

rian

ce;

API

C�

Ass

ocia

tion

ofPs

ycho

logy

Inte

rnsh

ipC

ente

rs;

LO

C�

locu

sof

cont

rol

scal

e;H

IM�

Hill

Inte

ract

ion

Mat

rix;

SQ�

Supe

rvis

ion

Que

stio

nnai

re.

throughout the United States differed. Threeself-report instruments have been used with reg-ularity: the Supervision Questionnaire (SQ) or amodification of it (Worthington & Roehlke,1979), a modification of the Counselor Re-sponse Form (CRF; Barak & LaCrosse, 1975),and the Barrett-Lennard Relationship Inventory(Barrett-Lennard, 1962). Researchers in threestudies directly measured supervisor behavior(Holloway & Wampold, 1983; Lambert, 1974;Rickards, 1984). Two used the Blumberg Cod-ing System (Blumberg, 1980) to categorize re-sponses, and one used Carkhuff (1969) ratings.Six of the studies have been correlational andone phenomenological; in one the researchertested hypotheses by means of a quasi-experimental design.

For studies in which researchers examinedtrainees at various levels of development, themethodology has been more varied. In 6 of 12studies, researchers used multiple training sites,though in 11 of the 12 the researchers sampleduniversity counseling centers. The nationalsamples have ranged from 3 sites (M. L. Fried-lander & Snyder, 1983) to 35 sites (M. L. Fried-lander & Ward, 1984). Because of wide geo-graphical range of samples, only one study in-volved measurement of actual behavior duringsupervision (Raphael, 1982). In most (7 of 12)the researchers used the Supervision Question-naire, its revision (Worthington, 1984a), oritems taken from it. Two researchers used mod-ifications of the CRF. Wiley (1982) and Reisingand Daniels (1983) created their own instru-ments. M. L. Friedlander and Ward (1984) cre-ated the Supervisory Styles Inventory and per-formed five independent studies of its reliabilityand validity, which constitutes the best psycho-metric data on an instrument designed specifi-cally to measure supervision. In 8 of the 12studies, hypotheses were tested in a quasi-experimental design.

What Is Known

1. There is some support for general devel-opmental models as proposed by Hogan (1964)and others (e.g., Stoltenberg, 1981, and Logan-bill et al., 1982). However, Holloway (1987)critically evaluated the extent to which the as-sumptions of a developmental model have beenmet by the theories and research purporting to

take a developmental perspective, and she tookissue with this conclusion.

2. In general, perceptions of supervisors andsupervisees have been broadly consistent withdevelopmental theories. Reising and Daniels(1983) tested some of Hogan’s (1964) ideas andshowed that from anxiety, need for techniques,and an unwillingness to be confronted to lowneed for work validation, counselors develophigh independence but some ambivalence as totheir role as a counselor. Reising and Danielsalso found that counselors did not identify spe-cific needs for supervision but rather describedstereotypical ideas of a good supervisor. De-spite their level of counselor development, theirideal supervisors were stereotyped, which sug-gests that counselors do not know how goodsupervision might differ for them as they gaincounseling experience. An alternative interpre-tation of Reising and Daniels’s findings is that itdoes not matter to counselors whether their su-pervisor supervises differently across levels ofdevelopment; however, this interpretation is notsupported by ratings of actual supervision,which show that actual supervision behaviordiffers with level of counselor development andthat different supervisor behaviors are related toperceptions of supervisor effectiveness at dif-ferent levels of counselor development.

Wiley (1982) tested aspects of Stoltenberg’s(1981) theory. She operationalized Stolten-berg’s four levels of counselor development bydescribing each level in terms of phrases thatapplied to a counselor at that level. She thendescribed ideal environments for counselors ateach level. Throughout the United States, 71supervisors rated 107 of their supervisees on thelist of descriptive phrases. The supervisors alsodescribed the environment that they believedthey provided for each supervisee on a list ofdescriptive phrases. Wiley tested three mainhypotheses. She found that the level of super-visor-rated development of their superviseeswas related to the amount of supervised coun-seling experience of the counselor but not to thelevel of unsupervised counseling experience.She also found that the supervisors perceivedthemselves to be providing different levels ofsupervisory environment with supervisees ofdifferent levels of supervised counseling expe-rience but not with supervisees of different lev-els of unsupervised counseling experience.Last, in a crucial test of Stoltenberg’s theory,


she found that congruence of supervisee’s levelof experience and supervision environment wasunrelated to either supervisor’s or supervisee’ssatisfaction with supervision. Generally, whensupervisors did not match the supervision envi-ronment with the level of supervisee develop-ment, they differed by providing supervision ata level lower than the supervisee’s level ofdevelopment. There were few gross mis-matches, which suggests that supervisors mightintuitively match levels of counselor and super-vision environment. Wiley’s research is an ex-cellent test of a theoretical position—a too-rareoccurrence in the supervision literature—but itcan be criticized because supervisors rated boththe level of the supervisee and the level ofenvironment that they tried to provide. There isno assurance that they actually provided theenvironment that they believed they provided.In fact, the structure of the task, in which su-pervisors were asked to make the ratings ofsupervisee and supervision environment at thesame time, might have introduced demands forthe supervisors to think more along develop-mental lines than they might usually do.

Miars et al. (1983) also investigated the coun-selor complexity model (Stoltenberg, 1981).Experienced supervisors described their super-vision four times: for first-semester practicum,second-semester practicum, advanced practi-cum, and internship counselors (regardless ofwhether they had experience with supervisees atthat level). They described themselves as con-ducting supervision differently depending onthe level of the hypothetical student at thatlevel. The supervisors reported the most varia-tions across supervisee level in dimensions ofstructure, directiveness, instruction, and degreeof collegiality. Supervisors’ perceived supervi-sory environments paralleled Stoltenberg’s ex-pectations, though supervisors’ expectationswere less differentiated than Stoltenberg’s (cf.Yogev & Pion, 1984). Again, this study suf-fered from use of supervisors alone as partici-pants. Worthington and Roehlke (1979) andHeppner and Roehlke (1984) found that super-visors perceived supervision differently than su-pervisees. Furthermore, all supervisors weredrawn from a single university, in which there isgenerally a high consciousness of how counsel-ors change as they gain experience.

Heppner and Roehlke (1984), using threestudies, examined beginning practicum, ad-

vanced practicum, and intern counselor train-ees. The supervision dyads were composed ofbeginning practicum counselors with intern su-pervisors, advanced practicum counselors withmostly doctoral-level supervisors, and interncounselors with doctoral-level supervisors. Inthe first study, Heppner and Roehlke (1984)found that before the beginning of supervision,supervisees at different levels of experience had(a) essentially the same expectations about su-pervisor expertness, attractiveness, and trust-worthiness and (b) the same locus of controlscores. At the beginning of supervision, super-visees also did not differ in their perceptions ofthe expertness, attractiveness, and trustworthi-ness of their supervisors. Neither superviseecharacteristics nor initial perceptions of super-visors were related to the actual impact of su-pervision or to each other. This was also foundby M. L. Friedlander and Snyder (1983), whoused similar methodology with 82 trainees atbeginning practicum, advanced practicum, andinternship levels in three states.

In their second study, Heppner and Roehlke(1984) used the SQ (Worthington & Roehlke,1979) and three additional items to compare thesupervision behaviors perceived by superviseesof different levels of experience. They corre-lated each supervisor behavior with supervis-ees’ ratings of satisfaction, supervisor compe-tence, and perceived impact of supervision oncounseling ability. Beginning practicum coun-selors were more satisfied with supervisors whofostered a positive relationship with the super-visee. Advanced practicum students were moresatisfied with supervisors who facilitated devel-opment of additional counseling skills. Internswere more satisfied with supervisors whohelped them to develop better counseling skillsand allowed them to deal with personal issues ordefensiveness that affect counseling. Resultsmight be attributable to the nature of the sam-ple: It was from one university counseling cen-ter. Worthington (1984a) surveyed 237 counsel-ors from ten counseling centers nationwide. Hetoo determined SQ behaviors that were relatedto perceptions of satisfaction with supervision,of competence of the supervisor, and of impacton counseling ability. Grouping items accordingto 12 factor-analytic clusters, he found that (a)his factors were similar to many of those foundby Reising and Daniels (1983), (b) support andencouragement were useful at first practicum

150 WORTHINGTON

level and at the internship level, times whenissues of identity were at the forefront, (c) skillsof intervention and conceptualization were builtthroughout all five levels of counselor experi-ence that were investigated, and (d) perceptionsof satisfaction and supervisor competence werepredicted by support and teaching (Heppner &Roehlke, 1984; Worthington & Roehlke, 1979),but perceptions of supervisor impact on thecounselor’s ability were also related to evalua-tion and teaching advanced skills (cf. Heppner& Roehlke, who reported that essentially thesame factors were predictive of all three mea-sures of supervision effectiveness). Taken to-gether, Heppner and Roehlke’s and Worthing-ton’s studies provide limited but reasonablycongruent support for developmental models ofsupervision. Worthington’s findings also sup-port Hess’s (1986) notion of spirals, in whichthe same issues are addressed at progressivelydeeper levels throughout training (see also Lo-ganbill et al., 1982).

In their third study, Heppner and Roehlke(1984) examined some of the critical incidentsin supervision. They found that support wasvalued across the three levels of experience.Emotional awareness, confrontation, and com-petency issues constituted critical incidents forpracticum students far more frequently than forinterns. Interns reported critical incidents in-volving parallel process and discussion of trans-ference and countertransference issues morefrequently than did practicum students. Theseresults too lent support for the general flavor ofdevelopmental theories.

One hole in the investigation of developmen-tal theories is in understanding the master coun-selor stage. Zucker and Worthington (1986) ex-amined differences between interns and post-doctoral applicants for licensure who werebeing supervised. They studied a national sam-ple of 34 interns and 25 psychologists. Fewdifferences were found between the two levelsof experience. Psychologists received supervi-sion that was generally less evaluative than in-terns’ supervision. Advanced supervision of in-terns and postdoctoral psychologists consistedof support and rapport, feedback, confrontationand evaluation, and negative feedback. Zuckerand Worthington concluded that investigationof supervision of master counselors requiredsampling counselors who had more experiencethan recent graduates so that the experienced

professional counselor could be clearly differ-entiated from advanced trainees.

Virtually all of these studies of how supervi-sion changes as counselors gain experiencehave supported the general tenets of develop-mental theories, though most of the details ofthe theories have not been investigated. Ideasabout how counselors develop (see Hill et al.,1981) are more in line with developmental the-ory than are the mechanics of matching super-vision to the different levels.

3. The behavior of supervisors changes ascounselors gain experience. Raphael (1982)compared supervision with trainees in eithertheir first or second practicum against thosewith four or more semesters of practicum. Hedeveloped a nine-category system for classify-ing verbal responses of supervisors. For exam-ple, supervisors could focus on the client, thetherapy relationship with the client, the therapyrelationship with the therapist, the therapy rela-tionship with both, feelings and thoughts of thetherapist about the therapy session, the therapistapart from the session, the supervisory relation-ship, the supervisor, or “other.” Supervisors’statements were sampled from tapes of actualsupervision. Supervisors did not differ in fre-quency of statements with counselors of differ-ent experience levels; however, the distributiondiffered across level of experience. Supervisorsof advanced trainees made higher proportions ofstatements that focused on (a) the client, (b) theclient in therapy, and (c) the supervisor. Super-visors of beginning trainees made higher pro-portion of statements that focused on (a) thecounselor’s behavior in therapy, (b) the coun-selor’s feelings and thoughts about therapy, and(c) the supervisory relationship.

These findings are consistent with researchon perceptions of supervisors and superviseesabout supervisor behavior during supervision.For instance, in most studies of both supervis-ees’ and supervisors’ perceptions of their be-havior, researchers have found that supervisorsteach specific behaviors about therapy to begin-ning-level counselors (see studies reviewed inPoint 2). Beginning counselors are also usuallyinsecure and lack self-confidence (Heppner &Roehlke, 1984; Reising & Daniels, 1983;Worthington, 1984a), which suggests that su-pervision might focus on the counselor’sthoughts and feelings about therapy. However,most supervisors and beginning supervisees


perceive little focus on the supervisory relation-ship during supervision. Focus on the supervi-sory relationship usually is perceived to occur atinternship-level supervision. Focus on the clientand the therapist during therapy are often per-ceived to occur at the practicum level. However,focus explicitly on the supervisor is usuallycharacteristic of supervision of more advancedcounselors (intern and beyond) who share mu-tuality and collegiality.

In other studies researchers have examinedsupervisor behavior at only one level of coun-selor experience. Rickards (1984) examined 28beginning counselors, excerpting 20-min seg-ments from supervision sessions and scoringthem with the Blumberg System for AnalyzingSupervisor-Teacher Interactions (Blumberg,1980). The Blumberg system comprises 15 cat-egories: 10 supervisor and 5 supervisee behav-iors. Rickards (1984) found that perceptions ofsupervisor expertness, attractiveness, and trust-worthiness were moderately related to coun-selor supervision behavior. Criticisms and opin-ions given by supervisors were negatively re-lated to counselor perceptions of positivesupervisor qualities. Counselors who expressednegative social-emotional behavior and whofailed to ask for information were also per-ceived negatively by their supervisors. Hollo-way and Wampold (1983) used the Blumbergsystem with supervision dyads involving begin-ning counselors and found the same results.Their outcome measures, though, were not rat-ings of supervisor expertness, attractiveness,and trustworthiness but were self-report mea-sures of satisfaction with supervision by bothsupervisors and supervisees. Again, critical su-pervisor responses spoiled what were ordinarilyhigh ratings of satisfactions with supervision bysupervisees. Supervisors did not like periods ofsilence and did not like counselors to supportthem for giving opinions. When the supervisorfollowed the trainees’ expression of ideas withrequests for more ideas, satisfaction with super-vision was high for both supervisors and super-visees. However, use of supportive communi-cation produced mixed reactions from supervi-sors and supervisees. One might surmise thatboth parties liked to be shown that their ideaswere appreciated but did not necessarily like theother party to verbalize that appreciation.

4. The supervision relationship changes ascounselors gain experience. The relationship

between supervisor and supervisee is influencedby the supervisee’s perceptions of his or hersupervisor. Heppner and Handley (1981) foundthat counselors’ perceptions of their supervisorsas expert, attractive, and trustworthy were con-sistently correlated with a positive supervisoryrelationship and with satisfaction with supervi-sion. Dodenhoff (1981) also found that whencounselors liked their supervisors, supervisorsliked them and thought them to be more effec-tive. Client ratings of the counselors’ effective-ness did not substantiate the ratings of thesupervisors.

In only one paper did the authors investigatethe effect of experience level of counselors onthe supervision relationship. Worthington andStern (1985) studied supervision relationshipsof pre- and post-master’s-level counselors withpre- and postdoctoral-level supervisors at oneuniversity. Supervisees perceived their relation-ships with their supervisors to change through-out the semester. Generally, supervisees ratedtheir relationship with their supervisors lowestafter 5 weeks of the semester and higher after 10and 15 weeks into the semester. Supervisors didnot perceive a difference in their relationshipswith their supervisees at the three times duringthe semester. Supervisee experience affectedperceptions of the development of the relation-ships. Supervisors perceived that their relation-ships with master’s students were more helpfulat 15 weeks than at 5 weeks into the semester.Supervisors did not perceive their helpfulness todoctoral students to differ at any of the threemeasurement times. Counselors at the master’slevel perceived their relationships to be morepositive with their supervisors than counselorsat the doctoral level. In general, master’s-levelcounselors rated their supervisory relationshipsas steadily improving. Worthington and Sternspeculated that supervisors get to know for-merly unknown master’s students during super-vision but have ongoing relationships with doc-toral students. They also offered a second hy-pothesis that master’s-level counselors might bemore responsive to their supervisors’ sugges-tions, thus causing their supervisors to like themmore (see Dodenhoff, 1981). A final hypothesiswas that doctoral counselors are more likely tobe in the stage of orthodoxy (Hill et al., 1981)and thus are less likely to be open to suggestionsfrom supervisors who differ in theoreticalstance from them.

152 WORTHINGTON

Results from a variety of studies with a va-riety of methodologies are clear. Superviseesperceive that their supervisors give differenttypes of supervision to them when they are atdifferent levels of training. As counselors ad-vance, they perceive supervisors to confront,deal with personal issues, tackle client resis-tance and transference/countertransference is-sues, give negative feedback, and treat them likepeers more often. Furthermore, supervisors areperceived to give less instruction, provide lessstructure, monitor the behavior of the super-visee less, and be less directive.

What Is Missing

Although these findings are consistent withthe various developmental models—which areremarkably consistent with each other—research has failed to answer a number of im-portant questions.

1. Do changes in supervision as counselorsgain experience promote growth and improve-ment of the supervisee, or do they merely sat-isfy the supervisee? Few studies of supervisionhave included measures of clients’ improve-ment (cf. Dodenhoff, 1981; Lambert, 1974) orclients’ perception of the counselor.

2. Within specific theoretical approaches tocounseling, how do supervisor behaviors relateto counselor development over time?

3. Until now, researchers have been con-cerned with identifying good supervisor behav-iors: those that contribute to supervisee satisfac-tion, those that show supervisor competence,and those that affect the supervisee’s counsel-ing. A good theory of lousy supervisor behav-iors is missing. What can the supervisor do toprohibit movement from one stage of counsel-ing to another? What can the supervisor do tocontribute to dissatisfaction with supervision?How can poor counseling outcomes be engen-dered? Although this sounds somewhat face-tious, the need is real. We need to spend morecreative thought in identifying the things thatwe do to contaminate supervision (see Hutt,Scott, & King, 1983; C. D. Miller & Oetting,1966; Rickards, 1984; Rose, 1965; Rosenblatt& Mayer, 1975).

4. Most theories of development of the coun-selor focus on the stages through which thecounselor passes. Little attention has been givento how the counselor makes the transition from

one stage to the next and, specifically, to howthe supervisor can accelerate (or retard, if thatseems appropriate) the counselor’s progressfrom each stage to the next.

5. What is the supervision relationship like?Most studies have focused on the supervisionbehaviors or styles but not on the characteristicsof the relationship (cf. Hutt et al., 1983; Worth-ington & Stern, 1985).

6. What kinds of influence strategies are usedby supervisors, and how do they work on thesupervisee who becomes increasingly psycho-logically sophisticated as experience is ac-crued? Studies on influence strategies to datehave been concerned with applying Strong’s(1968) interpersonal influence model, whichspecified source characteristics of the influ-encer, to supervision. The support for thismodel has been weak ( Heppner & Handley,1981, 1982; Heppner & Roehlke, 1984; Rick-ards, 1984). More sophisticated influence strat-egies, developed in counseling (e.g., Kiesler,1986; Strong, 1986), should be applied to su-pervision and investigated. Theories are neededto describe the influence strategies of supervi-sion, comparing and contrasting them withpsychotherapy.

7. Gender matching has been found to affectthe supervision relationship; matching accord-ing to cognitive style has not influenced therelationship. Would matching on other variablesaffect the process and outcomes of supervision?One variable of potential importance is theoret-ical orientation. One would expect that duringthe early years of counselor training, matchingsupervisors and supervisees according to theorywould be of little importance because of thetheoretical flexibility of the neophyte counselor.By advanced practicum or internship, though,most trainees adhere to a counseling theory(Hill et al., 1981). Having a theoretical mis-match between counselor and supervisor mightproduce dissatisfaction with supervision, espe-cially if the supervisor is also strongly commit-ted to a theoretical stance. At the advanced ormaster counselor stages, matching according totheoretical persuasion might again have littleimpact on supervision.

8. Exactly what is it about supervised expe-rience in counseling that helps a counselor tobecome more proficient? Wiley’s (1982) find-ing that counselors get better with supervisedcounseling experience but not just counseling


Tab

le3

Stud

ies

ofSu

perv

isor

sas

The

yG

ain

Exp

erie

nce

Aut

hors

Sam

ple

Supe

rvis

orSu

perv

isee

Des

ign

Inst

rum

ents

The

ory

Sele

cted

findi

ngs

Ston

e(1

980)

Uni

vers

ityac

adem

icpr

ogra

mfo

rm

ost,

vari

ous

setti

ngs

for

expe

rien

ced

supe

rvis

ors

(all

cogn

itive

-beh

avio

ral)

10un

derg

radu

ates

,6

inte

rns,

11ex

peri

ence

dsu

perv

isor

s(2

orm

ore

year

s’ex

peri

ence

;8

PhD

,3

AB

D)

Non

eA

nalo

gue

AN

OV

AS:

IV�

supe

rvis

orex

peri

ence

,D

V�

plan

ning

stat

emen

ts

Plan

ning

stat

emen

tsco

ded

onSu

perv

isor

Plan

ning

Cod

ing

Syst

em

Cog

nitiv

e1.

Exp

erie

nced

supe

rvis

ordi

ffer

edfr

omun

derg

radu

ates

on2

of9

plan

ning

cate

gori

es;

prod

uctiv

ityan

dco

ncer

nw

ithsu

perv

isee

2.L

ittle

atte

ntio

nto

goal

s,su

perv

isor

need

s,or

high

eror

der

thin

king

Goo

dyea

r&

Rob

yak

(198

2)

Uni

vers

ityco

unse

ling

cent

erst

aff

67(2

0w

ithbe

havi

oral

orco

gniti

ve-

beha

vior

al,

40ec

lect

ic,

7ps

ycho

anal

ytic

)

Non

eA

NO

VA

s:IV

�su

perv

isor

theo

ry,

theo

ry�

no.

supe

rvis

ees;

DV

�fo

cus

onpe

rson

,sk

ills,

orco

ncep

tual

izat

ion

Self

-rep

ort

ofem

phas

ison

each

ofth

ree

foci

ofsu

perv

isio

n

Pre-

and

post

-PhD

supe

rvis

ors

did

not

diff

erin

any

of12

fact

orgr

oupi

ngs

onth

eSu

perv

isio

nQ

uest

ionn

aire

(rev

ised

)

Mia

rset

al.

(198

3)U

nive

rsity

facu

ltyor

coun

selin

gce

nter

staf

f

37(1

6m

ale,

21fe

mal

e)Ph

Dco

unse

ling

orcl

inic

alps

ycho

logi

sts

with

supe

rvis

ion

expe

rien

ce(M

�8.

2ye

ars)

Firs

t-se

mes

ter,

seco

nd-s

emes

ter

adva

nced

prac

ticum

stud

ents

;in

tern

s

MA

NO

VA

:IV

�su

perv

isor

post

-Ph

Dex

peri

ence

;D

V�

supe

rvis

ordi

ffer

entia

tion

acro

ssle

vels

ofco

unse

lor

expe

rien

ce

Lev

elof

Supe

rvis

ion

Surv

ey(6

5ite

ms

from

Stol

tenb

erg’

s(1

981)

theo

ryan

dSu

perv

isio

nQ

uest

ionn

aire

Cou

nsel

orco

mpl

exity

Supe

rvis

ors

ofdi

ffer

ent

coun

selin

gor

supe

rvis

ion

expe

rien

cedi

dno

tm

ake

diff

eren

tial

disc

rim

inat

ions

acco

rdin

gto

leve

lof

supe

rvis

ee

Wor

thin

gton

(198

4a)

10un

iver

sity

coun

selin

gce

nter

sfr

omsi

xge

ogra

phic

alre

gion

sw

ithin

cont

inen

tal

U.S

.

155

post

-PhD

and

82pr

e-Ph

Dsu

perv

isor

s23

7(6

7in

first

sem

este

r,31

inse

cond

sem

este

r,53

inth

ird

sem

este

r,47

info

urth

sem

este

r;37

inte

rns)

MA

NO

VA

:IV

�pr

e-vs

.po

st-P

hDsu

perv

isor

s;D

V�

12fa

ctor

grou

ping

sof

Supe

rvis

ion

Que

stio

nnai

re(r

evis

ed)

Supe

rvis

ion

Que

stio

nnai

re(r

evis

ed)

Em

piri

cal

anal

ysis

ofpe

rcei

ved

supe

rvis

orbe

havi

ors

1.Pr

e-an

dpo

st-P

hDsu

perv

isor

sdi

dno

tdi

ffer

inan

yof

12fa

ctor

grou

ping

son

the

Supe

rvis

ion

Que

stio

nnai

re(r

evis

ed)

Wor

thin

gton

(198

4b)

Lar

geur

ban

univ

ersi

ty82

(12

post

-PhD

,19

post

-MS,

29pr

e-M

S,22

unde

rgra

duat

es),

facu

lty,

staf

fof

coun

selin

gce

nter

,st

uden

ts

Aud

iota

ped

coun

selin

gse

ssio

nA

nalo

gue

4�

2(c

ouns

elin

gex

peri

ence

ofsu

perv

isor

�co

unse

lor

leve

l)M

AN

OV

A

Eig

httr

ait

labe

lsA

ttrib

utio

nth

eory

1.A

ttrib

utio

nof

caus

eof

supe

rvis

eetr

aits

decr

ease

dw

ithex

peri

ence

ofsu

perv

isor

,bu

tpo

st-P

hDsu

perv

isor

diff

ered

from

pre-

PhD

supe

rvis

oron

only

1of

16ad

ject

ives

2.N

oef

fect

for

perc

eive

dex

peri

ence

ofsu

perv

isee

Tab

le3

(con

tinu

ed)

Aut

hors

Sam

ple

Supe

rvis

orSu

perv

isee

Des

ign

Inst

rum

ents

The

ory

Sele

cted

findi

ngs

Mar

ikis

,R

usse

ll,&

Del

l(1

985)

Cou

nsel

ing

psyc

holo

gypr

ogra

mat

larg

em

idw

este

rnun

iver

sity

30(1

0fir

st-y

ear

grad

uate

stud

ents

,10

adva

nced

grad

uate

stud

ents

,10

PhD

sw

ith2–

18ye

ars’

supe

rvis

ion

expe

rien

ce)

Aud

iota

ped

coun

selin

gse

ssio

nof

begi

nnin

gco

unse

lor

One

-way

MA

NO

VA

S(e

xper

ienc

eof

supe

rvis

or)

Supe

rvis

ion

sess

ion

code

dvi

aSu

perv

isor

yPl

anni

ngC

odin

gSy

stem

1.N

odi

ffer

ence

acro

ssle

vels

inpl

anni

ngst

atem

ents

2.L

ow-

and

high

-exp

erie

nced

supe

rvis

ors

mad

em

ore

stat

emen

tsan

dev

oked

mor

esu

perv

isee

stat

emen

tsth

atdi

dno

nexp

erie

nced

supe

rvis

ors

3.Pl

anni

ngw

asno

tre

late

dto

inse

ssio

nve

rbal

beha

vior

4.Su

perv

isor

view

edsu

perv

isee

the

sam

ere

gard

less

ofle

vel

5.Su

perv

isor

with

expe

rien

cera

ted

high

erth

ansu

perv

isor

with

noex

peri

ence

Wor

thin

gton

&St

em(1

985)

Stud

ents

inM

San

dPh

Dpr

ogra

ms

ata

larg

em

idw

este

rnun

iver

sity

92(1

⁄2po

st-P

hD,

2⁄3

pre-

PhD

;3⁄4

mal

e,1⁄4

fem

ale)

86(2

⁄3pr

e-M

S,1⁄3

post

-MS;

1⁄2

mal

e,1⁄2

fem

ale)

MA

NO

VA

s:IV

�su

perv

isee

degr

eele

vel,

supe

rvis

orst

atus

,sup

ervi

see

sex,

supe

rvis

orse

x,ge

nder

mat

chin

g(e

ach

cros

sed

with

time)

;D

V�

(for

supe

rvis

or)

four

mea

sure

sof

rela

tions

hip

stre

ngth

,on

esa

tisfa

ctio

n,on

esu

perv

isee

com

pete

nce;

(for

supe

rvis

ee)

thre

em

easu

res

ofre

latio

nshi

pqu

ality

,th

ree

ofev

alua

tion

ofsu

perv

isor

Supe

rvis

ion

Que

stio

nnai

re(r

evis

ed):

6ite

ms,

Lik

ert

scal

es,

onsu

perv

isio

nre

latio

nshi

p

Em

piri

cal

exam

inat

ion

ofsu

perv

isio

nbe

havi

oran

dre

latio

nshi

ps;

sex

diff

eren

ces

1.Su

perv

isor

sra

tepr

e-M

Ssu

perv

isio

nre

latio

nshi

psto

chan

geov

ertim

e;po

st-M

Sre

latio

nshi

psdo

not

2.Su

perv

isor

stat

usas

facu

ltym

embe

ror

doct

oral

stud

ent

mad

eno

diff

eren

cein

rela

tions

hip

Zuc

ker

&W

orth

ingt

on(1

986)

25un

iver

sity

coun

selin

gce

nter

sth

roug

hout

U.S

.

59po

st-P

hDsu

perv

isor

s,84

%lic

ense

d,67

%w

ith6

orm

ore

year

s’ex

peri

ence

atsu

perv

isio

n

25po

st-P

hDps

ycho

logi

sts

bein

gsu

perv

ised

for

licen

sure

and

34in

tern

s

Five

MA

NO

VA

S:IV

�lic

ensu

rest

atus

ofsu

perv

isor

;D

V�

perc

eptio

nsof

supe

rvis

ion

effe

ctiv

enes

s,12

fact

orgr

oupi

ngs

ofSu

perv

isio

nQ

uest

ionn

aire

(rev

ised

),ra

tings

ofho

wtim

eis

used

insu

perv

isio

n

Supe

rvis

ion

Que

stio

nnai

re(r

evis

ed);

open

-en

ded

Supe

rvis

ion

Que

stio

nnai

reco

ncer

ning

time

insu

perv

isio

n

Em

piri

cal

anal

ysis

ofpe

rcei

ved

supe

rvis

orbe

havi

or

Lic

ensu

rest

atus

ofsu

perv

isor

mad

eno

diff

eren

cein

perc

eptio

nsof

effe

ctiv

enes

sof

supe

rvis

ion,

inw

hat

was

done

duri

ngsu

perv

isio

n,an

din

how

time

was

used

insu

perv

isio

n

Not

e.A

NO

VA

�an

alys

isof

vari

ance

;M

AN

OV

A�

mul

tivar

iate

anal

ysis

ofva

rian

ce;

IV�

inde

pend

ent

vari

able

;D

V�

depe

nden

tva

riab

le.

experience is provocative. It could suggest aparallel to Freud’s notion that one cannot knowone’s own unconscious but needs an objectiveanalyst to help. Perhaps counselors find it dif-ficult to know their own counseling and im-prove it. Counselors create cognitive maps oftheir experience (theories and models) thatguide their perceptions of their own behavior,goals, and understanding of how they counsel.Without supervision to challenge their cognitivemaps, few counselors will systematically audio-tape or videotape to inject an impartial view oftheir counseling. Few will allow their work tobe viewed by others. Clients’ comments aboutcounseling style are often dismissed as beingfrom a naive source. Counselors believe thatthey are actually counseling the way they saythey are, but, as most supervisors know, the wayin which a counselor talks about his or hercounseling is not perfectly correlated with theway in which he or she counsels. It is thesystematic analysis of a counselor’s behaviorfrom a different viewpoint than one’s own thathelps a counselor to change self-perception andbehavior.

The Developing Supervisor: EmpiricalResearch

What Is Known

1. There are differences in skillfulness insupervision across supervisors. In several inves-tigations of the effectiveness of supervision,researchers have found a variety of levels ofsupervisor competence and impact (see Cross &Brown, 1983; Heppner & Handley, 1981;Worthington, 1984a; Worthington & Roehlke,1979; Worthington & Stern, 1985; Zucker &Worthington, 1986). Furthermore, Hester,Weitz, Anchor, and Roback (1976) found thatperceived differences in supervisor skillfulnessare related to attraction of supervisees to super-visors (see Table 3).

2. Supervisors do not become more compe-tent as they gain experience. Once supervisorsreach the advanced practicum level of experi-ence, they can apparently supervise with effec-tiveness equal to that of post-PhD supervisors.Marikis, Russell, and Dell (1985) studied stu-dents in their first practicum, students in theiradvanced practica, and post-PhD counselors.

All acted as supervisors during a 30-min super-vision interview with a counselor. Post-master’sstudents were rated as effective as post-PhDsupervisors, though both groups were rated asmore effective supervisors than beginningpracticum counselors. Marikis et al. (1985)found that the two groups of experienced super-visors addressed counseling skills of the coun-selor and talked about themselves as supervisorsmore than did beginning practicum students.Overall, the more experienced supervisorstalked less than beginning counselors. In noneof the three comparisons did post-master’scounselors differ from post-PhD counselors.Worthington (1984a) surveyed 237 superviseesnationwide and also found that pre- and post-PhD supervisors in actual supervision of coun-selors from first practicum to internship werenot rated as differentially effective in terms oftheir competence, their impact on the super-visee, or satisfaction of the supervisee withsupervision.

Zucker and Worthington (1986) wonderedwhether licensure status of the supervisors ofinterns and postdoctoral candidates for licen-sure affected the quality of supervision. Theysurveyed 59 psychologists and interns nation-wide. The supervisors were generally very ex-perienced. (Only one third had less than 6 yearsof experience.) When years of post-PhD expe-rience was used as a covariate, licensure statusof the supervisor did not affect supervisees’ratings of the effectiveness of supervision. Also,in no case was the covariate (years of experi-ence) significant.

Worthington and Stern (1985) surveyed 92supervisors and 86 supervisees at one univer-sity. They found that neither supervisors norsupervisees rated supervisors as more effectivedepending on their status as either faculty (orsenior staff) or students (interns or advancedpracticum students). There were also no per-ceived differences in the quality or strength ofthe supervision relationships that were attribut-able to faculty or student status of thesupervisor.

In sum, whether supervision experience isconceptualized as degree level, licensure status,or student-faculty status, supervisors beyond themaster’s level do not appear to differ in effec-tiveness of supervision.

3. Supervisors change little in other ways as

156 WORTHINGTON

they gain experience. Miars et al. (1983) di-vided supervisors into those with low (1–5years), medium (6–11 years), and high (12years and beyond) levels both of post-PhDcounseling experience and of supervision expe-rience. They found that supervisors did notmake differential discrimination of counselorneeds as counselors gained experience.

Goodyear and Robyak (1982) divided super-visors into those with 0–8 years of post-PhDexperience versus those with 9 or more years ofexperience. They found no difference acrosslevels of experience in focus of the supervisoron the person of the supervisee, in the skills ofthe counselor, or in the conceptualization of theclient’s problems. Goodyear and Robyak thendivided supervisors into those who had super-vised fewer than 25 supervisees and thosewith 25 or more. Again, no main effects werefound for supervision experience. One interac-tion was found between supervisor theory andsupervisor experience. Behavioral counselorswith fewer supervisees focused more on coun-seling skills than either behavioral counselorswith more supervisees or eclectic supervisorswith fewer supervisees.

Worthington (1984b) had “supervisors” atthe undergraduate, pre-master’s, post-master’s,and post-PhD levels rate a counselor on thebasis of hearing 10 min of excerpts from anaudiotaped counseling session. There were cleardifferences across all levels of supervisor expe-rience in the attributions made to the counselor;however, most of the differences occurred be-tween the undergraduates and the other threelevels of experience. Only one of eight compar-isons showed differences between post-PhDand post-master’s supervisors.

Stone (1980) and Marikis et al. (1985) inves-tigated whether supervisors at different levels ofexperience made different planning statementsbefore supervision. Stone (1980) found that un-dergraduate students differed from post-PhDsupervisors in only two of nine categories ofplanning statements: focus on the superviseeand number of statements. Marikis et al. (1985)found that pre- and post-PhD supervisors didnot differ on any category of planning state-ments, but beginning practicum students diddiffer from the more experienced supervisors onthree of nine categories.

What Is Missing

1. Although evidence to date shows littleindication that supervisors improve with expe-rience, this does not mean that they do notchange with experience. The inexperienced su-pervisor might contribute different benefits tothe supervisee. For example, the new supervisormight promote identification with himself orherself because he or she is or has recently beena student. The new supervisor might be awareof the issues that the trainee is dealing with to agreater extent than the seasoned veteran super-visor is. The new supervisor might be enthusi-astic, energetic, and willing to devote extra ses-sion time to the supervisee more readily than theexperienced supervisor is. On the other hand,the experienced supervisor might have moretechnical expertise than the fledgling, beingmore facile with client assessment, counselinginterventions, or technical skills. Although eachsupervisor is perceived to be effective and com-petent by the supervisee, each might clearlyoffer different resources to the supervisee andcontribute to the counseling effectiveness of thesupervisee in different ways.

Unwilling as we might be to accept it, mostsupervisors simply might not improve with ex-perience. One reason for this might be thatsupervisors have little training in how to super-vise effectively and thus may perpetuate themistakes of their own supervisors. Wiley (1982)showed that with counselors, mere counselingexperience was insufficient to produce changein counseling ability. The same might be truewith mere supervision experience or mere lon-gevity as a post-PhD professional. Mere expe-rience might be insufficient to enable one toview one’s work objectively or to take differentperspectives on one’s work. Perhaps a supervi-sor of one’s supervision would promote im-provement. There are few mechanisms for pro-viding this supervision beyond the internshiplevel, when supervised practice of supervisionis even part of the internship. Hess and Hess(1983) found that in only about 40% of theinternship training sites is there any ongoingtraining of supervisors. In only one third of theplaces in which interns were allowed to super-vise, one-to-one supervision of supervision wasgiven. McColley and Baker (1982) found that


only about 50% of new supervisors had had anytraining in supervision. Only 20% had had acourse or seminar in supervision.

2. The empirical investigation of how super-visors change with experience is at a rudimen-tary level. Few researchers have taken thatquestion as a primary focus of research. This isprobably because there have been few theoret-ical works about supervisors’ development untilrecently.

3. Theories are still imprecise and general.The important variables affecting improvementin or even change in supervision must still beidentified. For example, no one has yet investi-gated the amount of supervised experience insupervision, the amount of involvement inthinking about supervision (through writing pa-pers, giving workshops, conducting classes, andso on), attendance at supervision-related con-tinuing education functions, or involvement inpeer consultation about one’s supervision asthey relate to changes in supervision practicesor outcomes.

4. There is little specification of what makesa supervisor effective and thus of how onebuilds the skills necessary to become effective.

5. There is little understanding about how asupervisor might change in conceptual ability orin cognitive style as the supervisor gainsexperience.

Conclusion

In reference to the opening metaphor, if gain-ing experience at counseling and supervision islike the aging of wines, then this review uncov-ered two types of wines: counselors and super-visors. One type of wine, the counselor, changesand improves with age. Counselor trainers andsupervisors pay attention to the counselor’s ag-ing and aid his or her development. The othertype of wine, the supervisor, does not clearlyimprove with age. Supervisors appear to beneglected or given minimal attention by mostprofessional environments, yet are expected tochange with age and to age with quality. Theyare like a fine wine, bottled wholly in sterileglass without a cork that allows the wine todevelop robustness. More attention is neededwithin the profession to the maturing of thiswine into fullness.

References

Alonso, A. (1983). A developmental theory of psy-chodynamic supervision. The Clinical Supervisor,1(3), 23–36.

Ard, B. N. (1973). Providing clinical supervision formarriage counselors: A model for supervisor andsupervisee. The Family Coordinator, 22, 91–97.

Barak, A., & LaCrosse, M. B. (1975). Multidimen-sional perception of counselor behavior. Journal ofCounseling Psychology, 22, 471–476.

Barrett-Lennard, G. T. (1962). Dimensions of thera-pist responses as a causal factor in therapeuticchanges. Psychological Monographs, 76 (43,Whole No. 562).

Bartlett, W. E., Goodyear, R. K., & Bradley, F. O.(Eds.) (1983). Special issue: Supervision in coun-seling II. The Counseling Psychologist, 11(1),7–79.

Bernard, J. M. (1979). Supervisor training: A dis-crimination model. Counselor Education and Su-pervision, 19, 60–68.

Bernard, J. M. (1981). Inservice training for clinicalsupervisors. Professional Psychology, 12, 740–748.

Bernard, J. M. (1982). Laboratory training for clin-ical supervisors: An update. Paper presented at theannual meeting of the American Psychological As-sociation, Washington, DC.

Blount, C. M. (1982). A developmental model ofsupervision: Implications for practice and re-search. Paper presented at the annual meeting ofthe American Psychological Association, Wash-ington, DC.

Blumberg, A. (1980). Supervision of teachers (2nded.) Berkeley, CA: McCutchan.

Carkhuff, R. R. (1969). Helping and human rela-tions. New York: Holt, Rinehart & Winston.

Cross, D. G., & Brown, D. (1983). Counselor super-vision as a function of trainee experience: Analysisof specific behaviors. Counselor Education andSupervision, 22, 333–341.

Crowne, D., & Marlowe, D. (1964). The approvalmotive. New York: Wiley.

Dodenhoff, J. T. (1981). Interpersonal attraction anddirect-indirect supervisor influence as predictors ofcounselor trainee effectiveness. Journal of Coun-seling Psychology, 28, 47–52.

Fleming, J. (1953). The role of supervision in psy-chiatric training. Bulletin of the MenningerClinic, 17, 157–159.

Forsyth, D. R., & Ivey, A. E. (1980). Microtraining:An approach to differential supervision. In A. K.Hess (Ed.), Psychotherapy supervision: Theory,research, and practice (pp. 242–261). New York:Wiley.

Friedlander, M. L., & Snyder, J. (1983). Trainees’expectations for the supervisory process: Testing a

158 WORTHINGTON

developmental model. Counselor Education andSupervision, 22, 342–348.

Friedlander, M. L., & Ward, L. G. (1984). Develop-ment and validation of the Supervisory Styles In-ventory. Journal of Counseling Psychology, 31,541–557.

Friedlander, S. R., Dye, N. W., Costello, R. M., &Kobos, J. C. (1984). A developmental model forteaching and learning in psychotherapy supervi-sion. Psychotherapy, 21, 189–196.

Gaoni, B., & Neumann, M. (1974). Supervision fromthe point of view of the supervisee. AmericanJournal of Psychotherapy, 24, 108–114.

Goodyear, R. K., & Robyak, J. E. (1982). Supervi-sors’ theory and experience in supervisory focus.Psychological Reports, 51, 978.

Grotjahn, M. (1955). Problems and techniques ofsupervision. Psychiatry, 18, 9–15.

Handley, P. (1982). Relationship between supervi-sors’ and trainees’ cognitive styles and the super-vision process. Journal of Counseling Psychol-ogy, 29, 508–515.

Hart, G. M. (1982). The process of clinical supervi-sion. Baltimore: University Park Press.

Heppner, P. P., & Handley, P. G. (1981). A study ofthe interpersonal influence process in supervision.Journal of Counseling Psychology, 28, 437–444.

Heppner, P. P., & Handley, P. (1982). The relation-ship between supervisory behaviors and perceivedsupervisor expertness, attractiveness, or trustwor-thiness. Counselor Education and Supervision, 22,37–46.

Heppner, P. P., & Roehlke, H. J. (1984). Differencesamong supervisees at different levels of training:Implications for a developmental model of super-vision. Journal of Counseling Psychology, 31, 76–90.

Hess, A. K. (Ed.) (1980). Psychotherapy supervision:Theory, research, and practice. New York: Wiley.

Hess, A. K. (1986). Growth in supervision: Stages ofsupervisee and supervisor development. The Clin-ical Supervisor, 4, 51–67.

Hess, A. K., & Hess, K. A. (1983). Psychotherapysupervision: A survey of internship training prac-tices. Professional Psychology: Research andPractice, 14, 504–513.

Hester, L. R., Weitz, L. J., Anchor, K. N., & Roback,H. B. (1976). Supervisor attraction as a function oflevel of supervisor skillfulness and supervisees’perceived similarity. Journal of Counseling Psy-chology, 23, 254–258.

Hill, C. E., Charles, D., & Reed, K. G. (1981). Alongitudinal analysis of changes in counselingskills during doctoral training in counseling psy-chology. Journal of Counseling Psychology, 28,428–436.

Hogan, R. A. (1964). Issues and approaches in su-pervision. Psychotherapy: Theory Research andPractice, 1, 139–141.

Holloway, E. L. (1987). Developmental models ofsupervision: Is it development? Professional Psy-chology: Research and Practice, 18, 209–216.

Holloway, E. L., & Wampold, B. E. (1983). Patternsof verbal behavior and judgments of satisfaction inthe supervision interview. Journal of CounselingPsychology, 30, 227–234.

Hutt, C. H., Scott, J., & King, M. (1983). A phenom-enological study of supervisees’ positive and neg-ative experiences in supervision. Psychotherapy:Theory, Research and Practice, 20, 118–123.

Kagan, N. (1980). Influencing human interaction—Eighteen years with IPR. In A. K. Hess (Ed.),Psychotherapy supervision: Theory, research, andpractice (pp. 262–283). New York: Wiley.

Kiesler, D. J. (1986). Interpersonal methods of diag-nosis and treatment. In J. O. Cavenar (Ed.), Psy-chiatry (Vol. 1, pp. 1–23). Philadelphia: Lippin-cott.

Kirchner, K. R. (1975). Expectations about individ-ual counseling supervision. Dissertation AbstractsInternational, 36, 445B. (University MicrofilmsNo. 75–16, 012)

Kurtz, P. D., Marshall, E. K., & Banspach, S. W.(1985). Interpersonal skill-training research: A 12-year review and analysis. Counselor Educationand Supervision, 24, 249–263.

Lambert, M. J. (1974). Supervisory and counselingprocess: A comparative study. Counselor Educa-tion and Supervision, 14, 54–60.

Leddick, G. R., & Bernard, J. M. (1980). The historyof supervision: A critical review. Counselor Edu-cation and Supervision, 19, 186–196.

Littrell, J. M., Lee-Borden, N., & Lorenz, J. A.(1979). A developmental framework for counsel-ing supervision. Counselor Education and Super-vision, 19, 129–136.

Loganbill, C., Hardy, E., & Delworth, U. (1982).Supervision: A conceptual model. The CounselingPsychologist, 10(1), 3–42.

Marikis, D. A., Russell, R. K., & Dell, D. M. (1985).Effects of supervisor experience level on planningand in-session supervisor verbal behavior. Journalof Counseling Psychology, 32, 410–416.

McColley, S. H., & Baker, E. L. (1982). Trainingactivities and styles of beginning supervisors: Asurvey. Professional Psychology, 13, 283–292.

Miars, R. D., Tracey, T. J., Ray, P. B., Cornfeld, J. L.,O’Farrell, M., & Gelso, C. J. (1983). Variation insupervision process across trainee experience lev-els. Journal of Counseling Psychology, 30, 403–412.

Miller, C. D., & Oetting, E. R. (1966). Students reactto supervision. Counselor Education and Supervi-sion, 6, 73–74.


Miller, R. (1982). Supervision: A conceptual model.The Counseling Psychologist, 10(1), 47–48.

Phillips, S. D. (Ed.) (1984). Special issue: Computersin counselor training. Counselor Education andSupervision, 24, 130–221.

Raphael, R. D. (1982). Supervisee experience: Theeffect on supervisor verbal responses. Paper pre-sented at the annual meeting of the AmericanPsychological Association, Washington, DC.

Reising, G. N., & Daniels, M. H. (1983). A study ofHogan’s model of counselor development and su-pervision. Journal of Counseling Psychology, 30,235–244.

Rickards, L. D. (1984). Verbal interaction and super-visor perception in counselor supervision. Journalof Counseling Psychology, 31, 262–265.

Rose, S. D. (1965). Students view their supervision:A scale analysis. Social Work, 10, 90–96.

Rosenblatt, A., & Mayer, J. E. (1975). Objectionablesupervisor styles: Students’ views. SocialWork, 20, 184–188.

Sansbury, D. L. (1982). Developmental supervisionfrom a skills perspective. The Counseling Psychol-ogist, 10(1), 53–57.

Stoltenberg, C. (1981). Approaching supervisionfrom a developmental perspective: The counselorcomplexity model. Journal of Counseling Psychol-ogy, 28, 59–65.

Stone, G. L. (1980). Effects of experience on super-visor planning. Journal of Counseling Psychol-ogy, 27, 84–88.

Strong, S. R. (1968). Counseling: An interpersonalinfluence process. Journal of Counseling Psychol-ogy, 15, 215–224.

Strong, S. R. (1986). Interpersonal influence theoryand therapeutic interactions. In F. J. Dorn (Ed.),Social influence processes in counseling and psy-chotherapy, (pp. 17–30). Springfield, IL: CharlesC Thomas.

Tyler, J. D., & Weaver, S. H. (1981). Evaluating theclinical supervisee: A survey of practices in grad-

uate training programs. Professional Psychol-ogy, 12, 424–437.

Wiley, M. O’L. (1982). Developmental counselingsupervision: Person-environment congruency, sat-isfaction, and learning. Paper presented at the an-nual meeting of the American Psychological As-sociation, Washington, DC.

Worthington, E. L. Jr. (1984a). An empirical inves-tigation of supervision of counselors as they gainexperience. Journal of Counseling Psychology, 31,63–75.

Worthington, E. L. Jr. (1984b). Supervisors as em-pathic observers of counselor behavior. Profes-sional Psychology: Research and Practice, 15,457–461.

Worthington, E. L. Jr., & Roehlke, H. J. (1979).Effective supervision as perceived by beginningcounselors-in-training. Journal of Counseling Psy-chology, 26, 64–73.

Worthington, E. L. Jr., & Stern, A. (1985). Theeffects of supervisor and supervisee degree leveland gender on the supervisory relationship. Jour-nal of Counseling Psychology, 32, 252–262.

Yogev, S. (1982). An eclectic model of supervision:A developmental sequence for beginning psycho-therapy students. Professional Psychology, 13,236–243.

Yogev, S., & Pion, G. M. (1984). Do supervisorsmodify psychotherapy supervision according tosupervisees’ levels of experience? Psychother-apy, 21, 206–208.

Zucker, P. J., & Worthington, E. L. Jr. (1986). Su-pervision of interns and postdoctoral applicants forlicensure in university counseling centers. Journalof Counseling Psychology, 33, 87–89.

Originally published inProfessional Psychology:

Research and Practice,1987, 18, 189–208. �

160 WORTHINGTON

changes in supervision as counselors and supervisors … supervision as experience is... · changes...

Documents