Title Page
 Table of Contents
 List of Tables
 Review of the literature
 Methods and procedures
 Analysis and results
 Biographical sketch

Group Title: exploration of the relationship of ego development theory to counselor development
Title: An exploration of the relationship of ego development theory to counselor development
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00097411/00001
 Material Information
Title: An exploration of the relationship of ego development theory to counselor development
Physical Description: viii, 168 leaves : ; 28 cm.
Language: English
Creator: Borders, Leslie DiAnne, 1950-
Publication Date: 1984
Copyright Date: 1984
Subject: Student counselors -- Training of   ( lcsh )
Ego (Psychology)   ( lcsh )
Developmental psychology   ( lcsh )
Counselor Education thesis Ph. D
Dissertations, Academic -- Counselor Education -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis (Ph. D.)--University of Florida, 1984.
Bibliography: Bibliography: leaves 156-167.
Additional Physical Form: Also available on World Wide Web
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Leslie DiAnne Borders.
 Record Information
Bibliographic ID: UF00097411
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000496924
oclc - 12045551
notis - ACR6151


This item has the following downloads:

PDF ( 6 MBs ) ( PDF )

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
    Table of Contents
        Page iv
        Page v
    List of Tables
        Page vi
        Page vii
        Page viii
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
    Review of the literature
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
    Methods and procedures
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
    Analysis and results
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
    Biographical sketch
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
Full Text








Both professional and personal acknowledgements are due

to a number of persons. Committee members, Dr. Margaret L.

Fong, Dr. P. Joseph Wittmer, Dr. Gregory J. Neimeyer, and

Dr. James Archer, Jr., have each provided needed encourage-

ment, critiques, enthusiasm, and expertise at many points

during the research process. Sixty-three counseling gradu-

ate students volunteered several hours of their valuable

time to complete the instruments in this study; 44 also

contributed a counseling audiotape. Thirty-two supervisors

completed evaluations for their students. Betty Black, Sam

Clark, Dr. Tom Harrison, Myrna Neims, and Sue Williams

contributed intense hours and ratings of professional qual-

ity. Alicia Schmitt assisted as an on-call tutor for com-

puter programming. Each of these persons was a vital link

in producing this dissertation, and their efforts are

greatly appreciated.

Professional acknowledgement is extended to Dr. Jane

Loevinger, Dr. Hans H. Strupp, and Dr. Robert Myrick for

permission to use their instruments in this research.

Personal appreciation is due to several persons in

particular: to Dr. Fong, who challenged and nurtured as my

mentor par excellence; to Peggy, who provided a realistic

perspective on numerous crises and reminded me to celebrate;

to Myrna, who "checked in" without fail to encourage; to

Tom, who coached me through the steps toward the final goal;

and to O, stable and calming mainstay, who shared needed

retreats. Other fellow students, friends, professors,

supervisors, and supervisees often provided insights and


Personal thanks are also extended to those persons "at

home" in North Carolina who nurtured me through my early

professional development, through the decision to move to

Florida and return to graduate school, and then encouraged

me from afar.

Acknowledgement is also extended to family: Rev. and

Mrs. Hugh L. Borders; Fonda, Tony, and LeeAnn Harris; Lynn,

Jay, James and Matthew Layer; and Ken Borders, who accepted

my choices and assumed I'd reach my goals; and to my

heritage of strong women who have each succeeded in their

own way.





LIST OF TABLES . . . . . . .

ABSTRACT . . . . . . . . .



Developmental Models of Supervision
Education . . . . . .
Ego Development . . . . .
Rationale for the Study . . .
Purpose of the Study . . . .
Statement of the Problem. . . .
Definition of Terms . . . .
Overview of Remaining Chapters. .


Supervision Education . . . .
Developmental Models of Supervision
Education . . . . . .
Ego Development . . . . .
Summary . . . . . . .



Subjects . . . . . .
Instruments . . . . . .
Data Collection . . . . .
Null Hypotheses . . . . .
Statistical Analysis . . .


Descriptive Data . . . ...
Results of the Statistical Analyses


. . . 8
. . . 14
. . . 15
. . 15
. . . 16
. . 16
. . 18

. *19



. . . 105
. . . 112


Limitations of the Study . . . .
Discussion . . .. . . . . .
Conclusions, Implications, and Recommenda-
tions for Future Research . . . .
Summary . . . . . . . . .






DATA SHEET . . . . . .


SUPERVISORS . . . . . . . .




REFERENCES . . . . . . . . . .





. 145

. . 147

. 148

















DEVELOPMENT. . . . . . . ... 110

EXPERIENCE . . . . . . . 111


DEVELOPMENT. . . ... . . . 116

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



Leslie DiAnne Borders

August 1984

Chairperson: Dr. Margaret L. Fong
Cochairperson: Dr. P. Joseph Wittmer
Major Department Counselor Education

This study explored the efficacy of the theory of ego

development as a theoretical basis for developmental models

of supervision. A review of the literature of supervision

education indicated a need for a theoretical explanation of

individual differences in counselor development and pointed

to the relevance of ego development to counselor development.

The level of ego development for 63 graduate counseling

students was assessed by the Sentence Completion Test. Mul-

tiple dependent measures of counseling variables were in-

cluded: the structure (cognitive complexity, integration,

meaningfulness) and the content of perceptions of clients

were measured by the Repertory Grid Technique; in-session

counseling behavior with clients, by the Vanderbilt Psycho-

therapy Process Scales; and individual supervisors' ratings

of overall effectiveness, by the Counselor Evaluation Rating


Scale. A series of five separate multiple regression analy-

ses were computed to test the relationship of students' lev-

els of ego development with the structural cognitive measures

of their perceptions and in-session behavior and overall ef-

fectiveness ratings. None of the computed partial correla-

tions were significant at the .05 set-wise criterion level.

A 3 (ego level) X 4 (content category) Chi-square test

of differences for the content of perceptions of clients was

significant, indicating counseling students at higher ego

levels used fewer physical descriptions and more interac-

tional style descriptions. Post-hoc analysis suggested the

interaction of cognitive complexity and integration scores

tended to be related to ego levels. A highly complex rela-

tionship may exist between students' ego levels and their

cognitions about clients.

There were near significant positive relationships of

counseling students' in-session behavior and supervisors'

overall effectiveness ratings with ego levels. Level of ex-

perience appeared to have a mediating effect on scores on

several variables.

The strength of the relationships between levels of ego

development and scores on the counseling variables provides

tentative support for the theoretical hypotheses of the re-

lationships between ego development and counselor development

and suggests further research in this area. The results also

provide some support for developmental models of supervision.

Limitations of the study, implications for counselor develop-

ment, and recommendations for future research are discussed.



Supervision has frequently been conceptualized as an

educational process by which supervisors continue to "teach"

counseling to their supervisees beyond classroom training

in basic skills (Leddick & Bernard, 1980). Theorists and

researchers have offered varied explanations for how stu-

dents learn to become counselors, how supervisors function

as educators in the learning process, and the desired learn-

ing goals of supervision (for example, see Bartlett, Good-

year, & Bradley, 1983; Hart, 1982; Hess, 1980a). Whatever

the role (i.e., teacher, counselor, consultant) ascribed to

the supervisor by theorists, the goal of supervision has

been an educational one: promoting student learning in one

or more areas such as counseling skills, self-awareness, or

counselor-client interactions.

These supervision theories have been criticized for

being primarily adjuncts to counseling theories, however,

rather than distinct theories of supervision education

(Bartlett, 1983); for being "metaphors of experience" rather

than scientifically established theories (Holloway &

Hosford, 1983, p. 75). This view has led some reviewers of

supervision literature to conclude there is a lack of a

theoretical base for supervision education (Holloway &


Hosford, 1983; Leddick & Bernard, 1930; Mahon & Altmann,

1977). Holloway and Hosford (1983), however, have noted

that recent developmental models of supervision do not

equate the counseling and supervision process and relation-

ships, but differentiate "between the process of counseling

and the process of becoming a counselor" (p. 73). They be-

lieve these developmental models offer a framework for a

systematic approach to research which can provide the

empirical evidence for a sound theory of supervision edu-


Developmental Models of Supervision Education

The developmental models describe supervision as a

process of successive stages and highlight the educational

nature of supervision. These models present counselor growth

as a progression of sequential, hierarchial stages, each of

which requires different supervision instruction strategies.

Developmental theorists vary in their base for describing

the stages of counselor growth: Littrell, Lee-Borden, and

Lorenz (1979) focus on the changing roles or functions of

the supervisor; Loganbill, Hardy, and Delworth (1982) de-

scribe the dynamics of the supervisee confronting develop-

mental tasks or issues; Blocher (1983) and Stoltenberg

(1981) discuss environmental and personal variables impact-

ing the learning process. Recent studies based on the

concepts of the developmental models of supervision evidence

some progress in a systematic approach to supervision

research, although few studies have attempted to actually

validate a model.

A recent study based on Stoltenberg's model (Miars,

Tracey, Ray, Cornfeld, O'Farrell, & Gelso, 1983) found

supervisors perceived themselves varying their behaviors

with supervisees at different experience levels. They

perceived themselves offering inexperienced supervisees

(first or second semester practicum) more structure, direc-

tion, support, and teaching, while emphasizing personal

development, client resistance, and transference/counter-

transference issues with more experienced supervisees

(advanced practicum or intern level). The authors concluded

the supervisors were clearly "making some kind of develop-

mental distinctions when describing the nature of their

supervision across various trainee levels" (p. 410). Even

though the variations closely paralleled the optimal super-

vision educational environments proposed by Stoltenberg

(1981), the authors could not specify from their data what

distinctions the supervisors were-using to make decisions

about their supervisory behavior. The supervisors also

seemed to make more "gross developmental distinctions" (p.

410) (two levels) than those postulated by Stoltenberg

(four levels).

Other research related to developmental models has

focused on counselors' perceptions of their supervisors.

Differences were found among counselors at various experi-

ence levels in several of these studies. Worthington and

Roehlke (1979), for example, found beginning counselors

(the only experience level included) preferred direct teach-

ing and structured supervision within a supportive relation-

ship. Cross and Brown (1983) reported counselors at differ-

ent experience levels perceived their supervisors as behaving

differently. Beginning counselors perceived their super-

visors as emphasizing tasks and methods while experienced

counselors felt their supervisors' behaviors created a more

supportive and more intense relationship. During post-hoc

interviews following a longitudinal study by Hill, Charles

and Reed (1981), counselors reported changes in their rela-

tionships with their supervisors over time, moving from

dependency and anxiety about the supervisors' judgments and

evaluations to a more peer-like, consultative relationship.

These studies provide some support for developmental

models of supervision; both supervisees and supervisors

report and/or evidence changes in the supervision process

over time. But the stages of the developmental models of

supervision of Stoltenberg (1981) and Loganbill et al. (1982)

are not based on experience levels alone. Instead they

emphasize individual variations in counselor progress

through the developmental stages and imply personal at-

tributes will affect a counselor's rate of progress.

The developmental models which highlight the dynamics

of the supervisee and the learning process offer some

explanations for the individual variations in development.

Loganbill et al. (1982) assume counselors will function at


different competency levels (stages) with each of the eight

developmental issues they describe. Blocher (1983) believes

one strength of his model is its capacity to adjust to

students' individual differences in learning styles and

current developmental stage. Stoltenberg (1981) also rec-

ognizes variation in counselors' motivations, needs, and

potential resistances within each stage of development. His

model is based on the premise "that there are qualitative

differences in addition to, and not accounted for by, mere

quantitative differences in skill level and the knowledge

of theories" (p. 59). He states that development will vary

"significantly from trainee to trainee" (p. 60), depending

"on the skills and attributes of the trainee" (p. 63).

The confounding effects of individual differences has

been illustrated by Reising and Daniels' (1983) attempt to

validate Hogan's (1964) developmental model, the theoreti-

cal basis for Stoltenberg's (1981) developmental model.

Supervisees at four experience levels rated statements they

might make about themselves and statements they might make

about their supervision needs. The statements were designed

to reflect issues in the first three levels of Hogan's

model. While the supervisees' response patterns supported

Hogan's model of counselor development in general, the

authors concluded that a simple stage model could not ade-

quately describe the complexity of the issues Hogan included.

The authors added, "supervisors may need to go beyond the

simple developmental model and examine how the complex

model's individual issues are organized within each trainee"

(p. 242).

Friedlander and Snyder (1983) included self-efficacy

expectations of counselors at various experience levels in

a study of outcome expectancies of supervision and expec-

tations of supervisor's attributes and role. High levels

of confidence and high levels of expectations that super-

vision would affect clients predicted high expectations for

expert and evaluative supervisors; level of experience, how-

ever, was not predictive of expectations. Like Reising

and Daniels (1983), the authors concluded "individual dif-

fnrences override level of experience" (p. 348). It seems

that experience level alone is not sufficient to explain the

complexities of counselor growth and development.

The attention to individual differences to help explain

counselor development during supervision would be a signifi-

cant contribution to theories of supervision. Reviewers of

supervision research have cited the need to investigate the

impact of the counselor's personal attributes on the learn-

ing process in supervision (Bartlett, 1983; Holloway &

Hosford, 1983; Lambert, 1980; Mahon & Altmann, 1977). The

developmental models, however, have not detailed the personal

attributes which may account for individual variation in

learning (i.e., Blocher, 1983; Stoltenberg, 1981). Stolten-

berg (1981), for example, believes a supervisor should

consider an individual counselor's unique cognitive, motiva-

tional, value, and sensory orientations when creating an

appropriate learning environment, but his model does not

specify how these orientations may affect individual varia-

tions in counselor development. Stoltenberg has urged "a

further delineation of the characteristics of counselors at

different levels" (p. 64) which can better indicate appro-

priate choices of supervision learning environments and can

serve as a basis for future studies evaluating the effec-

tiveness of different supervision techniques. He concluded,

"Once this task has been accomplished for all levels of

counselor trainees, the factors appearing to be most instru-

mental in effecting change during the supervision process

can be operationalized and subjected to empirical scrutiny"

(p. 64).

While the Loganbill et al. (1982) "conceptual model" is

systematic, integrative, and descriptive of supervisee

dynamics, Miller (1982) finds it limited by its failure to

delineate the internal psychological processes evolving in

counselors during development. He believes there are three

such underlying processes or personality traits: conceptual

and ego development, emotional flexibility and congruence,

and awareness. Concurrent development of the three in a

supervisee is essential, developing into a mature profes-

sional over at least a twenty-year span. Miller believes

a supervisor's use of "catalytic interventions will facili-

tate conceptual, emotional, and awareness shifts that will

be marked by movement through the various developmental

stages" (p. 48) which have not been adequately described by


Both Miller (1982) and Blocher (1983) name stage

theorists whose work serves as the basis for their develop-

mental approaches to supervision. Blocher (1983), however,

does not integrate these theories into a sequential model

of supervision, and Miller (1982) presents only a very

concise overview of an integration of stage theory and

counselor development. In addition, no studies to date have

attempted to link counselor development with an established

theory of personality development.

In sum, the individual personal attributes of a super-

visee which could be critical for supervisors to consider

in the learning process have not been specified in a de-

tailed, comprehensive developmental model by either theo-

rists or researchers. In addition, none of the developmental

models have incorporated in detail a theoretical basis (i.e.,

personality theory) which could explain what internal

psychological changes occur and how they occur in counselors

as they integrate theoretical knowledge from the classroom

with practical experience in practicum and internship


Ego Development

Loevinger's (Loevinger, 1966, 1976) concept of ego

development and stages offers a comprehensive theoretical

framework of personality development which could serve as

a basis for counselor development. Loevinger defines ego

development as the "master trait" which reflects a person's

frame of reference for perceiving, interpreting, and react-

ing to others and one's life experiences.

Loevinger has integrated theories of self, cognitive,

character, moral and interpersonal development (Hauser,

1976; Loevinger, 1976; Swenson, 1980a) into a stage theory

which is marked by increasing differentiation and integra-

tion of views of the world, others, and self, and a shift

from an external to an internal focus. For instance, a

person's interpersonal style moves from dependency to

manipulation to belongingness to mutuality. Loevinger

describes this development in a series of ten major stages

and transition levels labeled by somewhat descriptive terms,

as summarized in Table 1.

One's perceptions of the world, others, and self are

reflected in a cognitive, moral, and interpersonal style

characteristic of a level of ego development. Loevinger's

(Loevinger & Wessler, 1970) detailed descriptions of the

cognitive and interpersonal styles of persons at different

levels of ego development imply counselors at different ego

levels would have varying capacities to, among other things,

express empathy, respect a client's differentness, deal

with identity issues, and understand the interactive dynam-

ics of the counselor-client relationship.

A substantial body of literature has offered evidence

of construct validity for the theory of ego development,



Code Name and Characteristics of Stage

I-0 Presocial. The stage characteristic of newborn
infants and severely regressed psychotics, in
which a person is unable to distinguish self from
others. Behavior is motivated by immediate

I-1 Symbiotic. A prespeech stage in which a person
does not clearly differentiate self from others.
Interpersonal relations are symbiotic and behav-
ior is still motivated by immediate impulses.

I-2 Impulsive. A stage in which a person begins to
establish a separate identity; is dependent,
demanding, egocentric and conceptually simplis-
tic; and sees others as primarily sources of
supply. Behavior is governed by rules and pun-
ishment; problems are external; and recognized
emotions are in a limited range.

Delta Self-protective. A stage in which the main motiva-
tion is not getting caught and control is a major
theme. A person is opportunistic, manipulative,
and hedonistic, but also less impulsive and depen-
dent. Problems are blamed on others or circum-

Delta/3 (Unnamed). A transitional stage in which the most
predominant theme is concrete, traditional sex
roles. A person's obedience and conformity to
social norms is based in simple and absolute rules.
Emotions are quasi-physiological; cleanliness and
physical appearance are stressed.

I-3 Conformist. A stage in which a person's major
motivation is following the rules, being concerned
with reputation, social acceptance, appearance,
and disapproval. Others are described in stereo-
typic terms and moralistic cliches, and one's
inner life is viewed in generalized and banal
terms. Interpersonal interaction is described in
terms of behaviors.

TABLE 1. Continued.

Code Name and Characteristics of Stage

I-3/4 Self-aware. A transitional stage in which a per-
son allows exceptions to rules and perceives
multiple possibilities in situations, although
still in terms of stereotypic categories. The
increase in self-consciousness and self-awareness
allows some differentiation of self from the group.
Interpersonal interactions are described in terms
of feelings or traits. This is probably the modal
level of young adults (ages 18-25) (Loevinger &
Knoll, 1983).

I-4 Conscientious. A stage in which a person is pre-
occupied with achievement and responsibility, and
self-criticism and self-evaluation by internalized
moral principles. A person has a richly differ-
entiated inner life, a sense of a longer time
perspective, and a greater ability to differen-
tiate between others. Psychological causality is
understood. Interpersonal relations are based on
the deeper feelings and needs of others. This may
be the modal stage for most graduate students
(Swenson, 1980b).

I-4/5 Individualistic. A transitional stage in which a
person is increasingly aware of conflicts between
others' needs and one's own needs, and of their
own internal conflicts. Major themes are a
heightened sense of individuality and a concern
for emotional independence. Interpersonal rela-
tionships are valued and are seen as continuing
or changing over time. There is a distinction
between process and outcome.

I-5 Autonomous. A stage of cognitive complexity in
which a person is able to transcend polarities and
appreciate paradoxes. Individual differences are
cherished; conflict is accepted as part of the
human condition. A person recognizes the need of
others for autonomy and allows others to be her-
self or himself. A person has a high tolerance
for ambiguity; is preoccupied with self-fulfillment;
and wants to be realistic, objective, and unpreju-

TABLE 1. Continued.

Code Name and Characteristics of Stage

I-6 Integrated. A stage in which a person promotes
the growth and development of others, having ac-
cepted and learned to cope with complexities and
paradoxes. The major focus is achieving an inte-
grated identity. There is existential humor,
value for justice, and idealism, and reconcilia-
tion to one's destiny. Few people have achieved
this stage, which resembles Maslow's self-actual-
ized person.

Note. Sources: Loevinger, 1979; Loevinger & Wessler, 1970;
Swenson, 1977, 1980b.

including sequentiality- and descriptions of behaviors and

attitudes characterizing each stage (IIauser, 1976; Loevinger,

1979; Loevinger & Wessler, 1970).

Several authors have commented on ego development

theory's relevance to counseling and counselor development

(Blocher, 1983; Miller, 1982; Swenson, 1977, 1980a). Swenson

(1977), for example, views ego development as "the central

personality variable in interpersonal relations" (p. 41).

Both Miller (1982) and Blocher (1983) specify Loevinger as

one of the stage theorists whose work serves as the bases

for their developmental approaches to supervision. Miller's

(1982) developmental sequence closely parallels Loevinger's

stages, and his final stage, also named "integration," is

characterized by conceptual and ego complexity, emotional

flexibility and congruence, and full awareness. Blocher's

(1983) definition of a counselor at a high level of cogni-

tive functioning, though couched in information-processing

terms, also closely resembles Loevinger's description of the

Integrated (1-6) person. His more cognitively complex

counselor is able to take multiple perspectives and offer

empathic understanding to a wider variety of clients because

of their greater ability to differentiate and integrate a

wide range of information. Blocher, however, does not out-

line a stage sequence of counselor development in his ap-

proach. Stoltenberg's (1981) descriptions of counselor

characteristics at his four levels of development also

reflect characteristics of levels of ego development, al-

though Stoltenberg does not refer to Loevinger's theory.

Rationale for the Study

Ego development theory could provide a framework for

understanding, charting, and facilitating counselor develop-

ment within the supervision process. It describes cogni-

tive and interpersonal orientations which could have impor-

tant implications for counselor education and supervision.

It describes issues counselors at different ego levels may

be confronting within themselves, issues which may color

their perceptions of and attitudes toward clients. It

describes both basal and optimal developmental levels for

counselors' ability to understand clients and their issues,

and for their ability to "perform" certain counseling skills.

By viewing supervision as an educational process and

raising a supervisee's level of ego development as one of

its educational goals, we have a potential basis for a

comprehensive theory of supervision education. Ego devel-

opment may provide a theoretical basis for describing sequen-

tial stages of counselor growth in both the cognitive and

interpersonal domains. Theoretical application of ego

development to supervision to date has been fragmentary and

incomplete (i.e., Blocher, 1983; Miller, 1982), or implied

rather than explicitly stated (i.e., Stoltenberg, 1981).

Lacking is a comprehensive application of each level of ego

development to counselor cognitive and interpersonal orien-


What seems to be needed is a descriptive list of per-

ceptions and behaviors of counseling students at each level

of ego development. Such a detailed description could pro-

vide a representation of progressive stages of counselor

development and a way to tap into the internal changes in

perceptions and attitudes counselors experience during their

supervision experiences. It could provide a description of

desired outcomes--the counselor perceptions, attitudes, and

behaviors supervisors are aiming towards--and a way to assess

the effectiveness of supervision education. It could also

provide a measure to assess a counselor's progress and


Purpose of the Study

The purpose of the proposed study is to explore the

usefulness of applying the theory of ego development to

counselor development and supervision education. The study

will examine the theory's ability to differentiate between

counseling students at different ego levels in terms of

perception of their clients, behaviors with their clients,

and counseling effectiveness with their clients.

Statement of the Problem

The proposed study will address the need to describe

the effects of counseling students' personal attributes on

their learning process, and the need for a theoretical basis

to explain the psychological, perceptual and behavioral

changes in counseling students during the learning process.

More specifically, the study will attempt to answer the

following research questions:

1. Is there a relationship between students' levels of

ego development and the level of cognitive complexity of

their perceptions of their clients?

2. Is there a relationship between students' levels of

ego development and the level of cognitive integration of

their perceptions of their clients?

3. Is there a relationship between students' levels of

ego development and the level of meaningfulness of their

perceptions of their clients?

4. Is there a relationship between students' levels of

ego development and their behavior with their clients?

5. Is there a relationship between students' levels of

ego development and supervisors' ratings of students' effec-

tiveness with their clients?

6. Is there a relationship between students' levels of

ego development and the content of their perceptions of

their clients?

Definition of Terms

Ego development refers to the central concept of per-

sonality in Loevinger's (1966, 1976; Loevinger & Wessler,

1970) developmental stage theory which is marked by increas-

ing differentiation and integration of perceptions, a shift

from an external to an internal focus, and complimentary

changes in interpersonal behaviors.

Level of ego development refers to one of the ten stages

or transitional levels in the -stage theory and is measured

in this study by scores on the Sentence Completion Test

(SCT) (Loevinger & Wessler, 1970; Loevinger, Wessler, &

Redmore, 1970).

Level of cognitive complexity refers to the degree of

differentiation in a person's perceptions or constructs as

defined by Kelly (1955) and Landfield (1977). It is meas-

ured in this study by Differentiation (FIC) scores on the

Repertory Grid Technique (Fransella & Bannister, 1977).

Level of cognitive integration refers to the hierarchial

arrangement of a person's perceptions or constructs as de-

fined by Kelly (1955) and Landfield (1977). It is measured

in this study by Ordination scores on the Repertory Grid

Technique (Fransella & Bannister, 1977).

Level of meaningfulness refers to the degree of polari-

zation of the perceptions or constructs a person uses to

describe self or others, as defined by Kelly (1955) and

Landfield (1977). It is measured in this study by Extremity

scores on the Repertory Grid Technique (Fransella &

Bannister, 1977).

Behavior with clients refers to the counselors' in-

session interactions with clients, including levels of

warmth, emotional involvement, positive attitudes, and

attempts to examine clients' underlying psychodynamics. It

is measured in this study by scores on the Vanderbilt

Psychotherapy Process Scales (VPPS) (O'Malley, Suh, &

Strupp, 1983; Strupp, 1981).

Effectiveness with clients refers to counselors' per-

formance during supervision and counseling sessions and in-

cludes behavior, knowledge, and attitude. It is measured

in this study by individual supervisors' ratings of coun-

selors' effectiveness on the Counselor Evaluation Rating

Scale (CERS) (Myrick & Kelly, 1971).

Content of perceptions refers to four categories of

counselors' perceptions of their clients (i.e., physical

characteristics or factual information, interactional style,

roles or habitual activities, psychological or cognitive

attributes). It is measured in this study by a classifica-

tion system devised by Duck (1973).

Level of experience refers to the amount of training

the supervisee has completed, i.e., registration for a first

practicum, second practicum, or internship experience.

Overview of Remaining Chapters

Relevant theoretical literature and empirical research

is reviewed in Chapter II. The methodology used in the

study is described in Chapter III. Descriptive data and

results of the statistical analyses are presented in Chapter

IV. Discussion of the results, implications, and conclu-

sions are covered in Chapter V.


Chapter II describes the current developmental models

of supervision education and reviews research investigating

the dimensions of the developmental models, including related

literature exploring individual differences between super-

visees affecting their learning and performance. The chap-

ter also describes the theory of ego development and its

stages, reviews validation research of its theoretical as-

sumptions, and compares and contrasts ego development with

other developmental theories relevant to counselor develop-

ment and supervision education.

Supervision Education

Supervision has frequently been conceptualized as an

educational process by which supervisors continue to "teach"

counseling to their supervisees. Theorists and researchers

have offered varied explanations for how students learn to

become counselors and how supervisors function as educators

in the learning process. Leddick and Bernard's (1980)

historical review of dynamic, facilitative, and behavioral

supervision theories, for example, reveals a consistent

view of the supervisor as educator. Other authors have

discussed supervision in terms of the roles or functions of


the supervisor, including teaching, counseling, and consult-

ing (Bernard, 1979; Boyd, 1978; Gurk & Wicas, 1979; Hess,

1980b; Littrell et al., 1979). Hart (1982) has grouped

numerous supervision approaches into three categories:

skill development (supervisor as teacher), personal growth

(supervisor as counselor), and integrative (supervisor as

consultant) models. He also discusses four goals specific

to supervision's educative function: clinical knowledge and

skills, professional identity, case conceptualization, and

self-awareness. Other writers have described how super-

visors with different theoretical orientations to counseling

facilitate counselors' learning of skills and/or qualities

specified by the counseling theory (for example, see

Bartlett et al., 1983; Hess, 1980a). Thus, whatever the role

described for the supervisor by theorists, the goal of

supervision has been an educational one: promoting student

learning in one or more areas such as counseling skills,

self-awareness, or counselor-client interactions.

These supervision theories have been criticized, how-

ever, for being primarily adjuncts to counseling theories

rather than distinct theories of supervision education

(Bartlett, 1983); for being "metaphors of experience" rather

than scientifically established theories (Holloway &

Hosford, 1983, p. 75). This view has led some reviewers of

supervision literature to conclude there is a lack of a

theoretical base for supervision education (Holloway &

Hosford, 1983; Leddick & Bernard, 1980; Mahon & Altmann,

1977). Holloway and Hosford (1983), however, have noted

that recent developmental models of supervision do not

equate the counseling and supervision processes and rela-

tionships but differentiate "between the process of coun-

seling and the process of becoming a counselor" (p. 73).

They believe these developmental models offer a framework

for a systematic approach to research which can provide the

empirical evidence for a sound theory of supervision educa-


Developmental Models of Supervision Education

The developmental models of supervision education are

based in the work of cognitive-developmental stage theorists

such as Hunt (conceptual systems), Kohlberg (moral judgment),

Loevinger (ego development), and Perry (intellectual-ethical

development). The movement toward cognitive-developmental

theories has been described as a "genuine advance" in under-

standing personality development and individual differences

(Loevinger & Knoll, 1983). Swenson (1977, 1980b) believes

the study of cognitive processes will provide a general

paradigm for psychology and a single framework for psycho-

therapy. Miller and Parker (in Kohlberg & Wasserman, 1930)

describe the cognitive-developmental approach as a "paradigm

shift" in counseling and guidance; they invited major

theorists to speak at a special conference on the implica-

tions for the field in 1977.

The recent descriptions of supervision as a develop-

mental process highlight the educational nature of super-

vision. These models describe counselor growth as a series

of sequential, hierarchial stages, each of which requires

different supervision instructional strategies. The four

predominant developmental models vary in their base for

describing the stages: Littrell et al. (1979) focus on the

roles of functions of the supervisor; Loganbill, Hardy, and

Delworth (1982), the dynamics of the supervisee; Blocher

(1983) and Stoltenberg (1981), the dynamics of the learning


The developmental framework presented by Littrell et

al. (1979) focuses on the changing roles or behaviors of the

supervisor in responding to the supervisee's stage of devel-

opment. The authors suggest supervision moves through four

stages characterized primarily by the changing tasks of the

supervisor: establishing a working relationship, goals and

contract; counseling and/or teaching models; consulting

model; self-supervision model. Supervisees gradually assume

greater responsibility for their learning, so that in the

final self-supervision stage the supervisee conceptualizes,

implements, controls, and manages supervision. The model

of Littrell et al. thus describes changing roles, responsibili-

ties, and behaviors of the supervisor and supervisee rather

than changes in the supervisee as a person.

In their developmental model, Loganbill et al. (1982)

focus instead on the dynamics of supervisees. They propose

supervisees confront a series of eight "supervisory issues"

(i.e., competence, emotional awareness, autonomy, theoreti-

cal identity, respect for individual differences, purpose

and direction, personal motivation, professional ethics)

based on Chickering's developmental tasks of college stu-

dents. The authors view counselor development as continuous

and on-going, with counselors recycling through these issues

by stages of stagnation, confusion, and integration. They

suggest supervisors use facilitative, confrontive, concep-

tual, prescriptive, and catalytic interventions to facili-

tate counselor growth through the stages. Because of

individual differences (not specified by the authors),

counselors may function at different competency levels for

each developmental issue. The Loganbill et al. model does

not propose an overall stage sequence, and it does not of-

fer a theoretical basis (i.e., personality theory) to

explain how changes occur.

In contrast, Blocher's (1983) cognitive-developmental

approach to supervision draws from theories of the psychology

of learning (e.g., Lanning, Kell and Mueller, Strong,

Krumboltz) and cognitive development (e.g., Piaget,

Loevinger, Kohlberg, Perry, Harvey, Hunt and Schroder). He

defines supervision as "a specialized instructional process

in which the supervisor attempts to facilitate the growth

of a counselor-in-preparation, using as the primary educa-

tional medium the student's interaction with real clients

for whose welfare this student has some degree of

professional, ethical, and moral responsibility" (p. 27).

This specialized instruction is "psychological education in

the fullest and most complete sense of the term" (p. 28).

Blocher's (1983) cognitive-developmental approach to

supervision education is designed to encourage cognitive

growth of a counselor toward more complex, comprehensive,

and integrative thinking. Growth in the cognitive schemas

used for information-processing of human interaction is

viewed as a prerequisite for changes in feelings and behav-

iors. Higher order cognitive schemas are encouraged through

a developmental learning environment balancing challenge and

support, innovation and integration. Blocher believes one

strength of his model is its capacity to adjust to students'

individual differences in learning styles and current devel-

opmental stage. The model as presented does not specify

these adjustments or developmental stages of growth but

focuses on the dynamics of the learning environment which

will promote counselor cognitive growth.

Stoltenberg (1981) describes a developmental model

which combines supervision and instructional theories and

highlights the supervisor-supervisee relationship. His

model integrates Hogan's (1964) concise four-level develop-

mental model of supervision and Hunt's (1971) levels of con-

ceptual development, and then applies Hunt and Sullivan's

(1974) person-environment matching approach to instruction.

The resulting "counselor complexity model" is a description

of cognitive and personality characteristics of counselors

at each of the four levels and optimal matching learning

environments for supervision at each level. The focus of

the levels is on the issues of dependency and autonomy,

describing counselors as moving from a beginning level

characterized by dependency on the supervisor (Level 1)

through successive levels of dependency-autonomy conflict

(Level II), conditional dependency (Level III), to a will-

fully interdependent master counselor (Level IV). Optimal

learning environments are matched to the developmental needs

of counselors at each level and are designed to facilitate

a counselor's advancement to the next stage. These super-

vision environments are characterized by a decrease in the

structure and instruction provided by the supervisor, and

an increase in a peer-like, collegial relationship between

counselor and supervisor.

Stoltenberg (1981) specifies "no specific time table

of progress" (p. 60) and does not tie his developmental

levels to experience levels exclusively. His model thus

recognizes "the different motivations, needs, and potential

resistances of counselors at different levels or stages of

development" (p. 59). He states that development will vary

"significantly from trainee to trainee" (p. 60), depending

"on the skills and attributes of the trainee" (p. 63).

Stoltenberg (1981) delineates four factors a supervisor

should consider as a basis for discriminating between coun-

selors as unique persons and choosing appropriate inter-

ventions: the counselor's cognitive, motivational, value,

and sensory orientations. While the latter is more descrip-

tive of the counselor's learning style and so is more

directly applicable to choice of supervision instructional

mode, the first three orientations address relevant cogni-

tive and personal characteristics. A counselor's cognitive,

motivational, and value orientations are three variables

which may explain in part the individual variations in

counselor development this model assumes. Stoltenberg urges

"a further delineation of the characteristics of counselors

at different levels" (p. 64) which can better indicate ap-

propriate choices of supervision learning environments and

can serve as a basis for future studies evaluating the ef-

fectiveness of different supervision techniques. He

concludes, "Once this task has been accomplished for all

levels of counselor trainees, the factors appearing to be

most instrumental in effecting change during the supervision

process can be operationalized and subjected to empirical

scrutiny" (p. 64).

Miller (1982) has made a similar point. He has criti-

cized developmental models of supervision for their lack of

attention to the psychological processes evolving in coun-

selors during development. He calls attention to three

underlying processes (i.e., conceptual and ego development,

emotional flexibility and congruence, awareness) which he

believes are crucial to understanding and facilitating

counselor growth. In reviews of supervision research,

Holloway and Hosford (1983) and Lambert (1980) have called

attention to the need for descriptive, exploratory studies

to identify relevant supervisee variables, effective super-

vision techniques, and desired supervision outcomes.

In summary, while developmental models of supervision

education vary in their central focus (i.e., supervisors'

roles, supervisee dynamics, learning process), each describes

a common thread of counselor development. Counselors are

assumed to have different supervision needs, be concerned

with different issues about counseling, and have different

levels of self-awareness at various stages of development.

Most developmental models assume counselor development depends

not only on counseling experience level, but more importantly

on individual differences or personal attributes of the

counselors themselves. Although the developmental models

emphasize individual differences, delineating what the

relevant differences are is an incomplete task at present

(Miller, 1983; Stoltenberg, 1981).

Research Related to Developmental Models of Supervision

Few published studies to date have been tied directly

to a particular developmental model. In addition, most

supervision-related research has primarily used beginning

practicum students (Hart, 1982; Holloway & Hosford, 1983;

Lambert, 1980) and investigated the acquisition of basic

counseling skills (Lambert, 1980), so that little is known

empirically about supervision of counselors at advanced


levels of training (Lambert, 1980). Some cross-sectional

and longitudinal studies have addressed issues relevant to

developmental models. Most have investigated counselors'

preferences for or perceptions of supervision; only a few

have identified developmental changes in behaviors of coun-

selors at different levels of experience (Reising & Daniels,

1983). Like the developmental theorists, these research

studies often point to the need to attend to individual

differences of counselors.

A recent study based on Stoltenberg's model (Miars,

Tracey, Ray, Cornfeld, O'Farrell, & Gelso, 1933) found super-

visors perceived themselves varying their behavioi6 with

supervisees at different experience levels. They perceived

themselves offering inexperienced supervisees (first or

second semester practicum) more structure, direction, sup-

port, and teaching,while emphasizing personal development,

client resistance, and transference/countertransference

issues with more experienced supervisees (advanced practi-

cum or intern level). The authors concluded the supervisors

were clearly "making some kind of developmental distinctions

when describing the nature of their supervision across

various trainee levels" (p. 410). Even though the varia-

tions closely paralleled the optimal supervision educational

environments proposed by Stoltenberg (1981), the authors

could not specify from their data what distinctions the

supervisors were using to make decisions about their super-

visory behavior. The supervisors also seemed to make more

"gross developmental distinctions" (p. 410) (two levels)

than those postulated by Stoltenberg (four levels).

Counselors' views of supervision across experience levels

Counselors at various levels of experience have also

reported different perceptions of or preferences for super-


Littrell (1978), for example, factor analyzed the con-

cerns of beginning counselors. He found the prepracticum

counselors most concerned about learning techniques and

meeting client needs, moderately concerned about the coun-

selor role and adequacy, and least concerned about whether

clients like them.

Worthington and Roehlke (1979) investigated beginning

counselors' perceptions of effective supervision. First

practicum counselors indicated they preferred direct teach-

ing (i.e., modeling, feedback, literature) within a sup-

portive relationship. They felt they had improved their

counseling as a result of these qualities in addition to

supervisors' support and encouragement of risking new

behaviors and to develop their own styles. They described

competent supervisors as having greater experience, skill,

and self-assurance.

Worthington (1984) conducted a similar study of coun-

selors at five experience levels from several geographical

areas using a cross-sectional approach. The supervisory

behaviors related to satisfaction with supervision and

perceived competence of the supervisor were the same behav-

iors identified by the beginning counselors in the previous

study. There were some differences between experience

levels, such as the more frequent focus on content with

less experienced supervisees, more frequent use of consulta-

tion with internship supervisees, and more frequent en-

couragement to find their own style of counseling with more

experienced supervisees.

Heppner and Roehlke (1984) used the same supervisory

questionnaire (see Worthington & Roehlke, 1979) with begin-

ning practice, advanced practice, and internship students,

and reported similar differences by experience level for

ratings of supervisory behavior which correlated with stu-

dents' satisfaction. They also examined the critical inci-

dents or turning points classifiedd by the supervisory issues

of Loganbill et al., 1982) reported by students at each of

the three experience levels. These findings paralleled

the differences in preferences for supervisory behaviors.

Beginning and advanced practice students described critical

incidents related to self-awareness issues; internship

students, those related to personal issues affecting therapy.

Moskowitz (1981) compared supervision preferences of

counselors at several experience levels. Beginning clinical

psychology students preferred an imitative approach, empha-

sizing learning techniques and role modeling, while more

advanced students preferred a therapist-centered approach,

emphasizing exploration and attention to therapist role

difficulties. Beginning students also indicated a preference

for more direction and less focus on errors than advanced

students. Nash's (1975) results suggested similar sequen-

tial differences, with students moving from a preference for

advice, technical suggestions, and support to a preference

for more discussion of theoretical and countertransference

issues. In related literature, a review of teacher educa-

tion research indicated a similar pattern: student teachers'

preferences for directive supervision decreased over time

while experienced teachers consistently preferred nondirec-

tive supervision (Copeland, 1982).

Cross and Brown (1983) found counselors at different

experience levels also perceived their supervisors behaving

differently. Counselors indicated the frequency of their

supervisors'behaviors on Worthington and Roehlke's (1979)

list. Beginning counselors perceived their supervisors

emphasized tasks and methods (i.e., observations, audio-

tapes), while more experienced counselors perceived their

supervisors emphasized both a more supportive and more

intense relationship. The latter indicated they received

more confrontation and feedback, saw their supervisors as

more competent, and said supervision contributed more to

improved client outcome and increased counselor self-

confidence than did the beginning counselors.

In post-hoc interviews of a longitudinal study, coun-

selors reported similar changes in their relationships with

their supervisors, moving from dependency and anxiety about

the supervisor's judgments and evaluations to a more peer-

like, consultative relationship (Hill, Charles, & Reed,


Counselors' behaviors across experience levels

Other studies have identified developmental changes in

behaviors of counselors at different levels of experience.

Cicchetti and Ornston (1976) reported a series of studies

analyzing the structure and content of responses of novice

and experienced psychotherapists. Novices asked more ques-

tions during an initial interview, but with training the

number of questions they asked was not different from ex-

perienced psychotherapists. Content analysis of responses

to a filmed client revealed significant differences between

the two groups, with novices using more concrete statements

while experienced psychotherapists used more abstract com-

ments which did not focus on the exact words or actions of

the client. The experienced psychotherapists' responses

instead seemed to attempt an integrated understanding of

the client's messages about self.

Hill et al. (1981) conducted a three-year longitudinal

study of twelve counseling psychology students and found

changes in their verbal responses in brief counseling ses-

sions with volunteer clients. The students used more min-

mal encouragers (i.e., simple agreement or understanding)

but fewer questions (open and closed) over time. No dif-

ferences were found for the use of directives (i.e.,

information, direct guidance) and results for the use of

complex responses (i.e., interpretation, confrontation) were

inconsistent, perhaps a result of the brevity of the coun-

seling session, the researchers concluded. In post-hoc

interviews, the students reported they felt they had im-

proved most in higher order skills such as timing, appro-

priateness of intervention, client dynamics, and techniques

for special populations.

These studies provide some support for a developmental

model of supervision; both supervisees and supervisors report

and/or evidence changes over time. Other studies, however,

have pointed to the effect of individual differences of

counselors as mediating variables for level of experience,

an effect assumed in several developmental models (i.e.,

Blocher, 1983; Loganbill et al., 1982; Stoltenberg, 1931).

Individual counselor variation

The confounding effects of individual differences have

been illustrated by Reising and Daniels' (1983) attempt to

validate Hogan's (1964) developmental model. Practicum,

advanced practicum students, interns, and professional

staff rated statements they might make about themselves

(e.g., "I lack the experience to know what to do with my

clients") and statements they might make about their super-

vision needs (e.g., "I need a supervisor who will be con-

frontive with me"). The statements were designed to reflect

issues in the first three levels of Hogan's model. A factor

analysis revealed less experienced trainees were in general

more anxious, dependent, and technique-oriented, and less

ready for confrontation, while more experienced counselors

were more independent, as the-theory described. The pro-

fessional staff, however, also reported more commitment

ambivalence, contrary to expectations based on Hogan's model.

They identified doubts about their own skill and the help-

fulness of counseling. While the respondents' response

patterns supported Hogan's model of counselor development

in general, the authors concluded that a simple stage model

could not adequately describe the complexity of the issues

hogan included. They added, "supervisors may need to go

beyond the simple developmental model and examine how the

complex model's individual issues are organized within each

trainee" (p. 242).

Grayson (1979) reported tolerance of ambiguity related

more strongly to counselors' preferences for supervision

than experience level. For counselors ranging from no

practicum experience to those in their second year of intern-

ship, higher levels of tolerance of ambiguity correlated

with a preference for experiential supervision; low levels

of tolerance, with didactic supervision.

In addition, Friedlander and Snyder (1983) included

self-efficacy expectations of counselors at various experi-

ence levels in a study of outcome expectancies of super-

vision and expectations of supervisor's attributes and role.

High levels of confidence and high levels of expectations

that supervision would affect clients predicted high ex-

pectations for expert and evaluative supervisors; level

of experience, however, was not predictive of expectations.

Like Reising and Daniels (1983), the authors concluded

"individual differences override level of experience" (p.


Counselor skill acquisition. The need to attend to

individual differences has also been of concern to reviewers

of research on skill acquisition who have cited the need to

investigate the impact of counselors' personal attributes on

the learning process (Bartlett, 1983; Holloway & Hosford,

1983; Lambert, 1930; Mahon & Altmann, 1977). Counseling

students' cognitive complexity, for example, has been shown

to affect their learning. Rosenthal (1977) investigated

the effect of first semester counseling students' conceptual

level on learning confrontation skills taught by two ap-

proaches. Self-instruction materials were more effective

with high conceptual level students, while guided instruc-

tion was more effective with low level students. Rosen-

thal concluded counselor educators needed to consider the

possible interactions of students' personality characteris-

tics with training approaches.

Reviewers have also cited research suggesting counselor-

trainees neither retain nor transfer skills to other settings

(Lambert, 1980; Leddick & Bernard, 1980; Mahon & Altmann,

1977). Spooner and Stone (1977), for example, evaluated

tapes of student counselors during prepracticum, practicum,

and three months after training. In the follow-up tapes,

students used less confrontation and goal setting, con-

sidered more advanced skills,'and used more questions,

considered a basic skill.

Mahon and Altmann (1977) have concluded individual

variables may influence the effects of training more than

a specific training procedure. They recall that Rogers'

core conditions describe feelings and attitudes of the

helper, not the specific skills operationalized from these

qualities by Truax and Carkhuff and others. "Personal

qualities underlying and unifying 'skills' need as much or

more emphasis as the skills themselves" (p. 49), they

added. This conclusion echoes statements made by develop-

mental theorists (i.e., Blocher, 1983; Miller, 1982;

Loganbill et al., 1982; Stoltenberg, 1981) and reviewers of

supervision research (Holloway & Hosford, 1983; Lambert,

1980) who have also cited the need to attend to individual

differences of counselors.

Sources of individual counselor variation. Mahon and

Altmann (1977) proposed counselor educators use the tenets

of perceptual psychology and the "self as instrument" con-

cept (Combs, 1969), a concept which might explain individual

differences in counselors. They believe perceptions under-

lying behaviors--beliefs, attitudes, and values--are the

integrating learning forces. "It is not the skills them-

selves which are all important, it is the control of their

use, the intentions with which they are used; and their

flexibility or changeability that is so crucial" (p. 49),

they explained. According to this view, changes in per-

ceptions would be necessary for changes in behavior to be

real and long-lasting, and more and less effective counselors

would be expected to differ in their perceptions of self,

others, and the helping process. Such perceptual differences

have been found in research of counselors, teachers, and

pastors using an instrument based on Combs' (1969) concept

(Brown, 1970; Combs, 1969; Dedrick, 1972; Rotter, 1972;

Vonk, 1970). Other studies using measures of cognitive

complexity, a concept which in part reflects the degree of

perceptual flexibility, have found similar differences

between more and less complex counselors' verbal responses

to clients (Goldberg, 1974; Lichtenberg & Heck, 1979) and

their empathy skills (Heck & Davis, 1973; Lutwak &

Hennessy, 1982).

Research of counselor attitudes, a possible source of

individual differences, has also illustrated the impact of

counselors' perceptions of others. For example, stereo-

typing has been found to interfere with counselors' process-

ing of information about ethnic minority clients (Wampold,

Casas, & Atkinson, 1981) and homosexual versus heterosexual

individuals (Casas, Brady, & Ponterotto, 1983). In addi-

tion, Hirsch and Stone (1982) found trainee's attitudes

toward a particular counseling skill affected performance

of that skill. After brief training, student volunteers

with positive attitudes toward reflection of feeling gave

higher quality reflections to coached clients than volun-

teers with negative attitudes. Such a relationship was not

found for interpretive responses, however.

Studies of personality traits which might explain

individual differences in counselor effectiveness have pro-

duced mixed results (Kaplan, 1983). These studies used

global measures of personality, such as the California

Personality Inventory, Omnibus Personality Inventory, and

the 16 PF, and correlated them with overall effectiveness

ratings made by clients or supervisors. Effective counsel-

ing students had those general qualities expected in good

counselors; they were more introspective, aesthetically

sensitive, emotionally expressive, flexible, tolerant, in-

dependent, and capable of intimate relationships. Much like

Mahon and Altmann (1977), Kaplan (1983) concluded, "Overall

the research on personality attributes supports the notion

that it is not what trainees possess that is important, but

rather how they use it in interacting with the client"

(p. 223).


The need to attend to individual differences and per-

sonal attributes of counselors has thus been a theme in

developmental models of supervision and critiques of them,

in research classifying students by experience level, in

related research of counselors' skill acquisition, and in

research on counselors' personal attributes. Developmental

theorists, researchers, and reviewers of supervision research

believe an awareness of counselors' personal attributes will

enhance a supervisor's understanding of counselor learning

and choosing interventions for more effective supervision

education. Research studies have found both supervisors and

counselors believe supervisors vary their behaviors but have

not identified what criteria other than counselors' experi-

ence level supervisors may be using to make their decisions

about those behaviors. The individual personal attributes

which would be critical for supervisors to consider has not

been specified in a detailed, comprehensive developmental

model by either theorists or researchers.

Lambert (1980) and Holloway and Hosford (1983) have

called for such a "prescriptive model" of supervision educa-

tion, one which would "predict what types of supervision

techniques will result in what types of trainee outcomes for

which type of trainee" (Holloway & Hosford, 1983, p. 75).

They also cite the need to delineate the desired outcomes

of supervision, especially for the advanced student, who has

seldom been included in supervision research. It seems

ironic that while supervision has been identified as a cen-

tral, significant learning experience during a counseling

program by both educators (Banikiotes, 1977; Lambert, 1980)

and students (Hart, 1982; Hill et al., 1981), we have con-

ducted supervision without a clear knowledge of the critical

variables impacting learning or the goals and desired out-

comes of the learning process.

Holloway and Hosford (1983) believe developmental

models provide the starting point for building a prescrip-

tive theory. They cite the need for exploratory, descrip-

tive studies to identify critical variables which can then

be studied in experimental research designs.

Ego Development

The concept of ego development and the theory of ego

development stages (Loevinger, 1966, 1976; Loevinger &

Wessler, 1970) seem to provide a framework sophisticated

and complex enough to serve as a basis for a model of

counselor development. Loevinger refers to ego development

as the "master trait" or "central core of personality"

(Loevinger & Knoll, 1983), second only to intelligence in

explaining human variability (Loevinger, 1966). Ego devel-

opment reflects a "person's frame of reference" (Loevinger,

1979, p. 284) or describes "the framework of meaning which

one subjectively imposes on experience" (Hauser, 1976, p.


Levels of ego development describe a "sequence from

egocentric impulsivity, through rule-bound conformity, to

a self-aware orientation that fulfills its self-chosen

responsibilities, to value for individuality and acceptance

of inner conflict" (Loevinger & Knoll, 1983, p. 202). Ten

stages and transitional levels mark this progression, as

described in Table 2.

S ec

U 4 '4l1


*- Ck

7 ?l

C .5; r

* '.- 3

> {
l/l C


E .

a & 0 >

S-. ; .--.C

un <



0 u0





-2 t





C. C.)I
Q0u u Q

S& -C.
. 0 -



U1 0 L

T > E
C~ C
?, : C '3 -

c < c
A- >.-o

: 8- .^ g

S2 E E.

0o3 t3
U' rC CC

un <
3 C l

u~v < c

0 -

x E
-03 10

E E.
o -0

n- Cw O L n


L >
CEo 'n3J C




M 0

- 0


z ril

Relevance to Counseling and Counselor Development

Loevinger's (Loevinger & Wessler, 1970) descriptions

of ego development stages illustrate how persons, and there-

fore counselors, at each level perceive others and interact

with them. At the Impulsive (1-2) stage, "people are seen

as sources of supply" (Loevinger & Wessler, 1970, p. 57);

at the Self-protective (Delta) stage, people are viewed as

either taking advantage of oneself or as able to be manipu-

lated and exploited; at the Conformist (1-3) stage, people

are described in stereotypic, conventional terms (especially

sex roles); at the Self-aware (1-3/4) transition stage,

basic individual differences are acknowledged; at the Con-

scientious (1-4) stage, interpersonal relations are de-

scribed in terms of mutuality; at the Individualistic (1-4/5)

transition stage, interpersonal relations are viewed as a

process, changing over time and having interactive effects;

at the Autonomous (1-5) stage, each person's individuality

and uniqueness is cherished; at the Integrated (1-6) stage,

identity and self-fulfillment for self and others is


Besides the changes in relationships, Loevinger also

describes other aspects of ego development particularly

germane to counseling (Loevinger & Wessler, 1970). For

instance, at the Conscientious (1-4) level, a person is

able to identify underlying feelings and patterns of behav-

ior, and is beginning to formulate explanations in terms of

psychological causality. The Individualistic (1-4/5) person

distinguishes between process and outcome, inner and outer

life, appearance and reality. The Autonomous (1-5) person

has a high tolerance for ambiguity and sees complex and

circular forces in interpersonal relationships, and is con-

cerned with broad social issues. The Integrated (1-6) per-

son additionally understands paradox and thus is able to

transcend and reconcile conflicts.

In addition, Loevinger (1976) has outlined the simi-

larities and parallels of the stages of ego development and

the stages of the process of therapy as described by Rogers

(1961) (see Table 3). For example, the Self-protective

(Delta) person perceives problems as external and has no

sense of personal responsibility for them. Stage progres-

sion is marked by increased recognition and expression of

concern about contradictions between experience and self,

and increased acceptance of responsibility for problems.

The Autonomous (1-5) person owns changing feelings and

trusts the process of evolving and reformulating perceptions

of self and others. The Integrated (1-6) person resembles

Maslow's self-actualizing person (Loevinger, 1979).

In support of the parallel developmental sequences of

ego development and therapy, Atkins (1976, and in Loevinger,

1979) found female clients' perceptions of counseling were

related to their ego levels. When asked to discuss a prob-

lem related to being a woman, pre-Conscientious (1-4) female

clients focused on control of emotions and concrete aspects



"3 i







0 E-i






3 2

" 1 -
0 0.


C 4. -y -
Ig ~ ;


-'0 ^ --7

CJ: ~ ~ 14
0 I I .
S-u - y
^ ^ ^ $^^

- -
. .
i14 C

5 i' i
2 u;E 'i S "

3 _: ; .
o L=C


=L "2


a7 A3 -

14 2
C'~F _

E Q-

-H c-' a

* .4- ''- C
-e ->
= = .^ S .

n ^ j *

5 r ; -S
i ^ 3 ^ / ^

*'. *

0 i-

- C
C o
U -

> u"
7 0

0. -'2


0. ^
i 11


^ .4 I
S- C
^- 35

3 0
2 ft

- '
u C

14 v
- .0
14 C
.0 0

0; -
c a

I i I

i C

0 -

- 0 0
I -s ^^

C "C S

0 'C" .
^ u 5=
*y rP

,, < ~~





-~ ~ -C C

a oi
J0 U- r -" 0 -


E 2
-G v
S:t C I |
* S* i 5U^*=- -? C
J - 5 gE? CS .
-U U -~0 L

- a 0. -3 4
-1 C U I ~ .^ U I i

^~nV i .= 5 r

= > t
.- c -

: 51 to.- 1

> -E 2
1 CI

U T~-O U -~v ~ 'C

Er r)-
i- ,- t Ui U i

3 *= L^ & SS : iSS ?

E > -
- a C 2
- j 0 o

S- 0E c' U 0
- *.J - = u s = c = c 3 -
- k r/ 0^x i ^ k ^ j .
o 4-- 2

C U.- i- i-- ^ ^ 0- S O Q
U *E a3 ^ *1 U 2 ?.

Si a, c S 0 5
0 3 C 3 2 C
- ) 3 U E

-- '- 3~ S0 *^ 01 ^ 2 "
- to U U U 1. 1-, I-..gs 1 S 5 S 4
C i.
C L i L

^ 3 i U C >
U- t U -S C U 2 | ;

U o 3 C U ;
.. 5 = 3 I 3 ^ .;* 3 X J -

*= = ? = 2 5 B = - - 3 ^ o- s Z *- 0. bO f~ QJ
U J C Y C U *3~> U -

*^S ^ a. S S- a. ^i S^ j^ l' i. c ^ ; ^ ^ S -

c '- Ui g ^ ^ u M = 5 ^ o ^
- -~ o C
U 051 Uo .5 U c >
U~v U) I. I' 1 t

C > ol; 'C C b O -

30 U^ 0:0 U ^ S

*;~- C. S. a 31 ^ S^^ ; ^j .=5 5

U to ii 0 5 i= ?,"' '' C u U .T c2 = - o ^ U
U~v 0C U_ 'CC

A- A U C 0 U O c ^- -

-- o. U >.. I U S i c c - 3 -' U- --c c o 1' U U* *
U~ U

C U" '- 1 0 0. ? C r r - 1 t -
'S U .C 0 = U 5
- i, o, IX U C U C -

.1 U 1
U A ~ C 10 U Cu- U, i
Cn C

-r -. C0 LU v5 0 -
U U U I 2^ U 0-
0 U -U~ 0n",
.~ U.
C- 1 to-n U. C -I 0..=.
C U C. ... 0.U &~~:0 0C U' CC

U U .C, U 0.. U
C U ue m > U -

C7 ?o 2. U 3, U C U Ul 1 CCI U~ C
~L U 0.2 EU C U IU
'-C-" IJ U 0a U -
C- .0 .0. t, t, L i
50L~ 4L= iQ4
.= e~ "- 1CL00

of woman's role, home, family, and mother. Post-Conscien-

tious (1-4) females were concerned with relationship, ab-

stract aspects of woman's role, self-identity, and inde-

pendence. The two groups also differed in their conceptions

of the purpose of psychotherapy. Low ego level women

stressed working out solutions; realizing, alleviating, and

coping with feelings; and adjusting to society. High ego

level women emphasized self-understanding; the self as

hidden or internally blocked; the relationship between

therapist and patient; patterns of behavior and being stuck

in them; and connections between thoughts, feelings, and


Higher levels of ego development should not be equated

with better adjustment, however; they do not represent a

"conflict-free sphere" (Loevinger, 1966, p. 205). Loevinger

explains, "Every stage has its weaknesses, its problems,

and its paradoxes, which provide both a potential for mal-

adjustment and potential for growth" (p. 200).

Counselor educators' views of ego development

Several counselor education theorists have highlighted

ego development theory's relevance to counseling and coun-

selor development (Blocher, 1983; Miller, 1982; Swenson,

1977, 1980a,b). Swenson (1977), for example, describes ego

development as "the central personality variable in inter-

personal relations" (p. 41) and concludes it seems the most

applicable to counseling of the current general developmental

theories (Swenson, 1980b). He believes it provides a basis

for organizing, relating, and selecting appropriate theoreti-

cal interventions for clients at different ego levels

(Swenson, 1977, 1980b) and has implications for optimum

levels of ego functioning for counselors (Swenson, 1980b).

What would be the effects of differences between counselor

and client ego levels, he asks and theorizes counselors

might function better in settings where the counseling work

"matches" their ego functioning.

Both Miller (1982) and Blocher (1983) specify Loevinger

as one of the stage theorists who serve as the bases for

their developmental approaches to supervision. Miller

(1982), for example, cites Loevinger's theory when he dis-

cusses conceptual and ego development, but the descriptions

of his other two processes (emotional flexibility and con-

gruence, and awareness) also reflect dimensions incorporated

into Loevinger's theory. For instance, his awareness con-

tinuum (unawareness, denial, suppression, awareness, inte-

gration) closely parallels the development of awareness

described by Loevinger's stages of ego development. His

final two stages, autonomous and integrated, have the same

names as Loevinger's stages.

Blocher's (1983) description of the counselor at a high

level of cognitive functioning, the goal of his approach,

closely parallels Loevinger's description of the person at

the Integrated (1-6) level of ego development:

This functioning includes the ability to take mul-
tiple perspectives in order to achieve empathic
understanding with people who hold a variety of
world views, value systems and personal constructs.
It includes the ability to differentiate among and
manipulate a wide range and large number of rele-
vant facts and causal factors. Finally, it in-
volves the ability to integrate and synthesize in
creative or unusual ways large amounts of such
information to arrive at an understanding of the
psychological identity and life situation of a
wide range of other human beings. Still further
the counselor engages in this quest in active
collaboration with the client, and in the hope of
imparting some skill and understanding of the
process to the client. (Blocher, 1983, p. 28)

While Blocher (1983) details the behavior of the super-

visor who seeks to educate a counselor with these abilities

and qualities, he does not include a description of sequen-

tial changes in counselor's perceptions and behaviors

during this educational process. Instead he directs the

supervisor to the stage theorists he uses as bases for his

model: Piaget and Inhelder, Loevinger, Kohlberg, Perry,

and Harvey, Hunt and Schroder.

Stoltenberg's (1981) descriptions of counselor charac-

teristics at his four levels of development often suggest

ego development stages, although he does not specify

Loevinger as a source. For example, the Level I counselor,

unilaterally dependent on the supervisor, lacking self-

awareness, is "quite concerned with rules of counseling" and

"searching for the right way to do things" (p. 61), much

like the Conformist (1-3) person. The more empathic and

tolerant Level III counselor exhibits a more highly dif-

ferentiated interpersonal style, and acknowledges individual

differences, as Individualistic (1-4/5) and Autonomous (1-5)

persons do.

Research Support for Ego Development

Loevinger has integrated the theories of Piaget,

Sullivan, Erikson, Fromm, Kohlberg, Perry, and Harvey, Hunt,

and Sullivan (Loevinger, 1976; Loevinger & Wessler, 1970)

into a common thread of development. As a result the

theory of ego development has empirical support from the

work of psychoanalysts, sociologists, philosophers, and

psychologists represented in its integration (Loevinger,


To measure ego development, Loevinger and her associ-

ates at the Social Science Institute at Washington University

have created a sentence completion test and have conducted

on-going research to refine the items and the detailed

scoring manual. Studies of the validity of theoretical

assumptions of ego development have all used the Sentence

Completion Test (SCT) (Loevinger & Wessler, 1970; Loevinger,

Wessler, & Redmore, 1970) (see the Instruments section of

Chapter III for a description of the test).

Unitary dimension

Evidence for the unitary dimension of ego development

and the SCT is based in a principal components analysis

which yielded an eigenvalue of 8.8 for the first principal

component and 1.2 for the second (Loevinger & Wessler, 1970).

In addition, Lambert (in Loevinger, 1979) found no subset

of moral items on the SCT in correlation with Kohlberg's

test of moral maturity, and Blasi (in Loevinger, 1979)

found no subset of responsibility items.

Polar and milestone traits

Manifestations of levels of ego development include

both polar and milestone traits (Loevinger, 1966). Polar

characteristics exhibit a linear pattern, such as the con-

stant decrease in the tendency to categorize according to

stereotypes, while milestone sequences "tend to rise and

then fall off in prominence as one ascends the scale of ego

maturity" (Loevinger, 1966, p. 202). Research assuming a

linear relationship between ego development and some ex-

ternal criteria has sometimes ignored the important mile-

stone concept and a more probable curvilinear or nonmono-

tonic relationship (Hauser, 1976). Hoppe and Loevinger

(1977) found an expected nonmonotonic relationship between

level of ego development and conforming behaviors, illus-

trating a milestone sequence. Expected correlations of

ego development with other traits may be positive, negative,

or curvilinear, depending on the peak stage for a particu-

lar trait (Loevinger, 1979).

Bochini (1978), for example, reported decreasing scores

on self-centered values but increasing scores on other-

centered and relational values as ego level increased.

Lasker (1978, and in Loevinger, 1979) found the need for

achievement peaked at the Conscientious (1-4) stage while

the need for power peaked at the Self-protective/Conformist

transition (Delta/3) level. In contrast, a study of the

correlation between empathy and ego level reported an ex-

pected linear relationship. Zielinski (1973) pre- and

posttested 40 graduate students in a beginning course in

counselor education on Carkhuff's index of communication of

empathy and his index of discrimination of empathic under-

standing. On pretest he found a moderate relationship (.46)

between level of ego development and ability to communicate

empathy, but no significant relationship with ability to

discriminate empathy. In comparisons of pre and post

scores, students with high levels of ego development gained

in ability to discriminate empathic understanding, but

neither high nor low ego level students gained in ability

to communicate empathy. Carlozzi, Gaa, and Liberman (1983)

also found differences in empathy scores for students at

different ego levels, although they did not use a correla-

tional approach but compared high and low ego levels. The

researchers administered the Affective Sensitivity Scale,

a multiple-choice measure of empathic responses to filmed

vignettes, to 51 undergraduate dormitory advisors. Under-

graduates at higher levels of ego development had signifi-

cantly higher empathy scores than those at lower levels.

In Rock's (1975 and in Loevinger, 1979) study of 50

undergraduate women, two measures of self-insight had highly

significant, positive linear relationships with ego


development, even with age and intelligence partialled out.

Female students at or below the Conformist (1-3) stage

tended to be unreflective or reflective in nonpsychological

terms in their own interpretations of their TAT stories and

their videotaped behavior. Most females shifted to

psychologically-minded self-reflection at the Conscientious

(1-4) stage. Almost all the females at the Individualistic

(1-4/5) and Autonomous (1-5) levels described themselves in

complex psychological terms and showed a dynamic understand-

ing of the reasons for their behavior and inner conflicts.

Sequentiality of ego stages--deliberate psychological

Studies reporting changes in ego level following a

theory-relevant intervention provide evidence for the sequen-

tiality of ego development and SCT, and point to the theory's

relevance to counselor education. Researchers and educators

have devised "deliberate psychological education" curricula

in attempts to stimulate the social, personal, ego and moral

development of a variety of student populations. The pro-

grams included training in basic empathic skills and

presentations of concepts of various developmental theorists;

and they used didactic, reading, and experiential activities

and devised activities to tap into both cognitive and affec-

tive domains in students' learning. Experiential activities

usually included were devised to be practical and realistic

so that students could actually assume the role of tutor or

peer-counselor. Most of these studies have reported in-

creases in empathic skills, moral judgment, and ego devel-


The majority of these studies have used high school

students. Mosher and Sprinthall (1971), for example,

reported a one-third stage increase on Kohlberg's Moral

Judgment Scale and one full stage on Loevinger's SCT for

students in the experimental group (counseling class), but

no changes for the control group (regular psychology class).

The counseling group also gained significantly on empathy

ratings of tapes. Sullivan (1975) reported almost dupli-

cate results on mora&1 and ego development for his control

and experimental high school groups after a one-year


In a study emphasizing personal growth through role

taking (Rustad & Rogers, 1975), high school students in-

creased their moral and ego development stages and improved

their basic counseling skills. In follow-ups of a small

number of the students, the levels on all three measures

had been maintained. High school students who tutored

junior high students also increased their levels of ego

development and moral reasoning (Cognetta, 1977), as did

female high school students following a women's literature

course (Erickson, 1975, 1977b). In a junior college

psychology course which gradually decreased structure and

increased student-initiated learning, students' ego devel-

opment levels increased while the reverse curriculum and

control groups showed a slight decline (Exum, 1978).

Other studies have used deliberate psychological edu-

cation curricula with student teachers, counselors, and

social workers. For example, Oja and Sprinthall (1978)

taught students representing all three groups for one sem-

ester. Experimental groups did not significantly increase

their ego levels but did gain significantly on moral

thinking and cognitive complexity. They also improved

their ability to use indirect teaching methods, viewed by

the researchers as more effective than directive approaches,

and tended to become less dictatorial and arbitrary and

more empathic. Hurt (1977) found that teachers who received

extensive empathy training gained significantly on measures

of ego development and moral thinking. Glassberg (1978)

used a "peer supervision" curriculum based in deliberate

psychological education principles with English education

students. The experimental group showed significant in-

creases in ego level, moral thinking, empathic skills, and

indirect teaching skills. They also shifted from an ex-

ternal to a more internal locus of control.

Bernier (1980) reported on a pilot curriculum with 18

counselors and teachers in a graduate workshop in develop-

mental education which involved classroom, prepracticum, and

field-based practicum experiences. Students showed signifi-

cant pre-post increases in accurate empathy and moral

thinking, but not significant differences on conceptual

level and ego development.

Behavioral correlates of ego development

Other validation studies have focused on behavioral

correlates of ego development, although Loevinger (1979)

asserts ego levels cannot be translated into single,

specific behaviors, and so expects low correlations. In

reviews of these validation studies, Loevinger (1979) and

Hauser (1976) reported expected relationship between low ego

levels and certain behaviors. For example, adolescent boys'

conformity behaviors peaked at the Conformist (1-3) stage;

institutionalized delinquents' behaviors were negatively

correlated with ego stage; and children's behaviors reflect-

ing responsibility were positively correlated with ego stage.

In contrast, evidence of validity for high ego levels is

based in correlations with attitudes, philosophy of life,

self-insight, and ability to communicate empathy (Loevinger,


Ego Development and Other Developmental Theories

Although other theories (i.e., Perry, Kohlberg, Harvey,

Hunt, & Schroder, and Kelly) incorporated by Loevinger into

ego development are used in counseling research and they

describe similar developmental themes and patterns,

Loevinger's theory was chosen as the theoretical basis for

this study because of its broader base and application

(Loevinger, 1976; Swenson, 1980a). The following reviews

and brief descriptions of these theories illustrate the

bases for this conclusion.


Perry's (1970) developmental scheme, based on Kohlberg's

theory of moral development, focuses on the intellectual-

ethical development of traditional age college students.

Students are theorized to move through nine major stages in

their conceptions of knowledge and values, from a simplistic,

categorical view to a complex, pluralistic one. The devel-

opment scheme has been criticized because it does not en-

compass all the relevant aspects of development and has

limited applicability for other age ranges (Widick, 1977).

The scheme also lacks an established reliable and valid

assessment instrument (Widick, 1977). These criticisms

point to the developmental scheme's limited potential in

research with older graduate students.

Some counselor educators have used Perry's developmental

scheme as a theoretical basis for counseling practice and

counselor development. For example, Widick (1977) sug-

gested the content of Perry's developmental sequence identi-

fies both goals and interventions for deliberate psychologi-

cal education programs and workshops in anxiety management,

career/life planning, and peer-counselor training. Schmidt

and Davison (1983) proposed using reflective judgment, a

scheme of intellectual development extrapolated from Perry's

broader scheme, to describe important client variables and

to suggest prescriptive interventions for encouraging

client development towards greater cognitive complexity.

Cooper and Lewis (1983) used Perry's scheme to de-

scribe counseling students' typical cognitive transitions

as they try to cope with the diversity of counseling

theories. According to the authors, students move from a

dogmatic and rigid dualistic framework (i.e., right/wrong

theory), to a pragmatic acceptance of multiple theories

(i.e., whatever works), to a theoretical commitment which

recognizes the limitations of all theorizing. They also

suggested intervention strategies to facilitate students'

movement through these transitions. Researchers, however,

have not investigated the efficacy of the application of

Perry's scheme to counselor development.


Loevinger and Knoll (1983) have cited Kohlberg as the

leading figure in the cognitive-developmental movement, as

both the most popular and favorite target in the field.

Kohlberg's (1969) developmental theory of moral judgment

and its stage sequence has considerable validity support

(Kohlberg & Wasserman, 1980; Loevinger & Knoll, 1983; Rest,

1980) and has been shown to be an important factor in real-

life decision-making (Rest, 1980).

Measures of moral development have correlated posi-

tively with ego development (Loevinger, 1979), and both

Kohlberg (Kohlberg & Wasserman, 1980) and Loevinger (1976)

have noted the overlap of the two theories. Studies of

deliberate psychological education have shown parallel

increases in measures of each for high school students

(Cognetta, 1977; Erickson, 1975, 1977; Mosher & Sprinthall,

1971; Rustad & Rogers, 1975; Sullivan, 1975), and teachers

(Glassberg, 1978; Hurt, 1977): Bernier (1980) reported no

increases for ego development, but significant increases in

moral judgment for counselors and teachers in a graduate


Gilligan (1982), however, has challenged Kohlberg's

claim of universality for his stage sequence, arguing that

it only describes a masculine developmental pattern which

defines maturity in terms of autonomy and achievement and

ignores a more feminine concern with relationships. In

addition, Rest (1980) and Welfel and Lipsitz (1983b) have

identified a predominant cognitive component in moral judg-

ment in its emphasis on the logical analysis of ethical

dilemmas. Rest's (1980) Defining Issues Test (DIT), an

objective measure of moral judgment based on Kohlberg's

theory, has usually yielded significant positive correlations

with measures of cognitive development, but nonsignificant

correlations with affective and personality measures.

Loevinger and Knoll (1983) and Rest (1980) have identified

the difficulty in administering and scoring Kohlberg's

Moral Judgment Instrument, and Kohlberg has frequently

revised his stages, adding and deleting transitions and sub-

stages (Loevinger & Knoll, 1983). Finally, Sebes and Ford

(1984) have suggested moral development needs to be studied

within the context of broader theoretical frameworks,


investigating "its relationship to affective, interpersonal,

cognitive, and psychosocial development" (p. 380).

Few studies have investigated counselors' moral

judgment, focusing instead on issues related to counseling

ethics (Zahner & McDavis, 1980). Several writers have

used Kohlberg's theory as a basis for a model of ethical

decision-making in counseling (Van Hoose & Paradise,

1979) and for ethics education (Welfel & Lipsitz, 1983b).

In a review of ethics research, Welfel and Lipsitz (1983b)

concluded there are no data yet to support the efficacy

of ethics education on counselors' actual behavior with

clients. They cited the weaknesses of measures of ethi-

cal orientation and suggested future studies of ethics

should be based in cognitive developmental theories and

should include counselor personality variables.

In a comparison of both professional and paraprofes-

sional counselors at various levels of training, Zahner and

McDavis (1980) found the professionals had significantly

higher levels of moral development than the paraprofes-

sionals as measured by Rest's DIT. Counselors' level of

training revealed no differences in moral development

scores for either group.

Welfel and Lipsitz (1983a) investigated the relation-

ship of stage of ethical orientation and stage of moral

reasoning of counseling students with different levels of

experience. Undergraduate seniors majoring in human

development, and beginning, advanced, and doctoral counsel-

ing students completed the DIT and the Ethical Judgment

Scale, a new instrument also based on Kohlberg's theory.

Doctoral students scored significantly higher in ethical

orientation than did less experienced students (comparisons

of DIT scores were not reported), and contributions to

professional and social action organizations significantly

correlated with ethical orientation. In addition, stage of

moral reasoning and stage of ethical orientation showed a

significant correlation, although the researchers noted the

weaknesses of the Ethical Judgment Scale and the need for

further validity studies.

Harvey, Hunt, and Schroder

Harvey, Hunt, and Schroder's (1961) conceptual systems

theory specifies four systems or stages to describe a per-

son's style of information processing and ability to adapt

to a changing environment. Stages include cognitive vari-

ables (i.e., complexity, integration) and interpersonal

variables or interpersonal orientation (i.e., dependence-

independence, empathy). Higher conceptual systems are

characterized by greater flexibility, complexity, abstract-

ness, interpersonal maturity, self-understanding, and

empathic awareness (Hunt, 1971). The conceptual model has

substantial empirical support, including one study which

found significant correlations between conceptual system and

both moral judgment and ego development (Hunt, 1971; Hunt

& Sullivan, 1974). Two studies of deliberate psychological

education included measures of ego development, moral

thinking, and conceptual systems also. Teachers and coun-

selors showed significant gains in cognitive complexity in

one study (Oja & Sprinthall, 1978), but no significant

increases in the second (Bernier, 1980).

The original conceptual systems theory has been revised

separately by the theorists several times. Hunt, for ex-

ample, now refers to both four stages and three (ABC) stages

(Hunt & Sullivan, 1974), and he has focused on elaborating

a matching-model of conceptual system and instructional

method for more effective education rather than on the

developmental model itself. At least four different meas-

ures of conceptual systems are reported in the literature

(i.e., Conceptual Systems Test, Sentence Completion Test,

This I Believe Test, Paragraph Completion Test). These

measures assess a person's values and beliefs as a basis for

classification into a conceptual system. While the con-

ceptual systems theory describes relevant developmental

changes for counselors, the current variations in the stage

sequences and in the assessment instruments are confusing

for a researcher.

A fairly substantial body of literature has included

a measure of conceptual system as a variable in studies of

persons in various helping professions, including teachers

and counselors, and in studies of counseling-related be-

haviors. One study indicated that counselors' conceptual

levels are related to their learning styles. Rosenthal

(1977) reported a self-instruction approach for teaching

confrontation skills was more effective with high concep-

tual level counseling students while guided instruction was

more effective for low level students. These results

reflect the assumptions of Hunt's matching-model for educa-

tion and parallel results of numerous studies with non-

counseling students (Hunt, 1971; Hunt & Sullivan, 1974).

In related literature, researchers have reported a

relationship between conceptual system and teaching style.

Hunt and Joyce (1967) found a significant correlation

between teachers' conceptual system and their tendency to

use a reflective teaching style, one which encouraged

students to formulate their own theories and to express

themselves. Murphy and Brown (1970) found distinctive

teaching styles for student teachers at each of four con-

ceptual systems. More concrete and dependent student

teachers evidenced a lecturing style, while more abstract

and interdependent student teachers evidenced more

spontaneity in using pupil questions and ideas. They also

encouraged further pupil exploration and led discussions

toward generalizations in a sequential process. Bernier

(1980), in a brief summary of studies of conceptual level

of teachers, found that more complex teachers were more

flexible both cognitively and behaviorally and able to

respond to a wider range of pupils' feelings and experi-

ences. He also concluded that their pupils exhibited higher

cognitive levels and more cooperation and self-exploration.

Several researchers have investigated the relationship

between conceptual system and empathy skills. Lutwak and

Hennessy (1982), for example, rated tapes of actual coun-

seling interviews conducted by advanced undergraduates and

graduates in their first interview skills course. They

found highly significant differences between high and low

conceptual level students on ratings of accurate empathy.

The researchers suggested low level students may be respond-

ing from a narrow and dogmatic framework which may affect

the counselor-client relationship and alter the direction

of counseling.

Other researchers have found a more complex relation-

ship between conceptual system and empathy, reporting sig-

nificant interactions between conceptual level and other

variables. In Heck and Davis's (1973) analogue study, for

example, 40 counseling students' written responses to 12

client statements were rated on an accurate empathy scale.

High conceptual level students' responses were consistently

rated higher, but there was a significant interaction

effect between client analogues and students, so that

students' empathy levels were not constant across client


Kimberlin and Friesen (1977) gave brief (2 sessions)

empathy training to undergraduate students who then wrote

responses to videotaped client statements which were de-

signed to be either ambivalent or nonambivalent. Higher

conceptual level students wrote more empathic responses to


ambivalent client statements, but there were no significant

differences between high and low level students' responses

to nonambivalent statements. In addition, high level

students wrote significantly more responses which actually

addressed the ambivalence, identifying it or reflecting

both sides of the conflicting emotions. The researchers

concluded the low level students were only able to empathize

with clear-cut emotions, limited in their ability to process

complex, conflicting information. In a similar later study

(Kimberlin & Friesen, 1980), the researchers found the same

differences for both male and female students.

Blaas and Heck (1978) compared students categorized as

having high or low levels of cognitive complexity on five

instruments, including a conceptual systems measure. First

semester counseling students' interactions with two simu-

lated clients were rated on several process variables, in-

cluding empathic responses. The researchers found no sig-

nificant differences between high and low complex students

on any of the process variables but did report a signifi-

cant interaction between empathy ratings and counseling

task (i.e., client) for the low complex students. They

interpreted students' performance on the process variables

other than empathy as more a function of client differences

than their cognitive complexity level. The researchers

concluded cognitive complexity might not be "universally

influential" but affected by situational and environmental


Strohmer, Biggs, Haase, and Purcell (1983) investigated

the effects of level of cognitive complexity and anxiety on

first practicum counselors' empathic responses to disabled

and nondisabled videotaped clients portrayed by actors.

While students with higher complexity scores had higher

empathy ratings, the study also found a significant inter-

action between disability condition, cognitive complexity,

and anxiety. The researchers found that the more complex

students showed greater tolerance of anxiety-arousing stimuli

and, even more, were able to use the additional stimuli in

the disabled client situation to respond more empathically

than the less complex students. They described a curvi-

linear relationship between cognitive complexity and empathy

ratings, with low complex students rated most empathic in

situations with minimal levels of arousing stimuli (i.e.,

nondisabled client and low anxiety), high complex students

rated most empathic in situations with moderate levels of

arousing stimuli (i.e., either a disabled client or high

anxiety), and both groups rated least empathic in situations

with maximum arousing stimuli (i.e., disabled client and

high anxiety). The researchers suggested the relationship

between cognitive complexity and empathy was more complex

than previously considered.

Other studies using a measure of conceptual system

have investigated the cognitive processing of counselors.

Lichtenberg and Heck (1979) analyzed the interactional

structure of counselors' and clients' verbal responses in

tapes of actual counseling interviews. Students were clas-

sified by a total of five measures of cognitive complexity,

including a conceptual system measure. More complex

students evidenced a somewhat greater degree of variation

in their responses than did less complex students, suggest-

ing different information processing styles for high and

low conceptual level students. Holloway (1979) reported

counselors' conceptual level was moderately related to

their formation of clinical hypotheses for videotaped clients.

Goldberg's (1974) analogue study rated beginning coun-

seling students' responses on dimensions reflecting the

core conditions (i.e., affective-cognitive, understanding-

nonunderstanding, specific-nonspecific, exploratory-

nonexploratory) and analyzed them for their interaction

style. Students at lower conceptual levels used more

direct verbal behavior (e.g., asked questions, gave infor-

mation or directions), while high level students accepted

and used clients' ideas. The most frequent response of all

students was asking questions, but low level students'

questions asked for information or data, while high level

students' questions were more open-ended, encouraging

client exploration of beliefs and feelings and client self-

responsibility for their replies. In addition, high level

or more abstract students were more likely to respond to

client affect, to convey an understanding of the client's

perspective, and to deal with core rather than peripheral

issues. The researchers concluded low level students'

responses reflected actions designed to maintain control of

the interaction.

Finally, Gore (1978) found no relationship between 50

first practicum counselors' conceptual level and their

supervisors' ratings of their effectiveness on the

Counselor Evaluation Rating Scale (CERS).


Kelly's (1955) personal construct theory describes a

person's attempts "to discern order in the physical and

interpersonal realities they confront" (Neimeyer & Neimeyer,

1981, p. 190). Persons elaborate their construct systems

to various degrees over time, developing new constructs or

modifying existing ones in an attempt to more accurately

interpret and predict events. This process of elaboration

reflects a developmental pattern, moving toward increasing

levels of complexity and integration in perceptions of self,

others, and life events. The Repertory Grid Technique

(Rep Grid) (Fransella & Bannister, 1977) for studying

constructs yields measures of both cognitive complexity

and integration, and research studies comparing persons

with more and less complexity and integration reflect

developmental differences (Fransella & Bannister, 1977).

Kelly's theory, however, does not specify a developmental

sequence of stages or levels, focusing instead on the

processes of the evolution of the construct systems.

Research using variations of the Rep Grid (Fransella

& Bannister, 1977), one of the measures to be used in this

study, has also studied the relationship of cognitive com-

plexity to the work of persons in the helping professions.

Some studies have compared the behaviors of more and less

complex persons or analyzed the content of their constructs.

Others have identified a developmental pattern in the

changes of their constructs over time which reflect changes

assumed in both the theory of ego development and models of

counselor development.

Two studies have investigated processing of information

about clients. Philip and McCulloch (1968) reported a case

study of a male psychiatric worker who was given the names

of 50 of his former hospitalized patients (role titles or

elements) to use as the basis for eliciting constructs in

two Rep Grids. A cluster analysis identified two main

types of constructs: the impact of the patient on the

psychiatric social worker (i.e., feelings about the patient)

and case conceptualization (i.e., the assessment of social


Duehn and Proctor (1974) studied the relationship of

cognitive complexity to social work students' decision-

making processes about clients. While all students used

more pieces of information about clients presented as

similar to themselves than about clients presented as dis-

similar, less complex students used significantly fewer

pieces of information with dissimilar clients than did more

complex students. The more complex students also specified

a greater number of alternative interventions, regardless

of the degree of similarity of clients.

Other studies using a Rep Grid have addressed issues

relevant to professional identity and professional develop-

ment during training. Two relevant studies found a curvi-

linear relationship between cognitive complexity and amount

of training for teachers (Runkel & Damrin, 1961) and social

work students (Ryle & Breen, 1974). Training or an increase

in knowledge seemed to restrict at first but then enlarge

the subjects' cognitive systems. The untrained subjects

used a large number of dimensions to describe students'

problems (Runkel & Damrin, 1961) and their relationships

with social work clients, supervisors, parents, and self

(Ryle & Breen, 1974), but seemed to use them erratically,

responding to the situation rather than on the basis of

specialized differentiations. Teachers and social work

students with partial training seemed to oversimplify their

construct systems, perhaps attempting to apply a few newly-

learned specific dimensions to all students or relation-

ships. Those with high levels of training seemed to use

their higher number of relevant dimensions to make more

discriminating and specialized perceptions. In addition,

the three Rep Grids completed by the social work students

over the course of their training showed a steady decrease

in the tendency to make polar judgments (extreme ratings)

on the constructs. The social work students thus seemed

to move toward both greater complexity and less dichotomous


Ben-Peretz and Katz (1983) found significant differ-

ences between 145 first and third year female student

teachers' constructs describing curriculum materials. While

first year student teachers viewed curriculum materials in

terms of their more superficial aspects (e.g., "lack of

illustration"), third year student teachers looked for the

more sophisticated educational aspects and potentials of the

curriculum materials (e.g., methods of teaching, level of

cognitive domain). The researchers explained the results

in terms of a developmental process from relative simplicity

to more divergent and professionally-oriented complexity.

They concluded experience in the teacher education program

helped students develop toward higher levels of profes-

sional complexity, rebuilding relevant personal constructs

which would guide their educational decisions and actions.

Lifshitz (1974) analyzed the content of the constructs

of social workers using standard role titles for 12 rep-

resentative people to compare the constructs of good social

work students and their more experienced supervisors. He

found the students used more concrete descriptions (e.g.,

age, sex, profession) while their supervisors used more

abstract descriptions (e.g., diligent, responsible), more

intrapsychic and more interpersonal characteristics. In

addition, students identified their father, mate, or friend

as their professional models, but the supervisors' preferred

models were self, mate, or another social worker with high

ethical standards. In addition, the good social work

students tended to devalue needy people. Lifshitz inter-

preted these differences as support for a developmental

pattern as students gained experience, characterized by

increasing levels of abstraction and symbolization, inter-

nalization of values, seeing self as a model, and aspiration

toward professional goals.

While these studies illustrate the impact of cognitive

dimensions on teachers' and social workers' perceptions,

thought processes, and changes during training, no relevant

published studies using a Rep Grid have used counselors as



The descriptions and/or critiques of these related

theories point to the rationale for choosing Loevinger's

theory of ego development as a basis for this study. Most

obviously, Loevinger has integrated each into her theory

(Loevinger, 1976; Swenson, 1980a). She refers to other

models as separate aspects of the unified ego development

concept, the "central core of personality" (Loevinger &

Knoll, 1983, p. 205). Ego levels describe both cognitive

and interpersonal orientations, both crucial components of

counselor performance. The theory and the instrument

designed to assess the developmental levels has substantial

reliability and validity support (Hauser, 1976; Loevinger,

1979), and recent revisions of each have been refinements

rather than major changes. The Sentence Completion Test's

two-volume manual (Loevinger & Wessler, 1970; Loevinger,

Wessler, & Redmore, 1970) includes comprehensive, detailed

descriptions and examples of responses for each ego level

to aid in scoring. Sentence stems tap a variety of a per-

son's perceptions of experiences and relationships, as op-

posed to a focus on ethical and moral decisions (i.e.,

Perry and Kohlberg), values and beliefs (i.e., Harvey, Hunt,

& Schroder), or descriptions of self, others, and life

events (i.e., Kelly). For these reasons, ego development

has been chosen as the theoretical basis for this study of

counselor performance and development.

The reviewed research using these developmental theories

or measures based on them illustrate their relevance to

counselor development and supervision education. Since

the theory of ego development encompasses these theories,

the research suggests ego development may be useful in an

exploratory and descriptive study of counselor development.


Developmental models of supervision education indicate

a pattern of increasing cognitive complexity which impacts

a counselor's perceptions of their clients and behavior

with their clients. The developmental view theorizes these

changes are a result of both experience and individual

differences in the counselors themselves. Research of

developmental models and of individual counselor variables

support this view of counselor development, but develop-

mental models have not yet identified and described the

possible interaction of individual differences and experi-

ence during counselor development.

This study explored the efficacy of the theory of ego

development, a comprehensive and established personality

stage theory, to describe the differences in counseling

students' perceptions of clients, their behavior with

clients, and their effectiveness with clients.


The purpose of the proposed study was to investigate

the efficacy of the theory of ego development as a theo-

retical basis for counselor development. Counseling stu-

dents' level of ego development was the independent vari-

able. Because of the exploratory and descriptive nature of

the study, multiple dependent measures of students' coun-

seling performance were used. Students' perceptions of

their clients were assessed by measuring the cognitive

complexity, cognitive integration, and meaningfulness of

their perceptions. In addition, the content of students'

perceptions of their clients were categorized. Other

dependent measures assessed students' behaviors with their

clients, and individual supervisors' ratings of students'

effectiveness with their clients.

Students' age and level of experience (first or second

practice or internship level) were potential confounding

variables. Age has been positively correlated with level

of ego development (Hauser, 1976; Loevinger, 1979;

Loevinger & Wessler, 1970). Level of experience has differ-

entiated between counselors' preferred supervision style

(Moskowitz, 1981; Nash, 1975; Worthington & Roehlke, 1979),

perceptions of their supervisors (Cross & Brown, 1983),

relationships with their supervisors (Hill et al., 1981),

and their verbal responses to clients (Cicchetti & Ornston,

1976; Hill et al., 1981). Because of these reported rela-

tionships, students' age and level of experience were in-

cluded in the statistical analysis procedures in this



The subjects for this study were University of Florida

counselor education graduate students enrolled in their

first or second counseling practice or internship during

Spring Semester 1984, who were seeing individual clients

and who volunteered to participate. The individual super-

visors of the volunteer students were also asked to

participate in the study.


The counselor education program at the University of

Florida is fully accredidated by CACREP and NCATE. All

students are required to complete at least one counseling

practicum, regardless of their chosen program area or

track, in addition to a second practicum and internship in

settings appropriate to their chosen track. To enroll in

their first practice, all students except those in the

counseling psychology track must successfully complete

one-semester courses in Theories of Counseling and Prin-

ciples of the Counseling Relationship. These students are

placed in sites following an interview and selection process

with the hosts at their preferred sites. Requirements for

completing first practice include seeing a minimum of three

clients and reviewing at least 11 tapes with an individual

supervisor, while requirements for second counseling prac-

tica and counseling internships vary but are comparable.

To enroll in their first practice, counseling psychology

students must complete two semesters of theoretical counseling

classes. They may also elect to take the Principles of the

Counseling Relationship course. Tapes are not required for

all of their practice and internships.

All practice and internship students meet with their

individual supervisors for one hour per week and, in addi-

tion, receive group supervision for one-and-one-half hours

per week either with a faculty member, supervised doctoral

student, or an approved on-site supervisor. University

supervisors contact on-site hosts several times during

the semester. Students are evaluated by their respective

supervisors and hosts at the end of the semester.

Of the 66 graduate students who met the criteria for

participation in the study, a total of 63 volunteered for

one or more parts of the study. Of these 63, 45 were

female and 18 were male, a ratio representative of the

counselor education department's enrollment. The students

represented each program area offered by the department:

school counseling (12), agency counseling (21), counselor

education (4), school psychology (6), student personnel in

higher education (7), and counseling psychology (13).

Twenty students were pursuing a Ph.D. degree; 43, an Ed.S.

degree. The students' ages ranged from 21 to 51 (mean

age=29.95, SD=7.09).

During the time of the study, 27 of the students were

enrolled in their first practicum experience, 10 in their

second practicum, and 26 in their internship experience.

The practice or internship settings represented each stu-

dent's program area and included public schools, mental

health agencies, the University student counseling centers,

the University residence halls, the University student

services offices, career counseling centers, specialized

service centers (planned parenthood, crisis and suicide

prevention), centers for special populations (rape victims,

juvenile offenders, alcoholics), and support services for

hospitalized patients.

Individual Supervisors

A total of 32 individual supervisors volunteered to

participate in the study, providing ratings for 57 students.

Of these supervisors, 12 were counselor education faculty,

4 were adjunct faculty, 10 were on-site supervisors, and 5

were counselor education and counseling psychology doctoral

students. On-site supervisors had previously been ap-

proved by the counselor education department, and doctoral

student supervisors were being supervised by senior


One supervisor rated seven students; one rated six

students; two rated four students each; one rated three

students; six rated two students each; and twenty-one

rated one student each.


Four instruments were used in this study: the Sentence

Completion Test of Ego Development (Loevinger & Wessler,

1970; Loevinger, Wessler, & Redmore, 1970), the Repertory

Grid Technique (Fransella & Bannister, 1977), the Vanderbilt

Psychotherapy Process Scale (Strupp, 1981), and the Coun-

selor Evaluation Rating Scale (Myrick & Kelly, 1971).

Sentence Completion Test of Ego Development

The independent variable in this study, level of ego

development, was assessed by the Sentence Completion Test

of Ego Development (SCT) (Loevinger & Wessler, 1970;

Loevinger, Wessler, & Redmore, 1970). The SCT is a pro-

jective measure made up of 36 sentence stems (e.g., Raising

a family . ., When people are helpless . ., When I am

criticized . ., When they talked about sex I . ., Being

with other people . ., A man's job . .), with comparable

forms for boys and girls and for men and women. Subjects

are asked to complete each sentence stem in any way they

wish, choosing anyone for references to persons in the

sentence stems.

Scoring procedures

The scoring procedure for the SCT assumes there is a

core level of functioning which can be derived from totaling

the ratings for each sentence stem (Hauser, 1976; Loevinger

& Wessler, 1970). Responses for each item are removed from

individual protocols and pooled for rating stem-by-stem

rather than by individual total protocol. Each response is

rated independently by two trained raters using the item-by-

item scoring manual (Loevinger, Wessler, & Redmore, 1970)

which contains both categories of responses and illustrative

responses for each ego level. Individual protocols are then

reassembled and a cumulative frequency distribution of the

item scores (ego levels) is tabulated. Total protocol rat-

ings (TPR) are determined by comparing the cumulative fre-

quency distribution to the "automatic ogive rules" (for

relatively new raters) or "borderline rules" (for more ex-

perienced raters) in the manual (Loevinger & Wessler, 1970).

The rules specify the overall ego level of individual pro-

tocols. For example, a total protocol rating of 1-4

(Conscientious) is given if there are no more than 24

(cumulative total) 1-3/4 (Self-aware) items ratings. The

cumulative frequency distribution can also be converted to

continuous scores by multiplying the number of scores for

each ego level by assigned numerical values (1-10 for

levels 1-2 through 1-6) and then adding them for an item

sum rating. This continuous score has the advantage of

enabling researchers to use regression analyses (e.g., Cox,

1974; Hoppe & Loevinger, 1977), although it has a disadvan-

tage of being more highly correlated with verbal fluency

than the TPR ratings based on the ogive rules (Hauser,

1976; Loevinger & Wessler, 1970). Ranges of continuous

scores for each ego level have been specified by several

researchers (e.g., Hoppe & Loevinger, 1977; Schenberg, cited

in Hoppe & Loevinger, 1977).

The SCT manual was developed using samples of female

populations ranging in age from 11 to 50+, both black and

white, of all marital statuses, who had grade school through

graduate college level education. They were employed in a

variety of settings or were volunteer workers, and included

some psychiatric patients. One group of 22 graduate stu-

dents in counseling were included in the samples for the

evaluation of the manual. Subsequent revisions of the SCT

and supplements to the manual have used an even wider data

base, including males from backgrounds comparable to and as

varied as the original female populations.

The scoring manuals for the SCT include a set of com-

prehensive and detailed self-trained graduated exercises for

new raters. Practice responses for each of the 36 items and

complete protocols are followed by answer keys with explana-

tory notes. Reliability coefficients between personally

trained and self-trained raters range from .76 to .92 in a

variety of studies, indicating that self-trained raters can

achieve the same degree of reliability and produce comparable

overall ratings as raters involved in the construction of

the manual (Hauser, 1976; Loevinger, 1979; Loevinger &

Wessler, 1970).

Reliability and validity

Hauser (1976) reviewed initial reliability studies and

although he found some influence of situational factors on

subjects' scores, he concluded there was acceptable evidence

for test-retest, split-half, and internal consistency re-

liabilities for the SCT. Loevinger (1979) later reported

internal consistency coefficients of about .85 in additional

studies. Loevinger and Wessler's (1970) original validity

studies found a relationship between SCT scores and age

(.74 for boys, .69 for girls), with expected progressive

age increases for all ego levels. Hauser's review (1976)

pointed to similar age trends.

Loevinger and Wessler (1970) found no significant

correlations between SCT scores and verbal fluency or intel-

ligence. Hauser (1976) also concluded the SCT was not

simply measuring these factors, although he suggested the

relationship of IQ and ego level needed further clarifica-

tion. Loevinger (1979) found extreme variability in the

relation between ego level and IQ, even for similar samples,

in later studies.

Hauser (1976) also found tentative evidence for struc-

tural validity of the SCT, both for the single factor nature

of the theory ("master trait") and for the organization of

psychological variables at specific stages. Loevinger

(1979) discussed evidence for the construct validity of the

SCT in a comprehensive review, including some unpublished

studies. She found general but not decisive support for

theoretical sequentiality, the developmental process of

the stages, in cross-sectional and longitudinal studies.

There were substantial correlations between the SCT and

external criteria such as other developmental stage tests

(e.g., Kohlberg's moral maturity) and behavioral measures

(e.g., conformity, delinquent and helping behaviors).

Correlations with isolated traits specific to stages ranged


Evidence of validity for low ego levels is primarily

behavioral (e.g., delinquent and deviant behaviors) while

evidence for high ego levels is based in correlations with

attitudes, philosophy of life, self-insight, and ability to

communicate empathy (Loevinger, 1979). Loevinger (1979)

found there was generally positive support for external

validity, although she added that ego development theory

does not clearly define what are positive results. She

concluded the SCT had adequate validity for research pur-

poses if administered and scored properly (i.e., not mailed

to subjects, scored by two trained raters). When used

clinically, she cautioned it should be used in conjunction

with other diagnostic data.

Administration and scoring

Loevinger's instructions for administering and scoring

the SCT were followed in this study (Loevinger & Wessler,

1970; Loevinger, Wessler, & Redmore, 1970). Subjects were

given the SCT in small group settings supervised by an

experienced administrator and trained scorer. The most

recently revised SCT, Form 81, was used. Responses to

individual items on the individual protocols were typed on

separate sheets in random order to insure anonymity. All

responses were scored item-by-item following the scoring

manual by two of three experienced raters (two doctoral

candidates in clinical psychology, one doctoral candidate

in counselor education); one rater scored all the responses.

The raters had 79% perfect agreement on the item responses,

and 93% agreement within a half-step for item responses.

These percentages were quite comparable to those reported

by other investigators (i.e., Hoppe & Loevinger, 1977;

Nettles & Loevinger, 1983). When there was disagreement,

a final rating for each item was determined by consensus or

by a third rater.

Individual protocols were scored using the "automatic

rules" for assigning overall ego levels (TPR ratings). An

item sum score was also tabulated for each protocol, using

the final item ratings, by giving each item score

numerical values as follows: I-2=1, 2/A=2, A=3, A/3=4,

I-3=5, I-3/4=6, I-4=7, I-4/5=8, I-5=9, I-6=10.

Repertory Grid Technique

The students' cognitive perceptions of their clients

were measured by the Repertory Grid Technique (Rep Grid)

(Fransella & Bannister, 1977). The Rep Grid is a technique

based in Kelly's (1955) theory of personal constructs and

his Role Construct Repertory Test. Kelly proposed that

each person continually evolves a unique system of cogni-

tive dimensions or "personal constructs" for interpreting

and predicting events and behavior. This hierarchially

organized system of bipolar personal constructs represents

the individual's attempts to find order and meaning in his/

her experience in the world. The constructs are the ab-

stract verbal labels the individual gives to recurring

themes or the generalizations he/she construes from events

(Neimeyer & Neimeyer, 1981).

The grid technique provides a method to elicit from an

individual a representative sample of the unique personal

constructs he/she uses to interpret and predict behavior

(Fransella & Bannister, 1977; Neimeyer & Neimeyer, 1981).

Typically, subjects are given a standard list of "role

titles" (e.g., your mother, a person with whom you usually

feel most uncomfortable, the happiest person you know

personally) and asked to name the person they know who best

fits that description. This list of "representative

people," called "elements," are presented to the subjects

in triads for comparison. For each triad the subject is

asked to describe a way in which two of them are alike

while different from the third. These elicited constructs

are the descriptions which comprise the unique personal

construct system the subject uses with the particular group

of elements elicited by the role titles. The Rep Grid thus

provides a vehicle for looking at a person's subjective

world (Neimeyer & Neimeyer, 1981), his/her unique interpre-

tations of life experiences, the concepts used to predict

events and behaviors of self and others.

The grid technique evolved from Kelly's work in psycho-

therapy, and so emphasized interpersonal relationships

(Fransella & Bannister, 1977). Much research has explored

the impact of therapists' and clients' personal construct

systems on therapy outcomes (Carr, 1980; Landfield, 1971;

Neimeyer & Neimeyer, 1981). The grid technique has been

applied, however, to a wide variety of areas, including

education, child development, social work, linguistics,

politics, social anthropology and urban planning (Fransella

& Bannister, 1977; Neimeyer & Neimeyer, 1981). These

studies have varied the elements used (e.g., people, maps)

in the grid to fit the context of the investigation.

Reliability and validity

Fransella and Bannister (1977) have outlined the prob-

lems of establishing traditional measures of reliability

and validity for the Rep Grid. Some of these problems stem

from the fact that the Rep Grid is a technique rather than

a rigid, standardized test. Other problems stem from its

basis in personal construct theory itself, which does not

assume stability for all types of constructs, for all

populations, or for all elements. In addition, since the

Rep Grid does not have a specific content (i.e., one set of

"items"), validity must be considered in the context of the

particular form being used. A variety of grid formats have

been used (e.g., rank order, rating grid, implications grid)

in different grid matrix sizes (5X5 to 45X45), and they

have been analyzed with diverse statistical procedures

(e.g., forms of cluster analysis, direct measurement of

matching between particular constructs, overall measures of

structure) (Fransella & Bannister, 1977; Neimeyer &

Neimeyer, 1981).

Various tests of reliability have yielded test-retest

coefficients in the range of .30 to .98, including both

individuals and groups, for a variety of measures of con-

structs (e.g., mal-distribution, intensity, factorial

similarity) (Fransella & Bannister, 1977). Certain kinds

of constructs seem to be used more consistently than others,

and different populations vary widely in the stability of

constructs in repeat Rep Grids. For example, one study

found test-retest coefficients of .6 to .8 for normal and

psychiatric populations but a coefficient of .2 for thought

disordered populations (Fransella & Bannister, 1977).

In terms of validity, the Rep Grid has discriminated

between clinically diagnosed groups,pre- and posttreatment

groups, and between normal and psychiatric groups (Fransella

& Bannister, 1977). Research has also supported the theo-

retical description of changes a person makes in his/her

construct system in reaction to repeated invalidation or

validation of the constructs (Fransella & Bannister, 1977).

Administration and scoring

For the purposes of this study, students were asked to

name clients as elements in the grid. Comparison of these

elements thus provided a sample of constructs the students

used to interpret and predict the behavior of their clients.

Students were asked to name actual clients since past

studies have indicated a person regards his/her own con-

structs as more important, meaningful, and useful for de-

scribing themselves and others than constructs provided by

the experimenters (Adams-Webber, Schwenker, & Barbeau,

1972; Fransella & Bannister, 1977; Landfield, 1971).

To ensure a representative sample of clients, the role

titles for naming clients were based on the list suggested

by Kelly (in Fransella & Bannister, 1977). This list was

reviewed by two Ph.D. psychologists who have conducted ex-

tensive research with the Rep Grid, and who have both

teaching and therapy experience with personal construct

theory. From a list of 21 possible role titles, each re-

viewer selected the 12 he considered the most representative

of Kelly's list and which had the least overlap. The final

list of eight role titles was selected to provide a balance

of "positive" (e.g., The client who is/was your greatest

success) and "negative" (e.g., The client who is/was the

hardest for you to understand) role titles for clients.

Each of the eight was selected by at least one of the


Rating Grid. The rating grid format (described below)

was used since it allows greater flexibility of response

(Fransella & Bannister, 1977) and analysis of both construct

content and structural interrelationships between constructs

(Neimeyer & Neimeyer, 1981).

In a semistructured group interview format, students

were asked to name the eight clients (elements) who fit each

of the descriptions (role titles). They then compared a

series of eight triads of these clients. These eight triads

were presented in the same sequence for each subject. Fol-

lowing the presentation of each triad, subjects were asked

to 1) identify a construct which described how two in the

triad are similar, 2) specify the opposite of each elicited

construct, providing the polar dimension, and 3) denote

their preference between each pair of constructs, giving a

positive sign (+) to the preferred construct and a negative

sign (-) to the other construct (e.g., + willing to be

vulnerable, keeps distance from feelings; + straightfor-

ward, manipulative).

The students then used the constructs to describe the

clients previously named. They rated each of their clients

on a Likert-type scale ranging from +3 (the preferred,

positive pole construct strongly describes the client) to

-3 (the negative pole construct strongly describes the

client). A zero rating was used to indicate either that the

construct did not apply to that client or that the two con-

structs were equally descriptive of the client (Appendix A).

The Rep Grids were analyzed by the ELTORP II computer

program developed by Landfield, Page, and Lavelle. This

analysis provides three relevant continuous scores for an

individual's construct system specific to the role titles


1. Cognitive differentiation/complexity. The Func-

tionally Independent Construction (FIC) score (Landfield,

1971) indicates the number of functionally independent or

separate construct clusters used by the subject. Clusters

are determined by patterns of interrelationships in the

ratings for both persons (elements) and constructs. The

total FIC score is the sum of clusters of functionally

equivalent constructs and clusters of functionally equivalent

persons (Landfield, 1977). Functionally equivalent clusters

indicate those constructs which are dependent on each other;

the subject does not differentiate between these constructs

when rating clients, but instead rates clients similarly on

these constructs. A high FIC score indicates the subject

is using the constructs independently, or is differentiating

between them when rating clients (Neimeyer & Neimeyer,

1931). Low FIC scores indicate a low degree of cognitive

complexity, while a high FIC score indicates a high degree

of cognitive complexity.

2. Cognitive integration/Ordination. The Ordination

(ORD) score is based on the hierarchial arrangement or

organization of the construct system assumed by personal

construct theory (Fransella & Bannister, 1977; Neimeyer &

Neimeyer, 1981). Using the absolute values of the nonzero

ratings, both the number of ratings used and the range of

ratings used are computed. The ORD score is the product of

these two indices. The computer analysis provides an ORD

score for each construct and client and an overall average

ORD score (Landfield, 1977). The average score was used in

this study.

The ORD score indicates to what degree of flexibility

a subject uses constructs to rate clients (elements)

(Fransella & Bannister, 1977; Neimeyer & Neimeyer, 1981).

Thus it describes to what degree a subject considers "shades

of meaning" between the poles of a construct (Landfield,

1977). The more integrated subject is using a more elabor-

ate construct system, one with integrative, "superordinate"

constructions. A low ORD score indicates the subject sees

few connections between events, while an extremely high

score indicates the subject construes interrelations between

almost all events (Landfield, 1977).

3. Meaningfulness/Extremity. The Extremity Score

(EXTR) has been used as an indication of the "meaningful-

ness" of the subject's constructs and elements (Neimeyer &

Neimeyer, 1981), with the more extreme ratings (+3 and -3)

reflecting more meaningfulness. The sum of the absolute

values of all ratings, the EXTR score indicates how polarized

the ratings are (Landfield, 1977). More extreme ratings

have been usually found on constructs elicited from sub-

jects than on constructs supplied to subjects (Fransella &

Bannister, 1977).

Content analysis. Students' constructs were also sub-

jected to a content analysis. Constructs were classified

into one of four content categories established by Duck

(1973) during a series of studies investigating the develop-

mental stages of friendship formation. As described by

Duck, these content categories represent a continuum of

more literal and concrete to more conceptual and abstract

constructs. They describe progression up a hierarchy from

viewing others "in terms of 'stereotypes' to a greater

individuation and differentiation of them" (p. 141) during

the formation of personal relationships. His four content

categories in progressive order are the following:

1. Physical characteristics or factual information

(e.g., tall-short; long hair-short hair; American-Spanish).

2. Interactional style (e.g., talkative-quiet; warm-

aloof; gesticulates-restrains gestures).

3. Roles or habitual activities (e.g., parent-child;

teacher-student; sings and plays guitar-can't do either).

4. Psychological characteristics, including person-

ality and cognitive attributes (e.g., sensitive-insensitive;

ambitious-not ambitious; interested in people-interested in


According to Duck (1973), psychological constructs are

more cognitively complex and require a more integrated use

of these deeper-level constructs in describing others.

He also describes a person who more frequently uses psycho-

logical constructs as one who has more potential for

developing social skills and abilities.

In Duck's (1973) series of studies, two or three inde-

pendent raters achieved interrater reliability in the range

of .75 to .90. In this study, two trained counselor educa-

tion graduate students independently rated each construct

of all the students. The raters had 78% perfect agreement;

on disagreements, consensus was reached for all of the con-

structs. The total number of constructs classified into

each category was used as a categorial variable measure of

the content of the students' descriptions (constructs) of

their clients (elements).

Appropriateness to the study

The grid technique was developed in the context of

Kelly's (1955) personal construct theory. Fransella and

Bannister (1977) have cautioned against the use of the

technique independent of the theory. Loevinger (1976)

specified Kelly as one of the theorists she incorporated

into ego development theory, and personal construct theory

and ego development theory describe similar patterns of

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs