 

MSU ‘

LIBRARIES
n

Vu—

 

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

$193355

 

 

 

 

 

 

EGO DEVELOPMENT AND COUNSELING:
THE EFFECTS OF COUNSELORS' AND CLIENTS' EGO DEVELOPMENT LEVELS
UPON THE EXPRESSED EMPATHY AND PREFERENCES OF COUNSELORS

By

Thomas Joseph McIntyre

A DISSE RTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Counseling and Educational Psychology

1985

To Eddi, with my love

ACKNOWLEDGEMENTS

My thanks to John Powell for his faith, encouragement and goodwill.

To Dick Johnson, Chris Clark and Ralph Kron for their understanding,
suggestions, and mercy.

To Joe Schroeder, Dirk Lorenzen, Joe and Marilyn Parke-Jeney, Rodger
Rinnard, Jack Loynes and Chip West for their assorted invaluable aid in the
initial planning, data collection, ratings, data analysis and housing.

To the United Dissertation and Comprehensive Workers, Solidarity!

To the Bluesmobile, for rising from the ashes to .make one last roadtrip
again and again.

To Brando, Ruby Falls Begonia, Ozzie and Harriet, and Tortillas for their
diversion and sustenance.

And one final acknowledgement and thanks to my family. for their

encouragement.

ABSTRACT
EGO DEVELOPMENT AND COUNSELING:

THE EFFECTS OF COUNSELORS' AND CLIENTS' EGO DEVELOPMENT LEVELS
UPON THE EXPRESSED EMPATHY AND PREFERENCES OF COUNSELORS

By

Thomas Joseph McIntyre

The study investigated the relationship between counselors' and clients'
ego development levels and counselors' expressed empathy and expressed client
preferences using a counseling analogue method. The research was based (on the
theories of ego deve10pment (Loevinger, 1976) and empathy (Rogers, 1975). The
hypotheses were a) counselorsl ego development is positively related to their
expressed empathy, b) counselors' expressed empathy varies among
deveIOpmental levels of analogues, c) an interaction exists between counselor
and analogue developmental levels so that expressed empathy peaks when
counselor and analogue develOpmental levels match, and d) counselors prefer
analogues at one level higher and at the same level of ego development as the
counselors. Hypotheses b, c and d were conﬁrmed.

Subjects, 42 master's level counseling students, completed the Washington
University Sentence Completion Test of ego deve10pment, responded in writing
to four typewritten analogues (representing clients at the Conformist,
Self-Aware, Conscientious and Individualistic ego development levels), and rank
ordered their preferences for the analogues. Subjects were assessed at the
Self-Aware, Conscientious and Individualistic levels. Subjects' counseling
responses were rated for expressed empathy (the dependent variable) with the
ReSponse Empathy Scale.

A 3 x 4 Analysis of Variance for repeated measures demonstrated a

non-significant main effect for subject level, a significant main effect for

analogue level (2 < .05, non-directional), and a significant interaction (3 < .05)
between subject and analogue levels. A Chi sguare indicated the subjects
preferred analogues whose developmental level was the same or one higher than
the subjects'. The data showed the subjects' expressed empathy peaked in
response to analogues at the same developmental level. Post hoc tests indicated
a) lower (2 < .05) expressed empathy in response to the Conformist analogue
than the Conscientious analogue, b) the Self-Aware subjects had lower
expressed empathy (p < .05) with the Conformist and Individualistic analogues
than with the Self-Aware and Conscientious analogues, c) the Individualistic
subjects had higher expressed empathy (p < .05) with the Individualistic analogue
than did the Self-Aware subjects, and d) the overall trend in the preference
data was subjects across levels indicated lowered preference for the Conformist
analogue.

Implications and recommendations for counselor training, supervision,

practice and research were discussed.

TABLE OF CONTENTS

List of Tables
List of Figures
Chapter 1: The Problem
Purpose
Theory
Hypotheses
Overview
Chapter 2: Review of Literature
Loevinger's Theory of Ego Development
Stages of Ego Development
Measurement of Ego Development
Summary of Loevinger's Theory
Empathy
The Empathy Concept
The Measurement of Empathy
Response Empathy Scale
The Use of Counseling Analogues in Research on Counseling
Ego Deve10pment and Empathy
The Theoretical Relationship
Empirical Evidence of the Relationship of Ego Development to
Empathy
Summary of Ego Development and Empathy
Summary of the Review of Literature

iv

vi

11

ll

11

16

19

25

30

30

33
37

38

Chapter 3: Design of the Study
Sample
Procedure
Measurement of Ego Development
Client Analogues
Measurement of EXpressed Empathy
Design
Testable Hypotheses
Statistical Analyses
Summary of Method
Chapter 4: Results
Expressed Empathy
Subjects' Preferences for Client Analogues
Expressed Empathy by Sex of Subjects
Chapter 5: Discussion
The Parameters of Generalizability
Review of Results and Discussion
The Main Effect for Subject Level
The Main Effect for Analogue Level
The Interaction Between Subject and Analogue Levels
of Ego Development
Trends in Counselor Preferences
Recommendations
Appendix A: Washington University Sentence Completion Test
Appendix B: Client Analogues
Appendix C: Revised ReSponse Empathy Scale

List of R eferences

41

41

41

42

43

44

45

46

47

48

49

49

S4

55

56

56

58

58

60

61

62

64

68

71

76

79

LIST OF FIGURES

Figure 4.1: Mean RES Total Empathy Scores by Subject Level

Figure 4.2: Mean RES Score Across Subjects by Analogue Level

Figure 4.3: Mean RES Total Empathy Scores by Subject Levels

vii

SO

51

S3

Chapter 1: The Problem

Since Loevinger's elucidation of her concept of ego
deve10pment (1966) and the publication of an assessment technique for this
construct (Loevinger, Wessler & Redmore, 1970) there have been a growing
number of investigations involving ego deve10pment. These investigations have
had various emphases. Some studies have sought evidence for the validity of ego
development theory, some studies have examined correlations between measures
of ego deve10pment and other constructs, and other studies have sought to

induce deve10pmenta1 change in subjects by way of experimental interventions
(Loevinger, 1979). Some studies have also related level of ego development to a
counselor's level of empathy.

Historically, a counselor's empathy has been considered a key ingredient
leading to successful outcome of counseling for a client. Recently empathy
theory has been broadened to consider empathy as a multifaceted process
occurring within a relationship rather than a constant ability which a counselor
exercised.

Swensen (1980) stated that ego deve10pment theory may form a basis for a
unified approach to the understanding of counseling. A study relating ego
development of counselors to their general empathic ability is part of this basis
(C arlozzi, Gaa & Liberman, 1983) as is an effort to use ego development theory
as a way of understanding client dynamics (Young-Eisendrath, 1982). What has
not been considered is how a counselor's level of expressed empathy may be a
result of and vary according to both the counselor's and client's levels of ego
development. This study investigates this question and therefore is an
opportunity to understand the developmental level of counselors, client dynamics
and counseling process simultaneously. Furthermore, this study seeks to establish
whether the relationship between counselor and client level of ego development

1

2

may result in differences in a counselor's preference for a client. This
asymmetry in preference was established in the case of moral development
theory (Rest, Turiel & Kohlberg, 1969). The uncovering of an asymmetry in
preference due to the relationship of counselor to client level of ego
development would be an important contribution to the understanding of
counseling and the client selection process.

Both Loevinger's theory of ego development (1966, 1976) and Rogers' ideas
about empathy (1957, 1975) deal with a person's internal frame of reference. A
clariﬁcation of the relationship between empathy and ego'development may be
of interest to those involved in counseling and the training and supervision of
counselors since empathy is considered important in successful counseling. In
addition, ego developmental theory is thought to present a potential addition to
the understanding of the counseling process, but there is little empirical basis
for this hope as of yet.

Purpose

The present study examines three interrelated tOpics. First, it is an
attempt to confirm the relationship between ego developmental level and
expressed empathy. Secondly, it focuses upon the interaction between ego
deve10pmenta1 level of counselor and client and the counselor's expressed
empathy through which he or she demonstrates understanding of the client's
internal frame of reference. With this aspect the present study makes a unique
contribution. Finally, an attempt is made to understand how a counselor's
preference for a client relates to counselor and client level of ego development.
Theory

The frame of reference a person uses to bring coherence and meaning to
his or her experience of self and others is the focus of the theory of ego

development which Loevinger (1966, 1976) has constructed. It is this theory

3
which Swensen (1980) claims has the potential of providing a framework for the

understanding of counseling. Yet there appears to be little research in existence
which would bring substance to Swensen's claim (Loevinger, 1979; Carlozzi et
a1., 1983).

Carlozzi et a1. (1983) hypothesizes that there is a relationship between ego
deve10pment and empathy. The construct of empathy (Rogers, 1957) is a thread
which has run through much of counseling practice and research for decades.
According to Rogers' definition, "Being empathic is to perceive the internal
frame of reference of another with accuracy and with the emotional components
and meanings" (1959, p. 210).

It may be inferred that the individual's internal frame of reference is a
point at which the relationship between ego deve10pment and counseling may be
examined. Research has just begun to investigate the relationship between a
counselor's ego deve10pment, which determines that counselor's frame of
reference, and measures of the counselor's empathic ability which contributes to
the counselor's understanding of a client's frame of reference (Carlozzi, et a1.,
1983; Zielinski, 1973, cited in Loevinger, 1979). Of course, it was the client's
internal cognitive and affective frame of reference which Rogers (1957, 1959,
1975) focused upon. What the literature has not addressed yet, except for one
inconclusive example (Young-Eisendrath, 1981, cited in Young-Eisendrath, 1982),
is the question of how frames ofreference of both counselor and client, as
determined by ego development, effect the counselor's eXpressed empathy.
Through this expressed empathy the counselor demonstrates an understanding of
the client's frame of reference, but this understanding may be shaped according
to the counselor's own frame of reference.

The present study, then, may contribute to the literature by clarifying the

relationship between counselor and client levels of ego development and the

4

counselor's expressed empathy. An understanding of this relationship might have
implications for the content of counselor training, provide a potential focus for
counselor supervision, and form one basis for optimal counselor and client
matching.

Hypotheses

1. A positive relationship exists between counselors' level of ego
deve10pment and level of expressed empathy.

2. Counselors' expressed empathy will vary in relation to clients' ego
development level.

3. Counselors will demonstrate a higher degree of eXpressed empathy in
response to a client at the same level of ego development compared to clients
at levels higher or lower than the counselors'.

4. Counselors will assign a higher preference to clients of equal ego
developmental level than those at levels lower or more than one level higher
than the counselors'.

Overview

Two important tapics remain before the reader can put the results of this
study in perspective. In Chapter 2, the literature which is relevant to this study
and its implementation is examined. This includes a look at the theories of ego
development and empathy, the few studies which have so far examined their
relationship, studies dealing with the possibly mediating characteristic of
preference, and parallel studies dealing with the theories of conceptual level
and moral deve10pment. In addition, areas of the literature which deal
speciﬁcally with methodological aspects relevant here, such as the use of
counseling analogues and of empathy measures, are reviewed.

Next, in Chapter 3 the methodology is set forth. Characteristics of the

sample, the instruments, methods of ratings and reliabilities obtained,

5
hypotheses, design and statistical analyses are described. It is hoped that the

reader will arrive at the end of Chapters 2 and 3 with a thorough understanding
of the foundation of previous research and theory and the subsequent method
upon which the results of this investigation stand.

In Chapter 4 the results of the statistical tests of this study's four
hypotheses are presented. The results of post hoc tests which contribute to an
overall understanding of this study's ﬁndings are also presented. The ﬁndings
and their implications are discussed in Chapter 5 along with recommendations

for practice, research and training.

Chapter 2: Review of Literature

Several areas of theory and research are relevant to the present
investigation. Therefore, in the following review, the reader becomes acquainted
ﬁrst with the theory of ego development, its measurement, reliability and
validity, and then with the theory of empathy, and its importance, variability
and measurement. Recent re-conceptualizations of empathy and a focus on the
uses of counseling analogues in empathy research are included here also. Finally
these two theories are related to each other and the basis in previous literature
for the present hypotheses discussed.

Loevinger's Theory of Ego Development

 

Jane Loevinger (1966, 1976) has developed a synthetic theory of ego
development which, while having links to a broad spectrum of earlier theories of
personality, can be seen as a unique inﬂuence in current psychological
understanding. Swensen (1980) noted this theory as having the potential to
provide a general framework for the understanding of counseling.

Loevinger conceived of the ego as a uniﬁed framework which the
individual uses to perceive and interpret the world. This framework included
impulse control, character development, interpersonal style, conscious
preoccupations and cognitive style (Loevinger, 1976). In addition, Loevinger
postulated that this ego deve10ps through qualitatively different steps. Ego
development is seen as a process parallel to, but separate from psychosexual,
intellectual and physical development.

Loevinger's...conception of ego development assumes that each person has

a customary orientation to himself and to the world and that there is a

continuum (ego deve10pment) along which these frames of reference can be

arrayed. "In general, ego development is marked by a more differentiated
perception of one's self, of the social world, and of the relations of one's
feelings and thoughts to those of others" (Candee, 1974, p.621). It is the
sequence of steps along this continuum of differentiation and complexity

that Loevinger and her co-workers have delineated as stages of ego
development. (Hauser, 1976, p.928-929)

7

These ideas have not remained solely in the realm of theory. Crucial to
Loevinger's work has been the arrival at a method of assessment, which led to
theory testing and modiﬁcation. Loevinger, et a1. (1970) developed a projective
sentence completion test which assesses an individual's ego development as a
uniﬁed construct. This measure, the Washington University Sentence Completion
Test (W USC T), repeatedly has been shown as reliable, and an accompanying
self-training program produces raters equally reliable to those tutored or
previously experienced ones (Loevinger, 1979; Hauser, 1976; Holt, 1980). The
WUSCT has been shown to have sufﬁcient validity for research purposes
(Loevinger, 1979).

_S_t_a_gg of Ego Develment

Loevinger delineated .10 stages and transitional levels of ego development.
Nine of these are measurable with the WUSC T, the ﬁrst is not. The transitional
levels are transitional only in the theoretical sense, as individuals may remain
stable at these levels over time.

The stages and associated symbols are: Presocial, I-l; Impulsive, I-2;
Self-protective, Delta; Ritual-Traditional, Delta/3; Conformist, I-3; Self-aware,
also known as the Conscientious/Conformist Transition, I-3/4; Conscientious, 1-4;
Individualistic, I-4/ 5; Autonomous, I-5; and Integrated, I-6.

Qmmon stﬁes for adolescents and adults. Holt (1980), in a national survey
using a 12-item short form of the WUSCT, found the modal level for young
adults to 'be the conscientious/conformist level, I-3/4. Other studies have agreed
and have indicated I-3/4 to be the modal level for adults (Hauser, 1976).
Additionally, Holt's data indicated that 832 of college men and 892 of college
women were assessed at stage 1-3, I-3/4, I-4 or I-4/5. The corresponding stage
names are the conformist, conscientious/conformist, conscientious and the

individualistic sta ges.

8
A closer look at the common adolescent and adult stages of Loevinger's
theory is useful as the interaction of these four stages of development were
examined within this study. - '

Conformist stage. The conformist stage, I-3, is one which most people

 

reach during childhood or adolescence. A person functioning at this stage is
very concerned with following rules and with the disapproval and shame
involved with breaking rules. Relatively little differentiation of internal states
is apparent and the affect reported is generally banal. The focus of interactions
with others is on actions and events instead of feelings and motives (Loevinger,
1976; Hauser, 1976).

Self-aware level: The conscientious/conformist transition. Beginning at the
self-aware level, I-3/4, an individual gains the capacity for introspection,
psychological causation, self-awareness and self-criticism. It becomes apparent
to the individual that the family or social groups can no longer provide
workable, clearcut rules which corresponding to reality. Therefore the rightness
of an action is judged according to relative aSpects of the situation, such as
time and place. These contingencies are simplistic, however, and are to become
more differentiated at higher levels (Loevinger, 1976; Hauser, 1976).

Conscientious stage. Idealism and "an excessive and sometimes intrusive

 

sense of responsibility for the welfare of other people", (Young-Eisendrath,
1982, p.332) are two notable characteristics of those persons at the
conscientious stage, I-4, of ego development. Obligations, self-criticism and
achievements measured by inner states and rules occupy consciousness. Feelings
and motives rather than actions alone are the currency of this individual's view
of social interaction. Non—stereotyped individual differences are perceived and
"social interaction is experienced as more vivid and meaningful than in earlier

stages" (Hauser, 1976, p.931).

9

Individualistic level. Greater conceptual complexity, a concern for

 

emotional independence and an increased sense of individuality characterize
those at the individualistic transition level, I-4/5. At this point persons are
more aware of inner conﬂict as well as conﬂicts between one's own needs and
those of others. An important change from the conscientious stage is that
persons are more accepting of themselves and others and are more tolerant of
differing viewpoints and Opinions of others. It is still difﬁcult, however, for an
individual to tolerate internal conﬂict which has increasingly entered
awareness, or to tolerate the notion that conﬂict and paradox are a part of the
human condition (Swensen, 1980; Loevinger, 1976).

Measurement of Egg Development

 

Washiﬂton University Sentence Completion Test. Loevinger's concept of

 

development is assessed via the W080 T, a sentence completion test of 36 items,
with separate forms for men and women (see Appendix A). The WUSCT "has
been carefully constructed and standardized in terms of its form, administration,
and scoring procedures" (Hauser, 1976). Since the publication of a scoring
manual (Loevinger, et al.,1970) it has been extensively used for research
purposes. The scoring manual consists of individual manuals for each sentence
stem, a self-training program and the ogive rules by which the cumulative
frequencies of the individually scored items are examined and converted into a
single total protocol rating (TP R). The extensive self-training program produces
raters as reliable as tutored or experienced ones (Loevinger, 1979; Hauser, 1976;
Holt, 1980).

Reliability of the scoring procedure. When the original WUSCT scoring

 

manual was validated Loevinger reported a median product-moment correlation
coefﬁcient between experienced raters of E = .75 (Loevinger, 1979). The median

correlation for TPR's was 5 = .86. Subsequent ratings of the same sample of 543

10
protocols of a diverse group of women and girls by four self-trained raters
produced results indistinguishable from that of the experienced raters. Various
pairs of experienced and self-trained raters produced a median interrater
correlation on item ratings of 5 = .76 (Hauser, 1976).

Holt (1980) reported reliabilities obtained in the scoring of a national
sample of 966 persons between the ages of 16 and 25 years. The 12-item
WUSCTs were scored using a team of graduate and advanced undergraduate
students at New York University. Reliabilities computed with intraclass
correlations were reported. Interrater reliability on individual items ranged from
.72 to .90, median .825 (women), and .57 to .90, median .78 (men). This study
provided important additional evidence for interrater reliability as scoring done
by a group separate from Loevinger's and on a nationally representative sample
veriﬁed consistency of results.

Evidence for validity. The internal validity of the WUSCT rests on a
number of points. This measure has face or content validity as its projective
nature permits the subject to supply his or her own frame of reference. It is
this frame of reference which the test tries to measure. The conception of ego
deve10pment can be communicated coherently. The statistical homogeneity of
the WUSCT has been shown and research aimed at breaking the items into
subscales has failed. Thus, there is strong evidence that it measures a single
construct (Loevinger, 1979).

"Evidence for construct validity is substantial, but falls short of clear
proof of sequentiality" (Loevinger, 1979). In any case this theory can be viewed
as a series of character types. Evidence for sequentiality is supportive of
construct validity, as Loevinger summarized:

The proﬁle or distribution of item ratings for protocols at adjacent

stages are more alike than at non-adjacent stages. Mean ego level

increases with age for cross-sectional samples during adolescence.
Mean ego level increases with age for longitudional studies during

ll

adolescence. Test-retest correlations are signiﬁcantly positive for

longitudional studies, even over spans of 6 years. Attempts to raise

ego level eXperimentally in a few weeks have not succeeded, but

eXperiments that have lasted 6 to 9 months have had statistically

signiﬁcant success. Finally, peOple can lower their score more

reliably and decisively than they can raise it. (1979, p. 305-306)

In addition, Loevinger (1979) summarized evidence for external validity in
the form of signiﬁcant correlations of low to moderate magnitude between the
WUSCT and a variety of other measures. These included interviews, objective

tests, projective tests and behavioral measures.

Summary of Loevinger's Theory

 

Loevinger's theory of ego development (1966, 1976) has focused upon the
evolving internal frame of reference which an individual uses to perceive and
interpret the world. The location of individuals upon this common path of
development is assessed using a projective sentence completion test._ Research
has shown this test, the WUSCT, to be sufficiently reliable and valid for
research purposes.

my

Several areas of theory and research focusing upon empathy are relevant
to the present investigation. These areas are reviewed below. This review
includes a look at the deﬁnition and importance of empathy in counseling, the
commonly used measures of empathy and criticisms of these measures, the
characteristics and validity of the Response Empathy Scale, and the strengths
and limitations of counseling analogues employed in empathy research.

The Empathy ConceLt

Rogers' view. In Rogers' formal statement deﬁning his concept of empathy
were several key features. He referred to a "state of empathy" (1959, p.210) as
if it were something constant. This state had to do with perceiving another's
"internal frame of reference...with accuracy and with the emotional

components" (1959, p.210), and the empathic person was to maintain an 'as if'

12
quality. It was this concept, along with Rogers' views on congruence and
warmth (1957), which generated much research in following years (Gladstein,
1983).

Rogers' (1975) re-examination of empathy underlined his Opinion that the
continued use of empathy as a concept is valuable in training and research, and
as a key part of human communication. Rogers also introduced a reﬁnement to
his views.

No longer did Rogers (1975) refer to a state of empathy. Rather, Rogers
presented empathy as a process with several facets. First, one must sense
another's "private perceptual world" (Rogers, 1975, p.4). Next, one must
communicate these sensings. Finally, this communication is received and one can
be guided by the other's response.

Reconceptualizations 1y Barrett-Lennard and Gladstein. Two subsequent

 

viewpoints, Barrett-Lennard (1981) and Gladstein (1977, 1983; Feldstein 8
Gladstein, 1980) added depth and clarity to Rogers' (1975) deﬁnition.
Barrett-Lennard examined the process of empathy Rogers refers to. Three
distinct and temporally sequential phases are speciﬁed in what he called the
empathy cycle. These are empathic resonation (Phase 1), expressed empathy
(Phase 2), and received empathy (Phase 3). Barrett—Lennard (1981) went on to
locate previous empathy measures in relation to the cycle. He predicted that
only moderate correlations can be expected between measures of adjacent
phases of the empathy cycle. One use of this cyclic conception may be to
specify what part of the cycle is being investigated in a particular research
effort.

Gladstein (1983) reviewed the various theoretical models used in the
deﬁnition and measurement of empathy. He noted two general elements in the

many empathy deﬁnitions employed in the counseling, developmental and social

l3
psychology literatures: cognitive and affective empathy.
Cognitive empathy refers to intellectually taking the role or perSpective of
another person. That is, seeing the world as the other person does.
Affective empathy refers to responding with the same emotion to another
person's emotion. That is, feeling the same way as another person
does...Rogers' definition (1975, p.4) included both. (Gladstein, 1983, p.468)
Gladstein's (1983) review was stimulated in part by the diversity of
outcomes in the literature on empathy. His review concluded that empathy may
or may not lead to helping behavior. Therefore, he said it is essential to
determine when each element, cognitive and affective, is or is not important.
He further noted that based on theoretical consensus it appeared that affective
and cognitive empathy would be important in initiating and building a counseling
relationship. Gladstein also mentioned cognitive role-taking empathy as
important in the problem-identification and exploration periods of counseling.
Gladstein (1983) stated that empathy is a multi-stage process agreeing with
Rogers' (1975) and Barrett-Lennard's (1981) comments. He placed the affective
component prior to the cognitive component temporally. It is apparent that both
affective and cognitive empathy would occur within the counselor prior to
Phase 2 of Barrett-Lennard's cycle. Resulting communication, expressed
empathy, takes place in Phase 2. It is also important to note that
Barrett-Lennard's definition of cognitive empathy included an understanding of
the "client's feelings and thoughts ﬁom their frame of reference" (Feldstein &

Gladstein, 19 80, p. 54).

The importance of empathy. The above viewpoints concerned the nature of

 

empathy, but did not address its importance in counseling. Rogers (1957) called
empathy one of the "necessary and sufficient conditions for therapeutic
personality change." Rogers (1975) concluded from his review of the literature
that "empathy is clearly related to positive outcome" (p.6). He based his

conclusion of positive outcome in part upon studies which have found a positive

14
relationship to exist between empathy measured from a counselor's or
therapist's actual interview behavior and outcome measures (Kurtz & Grummon,
1972; Mullen & Abeles, 1971). He reasoned that this positive outcome is due to
empathy's effect in dissolving alienation, and in increasing the client's sense of
self-value and identity.

Gladstein (1983) noted that challenges had been raised regarding the above
conclusions and that research evidence for empathy's importance was mixed. He
pointed out that mixed ﬁndings might have been due to the differing and
sometimes obscure ways in which theorists and researchers had deﬁned and
measured empathy. He noted some conceptual reﬁnements which could prove
helpful to clear up the mixed results. Gladstein concluded, however, that
empathy "can be helpful in certain stages, with certain clients, and for certain
goals" (1983, p.467). He did note cognitive and affective empathy as important
in the early stages of counseling, particularly when the counseling goals were
self-exploration. He was less supportive of the importance of empathy,
especially of affective empathy, when the goals involved Specific
problem-solving or action-oriented decision making.

Empathy's variability. Rogers (1959) originally spoke of a state of empathy.

 

His more recent view (1975) and those of others (Barrett-Lennard, 1981;
Gladstein, 1983) explained empathy occurring as a multi-phasic process. The
magnitude of empathy may vary, particularly as expressed and received
empathy. Several reports have conﬁrmed that a counselor's level of eXpressed
empathy may vary.

Truax (1966) found that a counselor's empathic behavior varied in a manner
contingent upon the patient's behavior. Curman (1973) measured the level of
therapeutic conditions of empathy, warmth and genuineness of three high- and

three low-functioning therapists. The levels of both groups were found to vary

15
across and within sessions. Changes across time within sessions were noted such
that the therapists' empathic behavior and other conditions peaked in the middle
and'end of the session.

As noted by Rogers (1975), the hvel of a counselor's empathic behavior
may vary positively according to the counselor's psychological adjustment.
Bergin and Soloman (1970) compared the scores of 18 post-internship psychology
graduate students on the MMPI and the Truax accurate empathy scale. A
negative correlation was noted between the MMPI Depression and Psychasthenia
scales and the scores for empathic behavior. Vesprani (1969) noted the same
relationship with 33 untrained college student subjects.

Summary of the empathy concept. Some speciﬁc points deserve emphasis.
Empathy is a process not a state. Barrett-Lennard's Phase 1 empathic
resonation is composed of both affective and cognitive elements. Empathy's
behavioral component, Phase 2, expressed empathy, may vary. Empathy is
important in counseling in that at the least it is sometimes helpful.

The reconceptualizations reviewed above demand consideration in planning
research and received it in the present study. Some questions need to be
answered. Which element of Phase 1 empathy is involved, cognitive and/or
affective; at what point is the cycle being examined, empathic resonation,
expressed empathy or received empathy? Is the measurement one which can
focus upon the particular element and phase involved; and, ﬁnally, is empathy
being examined at a point in counseling at which it might be expected to
matter?

The answers to the above questions can help focus a research project,
conﬁrm the choice of measurement, and make sense of diverse research ﬁndings
in the empathy literature. The present study, for instance, examined the result

of cognitive role-taking empathic resonation evident in expressed empathy. This

16
study used an analogue representative of a time in counseling (initial counseling
contact) when empathy is important, and used a measure which focused upon
verbal Phase 2 expressed empathy.

The Measurement of Empathy

 

It has been noted above that empathy may vary. This is an important point,
as the present study investigated the variability of counselors' expressed
empathy relative to a client and counselor characteristic, ego development. In
order to accurately measure this variability and thus provide a test of this
study's hypotheses, it was essential that the subjects' expressed empathy be
measured in a reliable and valid fashion. The review below, of the various
measures of empathy available, resulted in the selection-of the needed measure.

Commonly used measures. Following Rogers' (1957, 1959) conceptual

 

advances, a number of different measures were deve10ped to examine empathy
in counseling. The Barrett-Lennard Relationship Inventory (BL RI)
(Barrett-Lennard, 1962) attempted to measure empathy via the perceptions of a
relationship's participants. The Truax Relationship Inventory (TRI) (Truax &
Carkhuff, 1967) attempted the same. Both measures are paper and pencil
instruments completed by persons in a counseling relationship. Feldstein &
Gladstein (1980) noted that both are subjective measures clearly linked to
Rogers' empathy construct (Rogers, 1957, 1959). Both also appear to focus upon
the empathy cycle at Phase 3 (Barrett-Lennard, 1981).

Objective measures using external, independent judgments of counseling
mtewﬁw behavior were also deve10ped (Feldstein & Gladstein, 1980). These
include Truax's Accurate Empathy Scale (AE)'(Truax & Carkhuff, 1967) and
Carkhuff's (1969) Empathic Understanding in Interpersonal Process Scale (EU). A
more recently developed objective measure is the Revised Response Empathy

Scale (RES) (Elliott, Reimschuessel, et a1., 1982). All three are attempts to

17

measure what Barrett-Lennard (1981) referred to as Phase 2 empathy.

In contrast, the Affective Sensitivity Scale (ASS) (Kagan, Krathwohl and
Associates, 1967), examines processes identiﬁed with Phase 1 empathy
(Barrett-Lennard, 1981), but it does not examine those processes within any
one particular relationship. The ASS is administered by asking subjects to
respond to multiple choice lists of affect descriptions following videotape
segments of an interview.

Criticisms of empathy measures. A number of criticisms have been leveled

 

at the commonly used measures of empathy. Chinsky and Rappaport (1970)
focused upon Truax's AE scale. They questioned the meaning of the scale. They
noted that a validation study of the AE (Truax, 1966, cited in Chinsky &
Rappaport, 1970) found no signiﬁcant differences between ratings of therapist
behavior with client responses present and ratings with client responses absent.
Yet the defining points of the AE scale Specifically included reference to client
responses. Chinsky and Rappaport (1970) concluded that the AE scale was
measuring qualities other than those reﬂected in the deﬁnition of the scale.
Fﬁdman and Stone (1978) also found that the Carkhuff EU scale was not
sensitive to different rater stimulus conditions. They found ratings of counselor
behavior from conditions with client responses present to be comparable to
ratings with client responses absent.

Gormally and Hill (1974) questioned whether Carkhuff's EU scale was
measuring a global counselor characteristic beyond those deﬁned by the scale.
They based this criticism ﬁrst upon factor analyses which showed one factor
accounted for 792 of the variance in one study and 892 in another. Secondly,
strong correlations had been found between Carkhuff scales measuring empathy,
warmth and genuineness and global therapist characteristics. ‘

Chinsky and Rappaport (1970) also noted that Truax & Carkhuff's (1967)

18

indicated an inverse relationship between reliability obtained for the AE scale
and the number of therapists rated in any particular study. They speculated that
this too might have been due to the scale measuring therapist qualities other
than those Speciﬁed. Finally, Chinsky and Rappaport pointed out that the
Pearson 3's reported as reliability ﬁgures by Truax (1966) were spuriously
inﬂated by the use of an 3 in the calculations based upon the number of
repeated measures of the same therapists rather than the number of therapists
itself. Thus the assumption of independent measures was violated. Their
criticisms, then, questioned both the meaning and hence the construct validity,
of the AE scale and its reliability.

Kurtz and Grummon (1972) compared empathy ratings using six different
measures of counselor empathy. They compared these counselor empathy ratings
to ratings of client perceived empathy and a well constructed outcome measure.
Their data were collected from actual therapy cases. They found no signiﬁcant
correlations between the empathy measures, including the BLRI, Carkhuff's EU
scale and Kagan's ASS. They concluded from the results that the six measures
were assessing six different variables. In addition, no relation between the
measures and either client perceived empathy or outcome was established,
except for a moderate correlation between Carkhuff's EU scale and client
perceived empathy, (Barrett-Lennard's (1981) Phase 3 empathy). The EU scale
was the only one of the six empathy measures to be based upon actual therapist
interview behavior. The conclusions of Kurtz and Grummon (1972) questioned
the construct validities of all six empathy measures used. Based on these
ﬁndings need for an empathy measure with demonstrated construct validity is
indicated.

Feldstein and Gladstein (1980) derived from Rogers' (1975) theoretical

discussion six criteria applicable for evaluating the construct validity of

19

empathy measures. The three criteria they considered the most important
included a focus upon a) cognitive empathy, b) affective empathy, and c) the
counselor's communication of his or her empathic experience. Of the four
measures reviewed: Carkhuff's EU, Truax's AE and TRI, and Barrett-Lennard's
BL RI, none were found satisfactory according to their criteria.

Criticism is warranted regarding the Feldstein and Gladstein (1980) review
in light of Barrett-Lennard's (1981) description of the empathy cycle. It may be
too much to expect for any one instrument to be able to measure all three
phases of the empathy cycle simultaneously. To accomplish the above,
simultaneous access to the counselor's internal experience, the counselor'stand
client's interview behavior, and the client's internal experience would be
required. However, different measures of the three phases as separate entities
might be sought. One or more measures could tap into the cognitive and/or
affective components of the counselor's empathic resonation at Phase 1, another
could focus upon the counselor's expressed empathy at Phase 2, and yet another
could look at the client's perception of received empathy at Phase 3. An.
important step in the validation of the measures would be determining what
relationships exist between measures of the different phases (Barrett-Lennard, ,
1981).

The RES (Elliott, Reimschuessel, et a1., 1982) has been held for discussion
until this point largely because its deve10pment and validation are too recent
for it to have been reviewed and appear in the critical reviews above. However,
it was apparent that the authors took into consideration the preceding research
critiques and theoretical reconceptualizations. A review of the RES

demonstrated its relative strengths and limitations.

20

Response Empathy Scale

 

The RES was developed (Elliott, Filipovich, et a1., 1982) and revised
(Elliott, Reimschuessel, et a1., 1982) partly in response to Barrett-Lennard
(1981). Some features of this new measure are: a) its speciﬁc focus on Phase 2,
expressed empathy, b) its subscales measure different aspects of expressed
empathy, c) it is designed for use with individual counselor responses. The
reliability and validity of the RES rests upon two studies conducted by Elliott
and his associates.

Initial validity study. Elliott, Filipovich, et a1. (1982) attempted to

 

investigate three issues regarding the RES. a) Estimates of rater reliability
were sought at the level of the total RES and its subscales; b) Using factor
analysis, they sought to determine the core elements and underlying components
of response empathy when measured in this way; and, ﬁnally c) evidence for the
RES' validity was sought in its relationship to clients' evaluations of being
understood.

The data were sampled ﬁom 28 thirty-minute interviews in which 28
volunteer undergraduate clients were asked to discuss 'a personal problem. The
counselors were 12 internship level graduate students and three faculty members
in clinical psychology. Counselors' theoretical orientations were described as
client-centered or psychodynamic. Three samples of four to seven counselor
responses (episodes) were videotaped within each interview. Five raters used the
RES to rate counselor responses, rating each on the nine subscales before going
on to the next response. An adaptation of Kagan's (1975) videotape recall
technique was used to obtain client perceptions of being understood after the
interview. Clients were asked how understood they felt after each counselor
response and rated their answers on a 6-point scale. Counselor responses were

rated separately using audiotapes and transcripts with the RES.

21

Interrater reliability for total empathy scores of the RES was .91 using
Cronbach's alpha. A high degree of internal consistency for the RES was
indicated by an interitem reliability for total empathy of .82 (Elliott, Filipovich,
et a1., 1982). A

Factor analysis extracted two factors accounting for 622 of the total
variance. Factor 1, described as a Depth/Expressiveness factor, had strong
loadings for the Words, Centrality and Here and Now subscales. It accounted for
442 of the total variance. Factor‘ 2, described as an Empathic Exploration
factor, had strong loadings for the Frame, Centrality, Manner and Impact
subscales. It accounted for 182 of the total variance. The factor analysis also
provided a measure of which components were most central to empathy as
measured by the RES. The Centrality subscale, followed by the Frame and the
Words subscales were most central; Voice and Impact appeared to be the most
peripheral (Elliott, Filipovich, et a1., 1982).

Validity of the RES was investigated using correlations between observer
ratings (RES) and client's ratings of being understood. Correlations were
calculated at three levels of analysis: a) the individual counselor response, b)
episodes of four to seven counselor responses, and c) all counselor responses in
the three episodes per session. The relationship at the individual reaponse level
was small but signiﬁcant, 5 = .26. A medium sized effect was apparent for total
empathy at the episode level, 5 = .42. The association increased again when the
level of the session was examined, _r; =- .53 (Elliott, Filipovich, et a1., 1982). The
presence of this relationship is strong evidence for validity of the RES. The
investigators considered the RES to be measuring what Barrett-Lennard (1981)
referred to as Phase 2, expressed empathy, and considered the clients' ratings
to be measuring Phase 3, received empathy. Therefore a. relationship between

measures of these two phases in the empathy cycle was expected

22

(Barrett-Lennard , 1981 ).

Description of the revised Response Empathy Scale. In order to improve

 

reliability and clarity, some changes suggested by Elliott, Filipovich, et a1.
(1982) were incorporated into the revised RES. Two of the original subscales
were drapped: Accuracy and Here and Now. The EXploratory Manner subscale
was divided into two new scales: Collaboration and EXploration. The Client
Frame subscale was redeﬁned to become Client Feelings. The Expressiveness
subscale was deﬁned from the previous Voice Quality and Choice of Words
subscale. A Verbal Allowing subscale was added (Elliott, Reimschuessel, et a1.,
1982).

The revised RES has eight subscales: Client Feelings, Perceptual Inference
and Clariﬁcation, Topic Centrality, Expressiveness, Collaboration, Verbal
Allowing vs. Crowding, Exploration, and Impact on Exploration. Each subscale is
a 5-point (0-4) rating scale with descriptive anchors referring to speciﬁc
counselor behaviors. The Allowing and Impact subscales can be used only with
taped interviews. The six other subscales require only the initial client
statement or response and the counselor's reaponse, and can be used with
writtenresponses (Elliott, et a1., 1981).

Validity study for the revised Response Empathy Scale. Following revision

 

of the RES a second validation study was undertaken to assess the reliability of
the revised scale when applied to a general sample of outpatient counseling.
The design‘of this study was also inﬂuenced by Barrett-Lennard's (1981)
empathy cycle model. In addition to collecting data on expressed empathy
(Phase 2) and client received empathy (Phase 3), this study also investigated
counselor empathic resonation (Phase 1) (Barrett-Lennard, 1981).

Sixteen pairs of clients and counselors in a variety of outpatient settings

were used as subjects. The counselors were divided in experience. One half of

23
the counselors were practicing psychologists, psychiatrists and social workers,
while the other half were psychiatric residents and psychology interns. The
clients ranged in age from 18 to 45 years. Any with severe diagnoses were
excluded. Half of the dyads were sampled after brief contact (6-12 sessions);
the other half were sampled after extended contact (20-130 sessions) (Elliott,
Reimschuessel, et a1., 1982).

The methodology was as follows. Sessions were audiotaped in three 5-10
minute segments. The counseling dyads were asked to carry on their customary
inteviews. RES and client received empathy data were collected as in the
previous validity study (Elliott, Filipovich, et a1., 1982). This time counselors
also participated separately in a modiﬁed audiotape recall technique (Kagan,
1975) in which they rated a 7-point scale following each of the ﬁrst four
counselor responses in an episode. This rating was in response to the question
"'when you said that, did you feel you misunderstood or understood your client"
(Elliott, Reimschuessel, et a1., 1982)?

The revised RES was shown reliable by an alpha = .95 for interrater
reliability for total empathy. A high degree of internal consistency for the RES
was demonstrated by an interitem reliability of alpha = .81. Factor analysis
revealed three factors accounting for 742 of the total variance. Factor 1,
described as Depth/Centrality, accounted for 482 of the total variance. Factor
2, described as an Exploration factor, and Factor 3, described as a
Collaboration factor, accounted for 262 of the total variance together. The
communality values of the factor analysis showed the Centrality subscale to be
the most central component; only the Allowing subscale appeared to be
peripheral (Elliott, Reimschuessel, et a1., 1982).

The association between RES scores (Phase 2 empathy) and client received

empathy (Phase 3) was examined. At the response level a small but signiﬁcant

24

relationship between the RES total empathy score and clients' ratings of _r_'_ = .18
was obtained. This relationship was positive but insigniﬁcant at the episode and
session level. At all three levels the Collaboration subscale was positively and
signiﬁcantly correlated with clients' ratings, 3 = .32, .40, .64 (Elliott,
Reimschuessel, et a1., 1982).

The association between RES scores (Phase 2) and counselor ratings (Phase
1) was also investigated. This association was signiﬁcant and positive at both
the response and episode levels, E = .34 and _r_' 3 .52 respectively. Only the
correlation between the Collaboration subscale and counselor ratings attained
signiﬁcance at the session level. The lack of significance in the correlations
between the RES total empathy score and counselor ratings at the session level,
and with client ratings at the episode and session levels may well be due to
declining statistical power due to reduced 3. The p was lowest at the session
level, three times greater at the episode level and greater still at the response
level.

Summary of ReSponse Empathy Scale characteristics. The revised RES has

 

high interrater reliablity and appears to measure what it purports to, with the
possible exception of the Allowing subscale. Expressed empathy as measured by
the scale is related positively to measures of counselor empathic resonation and
client received empathy. This evidence for external validity has been shown in
both analogue and actual counseling settings.

The RES' measure of expressed empathy appears to consist of three
underlying factors. These core dimensions are a) "the counselor addressing
feeling-laden core client issues...by means of interpretive inferences
communicated in an expressive manner"(Elliott, Reimschuessel, et a1., 1982), b)
counselor behaviors fostering client exploration, and c) the counselor's

communication of a collaborative effort.

25

An important attribute of the RES is its focus on individual counselor
responses in terms of specific counselor behaviors. Because of this and its
construct validity as shown by factor analyses, it avoids some of the difﬁculties
linked to the Truax AE scale. In addition, its deﬁnition and demonstrated
validity in relation to Barrett-Lennard's recent empathy cycle model is an
advantage.

There are some limitations. The RES does appear to be largely a measure
of verbally communicated expressed empathy; this limitation was acknowledged
by Elliott and his associates. Because of the way in which the scale is used, the
independence of the subscale ratings relative to each other for individual
responses is also not clear. Thus, it may be advisable to use the total empathy
score obtained from the RES as the chief unit of analysis when following their
method. The construct validity of the Allowing subscale is in question. Also the
association between the RES and client received empathy at the session level
has not yet been ﬁrmly established. A larger 2 may be needed to establish this
relationship. Further validation studies could help to decide these issues.

At present, the RES appears to have sufﬁcient reliability and validity for
its use in measuring total expressed empathy in counseling research. A
reasonable unit of analysis would be the independently rated response or the
mean of response ratings in an independently rated episode of counselor
responses.

The Use of Counseligg Analogues in Research On Empathy

 

Research investigations of empathy frequently have employed experimental .
tasks which are analogues of actual counseling situations. The present study
used typewritten analogues to simulate clients at varying levels of deve10pment.
Several research reviews and empirical studies which have focused upon

counseling analogues are discussed below. The objective of this discussion is to

26

delineate the strengths and limitations of this type of research approach.

Strong'_s and Munley's reviews. Strong (1971) discussed the potential
usefulness to counseling theory and practice of experimental laboratory
research. He pointed out that this methodology gave the researcher more
control over variables of interest. This control could lead to the formation of
causal inferences not possible in correlational methods. This could "provide more
direct, unambiguous answers than would have been possible using traditional
correlational methods and 'real' clients" (Strong, 1971, p.109). Strong stated
that some phenomena of interest to the counseling profession could be
investigated in a laboratory setting which could not be examined practically or
ethically in a counseling setting. One obvious limitation of this methodology was
the difﬁculty in generalizing its results to a naturalistic counseling setting
(Strong, 1971).

Strong (1971) noted that the relevance of this method was found either in
'its implications being considered important in practice or in its direct
application to counseling. The chances of direct application were increased
when the conditions under which the research took place approximated the
conditions of the counseling setting. Strong listed ﬁve boundary conditions to be
considered in estimating the similiarity of the research setting to practice and
hence its genaralizability.

1. Counseling is a conversation between or among) persons....2. Status

differences between or among interactants constrain the conversation....3.

The duration of the contact between interactants varies, and at times is

extended....4. Many clients are motivated to change....5. Many clients are

psychologically distressed and are heavily invested in the behaviors they

seek to change. (1971, p.108)
Generalizability of research increased if its conditions matched these boundary
conditions (Strong, 1971).

Munley (1974) agreed that the major limitation of the analogue method of

counseling research was the uncertainty of generalizability. But Munley stated

27
its chief advantage was in the method's ability to investigate "Specifics of the
counseling process, eSpecially interaction effects between counselor
characteristics and behavior and client characteristics" (1974, p.328).

Because of the methodological advantages of analogue research, Munley
(1974) stated that the need for it to simulate closely the naturalistic setting
was of less importance than initially thought. However, in order to facilitate
generalizability, Munley (1974) made some recommendations. He advised that the
independent variables should be measured and manipulated with techniques and
within a range similiar to that of the actual counseling setting. He suggested
that the boundary conditions set out by Strong (1971) be considered in deciding
relevance of analogue method results. In addition he recommended that research
be undertaken comparing the results obtained in different kinds of analogues as
well as between analogues and naturalistic settings (Munley, 1974). An example
of this would be research comparing results of the same measure obtained in
written, audio and video analogues.

Munley (1974) reviewed analogue research methods in studies appearing in

the Journal of Counseling Psychology. He classiﬁed these studies according to

 

the nature of the analogue and dependent variable. The category most relevant
to the present investigation is that of audiovisual analogues with a dependent
variable of counselor behavior. Munley described the general format of this kind
of analogue.

The general experimental model of audio-visual studies, where counselor
behavior is the dependent variable, is to (a) present a counselor or group
of counselors with a client stimulus tape or film, (b) have the counselor
subjects imagine themselves in the role of counselor to the client shown,
and (c) ask for certain response behaviors at different points during the
tape. Client behaviors and/or characteristics may be employed as
independent variables and investigated in relationship to counselor
behavior. (1974, p.321)

Munley (1974) noted that this analogue method has some advantages. Large

numbers of subjects could view standardized stimulus materials with

'28
standardized experimental manipulations. The effect of client characteristics
upon a large number of subjects could be studied. Finally, research methods not
feasible in the actual counseling setting could be used.

Munley (1974) also analyzed this category of analogue study in terms of
Strong's (1971) boundary conditions. He noted that none of the studies reviewed
met the boundary conditions of actual counseling. Counselors in these studies
participated only vicariously in the relationship. While status differences were
present, Munley considered their effect to be minimal due to the short length of
contact. He noted that this short contact time may also limit the
generalizability of the ﬁndings to ﬁrst session counselor behavior. Finally,
counselor responses in all studies were written, and some responses were
restricted further by a multiple choice format.

Comparison of written with audio and audio/visual analggpe formats. One
recommendation of Munley's (1974) review was that a comparison be made
between results obtained using different types of analogue formats. As the
present investigation employed a written analogue, a review of research
comparing written with audio or audio/visual formats was appropriate.

Two studies compared the levels of expressed empathy obtained from
counselors when presented with different stimulus materials. Fridman and Stone
(1978) compared expressed empathy ratings using Carkhuff's EU scale from
subjects presented with video, audio and written client stimuli. No significant
differences in empathy ratings were found between stimulus conditions. Melnick
(1975) .also investigated the effects of differing methods of analogue
presentation. Subjects were graduate level counseling students who were
presented with role-played, videotaped and transcribed analogue clients.
Responses were rated using Carkhuff's EU scale for expressed empathy and for

affective response and exploration. The results from the videotaped and

29

transcript conditions were positively related. Signiﬁcant Pearson _r_"s were
obtained between the sets of dependent variable ratings of .90, .60 and .73
respectively. Melnick concluded that "these two methods are interchangeable as
stimuli to elicit counseling responses for evaluation purposes" (1975, p.111).

Melnick did ﬁnd that signiﬁcantly higher empathy scores were obtained
from the role-played condition. However, no signiﬁcant differences were found
between the three conditions in the ratings for affective. and exploratory
response. Melnick (1975) also noted differences in the dependent variable
according to whether the problem discussed was social-personal or
vocational-educational.

Two studies also compared different methods of obtaining counselor
responses. Carkhuff (1969) reported signiﬁcant correlations between empathy
ratings for written reSponses and for oral responses. The relationship was found
for high scoring subjects only. Therrien and Fischer (1978) extended Carkhuff's
(1969) ﬁndings to both high and low scoring subjects with a group of adult
subjects mixed as to education, age and amount of previous human relations
training. Using the Truax AE scale, they compared ratings of written versus oral
responses. The two groups of ratings were highly related with 5 = .93 over both
high and low scorers.

Stokes and Tait (1978) compared written and oral measures using
undergraduate and trained paraprofessional counselor subjects. Responses were
rated for level of reﬂective skills, verbal following and general quality. The
two conditions were a transcript of a client with written responses asked for
and a ﬁve minute videotaped client with oral responses. Strong positive
correlations were found between the sets of dependent variable ratings from the
two conditions. These correlations ranged from 5 = .53 to E = .869. The

relationships were greater for the paraprofessional subjects.

30

Hardin and Yanico (1981) compared naive subjects' perceptions of a male
counselor's characteristics in a videotaped, audiotaped and transcribed analogue
of the counselor interacting with a client. No signiﬁcant main or interaction
effects of mode of presentation upon four dependent variables were reported. In
this way the different methods were shown to be comparable. These variables
were counselor attractiveness, subject willingness to talk with the analogue
counselor, interest in the interview and identiﬁcation with the client. Some
sex-related differences were noted although the study's design did not permit
complete exploration of this effect. Male subjects rated the transcribed
counselor as less trustworthy than the counselor in other conditions and lower
than women subjects did. Additional research systematically varying the sex of
counselor and client would be needed to explore this effect completely. One
implication, though, is that raters of written counselor responses ought to be
blind to the sex of the counselor in order to avoid this effect.

Summary. The use of analogue methods in counseling research has the chief
advantage of permitting experimental control not usually possible in practice.
This is particularly true when investigating interaction effects between
counselor and client characteristics (Munley, 1974). The method's chief
limitation is the uncertain generalizability of results. This generalizability
increases as the conditions under which the research is conducted approach the
conditions of actual counseling practice. Different analogue formats have been
compared and written methods of stimulus presentation and response recording
have been shown to be comparable to audio and audio/visual methods.

Ego Development and Empathy

The Theoretical Relationship

 

Carlozzi, et a1. (1983) stated that empathic ability might be expected to

be related to level of ego development. They based this hypothesis upon the

31

structure of Loevinger's (1976) developmental framework.

The hierarchical sequence of ego development suggests increasing social

sensitivity, maturity, self-other differentiation, and decreasing

self-centeredness as one progresses to higher stages. Furthermore, ability
to empathize, which consists of suspending judgment and detecting and
responding to the feeling expressed by another, is conceptually related to

impulse control and interpersonal relations. (Carlozzi, et a1., 1983, p.113)
Carlozzi, et a1. (1983) also noted that empathy was considered to be related to
Piaget's theory of cognitive development and Rohlberg's theory of moral
development. Both of these theories have been conceptually and empirically
related to Loevinger's theory of ego development (Loevinger, 1976; 1979). Yet,
few studies have been conducted in which an empirical relationship is examined
between empathy and ego development (Carlozzi, et a1., 1983).

Swensen (1980) viewed ego development theory as having the potential to
provide a general model for counseling and psychotherapy. He noted, however,
that the relationship between ego development level and any given piece of
behavior is not likely to be simple. Swensen recommended that in seeking to
apply ego development theory to counseling an approach be taken which takes
into account both personal characteristics as exempliﬁed by ego deve10pment
level and characteristics of the enviornment. He found the origin of this
approach in Lewin's B = f(P,E) formula. Recent research which has examined
the interactions between person and enviornment in this way follows the
aptitude x treatment interaction (AT1) approach advocated by Cronbach (1957).

Swensen (1980) also stated that "another question...raised by the concept
of ego development...is the optimal differences in ego stages between the
therapist and client" (p.387). This was a question which Loevinger (1976) raised.
She cited the work of Grant and Grant (1959) as indicating the differential
effectiveness of therapists and juvenile delinquent combinations according to

the ego level of those involved. Using their own concept of ego development,

Grant and Grant (1959) found that:

32
subjects with high ego level responded best to teams of higher predicted-
efﬁciency, but the relation was reversed for offenders of low ego
1evel...ego level was presumably one component in predicting effectiveness.
(Loevinger, 1976, p.428)

Swensen (1980) hypothesized that therapists of lower ego level would be
less able to understand their clients than those therapists of higher level. This
was because they "transform the concepts of more complex levels into concepts
that are at their own (simpler) level" (Swensen, 1980, p.387). Swensen also
speculated that the optimal combination might be for a therapist or counselor to
be at a stage of ego development one level higher than the client, although he
did not explain this. Presumably, he spoke of the theory of the pacer, common
in cognitive deve10pmental literature, which Loevinger (1976) mentioned.

Loevinger (1980) agreed with Swensen (1980) that therapists should not be
at a lower ego level than the clients they work with. But she left open the
question of optimal match or mismatch, with the therapist one stage higher, as
Swensen suggested. Loevinger did see the need for therapists to "develop a
deeper understanding of clients whose own level is quite far below their own"
(1980, p.389). It appears from this viewpoint there may be something which gets
in the way of a more advanced therapist deeply understanding a client at a
stage of development which the therapist has long since passed through.

Young-Eisendrath (1982) followed up on Loevinger's viewpoint in the
development of a training program for social work students in which they
learned about characteristics of persons below their own level of deve10pment.
She reported anecdotally that post-conformist social work students "dislike the
person who sees the world from the conformist point of view and resist being
empathic with such a client" (Young-Eisendrath, 1982, p.332).
Young-Eisendrath's observation focused not only on the change in empathy due
to ego level mismatch but the factor of preference or liking as well.

Greenson's writings (1967, 1978, cited in Gladstein, 1983) provided

33

additional justiﬁcation for expecting a positive relationship between ego
deve10pment and empathy. While he stated that emotional contagion or feeling
like the client is part of being empathic, he also stated that a "capacity for
controlled and reversible regressions" (Greenson, 1967, p.369, cited in Gladstein,
1983) was essential, and that without it empathy lost its effectiveness.
Loevinger (1976) clearly placed regressions in the service of the ego as a
characteristic of post-conformist levels of ego development. Therefore, if this
regressive ability aids empathy, as Greenson (1967) believed, those persons at
higher ego levels would be expected to be more empathic.

Empirical Evidence for the Relationship ofigo DeveIOpment to Empathy

Ego deve10pment studies. On the basis of theory an empirical link was

 

suspected between ego development level and empathy. Such a link has been
tentatively established. Zielinski (1973, cited in Loevinger, 1979) found a
positive moderate correlation between the ego development level of graduate
students in a beginning counselor education course and their ability to
communicate empathic understanding as measured by Carkhuff's Index of
Communication. Carlozzi, et al. (1983) found that dormitory advisors whose own
ego development is at the I - 3/4 level and above scored higher on Kagan's ASS,
a measure of empathic resonation (Barrett-Lennard, 1981), than advisors at the
I-3 level or lower.

Young-Eisendrath (1981, cited in Young-Eisendrath, 1982) investigated the
relationship between counselor's ego level, client ego level and measures of
empathy in a client analogue study. Subjects were 34 recent graduates and
students in Master's level counseling and social work programs who participated
in either empathy skills training or training in ego deve10pment. Subject ego
level was assessed with the WUSC T. The client analogues were two 45-minute

videotapes of clients. One client had been assessed via the WUSCT at the

34

conformist stage. The other had been assessed with the WUSCT as
post-conscientious (I - 4/5 level or greater). The clients' videotapes were
considered exemplary of their ego deve10pment functioning. After watching the
videotapes, subjects were asked to respond to the Rokeach Value Survey and a
forced choice questionnaire as the client might. They also wrote brief essays
describing the clients. These essays were analyzed for agreement with the ego
deve10pment scheme and for ability to synthesize into a whole picture. The only
signiﬁcant group difference found was a higher agreement with the conformist
client on the value survey by the experimental group. Results indicated that
post-conscientious subjects described the client analogue at the
post-conscientious level better than did other subjects. Young-Eisendrath
concluded that problems with the measures may have contributed to the lack of
signiﬁcant ﬁndings. This was not surprising as these measures were rarely used
empathy measures. Kurtz and Grummon's (1972) results questioned the use of
similiar techniques based on their lack of relationship to either counseling
outcome or client-perceived empathy. In any case, Young-Eisendrath's measures
could be considered measures of Phase 1, empathic resonation (Barrett-Lennard,
1981) and of comprehension of ego development stage-related functioning.

Two studies have investigated subjects' comprehension of the ego
functioning of the various stages. Redmore (1976) found that low-level subjects
couldn't fake higher level ego development on the WUSCT even when asked
following instruction on ego development concepts. Higher level subjects were
able to respond to the WUSCT with lower level responses, however. Blasi (1976)
also showed that lower level subjects could not comprehend roles belonging to
higher ego levels. From this "asymmetry of comprehension" (Loevinger, 1979,
p.292) it could be expected that'lower level counselors would show decreased

empathy when interacting with higher level subjects, compared to their empathy

35
with equally low level subjects. These results were in keeping with those of
investigations of comprehension of stages of moral reasoning. Rest, Turiel and
Kohlberg. (1969) found that subjects had increasing difficulty comprehending
moral reasoning indicative of stages above their own.

Rest, et a1. (1969) also found that subjects tended to prefer the moral
reasoning of the highest level which they could comprehend. These results might
be extended to the realm of ego development with the hypothesis that a
person's preference for another's ego functioning and hence empathy might vary
according to the interaction of ego levels. Indeed, an incidental finding of the
Young-Eisendrath (1981, cited in Young-Eisendrath, 1982) study of empathy and
ego deve10pment was that subjects' preference for the post-conscientious client
analogue and stereotyping of the conformist client analogue signiﬁcantly
affected the subjects' ability to infer the clients' frame of reference. Mullen
and Abeles (1971) investigated the relationship between empathy, liking and
therapy outcome. They found that liking as measured by the Nonpossesive
Warmth Scale was unrelated to outcome. But they argued that the scale used
indicated a general level of liking, instead of a therapist's liking for a speciﬁc
client. It is obvious that measures indicating specific liking would be necessary
in investigating liking or preference as a possibly mediating factor in empathy.

Empathy and conceptual systems theory. The developmental theory

 

articulated by Harvey, Hunt and Schroeder (1961) has spawned a large body of
research studies (Hunt, 1977). Among these have been studies which have
investigated the effects of a counselor's conceptual level upon his/her empathic
behavior and other interview behavior. As there have been few reported studies
on the relationship between ego deve10pment and empathy, a review of this
parallel research from the viewpoint of a related developmental theory may be

useful.

36

Goldberg (1974) used an analogue counseling task with master's level
counseling students and demonstrated a signiﬁcant difference in the subjects'
response patterns as a function of their conceptual level. Focused in more
clearly on expressed empathy, Lutwak and Hennessey (1982) showed that the
empathic responding in interview behavior of undergraduate and graduate
students was signiﬁcantly and positively correlated to the subjects' level of
conceptual functioning.

Rimberlin and Friesin (1980) provided indirect evidence of an association
between ego development and expressed empathy in a study involving conceptual
level. They investigated subjects' empathic responses to ambivalent affect in
analogue research. Ambivalent affect was considered by Loevinger (1976) to be
indicative of post-conformist levels. In Kimberlin and Friesin (1980)
undergraduate "helpers" responded to videotaped analogue clients which
demonstrated ambivalent and non-ambivalent affect. Subjects' responses to the
analogues were rated using Carkhuff's EU scale.~ Low conceptual level subjects
showed a significantly lower level of expressed empathy in response to
ambivalent affect.

Another study addressed the matching and mismatching of the conceptual
level of counselors and client analogues. Heck and Davis (1973) demonstrated
that levels of empathic behavior could be signiﬁcantly affected by the
interaction of counselor and client analogue developmental levels. In this study,
master's level counseling students responded to two analogues consisting of 12
client responses each. One analogue was of high conceptual level and one was
of low conceptual level. The counselors were rated at high or low conceptual
level independent of the experimental task. A two-way ANOVA for repeated
measures yielded a significant main effect for counselor level and a significant

interaction effect across counselors and analogues. The mean empathy ratings

37
revealed the nature of this interaction. High level counselors were rated high in
expressed empathy with high level analogues, but were rated lower with low
level analogues. Low level counselors were rated moderately for expressed
empathy with low level analogues but were rated lower with high level
analogues. This study showed both a main effect and an interaction effect
related to counselor and client analogue developmental levels. As this
deve10pmental construct is related to ego development (Loevinger, 1976, 1979)
the results lead to the hypothesis that both a main effect and an interaction
effect might be expected due to the impact of both counselor and client ego
deve10pmental levels upon expressed empathy.
Summary of‘ELo DevelOpment and Empathy
Ego development has been related theoretically and empirically to

empathy. There have been few studies in this area, however, and none focused
Speciﬁcally on the interaction between ego levels of counselor and client and
the resulting effect upon expressed empathy. Loevinger (1976) and Swensen
(1980) cautioned against expecting a simple relationship between ego level and
behavior. Swensen advocated a research strategy similiar to Cronbach's (1957)
ATI approach.

Hauser (1976) agreed with this strategy.

Rather than predict correlations between ego deve10pment level and

behavior, it may be more fruitful to look for behaviors which are

generated by the interplay between certain situational cues and certain

ego deve10pment levels....Links between ego development level and action

can be effectively studied only by experimental conditions and dependent

variables, which are based on specific predictions derived ﬁ'om

theoretically described characteristics of each stage. (p.939-940)

Systematic variation in the expressed empathy of counselors due to the
interaction between counselor and client level of ego development has been

theoretically predicted. Studies which have investigated the relationships

between ego deve10pment and empathy have not yet taken the form nor focus

38
with which to study this interaction empirically. It is this interaction between
internal frames of reference of the participants in a counseling relationship and
its resultant impact upon the counselor's ability to respond empathically which
was the focus of this investigation. 7
Summary of the Review of Literature

Several areas of theory and research formed the basis for this
investigation. The theory of ego development has described the internal frame
of reference through which an individual perceives and interprets the world.
This theory has been complemented by an assessment technique with sufﬁcient
reliability and validity for research purposes.

Theory and research was also reviewed above dealing with the empathy
cycle. A counselor's behavioral manifestation of empathy, expressed empathy,
occurs within this cycle following the counselor's internal empathic resonation.
The content of the empathic resonation may be affective and/or cognitive in
nature.

Research into the empathy cycle has used a variety of measures. The RES,
one such measure, has been shown to have sufﬁcient reliability and validity for
its use in research. Research into the empathy cycle also has used analogues of
counseling interaction. The use of counseling analogues limit a study's
generalizability, but also allow the investigation of interactions between
variables not possible otherwise.

The theories of ego development and empathy have been related
conceptually because both deal with an individual's internal frame of reference.
This investigation examined the effects of the interaction between counselor
and client levels of ego deve10pment upon a counselor's expressed empathy and
upon a counselor's expressed preference for a client. The client analogues used

in this study allow the investigation of this interaction. The kind of expressed

39
empathy focused upon was verbally expressed, cognitive role-taking empathy.

This investigation's hypotheses had their source in the literature reviewed
above. A positive relationship was hypothesized between a counselor's level of
ego development and a counselor's level of expressed empathy. Theoretically,
this positive relationship was expected because Loevinger's (1976) description of
the stages of ego deve10pment characterized individuals at higher stages as
possessing greater openness, conceptual compleidty and ability to regress in the
service of the ego. Empirically, this expectation was supported by the
investigations of Carlozzi, et a1. (1983) and Heck and Davis (1973). These
studies showed a positive relationship between empathic resonation and ego
development, and between expressed empathy and conceptual level, respectively.

The expressed empathy of a counselor also was hypothesized to vary among
the ego development levels of the client analogues. These analogues were
constructed so as to represent clients demonstrating internal frames of
reference which varied qualitatively. It was expected that these four different
stimuli, the client analogues, would elicit different behavior, expressed empathy,
from the counselors.

The hypothesis that an interaction between counselor and client levels of
ego development would be evident in the levels of expressed empathy was based
on the assumption that a counselor would demonstrate the highest level of
expressed empathy when interacting with a client of equal ego development
level. In part this hypothesis was based upon Heck and Davis' (1973) ﬁnding of a
similiar interaction. between the conceptual levels of counselor and client
analogue. This assumption also resulted from two expectations. A counselor was
expected to manifest lower eXpressed empathy with a client of higher ego
deve10pment level. This was because it would be difﬁcult for an individual to

comprehend the more complex frame of reference used by another person at a

40
higher level of ego deve10pment (Loevinger, 1976, 1979; Redmore, 1976; Blasi,
1976; Young-Eisendrath, 1982).

A counselor was also expected to manifest a lower level of expressed
empathy when interacting with a client at a lower level of ego development
than the counselor's own. This would be perhaps because of the counselor's
asymmetry of preference, that is, a counselor would prefer the highest level of
ego deve10pment which that counselor could comprehend. The counselor's
comprehension would be limited by the level of ego deve10pment which the
counselor had reached. A similiar asymmetry of preference was observed in the
case of moral development (Rest, et a1., 1969), a theory which has been related
both conceptually and empirically to ego development theory (Loevinger, 1976;
1979). Evidence of an asymmetry of preference involving ego development was
observed anecdotally (Young-Eisendrath, 1982) and as an incidental ﬁnding to an
empirical investigation (Young-Eisendrath, 1981, cited in Young-Eisendrath,
1982). For these reasons, a counselor also was expected to indicate a lower
preference for a client whose level of ego deve10pment was lower than or more
than one level higher than the counselor's own level.

In Chapter 3 the method by which these hypotheses were tested is
described. The results of these tests are presented in Chapter 4 and discussed

in C hapter 5.

Chapter 3: Design of the Study

This study investigated the relationship between counselor and client levels
of ego development and the expressed empathy and expressed client preferences
of counselors. The subjects, Master's level counseling students, completed a
measure of ego deve10pment and responded in writing to four typewritten client
analogues. These analogues were previously standardized at four sequential
levels of ego deve10pment. Independent ratings by pairs of trained raters
provided indices of both ego development and expressed empathy. The subjects
also rank ordered their preferences for the client analogues. The eXpressed
empathy data were analyzed with a 3 x 4 repeated measures Analysis of
Variance; the preference data were analyzed with a Chi square.

Sample

Subjects were 42 masters level counseling students at a large midwestern
university. The 31 women and 11 men ranged in age from 22 to 50 years.
Median age was 27 and the mean age was 28.56 years.

Subjects were initially contacted by way of a letter which was distributed
in approximately six graduate counseling classes. The letter requested their
participation in counseling psychology research but did not reveal the nature of
the hypotheses. Subjects indicated interest by returning the letter with their
name and phone number attached. It was apparent that a number of potential
subjects who read this letter declined to participate. Those subjects indicating
an interest were then contacted and they completed the instruments individually
and in small groups as scheduling permitted.

Procedure

Subjects completed the instruments in two sessions, each lasting

approximately 45 minutes. In the ﬁrst session subjects completed the sentence

completion test. In the second session subjects responded to the four client

41

42

analogues and rank ordered their preferences for the clients.

Measurement of Ego Development

 

Washinixon University Sentence Completion Test (WUSCT). Form 11-68

 

Women and Form 11-68 Men of the WUSCT were used to assess ego development
level. This sentence completion test consisted of 36 sentence stems on three
pages which the subjects were asked to complete (see Appendix A).

Scoring of sentence stems. Two raters scored the WUSCT items separately.

 

One rater was a post-internship doctoral student in counseling psychology; the
other was a doctoral level psychologist. Both were eXperienced with projective
tests. Both raters completed Loevinger's self-training program (Loevinger, et
a1., 1970) before beginning the scorings.

WUSCT items were prepared for scoring following Loevinger's (1970)
instructions. All items from the same stem were typed in random order on the
same page or two. Code numbers were removed during scoring to ensure
independent ratings. Scoring manuals were consulted during the scoring process
(Loevinger, et a1., 1970; Redmore, Loevinger & Tamashiro, 1978; Browning,
1978).

Interrater reliability for the Washington University Sentence Completion
E. Reliability for the total protocol ratings (TP R) obtained from the item
ratings of each rater was calculated using intraclass correlations ggg) (Schrout
& Fleiss, 1979). Although Loevinger, et a1. (1970) used product moment
correlations for this index, Holt (1980) argued that the most appropriate method
for determining interrater reliability for an ordinal scale such as Loevinger's
' was with an _I_C__C_.

Loevinger, et a1. (1970) reported a median interrater reliability for TP R's
of}: = .86. Interrater reliability for the TPR's in the present study was& =

.773.

43

Total protocol rating. Following standard procedure, raters reached
agreement on discrepant item scores before ﬁnal TP R's for each subject were
determined. Loevinger, et a1.'s (1970) automatic ogive rules were used to
determine the TPR from a frequency distribution of item scores for each
subject's WUSC T. The distribution of TP R's in the present study may be found

in Table 3.1.

TABLE 3.1

Distribution of Ego Development Levels of Subjects

 

E go Development Level

I - 3/4 . _I__-_4 I - 4/5
Subjects
W omen 9 12 10
Men . 5 ' 2 4
Total 14 14 14

 

Client Analogues

 

Development. Four client analogues (see Appendix B) were developed and

 

revised in a pilot study following Loevinger's (1976) descriptions of ego
development levels I - 3, I - 3/4, I - 4, and I - 4/5. Each analogue began as an
intake report similiar in form to that used in some university counseling centers.
Following lengthy quotations from the analogue client was a space for the ﬁrst
counselor response. Following this ﬁrst response space were two more client
statements and counselor response spaces. This second and third client
statement were randomly sorted as a pair across the analogues which the

subjects responded to. In this way the description and initial statements of the

44

client were the independent variable, and the last statements themselves were
not.

The analogues were written to be characteristic of the impulse control,
character development, interpersonal style, conscious preoccupations and
cognitive style of a woman at a particular level of ego development. There
were, however, several common elements to all the analogues. Each client was a
woman university student who discussed her interpersonal presenting problem,
one other relationship, and her views about self, school and career. The
analogues were approximately equal in length.

Ratings of analpgue developmental levels. Client analogues were read and
rated for ego development level prior to use by two raters other than the
author. One rater was a counseling psychology doctoral student; the other was a
doctoral level psychologist. Both raters had been instructed in ego development
stage theory and made their ratings independently and in randomized order.
Ratings were in 1002 agreement with each other and with the intended
deve10pmental levels. Analogues were rated at the I - 3, I - 3/4, I - 4, and I -
4/5 levels of ego development.

Presentation of analcﬁues to subjects. Counselor subjects were given the

 

four analogues together in randomized order and asked to read and respond to
each client in the space provided as if he or she were that woman's counselor.

Measurement of Expressed Empathy

 

Revised Re3ponse EmLathLRatinLScale. The RES (see Appendix C) was

 

used to obtain a measure of expressed empathy for each subject's responses to
each analogue. The six subscales used were: Client Feelings, Perceptual
Inference and Clariﬁcation, Tapic Centrality, Expressiveness, Collaboration, and
Exploration. The Verbal Allowing vs. Crowding and Impact on Exploration

subscales were not used as they necessitated a response to the counselor's

45

response from a 'live' client. Each subscale has a ﬁve'point (0-4) rating scale
with descriptive anchors referring to speciﬁc counselor behaviors.

The mean of the two raters' total scores for the six subscales across the
three counselor responses to each analogue were used in the statistical analysis.
Thus, each subject's score had a possible range of 0 - 72.

Response Empathy Scale ratings. Subjects' responses were put in 16 groups

 

according to the analogue level and the particular pair of client statements at
the end. These responses were then randomly sorted within groups and the order
of the groups randomized for each rater. Raters scored them independently and
blind to the subject's developmental level. The two raters were doctoral
counseling psychology students. Both had previously used a similiar scale (i.e.,
Carkhuff's EU scale) and were trained with the RES until familiar with it.

Response Empatlm Scale interrater reliability. Elliott, Filipovich, et al.

 

(1982) and Elliott, Reimschuessel, et a1. (1982) reported interrater reliabilities
in the RES' validation studies of alphas .91 and alpha = .95 respectively for
the total RES score. In the present investigation interrater reliabilities were
calculated for the RES using a Pearson 5 for the ratings of responses to each
analogue.

Reliabilities were _r_ = .778 (I - 3 analogue), 5 = .832 (I - 3/4 analogue), 5 =
.823 (I - 4 analogue), and 5 = .875 (I - 4/5 analogue). The mean reliability was E
= .827.

The study's design was comparable to that'of an aptitude x treatment
interaction design (Cronbach, 1957). Both subject level and analogue level of
deve10pment were considered ﬁxed independent variables. The expressed
empathy score from the RES was the dependent variable.

The effect of subject level (at three levels) and analogue level (at four

46

levels) upon the outcome behavior of expressed empathy was examined. Subjects
at all levels reaponded to client analogues at all levels.

Testable Hypotheses

 

Hypothesis 1. Null hypothesis: no difference will be found in expressed

 

empathy of subject responses as measured by RES total empathy rating between
the subject groups at levels I - 3/4, I - 4, and I - 4/5.

Alternate hypothesis: The I - 4/5 group mean RES total empathy score will
be greater than the I - 4 group mean; the I - 4 group mean will be greater than
the I - 3/4 group mean. This was a directional hypothesis.

Hypothesis 2. Null hypothesis: No difference will be found in the RES total
empathy scores between subjects' responses across the four analogues.

Alternate hypothesis: There will be differences in the RES total empathy
scores for subjects' responses among the four analogues. This was a
non-directional hypothesis.

Hypothesis 3. Null hypothesis: The ego development level of the subjects
will not have an interactive effect with the ego development level of the
analogues upon the RES total empathy scores of the subjects' responses.

Alternate hypothesis: The ego development level of the subjects will have
an interactive effect with the ego deve10pment levels of the analogues upon the
RES total empathy scores of the subjects' responses.

Hypothesis 4. The subjects will not demonstrate a high or low preference

 

for any analogue relative to the other analogues as demonstrated by the
subjects' rank ordering of the analogues by preference.

Alternative hypothesis: The subjects will rank higher those analogues at
the subjects' own deve10pmental level or a level one step advanced compared to

the other analogues.

47
Statistical Analyses

 

Analysis of variance (ANOVA). A 3 x 4 ANOVA for repeated measures

 

(Winer, 1962) was computed to test Hypotheses 1, 2, and 3. The F test for
Factor A was the test for Hypothesis 1; the _I: test for Factor B was the test
for Hypothesis 2; and, the _F_ test for Factor AB (interaction) was the test for
Hypothesis 3. A Neuman-Keuls post hoc analysis was performed in order to
ascertain the source of the variance within factors (Myers, 1979).

The strategy of using an ANOVA and focusing upon the signiﬁcance of the
interaction between client analogue and counselor levels of ego development
was selected to facilitate comparability with the one study most similiar to the
present study (Heck 8 Davis, 1973).

A multiple regression approach could have been adOpted for this analysis
with total mathematical equivalence (Cohen 8 Cohen, 1975; Ward 8 Jenkins,
1973). If the ego development levels were of a ratio or interval quality and if
the process of forming the levels of A and B in the ANOVA resulted in the
collapsing of distinguishable classes into a class equatable to a level, then the
linear model of multiple regression would have led to increased power and
therefore greater parsimony.

The theory and measurement technique of ego development, however,
results in ordinal values at best and arguments that the scale is nominal are
tenable (Allen 8 Yen, 1979). In addition the present study does not involve the
loss of variability through the collapsing of data into classes. Therefore the
ANOVA and multiple regression approaches are equivalent from a mathematical
perspective, but the ANOVA approach has greater conceptual clarity given the
structure of the data.

Analysis of preference data. A Chi sguare was computed to test Hypothesis

 

4. An additional Chi sguare was computed to examine other overall trends in

48
the preference data.

Summary of Method

 

The hypothesis of chief interest was Hypothesis 3, regarding the
interaction effect upon counselor behavior of a counselor and client
characteristic. In order to allow a test of this hypothesis, an analogue method
of research was selected. This method was expected and had previously been
shown to allow for the test of similiar interactions.

The client analogues were developed and standardized at a range of ego
development levels which would commonly be found among late adolescents and
adults seeking counseling. The counselors chosen as subjects would likewise
often see clients at these levels in practice. The form of the analogues was
meant to be comparable to an initial counseling contact. It is thought that in
initial contact the dependent variable, expressed empathy, isian important
ingredient in successful counseling for social/personal issues similiar to the
analogues' presenting problems.

Measures were used in a reliable fashion which have demonstrated
reliability and validity. Statistical analyses were selected which provided
powerful and conceptually meaningful tests of the hypotheses. The overall
design of the study was one which allowed a test of the hypotheses with very

high internal validity .

Chapter 4: Results

The results of the statistical analyses which allowed tests of this study's
hypotheses are presented in this chapter. Initially, the values obtained from the
3 x 4 Analysis of Variance (AN OVA) are examined. The factors for main effects
and interaction of the ANOVA are then discussed along with apprOpriate post
hoc tests. The outcomes of the hypotheses dealing with expressed empathy are
also decided here on the basis of the ANOVA. Finally, the results of the gh_1
Sguare are presented, along with a post hoc test, and the test of the fourth

hypothesis decided.

Expressed E m pathy

 

A 3 x 4 ﬁxed effects Analysis of Variance (AN OVA) for repeated measures

tested Hypotheses 1', 2, and 3. The AN OVA summary can be seen in Table 4.1.

Table 4.1

ANOVA Summary

 

 

 

df MS F
M
A (subjects) 2 124.79 1.95, ns
S/A 39 63.87
B (analogues) 3 68.05 3.61, p < .05
AB (interaction) 6 50.57 2.69, p < .05
SB/A 117 18.84

 

Test for Hypothesis 1. The F test for Factor A (subjects) did not reach

 

signiﬁcance at the p < .05 level. Therefore the alternative hypothesis was
rejected in favor of the null hypothesis. The Response Empathy Scale (RES)

49

50
total empathy scores did not differ signiﬁcantly between subject groups at the I

- 3/4, I - 4, and I - 4/5 levels of ego development. The actual differences

between the means were in the expected order, however, as seen in Figure 4.1.

 

RES Scorei
44
43.5
43 43.05
42.5
42 42.08/
41.5
41
40.5
40 40.12
39.5

 

 

 

Subject Ego
Develoiment I - 3/4 I - 4 I - 4/5

Level Note: Non-signiﬁcant differences in means (p > .05)

 

 

Figure 4.1

Mean RES Total Empathy Scores by Subject Level

 

Test for Hypothesis 2. The F test for Factor B, E = 3.61, was signiﬁcant
at the p < .05 level. Therefore, the null hypothesis was rejected and the
alternative hypothesis was accepted. The RES total empathy scores for the
subjects did differ across the analogue levels I - 3, I - 3/4, I - 4, and I - 4/5.

Test for Hypothesis 3. The _F_ test for Factor AB (interaction), 2: = 2.69,

 

was signiﬁcant at the p < .05 level. Therefore the null hypothesis was rejected
and the alternative hypothesis was accepted. The ego development level of
subjects interacted with the ego deve10pment level of analogues.

An examination of the cell means (Table 4.2, Figures 4.2, 4.3) and the

results of the post hoc analysis shows the nature of the above results.

51
Table 4.2

Mean RES Scores For Subjects By Subject Level and Analogue Level

 

 

Ego DeveIOpment Level of Analogue

 

 

 

 

 

 

 

 

 

 

 

1-3 1-3/4 1-4 1-4/5

Supject Level
I - 4/5 40.21 42.57 44.95 44.47
I - 4 42.30 41.22 44.18 40.63
I - 3/4 38.61 42.24 41.00 38.63
All Subjects 40.37 42.01 43.38 41.24
RES Scorg

44

43.5 4 . 8

43

42.5

42 4 01

41.5

41 41.24

40.5

40 40.37

39.5
Ego
Development I - 3 I - 3/4 I - 4 I - 4/5

 

 

Level of Client Analogue
Note: Differences in means were signiﬁcant (p < .05)

 

Figure 4.2

Mean RES Score Across Subjects By Analogue Level

52

Post hoc analysis of Factor B (analogues). A visual examination of the data
in Figure 4.2 indicates the sources of the statistically signiﬁcant main effect B
among analogue levels. The mean RES total empathy scores are most different
betwen analogues I - 3 and I - 4. The means for analogues I - 3/4 and I - 4/5
are most similiar and fall between the means of the other two analogues.

A Neuman—Keuls post-hoc analysis for Factor B showed that the mean RES
total empathy scores for analogues I - 3 and I - 4 are signiﬁcantly different
from each other and thus contributed the most to this main effect. No other
signiﬁcant differences were revealed by this post hoc test.

Post hoc analysis for Factor AB (interaction). A visual examination of the

 

mean RES scores in Figure 4.3 and Table 4.2 reveals the nature of the
signiﬁcant interaction. For all subject groups the highest mean RES scores were
in response to the analogue at the developmental level correSponding to that
subject group's level.

A post hoc test of simple main effects of the interaction provides a
further breakdown of this interaction. First the results were examined by
subject group. The only subject group showing signiﬁcant differences between
mean RES scores among analogues was the I - 3/4 group. This group had a
significantly lower score with analogues I - 3 and I - 4/ 5 than with analogues I
- 3/4 and I - 4. One other group approached signiﬁcance with this test. This
was the I -'4/5 subject group which had a lower mean RES total empathy score
with the I - 3 analogue than with the three other analogues. These two
differences can be seen as the chief contributors to the signiﬁcant interaction
for ANOVA Factor AB although all differences contributed to the overall
interaction.

A test of simple main effects was also performed by analogue. The only
significant difference with this test was for analogue 4/ 5. With this analogue

subjects at level I - 3/4 had significantly lower RES total empathy scores than

53
subjects at level I - 4/5. It is only with analogue I - 4/5 that a main effect

similiar to that hypothesized for Factor A is apparent at a statistically
signiﬁcant level. The nature of the interaction prevents this effect from

attaining signiﬁcance overall in Factor A of the ANOVA.

 

RES Score
45

44.5
44
43.5
43
42.5
42
41.5
41

40.5

 

4/5 .‘
4o

39.5
39
38.5 3/4 374

38

 

 

Ego Development
Level of I — 3 I - 3/4 I - 4 I-- 4/5
Analogue

 

Note: Above interaction was signiﬁcant (p < .05)

 

Figure 4.3

Mean RES Total Empathy Scores by Subject Level

 

54
Subjects' Preferences for Client Analogues

 

Test of Hypothesis 4. A Chi sguare (Table 4.3) was computed to examine
the differences in preference of the counselor subjects for the client analogues
at the correSponding ego level and for the level one step advanced compared to
the other client analogues. The result was Elli. = 4.457 (if = 1; p < .05).
Therefore the null hypothesis was rejected in favor of the alternative
hypothesis. Subjects indicated a higher preference for those analogue clients at
the level correSponding to the subjects own or at a level one step advanced
compared to other levels. Due to ommissions by three subjects an 3 I 39 was

used.

 

 

Table 4.3

Chi Square
0 E. O E
39 32.5 26 32.5
39 45.5 52 45.5

 

 

 

 

 

 

Chi Sguare = 4.457, (d_f =- 1; p < .05)

 

Post hoc analysis of the preference data. A second Chi sguare revealed

 

the overall trend in this data. This was a 2 x 4 Chi Sguare (Table 4.4). The
result was EL].- = 34.77 (d_f = 3; p < .001). This highly signiﬁcant result showed a
signiﬁcant difference in preference overall. The main contributor, as determined
by the cell contributions (Table 4.5), to this difference was that all subject
groups ranked client analogue I - 3 signiﬁcantly lower than the other client
analogues. The conformist client analogue was significantly less preferred than

the post-conformist client analogues by the post-conformist counselor subjects.

55
Table 4.4

Post Hoc Chi Square

 

 

O E O E

4 19.5 35 19.5
28 19.5 11 19.5
24 19.5 15 19.5
22 19.5 17 19.5

 

 

 

 

 

 

Chi Sguare = 34.769 (d_f = 3; p < .05)

 

Expressed Empathy by Sex of Subjects

Malesubjects in the present study had a mean RES total empathy score
across analogues of 41.15 (p = 11). Female subjects had a mean RES total
empathy score across analogues of 41.97 (p = 31). Although this study's design
did not permit a test of signiﬁcance of this difference post hoc, the difference
in means appeared small considering the magnitude of the scores and the size of

other differences obtained.

Table 4.5

Cell Contributions To Post Hoc Chi Scmare

 

 

I - 3 12.32 12.32
Analogue I - 3/4 3.7 3.7
Level I - 4 1.04 1.04

 

 

 

 

 

Chapter Five: Discussion

This chapter is concerned with the interpretation of the results of this
study. Because of the nature of the design, an analogue counseling situation,
the question of generalizability will be addressed ﬁrst to prepare for a focus
upon speciﬁc ﬁndings as viewed in the light of the generalizability issue.

The ﬁndings will be reviewed one by one: the absence of a main effect for
subject level; the main effect for analogue level; the interaction; and, the
trends in subject preference. Reasons for these ﬁndings and implications
Speciﬁc to individual ﬁndings will also be explored. Finally, the overall
implications of this study for practice, training and research will be discussed.

The Parameters of Generalizability

 

The range of ego development levels. The subjects, Masters level

 

counseling students were assessed to be at levels I - 3/4, I - 4 and I - 4/5 of
ego develOpment. As these were the subject levels sought and expected, these
levels are considered ﬁxed by the design and therefore, not random effects. It
is possible, although less likely, to ﬁnd a counselor whose level falls outside of
this range (e.g. Delta or I - 5). This study does not address the behavior of
counselors outside of the ﬁxed range chosen and it is a limitation to the study's
generalizability. However, the ﬁxed range chosen is also a strength, as a large
percentage of counselors in training and in practice may be found at levels I -
3/4 through I - 4/5. Holt (1980) showed 70% of young adults with some college
eXperience at one of these three levels. An additional limitation to the study's
generalizability is the experience level of the subjects.

The ego development levels of the analogue clients were ﬁxed at I - 3, I -
3/4, I - 4 or I - 4/5 as it would be expected that a large percentage of persons
seeking counseling would be at one of these levels. Holt (1980) showed
approximately 852 of young adults with some college experience at one of these

56

57
four levels. This limits the study's generalizability to these speciﬁc levels of

clients.

The measures used to assess ego development and expressed empathy, the
WUSCT and the RES, are measures which have been used with actual counseling
clients. It is expected that this will be an advantage in comparing the present

results with future research in a more naturalistic setting.

 

Generalizability and the analgpe; This study has utilized analogue clients
in the form of typewritten descriptions and quotations instead of taped,
role-played, or actual clients. The major advantage of the analogue was that it
allowed experimental control not usually found in a naturalistic setting. This
strength was utilized in order to investigate an interaction requiring repeated
matches and mismatches of counselor and client characteristics.

The use of the analogue method also poses the major restraint upon the
generalizability of this study's results. In effect, the situation is one in which
very high internal validity was obtained at the expense of lowered external
validity. Both Strong (1971) and Munley (1974) point to this as the major
disadvantage Of the analogue method.

In terms of Strong's (1971) boundary conditions the analogue used involved
a constricted conversation of short duration in which the participants were of
unequal status which differs signiﬁcantly from actual counseling practice.
Counseling practice involves unconstricted conversation where both participants
are live and free to interact, and both the contact and the status differences
would take place over a more extended and Open ended period of time.

The analogue in this study does not meet Strong's (1971) boundary
conditions and thus is not directly generalizable to a counseling situation. There
is evidence, reviewed earlier, that a written analogue is comparable to a taped

analogue. However, extensions of this study into a practice setting will be

58
necessary to establish fully the application of its ﬁndings.

The nature of the analogue task. There are additional limitations inherent

 

in the particular analogue used. This analogue correSponds but is not equivalent
to an initial counseling contact with a female client who presents a
personal/social problem. It does not correSpond to later phases Of counseling or
to counseling for other presenting problems. It is not expected that the results
would have been different for a male analogue client. However, the sex of the
analogue was ﬁxed and thus the results do not generalize specifically to male
clients. Extensions of this study should focus upon initial contacts with persons
presenting personal/social presenting problems to maintain comparability.

Review of Results and Discussion

 

The Main Effect for Subject Level

The ﬁnding. The hypothesis that the expressed empathy of the subjects
would relate positively and signiﬁcantly with the their level of ego deve10pment
was not supported. The group means of the RES scores were in the expected
order but this did not reach significance. There are two possible explanations
for this lack of a significant effect. It may be that either the study's conditions
were not powerful enough to elicit this effect or that it is unrealistic to eXpect
a simple positive relationship to exist across experimental conditions.

Discussion and implications. Support for the lack of power as the cause of

 

the lack of signiﬁcance is found in the actual group means assuming the
expected order. That is, subjects at I - 4/5 had higher RES scores than did
subjects at I- 4; subjects' at I - 4 had higher RES scores that subjects at I -
3/4. One might argue that simply increasing the p of the sample would elicit
this effect. However, the main effect for subjects did not reach signiﬁcance at
the p < .05 level and the means may logically be explained as a chance

occurrence. Furthermore, the main effect for analogues and the interaction did

59

reach signiﬁcance with the same p, and SO the explanation concerning lack of
statistical power is weakened.

A different reason for the absence of a signiﬁcant main effect for
subjects may provide a better explanation. Loevinger (1976), Hauser (1976) and
Swensen (1980) argue that it is not expected that a construct as complex as ego
development would covary positively with any set of behaviors. Instead,
curvilinear relationships and interactions are expected. Using this reasoning, the
absence of the main effect for subjects is expected. This reasoning is reinforced
by the ﬁndings regarding the interaction. First, a more complex relationship, an
interaction, was found instead of the simpler main effect. Secondly, a post hoc
analysis found that when the analogue level was held constant at I - 4, I - 4/ 5
subjects had signiﬁcantly higher RES scores than I - 3/4 subjects, and I - 4
subjects' scores appeared in an intermediate position. This demonstrates a main
effect for subjects evident only when the variable which the subject
encountered was Speciﬁed and held constant.

The four RES score means across analogues for each subject group assumed
curvilinear patterns. When these three curvilinear relationships were analyzed
together, the main effect was eliminated whereas the interaction was evident.
This result may eXplain why Zielinski (1973, cited in Loevinger, 1979) and
Carlozzi, et a1. (1983) found positive relationships between WUSCT levels and
measures of empathy. In both studies the ego deve10pment level of the stimulus
materials was not established or systematically varied. One may presume that
the ego deve10pment level of those materials was constant if it was evident at
all. In the present study it is apparent that the introduction and systematic
variation of an important client characteristic, ego development, resulted in the
absence of a' main effect for subjects.

The cautions of Loevinger (1976), Hauser (1976) and Swensen (1980) are

60

supported. It is advisable that future research investigating the relationship
between ego deve10pment and the empathy cycle either a) focus upon
interactions or curvilinear relationships, or b) hold constant and report the ego
development level of the experimental condition when investigating a main
effect.
The Main Effect for Analpgue Level

The ﬁnding. The subjects' RES scores did vary signiﬁcantly among the
analogue levels. The post hoc analysis showed that the chief contributor to this
effect was that RES scores in response to the I - 4 analogue were signiﬁcantly
higher than the RES scores in reSponse to the I - 3 analogue. The RES scores
for both the I - 3/4 and the I - 4/5 analogues were in an intermediate position.
This ﬁnding shows both that the analogue levels were a signiﬁcant variable, and
that the post-conformist subjects had signiﬁcantly higher RES scores with the
post-conformist conscientious client analogue than with the conformist client
analogue.

Discussion and implications. This ﬁnding is a validation of the use of the

 

client analogues as the main effect shows that signiﬁcant variance resulted
ﬁom their use. Because the length of the analogues, the sex, and the topics
discussed were held constant, and the second and third statements which the
subjects responded to were randomly sorted across analogues, it is reasonable to
conclude that it was the ego development levels of these analogues which
created this variance.

This ﬁnding may also be taken as an indication that clients in an initial
counseling contact may be afforded varying amounts of expressed empathy by
counselors in relation to the ego deve10pment level of the clients. This may be
eSpecially true in the case of conformist clients being shown less expressed

empathy.

61

This relation between client ego development level and counselor expressed
empathy will need to be replicated in a naturalistic setting so as to ascertain if
it exists in such settings. To determine this, further research in a counseling
setting could be undertaken in which the relationship between the ego
development levels of clients and the expressed empathy of counselors in initial
contacts is investigated. Research might be unobtrusively conducted at agencies
where tape recording of sessions is often used and where written information
from clients is routinely obtained.

The Interaction Between Subject and Analogue Levels of ELO DeveIOpment

The ﬁnding. The signiﬁcant interaction factor in the ANOVA showed that
the RES scores were determined by both subject and analogue levels Of ego
deve10pment. The post hoc analysis revealed that two additional statements may
be made. A) I- 3/4 subjects had significantly lower RES scores with analogues
lower or more than one step higher than I - 3/4 (I - 3 and I - 4/5 analogues)
when compared to their RES scores with analogues at or one Step higher that I
- 3/4 (I - 3/4 and I - 4). B) Subjects at I - 4/5 had signiﬁcantly higher RES
scores with analogue I - 4/5 than did I - 3/4 subjects. I - 4 subjects' RES scores
were at an intermediate position. This result was discussed earlier in relation to
the main effect for subjects.

Discussion and implications. This ﬁnding conﬁrms the hypothesis of chief

 

interest in this study and as such this study can be considered a success. Each
group of subjects showed the greatest expressed empathy as measured by RES
scores with analogue clients of equal ego development level. The level Of ,
eXpressed empathy for every subject group declined with analogue clients higher
or lower than the subjects' own ego development level.

This ﬁnding suggests that counselors will demonstrate considerable less

empathy when dealing with clients developmentally different from themselves.

62

An examination of the means in Figure 4.3 also suggests that this effect of
decreased expressed empathy may be more pronounced when the counselor is
dealing with a client whose ego development is more than one level different
from the counselor's own, at least for counselors at I - 3/4 and I - 4/5.

Research conﬁrming this interaction in a naturalistic setting may be
difﬁcult because of the number of matches and mismatches of levels required.
However, the results from this study indicate that this sort of research may be
worthwhile. An alternate to a design including the range of levels examined
here, may be a study to investigate and compare speciﬁc matches and
mismatches rather than the entire range possible. For instance, the post hoc
analysis of the interaction Showed that I - 3/4 subjects had signiﬁcantly higher
expressed empathy with analogues at the same developmental level and one
level higher than with other levels.
Trends in Counselor Preferences

The ﬁndings. The fourth hypothesis was conﬁrmed as subjects indicated a
higher preference for client analogues at their own level or at one level higher
compared to client analogues at a lower level or more than one level higher
than the counselor's own. An additional analysis revealed that the overall trend
in the subjects' preferences was that the conformist client analogue was less
preferred by all the subject groups. All subjects were post-conformist. The
signiﬁcance of this second ﬁnding (p < .001) when compared to that of the ﬁrst
analysis (p < .05) suggests that this incidental result may overshadow the formal
test of hypothesis four in importance.

Discussion and implications. There are three studies which Speciﬁcally

 

. relate to the present study's investigation of preference. Rest, et a1. (1969)
found that subjects preferred the moral reasoning at the highest level which

they could comprehend. This level was often one level higher that the subject's

63

own. This is the primary source for the exact nature of hypothesis four.
Therefore, this study has conﬁrmed this asymmetry of preference using ego
development theory.

Mullen and Abeles (1971) investigated the relationship between empathy,
liking and counseling outcome. Whereas they found no relationship between
liking and outcome, it was concluded that this might have been because the
measure of liking, the Nonpossesive Warmth Scale, tended to measure an overall
warmth of the therapist rather than a therapist's particular liking for a speciﬁc
client. The present study obtained a measure of the subjects' Speciﬁc
preference for speciﬁc client analogues which may have aided in obtaining a
Signiﬁcant ﬁnding. An additional difference between the two studies is that the
present one does not relate preference to outcome.

Young-Eisendrath (1981, cited in Young-Eisendrath, 1982) obtained the
result most similiar to this ﬁnding regarding preference. Young-Eisendrath
(1981, cited in Young-Eisendrath, 1982) reported an incidental ﬁnding that the
subjects, also Masters level counseling students, preferred the
post-conscientious client analogue to the conformist client analogue.
Young-Eisendrath reported that this difference in preference affected the
subjects' ability to infer the client analogue's frame of reference.

The present study conﬁrms Young-Eisendrath's (1981, cited in
Young-Eisendrath, 1982) ﬁnding. The subjects, post-conformist Masters level
counseling students, demonstrated a signiﬁcantly lower preference for the
conformist client analogue relative to the post-conformist client analogues. In
addition, this study showed that subject preference was a function of both
subject and analogue levels Of ego development.

These ﬁndings establish a relationship between ego deve10pment and

counselor preferences for clients based upon the counselor's initial knowledge of

64

the client. These ﬁndings are more directly applicable to counseling practice
than the ﬁndings involving expressed empathy. This generalizability is based
upon the nature of the task by which the subjects' preferences were
established. Subjects read descriptions of four clients and then rank ordered
their preferences for these clients. The subjects were asked to rank their
preferences as if they were counselors with schedules which did not allow them
to see all the clients. This task directly resembles client selection in many
organizations in which counselors are allowed to choose their caseload on the
basis Of the contents of an intake report.

Therefore, these ﬁndings regarding counselor preferences and its
generalizability to client selection raises an issue previously unestablished. If
post-conformist counselor subjects in this study preferred post-conformist
clients signiﬁcantly more than the conformist client, it is reasonable to expect
that this difference in preference leading to differential selection of clients
would occur in an organization with a similiar client selection procedure, with a
similiar mix of client developmental levels, and with counselors of similiar
experience and ego deve10pment level. Investigations conﬁrming this effect in
an organization providing counseling could be done using counselor WUSCT
ratings, client selection data normally kept by the organization and independent
ratings of intake reports for client ego deve10pment level.

Recommendations

 

The application of aptitude x treatment interaction (ATI) ﬁndingp. This

 

study most resembles an ATI approach in its research design. The subjects'
WUSCT levels are the aptitude, the four analogues are the treatments, and the
dependent variable, RES scores, provides evidence of interaction. Therefore, the
views of Cronbach (1957, 1975) are helpful to consider before seeking general

implications and recommendations of this Study.

65
Cronbach (1975) stated that ATI ﬁndings have two general applications.

The ﬁrst is the use of the knowledge of the empirical relationships in new
situations. Cronbach (1957) stated that ATI results can be used empirically in
order to achieve maximally beneﬁcial matchups between persons and treatments.
Cronbach (1975) moderated this recommendation. He pointed out that an
empirical relationship established in one situation between an aptitude and
treatment must be re-established in new situations before being used reliably
for matching. ATI results are not automatically generalizable.

There is a second use for ATI results. "The other reasonable aspiration is
to develop explanatory concepts, concepts that will help people use their heads"

(Cronbach, 1975, p.126).

Recommendations for research. (A) Research is needed to conﬁrm the

 

existence of this study's ﬁndings in a counseling practice setting. This is
necessary because of the artiﬁcial nature of the analogue task. (B) Research is
needed to extend the present ﬁndings to a greater range of client ego
development levels, counselor ego deve10pment levels and counselor experience.
(C) Research is needed to ascertain the impact of the interaction of counselor
and client ego development levels upon client perceived empathy and upon
counseling outcome. (D) Research investigating components of the empathy cycle
Should hold constant and report the ego development levels of counselor and
clients or include these levels as a variable.

Recommendations for practicye, traininj and supervision. If the
generalizability of the present ﬁndings were established, the use of this
empirical relationship fOr optimal matching of counselors and clients might be
undertaken. In each setting, however, evidence would be needed as to the
presence of the empirical relationship. The waiting lists in organizations

providing counseling services might be examined by their Staffs to see if a

66

disproportinately high number of conformist clients were waiting for services
due to differential selection.

Counselor training programs using empathy training might systematically
include and vary the ego development levels of the materials used in this
training. Counseling students could be made aware of ego development theory
and its possible impact upon counseling process in a systematic fashion. This
might help the counseling students to become more aware of ego deve10pment
levels as a source of counter-transference and the misunderstanding of clients
when and if this occurs. Young-Eisendrath (1982) has begun such a program
within a social work school. Ego deve10pment levels Of client and counselor
might be a fruitful issue to consider in the supervision of counselors.

Implications of the interaction between counselor and client levels which
this study uncovered can be usefully looked at from two additional viewpoints.
These are the viewpoints Of high-level and low-level counselors.

Counselor training including the ItOpic of ego development could help to
raise the consciousness of high-level counselors to the possibility of an
important counter-transference issue. The present study showed that counselors
demonstrated lowered expressed empathy and lowered preference toward clients
whose level was below the counselors'. In practice, a Conscientious or
Individualistic level counselor dealing with a Conformist level client might
dislike this person's viewpoint and be less empathic and less helpful as a result.
This same counselor, dealing with a Self-Protective level client might resist
aiding this person to attain a Conformist level of deve10pment (the next stage)
even when this might be an apprOpriate treatment goal.

A counselor who had previously been alerted to this possibility and to his
or her personal reaction to low level clients might consider a) referral, b)

supervision, c) practice in viewing the client's world through this more primitive

67

viewpoint in order to understand better the client's outlook.

The other side of this issue is the case of a low level counselor. A
counselor at the Conscientious/Conformist Transitional level may have
Signiﬁcant difﬁculty comprehending the frame of reference of a high level
client. Thherefore, referral, supervision and participation in life experiences
designed to enhance the counselor's personal deve10pment are options here. This
same counselor, however, would not experience this difﬁculty comprehending
clients at the same level or below, and may have a higher preference for these
clients than a high level counselor. This could be advantageous. The use of
peer-counselors from the same group as a client population (e.g. high school
students) would be a case in which the counseling outcome could possibly be
improved through the use of peer counselors or paraprofessionals. This would be
because peer counselors, for example, may prefer and be very ﬂuent in the
frame of reference used by the bulk of the client population compared to

trained and possibly older and more deve10ped professionals.

APPENDICES

APPENDIX A

Washigton University Sentence Completion Test

 

Sentence Completion For Women (Form 11-68)

Code #-—--— Date
Instructions: Complete the following sentences.
1. Raising a family

2. A girl has a right to

3. When they avoided me

4. If my mother

5. Being with other people

6. The thing I like about myself is

7. My mother and I

8. What gets me into trouble is

9. Education

10. When people are helpless

11. Women are lucky because

12. My father

13. A pregnant woman

14. When my mother spanked me, I

15. A wife Should

16. I feel sorry

17. Rules are

18. When I get mad

19. When a child will not join in group activities
20. Men are lucky because

21. When they talked about sex, I

22. At times she worried about

23. I am

24. A woman feels good when

25. My main problem is

26. My husband and I will

27. The worst thing about being a woman
28. A good mother

29. Sometimes she wished that

30. When I am with a man

31. When she thought of her mother, she
32. If I can't get what I want

33. Usually she felt that sex

34. For a woman a career is

35. My conscience bothers me if

36. A woman should always

68

69

Sentence Completion For Men (Form 11-68)

Code #-—-—- Date
Instructions: Complete the following sentences.

1.
2.
3.

Raising a family
When a child will not join in group activities
When they avoided me

4. A man's job

5.

Being with other peOple

6. The thing I like about myself is
7. If my mother

8.

Crime and delinquency could be halted if

9. When I am with a woman

10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.

Education

When people are helpless

Women are lucky because

What gets me into trouble is

A good father

A man feels good when

A wife Should

I feel sorry

A man should always

Rules are

When they talked about sex, I
Men are lucky because

My father and I

When his wife asked him to help with the housework
Usually he felt that sex

At times he worried about

If I can't get what I want

My main problem is

When I am criticized

Sometimes he wished that

A husband has a right to

When he thought of his mother, he
The worst thing about being a man
If I had more money

I just can't stand people who

My conscience bothers me if

He felt proud that he

70

Examples of WUSCT sentence completions scored at DELTA, I - 3, and I - 4
levels of ego development (Hauser, 1976).

DELTA, Self- Protective Stage:
When peOple are helpless....."I don't like to be bothered with them."
When they talked about sex, L...."listep, 'cause I won't 83': in trouble, also I'll

know what I'll be doing;

 

I - 3, Conformist Stage:

When peOple are helpless....."I try to help them."

 

When they talked about sex, I....."joined riLht in."

 

I - 4, Conscientious Stage
When people are helpless....."I sympathize, unless they are unwﬂﬁnLto help

themselves."

 

When they talked about sex, I....."was interested and respected what they said."

APPENDIX B

Client Analogues

 

The following analogues were developed for the present study by the
author. They were constructed so as to simulate a initial contact with a client
at a varying level of ego deve10pment. Each client analogue discussed, in
typewritten quotations, her presenting problem (a troubling interpersonal
relationship, one other relationship, her attitudes toward school at the
university, her views about career and about herself. All information included
was varied according to the ego deve10pment stage descriptions of Loevinger
(1976). After construction, these analogues were independently rated by two
trained raters, in random order. The ratings were in 100 Z agreement with each
other and with the intended stages. The ego development levels of the following
analogues are:

Ann B. I - 3, Conformist Stage

Maria R. I - 3/4, Conscientious/Conformist Transition
Susan G. I - 4, Conscientious Stage

Julie K. I - 4/ 5, Individualistic Level

The instructions were:
My self selected code number is
Instructions For Part II

I would like you to read descriptions of four counseling clients. After each are
3 client statements followed by spaces marked "Counselor Response." Imagine
that you are the counselor and respond, in writing, in a way you feel would be
helpful. Also, try to respond to each client based upon everything you have
learned about her, and not just on the basis of a single statement.

After the subjects read the instructions and completed the analogues (in
random order), they were asked to indicate their preferences for the client
analogues in the following way:

Assume that you are a counselor with a busy schedule and that you only have
time to see 2313 Of these clients for counseling. Please rank order these clients
based upon whom you would most like to see, not necessarily who needs
counseling the most, but rather whom you'd prefer to counsel with your last
available time.

Please Rank Order
Most Preferred
2nd Most Preferred
3rd Most Preferred
Least Preferred

(Client's names)

 

 

 

Thank you.

71

72
Client: Ann B.

Presenting Problem
"...feeling sad about breaking up with my boyfriend." Since the breakup
Ann has been quite upset and has had low motivation.

Notes:

Recently Ann broke up with her boyfriend. This was a "long story", she
said, which she felt "...really sad" about. Ann Stated that last summer, when he
wasn't working or in school, her boyfriend started wearing his hair differently.
"...It looked so weird, but he didn't seem to care. I mean, no one I've ever
known has had hair. like that—and none of the guys who date my girlfriends,
either. I just didn't feel its right!

Finally her boyfriend broke off the relationship. Ann said she feels lousy
about this and that now on weekends she's the only one of her girlfriends
without a date. She feels ashamed at these times.

A girlﬁend of Ann's had suggested that she speak with a counselor and
that this would make her feel better. She and this woman are quite close. Ann
said "...we're really best friends. I mean, we always see things the same way
and everything. And when school or the classes we're in are confusing we can
talk together and ﬁgure it out." Ann is not quite sure about her career
direction, but she says that she knows she'll ﬁnd the right one with a bit more
time. ‘

Ann also said that she is the kind of person who "...likes to be part of
something, my family, my group of friends, and I like to be friendly and nice
with all my friends. I'm really a social type of person, you know, I always try to
look good whatever I'm doing. That's why its been so hard breaking up with my

boyfriend."

Ann said that our interview was like some of her courses at the university,
where "...the prof's ask different questions and get you to look at different
areas, almost like they're uncertain, until you, I mean we ﬁnd the solution. Its
like the saying my best friend has 'you have to go down a rough road before
you ﬁnd the answer.‘ Do you know what I mean?"

(C ounselor R eSponse)

"The way I'm feeling is really bothering the hell out of me!"

(C ounselor R eSponse)

"Well, that's the situation, what do you think?"

(C ounselor R eSponse)

73
Client: M aria R.

PresentipiProblem

She's been feeling generally uncomfortable and self-conscious around her
ﬁ'iends lately, especially about her opinion relative to theirs. Sometimes this
leads to feelings of panic.

Notes

During our interview Maria seemed always concerned about what she would
say. It seemed that she would become uncertain after she began speaking,
almost embarrassed about what she was saying or the way in which she was
saying it.

Maria's uncomfortable feelings have to do with herself in relation to her
friends, although she's vague about this. Since the time when she entered the
university, she has become aware that she didn't always follow what she used to
think were the "right" rules of conduct in school and around her friends.

For instance "...I was taking a course that was pretty easy but very
boring, and I just stopped doing a lot of the work. I ﬁgured that I'd get by and
I did. It was different, though, not applying myself to school like I used to. And
in other courses, I've started seeing that there's a lot of different ways to look
at things, which is different, too. It's like the way I grew up isn't the only
way."

"When I came here I thought that I'd ﬁnd the right guy and marry, and
ﬁnd a career that ﬁt, but now...I just don't know what to think, there's so
many different ways."

"And then this guy I'm seeing wants me to make up my mind about
marriage and everything, but I just feel confused. I guess it's alright for him
right now, 'cause he's older. He really gets frustrated about this, but I just end
up feeling lonely (she became tearful at this point). I'm still seeing him, even
though I don't intend to marry him...why, even that's a change!" She looked
fearful after she said this.

I've talked about this with my friend, Jill, but we just end up with the
same confusion. It's good to have her to talk with, though. We've always had a
rule: if one Of us needs to talk out something, the other listens, even if she's
involved in something else. I guess that seems like a really feminine thing to
do."

Right now, you could say, I'm the kind of person who's trying to sort out
which way to go and how I feel about it. Does it seem that way to you, too?"
(Counselor Response)

"This feeling I have...I just can't stand it!"
(Counselor ReSponse)

"I keep wondering...what do you think of all this?"
(Counselor ReSponse)

74
Client: Susan G.

Presenting Problem

She's depressed and feels responsible because a close friend had a "nervous
breakdown" recently. Since then she's been eating much less and has had a
great deal of trouble getting to sleep at night.

 

Notes

Susan took an active part in our interview and tried to tell me the things
she felt were important to know about her and her situation. She was quite
tearful throughout our meeting.

She started by telling me about who she was as a person. "I've set up my
own standards in my life. I still try to take others into account, of course, even
if they have a different point of view. But, overall, I use my own perspective.
This helps, like in coursework here at the university. A lot of subjects I study
don't give deﬁnite answers, because that's how the world is. Even then, I've
found school useful. Now I'm able to view situations in a relative sort of way
and then I decide how to act. My boyfriend is a lot like this too—that's one
thing that attracts us to each other—that we share a similiar way of ‘
understanding the world. We both have some long term goals for our careers and
lives and, for now, we ﬁnd we can support each other in our different goals."

"Well, one of the standards I have is to help Others if I'm able. This may
sound crazy, but I sort of decided that I'd always try to do this. It was like
that with my roommate. She was real depressed and having a hard time of it. I
found that I was able to help her see a different perspective about things. That
felt really satisfying to me, real good, being able to help."

"But then things in her life got tougher—its complicated because some
things for her go back to her family and all. She dropped out and I didn't hear
from her. Then last week I found out that she was in the hospital. (Susan
began sobbing at this point and looked away.) She had a nervous breakdown. I
felt horrible! I just keep saying to myself that somehow I could have reached
out after she left school. Maybe that could have changed things. I feel so
responsible. I don't know what to do."

(C ounselor R esponse)

"My feelings about this just keep eating away at me!"

(C ounselor R esponse)

"...uhm, how do you think I should approach this?"

(C ounselor R eSponse)

75
Client: Julie K .

PresentingyProblem

She's been involved in a relationship which is conﬂictual for her. She
decided to seek counseling in order to more fully understand her feelings and
the situation before acting.

 

Notes

Although she was visibly distressed, Julie seemed to be actively struggling
with the issues she spoke about. She also seemed to carefully consider and
respond to the interviewer's comments.

Julie said, "I wish my life were as straightforward as some of my courses
seem. You know, applying different principles to situations depending upon the
circumstances, like in physics. But its not.

Julie sounded like she had deve10ped a number of strong relationships with
those around her. But one was particularly troubling. In the past year Julie's
boyfriend has apparently been experiencing a personal career upheaval.

She said, "I could tell that he was going through some changes because his
appearance had changed. There were some feelings inside of him which
conﬂicted. That was OK in a way, because I've experienced that sort of
conflict, too. So I've tried to understand what he's going through and support
him, but I've only got so much energy. You see, there's some things that I've
committed myself to achieving as an individual. But I'm just not sure that I'll be
able to work toward these goals I've set, and be there for him as much as he
needs..." She seemed to have some really complex feelings about this.

"It's not so much a practical matter, since he supports himself and has his
own apartment, but more an emotional one. That's partly why it's so frustrating
to me. Sure, things might be easier if, say, my academic program at school were
less pressured or something, but this is still a conﬂict I feel inside me. I want a
deep relationship, one in which the two of us can develop a real closeness yet
develOp ourselves as individuals, too.

"But develOping myself as an individual takes a lot of determination and
endurance. I know, because its a commitment I made to myself some time back,
to live life as fully as I can while keeping to my own values. Still, it's really
been feeling like a bind, since I'm not sure if I can manage both areas as one
person. And that hurts, thinking that I might have to give up one. Its a paradox
to me; as Ilearn more about life, it only seems to get more complex. Does all
of this make sense?"

(Counselor ReSponse)

"I hate feeling this way...I hate it!"
(Counselor ReSponse)

"I ﬁgured it would be helpful to ﬁnd out your reaction."
(Counselor Response)

APPENDIX C

Revised Response Empathy Scale

 

Elliott, R., Reimschuessel, C., Filipovich, H., Zapadka, J., Harrigan, L., and
Gaynor, J. (1981). Department of Psychology, University of Toledo, OH
43606

Subscales L 2J 3, 4, 54 7.

 

Score each helper response on each of the following six dimensions. Leave
E blank items.

1. Client Feeling: To what extent does the helper address the client's feelings?

 

4 - Refers to speciﬁc client feelings (e.g., hurt, depressed, angry, anxious).

3 - Reference is to generalized or vague client feelings (e.g., bad, upset,
disturbed, okay) or to motivations (e.g., wants or desires).

2 - Reference is to client thoughts, cognitions or Opinions (e.g., "Do you
think that was the right way to handle it?")

1 - Reference is to client behaviors or actions (e.g., "Why did you do
that?")

0 - Reference is to other people's feelings, cognitions and behaviors (e.g.,
including therapist's, signiﬁcant others', etc.).

Note: For long responses, identify all ratable aSpects of the response and
average; use judgement for rounding up or down.

2. Percgptual Inference 8 Clariﬁcation. Does the helper make inferences to tell
the client something the client hasn't said yet to add to the client's frame of
reference or to bring out other implications?

 

4 - Significant inference or clariﬁcation present (e.g., interpretations).

3 - Moderate inference or clariﬁcation present (e.g., reﬂections of
implications, signiﬁcant summaries).

2 - Slight inference or clariﬁcation present (e.g., paraphrase or nonverbal
reﬂection).

0 - No inference or clariﬁcation present (e.g., echoic reﬂections,
information questions, advisement, guesses about factual Situations, etc.).

3. Centrality of TOpic: Does the helper refer to what is important to the client?
Does the helper's response relate to the helpseeker's basic complaint or
problem? This dimension is to be rated on the basis of content.

 

 

76

77
(3. Centrality of Topic, continued)

4 - Deﬁnite, clear reference to client's feelings and perceptions connected
with the client's major complaints; refers to conﬂicts, central issues, basic
relationships.

3 - Topic moving toward major concern.

2 - TOpic not clearly central; background facts, possibly related to major
concerns.

1 - Points of clariﬁcation of meaning; repairing mishearings,
misunderstandings.

0 - Topic deﬁnitely pp_t central (e.g., "side tracks," talk about the helping
process; third-party talk, general talk, "mhm.")

4. Expressiveness: Does the helper communicate in an appropriately expressive
manner, using language and voice, including:

a) language which evokes imagery in the rater,

b) creative, novel language (words and/or syntax),

c) modiﬁers (adjectives/adverbs) which intensify or upgrade (vs. attenuate)
meaning of message,

d) interjections (!), e.g., Wow! Gee!,

e) inﬂected voice (i.e., with variations in pitch and/or volume).

f) voice which communicates a speciﬁc emotion, e.g., sadness, excitement,
prizing, caring, surprise, anger.

Except - where language or voice distracts or evokes confusion in the rater.

4 - Greatly expressive; multiple or strong cues.

3 - Moderately expressive; singular or less-strong cues.

2 - Ordinary language or voice.

1 - Mildly distracting or confusing; uhuh; echoic reﬂections, repetitions.

0 - Moderately or extremely distracting or confusing; moderate or exu'eme
repetition.

5. Collaboration: Does the helper's manner communicate a sense of working
together in a collaborative process? This item is to be rated on the basis of
intention; do not use impact to rate this.

4 - Clearly communicates a sense of collaboration, mutuality, a "working
together" or "we're in this together attitude"; personalizing reSponseS that
communicate a sense of sharing and give support and strength to the client
(e.g., collaborative responses, completers); possible references to the T-C
relationship.

3 - Perception check in self-disclosure format; personal, peer informal
helping equalizer, friendship (e.g., "reallyl", "Let's talk"); reference to
shared history, themes.

78

2 - Possibly communicates collaboration (e.g., sympathizers - "That's hard",
"mhm", laughter intended to match client, ordinary reﬂections).

0 - Fails to communicate a sense of collaboration (e.g., session ending,
disagreement, unsolicited advice, cutting client off).

7. Exploration: Does the helper actively encourage the client's exploration by
the content of what the helper says? DO not use impact to rate this.

4 - Clear, active push to encourage client to explore (e.g., exploratory
questions, "How do you feel about that?", process advisement in the form
of "tell me" questions, tag questions at the end of signiﬁcant summary
reflections and interpretations.

3 - Exploratory questions at beginning of session to gather information.
2 - Possibly encourages client to explore; passive facilitation of
exploration (e.g., pausing; tentativeness; ordinary content reﬂections; using

self-disclosure or question format or frame for something else; "mhm".)

0 - Discourages exploration (e.g., closed questions or questions seeking
speciﬁc information; "topic jumping;" session ending).

LIST OF REFERENCES

LIST OF REFERENCES

Allen, M. J., 8 Yen, W. M. (1979). Introduction to measurement theory.
Monterey, CA: Brooks/C ole.

Barrett-Lennard, G. T. (1981). The empathy cycle: Reﬁnement of a nuclear
concept. Journal of CounselinyPsychology, 28, 91-100.

 

Bergin, A. E., 8 Soloman, B. (1970). Personality and performance correlates of
empathic understanding in psychotherapy. In T. Tomlinson 8 J. Hart (Eds.),
New directions in client-centered therapy. Boston: Houghton-Mifﬂin.

 

Blasi, A. (1976). Personal responsibility and ego development. In R. deCharms
(Ed.), Enhancinimotivation: Change in the classroom. New York: Irvington.

 

Browning, D. (1978). Preliminary manual for WUSCT items. Unpublished
manuscript, New York University.

Candee, D. (1974). Ego development aspects of new left ideology. Journal of
Personality and Social Psychology, 32, 620-630.

 

Carkhuff, R. R. (1969). Helping and human relations(Vol. 1). New York: Holt,
Rinehart and Winston.

 

Carlozzi, A. E., Gaa, J. P., 8 Liberman, D. B. (1983). Empathy and ego
development. Journal of CounseliniPsycholojgy, 3_0, 113-116.

 

Chinsky, J. M., 8 Rappaport, J. (1970). Brief critique of the meaning and
reliability of "accurate empathy" ratings. Psycholoiical Bulletin, 13,
379-382.

 

Cohen, J., 8 Cohen, P. (1975). Applied multiple regryession/correlation analysis
for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum.

 

 

Cox, L. N. (1974). Prior help, ego development and helping behavior. Child
Development, 42, 594-603.

 

Cronbach, L. J. (1957). The two disciplines of scientific psychology. American
Psychologist, 2, 671-384

Cronbach, L. J. (1975). Beyond the two disciplines of scientiﬁc psychology.
American Psychologist, _3_Q, 116-127.

Elliott, R., Filipovich, H., Harrigan, L., Gaynor, J., Reimschuessel, C., 8
Zapadka, J. (1982). Measuring response empathy: The development of a
multicomponent rating scale. Journal of CounseliniPsychology, 2_9_, 379-387.

 

79

80

Elliott, R., Reimschuessel, C., Filipovich, H., Zapadka, J., Harigan, L., 8
Gaynor, J. (1981). Revised Response Empathy Rating Scale. Unpublished
manuscript, Department of Psychology, University of Toledo, Toledo, OH.

 

Elliott, R., Reimschussel, C., Filipovich, H., Zapadka, J., Harrigan, L., 8
Gaynor, J. (1982). Validation of the Revised response empathy scale:
Comparison with client and counselor perceptions. Unpublished manuscript,
University of Toledo, Toledo, OH.

 

English, R. W., 8 Jelenevsky, S. (1971). Counselor behavior as judged under
audio, visual, and audiovisual communication conditions. Journal of
Counselinj Psycholm, L8, 509-513.

 

Feldstein, J. C., 8 Gladstein, G. A. (1980). A comparison of the construct
validities of four measures of empathy. Measurement and Evaluation in
Guidance, ﬂ, 49-57.

 

Fridman, M. S., 8 Stone, 8. C. (1978). Effect of training, stimulus context, and
mode of stimulus presentation on empathy ratings. Journal of Counseling

Psychology, 22, 131-136.

Gladstein, G. A. (1977). Empathy and counseling outcome: An empirical and
conceptual review. Counseling thologist, _6_, 70-79.

 

Gladstein, G. A. (1983). Understanding empathy: Integrating counseling,
deve10pmental, and social psychology perspectives. Journal of Counselinj
Psychology, ﬂ, 467-482.

Goldberg, A. D. (1974). Conceptual system as predisposition toward therapeutic
communication. Journal of Counseling—Psychology, _2_1, 364-368.

 

Gormally, J., 8 Hill, M. (1974). Guidelines for research on Carkhuff's training
model. Journal of Counseliﬁ Psycholcgy, 2;, 539-547.

Grant, J. D., 8 Grant, M. Q. (1959). A group dynamics approach to the
treatment of non-conformists in the Navy. Annals of the American Academy
of Political and Social Science, 322, 126-135.

 

Greenson, R. R. (1967). The technique and practice of psychoanalLsisWol. 1).
New York: International Universities Press.

Greenson, R. R. (1978). Explorations'in psychoanalysis. New York: International
Universities Press.

 

Gurman, A. S. (1973). Instability of therapeutic conditions in psychotherapy.
Journal of Counseliri Psychology, A), 16-24.

 

Hardin, S. L, 8 Yanico, B. J. (1981). A comparison of modes of presentation in
vicarious participation counseling analogues. Journal of Counseling

Psychology, a, 540-543.

 

Harvey, O. J., Hunt, D., 8 Schroeder, H. (1961). Conceptual systems and
personality organization. New York: Wiley.

 

 

81

Hauser, S. T. (1976). Loevinger's model and measure of ego deve10pment: A
critical review. Psycholcgtical Bulletin, _8_3_, 928-955.

 

Heck, E. J., 8 Davis, C. S. ( 1973). Differential expression of empathy in a
counseling analogue. Journal of Counseling Psycholggy, 20, 101-104.

 

Hicks, S. W., Cormier, W. H., 8 Bounds, W. G. (1974). Readirgstatistics and
research. New York: Harper 8 Row.

 

Holt, R. R. (1980). Loevinger's measure of ego development: Reliability and
national norms for male and female short forms. Journal of Personality and
Social Psychology, 32, 909-920.

 

Hunt, D. E. (1977). Conceptual level theory and research as guides ~to
educational practice. Interchange, 8(4), 78-90.

Kagan, N. (1975). Interpersonal process recall: A method of inﬂuencinLhuman
interaction. Mason, MI: Mason Media.

 

Kagan, N., Krathwohl, J., 8 Associates (1967). Affective Sensitivity Scale.
Mason, MI: Mason Media.

Kimberlin, C. L., 8 Friesen, D. D. (1980). Sex and conceptual level empathic
responses to affect. Counselor education and supervision, l_9_, 252-258.

 

Kohlberg, L. (1969). Stage and sequence: The cognitive-developmental approach
to socialization. In D. A. Goslin (Ed.), Handbook of socialization theory and
research. Chicago: Rand McNally.

 

Kurtz, R. R., 8 Grummon, D. L. (1972). Different approaches to the
measurement of therapist empathy and their relationship to therapy
outcomes. Journal of Consulting and Clinical Psychology, 3_9, 109-115.

 

Loevinger, J. (1966). The meaning and measurement of ego develOpment.
American Psychologist, a, 195-206.

Loevinger, J. (1976). Ego deve10pment: Conceptions and theories. San Francisco:
Jossey-Bass.

 

Loevinger, J. (1979). Construct validity of the sentence completion test of ego
development. Applied Psychological Measurement, 3, 281-311.

 

Loevinger, J. (1980). Some thoughts on ego development and counseling.
Personnel and Guidance Journal, 28, 389-390.

 

Loevinger, J., Wessler, R., 8 Redmore, C. (1970). Measu'pg egp
develgpment(Vols. 1-2). San Francisco: Jossey-Bass.

 

 

Lutwak, N., 8 Hennessey, J. J. (1982). Conceptual systems functioning as a
mediating factor in the development of counseling skills. Journal Of
Counseling Psychology, a, 256-260.

 

Melnick, R. R. (1975). Counseling response as a function of method of problem
presentation and type of problem. Journal of Counseliﬂ Psychology, g2,
108-112.

 

82

Mullen, J., 8 Abeles, N. (1971). Relationship of liking, empathy, and therapist's
‘experience to outcome of therapy. Journal of Counseling Psycholm, l8,
39-43.

Munley, P. H. (1974). A review of counseling analogue research methods. Journal
of CounseliﬂPsychology, 2_1, 320-330.

 

Myers, J. L. (1979). Fundamentals of experimental desigp(3rd ed.). Boston: Allen
8 Bacon.

Redmore, C. (1976). Susceptibility to faking of a sentence completion test of
ego development. Journal of Personality Assessment, 39, 607-616.

Redmore, C., Loevinger, J., 8 Tamashiro, R. T. (1978). Measurig ego
deve10pment: Scoriananual for men and boys. Unpublished manuscript,
Washington University, St. Louis, MO.

Rest, J. R. (1973). The heirarchical, nature of stages of moral development.
Journal of Personality, ﬂ, 86-109.

Rest, J. R. (1980). Development in judgipgymoral issues. Minneapolis: University
of Minnesota Press.

 

 

Rest, J. R., Turiel, E., 8 Kohlberg, L. (1969). Level of moral deve10pment as a
determinant of preference and comprehension of moral judgments made by
others. Journal of Personality, 31, 225-252.

Rogers, C. R. (1957). The necessary and sufficient conditions of therapeutic
personality change. Journal of Consulting Psychology, 2i, 95-103.

Rogers, C. R. (1959). A theory of therapy, personality and interpersonal
relationships as developed in the client-centered framework. In S. Koch
(Ed.), Psycholpgy: A study of a science: formulations of the person and the
social context(Vol.3). New York: McGraw-Hill.

 

Rogers, C. R. (1975). Empathic: An unappreciated way of being. Counseling
Psychologist, 5(2), 2-10.

Schrout, P. E., 8 Fleiss, J. L. (1979). Intraclass correlations: Uses in assessing
rater reliability. Psychological Bulletin, 8_6_, 420-428.

Stokes, J. P., 8 Tait, R. C. (1978). Comparability of written and oral measures
in evaluating counselor training. Counselor Education and Supervision, 11,
222-229.

Strong, S. R. (1971). Experimental laboratory research in counseling. Journal of
Counseling Psychology, l_8_, 106-110.

. Sutton, P. M., 8 Swensen, C. H. (1983). Reliability and concurrent validity of
alternative methods for assessing ego development. Journal of Personality
Assessment, 4_7, 468-476.

Swensen, C. H. (1980). Ego deve10pment and a general model for counseling and
psychotherapy. Personnel and Guidance Journal, 58, 382-388.

 

83

Therrien, M., 8 Fischer, J. (1978). Written indicators of empathy in human-
relations ﬁ'aining: A validational study. Counselor Education and

Supervision, _1_7_, 272-27 7.

 

Truax, C. B. (1966a). Inﬂuence of patient statements on judgements of therapist
statements during psychotherapy. Journal of Counseling Psychology, a,
335-337.

Truax, C. B. (1966b). Reinforcement and nonreinforcement in Rogerian
psychotherapy. Journal of Abnormal Psychology, A, 1-9.

Truax, C. B., 8 Carkhuff, R. R. (1967). Toward effective counseling and
psychotherapy. Chicago: Aldine.

 

Vesprani, G. J. (1969). Personality correlates of accurate empathy in a college
companion program. Journal of Consultipg and Clinical Psychology, a,
722-727.

Ward, J. H., 8 Jennings, E. (1973). Introduction to linear models. Englewood
Cliffs, NJ: Prentice-Hall.

Winer, B. J. (1962). Statistical principles in experimental desigp. New York:
McGraw-Hill.

Young-Eisendrath, P. (1981). Comparison of brief training in ego development
and in empathy skills for beginning psychotherapists. (Doctoral dissertation,
Washington University, 1980). Dissertation Abstracts International, 41B,
3159.

 

Young-Eisendrath, P. (1982). Ego development: Inferring the client's frame of
reference. Social Casework, {33, 323-332.

 

Zielinski, C. E. (1973). Stages of ego development as a correlate of ability in
discrimination and communication of empathic understanding. (Doctoral
dissertation, University of Houston, 1973). Dissertation Abstracts
International, 34A, 1635.

     

ll!

1011111111101