COUNSELING PROCESS AND
WW5 WED TO CLEEMY’S PERECEPTION
Q? SELF AND CWNSEWR

That: fan- ‘rke Dawn a)? pit. D.
MICHIGAN STATE UNIVERSITY
Arthur <3e Balms
1966

LIBRARY

Michigan State

 

llllllllllllllllllilllllllllllllllllillillHllllllllilll’lllll

31293 10464 6777

This is to certify that the

' thesis entitled

COUNSELING PROCESS AND OUTCOME RELATED
TO
CLIENT'S PERCEPTION OF SELF AND COUNSELOR

presented by -

ARTHUR c ,' BALLAS

has been accepted towards fulfillment
of the requirements for

Ph D degree in Education

 

M ,

Date August 24. 1966

I‘

0-169

ABSTRACT

COUNSELING PROCESS AND OUTCOME RELATED
TO CLIENT'S PERCEPTION OF
SELF AND COUNSELOR

by Arthur C. Ballas

Client-counselor relationship and sources of their
mutual influence provide an important area for study in
counseling. This study was concerned with specific aSpects
of the client-counselor relationship and the client's per-
ception of the counselor as a referent. Its purpose was to
investigate client-perceived counselor potency, activity,
evaluation, and similarity of the client's self-percept and
his counselor percept as variables influencing process and
outcome. A form of the Semantic Differential was used to
measure the client's percept of himself and percept of coun-
selor. Judged changes of MMPI responses (taken before and
after counseling) were used as criteria for successful,
partly successful or unsuccessful psychological change.

The hypotheses examined were as follows (referring
to successful cases of counseling, as compared to partly
successful and unsuccessful cases).

1. A client's self-percept will become more like his

counselor-percept.

Arthur C. Ballas

2. A client's percept of counselor potency Will not
change over the term of counseling.
3. A client's self-percept will become more like

the counselor's self-percept.

Twenty-two client-counselor pairs from the Michi-
gan State University Counseling Center were used in the
sample. The profiles of client responses to the MMPI be-
fore and after counseling were used as the criteria for
counseling outcome (i.e., successful, partly successful or
unsuccessful). Client responses to the concepts "me" and
"counselor" on the Semantic Differential for pre, middle,
and terminal phases of counseling were utilized as the
measure of perceived client-counselor similarity, and
client perception of counselor potency. Pre-counseling
Semantic Differential responses to the concept "me" were
also obtained for the counselors.

Analysis of the data involved the use of the
Kruskal-Wallis one-way analysis of variance of the Mann-
Whitney U Test, as tests of significance.

The results did not support the hypotheses re-
lating successful counseling outcome to increased client
perception of similarity to counselor. The evidence
seemed to indicate, rather, that improvement was associ-
ated with increased client perception or differentiation
of self and counselor. This client perception of dis-
similarity was discussed in terms of autonomy as a purpose

and function of counseling.

Arthur C. Ballas

The hypothesis that client perceived counselor
potency remains invariant over the course of successful
counseling was not supported by the results. However, the
relative positions and disparity of client-perceived potency
of self and potency of counselorwere indicated as a possible
relational variable to counseling outcome. That is, the
data suggested that unsuccessful counseling outcome was
characterized by a client's perception of self potency as
being equal to, or greater than, the perceived potency of

the counselor.

COUNSELING PROCESS AND OUTCOME RELATED
TO CLIENT'S PERCEPTION OF
SELF AND COUNSELOR
By
35‘3

(

Arthur Cf Ballas

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Counseling, Personnel Services
and Educational Psychology

1966

Dedicated to my wife,

Birdie

ACKNOWLEDGMENTS

My thanks are many and yet it is most difficult
to state them in a manner that expresses my appreciation
and sense of indebtedness to those who made the completion
of this thesis possible.

Dr. Bulford Stefflre, my advisor and thesis chair-
man.gave unhesitatingly of his valuable time throughout the
writing. His c00peration and relationship toward me was an
eloquent manifestation of what counselors purport to do,
i.e., express an interest and empathic understanding with
those whom they counsel.

To Dr. B. Kell and Dr. F. Blackington, two of my
committee chairmen, I am also thankful for their helpful
comments, encouragement, and their willingness to give of
their time to react most expeditiously to my requests.

Dr. F. L. Erlandson and Dr. S. Plyler most gener-
ously gave of their time and skill to judge the MMPI pro-
files. Without them the study could not have been conducted
as it was designed. Bach gave selflessly of his time.

I am thankful to Dr. J. Powell for editing and
commenting on the writing. His encouragement and comments
were most helpful.

Finally, I wish to offer a very special thanks to
Mr. Larry Thornton. His help on the technical aspects of

the analysis were crucial to the completion of the study.

iii

TABLE OF CONTENTS
Chapter Page
I. INTRODUCTION . . . . . . . . . . . . . . . . . 1

Purpose . . . 1

The Problem . 2

Hypotheses 3

Assumptions 6

Limitations 7

Summary . 8

II. REVIEW OF THE LITERATURE . . . . . . . . . . . 10

The Counselor as a Referrent or Role Model. 10

The Counseling Relationship . . . . . . 13

Relationship and Therapeutic Change . . . . 15

Counselor Variable. Similarity . . . . . . 18

Counselor Variable: Potency . . . . . . . 23

Summary . . . . . . . . . . . . . . . . . . 25

III. METHOD . . . . . . . . . . . . . . . . . . . . 28

The Sample . . . . . . . . . . . . . . . . 28

Procedure . . . . . . . . . . . . . . . . . 30
Instruments: The MMPI and Categorization

of Subjects . . . . . . . . . . . . . . . 32

Instruments: The Semantic Differential

and Manner of Utilization . . . . . . . . 35

IV. RESULTS . . . . . . . . . . . . . . . . . . . 40

Summary . . . . . . . . . . . . . . . . . . 63

V. DISCUSSION . . . . . . . . . . . . . . . . . . 66

Relation of the Results to the Hypotheses . 66

Some Theoretical Implications . . . . . . . 70

Implications for Future Research . . . . . 73

Summary . . . . . . . . . . . . . . . . . . 7S

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . 78

APPENDIX . . . . . . . . . . . . . . . . . . . . . . 82
The Fourteen Scales of the Semantic Differen-

tial Used in the Study . . . . . . . . . . . 82

iv

Table

10.

11.

12.

LIST OF TABLES
Page

Listing of clients and counselors, and the
number and times of administration of the
Semantic Differential for each subject . . . . 29

The ratings of client changes on the MMPI . . . 33

The final grouping and classification of the
MMPI ratings . . . . . . . . . . . . . . . . . 3S

Semantic Differential mean D scores between
clients' concept "Me" and clients' concept
"Counselor" . . . . . . . . . . . . . . . . . 40

Semantic Differential mean raw scores for the
factors potency, activity and evaluation . . . 4S

Kruskal-Wallis one—way analysis of variance
of Semantic Differential potency factor
scores . . . . . . . . . . . . . . . . . . . . 46

Mann-Whitney U Tests of Semantic Differential
potency factor scores . . . . . . . . . . . . SO

Kruskal-Wallis one-way analysis of variance
of Semantic Differential activity factor
scores 0 o 0 O I O O O I O O O I O O O O O O o 52

Mann-Whitney U Tests of Semantic Differential
activity factor scores . . . . . . . . . . . . S4

Kruskal-Wallis one-way analysis of variance
of Semantic Differential evaluation factor
scores 0 0 O 0 O O O O O 0 O O 0 0 O O O O O 0 S7

Mann-Whitney U Tests of Semantic Differential
evaluation factor scores . . . . . . . . . . . 59

Semantic Differential mean D scores between
client concept "Me" and counselor concept
"Me" 0 I t: O t‘ O O O I O O O o O t‘ D G o 0 a O 62

Figure

LIST OF FIGURES

A graphic representation of the D scores
between different pairs of client percepts
as used in the study . . .

Graphs of D scores between the differently
combined pairs of clients' percpets of
"me" and "counselor"

Graphs of the potency factor scores for the
successful, partly successful and unsuccess-
ful counseling cases . . . . . . . . .

Graphs of the activity factor scores for the
successful, partly successful and unsuccess-
ful counseling cases . . . . . . . . . .

Graphs of the evaluation factor scores for
the successful, partly successful and un-
successful counseling cases . .

Graphs of D scores between clients' three

percepts of "me" and the counselor's per-
cept "me".

vi

Page

37

42

46

52

56

I. INTRODUCTION

Client-counselor relationship and sources of their
mutual influence are variables receiving increasing atten-
tion in counseling research. Counseling may be viewed as
an interactive interpersonal experience and it is generally
accepted that it involves a two-way interaction between
client and counselor who, in turn, serve as the parameters
to the counseling process.1 This study is concerned with
the client's perception of the counselor. Its focus is
upon the counselor as he is viewed by the client, and it
examines how this perception relates to or influences the

counseling process.

Purpose

A review of the literature suggests that there
are several aspects of a client's perception of his coun-
selor which appear to be influential to the counseling
process. First, is the counselor viewed as a model? That

is, does he provide a stable referent against which the

 

1The terms "counseling and psychotherapy" and
"counselor and therapist" are not clearly differentiated
in the research literature. Since most authorities agree
to overlap in the use of these terms, and since frequent
references to research using these terms will be made in
the present study, they shall be used interchangeably.

client can evaluate himself and the world, and a reference
with whom the client can judge his own behavior? Second,

if he does view the counselor as affording a frame of refer-
ence will he change so that he becomes more or less like
him? Third, does his perception of the counselor as strong
or weak, competent or incompetent affect the counseling out-
come?

The focus of this study is upon the client's per-
ception of the counselor. A client may perceive his coun-
selor as a referent, as a role model, or as one who provides
the psychological context necessary as a condition of change.
Specifically, the purpose of the study is to examine client-
perceived counselor potency, activity, evaluation, and simi-
larity of the client's self-percept and his counselor-percept
as variables influencing counseling process and outcome.

The study involves the use of self-referred college
students at a university counseling center as subjects. The
Minnesota Multiphasic Personality Inventory (MMPI) adminis-
tered before and after counseling will be used as the cri-
terion of outcome or change. The clients' self-percepts
and the clients' perceptions of the counselor are measured
with the Semantic Differential which was administered before

counseling and after every fourth interview.

The Problem

 

It is implicit, if not explicit, that a client

enters a counseling relationship so that he may achieve

psychological and/or behavioral change. Some studies have
dealt with measures of change as a means of evaluating coun-
seling outcome. Other studies have been concerned with the
delineation or isolation of factors upon which change is
contingent. Cartwright and Cartwright (1958), for example,
discuss the importance of a client's faith in his therapist
as an important variable related to client change. The
therapeutic relationship as an effective factor in psycho-
therapy was indicated in a study by Carkhuff and Truax (196$).
Client-perceived counselor potency and change in psycho-
therapy was studied by Borrelli (1965). Considerable at-
tention has also been given to change in counseling as a
function of client-counselor similarity.

Changes during counseling and the factors related
to change thus are a problem in understanding the counseling
process. The reported findings suggest that client-perceived
counselor potency and client-counselor similarity are impor-
tant influences in the counseling process and client change.

In this study an investigation was made of the
effect of client-perceived counselor potency, and similarity
of clients' self-percept and counselor-percept upon the pro-

cess and outcome of counseling.

Hypotheses

This study was conceived in a way permitting direct
examination of two hypotheses and the generation of an ex-

ploratory hypothesis. The two major hypotheses are: (1) that

a client's self percept will become more like his counselor-
percept over the course of successful counseling; (2) that
client-perceived counselor potency remains invariant during
the process of counseling in successful cases.

The third hypothesis is that client's self-percept
will shift toward the counselor's self-percept as a result
of successful counseling.

Stated in testable form the hypotheses--referring
to successful cases of counseling, as compared to partly -
successful and unsuccessful cases--are as follows:

1. A client's self-percept will become more like his
counselor-percept.

2. A client's percept of counselor potency will not
change over the term of counseling.

3. A client's self-percept will become more like the
counselor's self-percept.

The sample of twenty-two subjects for this study
was drawn from students seeking counseling of a "personal-
social" nature at the Michigan State University Counseling
Center. The client's percepts of himself and counselor
were obtained from his responses to the concepts "me" and
"counselor" on a form of the Semantic Differential which
was administered before counseling, following every fourth
interview, and at termination. Changes in these percepts
(i.e., the percepts "me" and "counselor") with respect to

each other over the course of counseling were measured with

the D statistic. Further comparisons were made using the
factors of potency, activity and evaluation (which account
for most of the variance of the Semantic Differential).
These factors were scored separately for each of the two
concepts (i.e., "me" and "counselor") and compared at three
different times, viz., pre, middle and post counseling.
Statistical analysis of these data were made using the
Kruskal-Wallis one way analysis of variance and the Mann-
Whitney U statistic.

The two profiles of MMPI scores for each subject
(taken before and after completion of counseling) were com-
pared and rated for change by two experts. These ratings
were used to classify the subjects as having changed posi-
tively, negatively, or not at all during counseling. This
served as the criterion for dividing the group of subjects
into three categories, viz., successful, partly successful,
and unsuccessful.

The terms self-percept and counselor percept, as
used in this study, shall refer to the client's responses
to the concepts of "me" and "counselor," respectively, on
the Semantic Differential. Potency, activity and evaluation
are the three major factors extracted from Semantic Differ-
ential data. Their meaning may best be indicated by the
specific scales that have high factor loadings for the par-

ticular factor in question. Potency scales are large-small,

thin-thick. Activity is identified by: fast-slow, active-
passive, sharp—dull. Some evaluation scales are: good-bad,
fair-unfair, free constrained.

Successful counseling cases, as indicated pre-
viously, were identified using pre and post MMPI profiles

as criteria.

Assumptions

 

One of the major assumptions of the present study
lies in the use of the Semantic Differential as an instru-
ment for the description of self and others. However, re-
peated factor analyses of Semantic Differential data yield
the same three factors as accounting for most of the variance,
viz., potency, activity and evaluation. Thus, as Osgood,
gt_gl. (1957, p. 325) state, ". . . most of the variance in
human semantic judgments could be explained in terms of a
relatively small number of orthogonal factors." This tends
to lend support to the apprOpriateness of the Semantic Dif-
ferential as an instrument for the description of self and
counselor as used in this study. Further, it is assumed
that the client's responses to the concepts "me" and "coun-
selor" represent his percept of himself and his percept of
his counselor.

A second assumption of this study is that the
MMPI affords a proper index of a client's "psychological"

change over the course of counseling and that valid indexes

of such change can be derived from the judges' ratings of
pre and post profiles. It is further assumed that such
changes in MMPI responses can be attributed to counseling.

Another assumption is that pre, middle and post
Semantic Differential Scores are comparable for subjects
with different numbers of counseling interviews. The
rationale is that successfully terminated clients are com-
parable despite differences in number of counseling inter-
views. Stated differently, in terms of counseling outcome,
a successfully terminated client of twelve interviews can
be considered similarly with a successfully terminated

client of twenty interviews.

Limitations

 

The present sample of twenty—two subjects does
not provide an Optimally large number of cases for division
into the comparison groups of successful, partly successful
and unsuccessful.

Differing numbers of interviews for different
clients obviate direct comparisons over particular inter-
views and further limit statistical analysis of the data.

The perceived "potency" of the counselor is of
particular import to the present study. However, only two
of the sixteen Semantic Differential scales yielded suf-
ficiently high factor loadings on potency to permit their

use. These were the scales large-small and thin-thick.

Responses to these adjectives could conceivably have been
influenced by the counselor's actual physical appearance.
They provide little breadth or range toward a definition

of potency, and a single deviant or spurious score could

markedly affect the mean scores derived therefrom. Thus,
the meaning that can be ascribed to potency is limited

accordingly.

Summary

Client-counselor relationship and sources of
their mutual influence provide an important area for study
in counseling. This study is concerned with specific as-
pects of the client-counselor relationship and the client's
perception of the counselor as a referent. Its purpose is
to investigate client-perceived counselor potency, activity,
evaluation, and similarity of the client's self-percept and
his counselor percept as variables influencing process and
outcome. A form of the Semantic Differential was used to
measure the client's percept of himself and percept of coun-
selor. Judged changes of MMPI responses (taken before and
after counseling) were used as criteria for successful,
partly successful or unsuccessful psychological change.

The hypotheses examined are as follows (referring
to successful cases of counseling, as compared to partly
successful and unsuccessful cases).

1. A client's self-percept will become more like his

counselor-percept.

2. A client's percept of counselor potency will not
change over the term of counseling.
3. A client's self-percept will become more like the

counselor's self-percept.

Chapter II provides a review of the pertinent
literature as a background to the study. Chapter III de-
scribes in detail the method and procedure of the study.
The results are presented and eXplained in Chapter IV. The
final chapter discusses the results, their meaning, impli-
cations, and indications for further research. It is fol-

lowed by a summary of the study.

II. REVIEW OF THE LITERATURE

The very nature of the counseling interview makes
the counseling relationship a unique interpersonal experi-
ence. It is a dyadic relationship whose successful outcome
(i.e., psychological and/or behavioral change in the client)
is contingent upon both members. Though each member is im-
portant to this relationship, it is the counselor who brings
his professional training to the situation and utilizes it
in a manner to facilitate client change. The way in which
the counselor is perceived and related to by the client
emerges as an important aspect to the understanding of coun-

seling.

The Counselor as a Referent
or Role Model

 

 

The counseling relationship forms a milieu within
which the client hOpes to change or modify his behavior.
Change is relative and the counselor may serve, among other
things, as a referent or role model. Rogers (1951) notes
that men are continually checking their perceptions against
one another in order to make them a more reliable guide to
reality. Each perception is in the form of an hypothesis
to be checked against further perceptions. The implication

here is that man checks these perceptions internally. It

10

11

would logically appear to follow that in a counseling re-
lationship the client can check his perceptions externally
against the counselor, or against the perceptions of the
counselor--using him as a referent. Goodstein in reference
to the counselor as a role model says:
. . . the counselor has become an important person to
the client and can often become a model for the client
to imitate. The counselor is typically seen as a well-
adjusted, successful professional person who is well
rewarded, both intrinsically and extrinsically, for
his efforts. The therapist's nurturance of the client
would ordinarily facilitate such social imitation in a
manner not unlike the classical conception of identi-
fication. (in Stefflre, 1965, pp. 174-175)

Bandura (1964) sees modeling as one of the impor-
tant influences in mediating behaviors. He indicates that
studies dealing with subject's observation of the behavior
of models suggest three rather different effects, each of
which may be reflected in an increase in the number, range,
and intensity of the observer's matching responses. First,
the subject may acquire new responses that did not previously
exist in his repertory. Second, observation of models may
strengthen or weaken inhibitory responses. Third, it is
possible that observation of a model sometimes elicits pre-
viously learned matching responses in the observer simply
because the perceiving of acts of a certain kind serves as
a "releaser" for responses of that same class. In direct

consideration of such effects in psychotherapy he states

that:

During the course of traditional therapy the
client is exposed to many incidental cues involving the
therapist's values, attitudes, and patterns of behavior.

12

They are incidental only because they are usually con-
sidered secondary or irrelevant to the task of resolving
the client's problems. Nevertheless, some of the
changes observed in the client's behavior may result,
not so much from the intentional interaction between
the client and the therapist, as from active learning
by the client of the therapist's attitudes and values
that the therapist has not directly attempted to teach.
(1964, p. 246)

Rosenthal (1955) in a study of changes of moral
values following psychotherapy found that in Spite of the
usual precautions taken by therapists to avoid imposing
their values on clients, the patients who were judged as
showing the greatest improvement changed their moral values
(in the areas of sex, aggression and authority) in the direc-
tion of the values of their therapists, whereas patients who
were unimproved became less like the therapist in values.

Landfeld and Nawas (1964) found that improvement
in psychotherapy is accompanied by a shift in the present
self of the client toward the ideal of the therapist within
the framework of the client's language dimensions.

Farson (1961), using the term "introjection" to
describe the forced adoption by one individual of the per—
sonality characteristics of another, investigated the degree
to which a client's self-descriptions tend to become more
congruent with the self-descriptions of his therapist. His
findings suggest that it is possible for a client to achieve
an adjustment in therapy which is independent of the per-

sonality of his therapist. The findings also imply, how-

ever, that the less adjusted, less competent therapist tends

13

in his relationship with his clients to induce conformity
in himself. These conclusions, though inconsistent, make
the question of counselor personality as a molding or
modeling influence more intriguing-~though perhaps less
clear.

If then counseling can be viewed as a learning
situation in which the counselor's values are ad0pted, even
though indirectly, it follows that the counselor and his
client's perception of him may be crucial to the counseling
outcome.

Client changes are thought to occur as a result
of counseling, and counseling is a uniquely intimate dyadic
interaction. The counseling relationship, therefore, emerges

as a critical factor in the outcome.

The CounselingRelationship

 

Counseling has been variously defined and often
used synonomously with psychotherapy. In this study it
shall refer to a dyadic professional relationship between
client and counselor whose purpose is to help the client
resolve his conflicts, reduce his anxiety or help him to
effect behavioral changes so that he may adequately deal
with his concerns and environment.

A good counseling relationship can be defined
many ways. However, as Shoben (1949) has indicated, there

is considerable agreement on this issue. The characteristics

14

most frequently cited as desirable are the counselor's
warmth, acceptance, permissiveness, respect for the client,
understanding, interest in the client, and liking for the
client. Rogers (1957, 1959) made the additional stipulation
that, in successful therapy the client must be able to per-
ceive these counselor qualities, and he also asserted
(Rogers, 1954) that the client must like and respect the
counselor.

Respect for someone implies a perception of ade-
quacy or strength, rather than inadequacy or weakness. The
word connotes a positive attitude of esteem. Accordingly,
the client's respect for the counselor is of particular im-
port to the present study as it may relate to perceived
potency of the counselor.

Fiedler (1950), in a frequently quoted study,
concluded that the concept of an ideal therapeutic relation—
ship is not described differently by therapists of different
schools. Interestingly, his study also indicated that lay
people perceived and described the ideal therapeutic rela-
tionship as well as trained therapists. He stated that his
data rather support the hypothesis that a good therapeutic
relationship is very much like any good interpersonal re-
lationship.

Fiedler's study has been criticized in several
respects. The items in his Q deck, for example, appear to

represent extreme positions and therefore would tend to be

15

sorted similarly by subjects who differ in their views of
Icounseling--thus, resulting in spuriously high correlations.
The training and experience variables do not appear pr0perly
controlled. Finally, it may be questioned that responses

to a scale may be a function of its content rather than a
function of the respondent. In spite of these criticisms
Fiedler's study implies that the ability to describe an
ideal therapeutic relationship may be a function of ex-
perience as Opposed to allegiance or indoctrination to a
particular school. That lay subjects described the ideal
relationship similarly to experts further implies that such
a relationship is aspired to as ideal in our society.
Present knowledge appears to support Fiedler's hypothesis
that the therapeutic relationship may be only a variation

of good interpersonal relationships in general.

Relationship and Therapeutic Change

 

Aspects of client-counselor relationship have
been studied in a number of ways. Counselor ratings of
client improvement has received some attention in studies
by Gorlow, Hoch, and Telschow (1952), Seeman (1954), and
Snyder (1961). These studies indicated a strong relation-
ship between judged progress of a client and liking a
client by the counselor.

Client ratings of counselor and improvement
(Feifel and Eells, 1962; Grigg and Goodstein, 1957; Lipkin,

1948) yielded similar results of a correspondence between

16

liking and progress. However, as Gardner (1964) indicates,
these studies suffer from interdependent measures. "It is
difficult to obtain independent measures of the two vari-
ables under consideration. When one judges a helping re-
lationship to be good--one tends to experience a feeling

of satisfaction which in turn acts as a set for perceiving
progress toward therapeutic goals." (Gardner, p. 428)

Ratings by outside judges of therapeutic change
and the quality of the therapeutic relationship have advan-
tages over those described above in that the client-counselor
relationship is rated by others who are not involved. Holt
and Lorborsky (1952), Fiedler (1950) and Van der Veen (1961)
conducted studies using outside judges. Again the results
tended to support the relationship between the quality of
the therapeutic relationship and therapeutic change.

Carkhuff and Truax (1965) conducted a study which
indicated that the therapeutic relationship, as established
by the therapist, is of paramount importance to patient im-
provement in psychotherapy. Lay hospital personnel were
trained in what was termed "effective therapeutic dimensions."
These dimensions were posited as necessary to effect thera-
peutic change. They were: (1) therapist accurate empathic
understanding; (2) therapist communication of warmth and
positive regard; (3) therapist genuineness of self-congruence;
and (4) patient depth of self-exploration. ”The therapist's

role was to communicate a warm and genuine concern and depth

17

of understanding." (Carkhuff and Truax, p. 428) The re-
sults indicated that lay therapists did provide uniformly
significant improvement to their patients, thus implying
that the nature of the established therapist-patient rela-
tionship was the crucial variable in patient improvement.

Although the results of various studies concern-

ing the counseling relationship and client change (i.e.,
therapeutic change) are not unequivocal, a preponderance

of the researchers appear to agree that a positive relation-
ship exists between these two variables. Present knowledge,
therefore, tends to support the importance of an "ideal
therapeutic relationship" in contributing to "therapeutic
change." Gardner indicates in his review:

The evidence that the quality of the therapeutic
relationship is a correlate of therapeutic change lies
not in the conclusive results of any one study but
rather in the repeated findings of a series of studies,
most of which contain one or more serious defects.
Methodology varies greatly, and absence of precise
definitions often makes it difficult to discern whether
the "good relationship” of one study contains the same
elements as that of another study or different. In a
sense, the diversity of procedure strengthens the force
of the conclusion. Null results do not cluster in any
one methodological cell, and, in all types of design,
positive results occur far more frequently. (p. 431)

In the present study two aspects of client-counselor

‘relationship shall be examined in respect to therapeutic
<:hange. Specifically, the relationship of client-perceived
cnounselor potency, and similarity of a client's self—percept

tc: his counselor percept will be studied as they relate to

61 successful counseling experience.

18

Counselor Variable: Similarity

 

Studies concerned with counselor variables have
dealt with such aspects as counselor personality character-
istics correlated with the quality of the counseling rela-
tionship. (Ashby, Ford, Guerney, and Guerney, 1957; Brams,
1961; Fiedler and Senior, 1952; Streitfeld, 1959). In
general, the results have not indicated positive relation-
ships.

Of particular bearing on the present study are
those investigations dealing with client-counselor similarity.
These investigations can be divided into two kinds, viz.,
those concerned with "assumed similarity" and those concerned
with "real similarity."

The assumed similarity construct (Fiedler, 1951)
refers to the degree to which a counselor considers himself
similar in personality to his client. A positive correlation
between assumed similarity scores and the degree to which
counselors liked their clients was reported by Hunt, Ewing,
LaForge, and Gilbert (1959).

However, the authors of this study are explicit
in presenting their study as an "approach to research" and
not as a conclusive study. The data are, therefore, sug-
gestive rather than conclusive.

Real similarity has been assessed in various ways.
It refers to the degree of similarity between client and

counselor as assessed by measures such as self report

19

inventories (e.g., Edwards Personal Preference Schedule,
MMPI, Semantic Differential). Axelrod (1961) using the
Rorschach as a criterion of therapist-patient similarity
found that ideation was a statistically significant factor
in a comparison of improved and unimproved patients. To
that extent he felt it affirmed his hypothesis that similar
patient-therapist personality characteristics are positively
related to therapeutic progress. Though not statistically
significant the factors of flexibility, breadth of interests,
and intelligence also indicated a trend in support of his
hypothesis. A major criticism to Axelrod's study lies in
the selection of his criterion groups of improved and un-
improved patients. Each therapist participating in the
research was asked to choose his two most improved and two
most unimproved patients. It has been indicated that liking
of a client and judged improvement of a client are corre-
lated. Consequently, Axelrod's use of the therapists'
judgments of improved and unimproved patients may be per-
ceived as a major weakness. More objective criteria of
selection would have been desirable.

Carson and Heine (1962) compared patients and
therapists on the basis of MMPI responses. They divided
their patient-therapist dyads into five groups ranging from
very high to very low similiarity. The hypothesis that

there is a curvilinear relationship between therapeutic

20

success and patient-therapist similarity was supported.

They reasoned that:
If psychotherapeutic success depends upon the thera-
pist's being able to achieve an Optimum balance between
empathy and objectivity in dealing with his patient,
then the relationship between patient-therapist simi-
larity and therapeutic success might well be of curvi-
linear form. With very high similarity the therapist
might be unable to maintain suitable distance and ob-
jectivity, whereas in the case of great dissimilarity
he would not be able to empathize with, or understand,
the patient's problems. (Carson and Heine, p. 38)

To this writer the Carson and Heine study seemed
to be well-designed, executed, and free of any major weak-
nesses. If its findings are corroborated by further inves-
tigation matching of clients and counselors could become an
accepted practice.

Mendelsohn and Geller (1965), using the Myers-
Briggs Type Indicator as a measure of client-counselor simi-
larity, obtained results tending to support Carson and
Heine's findings. They indicated that their results sug-
gested ". . . that some middle level of similarity allows
for the optimal balance of empathy and understanding as
well as objectivity on the part of the counselor.” (Mendel-
sohn and Geller, p. 71) In their study client-counselor
similarity on the Myers-Briggs Type Indicator was related
to greater length of counseling contact. Client evaluation
of outcome as satisfactory, (based primarily on the client's
satisfaction relative to the acquisition of objective infor-

mation) was related curvilinearly with client-counselor

similarity. Comfort-Rapport, (i.e., the degree to which

21

the client felt comfortable in the counseling relationship,
his feelings of being understood, etc.) resulted in both a
linear and curvilinear relationship with client-counselor
similarity depending on the sample. It was linear with the
freshman sample, curvilinear with the nonfreshman sample.
It was concluded that outcome varies with the criterion
used. Accordingly, Mendelsohn and Geller suggest that much
of the inconsistency in the results of studies using simi-
larity as a criterion could be explained by the use of dif-
fering outcome criteria as well as of differing samples.

A similar criticism may be made against their
study. The question may be posed as to what are acceptable
outcome criteria? Further, of Mendelsohn's and Geller's
sample of seventy two clients only two had as many as five
sessions, four had four sessions and the remainder had
three or less. It seems reasonable to question the type or
manner of client change that occurs under such brief condi-
tions.

Ourth (1964) examined the relationship between
degree of client-counselor similarity and length of stay or
degree of improvement. He developed his own dimension la-
beled Internal-External (I-E) orientation for use as a cri-
terion of similarity. The I-E dimension was designed to
measure "whether a person could use self-synthesizing pro-
cesses as the basis for personal validation (and therefore

be internally oriented) or whether the source of his sense

22

of validity was some external referent (i.e., either a dog-
matic code or the contemporary environment)." The degree
of therapist-client I-E similarity was not found to be re-
lated to the client's length of stay in therapy or his
judged improvement. SchOpler (1958) also failed to find
evidence of a relationship between psychotherapeutic out-
come and patient-therapist similarity. The question of
suitable or standard outcome criteria suggests an explana-
tion for these negative results which are in contradiction
to other findings. Tuma and Gustad (1957), on the other
hand, reported positive results where close resemblence
between clients and counselors was associated with rela-
tively better criterion performance by clients. The depend—
ent variable in their study was client learning about self,
based on a Self-Knowledge Inventory.

The discrepancies in many of these studies of
similarity, suggests Gardner, ". . . strongly support
Levinson's (1961) thesis that patient-therapist similarity
cannot be thought of as a unitary trait. Some similarities
may facilitate good relationships and therapeutic progress,
while others may be sources of impasses." (Gardner, p. 434)

One of the purposes of the present study is to
examine client-counselor similarity in a different way. One
aspect shall be a comparison of the client's self-percept
with his counselor—percept as it relates to counseling out-

come and process. Rather than considering these percepts

23

as being static or fixed it is hypothesized that they will
change in respect to each other during counseling. More

specifically, it is hypothesized that successful outcome as

 

compared to partly successful and unsuccessful outcome, will

 

be characterized by a shift in the client's self;percept in

 

the direction of his counselor-percept. Ewing (1954) re-

 

porting the results of his study indicated that clients
deemed most improved changed their rating of the self figure
so as to be more like their counselor figure. He also re-
ported that clients "deemed most improved" changed their
rating of the counselor figure to be more like their second
rating of the self figure. These results imply that posi-
tive therapeutic change is related to increased congruence
of a client's self-percept and his counselor-percept.
Stated in another way, the client's self-percept becomes
more similar to his counselor percept following successful
counseling. In addition, a similar hypothesis and compari-
son shall be made regarding the client's self-percept and
the counselor's self-percept. It is hypothesized that the

client's self percept will change in the direction of the

 

counselor's selfapercept in successful cases of counseling

 

as compared to partly successful and unsuccessful cases.

 

Counselor Variable: Potency

 

The variable client perception of counselor strength

has received relatively little attention in empirical studies.

24

Yet its importance, whether explicit or implicit, has been
suggested by a variety of writers.

Rogers (1952) considers the client's perception
of the counselor, his attitudes, and procedures as crucial
to the counseling outcome. He has also considered it es-
sential that the client like and respect the counselor
(Rogers, 1954).

Stransky (1946) implies that the therapist must
be perceived as potent in that he views the nature of rap-
port between client and psychotherapist as a "subordination-
authority" relation. Maeder (1955) implies the importance
of potency even more forcefully in that he sees the thera-
pist-patient relationship as the "archtype of the savior."
Studies of the "placebo effect" in psychotherapy by Rosen-
thal and Frank (1956) indicate that improvement may be a
function of ". . . the patient's fa1th in the efficacy of
the therapist and his technique." (Rosenthal and Frank,
p. 300). Rosen e£_al. (1961) in a study of perceived sources
of social power demonstrated that pe0p1e regarded as helpful
are perceived to possess power. Kell and Mueller state that
"Adequacy is the counselor's passport to unraveling his
client's conflicts and effecting change." (1966, p. 86)
They view client-perceived counselor strength as a pre-
requisite for client improvement. Borelli (1965) found a
direct relationship between clients' perception of counselor

potency and changes in the Hysteria scale of the MMPI. This

25

overview lends some support to the significance of counselor
potency (adequacy or strength) as an important factor in
counseling.

In the present study client-perceived counselor
potency will be examined in respect to counseling process
and outcome. It is reasoned that a client in successful
counseling perceives his counselor as a source of strength
who is capable of helping him to resolve his difficulties.
Perceiving the counselor as inadequate would cause a de-
generation of the relationship resulting in failure. Ac-
cordingly, the hypothesis to be tested is that a client's

percept of counselor potency will remain invariant during

 

the series of interview in successful counseling as com-

 

pared to partly successful and unsuccessful counseling.

 

Summary

The counseling relationship is a unique, dyadic
interpersonal encounter through which the client hopes to
change or modify his behavior. His counselor is a major
variable in facilitating or making such change and may be
used as a referent against whom the client can check his
perceptions. The counselor may thus serve as a role model
influencing client changes. Accordingly, the counseling
relationship may be perceived as critical to counseling

outcome 0

26

The characteristics of the counseling relation-
ship most frequently cited and generally accepted as de-
sirable are the counselor's warmth, acceptance, permissive-
ness, respect for the client, interest in the client, liking
for the client, and understanding. Research, though equivo-
cal, tends to support a relationship between the quality of
the therapeutic relationship and therapeutic change.

In the present study two aspects of client-counselor
relationship in respect to therapeutic change will be examined,
viz., similarity of client self-precept to counselor-percept,
and client-perceived counselor potency.

Research studies of client-counselor similarity and
outcome are inconclusive. Some of the differences may result
from the different outcome criteria used in the various
studies.

The present study is an examination of the similar-
ity between a client's self-percpet and his counselor percept
as it relates to counseling process and outcome. It is hy-
pothesized that, in the course of successful counseling, as
compared to partly successful and unsuccessful counseling,
the client's self-percept will become more like his counselor-
percept.

Client-perceived counselor potency is a second as-
pect of client-counselor relationship to be examined.

Although there is a paucity of research regarding counselor

potency (competence or adequacy) the literature suggests

27

that client-perceived counselor potency may be integral to
successful counseling outcome. The hypothesis examined in

the present study states that client:perceived counselor

 

potency remains invarient (i.e., stable) throughout the
counseling process in successful cases as compared togpartly

successful and unsuccessful cases.

 

III. METHOD

The Sample

 

The data used in this study are a part of an on-
going research project at the Michigan State University
Counseling Center. Fifty-four students seeking counseling
of a personal-social nature agreed to participate in the
research. Each counselee was administered a form of the
Semantic Differential before counseling, following every
fourth interview, and at termination. They were asked to
respond to twenty-one concepts composed of sixteen bi-polar
scales. Two of these concepts, viz., "me” and "counselor"
are the ones investigated in this study. MMPI profiles
were also obtained before and after counseling for a number
of these clients.

For the purposes of this study only those subjects
with pre and post MMPI Profiles, and at least two Semantic
Differential administration could be utilized. This usable
sample consisted of twenty-two self-referred undergraduate
students composed of twelve females and ten males. Their
counselors were advanced graduate practicum students, in-
terns, or Counseling Center Staff members composed of fifteen
males and eight females. Table 1 presents these data and the
number and times of administration of the Semantic Differen-
tial for each subject.

28

29

Table 1. Listing of clients and counselors, and the number
and times of administration of the Semantic Dif-
ferential for each subject.

 

 

Semantic Differential

Client Counselor Interview Administered

 

 

 

I.D. No. Level Sex Pre 4 8 12 16 20
Mr. A 19 Intern M x x x x
Miss B 53 Intern F x x x x x
Mr. Ben 43 Intern M x x x x
Miss Beth 56 Practicum F x x x x x
Miss Betsy 98 Practicum M x x x x #
Mr. Bill 13 Staff M x x x x
Miss Clara 94 Practicum F x # # x
Mr. E. 79 Intern M x # x x x x
Mr. Earnest 97 Practicum M x x # x
Mr. Ed 19 Staff F x x x x
Miss Elaine 40 Staff M x x x x x x
Mr. Hans 85 Intern M x x x x x x
Miss Jan 70 Intern F x x x x x #
Miss Jane 33 Staff M x x x x x x
Mr. Jim 10 Staff M x x x x x
Miss June 74 Intern M x x x x x
Miss Kate 41 Intern F x x x x x
Miss Lo 19 Staff F x x x
Miss Margie 59 Practicum M x # x # x
Miss Pear 11 Staff M x x x
Miss Pen 78 Intern M x x x x
Mr. Pete 79 Intern M x # x x x
Mr. Sam 78 Intern F x x x

 

x - administered
# - data missing

30

There was no reason to believe that sex differ-
ences would Operate differently. However, as a precaution
all individual cases were scored and graphed separately.
No evidence was found to indicate that sex differences did

Operate differently.

Procedure

 

All students requesting personal counseling were
screened by an intake counselor. During the intake inter-
view the counselor assessed the problem and decided whether
or not the client was a suitable candidate for participation
in the research. The following criteria of suitability were
used:

1. The student must not have been seen in a counsel-
ing relationship at any previous time--at the

Michigan State University Counseling Center or

elsewhere.

2. He had to be an undergraduate student.

3. His problems had to be judged of a nature that

were suitable for the agency.

If these criteria were met the student was asked to partici-
pate in the research project. He was told that his commit-
ment to the project would require approximately two hours
each of pre and post testing, a monthly testing of about
forty minutes during the period of counseling, and permis-
sion that the interviews be tape-recroded. If these re-

search conditions were agreed to, the student was accepted

31

as a research client. He was then taken to the testing
room where preparations were made for administration (be-
fore his first interview) Of the MMPI, a Semantic Differ-
ential, and a Personal Data Sheet. The client was then
assigned to one of the research project's counselors whose

free appointment time matched his.

Instruments: The MMPI and
Categorization ofESubjects

 

 

The MMPI was develOped in 1940 for the purpose
of diagnosing psychOpathology (Dahlstrom and Welsh, 1960).
The inventory consists of 550 affirmative statements which
the subject is asked to classify into three categories:
Ture, False, and Cannot say. In its most common form the
MMPI consists of nine clinical scales and three validating
scales. However, it has become one of the most extensively
researched instruments of its kind and now has over two
hundred scales. In order to aid assessment in this study
two additional (i.e., additional to the nine common clini-
cal scales) were scored. These were the Anxiety (A) and
Ego Strength (Es) scales (Welsh and Dahlstrom, 1956).

In the present study changes on the MMPI were
used as a criterion of psychological change. Scores on the
MMPI were the basis upon which a client was judged to have
changed positively, negatively, or not at all, over the
period of counseling. Two judges, both of whom are con-

sidered expert in the use and interpretation of the MMPI,

32

were asked to compare the pre-counseling and post-counseling
profiled MMPI scores of each subject. Specifically, they
were asked to rate client change on a five point scale rang-
ing from "unsatisfactory" to "satisfactory."

In order to test for intra-judge reliability each
judge was asked to rate the profiles twice. The span of
time between each series of ratings for each judge was one
week. As indicated in Table 2, intra-judge reliability
was high. For example, a comparison of the first and second
series of ratings of Judge I reveals that identical ratings
were given to twenty-one of the twenty-three subjects and
essentially the same rating (4 and 3 +) was given to one of
the remaining two (Miss B). Thus the first judge was con-
sistent on twenty-two of twenty-three judgments. Judge II
gave identical numerical ratings to sixteen subjects, and
essentially the same rating to six. He differed markedly
only with Mr. Jim, judging this change to have been partly
satisfactory the first time and partly unsatisfactory the
second time., Thus, he also was consistent with twenty-two
of twenty-three judgments.

Inter-judge reliability was comparably high.
Thirteen subjects were given identical numerical ratings
on each of two judgments by each judge. Thus, four numeri-
cal ratings for these clients were equal (as an example see
Table 2--ratings for Mr. A and Mr. Ben). Four subjects

received three out of four numerically equal ratings and

33

Table 2. The ratings of client changes on the MMPI.

 

 

Ratings of Judge I Ratings of Judge II Final

 

 

Client Judged
lst 2nd 3rd lst 2nd 3rd Ratings
Mr. A 4 4- 4 4 4
Miss B 4 3+ 3 2 1 2 1
Mr. Ben 4 4 4 4 4
Miss Beth 3 3 3 3 3
Miss Betsy 4 4 3 4 4
Mr. Bill 5- 5 5 4 5
Miss Clara 1 l l l 1
Mr. E. 4- 4— 4 4 4
Mr. Earnest 4+ 4 4 4 4
Mr. Ed 5 5 5 5 5
Miss Elaine 5- 4- 5 4 4
Mr. Hans 4- 4+ 4 4 4
Miss Jan 4— 4 3+ 3- 2 4 4
Miss Jane 3+ 3 4 4 4
Mr. Jim 3 2 2 4 2 2 2
Miss June 5+ 5+ 5 5 5
Miss Kate 5 5 5 5 5
Miss Lo 4 5 5 5 5
Miss Margie 5 5 5 4 5
Miss Pear 2 2 l l 2
Miss Pen 1 l l l 1
Mr. Pete 4 4- 4 4 4
Mr. Sam 4 4 4 4 4

 

Ratings: 1 - satisfactory; 2 - partly satisfactory; 3 - no
change; 4 - partly unsatisfactory; 5 — unsatisfactory.

34

the discrepant rating differed by only one unit. For ex-
ample, Miss Betsy was given a rating of 4 on both judgments
by Judge I and a rating of 3 and 4 on the first and second
ratings, respectively, by Judge 11. Three subjects (Miss
Jane, Miss Elaine, Miss Pear) differed on two of the four
ratings. However, this difference was only by one unit and
it was in the same direction relative to satisfactory or
unsatisfactory change. That is, the scores were either 4
or 5 (representing satisfactory change), or 1 or 2 (repre-
senting unsatisfactory change). Thus, twenty of the twenty-
three clients were rated similarly as having changed posi-
tively, negatively, or not at all. For the three remaining
subjects each judge was independently asked to rate his pro-
files a third time, (one week later). These ratings are
also indicated in Table 2, viz., Miss B, Miss Jan and Mr.
Jim. No agreement was reached on Miss B. Consequently,
this case was excluded from the sample. Miss Jan was judged
similarly four of the six times. The third judgments of 3+
and 4, respectively, were essentially the same and a final
rating of 4 was agreed upon by the judges and the writer.
Mr. Jim was treated similarly and the final rating of 2 was
agreed upon. Again, note that four of six judgments for
Mr. Jim were identical.

For the purposes of this investigation the final
ratings were divided into three categories, viz., unsatis-
factory, partly satisfactory and satisfactory. This cate-

gorization is presented in Table 3.

35

Table 3. The f1na1 grouping and classification Of the MMPI

 

 

 

 

ratings.
Final MMPI Ratings
Unsatisfactory Partly Satisfactory Satisfactory
5 11 6

 

Using these divisions, six of the usable sample of twenty-
two subjects were judged to have been successful counseling
cases, eleven were considered to have been partly success-
ful and five were classed as unsuccessful. This was the

final grouping used in the testing of the hypotheses.

Instruments: The Semantic Differ-
ential and Manner of Utilization

 

 

The Semantic Differential was developed by Osgood
and his associates (1957) as a research tool for the measure-
ment of meaning. It is essentially a controlled association
and scaling procedure. Each concept to be differentiated
is rated on a series of seven point scales composed of ad-
jectives that are polar Opposites. Two of the scales used
in this study are presented below. (See Appendix A for all

the scales used).

active : : : x : : : passive

_‘¥

bad x : : : : : : good

 

36

The respondent is asked to rate a concept (e.g.,
mother) as being "very closely related" to one end of the
scale, "quite closely related," "slightly related," or
"neutral" on the scale. The "neutral" position is repre-
sented by the middle space (indicated by an x on the active-
passive scale in the example on the previous page). A
"very closely related" position is represented by the ex-
treme right or left position (see the bad-good scale, p. 35).

Osgood has concluded, on the basis of extensive
research, that the great majority of the factorial structure
is-composed primarily of three factors--activity, evaluation,
and potency. The potency variable is of particular import
for the purposes of testing one of the hypotheses of this

study, viz., that client perceived counselor potency is in-

 

variant over the course of successful counseling.

Measures of the client's perception of himself
and his counselor were Obtained from his responses to the
Semantic Differential concepts of "me" and "counselor."
These measures were obtained before the first counseling
interview (i.e., following the intake interview) and every
fourth interview thereafter, as indicated in Table l. The
counselors' responses to the concepts "me" and "counselor"
were also obtained before counseling took place.

In semantic space if two concepts are close to-
gether they are alike for the individual making the judg-

ments. Conversely, if two concepts are far apart they

37

differ in meaning. Osgood and his associates use the D

statistic (a measure of distance) as a measure of the re-

lation between any two concepts (Osgood gp_al., 1957).

This measure (i.e., the D statistic) was used in the pres-

ent study to assess similarity or dissimilarity between

the concepts under investigation (viz., "me" and "counselor").
In order to test the first hypothesis, that a

client's self-percept will become more like his counselor

 

percept in successful cases of counseling as compared to

 

partly successful and unsuccessful cases, a number of D

 

measures between various pairs of concepts were made.

These involved three types of comparisons which are graphi-
cally represented in Figure l for a client who responded
to the Semantic Differential three times. The same pro-
cedure was used for each subject with the maximum number
of Semantic Differential administrations ranging to six

for any one subject.

Figure l. A graphic representation of the D scores between dif-
ferent pairs of client percepts as used in the study.

 

 

 

Time I Time II Time III
lst .
Comparison .F_“D__7 i——4}__1 .F_—D—_7
2nd _
Comparison + D ¢

D

+ +
Concepts Me Cslr Me Cslr Me Cslr
3rd 1* D 1.
Comparison

38

The first comparison was made by computing D scores
between the concepts "me" and "counselor" for each time Of
testing (i.e., Time I, and II in the previous example).
Diminishing scores indicated movement of the concepts toward
each other, i.e., increased congruence of the concept "me"
with the concept "counselor." However, this was not suffi-
cient information to indicate if one or both concepts had
shifted. To determine such movement two more comparisons
were necessary. Therefore, the second comparison was made
using the first "me” as a referent and comparing it with
the subsequent "counselor" percepts. Differences in this
case could be attributed to movement of the "counselor"
variable (i.e., the client's percept of counselor). The
third comparison also used the first "me" concept as a
referent. This time, however, it was compared with the
subsequent "me" concepts, thus permitting determination
of a directional shift for this variable.

For the testing of the third hypothesis, that a
client's selfjpercept will become more like his counselor's
self-percept in successful cases of counseling as compared
to partly successful and unsuccessful cases, similar scoring
procedures were utilized. With the counselor's self-percept
used as a referent, each client's concept of "me" over the
course of counseling was compared with it. Decreasing
scores tended to support the hypothesis, while increasing
scores indicated movement away from the counselor's self-

percept.

39

The second hypothesis, that a client'sgpercept of

 

counselor potency remains invariant over the course of suc-

 

cessful counseling as compared to partly successful and un-

 

successful cases, was studied in a different manner. In

 

this instance it was necessary only to compare the scores
on those scales measuring potency. A factor analysis of
the Counseling Center research sample by Borrelli (1965)
established the factorial meaning of the various scales
(see Appendix B). Two scales: large-small, and thin-
thick were shown to measure the potency factor. A com-
parison of these scale scores was used as the criterion
of change or stability of this variable.

The activity and evaluation factors for each
subject were also scored and compared for the percept "me"
and the percept "counselor." Activity was composed of
four scales while Evaluation was composed of eight scales
(see Appendix B).

D scores and raw scores, as indicated above,
were computed separately for each of the three groups,
i.e., the successful, partly successful, and unsuccessful
groups. These data were statistically analyzed using the
Kruskall-Wallis method of one way analysis of variance and

the Mann-Whitney U test.

IV. RESULTS

In order to test the first hypothesis that a

client's self-percept will become more like his counselor

 

percept in successful counseling cases as compared to

 

partly successful and unsuccessful cases, D scores were

 

computed between the concept "me" and the concept "coun—
selor" as indicated in Chapter III. To permit a statisti-
cal comparison Of the data between the three groups of
successful, partly successful and unsuccessful client's
mean D scores were derived for first, middle and terminal
administrations of the Semantic Differential. These three
scores were selected to represent before, middle and end
of counseling and are presented in Table 4 in the columns

headed I, II, and 111.1

Table 4. Semantic Differential mean D scores between
clients' concept "Me" and clients' concept
"Counselor."

 

 

D Score—Me/Cslr D Score-Me/Cslr T D Score-Me/Me T
Group Time Administered Time Administered Time Administered
PI' II III I II III I II’ III

 

 

 

S 7.85 9.58 7.85 7.85 8.61 8.65 5.85 7.81 7.01
PS 7.69 7.32 6.13 7.69 7.62 8.05 5.54 5.91 6.49
U 7.42 8.49 6.72 7.42 7.14 6.69 4.07 5.32 4.84

 

1Hereinafter the scores for pre, middle and termi-
nal administrations of the Semantic Differential shall be
represented as I, II and III, respectively.

40

41

The first group of scores, labeled Me/Cslr, are
the mean D scores between the concept "me" and the concept
"counselor" for each of the three times. Decreasing scores
from I to 111 indicate convergence of the "me" and "counse-
lor" percepts over the course Of counseling, while an in-
crease represents divergence.

The second group of scores labeled Me/Cslr T (con-
cept "me" with concept "counselor" over time) are the mean
D scores between the first client percept "me" and the sub-
sequent counselor peGCets. In this instance the first
client percept "me" was held constant and compared with
each of the subsequent client percepts of "counselor."
Thus, any change in these scores represents a change or
movement of the client's "counselor" percept.

The third group labeled as Me/MeT (concept "me"
with concept "me" over time) is a comparison of the first
"me" percept with the subsequent "me" percepts. As in the
preceding group, the first client percept "me" was held
constant and compared with each of the subsequent percepts
"me." Changes in these scores represent changes in the
client's self-percept over the course of counseling.

The statistical treatment of the scores in Table
4 is discussed belowiJiconjunction with Figure 2 which is
a graphic representation of these scores. An inspection
of Figure 2a, which is a depiction of the convergence or

divergence of the percepts "me" and "counselor," reveals

 

42

no marked differences between the three groups. A Kruskal-
Wallis one-way analysis of variance indicates that there is
no significant difference between the three groups in re-
spect to the sums of ranks between scores I and III (i.e.,
first score and terminal score) of this Me/Cslr ("me" -
"counselor") comparison (H = 1.4, df = 2, p < .50), nor are
there any significant differences between scores I and II
on this variable (H = 3.1, df = 2, p < .30). Therefore, it
does not appear that variation (i.e., convergence or diver-
gence) of a client's self-percept differs with respect to
membership in the successful, partly successful, or unsuc-

cessful groups.

Figure 2. Graphs Of D scores between three differently com-
bined pairs of clients' percepts of "me" and
counselor.

Me/Cslr Me/Cslr T Me/Me T

 

 

10» 10?
/.—.—-"‘j
A frog...
CO... d”.
l ’4"
S-b 5"’;."..‘\
1.. ‘.o
.0
l n 1 n l 4

 

 

 

 

 

Successful Group, N = 6
-------- Partly Successful Group, N = 11
........ Unsuccessful Group, N = 5

43

In Figure 2b, which represents change or stability
of the counselor percept, little disparity appears between
the three groups. Differences between scores I and III were
computed and ranked as before. A Kruskal-Wallis one-way
analysis of variance again indicated no significant differ-
ence between the three groups (H = 3.3, df = 2, p < .20).
Thus, there appears to be no difference among the three
groups of successful, partly successful, and unsuccessful
cases with respect to the perceived stability of the coun-
selor percept over the course of counseling.

The change or stability of the "me" percept is
graphed in Figure 2c. One-way analysis of variance, applied
similarly as before, indicated no significant differences
among the three groups between scores I and III, i.e., be-
tween pre and end of counseling scores. (H = 4.4, df - 2,

p < 0.15). It may therefore be concluded that the vari-
ability (i.e., change) of the client's self-percept over
the course of counseling does not differ statistically
among the three groups of successful, partly successful
and unsuccessful counseling cases.

The second hypothesis that clientjperceived coun-

 

selor potency remains invariant over the course of success-

 

ful counselingas compared to partly successful and un-

 

successful counseling was tested by comparison of the potency

 

scores of each group.

44

A tabulation of mean raw scores for the factors
of potency, activity and evaluation are presented in Table 5
for the percepts "me” and "counselor.” As before, scores
for the three groups represent the first, middle and final
administrations of the Semantic Differential (columns I, II,
and III). These scores are graphically presented as Figures
3, 4 and 5, representing the factors of potency, activity
and evaluation, respectively.

One way analyses of variance were computed for
differences between the sums of ranks of a number of the
potency scores for the three groups. The results of these
tests are shown in Table 6.

In Table 6 the column labeled "SD Concept" indi-
cates the Semantic Differential concept or concepts ("me"
or "counselor" or both) being considered. The next column
(i.e., Scores Compared) lists the specific scores being
considered, where Roman numeral 1, II or III indicates the
first, middle or terminal administration of the Semantic
Differential, and the letter immediately following each
numeral represents the concept "me" or the concept "counse-
lor." For example, the first row states that the concept
being considered is "me." It is a statement of the test
of significance (p < .50) of the first client percept of
self potency (IM) relative to differences in the score
between the three groups. Rows two and three are state-

ments of the tests of significance of the middle and terminal

45

 

o.mm w.vm N.wm ¢.vm 0.5m o.mm
N.nm v.mm w.vm o.mm 0.0m N.mm
o.wm w.mm v.5m o.nm m.wm N.vm

0.5H N.5H m.wH N.~H o.mH N.o~
m.HN o.o~ u.wH o.mH m.o~ o.mH
w.wa n.NN n.wH m.om o.NN m.mH

e.m o.m e.m N.w 0.5 ~.w D
u.m H.w m.m m.n m.m H.u mm
m.m m.n o.oa m.n m.m m.m m

 

HHmu oz uHmu oz pamu oz

 

HHH HH H

hamu

oz “Hmu 0: hamo

0:

 

 

wopoumflcflew< mafia

HHH

HH

HHmu oz nHmo o: pamu O:

 

 

 

:Ofiuwzam>m

vehoumflcfiew< oEHH

HHH HH H

 

 

Refi>uuu<

macho
voyeumfi2H6u< mafia

 

xocowom

 

 

.:0flumsfim>o paw xufi>fiuom .xocou0d mucoumm on“ how monoom 3mg :moE Hmflucouommflm ufipcmEOw .m manmh

Figure 3.

46

Graphs of the potency factor scores for the

successful and unsuccessful counseling cases.

10 q

 

I l
I II III

a.
Successful Group

10.

 

W.

h”—'
”

r,

II III

b

Partly Successful

Group

Client's "counselor" percept
-- ------- Client's "me" percept

lO‘f

 

 

S-H—
5
I II III
c.
Unsuccessful
Group

 

 

 

Table 6. Kruskal-Wallis one-way analysis of variance of Se-
mantic Differential potency factor scores.
SD Concept Scores Compared Results
1. Me IM p < 50
2. Me IIM p < 70
3. Me IIIM p < 20
4. Counselor IC p < 70
5. Counselor IIC p < .50
6. Counselor IIIC p < .50
7. Me IM-IIIM p < .80
8. Counselor IC-IIIC p < .70
9' Me IC-IM < 10
Counselor p ’
10. Me
Counselor IIC-IIM p < .30
11. Me
Counselor IIIC-IIIM p < .30

 

I - lst SD administration
II - Middle SD administration
III - Terminal SD administration

M - Me concept

C - Counselor concept

47

client percepts of self potency, respectively (IIM, p < .70;
IIIM, p < .20). The differences of the magnitudes between
corresponding scores (i.e., first, middle and terminal) be-
tween the successful, partly successful and unsuccessful
groups for client-perceived potency of self do not appear
to be significant.

The fourth, fifth and sixth rows of Table 6 dupli-
cate the comparisons of the first three rows with respect
to the client's percept of counselor potency. The results
(IC, p < .70; IIC, p < .50; IIIC, p < .50) indicate that
the differences of the magnitudes between corresponding
scores.(i.e., first, middle and terminal) between the suc-
cessful, partly successful, and unsuccessful groups for
client-perceived counselor potency do not appear to be sig-
nificant.

The analysis of groups across times (as in the
first six rows of Table 6) did not result out of the hy-
potheses but were rather a means of determining if the
findings were to be attributed to differences Of status
in the groups. Inasmuch as they did not appear to be dif-
ferent, the assumption of equivalence of groups was made
throughout.1

Row seven states that the concept being considered
is "me," and that the difference between the first and ter-

minal "me" scores (IM-IIIM) is being tested for significance

 

1Similar analyses and assumptions were made in re—
Spect to the activity and evaluation factors as will be in-
dicated later.

48

of difference between the three groups. The resulting prob-
ability, p < .80, of course, indicates no statistical sig-
nificance for the differences of these two scores between
the three groups. That is, the difference between the
clients' "me” percepts between the first and final test
administration is not statistically different among the
three groups. Therefore, it does not appear that the vari-
ability or stability of client-perceived self potency is
different among the successful, partly successful or un-
successful groups.

The eigth row of Table 6 shows the comparison
of the client's first and terminal counselor percept (IC-
IIIC; p < .70). The result indicates no statistical sig-
nificance among the successful, partly successful and
unsuccessful groups with respect to the difference between
these two scores. Thus, the variability of clients' per-
ceived counselor potency does not appear to be different
among the three groups with respect to first and terminal
scores.

The ninth row Of Table 6 indicates the compari-
son of the first "counselor" percept and "me" percept (IC-
IM) scores of the potency factor among the three groups.
The difference between these scores is not significant
(p < .10). However, the probability is considered suffi-
ciently high to warrant a closer examination (to be dis-

cussed later). It thus appears that clients' perceived

49

differences in potency between themselves and counselors,
prior to counseling, may be different with reSpect to the
successful, partly successful and unsuccessful groups.

The tenth and eleventh rows of Table 6 compare
the same two scores described for row nine but for dif-
ferent times, i.e., for middle and terminal administrations
of the Semantic Differential, respectively (IIC-IIM; IIIC-
IIIM). The resulting probability of .30 in each case is
not statistically significant. However, as in row four,
it is considered sufficiently high to justify closer
scrutiny as will be discussed later. These results suggest
'that client-perceived potency differences between self and
counselor during middle and final stages of counseling may
be different for the successful, partly successful and un-
successful groups.

Table 7 is the tabulation of Mann—Whitney U tests
performed on the potency data. These tests were made as a
result of some of the suggested differences indicated by
the one-way analyses of variance discussed relative to
Table 6. In Table 7 the first row is interpreted as
follows: the letters S-U under the column headed "Groups
Compared" indicate that the comparison being made is be-
tween the successful and unsuccessful groups. The Specific
"Scores compared” between these two groups, (IC-IM), are the
differences between the first "counselor" concept potency

score and the first "me" concept potency score for the

50

Table 7. Mann-Whitney U Tests of Semantic Differential
potency factor scores.

 

 

 

 

Groups Compared Scores Compared Results
S-U (IC—IM) p < .165
PS-U (IC-IM) p > .05
S-U (IIC-TIM) p < .214
PS-U (IIC-IIM) p > .05
S-U (IIIC-IIIM) p < .381
PS-U (IIIC-IIIM) p > .05
S - Successful
PS - Partly Successful
U - Unsuccessful

successful and unsuccessful groups. The resulting prob-
ability, p < .165, is interpreted to indicate that there
is no significant difference between these two potency
scores with respect to the successful and unsuccessful
groups.

The second row is a comparison of the same two
scores (i.e., IC-IM) with respect to the partly successful
and unsuccessful groups (PS-U). The resultant probability
(p > .05) is not significant. Therefore, it does not ap-
pear that the perceived difference in client-counselor
potency before counseling is different for the part1y.suc-
cessful group as compared to the unsuccessful group.

Rows three and four repeat the comparisons of
rows one and two, respectively, for the middle potency

scores (IIC-IIM). Row three indicates no statistical

51

difference between these two scores (p < .214) between the
successful and unsuccessful groups (S-U). Row four indi-
cates no statistical difference (p > .05) between these
two scores relative to the partly successful and unsuccess-
ful groups (PS-U). Perceived potency differences between
self and counselor, therefore, do not appear to be different
during the middle phase of counseling for either the suc-
cussful, partly successful or unsuccessful counseling cases.

Rows five and six repeat the above comparisons
for the terminal phase of counseling (IIIC-IIIM). The re-
sulting probabilities (p < .381, S-U; p > .05, PS-U) are
not significant. Therefore, this may be interpreted to
indicate that perceived potency differences between self
and counselor at the terminal phase of counseling do not
appear to be differentiated among the successful, partly
successful or unsuccessful counseling groups.

Figure 4 represents the scores of the Semantic
Differential on the activity factor as listed in Table 5.

The results of the statistical analyses per-
formed on the activity scores are summarized in Table 8.
They are presented in the form and manner described for
the potency factor scores in Table 6. For example, the
first row states that the concept being tested is "me."
It is a statement of the test of significance (p < .70)
of the first client percept (IM) of self for the activity

factor relative to differences in the score between the

52

Figure 4. Graphs of the activity factor scores for the suc-

cussful, partly successful and unsuccessful coun-
seling cases.

251- 251' 25]-

 

 

 

 

 

 

 

l l I l I J
I II III I II III I II III
a. b. c.
Successful Group Partly Successful Unsuccessful
Group Group

 

Client's "counselor" percept
--------- Client's "me" percept

Table 8. Kruskal-Wallis one-way analysis of variance of
Semantic Differential activity factor scores.

 

 

 

SD Concept Scores Compared Results
Me IM p < .70
Me IIM p < .50
3. Me IIIM p < .20
4. Counselor IC p < .30
5. Counselor IIC p < .70
6. Counselor IIIC p < .20
7. Me IM-IIIM p < .05*
8. Counselor IC—IIIC p < .20

 

53

three groups. Rows two and three are statements of the

tests of significance of the middle and terminal client
percepts of activity of self (IIM, p < .50; IIIM, p < .20).
The differences of the magnitudes between corresponding
scores (i.e., first, middle and terminal) between the success-
ful, partly successful, and unsuccessful groups for client-
perceived activity of self do not appear to be significant.

The fourth, fifth and sixth rows of Table 8
duplicate the comparisons of the first three rows with
respect to client's percept of counselor activity. The
results (IC, p < .30, IIC, p < .70; IIIC, p < .20) indi-
cate that the differences of the magnitudes between cor-
responding scores (i.e., first, middle and terminal)
between the successful, partly successful and unsuccess-
ful groups for client-perceived counselor activity do
not appear to be significant.

The seventh row states that the differences
between the client's first and terminal self-percept
activity scores (IM—IIIM) are significantly different
(p < .05) among the three groups of successful, partly
successful and unsuccessful counseling cases. It may
be inferred from this that the activity variable Oper-
ates differentially among the three groups. A closer
statiStical examination will be presented and discussed

later.

54

The eighth row of Table 8 indicates that there
is no statistically significant difference (p < .20) be-
tween the three groups regarding differences between first
and terminal counselor percepts (i.e., IC-IIIC) on the
activity factor. It does not therefore appear that per-
ceived counselor activity from before counseling to end
of counseling differs between successful, partly success-

ful and unsuccessful counseling cases.

Tabel 9. Mann-Whitney U Tests of Semantic Differential
activity factor scores.

 

 

 

Groups Compared Scores Compared Results
SU (IM-IIIM) p < .015*
PS-U (IM-IIIM) p > .05
S-PS (TM-IIIM) p > .05
S-U (IIC-IIIC) p < .268
PS-U (IIC-IIIC) p > .05

 

Table 9 is a tabulation of the results of a
closer statistical examination (the Mann-Whitney U Test)
of activity factor scores as they differ among the three
groups. The first row is a comparison of the first and
terminal client self-percepts (IM-IIIM) for differences
between the successful and unsuccessful groups (S-U) re-
garding activity factor scores. The result of p < .015

indicates that the difference between these two scores

55

is highly significant in the comparison of the successful
with the unsuccessful group. Therefore, it appears that
client-perceived change of client activity is different in
successful counseling cases as compared to unsuccessful
counseling cases. The perceived activity of self increases
from beginning to end of counseling for the successful cases
whereas it decreases for the unsuccessful group.

Rows two and three of Table 9 are a comparison of
these same scores (IM-IIIM) between different group pairs--
row two in respect to the partly successful and unsuccess-
ful groups (PS-U), row three in respect to the successful
and partly successful groups (S-PS). Neither of these
scores is significant (p > .05), indicating that the dif-
ferences between the first and terminal self-percept (on
the activity factor) are not different in respect to a
comparison of the partly successful with the unsuccessful
group, nor in respect to a comparison of the successful
with the partly successful group. Thus, client-perceived
changes in activity of self from before to end of counsel-
ing appear different in the successful group as compared
to the unsuccessful grOUp. However, there are no such
apparent differences in a comparison of the partly success-
ful with the unsuccessful group, nor in a comparison of
the successful with the partly successful group.

Rows four and five of Table 9 respectively com-
pare the middle and terminal counselor-percept activity

scores (IIC-IIIC) between the successful and unsuccessful

56

groups (S-U; p < .268), and between the partly successful
and unsuccessful groups (PS-U; p > .05). As indicated,
neither comparison resulted in statistical significance.

It does not therefore appear that client-perceived coun-
selor activity from the middle to the end of counseling
varies between successful, partly successful and unsuccess-

ful counseling cases.

Figure 5. Graphs of the evaluation factor scores for the
successful, partly successful and unsuccessful
counseling cases.

 

 

 

 

 

 

 

I II III I II III 'I II III
a. b. c.
Successful Group Partly Successful Unsuccessful
Group Group

 

Client's "me" percept
---------- Client's "counselor" percept

Figure 5 represents the scores of the evaluation
factor listed in Table 5. The statistical treatment of
these scores is presented similarly to those of the potency
and activity factors as described previously. Table 10 is

a tabulation of these data.

57

Table 10. Kruskal-Wallis one-way analysis of variance
Semantic Differential evaluation factor scores.

 

 

 

SD Concept Scores Compared Results
1. Me IM p < .70
2. Me IIM p < .50
3. Me IIIM p < .70
4. Counselor IC p < .30
5. Counselor IIC p < .10
6. Counselor IIIC p < .20
7. Counselor IC-IIIC p < .20
8. Me .
Counselor IC'IM p < '30
9. Me
Counselor IIC-IIM p < ’10
10. Me .
Counselor IIIC-IIIM . p < .30

 

The analyses of variance computed with the evalu-
ation factor scores are shown in Table 10. They are pre-
sented similarly to the manner in which the analyses of
variance were diScussed relative to the potency and activity
factor scores in Tables 6 and 8, respectively. The first
row states that the concept being considered is "me." It
is a statement of the test of significance (p < .70) of the
first client percept of self (IM) on the evaluative factor
relative to differences in the score between the three
groups. Rows two and three are statements of the tests of
significance of the middle and terminal client percepts of

.self on the evaluation factor, respectively (IIM, p < .70;

S8

IIIM, p < .70). The differences of the magnitudes between
corresponding scores (i.e., first,midd1e and terminal) be-
tween the successful, partly successful and unsuccessful
groups for client perception of self on the evaluation
factor do not appear to be significant.

The fourth, fifth and sixth rows of Table 10 du-
plicate the comparisons of the first three rows with re-
spect to the client's percept of counselor on the evaluation
factor. The results (IC, p < .30; IIC, p < .10; IIIC,

p < .20) indicate that the differences of the magnitudes
between corresponding scores (i.e., first, middle and ter-
minal) between the successful, partly successful and un-
successful groups for client perception of counselor on
the evaluation factor do not appear significant.

Row seven of Table 10 compares the first and
terminal clients' counselor-concept evaluation scores (IC-
IIIC) with respect to differences between the successful,
partly successful, and unsuccessful groups. The differ-
ences between these two scores between the three groups
are not significant (p < .20). Thus, it does not appear
that changes in clients' evaluation of counselor from
beginning to end of counseling is different with respect
to successful, partly successful or unsuccessful counseling.

Rows eight, nine and ten of Table 10 represent
comparisons of the differences between clients' self-percepts

and counselor percepts relative to the evaluation factor

59

between the three groups. These differences are compared
for the beginning, middle and terminal stages of counseling
respectively (IC-IM, p < .30; IIC-IIM, p < .10; IIIC-IIIM,

p < .30). Although there are no significant differences in
these comparisons the .10 probability of the middle score

is deemed sufficiently high to justify a closer statistical
examination (to be discussed and presented later). The re-
sults of the one-way analyses of variance appear to indicate
that there may be differences between the successful, partly
successful and unsuccessful groups in respect to client-

perceived evaluative differences between self and counselor.

Table 11. Mann-Whitney U Tests of Semantic Differential
evaluation factor scores.

 

 

 

Groups Compared Scores Compared Results
S-U (IIC-IIM) p < .063
PS-U (IIC-IIM) p < .05*
S-PS (IIC-IIM) p > .05
S-U (IIIC-IIC) p < .089
PS-U (IIIC-IIC) p < .025*

 

Table 11 summarizes the results of the Mann-
Whitney U tests applied to the evaluation factor data. The
table presents this information as described for the potency
and activity scores in Tables 7 and 9, respectively. The
first three rows in Table 9 respectively compare the success-

ful with the unsuccessful group (S-U), the partly successful

60

with the unsuccessful group (PS-U), and the successful Wlth
the partly successful (S-PS) for significance of the dif-
ferences between the client's "me" and "counselor" percepts
at the time of middle testing (IIC-IIM). These differences
are significant in the comparison between the partly success-
ful and unsuccessful grOUps (p < .05) and they approach sig-
nificance (p < .063) in the comparison between the success-
ful groups (S-U). They are not significantly different
between the successful and partly successful groups as shown
in row three of Table 9 (IIC-IIM, p > .05). These results
appear to indicate that the differences in clients' evalu-
ation of selves and counselors during the middle stages of
counseling is different for unsuccessful cases of counseling
as compared to successful or partly successful cases. That
is, the perceived disparity of self and counselor is much
greater in the case of unsuccessful cases as compared to
successful or partly successful cases.

The last two rows of Table 11 indicate the statis-
tical comparison of the difference between the final and
middle client concept of "counselor" on the evaluation
factor (IIIC-IIC). In the penultimate row the successful
and unsuccessful groups are compared (S-U). These two
groups are not considered statistically different (p < .089)
in terms of the difference between final and middle evalu-
ation scores of the client concept of "counselor." However,

the .089 level of confidence may be considered as approaching

61

significance. The final row indicates that there is a sta-
tistical difference in terms of these scores between the
partly successful and unsuccessful groups (PS-U, p < .025).
These results are interpreted to indicate that the changes
of a client's evaluation of counselor from the middle to

the terminal stage of counseling may differ for unsuccessful
cases as compared to partly successful or successful cases
Of counseling. For the successful and partly successful
cases counselor evaluation rises and becomes more divergent
from the client's self-evaluation whereas, for the unsuccess-
ful cases, the client's percept of counselor evaluation de-
creases and converges with his self-evaluation.

The third hypothesis viz., that the client's self-

 

percept will become more like the counselor's self-percept

 

in successful counseling cases as compared to partly success-

 

ful and unsuccessful cases, was tested by computing the D

 

scores between each client's percept "me" (i.e., for first,
middle and terminal Semantic Differential responses) and
the counselor's first and only percept "me." That is, each
of three client self-percepts was compared with the single
counselor self-percept. The mean D scores between the
clients' three percepts of "me" (i.e., for Semantic Differ-
ential responses I, II and III) and the counselor's percept
"me" are presented in Table 12. Figure 6 is the graphic

representation of these scores.

62

Table 12. Semantic Differential mean D scores between
client concept "Me" and counselor concept ”Me."

 

 

Time Administered

 

 

Group

I II III
Successful 6.4 8.6 7.8
Partly Successful 8.0 8.2 7.8
Unsuccessful 6.4 7.8 6.9

 

Figure 6. Graphs of D scores between clients' three per-
cepts Of "me" and the counselor's percept "me."

10

 

Successful Group
--------- Partly Successful Group
......... Unsuccessful Group

 

 

1
I II III

An inspection Of the data in Table 12 indicates
little difference between the respective scores of the
groups. Nevertheless, a one-way analysis of variance was
applied to the sums of ranks between scores I and II (H =
3.6, df = 2, p < .20) and I and II (H = 2.2, df = 2, p <
030). The results indicate that differences between these

scores (i.e., between I and II, and between I and III) are

63

not significant among the three groups. These results indi-
cate that there is no difference between successful, partly
successful and unsuccessful clients regarding change of self-
percept relative to counselor-percept over the course of

counseling. The third hypothesis is therefore rejected.

Summary

The Semantic Differential data of this study were
divided into three groups of scores representing the begin-
ning, middle, and terminal phases Of counseling. The first
hypothesis (i.e., predicted convergence of self-percept and
counselor-percept over the course of counseling for success-
ful counseling cases as compared to partly successful and
unsuccessful cases, was examined by comparing the D scores
for the client's percepts of "me” and "counselor" over the
three times. A one-way analysis of variance indicated no
significant differences between D-scores of the successful,
partly successful, and unsuccessful counseling cases.
These results failed to support the hypothesis that a client's
self-percept will become more like his counselor-percept in
successful cases of counseling, as compared to partly success-
ful and unsuccessful cases.

The second hypothesis (i.e., invariance of client-
perceived counselor potency over the course of counseling)
‘was examined through analyses of the raw potency scores.

(One-way analyses Of variance were first applied as tests of

64

significant differences among the three groups (i.e., success-
ful, partly successful and unsuccessful cases). Where the
results of these first analyses indicated significance or
near significance a closer statistical examination of the
potency scores was made by application of the Mann-Whitney
U test. No significant differences were found. Thus, the
hypothesis that a client's percept of counselor potency will
not change over the course of counseling for successful cases
of counseling, as compared to partly successful and unsuccess-
ful cases was not supported. Q

The third hypothesis (i.e., the predicted conver-
gence of client self-percept with counselor self-percept over
the course of counseling) was examined by comparing the total
scale scores of the client's percept "me" over the three
times (i.e., pre, middle and terminal phases of counseling)
with the one, and only one, counselor-percept "me." Tests
of one-way analysis of variance applied to the D scores of
these comparisons indicated no significant differences.
Thus, the hypothesis that a client's self-percept will be-
come more like the counselor's self-percept in successful
cases of counseling, as compared to partly successful and
unsuccessful cases, was not supported.

Additional analyses were made in order to examine
more extensively the data for any trends tending to either
support the convergence hypothesis or to more conclusively

refute it. One-way analyses of variance were performed on

65

the factor scores for activity and evaluation. The results
justified closer testing with the Mann-Whitney U test, re-
sulting in several significant differences among the three
groups of successful, partly successful and unsuccessful
cases. Specifically, there was a significant difference
between the successful and unsuccessful groups in the

change of the client's self-percept on the activity factor.
These results indicated that client-perceived activity in-
creased over the counseling period for the successful groups
whereas it decreased over this time period for the unsuccess-
ful group. Similar analyses were made on the evaluation
factor scores resulting in significance of differences be-
tween the clients' self-percept and counselor-percept between
the three groups. These differences indicated that, on the
evaluation factor, successful and partly successful clients
tend to see themselves as more similar to their counselors
whereas unsuccessful clients perceive themselves as less
similar during the middle phase Of counseling. The unsuccess-
ful clients manifested a tendency to bring their self-percept
and counselor-percept into convergence at the terminal phase
of counseling, whereas the successful group indicated diver-
gent trends of self-percept and counselor-percept at the end

of counseling.

V. DISCUSSION

Relation of the Results
to the Hypotheses,

 

Hypothesis 1: A client's self-percept will become more like
his counselor-percept in successful cases of
counseling, as compared to partly successful
and unsuccessful cases.

The results do not support this hypothesis in
terms of the comparisons made between the client's percept
"me" and the client's percept ”counselor" on the Semantic
Differential. In fact, graphs of the scores depicting these
comparisons (see Figure 2) are strikingly similar for the
three groups (i.e., the successful, partly successful, and
unsuccessful counseling cases). There were indications,
however, that the counselor percept remains relatively
stable over the course Of counseling in all three groups.
This suggests that the counselor may indeed be perceived
as a referent by all clients but that this referent is used
differentially. For example, there did appear to be con-
siderable variability of the self-percept over the period
Of counseling in all of the groups. However, the greatest
variability was manifested by the unsuccessful group. It
therefore appears that the degree of change of one's self—

percept during counseling may be more closely related to

66

67

counseling outcome than to the manner of change with re-
Spect to one's counselor-percept.

The negative results of this study relative to
the first hypothesis are in direct contradiction to the
Ewing study (1954) and the Sapolsky study (1965) mentioned
in Chapter II. These two studies, it may be recalled, in-
dicated that improvement of clients in therapy was associ-
ated with a change of the self-concept to become more like
the perception of the counselor. The patent question is,
why do the results of the present study differ from the
independent findings of Ewing and Sapolsky? One reason
for the differences may be that different outcome criteria
with respect to client improvement were utilized. Ewing
used the counselor's "estimate of client improvement;"
Sapolsky used the judgment of supervisory staff psychia-
trists as the criterion of improvement. In the present
study, judgments of change of MMPI scores from before to
end of counseling were utilized as measures of outcome.
However, even had similar outcome criteria been used, the
measures of perceived client-counselor similarity were
different in the Ewing study. Ewing used a 100 item de-
scription of traits as his instrument while the Semantic
Differential was used in this study. Sapolsky, on the
other hand, also used the Semantic Differential as a
measure of similarity and although his scales did not cor-

reSpond with the scales used in the present study, the

68

three factors--potency, activity and evaluation--are indige-
nous to the Semantic Differential and therefore comparable.
Thus, the differences between the present study and the
Sapolsky study may best be explained by the different out-
come criterion used in each. It would seem that direct
comparison of such studies of similarity could be properly
made only if identical methods and instruments were used
to assess the similarity variable. Of these criticisms,

it seems to this writer that the failure to use multiple
outcome criteria in any studies of this type (including

the present study) is a major objection. Counseling out-
come is not only difficult to measure but difficult to
define. Accordingly, this writer agrees with, and en-
dorses the statement made by Goodstein and Grigg regarding
evaluative criteria of counseling outcome or effectiveness.
”Any completely satisfactory evaluation should involve
multiple criterion measures, including evaluations of
personal and social adjustment by the client and others,
actual performance records, self-concept and other person-
ality dimensions, and client satisfaction with the counsel-
ing process." (1959, p. 19).

Differences between studies may also be a function
of the instruments used and the level of training and rela-
tive competence of the counselors. In the present study a
preponderence of counseling trainees (i.e., interns and
practicum students) were used. These differences may sig-

nificantly influence the outcome.

69

Hypothesis 2: A client's percept of counselor potency will
not change over the term of counseling in
successful cases of counseling as much as it
does in partly successful and unsuccessful
cases.

The results do not support this hypothesis, for
client-perceived counselor potency appears to be invariant
for all three groups over the course of counseling. The
results (although not statistically significant) suggest,
however, that client-perceived differences between potency
of self and potency of counselor are different in several
respects for the unsuccessful cases as compared with the
successful and partly successful cases. Initially, for
example, the unsuccessful cases perceived themselves as
more potent than their counselors, and although these
positions were reversed at the middle stage of counseling
(i.e., client-perceived counselor potency was elevated
above client-perceived potency of self) the terminal scores
were reversed again, with the client perception of counselor
potency lower than the perceived potency of self. These re-
sults suggest that successful counseling outcome may be as-
sociated with client-perceived potency of counselor. That is,
the client may need to perceive his counselor to some degree
as more potent than himself in order to "improve" in coun-

seling.

70

Hypothesis 3: A client's self-percept will become more

like the counselor's self—percept in success-

ful cases of counseling, as compared to

partly successful and unsuccessful cases.

This hypothesis was not sustained by the results.

Little, if any difference, was apparent among the three
groups in respect to the relative changes between the client's
self-percept and the counselor's self—percept over the course
of counseling. That there were no differences suggests that
the groups may not have been validly differentiated. That
is, the outcome criterion used to categorize the clients
into successful, partly successful and unsuccessful cases

may not have been sufficiently or accurately discriminating.

Some Theoretical Implications

 

The different results of studies dealing with
client-perceived counselor similarity as related to coun-
seling outcome appear to make this an equivocal issue.
Evidence that the client may use the counselor as a refer-
ent or model for change was not found in the present study.
Accordingly, rather than serving as a model or referent
for change, the counselor may instead, in some manner, pro-
vide the milieu and conditions by which the client can
achieve psychological or behavioral change.

One of the generally accepted criteria of success-
ful counseling outcome is improved or increased client auton-

omy. By definition autonomy does not mean dependence upon

71

or mimesis of Others. Thus, one may argue that in success-
ful cases of counseling the client should perceive himself
as more different from, rather than more similar to his
counselor. This notion receives considerable support from
Parson's (1961) study which was concerned with the "limiting
or molding" influence of the counseling relationship upon
the client. The results indicated that counselors judged
as most adjusted and most competent by their colleagues
were the least likely to have their clients resemble them
after therapy. This implies that clients of competent
counselors achieve a degree of autonomy rather than be-
coming identified with their counselors.

Results of the present study lend some support
to the notion of client-counselor differentiation rather than
client-counselor similarity as a function of counseling
success. For example, it was noted that client-perceived
activity of self, and client perceived activity of counselor
suffered a somewhat parallel decline over the course of
counseling for the unsuccessful group. In the successful
grOUp, client-perceived counselor activity decreased while
perceived activity of the self increased during counseling.
Thus, in terms of perceived activity with respect to self
and counselor there was similarity and degeneration mani-
fested in the unsuccessful group. In other words, the
client perceived both himself and his counselor as becoming

progressively less active as counseling proceeded--implying

72

an attenuation of client-counselor interaction and resulting
in the ultimate degeneration of the relationship. For the
successful group, on the other hand, perceived client ac-
tivity increased, while perceived counselor activity de-
creased over the course of counseling. The implication
here is that a successful counseling relationship is char-
acterized by increased client participation (i.e., activity)
and involvement in counseling. The accompanying decrease

in perceived counselor activity may be a manifestation of

a client's increasing differentiation of himself from his
counselor.

An equally significant difference was evidenced
between the unsuccessful and successful groups on the evalu-
ation factor scores. During the middle stage of counseling,
for example, the unsuccessful group displayed very disparate
percepts of self and counselor--indicative Of client-perceived
dissimilarity with counselor. At the end of counseling, how-
ever, these two percepts converged markedly in the unsuccess-
ful cases, but diverged for the successful and partly success-
ful cases. Again, a reasonable interpretation may be that
successful clients perceive themselves as becoming different
from their counselors, perhaps more autonomous, as they ap-
proach successful termination. On the other hand, the un-
successful clients perceive their counselors as being more

like themselves--i.e., less differentiated.

73

In conclusion, the results of this study tend to
indicate that client improvement in counseling is associated
with client perceived dissimilarity with counselor rather
than similarity, as hypothesized.

Finally, client and counselor will come to a full
recognition, separately in many ways perhaps, that
they have meant and will continue to have importance
to each other. They can agree to separate with
awareness in each person that they are commonly
human. Yet each of them is unique since genuine
sharing has led to autonomy. (Kell-Mueller; 1966,
p. 144)

Implications for Future Research

 

To the writer two aspects of this study pose in-
triguing questions. The first Of these questions is related
to the client's perception of counselor potency. In the
present study the relationship of client-perceived potency
of self with that of client-perceived potency of counselor
was graphically (see Figure 3) very different for the un-
successful group as compared to the successful and partly
successful groups. Although the difference was not found
to be statistically significant there appears to be suf-
ficient indication to justify further exploration of this
variable as it related to counseling outcome. If indeed,
the disparity between perceived potency of self and per-
ceived potency of counselor is related to counseling pro-
cess and outcome, it could serve as a useful predictor
variable at the outset of counseling and permit the impli-

mentation of measures that would improve the probability

74

of counseling success. A larger sample than that used in
the present study in addition to empirical establishment

of a larger number of potency scales (N.B., only two scales
of the Semantic Differential used in the present research
represented the potency factor) could be a significant im-
provement in researching this question.

The second question of interest is in respect to
the evaluation factor. Would replication yield results
similar to those in the present study? Specifically, are
successfully terminated clients different from unsuccess-
fully terminated clients in respect to their evaluation of
themselves and their counselors? The convergence of self
and counselor (i.e., percept of self and counselor) for the
unsuccessful group at termination is in marked contrast to
the divergent trend manifested in the successful and partly
successful groups. Replicatory confirmation would lend
credence to the notion of client-perceived self-counselor
dissimilarity as being positively related to successful out-
come and thus support the thesis that the client's achieve-
ment of autonomy is a meaningful goal of counseling.

Finally, multiple outcome criteria, as mentioned
earlier, would permit a more acceptable and defensible cate-
gorization of unsuccessful and successful counseling cases.
Included in these criteria it would seem valuable to use

client evaluation of assessment of outcome.

75

Summary

This study was conceived to examine the client's
percepts of himself and percepts of his counselor as they
relate to counseling process and outcome. Of particular
research interest were the aspects of client-perceived
counselor similarity and client-perceived counselor potency
as they Operate during the counseling process, and how they
relate to counseling outcome. Previous studies have indi-
cated that successful counseling outcome is accompanied by
a shift in the client's self-percept. The rationale for
the present study was based upon the findings of such
"similarity" studies. If, for example, a client's improve-
ment during counseling involves a shift or a change in his
self-percept toward that, or more like his counselor percept,
it implies use of the counselor as a model. Since change is
relative and the counseling situation is the medium through
which change is accomplished it seemed reasonable to assume
that the counselor, in addition to providing the psycho-
logical context necessary as a condition for change, might
afford the client a relatively stable frame of reference
with respect to change. Thus, it was postualted that the
.client's perception of his counselor would necessarily be
stable during the course Of successful counseling. Further,
it was postulated that a client's change in self-percept
would be directional in reSpect to his counselor-percept,

i.e., it was hypothesized that a client's self-percept

76

becomes more like his counselor percept over the course of
successful counseling.

The concept of counselor stability suggested a
second hypothesis--that client-perceived counselor potency
remains invariant over the course of successful counseling.
It was reasoned that, in a helping relationship such as
counseling, client-perceived counselor potency (i.e., com-
petency) would be a necessary condition for client change
in that perceived counselor inadequacy would result in an
ineffectual relationship.

The third exploratory hypothesis compared change
of client self-percept with respect to counselor self-percept.
It stated that a client's self—percept becomes more like his
counselor's self-percept in successful counseling cases.

The study was designed in a manner that tested the
hypotheses in respect to three groups of counseling cases
viz., a successful group, a partly successful group and an
unsuccessful groups.

Twenty-two client-counselor pairs from the Michigan
State University Counseling Center were used in the sample.
The profiles of client responses to the MMPI before and after
counseling were used as the criteria for counseling outcome
(i.e., successful, partly successful or unsuccessful).

Client responses to the concepts "me" and "counselor" on the
Semantic Differential for pre, middle, and terminal phases

of counseling were utilized as the measure of perceived

77

client-counselor similarity, and client perception of coun-
selor potency. Pre-counseling Semantic Differential re-
sponses to the concept "me" were also obtained for the
counselors.

Analysis of the data involved the use of the
Kruskal-Wallis one-way analysis of variance and the Mann—
Whitney U Test, as tests of significance.

The results did not support the hypotheses re-
lating successful counseling outcome to increased client
perception of Similarity to counselor. The evidence
seemed to indicate, rather, that improvement was associ-
ated with increased client perception or differentiation
of self and counselor. This client percpetion of dis-
similarity was discussed in terms of autonomy aS‘a pur-
pose and function of counseling.

The hypothesis that client perceived counselor
potency remains invariant over the course of successful
counseling was not supported by the results. However,
the relative positions and disparity of client— perceived
potency of self and potency of counselor was indicated as
a possible relational variable to counseling outcome.
That is, the data suggested that unsuccessful counseling
outcome was characterized by a client's perception of self
potency as being equal to, or greater than, the perceived

potency of the counselor.

BIBLIOGRAPHY

Ashby, J. D., Ford, D. H” Guerney,B. 6., Jr., and Guerney,
Louise F. Effects on clients of a reflective
and a leading type of psychotherapy. Psychol.
Monogr., 1957, ll (24, Whole No. 453).

Axelrod, J. An evaluation Of the effect on progress in
therapy of simiIaritieS and differences—between
the personalities of patients and their thera-
pists. Unpublished doctoral dissertation, New

or niversity, 1951.

Bandura, A. and Walters, R. H. Social learning and per-
sonality development. New York: ’HOIt, I964?

 

Borrelli, N. J. The client's perception of therapist
potency andichanges in_psychotherapy. Unpub-
lishéd doctoral dissertatibn, Michigan State
University, 1965.

Brams, J. M. Counselor characteristics and effective com-
munication in counseling. J. counsel. P§ych01.,
1961, 8, 25-30.

 

Carkhuff, R. R. and Truax, C. B. The effects of lay group
counseling. J. consult. Psychol., 1965, 29,
426-431.

Carson, C. D. and Heine, R. W. Similarity and success in
therapeutic dyads. J. consult. Psychol., 1962,
£9- (1), 38-430

 

Cartwright, D. S. and Cartwright, Rosalind D. Faith and
improvement in psychotherapy. J. counsel. Psychol.
1958, 5, 174-177.

Dahlstrom, W. G. and Welsh, G. S. An MMPI handbook: A
guide to use in clinical practice and'researEh.
Minneapolis: University ofiMinnesotaiPress{7I960.

 

, and Welsh, G. S. Basic readings on the MMPI in
psychology and medicine. Minneapolis: *University
of Minnesota Press, 1956.

 

Ewing, Thomas N. Changes in attitude during counseling.
J. counsel. Psychol., 1954, l, (4), 232-239.

\

78

79

Farson, R. E. Introjection in the psychotherapeutic relation-
ship. J. counsel. Psychol., 1961, 8, 337-342.

 

Feifel, H. and Eells, Janet. Patients and therapists assess
the same psychotherapy. Paper read at American
Psychological Association, St. Louis, September,
1962.

Fiedler, F. E. The concept of an ideal therapeutic rela-
tionship. J. consult. Psychol., 1950, 14, 239-
245.

 

A method of objective quantification of certain
countertransference attitudes. J. clin. Psychol.,
1951, 1, 101-107.

 

, and Senior, K. An exploratory study of unconscious
feeling reactions in 15 patient-therapist pairs.
J. abnorm. soc. Psychol., 1952, 41, 446-453.

 

Gardner G. G. The psychotherapeutic relationship. Psychol.
’ Bull., 1964, g; (6), 426-437.

Gorlow, L., Hoch, E. L., and Telschow, E. F. The nature of
non-directive group psychotherapy: An experimen-
taliinvestigation. NeWSYOrk: Teachers COllege,
Columbia University, 1952.

 

 

Grigg, A. E. and Goodstein, L. D. The use of clients as
judges of the counselor's performance. J. counsel.
Psychol., 1957, 4, 31-36.

 

Holt, R. R. and Luborsky, L. Research in the selection of
psychiatrists: A second interim report. Bull.
Menninger Clin., 1952, 16, 125-135.

 

Hunt, J. Mc V., Ewing, T. N., La Forge, R. and Gilbert, W.
M. An integrated approach to research on thera-
peutic counseling with samples of results. J.
counsel. Psychol., 1959, 6, 46-54.

 

Kell, B. L. and Mueller, W. J. Impact and change: A study
of counseling relationships. Neinork: Appleton-
Century-Crofts, 1966.

 

Landfeld, A. W. and Nawas, M. M. Psychotherapeutic improve-
ment as a function of communication and adOption
of therapist's values. J. consel. Psychol., 1964,
ll (4), 336-341.

 

Levinson, D. J. The psychotherapist's contribution to the
patient's treatment career. Paper read at Confer-
ence on Psychotherapy Research, Cahpel Hill, North
Carolina, 1961.

80

Lipkin, S. The client evaluates nondirective psychotherpay.
J. consult. Psychol., 1948, 22, 137-146.

 

Meader, A. Uber die zwischenmeashliche Beziehung in der
Psychotherapie. (Interpersonal relationships in
psychotherapy). Acta.ppsychother. Psychosom.

Ortho aeda o ., 1955, 2, 242-500. Abstract 2971,
4674, V01. 3%, 1956.

Mendelsohn, G. A. and Geller, M. H. Structure of client
attitudes toward counseling and their relation to

client-counselor similarity. J. counsult. Psychol.,
1965, 29, (1), 63-72.

 

 

Osgood, C. E., Suci, G. J. and Tannenbaum, P. H. The measure-
ment of meaning. Urbana, Illinois: Univer51ty of
Illinois Press, 1957.

 

 

Ourth, L. L. The relationship of similarity in therapist-
client pairs to client's stay and improvement in
psychotherapy. Disser. Abstr., 1964, 24(a),
3839-3840.

 

Rogers, C. R. Client-centered therapy. Boston: Houghton-
Mifflin, 1951.

 

The characteristics of a helping relationship.
Personnel 8 Guid. J., 1952, 22, 6-16.

 

An overview of the research and some questions
for the future. In C. R. Rogers and Rosalind F.
Dymond (Eds.), Psychotherppy and personality
change. Chicago: *University or Chitago Press,
1954, 413-430.

 

. The necessary and sufficient conditions of thera-
peutic personality change. J. consult. Psychol.,
1957, 22 95-103.

 

Therapist-patient relationship. In E. A. Rubin-
stein and M. B. Parloff (Eds.), Researchgin psycho-
therapy. Washington, D. C.: National PubliShing,
1959, 273.

 

Rosen, S., Levinger, G., and Lippet, R. Perceived sources
of social power. J. abnorm. soc. Psychol., 1961,
92, 439-441.

 

Rosenthal, D. Changes in some moral values following psy-
chotherapy. J. consult. Psychol., 1955, 29, 431-
436.

 

Sapolsky,

Siegel, S.

Schopler,

Seeman, J.

Shoben, E.
Snyder, W.
Stefflre,

Stransky,

81

and Frank, J. D. Psychotherapy and the placebo
effect. Psychol Bull., 1956, 99, 294-302.

 

A. Relationship between patient-doctor compati-
bility, mutual perception and outcome of treatment.
J. abnorm. Psychol., 1965, 29 (1), 70-76.

 

Nonparametric statistics for the behavioral
sc1ences. New York: McGraw Hill, 1956.

 

J. H. The relation of patient-therapist personality
similarity to the outcome Ofipsychotherapy. Unpub-
lished doctoral dissertation, University of
Colorado, 1958.

 

 

Counselor judgments of therapeutic process and
outcome. In C. R. Rogers and Rosalind F. Dymond
(Eds.),.Psychotherapy and personalipy changg.
Chicago: University of ChicagoiPress, 1954, 99-
108.

 

J., Jr. Psychotherapy as a problem in learning
theory. Psychol. Bull., 1949, 44, 366-392.

 

U. The psychotherpay relationship. New York:
Macmillan, 1961.

 

 

B. Theories of counseling. New York: McGraw
Hill, 1965.

E. Autoritat a Subordination in der psycho-
therapie der verschiedenen Lefensphasen.
(Authority and subordination in the psychotherapy
of the various phases of life.) Wien: Klin
Wschr., 1946, 99, 39-41. Abstract 4027, 1947.

Streitfeld, J. W. Expressed acceptance of self and others

by psychotherapists. J. consult. Psychol., 1959,
‘29, 435-441.

 

Tuma, A. H. and Gustad, J. W. The effects of client and

counselor personality characteristics on client
learning in counseling. J. counsel. Psychol.,
1957, 4, 136-141.

 

large

thin
colorless
easy

safe

sharp
Optimistic
free

fair
active

bad
destructive
Slow

changing

APPENDIX

The Fourteen Scales of the Semantic
Differential Used in This Study

 

 

 

 

O.

 

 

 

 

 

 

 

 

 

 

82

small

thick
colorful
difficult
dangerous
dull
pessimisitc
constrained
unfair
passive
good
productive
fast

stable

 

Factors
potency
potency
activity
evaluation
evaluation
activity
evaluation
evaluation
evaluation
activity
evaluation
evaluation
activity

evaluation

"I71111111111111All“