Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Therapeutic Movement As A Function Of Awareness Of Goals
(USC Thesis Other)
Therapeutic Movement As A Function Of Awareness Of Goals
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THERAPEUTIC MOVEMENT AS A FUNCTION OF
AWARENESS OF GOALS
by
Marion Crumley Luenberger
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
CEducation)
August 19 72
INFORMATION TO USERS
This dissertation was produced from a microfilm copy of the original document.
While the most advanced technological means to photograph and reproduce this
document have been used, the quality is heavily dependent upon the quality of
the original submitted.
The following explanation of techniques is provided to help you understand
markings or patterns which may appear on this reproduction.
1. The sign or "target" for pages apparently lacking from the document
photographed is "Missing Page(s)". If it was possible to obtain the
missing page(s) or section, they are spliced into the film along with
adjacent pages. This may have necessitated cutting thru an image and
duplicating adjacent pages to insure you complete continuity.
2. When an image on the film is obliterated with a large round black
mark, it is an indication that the photographer suspected that the
copy may have moved during exposure and thus cause a blurred
image. You will find a good image of the page in the adjacent frame.
3. When a map, drawing or chart, etc., was part of the material being
photographed the photographer followed a definite method in
"sectioning" the material. It is customary to begin photoing at the
upper left hand corner of a large sheet and to continue photoing from
left to right in equal sections with a small overlap. If necessary,
sectioning is continued again — beginning below the first row and
continuing on until complete.
4. The majority of users indicate that the textual content is of greatest
value, however, a somewhat higher quality reproduction could be
made from "photographs" if essential to the understanding of the
dissertation. Silver prints of "photographs" may be ordered at
additional charge by writing the Order Department, giving the catalog
number, title, author and specific pages you wish reproduced.
University Microfilms
300 North Zeeb Road
Ann Arbor, Michigan 48106
A Xerox Education Company
I
I
73-747
LUENBERGER, Marion Crumley, 1915-
THERAPEUTIC MOVEMENT AS A FUNCTION OF
AWARENESS OF GOALS.
University of Southern California, Ph.D., 1972
Education, guidance and counseling
University Microfilms, A XEROX Company, Ann Arbor, Michigan
THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED.
U N IV E R S IT Y O F S O U T H E R N C A L IF O R N IA
THE GRADUATE SCHOOL
U N IVER SITY PARK
LOS ANGELES. C A LIFO R N IA 9 0 0 0 7
This dissertation, written by
M-AR-I-ON—C-RUMLEY—LUEN^ERGE-R......
under the direction of Aer—- Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by The Graduate
School, in partial fulfillm ent of requirements of
the degree of
D O C T O R O F P H IL O S O P H Y
Dean
Date.
DISSERTATION COB
Chairman
PLEASE NOTE:
Some pages may have
indistinct print.
Filmed as received.
University Microfilms, A Xerox Education Company
ACKNOWLEDGMENTS
I wish to acknowledge my gratitude to the many indi
viduals whose helpful cooperation made this research study
possible. I am indebted to the following professional
therapists for their interest, support, commitment of time,
and faithful carrying out of the research design: Gerda
Friend, M.S.S.S.; Drs. Allen P. Webb, Theodore M. Johnson,
A. D. Hart, and James V. Van Camp; and fellow doctoral
candidates LaWanda Katzman and Geraldine Pickart. To
them I extend deeply felt appreciation.
I am sincerely grateful to the members of my Disser
tation Committee for their helpful advice and assistance;
to my chairman, Dr. William Ofman for his many thoughtful
statements, stimulating suggestions, and confidence in me;
to Dr. Allen Webb, whose service as both a committee mem
ber and a therapist was a continual inspiration to me; and
to Dr. Milton Wolpin for his ready availability and good-
natured encouragement.
And finally I want to acknowledge gratitude to my
friends, colleagues, and family for their interest and
support, and especially to express heartfelt appreciation
to my husband, Fred, for xeroxing the hundreds of neces
sary pages and for his patience and willing cooperation
throughout my involvement in this study.
TABLE OF CONTENTS
Acknowledgments ....................................
List of Tables......................................
CHAPTER
I. INTRODUCTION ................................
The Study
Statement of the Problem
Hypotheses
Limitations in the Design
Delimitations
Definition of Terms
Organization of the Remainder of the Study
II. REVIEW OF THE LITERATURE ...................
Goals in Psychotherapy
Categorizations of Therapeutic Systems
and Goals
Congruence of Patient-Therapist Values
Motivation
Summary
III. THE METHOD ..................................
Description of the Sample Groups
Description of the Experimenters
The Experimental Period
Procedures of the Study
The Instruments
Development of the Lists of Goal Statements
Treatment of the Data
IV. THE FINDINGS ...............................
Experimental Findings
Additional Findings
V. SUMMARY, IMPLICATIONS, CONCLUSIONS AND
RECOMMENDATIONS.............................
Summary of the Method
Summary of the Findings
Implications and Conclusions
Recommendations iii
ADDENDUM.......................................... 134
BIBLIOGRAPHY ...................................... 145
APPENDICES........................................ 155
iv
LIST OF TABLES
Table Page
1. Number of Experimental and Control Subjects
for Each Experimenter.......................5 7
2. Sex and Age Range Distribution of Subjects . . 59
3. Educational Level of Subjects....................59
4. Concepts Measured by the Semantic Differential 77
5. Bi-polar Adjective Scales of the Semantic
Differential .................................. 77
6. Distribution of Judges' Categorizations of
Goal Statements.................................. 82
7. Pretest Personal Orientation Inventory I Scale
Means and Standard Deviations for Experimental
and Control Groups..............................86
8. Pretest Multiple Affect Adjective Check List
(Today Form) Means and Standard Deviations
for Experimental and Control Groups............. 87
9. Pretest Multiple Affect Adjective Check List
(General Form) Means and Standard Deviations
for Experimental and Control Groups............. 88
10. Pretest Semantic Differential Means and
Standard Deviations for Experimental and
Control Groups............................... 88
11. Comparison of Gain Scores for Experimental
Versus Control Groups on Inner-Direction
Scales of the Personal Orientation Inventory . 90
12. Comparison of Gain Scores for Experimental
Versus Control Groups on Today Form of the
Multiple Affect Adjective Check List ........ 91
13. Comparison of Gain Scores for Experimental
Versus Control Groups on General Form of the
Multiple Affect Adjective Check List ........ 91
v
14. Comparison of Gain Scores of Experimental
Versus Control Groups on the Semantic
Differential ............................... 93
.15. Comparison of Experimental and Control
Groups on Attainment of Experimenter's
Goal: Statistical and Descriptive ........ 94
16. Comparison of Experimental and Control
Groups on Movement Toward Goal: Statistical
and Descriptive............................. 9 5
17. Number of Value Shifts of Experimental and
Control Subjects Toward and Away from their
Therapist's Values as Measured by a Semantic
Differential ................................ 97
18. Number of Value Shifts of Total Experimental
and Control Groups Toward and Away from Ther
apists ' Values as Measured by a Semantic
Differential ................................ 98
19. Relationship of Experimenters' Orientation to
Goals Selected for Experimental and Control
Groups...................................... 100
20. Relationship of Experimental and Control
Groups' Goal Choices to the Goal Choice
Orientation of Their Experimenters ........ 10 2
21. Means and Standard Deviations of Me and My
Ideal Self Concepts......................... 105
22. Relationship of Therapists' Orientation to
Gain or Loss on the Personal Orientation
Inventory.................................... 10 7
23. Relationship of Subjects' Goal Choice to
Gain or Loss on the Personal Orientation
Inventory.................................... 10 8
24. Relationship of Therapists' Goal Choice to
Gain or Loss on the Personal Orientation
Inventory.................................... 108
25. Relationship of Therapists' and Subjects'
Mutual Goal Choices to Gain or Loss on the
Personal Orientation Inventory ............ 109
vi
26. Relationship of Experimental Subjects' and
Control Subjects' Choice of an Existential
Goal to Gain or Loss on the P OI.......... 109
27. Relationship of Therapists' Choice of an
Existential Goal for Experimental and Control
Subjects to Gain or Loss on the POI .... 110
28. Relationship of Mutually Chosen Existential
Goals to Gain or Loss on the POI between
Experimental and Control Groups .......... 110
29. Relationship of Therapists’ Orientation to
Gain or Loss on the General Anxiety Scale . Ill
30. Behavioristic Goal Choices................. 113
vii
CHAPTER I
INTRODUCTION
The ultimate goal of all psychotherapeutic treatment
or counseling is some form of improvement in a person, be
it freedom from distressing emotions, more effective beha
vior, or self-awareness. Goals such as these are central
in the psychotherapeutic process, yet, because of their
multiplicity and the very obviousness of their function,
they have historically been merely implicit global goals
inherent in the theory and process espoused by the thera
pist. Generally, they have not been explicitly defined by
the therapist or for or by the patient. Particularly, pa
tients have not been asked to write out a specific goal.
There seem to be two reasons for this emphasis upon theore
tical, global goals.
First, both public and research interest has focused
on the inner workings of the process of therapy--on what
takes place in the therapeutic session. The meaningfulness
of studies on the outcome of psychotherapy, on the other
hand, has frequently been challenged. In recent years the
quantity and quality of outcome research studies have in
creased, but, nevertheless, "goals of therapy have received
2
scant attention from researchers" (Meltzoff and Kornreich,
1970, p. 172).
Secondly, psychotherapy has been strongly rooted in
and for most of its short history has been saddled with the
disease concept. The patient was considered "mentally ill"
and the goal of psychotherapeutic treatment was for him to
become "well.” The patient presented himself to the psy
chologist or psychiatrist as to a physician, requesting
only that he be cured of his illness and suffering--a pain
which somehow was imposed upon him from without and which
in no way implied his responsibility.
Strong criticisms have been raised in recent years
against the medical model of which one of the most vocal
contributors has been Szasz (1961) . As a result, more em
phasis is being placed on how to achieve "mental health"
than on how to cure "mental illness." An exact definition
of "mental Health," however, continues to elude theoreti
cians. Jahoda (1958), in her survey of the literature found
nine broad concepts of ideal behavior which could be
equally valuable in defining "positive mental health."
This change of emphasis and the growing body of know
ledge from practical, theoretical, and research experience
in psychotherapy have broadened the base for the applica
tion of the field. An outstanding example of this is the
changing role-concept of the counselor in the schools.
3
Counseling and psychotherapy have come to be much more syn
onymous terms for a process as their implicit goals have
become more analogous.
As the new theories or systems in the field of psy
chology have emerged, there has been a shift in and a large
addition to the number of implicit goals of therapy. The
basic differences in these approaches are, as Patterson
(1966) has pointed out, the differences in each theorist’s
view of the nature of man. The implicit goals of each
have been recognized as a reflection of the founder’s value
system (Glad, 1959; Lowe, 1969). Differing goals and
values then permeate the various therapeutic methods and
are held by individual therapists. Each patient, too, has
his unique set of goals and values, and these may differ
from those of his therapist or counselor.
Because of the paucity of research in this area, this
study was an attempt to ascertain the function of a pa
tient's goal, explicitly delineated in writing, in thera
peutic movement, its relationship to the patient's value
system and to the therapist's goal and value system.
The Study
The Importance of the Study
The focus on goals in this investigation would appear
to be important from several aspects. Among these are:
4
1. The difficulty in defining and measuring
success in psychotherapy.
2. The continuing search for effective motiva
tion for psychotherapeutic movement.
3. The increasing numbers of persons who seek the
services of those in the helping professions.
The problem of defining "success" in psychotherapy is
a very difficult one and has plagued those engaged in it,
especially since Eysenck's (1953, 1960) bitter and contro
versial statement in which he claimed that research studies
showed the effects of psychotherapy to be extremely small
and, in comparison to non-treatment improvements, not
worthwhile. Therapists and counselors are intuitively
aware of their successes and failures but most often do not
or are unable to present empirical evidence. When evidence
is presented, it is still subject to the reviewers' ap
praisal, usually with conflicting results. An example of
this is that Truax and Carkhuff's (1967) conclusion after
surveying the research was an echo of Eysenck's, but Melt
zoff and Kornreich (1970) refute that conclusion, case by
case.
Part of the problem seems to be the fuzziness in
thinking about goals and the confusion which sometimes ex
ists between goals and process. A written goal would do
much to clarify thinking. If the patient writes out his
goal, the therapist is free to evaluate it in terms of his
perception of the problem, but he also becomes explicitly,
5
rather than intuitively, aware of where the patient is at
one particular moment in time. Furthermore, he may note im
mediately the similarity or the discrepancy between his
goal for the patient and the patient’s goal for himself.
Therapy is a problem-solving process and generally
proceeds through three steps: the presentation of a problem,
intervention methods used for solving it, and hopefully,
the resolution, which is then referred to as the realized
goal. If the goal were recognized for its true function at
the beginning of therapy as the outcome to be sought, then
objective instruments and observational techniques could be
employed to evaluate the outcome as well as direct the in
tervention into proven methods (Urban and Ford, 1971).
There are many theories of motivation : drive reduc
tion, incentive, achievement, growth potential, among them.
But studies on motivation in therapy have shown that this
is an elusive variable with initial high and low motivations
having little relation to outcome (Meltzoff and Kornreich,
1970). This investigation may point in the direction of
showing that giving a person responsibility for his goal
choices and continuing his awareness of them by continual
referral to the written goals will increase his motivation
to refine and to attain these goals.
The increasing number of persons who seek the ser
vices of those in the helping professions will necessitate
6
more efficient expenditure of their time and effort. The
hypothesized motivation of written goals may expedite the
process. Insight into the importance of the relationship
between the goals of the person seeking help and the goals
and values of the helper may also expedite the process or
aid in the placement of patients.
Statement of the Problem
The thesis of the present investigation is that it
is important that the patient specify in writing an expli
cit goal and be continually aware of his goal, his move
ment toward it and any changes or additions he may make in
it. Moreover, it is equally important that the therapist
specify and be aware of his goal for his patient at the
beginning of therapy.
The problem examined in this study was to determine
the degree of change, as indicated by measures on objective
instruments and reports of behavioral differences, which
took place in individual therapeutic and counseling pa
tients when they specified their desired outcome goals.
This specification consisted of their choosing their goals,
habitually referring to written statements of these, and
having the opportunity to revise these goals during the
experimental period.
This study also determined whether or not these
changes were related to a tendency for patients1 values
and goals to become congruent with the therapist's own
values and personality theory goal. It was further deter
mined whether or not such a relationship was necessary for
either the method of specified or nonspecified patient
goals to be effective.
Hypotheses
The hypotheses to be investigated were based on the
expected benefits of goal-setting in psychotherapy. They
were conceptualized as follows:
1. Subjects who write out and habitually become aware
of the personalized goal or goals they choose to
attain from individual psychotherapeutic treatment
will show more positive change than subjects who do
not write out and habitually become aware of such a
goal or goals on the following dependent variables:
1.1 Self-actualization
1.2 Anxiety, depression and hostility
1.3 Evaluative meanings of concepts
1.4 Verbally reported overt behavior change,
judged as to movement towards the goal
2. There will be a greater tendency for the values of
those subjects participating in the goal-setting
process to become congruent with the values of their
respective therapists than for the values of those
not participating in the goal-setting process to
become congruent with the values of their respective
8
therapists.
3. Subjects participating in the goal-setting process
will more often indicate at the conclusion of the
experimental period a preference for a general goal
which is congruent with the general goal selected
for each at the beginning of the experimental period
by their therapist than will subjects not partici
pating in the goal-setting process.
4. Subjects who regularly have the opportunity to re
vise their written goals will do so at least once.
Assumptions
For the purpose of this study it was basically as
sumed that therapy can bring about improvement or growth
in persons and that positive changes on the criterion mea
sures used in this study are indicative of such movement.
Further, it was assumed that the measuring devices used
would accurately measure the degree of change without re
gard to age, sex, and maturity. The subjects of this re
search would constitute a representative sample of the per
sons seeking individual therapeutic treatment or counseling.
Further, the following assumptions were made regard
ing the therapists: (1) They would follow the procedure
of the research design in a standardized manner; (2) aside
from the experimental variables, therapy would proceed ac
cording to the individual style of the therapist and in as
9
similar manner for each subject of the therapist as the
individual case permitted; (3) the value orientation of a
therapist has influenced his choice of personality theory;
his values will remain the same during the experimental
period, and his value system is an influence on the value
system of his patients.
Finally, it was assumed that the categorization of
personality theories into three broad types for this re
search was an adequate categorization to accommodate all
practicing professionals.
Limitations in the Design
Three limitations are inherent in the basic design:
1. The test instruments and verbal reports were
indirect measures of a person's values, affects,
and behavior.
2. Each participating professional had an indivi
dual variation in personality, style and
rapport.
3. Subjects varied in age, sex, occupations and
amount of education.
Delimitations
The subjects of this study were persons who applied
for therapy or counseling in a therapeutic setting of
their choice. These settings varied and consisted of pri
vate guidance and psychiatric clinics, a community college,
and private practices in psychology in Los Angeles and Or
ange Counties of Southern California, and of an out-patient
10
clinic in a psychiatric hospital in Pennsylvania.
The amount of therapy was limited to fourteen ses
sions, and the lower age limit of the subjects was fifteen.
As the participating professionals were those who
volunteered and were interested in the research, they were
not equally representative of the three types of person
ality theory subsumed in this study.
Definition of Terms
Therapeutic Movement: Changes in the attitudes, affects,
values and/or behavior of a patient as a result of therapy
or counseling which are inferred from results of objective
instruments and verbal reports. A larger inference is
made, but not empirically measurable, that these are
changes or growth whereby the patient accepts himself with
his limitations and potentialities and perceives his life,
situation, and relationships as more satisfying, richer,
and fuller.
Criterion Measures:
Personal Orientation Inventory (POI): A self-administered,
objective instrument designed as a comprehensive measure
of values and behavior seen to be of importance in the de
velopment of self-actualization.
Multiple Affect Adjective Check List (MAACL): a self
administered objective instrument which measures affects of
anxiety, depression, and hostility.
11
Semantic Differential Scale: A scaling instrument of se
lected terms designed to measure the evaluative meaning of
concepts.
Behavior Reports : Statements by the experimental subjects
or the therapist (for control subjects) at the beginning of
therapy which describe present behavior, the behavior de
sired which is relevant to the goal, and behavior at the
end of the experimental period.
Goal: The desired end result of purposeful action or
thought which is assumed to be initiated by
psychotherapeutic treatment.
Criterion Measure:
List of Goal Statements: An instrument developed from the
literature pertaining to statements or definitions of the
goals for each of the three broad theories of personality
and their related systems of psychotherapy.
Awareness : In this study connotes regularly aroused
cognitive recognition of and affective
association with a purposeful idea or goal.
Instrument:
Written Goal: The statement of the goal written out on a
4 x 6 card, taken home by the experimental subjects and
placed in a convenient or conspicuous place for daily re
ferral .
Value : A subjectively held belief as to the worth,
utility and/or desirability of man and his actions.
12
Criterion Measure:
Semantic Differential Scale: As described above.
Organization of the Remainder of the Study
The remaining chapters of this study are organized as
follows: Chapter II reviews the literature and research
which pertains to the study. It is divided into four main
sections: (1) the goals in psychotherapy; (2) categoriza
tions of therapeutic systems and goals; (3) congruency of
patient-therapist values; and (4) motivation.
Chapter III describes the basic design of the study.
It includes a description of the sample groups and of the
therapists who served as experimenters; the experimental
procedures; a report on the instruments, including two de
veloped for this study; and the manner in which the data
was treated.
Chapter IV contains the findings of the investiga
tion. The results of the various statistical tests are
reported and analysed as they related to the hypotheses
as stated in this first chapter. Additional relevant
findings are then reported and analysed.
Chapter V concludes the study with a brief summary
of the procedures and the findings. Implications of the
study are discussed and conclusions drawn. Finally, the
recommendations for further study are presented.
13
The Addendum gives examples of goals set by the ex
perimental subjects and presents the experimenters’ per
sonal reactions to the goal-setting process.
CHAPTER II
REVIEW OF THE LITERATURE
This chapter reviews the literature which provided
theoretical and functional support for the investigation.
The first section of the review presents the current empha
sis in the literature on the function of the goal and ex
amines the research into the application of the patients'
and the therapists' goals in therapy, including written
goals. The second section reviews the major theoretical
categorizations of therapeutic systems and goals. The
third section is devoted to a survey of the research on
congruency of the patient-therapist value system. Finally,
the fourth section explores the developing concepts con
cerning patient motivation.
Goals in Psychotherapy
A trend is developing in psychotherapeutic and coun
seling literature which varies from mild suggestion to
strong demands that specification of goals be a primary
focus of treatment, counseling, and research (Meltzoff and
Kornreich, 1970; Ohlsen, 1970; Urban and Ford, 1971). It
is felt that this is an area which has been too long ne
glected both in terms of efficiency of treatment and ac-
14
15
countability of the professionals (Ehrle, 1970).
Meltzoff and Kornreich (19 70) in their comprehensive
review of psychotherapeutic research state that the mean
ingfulness of studies on the outcome of therapy has fre
quently been challenged but, nevertheless, such studies
have steadily increased in number and quality, beginning in
1950, and especially in the 1960Ts. These authors point
out that despite this increase and despite the fact that
"the goals of therapy are central not only to patients 1
and therapists ' perceptions of success but also to the in
terpretation of the meaningfulness of research findings
. . . goals of therapy have received scant attention from
researchers" (1970, pp.172-173).
In a similar vein, Mahrer (1967), in the introduction
to the book on goals whichhe edited, states,
The literature on psychotherapy has relatively
little to offer on the goals of psychotherapy--
their identification, significance and organi
zation. On this point, clinicians, researchers,
and theoreticians have been curiously inarti
culate . . . In contrast to the paucity of
literature on goals, two bordering areas of in
vestigation have attracted considerable atten
tion, namely, methods and techniques of psycho
therapy and indices of improvement. However, the
methods and techniques are only working sub-
goals which point the way toward the ultimate
goals without telling us a great deal about
them. (1967, p. 1)
Hersch (1968), in delineating his reasons for the
"discontent explosion" in the mental health field attri
butes one reason for it to the disagreement over the
16
question, "What is the goal?". Previously, three goals of
treatment dominated the field: reduction of inner distress,
cure of mental illness, or personality reorganization. Pre
sently, the merit of these is questioned, and the emergence
of new goal concepts has contributed to further contention.
He states, "In sum, there is little current consensus con
cerning goals, and opposite viewpoints are prominent"
(1968, p. 500). He cites the publication of Mahrer's com
plete text on the subject of goals, which was referred to
above, as evidence of the divergent points of view in this
area.
Although they acknowledge that, in fact, there is
little agreement on what should constitute the objectives
of treatment programs as well as criteria measures, Urban
and Ford (1971) emphasize the importance of explicit spe
cifications of the outcomes one seeks. Delineation of spe
cific therapyobjectives, which can be implemented by de
limited and indicated procedures and are subject to empiri
cal test, constitutes a critical step in the therapeutic
process. Far too often, in practice and research, these
objectives never become formulated. As a result the
changes which occur do so without planning on the thera
pist's part and are "essentially fortuitous and cannot in
any way be ascribed to deliberate and planned intervention
by a therapist" (1971, p. 15).
17
Lowe (1969) touched upon this subject and perhaps
gives a further reason for the apparent reluctance of
therapists and researchers to specify goals when he said,
"Statements about goals have been distrusted because they
imply statements of teleology or purpose which are regarded
as alien to scientific objectivity." However, he further
states, "While the scientist may prefer not to discuss
ends or goals, he cannot long avoid doing so in an applied
profession" (1969, p. 90).
Goldberg, Lesser and Schulman (1966) also give a
clear statement of the objections they found in the litera
ture to the setting of goals. They cite writers who believe
that setting goals would impede, artificially alter, or
fail to be appropriate to emerging therapeutic relation
ships , resulting in an interference to the individualized
and free flowing process of that relationship. Some others
were found to be concerned about over-rigidity of therapeu
tic direction or "strait-jacketing" the therapist. Still
others believed that too early specification of objectives
may result in under or overestimation of the possibilities
of working therapeutically with families or individuals.
These authors believe that these objections are not
valid, as they state in the following:
... it does not necessarily follow that atten
tion to goal setting will result in the imposi
tion of arbitrary, rigid, restrictive treatment
18
aims. Rather, we view the selection of treat
ment goals as flexible guideposts about which
therapists can review and reformulate past and
current conceptions of a case and organize their
thinking regarding future treatment objectives.
It is our belief that the formulation of aims
and predicted results at the outset and at var
ious points in the life of a case leads not to
restrictiveness but to differentiated therapeu
tic planning based on sound diagnosis and eval
uation of psychopathology and ego strengths. It
can provide a baseline for the therapist to
gauge developments and attainments in a case
against expectation and revise treatment plan
ning if indicated. For research purposes the
inclusion of information regarding treatment
goals in any evaluation of the outcomes or re
sults seems indispensable. [1966, p. 127)
These above citations evidencing the concern for more
attention to the efficacy of goal setting in therapy stem
in part from the published research of the behavior modi
fication therapists which appears to demonstrate and docu
ment positive results in behavior change (Lazarus, 1961;
Paul, 1967). These and other studies, such as Cooke (1966)
have been carried out "usually under careful experimental
conditions and with rigorous and objective criterion mea
sures" (Meltzoff and Kornreich, 1970, p. 207).
Although the behavioral techniques were applied in
these careful studies to phobias, anxieties, and maladap
tive behavior, John Krumboltz (1966), who is a leading ex
ponent of behavior modification in the counseling field,
also strongly advocates behavioral, individualized goals
with adequate criterion measures. Anderson (1970) in re
viewing Krumboltz and Thoresen's book on behavioral coun
19
seling says that the book emphasizes "taking the insoluble
problems sometimes presented by clients and replacing them
with achievable goals" (1970, p. 78). Krumboltz himself
states that goals such as self-actualization and self-un
derstanding are too abstract; they may be worthwhile but
are not sufficient. He lists three criteria for counseling
goals:
1. The goals of counseling should be capable
of being stated differently for each indi
vidual client.
2. The goals of counseling for each client should
be compatible with, though not necessarily
identical to, the values of his counselor.
3. The degree to which the goals of counseling
are attained by each client should be ob
servable. (1966, pp. 4-5)
In the writings of the existentially oriented thera
pists, a consideration of goals, in addition to the impor
tant patient-therapist relationship, is becoming evident.
Carkhuff and Berenson (1967) speak of the two phases of
therapy. In the first phase, "the therapist communicates
empathy and positive regard . . . in order to establish a
genuine, fully sharing relationship in which the client can
feel himself freely and deeply. During the latter phase,
the emphasis is upon concrete considerations of problems
and their possible solutions within the context of this
genuine relationship." This emphasis is further defined
as "concrete and specific direction and goals implemented
20
by concrete steps" (1967, p. 142).
Douglas Q. Corey (1966) and others (Ofman, 1970;
Peris, 1969) have approached the goal concept from a dif
fering angle. They would say that all of the above implies
either directly or indirectly that behavior change is the
goal of therapy, whereas, in reality, change cannot be
willed to take place. Any goal-striving towards becoming
different or healthier is game playing. The purpose of
therapy, according to Corey, is "not to change the indi
vidual . . . , but to get the individual to fully embrace
that nature of life, the nature of himself, and the fact
that he has already chosen for good reasons to lead the
life he is leading" (1966, p. 110). The patient is expres
sing his goal in the way he Is and denying it.
This examination of the literature points to the dif
fering orientations concerning the general subject of goals
and goal setting in therapy and counseling and shows what
appears to be a growing consideration of the function of a
goal. It is pertinent now to review actual utilizations of
specific goals in practice and research.
The Use of Specific Goals in Therapy
Ford and Urban (1965) in their comparison of the
systems of psychotherapy ask, as a part of their summary,
the question, "Who selects the goals--the therapist or the
patient?" Although not so stated, their inference seems to
21
be that they are referring to specific goals, centered
around the therapist's orientation of a general, global
goal. In their view, Freud, the Ego-Analysts, Dollard and
Miller, Wolpe and Adler all accept the therapist as the
expert who sets the goal by establishing what the problem
is. Wolpe and Adler also suggest the therapist selects
the prefered new responses, while the others expect the
patient to make this selection.
The Existentialists, according to these authors, ex-
pect the patient to select the goals as a significant as
pect of treatment. Rogers, for instance,'believes there is
an innate response pattern which elicits behavior (the ac
tualizing tendency) and an innate regulatory mechanism
(organismic valuing) which will lead to sound, socially ac
ceptable choices if permitted to operate without interfer
ence. However, although Existentialists emphasize the pa
tient's choice of goals, on another level each theorist
also chooses the goal as existential assumptions about the
nature of man define the therapeutic goals.
Sullivan is presented as representing the third opin
ion that agreement by the patient and the therapist on some
explicit goals is a precondition to effective therapy.
"Although some changes in the patient's behavior can be
achieved without his awareness, they can be accomplished
more efficiently if the patient explicitly agrees to them
22
and can participate actively in the effort to realize them”
(Ford and Urban, 1965, p. 664). Coexistent with this be
lief is Sullivan's condition that the therapist, of course,
may express his objections to undesirable or irrelevant
goals, and he may have additional goals for later in
therapy.
Such agreement on goals by the patient and the thera
pist has been more recently advocated. For example, Fiske,
Luborsky, Parloff, Hunt, Orne, Reiser, and Tuma (1970), in
their list of suggestions for improving research investiga
tion call for research studies in which the outcome goals
of the patient and the therapist are congruent and the
goals of treatment are individualized. Meltzoff and Korn-
reich (1970) speak of perception of success in psychotherapy
as being based in part upon goals, and they point out that
if the therapist's and the patient's goals are not the
same, success for one may be failure for the other.
Ohlsen (1970) argues against the use of vague general
goals and stresses the importance of specific client goals
both for effective counseling and for appraisal of counsel
ing outcomes. He explicates both the client's and the
counselor's roles in the setting of goals :
When a counselor accepts a client's reasons for
seeking counseling, helps him translate these
reasons into clearly and simply stated changes
in behavior and attitudes, and helps him recog
nize and state new goals, it helps the counselor
23
identify therapeutic material and helps his
client obtain feedback on his own growth.
(1970, p. 42)
Furthermore he emphasizes that goals stated in specific,
measurable terms are necessary for defining adequate cri
teria and for the selection of appraisal instruments.
Ohlsen sees value in the development of taxonomies
for fitting specific goals to individuals such as Krath-
wohl, Bloom, and Masia (1964) have developed for the af
fective domain in education. Two such attempts in the
guidance area are being undertaken: The National Study of
Guidance and Accountability in Pupil Personnel Services:
A Process Guide for Development of Objectives for use in
California.
The primary purpose of the National Study of Guidance
(Wellman, 1970), which is in its preliminary stages, is
. . . to develop and test a research design
which will provide a structure for the collec
tion and analysis of data to determine the dif
ferential effectiveness of guidance in terms of
student behavioral outcomes, and to ultimately
provide the basis for a large scale evaluation
of guidance in elementary and secondary schools.
(1970, Part I, p. 1)
The undertaking of this study is a result of the en
actment of The National Defense Education Act of 1958. The
United States Office of Education was charged with the ad
ministration of Sections 601-504(a) of Title V of this Act
which included the responsibility for evaluation of guidance
24
programs. The extensive and detailed taxonomies of gui
dance objectives and variables being prepared under Well
man's direction give a multitude of general and specific
goals or objectives for guidance with their related cri
teria for secondary school students. These are formulated
so as to cover the major portion of guidance activities for
a majority of students thereby including both the univer
sality and the individuality of guidance. Additional in
dividualizing is recognized as necessary for more unique
needs of some students. Both process and outcome research
is expected to result from this study.
The Process Guide in California which was developed
by a task force of California educators is now in published
form, edited by Sullivan and O'Hare (1971). Two primary
sources were used in its preparation: the preliminary work
of Wellman and the writings of Krumboltz. The latter's
three criteria for counseling goals as cited above are in
herent in its formulations. Its purpose and methodology
are similar to those of the National Guide.
These projects are examples of the efforts being ex
pended at the national and state levels, which in itself
is a reflection of the growing concern over the question
of effectiveness of counseling and therapy in the public
service institutions with the issue of accountability to
the tax-paying community a motivating force.
25
Raymond A. Ehrle (1970), who is Director of Manpower
Services at Teledyne Packard-Bell in Washington, D.C., ex
presses this concern and contends that an individual-
process oriented approach to counseling is not the answer
in view of the pressing problem of obtaining effectiveness
plus efficiency. He says, "This situation is becoming in
creasingly obvious in professional discussions, in Con
gressional debates, and in tentative budget appropriations
at a time when, of necessity, the notion of effectiveness
plus efficiency is coming into fashion" (1970, p. 119). He
calls for a program of "performance contracting" wherein
commonly agreed upon goals of performance are described
and specified in advance. A by-product of this system
would be professional accountability, promotion of experi
mentation and demonstartion, and protection to the client.
Written Goals in Therapy
A survey of the literature reveals that there are
few research studies which explore the use of written
goals. Although a specific therapist may record his goal
for each patient or a patient record his own goal, there is
no evidence of this practice in any of the above cited re
views (with the possible exception of the Process Guides),
and there is little research evidence as to the value or
relevancy of such a practice. The few studies reported be
low are peripheral to the conception of written goals as
26
defined in this investigation, either as to their subject
population, the use made of the written goals, or the re
search design.
Winter, Griffith and Kobb (1968) conducted research
into the capacity for self-direction. A content analysis
was made of self-descriptive essays written by students,
thirteen of whom were subsequently successful and eleven of
whom were unsuccessful in attaining self-directed behavior
change goals. The results revealed that (1) high change Ss
more frequently stated goals with implicit recognition that
the goal had not yet been attained, (2) low change Ss more
frequently described themselves with little recognition of
alternative possibilities and (3) low change Ss showed more
tentativeness and uncertainty about themselves. This ex
periment was cross-validated in a second sample of nine
successful and 22 unsuccessful Ss. The authors state that
the results suggest successful self-directed change is mo
tivated by awareness of the cognitive dissonance created
when an individual commits himself to a valued goal that he
sees as different from his present behavior.
Goldberg, et al. (1966) describe the need they felt in
a child guidance clinic for "an evaluative instrument which
could help organize thinking-about goals in individual cases
and which would serve as a focus about which to integrate
diagnostic, evaluative and treatment planning hypotheses"
27
(1966, p. 125). Taking the position that all treatment
modalities should be goal-focused activities, they de
signed a detailed 4 x 4 "Guide to Goal Setting" on which is
listed for each case the primary goal, the mediating or
facilitating goals, behavioral goals, and obstacles to goal
realization. These are all related separately to the over
all family picture, the extra and intra-family dynamics and
to the intra-psychic pathology. In the testing-out stage
for this guide, they question that so complex a model is
necessary, but they remain convinced that the best use of
therapeutic resources is through the continuous setting
and reevaluation of treatment goals.
Vordenberg (1970) reported on a method he has found
in group therapy for eliciting from the group those prob
lem situations which are of greatest concern to the members
of the group. He asks each member to write out a brief
resume of his problem situation which he is willing to have
the group consider and discuss. The unidentified resumes
are read aloud and the group members rank their degree of
interest in each situation. He finds this method effective
in setting group goals and meeting individual needs as well
as getting the group started.
The Union College Character Research Project (1963)
has for many years used as one of its basic concepts the
setting of a written learning goal for each lesson in its
28
curriculum for character education. Parents set the goal
for young children and older youth write out their own. The
written records of these goals, plus a description of the
method used to attain them and the reported successes and
failures in attaining them, are forwarded to the research
laboratory for analysis and as data for extensive research.
Based on premises formulated by Ernest Ligon (1948) these
definite goals stated realistically have been shown to be
an essential concept for character education.
Pascal and Zak (1956) in a research study primarily
investigating the amount, of therapy most effective for
behavioral change specified in writing in each case record
before treatment began what the outcome measure would be
in terms of behavior to be changed. They evaluated the
effectiveness of therapy by whether their behavioral goal
was attained. Of thirty subjects, twenty-eight changed
behavior in the predicted direction. There were no differ
ences in the number of behavior changes for different
amounts of therapy.
Categorizations of Therapeutic Systems and Goals
It is now well accepted that a concept of the nature
of man is incorporated into the varying personality theories
and is related to the therapist's choice of a theory and to
his goals and values (Allport, 1962; Glad, 1959; Lowe,
1969; Patterson, 1966; Singer, 1970). There are many dif-
29
ferent personality theories which have developed into psy
chotherapeutic systems, and each could be assumed to have
its own particular goals. Lowe (1969) cites Harper's sur
vey of psychotherapies in which he found some thirty-six
different systems. This survey was made before the recent
proliferation of existential and learning theory types.
Many authors and researchers have categorized these into
types.
Gordon Allport's (1962) "three images of man" is pre
sented first, for his division forms the theoretical basis
for the categorization of the therapeutic systems into
three, broad, general types utilized in this study. All
port, recognizing that there are differences between psy
chotherapeutic theories, postulates as fundamental, the
difference as to the image of the nature of man. He pre
sents the three "images" or views which describe the nature
of man as seen by the differing theorists and says these
provide a focus for all psychological activity.
The first is man seen as a reactive being--a biologi
cal organism reacting to stimuli in his environment. His
experiences and past learning have determined his behavior
--environmental determinism. The concepts representing
this point of view include reaction, reinforcement, reflex,
respondent, reintegration and reconditioning--all "re-"
concepts refering to the antecedent conditions. Man as a
30
reactive being is the image of man assumed by the psycho
therapists with a naturalistic, learning or behavior theory
approach.
His second image is man as a reactive being in depth,
where man is seen as reacting to his innate drives, motives,
impulses, and needs rather than to his environment. This
is the view of depth psychology, including psychoanalysis
and the psycho-dynamics of neo-Freudians. The concepts
under this rubric are familiar psychoanalytic concepts,
such as repression, regression and resistance.
Allport calls the third image "man as being in the
process of becoming," a view of man as a personal self,
conscious, future oriented and committed. The representa
tive concepts are tentativeness and commitment; man is
committed t£ something--t£ his existence and ;to becoming.
This is the outlook of the client-centered and existential
therapists.
In another of his writings (I960) Allport distin
guishes the three categories from a somewhat different
frame of reference when he says:
Already in the 20th century three great revolu
tions have occurred in man's thinking about his
own mind. These are, first, Freudian psycho
analysis, with its discovery of the depth and
the emotion in mental life; second, Behaviorism,
with its discovery of the accessibility of mind
to objective study; and third, Gestalt psychology,
with its discovery of the essential orderliness
and self-regulation of mind. (1960, p. 3)
31
Ford and Urban (1965) also comment on the underlying
assumptions about the nature of man which operate implicitly
in the psychotherapeutic systems. They describe two gen
eral views, saying man is seen either as a pilot or a robot.
Man as a pilot chooses his course in life relative to
his personal characteristics, present situations, and goals
he seeks. He can be "responsible" for his own behavior.
The theories of Adler, Rank, Rogers and the existentialists
are subsumed under this rubric.
In the view of man as a robot, the automaticity of
behavior is emphasized. Man's ship of life is directed by
all the influences and situations it meets; it only seems
to be guided but in reality it is determined by outer or
inner forces. Man is not responsible for the course it
takes. Wolpe, Dollard and Miller, and Freud exemplify
theories of this view.
A similar dichotomy has been made by Patterson (1966).
He refers to the rational and affective approaches. The
rational approach emphasizes reasoning, insight and imper
sonal objectivity, whereas the affective emphasizes affect,
experiencing, and personal spontaneity (1966, p. 498).
Both Ford and Urban and Patterson recognize that
there is no theory which is a pure embodiment of their two
types. Ford and Urban believe there is need for both, and
Patterson sees the therapeutic processes as falling on a
32
continuum between the rational-affective, which, however,
appears to have a bimodal distribution or a dichotomy.
Research studies, using factor analysis, have been
conducted in an attempt to relate therapists’ methods,
techniques, purposes, and philosophy to the differing or
ientations. Wallach and Strupp (1964) used a seventeen-
item scale of Usual Therapeutic Practices to compare Ortho
dox Freudian, general psychoanalytic, Sullivanian and
client-centered therapists. On the major factor--the main
tenance of personal distance--the first group was the high
est in personal distance, the second group next highest,
and the remaining two about the same but lower than the
other two.
Sundland and Barker (1962) used a 133 item scale
called the Therapist Orientation Questionnaire. The 139
psychotherapists who participated were classified into
three groups --Freudians, Sullivanians, and Rogerians--and
compared on the scale. The three groups differed signifi
cantly on nine of the sixteen sub-scales with Sullivanians
in the middle. A factor analysis of the sixteen sub-scales
yielded six factors of which the major factor was a contin
uum from ’ ’analytic" to "experiential." (An interesting
corollary to this study is a finding in a unique follow-up
study. Sixty of the 139 psychologists had been classified
as inexperienced. Anthony (1967) sent the Therapist Orien
tation Questionnaire to these and received thirty-eight
33
responses. Among the many shifts in professed methods and
techniques used by the Freudians, Sullivanians, and Roger-
ians was an increase by all in setting goals and a shift
regarding client-self-understanding as an increasingly im
portant goal.)
Some contrary evidence and conclusions are present in
the literature in regard to the relationship between a
therapist's theoretical orientation and the techniques
and approach he utilizes in practice. Weissman, Goldschmid
and Stein (1971) received responses from 116 clinical psy
chologists and found that neither the therapeutic orienta
tion they received in their training nor their espoused
orientation utilized in practice was related to the tech
niques actually employed in present practice. They con
clude that the therapeutic "themes" of Equalitarian, Dog
matist, Authoritarian, Normalist, and Pragmatist as sug
gested by M. Lionells in an unpublished doctoral disserta
tion are more important than that of orientation in under
standing hinds of behavior utilized by therapists.
Nevertheless, studies by Strupp (1955, 1958a, 1958b)
and McNair and Lorr (1964a) do illustrate that theoretical
orientation is related to techniques and methods used. In
his summary of the review of these studies Patterson says,
"These studies support the existence of differences among
therapists" (1966, p. 493). Meltzoff and Kornreich (1970)
34
in their similar review summarize:
Available research indicates that there are
apparently genuine differences in approach
and technique as a function of "school" or
orientation of the therapist. Therapists tend
to conceptualize therapeutic problems, steer a
course, and use strategems that reflect their
orientations. (1970, p. 383)
Fielder's (1950) classic study explains the reasons
that despite these differences all approaches report suc
cess. He concluded from his study of therapists that the
ability to create the "ideal therapeutic relationship" was
probably more a function of the ability to form good inter
personal relationships than any particular orientation.
Patterson (1966) in his above review mentions that
none included the behavior therapists, but a recent study
by Ryan and Gizynski (19 71) showed that the "prevalence of
behavior modification techniques in the therapy was not sig
nificantly related to outcome, whereas a variety of pa
tients' personal feelings about their therapists were."
The results suggested that the important elements of these
behavioral therapies were interpersonal ones much as has
been demonstrated in psychodynamic psychotherapies as by
Fiedler, for example. However, the results also suggest
that the focus on change in the presenting problem may lead
to different conclusions than if one examines general
changes in the patients' adjustment and satisfaction.
35
Categorization of Goals Stemming from the Systems
The differing conceptions of the nature of man which
influence the therapist's choice of a psychotherapeutic or
ientation also inherently include differing conceptions of
the goals of therapy. Singer (1970), who has categorized
the systems into "tension-reduction" and "activity-seek
ing," says
These two contrasting conceptions concerning the
basic nature of man must bring in their wake dif
fering therapeutic aims, differing knowledge and
awareness toward which therapists will try to
help their patients. (1970, p.26)
The aims or goals of psychotherapy have been categor
ized both theoretically and empirically. Mahrer (1967), in
his summary of the statements about goals by fifteen con
tributors to his book, The Goals of Psychotherapy, proposed
that the general goals formed three families of psycho
therapeutic approaches: the biopsychological developmental
therapies, the psychological actualization therapies, and
reconstructive therapies. He has found similarities and
differences between the goals of these three. The similar
ities, which by and large are accepted by each, are reduc
tion of psychopathology, of symptomatology, of defenses, of
psychological pain and suffering, of anxiety, the increase
of pleasure and social commitment, and the achievement of
mediating goals.
The amount of similarity and overlap between the
36
types is the subject for criticism of Mahrer's classifica
tion in Seeman's (1969) review of the book. While respect
ing the classification principle itself, Seeman believes a
problem arises from the difficulty of discriminating the
relationship between processes of therapy and goals and be
tween ultimate and mediating goals.
Buhler (1968) recognizes that there have been many
attempts to define the goals of psychotherapy. She believes
that goals fall into three categories: (1) Adequate beha
vior and functioning, (2) improved inner experiences with
happiness or at least contentment with self-awareness and
(3) modes of existences --completeness or authenticity.
Lowe (1969) presents "the four most important alter
natives which contemporary society has constructed to re
place the mental health model, whose social relevance ap
pears to be steadily decreasing" (1969, p. 205). He claims
that since the demise of the "mental illness model" counsel
ors and psychologists have sought to maintain the objecti
vity by using the term "positive mental health," but his is
an illusion as there is no objective standard for so-called
positive adjustment. Today social values have become dif
fuse and relative. The result in therapy is a multiciplity
of value orientations in which practitioners use varied
means in their attempt to reach different goals.
Thus, for Lowe, a therapeutic system with its respec-
37
tive goals is, and should be recognized as such, an expres
sion of a value orientation. The goals are then values.
He sees the goals for the humanist (Maslow, Rogers, Ellis)
as freedom to be oneself, creative self-actualization, and
a sense of dignity and worth; for the naturalist (the beha-
viorist) the goals are an individual uninhibited by anxiety
and positively reinforced; for the social theorists (Sulli
van, Adler, Horney) the goals are improved human relation
ships and better culture; and for the existentialist (the
later Rogers, Frankl, Bugenthal) the goals are affirmation
of being and authenticity of being.
Glad (1958) is another author who relates the goals of
differing therapeutic systems to the value structure inher
ent in the system. All systems have the general goal which
he labels ’’good adjustment,” but each has its own definition
of this based upon its special value system. The four sys
tems analyzed by Glad are more specific than the broader
classifications above; they are Freud's psychoanalysis, Sul
livan's interpersonal psychiatry, Rank's dynamic relation
ship therapy, and Roger’s phenomenology. No behavior modi
fication therapist appears in this group. The goals for
each, based on its inherent value system are as follows:
Psychoanalysis values the psychosexually mature indi
vidual who controls affectionate and aggressive drives in
socially valuable ways and is relatively authoritarian,
38
wisdom-giving, and paternalistic. Interpersonal psychiatric
theory values the socially integrated personality who is in-
terpersonally affectionate, friendly in an equalitarian way
and accepts the conventional. Dynamic relationship theory
values the creative individual who is independent, unusual,
and unconforming. Client-centered theory emphasizes the
acceptance of one’s right to be understood in his personal
individuality with empathic respect for others.
By contrast, Gendlin (1967) argues that neither psy
chotherapeutic orientation or process is based on values,
but rather values are based on the process.
Researchers have used a different approach than the
authors in attempts to classify goals. These studies do
not directly relate psychotherapeutic orientation to the
choice of goal. Michaux and Lorr (1961) worked out a sys
tem for classifying goals as (1) reconstructive (personality
change with insight), (2) supportive (maintaining and
strengthening current adjustment), (3) relationship (better
ing social and familial relations), and (4) not classifiable.
Therapist statements of goals were classified as to type of
goal by four psychologists working independently, and data
from ninety-two cases were used. At the end of four months
of therapy, it was found that the greater the rated severity
of illness, the more likely was a supportive goal. Goals of
psychiatrists, psychologists, and social workers did not
39
differ from each other. Changes after four months were not
significantly related to goal type.
In a sequel to this study, McNair and Lorr (1964b)
obtained responses to a thirty-item Goal Statement Inventory
from 259 therapists from forty-four Veterans Administration
Clinics in relation to 523 of their patients. The thera
pists included sixty-seven psychiatrists, 103 psychologists,
and ninety-five social workers, averaging ten years of ther
apy experience. About one-half of the patients were neuro
tic; the others were psychotic or had personality or psycho-
physiological disorders. The thirty goals represented a
comprehensive list and encompassed those of Michaux and
Lorr's reconstructive, supportive and relationship types.
Factor analysis revealed that Michaux and Lorr’s qualitative
goal schema did not fit simple structure criteria. Three
factors were isolated. Factor I corresponded closely to the
reconstructive goal of Michaux and Lorr and represented the
conventional goals of intensive therapy with neurotics.
Factor II was a stabilization factor representing goals of
stabilizing current patterns and preventing worsening. These
two sets of goals are incompatible, with a correlation of
p=-.27. Factor III dealt with situational adjustment goals
stressing adjustment in various contemporary life situations.
The latter two factors were moderately related and were
thought possibly to be components of a second-order suppor-
40
tive factor.
These categorizations of the goals of therapy embody
both similarities and differences. Although there seems to
be a definite differentiation in ultimate goals between the
systems, the differences in more specific goals are less ob
vious. Patterson (1966) reasons that the seeming differ
ences in goals may be less than they first appear. Seman
tic details may foster the emphasis on differences. If ab
stract goals, such as self-actualization could be defined
in terms of behavior or accepted as denoting the goals of
responsibility, freedom and independence--all goals which
he says would be accepted by the behavior therapists --a
more unified statement of psychotherapy goals might be
developed (1966).
Lowe (1969) however, sees an unreconcilable disparity
between the four value orientations. He pairs each of
these and shows the extreme polarity between them. For ex
ample, there is a difference in values between "the choice
of feeling secure because one feels he has mastered certain
basic learning sets and the choice of that creative doubt
which results from existential crisis" (1969, p.207).
Glad (1959) perhaps states the consensus when he
says :
There is evidence and opinion that both change
in psychotherapy and personality process in
groups of normal people will depend upon the
theory of psychotherapy or leadership and the
41
kind of personality outcome the theory anticipates.
(1959, p. 5).
Congruency of Patient-Therapist Values
Reference has been made in the above review to the
influence his value system has on the therapist’s choice of
a therapeutic system and its goals. It also appears from a
review of the literature that patients tend to adopt the
values of their therapists.
Glad (1959, p. 6), who emphatically states that "a
personality theory becomes a value system for the person
who espouses it,” seems convinced that research findings,
informed opinions and logic all support the proposal that
the particular method in therapy produces its own value-
form in clients treated by it and personality change will
be toward the value system of the therapist. Additionally
or conversely, a patient's chance for success in therapy
is enhanced if he shares at the beginning of therapy a
value system similar to that of the therapist. For example,
a patient who is democratic in attitudes is apt to respond
positively to a client-centered therapist who is the kind
that Glad believes displays democratic attitudes. The ven
turesome, creative patient will respond best to a Rankian
therapist. But the democratically oriented therapist is
likely to meet failure with the patient who values
creativity.
42
One way for a therapist to strive to insure congru
ency in values is to instruct or educate his client in the
accepted "values of society." This is the point of view of
Williamson on the counseling of students as described by
Patterson (1966). Williamson believes that the counselor
has a professional obligation to be concerned with the
values represented in the goals of counseling and that
since human nature is potentially both good and evil, the
purpose of education and counseling is to enhance the good
and minimize the evil in man's nature. Hence the counselor
judges and assumes the responsibility for assisting the
client to develop his best potentiality.
Rogers (1967), of course, would desire to help a
client grow in his potential, but not through or toward
other-imposed values. For Rogers, as for Gendlin quoted
above, values are based on the process of therapy and
emerge from it. Although the client's feelings and convic
tions frequently change during therapy and he may give up
values introjected by others, this change comes about
through the process of his experiencing himself in the ac
cepting, valuing patient-therapist relationship. Rogers
lists the value directions he perceives as emerging from
this process, which are values he covets and, therefore,
by implication, a tendency toward congruence is opted.
Opinions gathered (Wolff, 1954) from a variety of
43
psychotherapists led to the conclusion that therapists be
lieved their theoretical value systems tended to be adopted
by their patients in successful treatment. A small body of
research has been conducted on this issue. But first let
us consider the results of two closely associated and plau
sible hypotheses. One is that patient-therapist personality
similarity is conducive to more effective results. Meltz-
off and Kornreich (1970) , in summarizing twenty-one such
studies, found "no solid evidence that patient-therapist
similarity or dissimilarity either aids, abets or hampers
effectiveness” (1970, p. 325). On the other hand, five
studies on the hypothesis that patient-therapist similarity
of personality has a tendency to increase as a result of
therapy indicate tentatively that such movement does take
place.
One study is reported that deals with client-counselor
similarity in values rather than personality (Cook, 1966).
Cook was concerned with the influence of value similarity,
as measured by the Allport, Vernon and Lindzey Scale of
Values, on changes in the client’s responses to evaluative
scales for four concepts on a Semantic Differential. Ninety
university students were seen in short-term counseling by
forty-two advanced counseling trainees. The results indi
cated a curvilinear relationship between value similarity
and the two concepts which showed significant change. In
other words, differences in the concepts of "education” and
44
"my future occupation" suggested a more positive change in
meaning by those with medium similarity to the counselor's
meanings of these concepts than did the differences of
those with high or low similarity.
A classic study on the congruency of values or the
tendency for patient values to shift to those ofhis thera
pist is Rosenthal's (1955). Twelve patients at the Henry
Phipps Psychiatric Clinic were given four tests before and
after therapy. Their therapists (psychoanalytically ori
ented) were given two of these tests. Patients who improved
tended to revise moral values of sex, aggression and author
ity in the direction of their therapists while moral values
of unimproved patients tended to become less like their
therapists. There are certain shortcomings in this study,
such as a small N, a large number of dropouts, and no indi
cation of any differentiation in the three value areas.
Nevertheless, the results of this study are widely quoted.
Landfield and Nawas (1964) assert that the findings of
their research indicate that Rosenthal's conclusion can be
accepted, but with reservation. Their hypothesis related
to values was "improvement is accompanied by a shift in the
present self of the client toward the therapist's ideal as
described within the client's language dimension." The
thirty-six patients and six therapists took Kelly's Role
Construct Repertory Test. Construct dimensions from these
45
were ranked and rated by each patient and his therapist.
Results showed that 801 of the most improved patients shif
ted toward the therapist's ideal and 12% of the least im
proved shifted away from the therapist's ideal. The au
thors caution that their therapists were of the "psychology
of personal constructs" approach, and it is their considered
opinion that results in similar studies are crucially de
pendent on the school or approach of the therapist.
Other researchers found that those patients who were
most religious were least likely to benefit from psycho
therapy. The rated lack of improvement among religious
patients may have been related to the large disparity of
values between patient and therapist (Rosenbaum, Fried-
lander, Jane, and Kaplan, 1956) .
In a comprehensive study, Welkowitz, Cohen, and Ort-
meyer (1967) investigated the hypothesis that therapists
and their own patients would have more similar value sys
tems than would therapists randomly paired with patients
who were not their own. The Morris Ways to Live and the
Strong Vocational Interest Blank were administered to 44
patients and 38 Ph.D. and M.D. therapists in two training
institutions. It was found, first of all, that the thera
pists. and their own patients were closer in value than
those randomly paired, and value similarity tends to in
crease with duration of therapy. Also, those patients
46
rated as "most improved" by their therapists were closer to
their therapists than patients rated "least improved." The
proposition that values move toward similarity in ongoing
therapist-patient dyads was not refuted.
Lowe (1969)} who refers to the above studies as sig
nificantly indicative of congruency, also points out that
attempts to replicate these results suggest that the rela
tionship between improvement and adoption of the therapist's
values may be neither simple nor straightforward. The ef
fect may vary from different types of therapists and dif
ferent types of clients. The dependent client is more
highly suggestible than the independent client, but general
ly clients form a class of dependent individuals. He also
illustrates the fact that no major system of psychology has
as its primary purpose that of changing the patients'
values. Rather the goal is improvement of personality or
behavior, and value judgments determine how this improve
ment is defined. The therapist may not be aware of the ex
tent of interaction between values and process, but when
congruence is lacking between the values and goals of the
patient and therapist, they find it difficult to agree on
therapeutic goals (1969, pp. 36-39). Buhler (1962) points
out that values are not identical with goals, and she quotes
a statement from the Cornell study of values:
Values are not the concrete goals of behavior,
but rather are aspects of these goals. Values
47
appear as the criteria against which goals
are chosen, and the implications which these
goals have in the situation. (1962, p. 31)
Motivation
The general area of motivation is so large and so
broad that the philosophical and psychological literature
abounds with attempts to answer the question, "Why do
people act the way they do?" Research has resulted in a
multitude of motivational theories. "It is clear that a
comprehensive, definitive psychology of motivation does not
exist. Nor would it be reasonable to expect that one
should . . . It is evident that the type of motivational
construct used is likely to be consistent with the philo
sophical origins of the larger theory" (Cofer and Appley,
1964, pp. 808, 40).
Psychoanalysis, behaviorism, and existentialism each
contain differing conceptions of motivation. Freud’s theory
begins with motivation--the biologically derived drives of
the erotic and the death instincts. Behaviorism rests on
the principles of drive-reduction and reinforcement. The
motivational construct of the existentialists is that there
is an innate forward-pulling force toward growth which is
described variously as purpose, goal-seeking, self-fulfill
ment, and self-actualization. Frankl (1963) , however, dis
agrees with this latter, believing they are effects rather
than intentions. For him the primary motivation is the
48
"will-to-meaning."
Awareness
The term "awareness," however, is a much more limited
motivational construct. As awareness of a goal was con
ceived in this study to be the cognitive conception and the
regularly aroused affective association with a written goal,
it is a therapeutic motivational technique for attainment of
the goal and, parenthetically, for changes, growth or im
provement in the patient. There are other terms used in
the literature and in research which are synonomous or
closely related. "Attention" is one.
William James, writing eighty years ago, stated
clearly his ideas of the definition, functions and charac
teristics of attention :
Everyone knows what attention is. It is the
taking possession by the mind, in clear and
vivid form, of one out of what seems several
simultaneously possible objects or trains of
thought. Focalization, concentration, or
consciousness are of its essence. It implies
withdrawal from some things in order to deal
effectively with others . . . The immediate
effects of attention are to make us (a) perceive,
(b) conceive, (c) distinguish and (d) remember.
(1966, pp. 5, 17).
In scientific research today the orienting reaction
or reflex is often equated with "attention." "The orienting
reflex is an operationally defined concept which corres
ponds in part to the conditions of usage of the conscious
ness-centered concept of ’attention'” (Maltzman and Raskin,
49
1966, p. 112). The orienting reaction has been an important
area in psychological research since Pavlov noticed the gen
eral alerting reflexes which a dog made when first presen
ted with an unconditioned stimulus, such as a tone. At the
present time the amount of research on the orienting reac
tion is extensive both in magnitude and breadth, and much
of this research has shown the importance of it to learning
(Luria, 1963; Zaporozhets, 1965). Maltzman and Raskin
(1966) say that they welcome this concept which overcomes
the use of such hypothetical variables as "learning,”
"drive," or "attention." To them it is the fundamental de
terminer of learning. Berlyne, who has researched the
orienting reaction and its relationship to the stimuli
which elicit it, says, "Since, as more and more writers are
recognizing the newer concept of 'arousal' has much in
common with the older concept of 'drive,' verification that
collative variables affect arousal processes have far-
reaching implications for motivation theory" (Berlyne and
McDonnell, 1965, p. 306).
"Expectancy" is another term used as a motivational
concept. Tolman (1932) was one of the first to develop the
"expectancy model" to explain his concept of "purposive be
havior." Basing his premises on learning theory, he never
theless, deemphasized the drive determinant of behavior and
emphasized the demands of goal objects.
50
Expectancy has become a part of the theory of achieve
ment motivation. The formula derived from experimental re
search is:
Motivation=F (Motive x Expectancy x Incentive)
There are two motives--to achieve and to avoid fail
ure. Atkinson and Feather (1966) who are the chief expon
ents of the theory of achievement motivation, point out the
similarity of their concepts and equations to those of
Tolman. All three recognize an inner, subjective variable
and place emphasis on the goal or outcome.
Allport (1960) prefers the term "intention,” believing
that the "expectancy model" is stimulus bound--that is, in
the presence of certain signs or cues the organism expects
a certain goal to appear if it follows its customary beha
vior, or, if it does not, the organism may vary its behavior.
"Intention," on the other hand, opposes the reactivity of
expectancy and is directed toward the future; it is what the
individual is trying to do even if he does not precisely
know what he is trying to do. It is part of the private
world of desire, aspiration and conscience and is linked to
value.
Patient Motivation
In therapy the term "expectancy" remains, and it has
long been thought to play a part in either the process or
outcome of therapy. Meltzoff and Kornreich (1970) review
51
studies of the function of expectancy on patient-therapist
relationship, on congruency of patient-therapist expectan
cies, and on outcome. In the latter studies, the results
have been inconclusive. Results have been more promising
in those on the development or maintenance of the thera
peutic relationship. These reviewers conclude that the
present state serves only to suggest hypotheses in this
area which may be a crucial one.
Additional studies on expectancies have been those of
Bednar and Parker. Bednar (1970) postulated that success
ful therapy can be viewed as a placebo effect, generated
by interpersonal persuasion that successfully arouses client
expectation for improvement. Therefore, it was hypothesized
that high persuasibility subjects whose expectances for
personal growth were experimentally heightened would show
more personal growth than low persuasibility subjects whose
expectancies were not heightened. Two studies (Bednar and
Parker, 1969; Kaul and Parker, 1971) failed to show signi
ficant results in the expectancy factor. It is possible
that using students as subjects rather than patients exper
iencing distress was a variable which influenced the re
sults .
Heinrich (1968) investigated the question as to whe
ther motivation by means of monetary rewards results in
better retention of nonsense syllables and found no signi
52
ficant results. Locke and Bryan (1967) investigated the
relation of task success to task liking and satisfaction
and as determinants of level of performance and boredom.
Ss were assigned to a "hard task" group and a "do your
best" group. The results showed that (a) specific hard
goals produce a higher performance level than a goal of
"do your best," (b) that hard goals yield less overall task
liking and satisfaction than easy goals, and (c) that spe
cific hard goals produce more interest (less boredom) than
"do your best" goals.
In another study, goal setting as a means of increas
ing motivation was investigated. Ss were ten male and ten
female college students who were screened as high and low
motivators. The task was to set a goal for the number of
three-two digit addition problems to be completed on each
of several trials. The control group was just told to "do
your best." Results showed that specific performance goals
can serve to raise the motivation and produce more positive
attitudes of Ss who are low in motivation. The high moti
vated Ss told to "do your best" showed little performance
increase over trials and also had less favorable attitudes.
These results seem to indicate that attitudes and perfor
mance are not necessarily related (Bryan and Locke, 1967).
Research as to validity of these latter three studies
in therapy situations is needed.
53
Krause (1967) has developed a check-list inventory,
the Client Behavior Inventory, which is designed to measure
client motivation. This fifty item inventory, which con
tains questions believed to have content validity in four
areas of behavioral indicators of motivation, such as
regularity of keeping appointments, recall of therapy ma
terial, and evidence of "homework," is filled out by the
therapist. This instrument has possibilities for further
research and validation.
Related to the question of motivation is the question
of self-responsibility. The patient's acceptance of re
sponsibility for himself and for his movement in therapy
is an important concept, especially for the existentialists.
Schroeder's (1960) study seems to be the only one which re
lates responsibility to motivation. Acceptance by the pa
tient of his responsibility for his troubles rather than
ascribing them to outside sources was measured by the Gil
bert Self-Interview Test and movement in therapy was judged
by the therapist. Schroeder hypothesized that those at the
extremes would prove more difficult than those who accepted
moderate responsibility. Difficulty was defined in terms
of taking a long time in treatment or terminating early be
cause of resistance or lack of motivation. The main re
sult was that high responsibility is associated with dif
ficult (long) therapy and high movement. Meltzoff and
54
Kornreich (1970), who report this study, conclude that "On
the face of it, the variable of perceived responsibility
appears to be an important one whose parameters merit fur
ther study" (1970, p.251).
"One of the axioms of psychotherapy is that an indi
vidual has to be motivated for it and want to change if any
change is to take place" (Meltzoff and Kornreich, 1970,
p. 250). One study is reviewed in which initial "need to
change" was positively related to subjects judged improved.
But three studies showed that initial motivation for change
was not ordinarily present in successful cases. The editors
state that these paradoxical results probably imply that
motivation is not fixed, that it is possible for highly mo
tivated patients to lose their initial desire and for poorly
motivated patients to increase theirs. Initial motivation
is apparently not a necessary or sufficient condition of
success.
Summary
The review of the literature was conducted in accor
dance with the purpose of this investigation which was
based on the hypothesis that awareness of goals serves an
important function in psychotherapy. The literature cited
stressed the increasing concern that more attention be
focused on goals. Both the opinions of theorists and the
experimental research on the utilization of patient and
55
therapist goals were examined. Global therapeutic goals
were considered in their relationship to personality theor
ies and to systems of psychotherapy. The evidence verified
the conclusion that the systems and their respective goals
fall consistently into from two to four general categories,
based philosophically on views of the nature of man.
Contemporary opinions concerning the interrelation
ship of a therapist's choice of therapeutic orientation and
his value system were included, and research supporting the
hypothesis that a tendency exists for patients to adopt
their therapists’ value system was cited. Finally, selected
literature and research from the broad area of motivation
were presented with special emphasis on methods which have
been suggested and others which have been utilized experi
mentally in attempts to heighten motivation.
CHAPTER III
THE METHOD
In this chapter the details of the research design are
described. This information is organized into the follow
ing divisions: (1) description of the sample groups, (2)
description of the experimenters, (3) the experimental per
iod, (4) the procedures of the study, (5) the instruments,
and (6) the treatment of the data.
Description of the Sample Groups
The sample population of this study was composed of
thirty-five persons fifteen years of age and older, who
had applied and been accepted for individual therapy or
counseling. Although no effort was made to categorize
their complaints or reasons for seeking therapy, they were
assumed to be representative of the population which seeks
help from clinics, psychologists and counselors for personal
problems. Any person considered psychotic and possibly in
need of hospitalization was not accepted for the study.
Also, any person who specifically sought or was deemed ap
plicable for short-term therapy (four to six sessions) was
not accepted as a subject.
Randomization of the experimental and control sub-
56
57
jects was controlled by each experimenter in the following
manner: as he began the investigation, the experimenter al
ternately designated each new subject scheduled for his
therapy or counseling as an experimental or control subject.
As the case loads of the experimenters varied and as some
were able to assume more new subjects than others, the seven
sub-groups of subjects varied in size from three to eight.
Table 1 shows the distribution of subjects among the experi
menters. The total number of subjects was thirty-five:
nineteen experimental and sixteen control.
TABLE 1
Number of Experimental and Control Subjects
for Each Experimenter
Experimenter Number 1 2 3 4 5 6 7 Total
Experimental Subjects 4 2 4 2 2 2 3 19
Control Subjects 3 2 4 2 2 1 2 16
Total 7 4 8 4 4 3 5 35
Demographical information was obtained for each sub
ject on the Subject Information Form. A copy of the form
is provided in Appendix A. The information thus obtained
related to the variables of age, sex, education, marital
status, race, occupation, and socioeconomic level which
were not controlled, but assumed to be randomized. Meltzoff
and Kornreich (19 70) in summarizing studies on these patient
58
background variables found they seemed to make little dif
ference in outcome. Low socioeconomic level is a possible
exception, but it was believed that few lower socioeconomic
level persons would be included in the study. Such was the
case, for all subjects stated they were of either lower-
middle or upper-middle socioeconomic status. Seven experi
mental and five control subjects were lower-middle; twelve
experimental and eleven control subjects were upper-middle.
One control subject was Mexican-American; all the
others were Caucasian. The median age of the experimental
group was 23 years with a range of 15-52. The median age
of the control group was 24.5 with a range between 17-43.
There were 13 males and 22 females. The distribution of the
age range and sex for experimental and control subjects is
given in Table 2.
Lower education was found in the studies by Meltzoff
and Kornreich (1970) to be possibly related to early drop
out from therapy. There were no subjects dropped in this
study, but two subjects (one E and one C) who were attending
high school did refuse to retake the POI. The educational
level distribution of subjects is given in Table 3.
Sixteen of the subjects (seven Es and 9 Cs) were full
or part-time students. Two from each group were listed as
housewives. Other occupations represented were clerks,
secretaries, pilot, actress, research analyst, librarian
59
TABLE 2
SEX AND AGE RANGE DISTRIBUTION OF SUBJECTS
Age Range 15-19 20-24 25-29 30-45
Over
45 Total
Experimental
Male 0 4 1 1 0 6
Female 4 3 4 1 1 13
Total 4 7 5 2 1 19
Control
Male 0 3 1 3 0 7
Female 3 2 2 2 0 9
Total 3 5 3 5 0 16
Total Subjects 7 12 8 7 1 35
TABLE 3
EDUCATIONAL LEVEL OF SUBJECTS
Less than H.S. 1-3 yrs. College Masters Total
H.S.GradGrad. college Grad. Degree (b)
Experimental 3 7 3 4 1 18
Control 2 4 9 0 1 16
Total 5 11 12 4 2 34
(a) One E and two Cs were below age for high
uation.
(b) One E did not state educational level
school grad-
and mechanical engineer. The majority of sublets were
single (twelve Es and ten Cs). Four of the experimental
60
group were married and three were divorced. Six control
subjects were married and none were divorced.
The subjects knew they were taking part in a research
project, but they were not told its purpose or details. Al
most all of the subjects found the tests interesting and
were eager to learn the results. They often made the usual
complaints about the ambiguity of certain statements or
words but checked the answers conscientiously, nevertheless.
Occasionally, an experimenter reported having had some dif
ficulty in persuading a subject to take the posttests, but
this occurred in only about three or four cases. As re
ported above, two of these definitely refused to retake the
POI. The investigator had slight contact with only two of
the subjects. The rest were unknown to her.
Description of the Experimenters
Seven experienced professionals served as experimen
ters in carrying out the research design with their own
patients as subjects. The investigator wasnot one of the
experimenters. Each of these persons carefully studied and
followed the directions, recorded the necessary information
and data, administered the tests, and encouraged, while not
leading, the experimental subjects in the goal-setting
process.
These professionals and the settings in which they
work are representative of personnel and facilities which
61
are available to persons seeking therapeutic or counseling
help; (1) a licensed clinical psychologist, specializing
in child and adolescent psychology in private practice,
(2) a psychological assistant, Ph.D. candidate, in the of
fice of a licensed clinical psychologist, (3) a psychiatric
social worker in a child guidance clinic, (4) a clinical
psychologist in a private clinic, (5) a clinical psycholo
gist from South Africa doing postdoctoral work in a college
sponsored counseling center, (6) a counselor of students,
Ph.D. candidate, in a community college and (7) a clinical
psychologist in the outpatient clinic of a psychiatric
hospital. The first six were located in Southern Califor
nia and the seventh in Pennsylvania.
Four of the experimenters were men and three were
women. Their years of experience as therapists or counsel
ors ranged from three to fourteen with a mean of 7.8 years.
The experimenters were asked to specify which of four
theoretical orientations--psychoanalytic, behavioristic,
nondirective, or existentialistic--would most nearly be
labeled as their own. Of the seven therapists, three had
no difficulty with this: two existentialism and one psy
choanalytic. The other four declared that they considered
themselves eclectic or pragmatic so that on occasion they
employed some or all of the theoretical methods. Neverthe
less, all designated one as first choice, resulting in
three of existentialism, three of psychoanalysis, and one
62
behaviorist.
In order to test Hypothesis 2 which states that the
values of the experimental subjects would tend to become
more congruent with the values of their therapists than
would those of the control subjects, the experimenters also
took the Semantic Differential as a measure of their values,
at the beginning of the experimental period.
Summary of Description of Sample and Experimenters
The sample population consisted of thirty-five per
sons who varied on several background variables but had in
common the variable of seeking therapeutic or counseling
help for some problem. They were independently and randomly
assigned to the experimental and control groups.
The experimenters also varied in age, sex, location,
and therapeutic orientation. Each, however, had had con
siderable training and experience as a therapist or coun
selor.
This variability among subjects and therapists allowed
for more generalization of results than is possible when
such representativeness is not present. "The more represen
tative the patients are of the population of patients,
therapists of the population of therapists, and therapy of
the population of therapies, the greater the generality of
the findings" (Meltzoff and Kornreich, 1970, p. 14). The
size of the sample, however, was a limitation on general
ization.
63
The Experimental Period
The experimental period was designated as fourteen
sessions of therapy or counseling with each subject. This
period was selected as being longer than that required for
short-term therapy but of sufficient time for the primary
purpose of goal attainment. A further specification was
that no subject could be considered dropped from the ex
periment who had completed eight sessions. Four subjects
(three Es and one C) terminated therapy in less than the
fourteen sessions; three considered their goals attained,
one became ill and one needed to travel. All of these took
the posttests and completed the experiment.
Fourteen sessions rarely constituted a commensurate
number of weeks. A few subjects were seen occasionally
twice a week, but most subjects missed several weekly ap
pointments because of holidays, vacation, illness, etc.
The median number of weeks for the experimental subjects
was 15.5 with a mean of 14.6. The median number of weeks
for the control subjects was 16.3 with a mean of 15.6.
The first three subjects were started on July 13,
1971 and the last two in early December, 1971. Subjects
were added each month in the interval with the majority
(12) beginning in September, 1971. The last posttesting
was administered in late March, 1972. The total experimen
tal period, therefore, covered eight months.
64
Procedures of the Study
A procedural kit was prepared for each experimenter
containing a Manual of Instructions, a form for information
about the experimenter, test booklets and manuals, and an
individual file folder with the necessary forms and test
answer sheets for each subject. The six forms developed for
recording goals and information were color coded to facili
tate their use. The color code is given in the Manual. A
copy of the Manual and the forms are presented in Appendix A,
The introduction of the goal setting process (the
experimental variable) was deemed crucially important. The
following suggestion of a possible way of presenting this
process was given to the experimenters in the Manual:
We will be working together for some time, and it
would be helpful to know what it is you may wish
to achieve through this therapy (counseling)--
what goal or goals you will be striving to attain.
We believe there may be merit in your stating (ver
balizing) a goal and reminding yourself of it often.
I would like your cooperation in working with me on
this plan by writing out your goal on this card,
taking it home and putting it in a place where you
will see and read it every day. Perhaps you could
attach it to your bathroom mirror, to a closet
door, or to the refrigerator. As we go along, we
will talk about your goal. You may find that you
want to modify it later, and you may do so when
ever you wish.
Specific instructions were also given for obtaining
statements of a behavior which the experimental subject
would consider indicative of a significant goal attainment
as follows:
65
The behavioral statements will necessarily be in
dividualized, but they must be stated with suffi
cient specificity that they could be verified be-
haviorally by others. For example; Do dishes to
gether three nights a week; apply for a new job;
make one new friend; express my feelings Ganger,
hurt, frustration, affection) toward a significant
other; refrain from my emotional outbursts often
enough that a significant other notices it; have
enough self-confidence or courage to ask for a
date, a raise, a teacher’s help, etc.
The three test instruments were administered some
time during the first or second session or before the be
ginning of the third session for the pretest and at the
last session for the posttest. Although suggestions for
scheduling both the pre and posttesting administrations
were given in the Manual, most generally the experimenters
took fifteen to twenty minutes at the first or second and
the last session for administration of the MAACL and the
Semantic Differential, allowing the subjects to take the
POI home and return it the following session.
The goal discussion for experimental subjects consti
tuted approximately five minutes of each session. The
point at which this discussion did occur during the hour
was left entirely to the discretion of the experimenter. He
inquired about the subject’s daily referral to the goal,
his progress towards its attainment and his possible desire
to change it. This procedural step was intended as a re
emphasis upon the importance of the goal and as a means for
strengthening the subject’s motivation for attaining it or
66
a new goal more appropriate to his growing self-awareness.
If the subject changed his goal, he wrote the new goal on
a card to take home for daily referral.
The control subjects were administered pre and post
testing on the three instruments and attended fourteen in
dividual therapy sessions in which there was no planned
discussion of goals.
At the final session, all subjects were posttested on
the three instruments. Also, all subjects checked on the
List of Goal Statements the one which most nearly described
the goal they had been or were still seeking to attain. In
requesting this goal checking of a control subject, the ex
perimenter explained that presumably the subject had had an
outcome goal in mind similar to one he would find on the
list. In some cases, subjects stated that more than one
goal applied, and they were asked to rank those for pertin
ency .
Experimental subjects additionally evaluated their
attainment of their personal goals and illustrated attain
ment or movement by statements of specific behavior. Ex
perimenters evaluated each of their subjects' attainment
of or movement towards the general goal they had previously
selected for each subject from the List of Goal Statements
for Experimenter Use.
All of the experimenters conducted their therapy or
67
counseling sessions in the standard fifty minute hour.
Other than the stipulated experimental procedures, each
therapeutic or counseling hour was conducted by each ex
perimenter in his own accustomed manner and in accordance
with his felt needs for the individual subject. No record
was made of the transactions during these parts of the
session.
The Instruments
As the primary purpose of this study was to assess
the degree and direction of change occurring within the
total sample on the variables of self-actualization, af
fects, values and behavior as a result of goal setting in
therapy, it was necessary to administer tests to each sub
ject to obtain data on the first three of these variables.
Three instruments were chosen for their relevance to these
variables. These instruments and the development of the
Lists of Goals Statements are described below.
Personal Orientation Inventory
The Personal Orientation Inventory (POI), developed
by Shostrom (1968), was used as the measure of change or
growth in self-actualization in the study. The POI is pur
ported to be a comprehensive measure of values and behavior
seen to be of importance in the development of self-actual
ization. The self-actualized person is described by
Maslow (1954), Rogers (1961), and Shostrom (Brammer and
68
Shostrom,I960) as one who is fully functioning, lives an
enriched life, and develops and uses all of his potential
ities. Such characteristics of a person might be the goal
of existential psychotherapy. This instrument was chosen,
therefore, to reflect movement toward goals chosen which
were of the existential orientation.
The POI is a 150-item paired-opposite forced choice
questionnaire and is essentially self-administering. Ad
ministrators are allowed to define words of necessary. The
items are printed in a re-usable test booklet, and the
subject records his answers on an IBM answer sheet. (See
Appendix B).
The items are scored twice, first for two basic
scales of personal orientation--inner directed support (127
items) and time competence (23 items)--and, secondly, for
ten subscales, each of which measures a facet of self-
actualization. The inner directed support scale (I) has
been found to be the best overall measure of the POI. Knapp
(1965) contended that "for the present purpose, the I scale
(inner directed) scores were used as the best single esti
mate of self-actualization" (1965, p.171). Damm (1969)
intercorrelated four potentially overall scales and all
other POI scales and concluded that the I scale raw score
(r-.95) or a combination of the raw scores of the I and Tc
(time competence)scales (r-.97) is the best overall measure.
69
For the statistical data of this study the I_ scale raw
score is used.
Additionally, however, all scales were hand scored
and a profile prepared for each subject on the pretest and
the posttest if requested by the therapist for his use in
psychotherapeutic treatment of the subject.
Shostrom reports only one study on the reliability
of the POI in which a total of forty-eight college students
were tested. Reliability coefficients ranged from .55 to
.85. That of the I_ scales was .84.
Research studies cited in the manual indicating the
ability of the POI to discriminate between self-actualized
and non-self-actualized persons and showing its correla
tions to other scales seem to establish its validity. Ber-
gin (1971), who reviews his research on many instruments,
says:
We are impressed with the potentialities of the
POI which measures life-orientation, self-actual
ization tendency, inner direction, and similar
dimensions usually considered to be qualities.
A series of studies relating it to the MMPI, the
Eysenck scales, therapeutic change and differences
between diagnostic groups reveals both its valid
ity and its ability to measure important dimen
sions not tapped by traditional scales. A good
measure of values is sorely needed in psychother
apy research, and perhaps this is it. (p.262)
Multiple Affect Adjective Check List
The Multiple Affect Adjective List (MAACL) is an
instrument developed by Zuckerman and Lubin (1965) to pro
70
vide valid measures of three clinically relevant negative
affects: anxiety, depression and hostility. These are af
fects which therapists of all orientations might expect to
be pertinent to their therapy cases, but the psychoanalyti-
cally oriented may accept insight into the antecedents of
these as a goal. This instrument was chosen for its possi
bility of reflecting movement toward a psychoanalytic goal.
The authors state that the MAACL measures affect as
a "state" rather than a "trait." Two forms of the test
containing identical positive and negative adjectives are
used: a "General" form and a "Today" form. The instruc
tions differ for each; for the former the subject is asked
to check words describing how he "generally” feels and for
the latter he is asked to check words describing how he
feels "now" or "today."
The MAACL is an extension of a check list (Affect
Adjective Check List) developed by Zuckerman (1960) which
was designed to measure anxiety only. When the two new
scales were added, the total number of items or adjectives
was increased from sixty-one to 132 (see Appendix C). The
total number of keyed words, however, is eighty-nine. The
keyed items were empirically selected; many of them have an
obvious connection in meaning with the affective state,
but, in some cases, words which appear to have obvious con
notations were not responded to. These were omitted from
the scoring. The authors cite considerable data relating
71
scores to clinical observations and to experimental mani
pulations. They caution, however, that the test is still
in a research phase and is not yet recommended for routine
applied use. The normative data is based on results from
536 subjects, representing job applicants, college stu
dents, and psychiatric patients. Significant differences
were found between the normal and patient samples. Norms
are available, though, only for the three scales on the
"Today” form and the Anxiety scale for the "General” form.
Internal reliability of the MAACL is high, but retest
reliability has not been established. High test-retest re
liability is not sought for the "Today" form, as consistent
responses would not reflect fluctuations of moods. The
"General" anxiety scale shows a reliability correlation of
.68, but no data has been collected on the Depression and
Hostility scales.
Studies on validity have been designed so that some
use the "General" and some use the "Today" form, sometimes
with the Anxiety scale only and at other times with all
three scales. Results appear to indicate that the "Today"
Anxiety scale validly measures hypnotically induced anxiety
and stress anxiety. The Depression scale is significantly
related to ratings of depression. The "General" Anxiety
scale correlates about .60 with the Taylor Manifest Anxiety
Scale. Average correlation with the MMPI scale is lower.
72
The reliability and validity of both forms and all
scales are not sufficiently clear nor significantly estab
lished, but information from additional studies is requested
by the authors.
The MAACL is essentially self-administering and is
brief, requiring from five to ten minutes. In scoring, the
plus items (negative adjectives) are scored if the subject
checks them, and the minus items (positive adjectives) are
scored if the subject does not check them. This method of
scoring is a partial control over a checking response set.
A lower posttest score indicates improvement. All tests
for this study were hand-scored, and a pretest-posttest
profile prepared for the therapist if requested.
The directions which appear on each MAACL form are
as follows:
Today Form
DIRECTIONS: On this sheet you will find words which de
scribe different kinds of moods and feelings. Mark an |[
in the boxes beside the words which describe how you feel
now--today. Some of the words may sound alike, but we
want you to check all the words that describe your feelings.
Work rapidly.
The words in the above are changed on the ''General"
form and read: you generally feel.
Bergin (1971), who was cited above, did not specifi-
73
cally review the MAACL, but he said, "Adjective check lists
are easy to administer and appear to have value in assess
ing change and in other types of evaluation" (p.261).
Semantic Differential
The semantic differential is a technique rather than
a test which was developed by Charles E. Osgood and his as
sociates (Osgood, Suci, and Tannenbaum, 1957) for the pur
pose of quantitively measuring the attitudes of individuals
towards words or concepts. They describe it as a combina
tion of associational and scale procedures. "There is no
absolute or fixed set of questions and answers to be used
for all situations, and then scored accordingly. Rather
each researcher . . . devises his own form of it according
to guiding principles established by Osgood's research"
(Nelson, 1964, p.3). A typical test based on this technique
includes a group of concepts which are to be rated by the
subject on a set of bipolar adjectives referred to as
scales.
Because of its flexibility in both application and
scoring, the semantic differential was used for three pur
poses in this study. First, it appears as the criterion
measure in Hypothesis 1.3 by which it will be ascertained
whether or not the experimental subjects show more positive
change in evaluative meanings of concepts than do the con
trol subjects.
74
Relevant to the semantic differential as a measure of
change in therapy, Osgood, et al. (1957, p.220) explain:
The significance of meaning as a critical var
iable in personality is most apparent perhaps
in the process of therapy itself, where the
principle changes that occur appear to be in
the meanings that various persons, events, and
situations have for the patient, and changes in
the interrelationships between these signifi
cances . . . The changes that take place during
therapy, then, should be reflected by changes in
the patients' meanings of such relevant concepts,
which, in tuxn should be reflected in judgments
on the semantic differential.
Secondly, the semantic differential was used for Hy
pothesis 2 which tests the tendency for the values of ex
perimental subjects to become congruent with their thera
pists' values. Osgood et al. (1957) see as one of the sig
nificant by-products of their work that the semantic dif
ferential is a new approach for "attitude" measurement. In
the opening paragraphs of their book, "attitude" is equated
with "value," for they say most social scientists would
agree "that one of the most important factors in social ac
tivity is meaning and change in meaning whether it be
termed 'attitude' or 'value,' or something else again"
(1957, p.l). Again it is stated that attitudes are learned
and implicit, have reciprocally antagonistic properties,
vary in intensity, and are predispositions toward an eval
uative response (1957, p.189-190).
Thirdly, it was recognized after the research for
75
this study had begun that the semantic differential as
constructed, would supply data on changes in the Self-Ideal
Self, which has been a source of conflicting research re
sults since Rogers (1951) first gave emphasis to the cor
relation between an individual's self-concept and ideal-
self-concept as a measure of adjustment.
Endler (1961),who conducted a research in therapy
using the semantic differential said in his conclusion:
In view of the results, it can be concluded that
changes in the meaning of the self-concept is a
promising criterion of improvement during ther
apy. The evaluative factor of meaning is the
most sensitive to change and seems to be an im
portant determinant of psychological adjustment.
(1961, p. 110)
Reliability and Validity of the Semantic Differential
Osgood (1957) discusses at length the problems of and
the research directed toward establishing reliability co
efficients. The evidence which he has amassed shows that
for individual subjects a shift of more than two scale
units probably represents a significant change or differaice
in meaning, and a shift of more than 1.00 or 1.50 scale
units in factor score (depending upon the particular factoi)
is probably significant. For group data, changes or dif
ferences in measured meaning as small as one-half of a
scale unit are significant at the .05 level. "These levels
of reliability should be satisfactory for most applications
of the instrument (1957,p.328). Weston (1966) in summariz
ing results of Osgood's and others' studies says that these
76
tend to show a high degree of reliability for this tech
nique ,
To determine whether the semantic differential mea
sures what it purports to measure (meaning) is difficult
since there is not an independent commonly accepted quan
titative criterion of meaning. Osgood et al. present sev
eral pages of validation studies (pp.140-166), and their
conclusion is that the semantic differential has high face
validity but that there is need for further studies of
this technique. Weston (1966) cites studies which demon-
.strate discrimination of subjects, measurement of induced
anxiety, and favorable validity.
Description of the Semantic Differential
The format of the semantic differential used in the
study was based on Osgood's approach with certain modifica
tions to meet the purpose of the study. Its design utilized
the assumption that the subjects would respond to the con
cept words or phrases in a manner which reflected their
basic feelings about each concept. The concepts chosen
were considered to be a representative sample of concepts
about which individuals have value feelings. The eleven
concepts are listed in Table 4.
One concept was printed at the top of a single page,
followed by the fourteen scales of bipolar adjectives,
which are listed in Table 5. These pages were stapled to
77
TABLE 4
CONCEPTS MEASURED BY THE SEMANTIC DIFFERENTIAL
SCHOOL OR WORK SEX
ME RELIGION
MY IDEAL SELF MONEY
PEOPLE IN GENERAL FREEDOM
LIFE FEELINGS
NATURE
TABLE 5
BI-POLAR ADJECTIVE SCALES OF THE
SEMANTIC DIFFERENTIAL
Good--Bad Responsible--Irrespon-
s ible
Cruel--Kind Foolish-Wise
Ugly--Beautiful Changeable--Stable
Valuable--Worthless Free--Not Free
Painful--Pleasurable Unfriendly--Friendly
Optimistic--Pessimistic Cooperative--Competitive
True--False Boring--Interesting
form a booklet with a face sheet containing the instruc
tions for marking a cell in each scale. (See Appendix D
for the face sheet.) The subject quantified the meaning
of each concept by rating his attitude towards it as is
78
shown in the following example:
LIFE
GOOD
T 3 5 ( >
BAD
CRUEL
KIND
6 5 4 3 2 1
Positive and negative adjectives were randomly placed to
the left as a control forresponse set.
The numbers printed under the six point scales above
did not appear in the booklet used by the subjects but in
dicate the scoring system and represent the following atti
tudes or feelings: LIFE IS
1) Extremely good or beautiful
2) Quite good or beautiful
3) Slightly good or beautiful
4) Slightly bad or ugly
5) Quite bad or ugly
6) Extremely bad or ugly
The lower the number the more positive the attitude. Eleven
concepts to be judged on fourteen scales each resulted in
154 scores per subject. The total scale score indicative of
the most positive attitude or value for a concept was four
teen; that indicative of a completely negative attitude or
value for a concept was eighty-four. A key for the inter
pretation of the total scale score for each concept was
79
given to each experimenter and is shown in Appendix D.
In Osgood's design, five, seven or nine cells were
used. Six were used in this design to provide a forced
choice for dichotomizing the responses (Friedman and Gladden,
1969) and compelling the subjects to deal with their feel
ings (Webb and Harris, 1963).
Generally, adjectives are selected to represent each
of the three primary factors --evaluative, potency, and ac
tivity--found in Osgood's factor analysis studies. Osgood
points out, however, that to measure attitude, sets of
scales which have high loadings on the evaluative factor and
negligible loadings on other factors would be used. He sug
gests the possibility of a refined method in which each
scale would be weighted in terms of its evaluative factor
loading for a particular set of concepts. But this would be
laborious and add little to the precision. Therefore, in
practice,scales representing other factors are included
"both to obscure somewhat the purpose of the measurement
and to provide additional information on the meaning of the
concept as a whole, aside from the attitude toward it"
(1957, p.191).
As the purpose of the study was to measure attitudes
toward valued or nonvalued concepts, the fourteen bi-polar
adjectives were all presumed to be representative of the
evaluative factor. Ten of these were from Osgood's evalua
80
tive adjuectives; the other four were from minor factors.
This method was used by Winter (1961) in a study to measure
the overall value discrepancy between students and profes
sors .
Development of the Lists of Goal Statements
The lists of Goal Statements were devised and devel
oped for the purpose of producing an instrument with which
the experimenters (therapists) could select a global, gen
eral goal for each subject. As the final lists contain
four goals from each of the three broad therapeutic orien
tations (psychoanalysis, behaviorism and existentialism),
it was expected that experimenters from these differing
orientations would select a goal representative of their
orientation. The subject, at the last experimental ses
sion, selected a goal from the paraphrased list. It was
hypothesized that the selection made by the experimental
subjects would be a goal of the same therapeutic category
as that of their experimenter.
Two lists were developed: (1) List of Goal Statements
for Experimenter's Use and (2) List of Goal Statements,
which was for use by the subjects. The latter is parallel
to the former except that the therapeutic language of the
former is paraphrased into lay language.
A longer list of eighteen goal statements (six from
each orientation) was compliled by culling the literature
81
for goal statements by authorities in personality theory
and by individual theorists. Resources used included per
tinent chapters in Ford and Urban (1965), Stefflre (1965),
Lowe (1969), and Patterson (1966); and original sources by
Krumboltz (1966), May (1958), Bugenthal (1965), Dollard
and Miller (1950) and Frankl (1963). These eighteen state
ments were also listed in terms considered to be more ap
plicable to the understanding of subjects.
Five practicing psychologists judged the two lists
(List of Goal Statements for Therapists' Use with Indivi
dual Patients and List of Goal Statements for Patients'
Use. These appear in Appendix E). Each statement in the
first list was judged as to its theoretical orientation,
ranked according to the importance the judges believed pro
ponents of each of the three theories would give to the
goals, and matched with the statements in the second list.
For specific instructions given to the judges see Appendix
E.
Fourteen of the eighteen statements were unanimously
placed in the expected category. The choices, however,
were not forced into the expected six selections for each
orientation. Instead, judges were asked to place not more
than eight or less than four in any one category. The y2
for the distribution of selected categories is given in
Table 6. The results indicate that the interjudge correla
82
tion was satisfactory as the y2 of differences was not sig
nificant at the .10 level.
TABLE 6
DISTRIBUTION OF JUDGES' CATEGORIZATIONS
OF GOAL STATEMENTS
Behaviorism Psychoanalysis Phenomenology
X2
21 32 37 4.46a
aNot significant at the .10 level of significance
The judges' matching of the Therapists' List with the
Patients' List was perfectly correlated (p=1.0).
The four highest ranked statements categorized as
psychoanalytic and phenomenological were used for the final
listing. As there were only three statements unanimously
categorized as behavioristic, two additional goal statements
were obtained by reference to Skinner (1953). These beha
vioristic goals and their paraphrased wording were categor
ized and matched unanimously by the judges and were ranked
as to importance with the three formerly chosen.
The final list of twelve goal statements, entitled
List of Goal Statements for Experimenter Use, identified as
to therapeutic orientation, and the paraphrased list, en
titled List of Goal Statements appear in Appendix E.
83
Treatment of the Data
The pretest and posttest scores of the _ I scale of the
POI, the two forms of the MAACL (six scales) and the eleven
concepts of the Semantic Differential were prepared for the
IBM 370-155 computer at the University of Southern Calif
ornia Computing Center. Fisher's t test was used on pre
test scores to determine the equivalence of groups prior to
the study. Gain scores were computed for the eighteen var
iables and analysed by _ t test also to determine whether sig
nificant differences existed between the experimental and
control groups on the gain scores at the .05 significance
level. The program utilized for the t _ test was BMDX 70
which yielded t values with means and standard deviations
for all groups.
The judgments as to goal attainment and movement
toward goals based on statements of behavioral change were
analysed by Fisher's Exact Probability Test (Siegel, 1956)
and evaluated by Latscha's tables (1953).
Both the individual seven sub-groups and the total
experimental and control groups were analysed for subject-
therapist congruency on values. The differences between
subject pretest and therapist scores and subject posttest
and therapist scores of the Semantic Differential's eleven
concepts were computed. The differences between these dif
ferences were dichotomized by the median. The resultant
84
contingency tables were evaluated by the y2 test for Two
Independent Samples (Siegel, 1956), using the formula:
x2 = N(/AD - BC/ - | )2
(A + B) (C + D) (A + C) (B + D)
to determine whether significant differences existed be
tween the experimental and control subjects of each thera
pist and between the total number of experimental and con
trol subjects at the .05 level of significance.
The relationship of subjects' selection of a general
goal to the general goal chosen by their therapists was an
alysed by Fisher's Exact Probability Test (Siegel, 1956),
utilizing Latscha's Tables (1953), to determine whether sig
nificant differences existed between the experimental and
control groups at the .05 level of significance.
The number of goal changes were tabulated and de
scribed.
The procedures described above were established and
implemented in order to determine the effects of patients'
frequent referral to written goals on positive movement in
therapy. Additionally, they were established and implemented
to determine whether the values and certain selected goals
of subjects who write out personal goals show a significant
tendency to become congruent with the values and selected
goals of their therapists. Chapter IV presents the results
of the statistical analysis and the findings of the study.
I
CHAPTER IV
THE FINDINGS
The findings and the results of the investigation are
presented in this chapter in two main sections. First, the
findings from the statistical analysis which relate to the
Hypotheses as stated in Chapter I will be shown. Secondly,
the data contained material for additional statistical an
alysis relevant to the effects of goal-setting in therapy.
These include the findings for the experimental and control
groups on (1) the additional outcome measure (referred to
in Chapter III) of movement toward a closer relationship
between subjects' concepts of Self-Ideal Self as a result
of therapy and (2) the relationship of the theoretical or
ientation of the therapist, of the subjects' selected gen
eral goal, and of the therapists' general goal to positive
movement of two variables. The second section of the chap
ter will explain and present these additional findings.
Experimental Findings
In order to evaluate meaningfully the outcome of the
experimental design, it was necessary to establish statis
tically the equivalence of the experimental and control
groups at the beginning of the study. Therefore Fisher's
85
86
test was performed on the pretest scores of the I _ scale
of the POI; the anxiety, depression and hostility scales of
both the Today and General forms of the MAACL; and the
eleven concepts of the Semantic Differential. A t value
of 2.04 would indicate for all these measures that differ
ences were evident between the two groups at the .05 level
of confidence.
Pretest mean scores and standard deviations for the I _
scale of the POI for experimental and control groups are
presented in Table 7. It was found that no significant dif
ference existed between the experimental and control groups
on this variable when the study began.
TABLE 7
PRETEST PERSONAL ORIENTATION INVENTORY I SCALE
MEANS AND STANDARD DEVIATIONS FOR EXPERIMENTAL
AND CONTROL GROUPS
Experimental Control
Standard Standard
Mean Deviation Mean Deviation t p
POI: Inner-
Direction 73.94 13.20 79.26 11.81 1.21 .23
(I)
Not significant at .05 level of confidence
N=18; one member was excluded from this evaluation
N=15; one member was excluded from this evaluation
87
Pretest means and standard deviations for the Today
and the General Forms of the MAACL as presented in Tables 8
and 9 again indicate that no significant differences are
found. Likewise, no significant differences between the
pretest means of the eleven concepts of the Semantic Differ
ential were found for the experimental and control groups
(Table 10).
TABLE 8
PRETEST MULTIPLE AFFECT ADJECTIVE CHECK LIST
(TODAY FORM) MEANS AND STANDARD DEVIATIONS
FOR EXPERIMENTAL AND CONTROL GROUPS
MAACL
Experimental
Standard
Mean Deviation
Control
Standard
Mean Deviation t
_
Today Form
Anxiety 12.63 4.17 10 .50 3.42 1.63 .11
Depress ion 20.10 4.16 18.31 7.50 0.89 .37
Hostility 9.73 4.03 10.43 3.57 0.54 .59
Not Significant at .05 level of confidence
a : N = 19
b: N = 16
88
TABLE 9
PRETEST MULTIPLE AFFECT ADJECTIVE CHECK LIST
CGENERAL FORM) MEANS AND STANDARD DEVIATIONS
FOR EXPERIMENTAL AND CONTROL GROUPS
_
Experimental Control
MAACL Standard Standard
Scale________ Mean Deviation_____Mean Deviation t p
General Form
Anxiety 10.41 3.41 9.80 3.96 0.47 .64
Depression 16.41 4.44 17.46 7.96 0.47 .64
Hostility 6.82 4.17 9.60 3.71 1.98 .06
Not significant at .05 level of confidence
a: N = 17; two members were excluded who did not take test.
b: N = 15; one member was excluded who did not take test.
TABLE 10
PRETEST SEMANTIC DIFFERENTIAL MEANS AND STANDARD
DEVIATIONS FOR EXPERIMENTAL AND CONTROL GROUPS
o ~ - .
Experimental Control
Standard Standard
Concept Mean Deviation Mean Deviation t
P
School or
Work 37.31 13.29 37.00 9.28 0.08 .93
ME 39.31 9.62 40.25 10.74 0.27 .78
My Ideal
Self 20.52 5.50 23.43 7.99 1.27 . 21
People in
General 41.42 13.04 45.31 11.58 0.92 . 36
Life 40.68 10.67 43.00 11.58 0.62 . 54
Nature 28.21 9.42 26. 56 9.15 0.52 .60
Sex 34.52 11.26 30.31 8.58 1.23 .23
Religion 36.26 16.99 39.50 20.73 1.49 .51
Money 41.00 14.31 41.31 9.87 0.07 .94
Freedom 31.68 7.41 32.00 6.42 0.13 .89
Feelings 38.63 9.28 41.37 3.44 0.70 .48
Not significant at the .05 level of confidence
a: N = 19 b: N = 16
89
As none of the t values reached 2.04 on any of the
measures, it was established that at the beginning of the
study the experimental and control groups were equal.
The Hypotheses and Results
Hypothesis Number 1. This hypothesis stated that
experimental subjects who write out and habitually become
aware of their personal goal or goals will show more posi
tive gain in therapeutic treatment than will control sub
jects who do not write out and habitually become aware of
personal goals while in therapeutic treatment. The hypo
thesis contained four sub-hypotheses, representing the
criterion measures.
Number 1.1. Experimental subjects will show more po
sitive gain on a measure of self-actualization than will
control subjects. The Inner Direction (.1 ) scale of the POI
was utilized as the measure of self-actualization, and the
gain scores of the experimental and control subjects were
evaluated by Fisher's t test.
The mean gain for the experimental group (N = 18) was
13.00 with a standard deviation of 11.31. The mean gain
for the control group (N = 15) was 3.46 with a standard de
viation of 13.38. The t value which resulted was 2.22,
which fs statistically significant with 31 degrees of free
dom at probability of .03 (see Table 11). The null hypo
thesis of no difference in gain scores between the experi-
90
mental and control groups is rejected at the .03 confidence
level. With these subjects goal-setting and habitual re
ferral to goals was effective in producing positive gains
in self-actualization values and attitudes.
TABLE 11
COMPARISON OF GAIN SCORES FOR EXPERIMENTAL VERSUS
CONTROL GROUPS ON INNER-DIRECTION SCALE OF THE
PERSONAL ORIENTATION INVENTORY
Mean
Standard
Deviation t
P
Experimental Group 13.00 11.31
Control Group 3.46 13.38 2.22* .03
*Significant at .03 level of confidence.
Number 1.2. Experimental subjects will show more
positive gain on measures of anxiety, depression, and hos
tility than will control subjects. The three scales on
each of the two forms of the MAACL were used as measures of
these affects. The two forms represent two differing time
sets: (1) how the subject feels "Today" and (2) how he
feels in "General."
The comparison for the "Today"form shows that there
are no significant differences between the experimental
group (N = 19) and the control group (N = 16). The null
hypothesis of no difference at the .05 confidence level is
accepted for the affects measured by the "Today" form
(Table 12) .
91
The three mean gain scores for this form do show more
positive gain for the experimental group than for the con
trol group but not at the stipulated significance level.
TABLE 12
COMPARISON OF GAIN SCORES FOR EXPERIMENTAL VERSUS
CONTROL GROUPS ON TODAY FORM OF THE MULTIPLE
AFFECT ADJECTIVE CHECK ; LIST
MAACL
Scale
Experimental
Mean Standard
Deviation
Control
Standard
Mean Deviation t
P
Today Form
Anxiety 4.73 6.90 1.93 4.73 1.37 .18
Depression 6.31 9.75 3.06 8.43 1.04 .30
Hostility 2.78 6.03 1.12 4.51 0.91 .37
Not significant at the .05 level of confidence.
On the General form of the MAACL a significant differ
ence between the experimental group (N = 17) and the con
trol group (N = 15) was found on the anxiety scale. The t
value of 2,44 is significant at the .02 level of probabil
ity. Neither the Depression nor the Hostility Scale reached
significance (see Table 13).
TABLE 13
COMPARISON OF GAIN SCORES FOR EXPERIMENTAL VERSUS
CONTROL GROUPS ON GENERAL FORM OF THE MULTIPLE
AFFECT ADJECTIVE CHECK LIST
Experimental Control
MAACL
Scale Mean
Standard
Deviation Mean
Standard
Deviation t
P
General Form
Anxiety 3. 70 4.56 - . 20 4.45 2.44* .02
Depression 4.17 7.65 1.93 4.66 0.98 .33
Hostility 0.52 3.74 1.00 4.58 0.32 .75
*Significant at .02 level of confidence.
92
The null hypothesis of no difference between the ex
perimental groups and control groups is rejected at the .02
level of confidence for general anxiety. It is accepted
for the affects of depression and hostility.
Number 1.3. Experimental subjects will show more
positive gain on evaluative meaning of concepts as measured
by the Semantic Differential than will control subjects.
The pretest-posttest gain scores for eleven concepts about
which individuals may have value feelings were analysed by
Fisher's i t test for differences between the experimental
group (n=19) and the control group (n=16) and are shown in
Table 14.
Table 14 reveals that no significant differences at
the .05 level of confidence were found. The null hypothe
sis of no difference between experimental and control
groups on measures of evaluative meanings of concepts is
accepted. It may be noted, however, that two concepts--
Life and Money--did reach a probability level of .10 and
.06 respectively.
The loss by the experimental subjects on the concept
of My Ideal Self requires interpretation which will be
given in the section of additional findings.
Number 1.4. Reported change in overt behavior will
be judged as more movement toward a goal for experimental
subjects than for control subjects. At the beginning of
93
TABLE 14
COMPARISON OF GAIN SCORES OF EXPERIMENTAL VERSUS
CONTROL GROUPS ON THE SEMANTIC DIFFERENTIAL
Concept
Experimental
Standard
Mean Deviation
Control
Standard
Mean Deviation t
P
School or
Work 4.00 12.10 - .62 13.38 1.07 .29
ME 6.53 12.16 1.37 11.12 1.30 .20
My Ideal
Self -.57 6.12 1.06 7.61 0.71 .48
People in
General 4.52 7.84 2.62 7.90 0.71 .48
Life 7.42 7.01 2. 56 10.02 1.68 .10
Nature 1.57 6.70 -2.56 8.28 1.64 .11
Sex 1.84 17.10 2.12 8.12 0.06 .95
Religion - .05 9.51 -.25 10.09 0.06 .95
Money 1.47 8.12 -3.87 8.53 1.90 .06
Freedom 5.63 9.87 1.18 8.80 1.39 .17
Feelings 7.21 9.63 2.87 12.07 1.18 .24
Not significant at the .05 level of confidence.
therapy or counseling, the experimenter selected a goal for
each experimental and control subject. He evaluated the
subject's movement toward this goal by indicating if the
goal was attained, not attained, or partly attained on the
basis of reported behavior. The results were tabulated and
are shown in Table 15. There are no significant differences
between the experimental and control groups.
Descriptively, however, a difference in the trend may
be noticed. Thirty-two percent of the experimental group's
I
94
TABLE 15
COMPARISON OF EXPERIMENTAL AND CONTROL GROUPS ON
ATTAINMENT OF EXPERIMENTER'S GOAL:
STATISTICAL AND DESCRIPTIVE
Attained
Partly
Attained
Not
Attained
Partly
Attained Attained Not
Experi
mental 6 13 0 32% 68% 0
Control 3 12 1 19% 75% 6%
Not significant at .05 level of confidence (Latscha Tables)
N = 19 experimental, 16 control
goals were evaluated "Yes" (attained) against nineteen per
cent of the control group's goals, leaving sixty-eight per
cent of the experimental group's goals partly attained and
seventy-five percent of the control group's goals partly
attained and six percent not attained.
The experimenter also evaluated the amount of
movement toward the goal for each subject: his selected
goal for control subjects and their personal goal for ex
perimental subjects. Again the results as indicated in
Table 16 show no significant difference between the groups,
but percentage figures indicate a trend in the direction
of the research hypothesis.
Experimental subjects evaluated their attainment of
their personal goals also. There was no comparative con
trol group for this measure. Results indicate that
95
TABLE 16
COMPARISON OF EXPERIMENTAL AND CONTROL GROUPS ON
MOVEMENT TOWARD GOAL: STATISTICAL AND DESCRIPTIVE
Definite
Movement
Some
Movement
Definite
Movement
Some
Movement
Experimental3 - 12 6 67% 33%
Control 9 7 56% 44%
Not significant at the .05 level of confidence.
a: One subject did not evaluate his goal.
forty-four percent stated they had attained their goals and
fifty-six percent stated that their goals were partly at
tained. Typical examples of written goals and behavioral
reports as to attainment and partial attainment will be
found in The Addendum.
Hypothesis Number 2
This hypothesis stated that there would be a greater
tendency for the values of the experimental subjects to be
come congruent with the values of their respective thera
pists than would the values of the control group. Basic
assumptions for this hypothesis were that the values of a
therapist reflect his theoretical orientation and that his
value system is an influence on the value system of his
patients. It was hypothesized that this influence would be
greater for experimental subjects involved in the goal-
setting process.
The Semantic Differential with its eleven concepts
96
was the measuring instrument utilized for this hypothesis.
The method used for analysing the subjects' shift toward
(closer congruency) the therapist's values was explained
in the section on treatment of data in Chapter III.
The total number of shifts above and below the me
dian with the resulting x2 for each experimenter-therapist
is shown in Table 17. A difference between his experimen
tal and control subjects significant at the .05 level of
confidence was found in the case of only one experimenter
(No. 3). A difference significant at the .10 level was
found to exist between the experimental and- control sub
jects of experimenter No. 4. Both of these experimenters
were existentially oriented, but No. 3 stated that he
utilized behavior modification techniques secondarily.
The results also show that the experimental and con
trol subjects of experimenters 1 and 2 made virtually an
equal number of shifts above and below their median. With
experimenters 5, 6, and 7, the control subjects made more,
but not significantly more, shifts toward the values of
their therapists than did the experimental subjects.
These observations of results would lead us to be
lieve that no significant difference would be found in the
comparison between the total experimental group and control
group. Table 18 verifies this, for the x2 value of .306
is not significant and merely indicates a slight tendency
97
TABLE 17
NUMBER OF VALUE SHIFTS OF EXPERIMENTAL AND CONTROL SUBJECTS
TOWARD AND AWAY FROM THEIR THERAPIST'S VALUES AS
MEASURED BY A SEMANTIC DIFFERENTIAL
Experimenter No.l,Median=-l.5
Toward Away x2
Experimenter No.2,Median=3
Toward Away x2
Experimental
Control
22 22
17 16 .009
Experimental
Control
10 12
10 12 .093
Not significant Not significant
Experimenter No.3,Median=l
Toward Away x2
Experimenter No.4, Median=2
Toward Away x2
Experimental
Control
26 18
16 28 3.68*
Experimental
Control
12 10
6 16 2.35
*Significant
(one tailed)
at .05 level Significant i at .10 level
Experimenter No.5,Median=0
Toward Away x2
Experimenter No.6, Median=0
Toward Away x2
Experimental
Control
8 14
13 9 1.39
Experimental
Control
10 12
6 5 .015
Not significant Not significant
Experimenter No. 7, Median-1
______________ Toward Away x2
Experimental 14 19
Control 12 10 .527
Not significant
in the direction of the hypothesis.
An additional noteworthy result is that in the total
group of thirty-five subjects, evaluating eleven concepts
each, there are fewer shifts (182) toward their therapists'
values than away from (203) those values.
98
TABLE 18
NUMBER OF VALUE SHIFTS OF TOTAL EXPERIMENTAL AND
CONTROL GROUPS TOWARD AND AWAY FROM THERAPISTS'
VALUES AS MEASURED BY A SEMANTIC DIFFERENTIAL
Toward Away N
x2
Experimental Group 102 107 209
Control Group 80 96 176
Total 182 203 385 .306
Not significant at .05 level of confidence.
Hypothesis Number 2 is concerned with the shifts in
each of the subgroups. Therefore, it consists of seven sub
hypotheses. The null hypothesis of no greater tendency for
the values of the experimental subjects than the values of
the control subjects to become congruent with the values of
their respective therapists is accepted for six of the sub
hypotheses. It is rejected for one sub-hypothesis (Experi
menter Number 3) at the .05 level of confidence.
Hypothesis Number 3 :
This hypothesis stated that the experimental group
would more often indicate at the conclusion of the experi
mental period a preference for a general goal which was
congruent with the experimenter's previously chosen goal
for them than would the control group. Congruency of goals
would be ascertained by selection from the Goal Statement
List of a goal in the same theoretical orientation. The
Goal Statement List contained twelve goals--four from each
99
of the three broad theoretical categories: existential,
psychoanalytic, and behavioristic.
First, a comparison was made to determine whether the
experimenters chose goals from their professed orientation.
For the total group of thirty-five subjects, thirty-five
goals were chosen by the experimenters. Nineteen of these
were of the experimenters’ professed orientation and six
teen were not. The division was equally distributed be
tween the experimental and control groups:
Experimental: 10 of the same orientation.
9 of different orientation.
Control: 9 of same orientation.
7 of different orientation.
Therefore, it seems it could be concluded that in this
population of therapists serving as experimenters, no con
sistency exists between espoused theoretical orientation
and individualized goal selected for a patient (subject) at
the beginning of therapy. However, modifications to this
are apparent.
Ten of the experimental group had experimenters with
existential orientation, seven had experimenters with psy
choanalytic orientation, and two had a behavioristic ex
perimenter. In the control group, nine had experimenters
with existential orientation, six had those with psycho
analytic orientation, and one had a behaviorist. These
figures, of course, constitute the possible number of goals
100
related to each orientation which could have been selected
for the subjects if the experimenters were completely con
sistent. Table 19 shows these figures and the number ac
tually chosen.
TABLE 19
RELATIONSHIP OF EXPERIMENTERS' ORIENTATION TO GOALS
SELECTED FOR EXPERIMENTAL AND CONTROL GROUPS
Experimenters'
Orientation______Existential Psychoanalytic Behavioristic
Experimental Group
Possible no. of
subject goals 10 7 2
No. chosen of same
orientation 8*
Q A * 2 * A A
No. chosen of dif
ferent orientation 2 7 0
Control Group
Possible no. of
subject goals 9 6 1
No. chosen of same
orientation 6 2 1
No. chosen of dif
ferent orientation 3 4 0
* Significant at .055 level
** Significant at .008 level
***Significant at .001 level
The choices for the experimental group show that ex
istential and behavioristic experimenters chose a signifi
cant number of like-oriented goals, but psychoanalytic ex
perimenters chose goals of different orientations to a very
101
significant degree.
The choices for the control group show consistency
only for the one behaviorist experimenter.
The conclusion above is, therefore, modified to state
that with experimental subjects consistency between thera
pists' orientation and choice of a goal for subjects did
exist with therapists of existential and behavioristic
orientation, while psychoanalytic orientation denoted a
choice of other-oriented goals.
The question raised by this hypothesis was whether
the goal selected by the subject at the end of the experi
mental period from the List of Goal Statements would be of
the same orientation as the goal selected by the experi
menter earlier. The experimenter's expressed orientation
is not a variable here.
The subject had four choices out of twelve on the
List to select a goal congruent with his experimenter's. If
the subject wished to check more than one, he was asked to
rank his selections. Two-thirds of the subjects did rank
from two to four goal choices. Fifty percent of the second
choices were of the same orientation as the first. Only
the first choice was utilized in the analysis.
The experimental subjects chose fourteen same-orien-
tation goals and four different-orientation goals. The
control subjects chose eight same-orientation goals and
102
and seven different-orientation goals. These results do
not reach significance at the .05 level of confidence on
Latscha's tables. The null hypothesis of no difference in
the groups betwe'en choices of same or different theoreti
cal orientations is accepted. The trend, however, is in
the direction of the research hypothesis (see Table 20).
Further findings on the relationships of goal choices to
movement in therapy will bo reported in the next section.
TABLE 20
RELATIONSHIP OF EXPERIMENTAL AND CONTROL GROUPS'
GOAL CHOICES TO THE GOAL CHOICE ORIENTATION
OF THEIR EXPERIMENTERS
Same Different
Group Orientation Orientation
Experimental 14 5
Control 8 7
Not significant at .05 level.
Hypothesis Number 4:
This hypothesis stated that experimental subjects
having the opportunity to revise their goals will do so at
least once. The findings do not support the research hy
pothesis, and the null hypothesis is accepted.
There were eleven revisions or additions to goals
made by eight of the nineteen experimental subjects. On
the other hand, a total of forty-one goals were set by the
nineteen subjects in the first sessions. This is a mean
103
of 2.16 goals for the subjects. Six set one only. Perhaps
the scarcity of additions or revisions was a function of
the number of original goals.
Apparently, setting and writing out goals at the be
ginning of therapy was well accepted by these subjects.
Whether a single goal or numerous goals was more advanta
geous to the therapeutic process was not reported to the
investigator by the experimenters. In some cases, it ap
peared that attainment of one goal was worked towards at a
time. But the numerosity of goals complicated the evalua
tive process and accounted for the large number of "partly
attained" evaluations.
Additional Findings
The findings relevant to the four hypotheses of the
study have been presented in the previous section of this
chapter. It became evident that additional information re
lated to the purpose of the study could be gleaned from
the data. This is presented here.
First, considerable research has been conducted into
change in the discrepancy between the concept of the self
and the ideal-self. Rogers (1951) was the first to empha
size correlation between an individual’s self-concept and
ideal-self-concept as a measure of adjustment. Two classic
studies in this area are Rogers' and Dymond's (1954) and
Butler and Haigh's (1954) in which significant decreases in
104
S - I discrepancy were found in client-centered therapy.
Since then this has been a popular criterion measure and
has also been the source of criticism.
Criticism has arisen as to the reliability of the
discrepancy, as to what it truly measures--self-esteem or
"Phillistine self-satisfaction1' (Loevinger and Ossorio,
1959), or conformity. Butler, in a recent study (1968),
found that the ideal concepts are more stable than self-
concepts; Varble and Landfield (1969) found that both the
self and the ideal-self change; and Frank and Heister (1967)
found the self-concept more stable over time.
Despite the criticism and some conflicting results,
most studies conclude by supporting Rogers' assumption. Al
though the Q-sort was the most generally used measuring
instrument at first, many new and different measures have
been introduced lately. Luria (1959) and Endler (1961)
both reported changes in the concept of the self to be a
central aspect of improvement with therapy and employed the
semantic differential to measure it.
The Semantic Differential used in this present study
includes the concepts of Me and My Ideal Self. The ques
tion to be determined is, then, is a more significant de
crease in Self, Ideal-Self discrepancy found in the experi
mental than in the control group?
Table 21 presents the pretest and posttest means,
105
standard deviations, and I t values for the differences be
tween the Me concept and the My Ideal Self concept, and the
means, standard deviations and £ value for the resulting
difference of these differences.
TABLE 21
MEANS AND STANDARD DEVIATIONS OF ME AND
MY IDEAL SELF-CONCEPTS
Experimental
Standard
Mean Deviation
Control
Standard
Mean Deviation t
P
Pretest
Differences 18.78 10.07 16.81 10 .80 0. 56 . 58
Posttest
Differences 11.68 10. 71 16.50 11. 53 1.30 . 20
Total
Differences 7.10 11.37 .31 14.00 .1.58 .12
Not significant.
Although none of the t values are significant, it may
be noted (in the third row of the table) that the experi
mental group’s mean of 7.10 is considerably higher than the
control group's mean of .31. The difference between these
means signifies that the experimental group showed a greater
tendency toward posttest congruency than did the control
group. The large variance between individual scores, which
is probably a result of differing individuals' scoring
habits, functioned to prevent the mean difference from
reaching significance above the .123 level.
106
The results of this comparison tend towards the con
clusion of previous research studies that the Ideal-Self
concept is more stable than the Self concept and to ex
plain the loss mean score on My Ideal Self for experimental
subjects as shown in Table 14. Comparison of the mean gain
scores (see Table 14) of the ME concept and the My Ideal
Self concept shows that these were larger on the ME concept
for both groups. The subjects evaluated themselves more
positively. The smaller pre-posttest mean gains for the My
Ideal Self concept, especially for the experimental group,
which shows a loss, indicate greater stability for this
concept. In other words, more changes were apparent in the
subjects' concept of their real selves than in their Ideal-
Self concept. The loss, rather than gain, in Ideal Self
reflects the tendency for closer congruency of the My Ideal
Self to the Me self as subjects tended to lower the ideal-
self concept to meet the real Me concept.
The second additional finding concerned the relation
ships between the significant gains by the experimental
subjects on the I_ scale of the POI and the General anxiety
scale of the MAACL and the theoretical orientation of the
therapists, of the subjects' goal choices, and of the ther
apists' goal choices.
Self-actualization, which is measured by the POI, is
an existential goal. To answer the question as to whether
there is a relationship between gain on the POI and
107
Cl) a therapist’s stated orientation, (2) a therapist's
choice of an existential goal regardless of his orientation
and (3) a subject's choice of an existential goal, the data
was analysed. Fisher's Exact Probability Test with
Fisher's and Latscha's tables was utilized.
The total subject population was analysed first for
the above three relationships. Tables 22, 23, and 24 show
the results. No significant relationship was found between
the therapists' orientation and gain or loss on the POI
(Table 22). Significant relationships were found between
both the subjects' existential goal choices (p = .012) and
the therapists' existential goal choices (p = .014) and
gain on the POI (Tables 23 and 24).
TABLE 2 2
RELATIONSHIP OF THERAPISTS' ORIENTATION TO GAIN OR
LOSS ON THE PERSONAL ORIENTATION INVENTORY
Gain Loss N
Existential Orientation 16 3 19
Other Orientation 10 4 14 2 subj ects did
not retake POI
Not significant
108
TABLE 23
RELATIONSHIP OF SUBJECTS' GOAL CHOICE TO GAIN OR
LOSS ON THE PERSONAL ORIENTATION INVENTORY
Gain Loss N
Existential Goal 17* 1 18
2 subjects did
Other Goal 7 6 13 not select a
goal from the
Lis t,
2 subjects did
not take the POI
*Significant at .012
TABLE 24
RELATIONSHIP OF THERAPISTS' GOAL CHOICE TO GAIN OR
LOSS ON THE PERSONAL ORIENTATION INVENTORY
Gain Loss N
Existential Goal 18* 1 19
2 subjects did
Other Goal 8 6 14 not take the POI
* Significant at .014.
In some cases both the therapist and the subject
chose a goal of the same orientation. If the mutual goals
were existential as opposed to mutual goals of either psy
choanalytic or behavioristic orientation, a significant re
lationship was found between mutual existential goals and
gain on the POI. Table 25 shows this relationship.
Do these significant relationships between the sub
jects' and therapist-experimenters' choice of an existential
109
TABLE 2 5
RELATIONSHIP OF THERAPISTS' AND SUBJECTS1 MUTUAL GOAL
CHOICE TO GAIN OR LOSS ON THE PERSONAL
ORIENTATION INVENTORY
Gain Loss N
Mutual Existential Goals 14* 0 14
Mutual Other-oriented Goals 3 4 7
^Significant at .01
goal and gain on the POI apply to both experimental and
control? The analysis of the data to answer this question
shows there is a significant relationship at .05 level of
confidence between the experimental subjects who wrote out
their personal goals and later chose an existential goal as
the one most nearly descriptive of their own to a gain in
self-actualization as measured by the POI. No significant
relationship was found for the control subjects on the
same relationship (see Table 26).
TABLE 26
RELATIONSHIP OF EXPERIMENTAL SUBJECTS' AND CONTROL
SUBJECTS' CHOICE OF AN EXISTENTIAL GOAL TO GAIN
OR LOSS ON THE POI
Experimental Control
Gain Loss Gain Loss
Existential Goal 11* 0 6 1
Other Goal 3 3 4 3
*Significant at .05 level. Not significant.
110
Table 27 shows a significant relationship between the
therapists' choice of an existential goal for experimental
subjects and their gain on the POI. No significance was
found for the control subjects.
TABLE 27
RELATIONSHIP OF THERAPISTS’ CHOICE OF AN EXISTENTIAL
GOAL FOR EXPERIMENTAL AND CONTROL SUBJECTS TO GAIN
OR LOSS ON THE POI
Experimental
Gain Loss
Control
Gain Loss
Existential Goal 13* 0
Other Goal 2 3
5 1
6 3
^Significant at .05 level Not significant
Significance at the .05 level of confidence was also
found when the therapist and the subject both choose an
existential goal for the experimental subjects (Table 28).
TABLE 28
RELATIONSHIP OF MUTUALLY CHOSEN EXISTENTIAL GOALS TO
GAIN OR LOSS ON THE POI BETWEEN EXPERIMENTAL
AND CONTROL GROUPS
Experimental Control
Gain Loss Gain Loss
Mutual Existential Goals 10* 0 4 0
Mutual Other-Oriented Goals 1 2 2 2
*Significant at .05 level Not Significant
Ill
In summary, this analysis shows that although the
therapist's stated orientation had no significant relation
ship to subject's gain on the POI, a significant relation
ship did exist for experimental subjects, but not for con
trol subjects, to gain in self-actualization and the choice
of an existential goal by the therapist, subject, or both.
A decrease in anxiety, as is measured by the General
form of the MAACL, was considered to be a psychoanalytic
goal. To answer the question as to whether there is a re
lationship between gain on the anxiety scale and a thera
pist's psychoanalytic orientation, his choice of a psycho
analytic goal or a subject's choice of a psychoanalytic
goal, an analysis, as above, of the data was carried out.
No significant relationship was found between the
therapists' orientation and gain on the Anxiety scale
(see Table 29) .
TABLE 29
RELATIONSHIP OF THERAPISTS" ORIENTATION TO GAIN
OR LOSS ON THE GENERAL ANXIETY SCALE
Gain Loss N
Psychoanalytic Orientation 10 3 13
Other Orientation 10 6 19 3 subjects
did not
take test
Not significant
112
When the relationships of gain to subjects' and ther
apists' goal choices were evaluated, no significant rela
tionship was found between either's goal choices and gain
on the Anxiety scale. There were few psychoanalytic goals
chosen, especially by experimental subjects (three) or for
experimental subjects by the experimenters (three). Four
were chosen by control subjects and all of these were as
sociated with loss on the Anxiety scale. Five were chosen
for control subjects by their experimenters and four of
these were associated with loss or no gain.
Apparently, psychoanalytically oriented experimenters
who chose a goal from another orientation or whose sub
jects chose a goal from another orientation contributed to
the significant gain on the Anxiety scale. Choice of a
psychoanalytic goal, however, appeared to result in a loss.
One consistent finding for this anxiety scale was
apparent in the choice by subjects and therapists of be
havioristic goals. The experimenter-therapist with the be-
haviorisitic orientation did not administer the test to his
subjects, which limits this analysis, but the results as
tabulated are shown in Table 30.
With two exceptions in control subjects' goal choices,
all choices of a behavioristic goal were associated with
lower general anxiety scores on the MAACL. The lessening
of anxiety may then be more a behavioristic goal than a
Filmed as received
113
without page(s)_____
UNIVERSITY MICROFILMS.
CHAPTER V
SUMMARY, IMPLICATIONS, CONCLUSIONS AND
RECOMMENDATIONS
This study was concerned with investigating the ef
fectiveness of awareness of specific goals by individuals
in therapy with therapists of varying theoretical orienta
tions. Awareness of a goal was defined as having the sub
ject keep a written statement of his goal or goals in a
conspicuous location for daily referral. Although litera
ture and research studies abound with references to thera
peutic goals in general terms, no study had been conducted
previously into the function of simple but habitual re
ferral by the subject to the aims he is attempting to at
tain through therapy.
Because it was generally conceded that goals are
value systems (Lowe, 1969) or values are aspects of goals
(Buhler, 1962) , this study also attempted to determine whe
ther a tendency exists for subjects to adopt the value
system of their therapists and, specifically, whether the
goal-setting and awareness process significantly increased
this tendency.
The review of the literature cited opinions and re
search which affirmed the consensus that the theoretical
114
115
orientation espoused by the therapist reflects his value
system, which in turn influences the goal he chooses for a
subject. A further question which this study, then,
sought to answer was, "Will individuals in therapeutic
treatment indicate, at the end of the experimental period,
that the goal they have been seeking to attain is of the
same orientation as their therapist's and, specifically,
does the goal-setting and awareness process significantly
increase the number of such congruent indications?"
Summary of the Method
The sample population of the study were thirty-five
individuals who sought counseling or therapy. Nineteen of
these were experimental subjects and sixteen were control
subjects. Seven therapists or counselors acted as experi
menters in carrying out the design with those subjects who
had applied to each for his services. Each experimenter
randomly assigned his subjects as experimental or control.
The size of the sub-groups varied from three to eight
subj ects.
Fourteen sessions of therapy constituted the indivi
dual experimental period. Sessions were held in the pri
vate offices of the experimenters during the eight month
period of July, 1971 to March, 1972.
Three test instruments were used to measure the
116
effects of the goal-setting process : (1) the Personal
Orientation Inventory, a measure of self-actualization,
(2) both the Today and the General forms of the Multiple
Affect Adjective Check List, a measure of anxiety, depres
sion, and hostility, and (3) a Semantic Differential, a
measure of the meaning of valued concepts, which was used
both to ascertain therapeutic movement and congruency of
subject-experimenter values. The experimenter also com
pleted this latter instrument.
Pretesting occurred during the first or second ses
sion of therapy or counseling. At some point during these
sessions, experimental subjects were asked to write on a
white 4 x 6 card, the goal or goals they hoped to attain.
They were then given instructions for daily referral to the
card. At each following session, the experimenter in
quired as to their progress toward the goal, their refer
ral to it, and their desire to revise or add to it.
Before the third session the experimenter selected
for each subject (both experimental and control), a goal
from a list of Goal Statements, consisting of twelve pre
judged goals of which four were considered to be pertinent
general goals of psychoanalysis, four of existentialism,
and four of behaviorism.
Control subjects did not set or write out personal
goals.
117
At the last session all subjects took the posttests
and selected a goal from the List of Goal Statements which
most nearly described their considered goal or aim in
therapy. They also related behavior or the experimenter
recorded noted behavior which indicated full or partial at
tainment of that goal, All subjects were judged at the
conclusion of the session by the experimenter for attain
ment, partial attainment or no attainment of the goal and
for amount of movement toward the goal.
The data from the three test instruments was eval
uated by Fisher's t test on the IBM computer. Other data
was analysed by Fisher's Exact Test and the y2 test for
two independent samples.
Summary of Findings
Equivalence of the experimental and control groups
at the beginning of the study was established, for the t
test of the pretest means on the eighteen variables showed
no significant differences at the .05 confidence level be
tween the groups.
Gain scores of the three instruments revealed a gen
erally consistent trend in support of the primary hypothe
sis that individuals who write out and habitually refer
to the personalized goal they choose to attain from therapy
will show.more positive change than individuals who do not
do so.
118
The gains for the experimental group were signifi
cantly greater at the .03 confidence level on the inner-
directed scale of the Personal Orientation Inventory, which
is an over-all measure of self-actualization. Further an
alysis with gains and losses on the POI revealed that the
choice of an existential goal by the therapist, by the sub
ject, or by both was related at the .05 confidence level to
gain on the POI with experimental subjects. The therapists'
preferred orientation of existentialism was not related to
the gains.
On the two forms of the Multiple Affect Adjective
Check List, five of the six scales showed a more positive
gain for experimental subjects than for control subjects,
one of which was significant. The gains on the Anxiety
scale of the General form resulted in a difference, signi
ficant at the .02 confidence between the experimental and
control groups. Gains on this scale, however, were not re
lated to psychoanalytic orientation or goal choices by
therapists or subjects. There were indications that they
may have been related to behavioristic goal choices by
therapist and subjects.
Eleven concepts were measured for evaluative feelings
by the Semantic Differential. None of these reached the
.05 significance level. Mean gain scores for eight of
these, however, revealed definite trends, indicating more
119
positive gain by experimental subjects, A small loss by
the experimental group for the concept of My Ideal Self was
a function of the tendency for change in the direction of
equality with the ME concept. Additional findings indica
ted that the Ideal-Self concept tended to be more stable
than the ME concept for both experimental and control
groups.
Both experimental and control groups had loss rather
than gain scores on the Religion concept, but the experi
mental group lost less. Only the concept of Sex, there
fore, of the eleven concepts showed slightly more (.28 of
a point) gain by the control group than by the experimental
group.. The concept of Money reached the .06 probability
level and the concept of Life reached the .10 probability
level for differences in gain scores by the experimental
group.
Goal attainment and movement in therapy as judged,
were not statistically significant, but descriptively, the
experimental group showed more goal attainment and movement
than the control group.
The assumption that subjects tend to adopt the val
ues of their therapist was the basis for the second hypothe
sis of the study, which stated that this tendency would be
more pronounced by experimental subjects. This was not
verified for the entire group. The direction of shifts in
120
values either toward or away from therapists' values was
found to be related solely to individual therapists and
their subjects. The total shift was not significant, and
the differences in direction balanced out at a result of
approximately no shift by either group. In one sub-group
of eight subjects, the four experimental subjects did shift
toward their therapist's values at the .05 level of confi
dence .
For the third hypothesis it was predicted that experi
mental subjects would, more often than control subjects,
select as most descriptive of their therapeutic goals, a
general theoretical goal of the same orientation as their
therapists' selection for them. Relevant to this hypothe
sis, it was significant that therapists of existential or
behavioristic orientations selected like-oriented goals for
experimental subjects but those of psychoanalytic orienta
tion did not. The prediction that the experimental sub
jects’ choices would be congruent with their therapists'
choice was not statistically significant at the .05 level.
The prediction that experimental subjects would change
or revise their goal at least once was not born out. In
stead, it was found that these subjects each specified an
average of 2.16 original goals and worked toward the at
tainment of these without revision.
121
Implications and Conclusions
This study has been concerned with determining whe
ther goal-setting and continued awareness of goals serves
a motivating function in positive therapeutic movement. It
has also been concerned with the relationship of value sys
tems and theoretical systems or orientations to goals and
goal-setting. The results have indicated that this tech
nique or process does serve a valuable function. The total
sample was not large, so a broad generalization of the re
sults cannot be made. On the other hand, the variations
among the subjects and between the experimenters and the
settings were many, so the results are not limited to one
clearly defined sample. The one common definition of the
sample was that they were individuals seeking therapy or
counseling. Their ages, education, sex, presenting prob
lems, and marital status varied. They sought out their
therapists from those in private practice, in clinics, or
in a college counseling center, located in differing geo
graphical areas. Despite these variations and after ran
domization of the sample, the two groups were equal on the
eighteen variables at the beginning of the study.
The measuring instruments utilized in the study pre
sented certain advantages and disadvantages. The Personal
Orientation Inventory proved to be not only an effective
measure for change or movement but a valuable therapeutic
122
tool. Only one scale--the inner-directed scale--was used
for the gain scores, for (as explained in Chapter III) it
is considered to be a good over-all measure of self-actual
ization as 127 of the 150 statements on the test are in
volved in the scoring for this scale.
A self-actualized person is an inner-directed rather
than other-directed person. The inner-directiveness of
self-actualization includes self-actualizing values, flex
ibility on applying those values, sensitiveness of and
spontaneity in expressing feelings, self-regard and self
acceptance, awareness, acceptance of one’s aggression, and
capacity for intimate contact. These are primarily goals
of existentialism and involve both interpersonal and intra
personal types. The statistically significant gains by the
experimental group on this test seem to imply that the
goal-setting process is especially effective for movement
toward the facets of self-actualization. A disproportion
ate number of existential goals were chosen by both the
therapists and the subjects, but this choice for and by the
goal-setting subjects also contributed significantly to
gain on the POI. This test, as a measure of existentialism,
was effective. Difficulties arose in persuading teenagers
to retake the test. Their objections were that it was too
long and that they did not understand it.
The use of two forms of the Multiple Affect Adjective
123
Check List was probably unnecessary. A comparison of the
adjectives checked on the two forms was of interest thera
peutically, but the profile form was confusing to the ex
perimenters. As a measure of change or movement, it is
doubtful if the ’ ’Today" form is beneficial. This form has
proved effective in short-term, specific experiences, but
it may be possible that a time set of "now" or "today" does
not reflect the deeper, more permanent inner state of a
person towards which therapy and this study were directed.
The scores on its three scales were all in the direction of
the research hypothesis, but may, in some cases have been
a reflection of a more comfortable feeling in the "now"
with their therapist.
The General form had two interesting results: the
significant gain score on the anxiety scale and the com
parative, though not significant, loss by the experimental
group on the hostility scale, one of the two of the eight
een variables where this trend appeared. By far the most
common theme in the personal goals of the experimental sub
jects was overcoming some feeling of anxiety or fear. Al
though depression or hostility may underlie these affects,
apparently experimental subjects more readily (or at a
fourteen session stage of therapy) recognized a change in
their feelings of anxiety.
The lack of any significant results on the Semantic
122
tool. Only one scale--the inner-directed scale--was used
for the gain scores, for (as explained in Chapter III) it
is considered to be a good over-all measure of self-actual-
izaticn as 127 of the 150 statements on the test are in
volves in the scoring for this scale.
A self-actualized person is an inner-directed rather
than other-directed person. The inner-directiveness of
self-actualization includes self-actualizing values, flex
ibility on applying those values, sensitiveness of and
spontaneity in expressing feelings, self-regard and self-
acceptance, awareness, acceptance of one's aggression, and
capacity for intimate contact. These are primarily goals
of existentialism and involve both interpersonal and intra
personal types. The statistically significant gains by the
experimental group on this test seem to imply that the
goal-setting process is especially effective for movement
toward the facets of self-actualization. A disproportion
ate lumber of existential goals were chosen by both the
therapists and the subjects, but this choice for and by the
goal-setting subjects also contributed significantly to
gain on the POI. This test, as a measure of existentialism,
was effective. Difficulties arose in persuading teenagers
to retake the test. Their objections were that it was too
long and that they did not understand it.
The use of two forms of the Multiple Affect Adjective
124
Differential was disappointing, for the differences in mean
gain scores on seven of the concepts were considerable and
well over the range which Osgood designates as probably
significant at the .05 level. The variances were very laige
however. The design of this particular semantic differen
tial may have included too many bipolar scales of the eval
uative factor. Or the large variances may have been a
valid indication of the variance between value systems of
the general population.
Previously, attention was called to the fact that for
eight of the concepts mean gain was in the direction of the
research hypothesis and explanations given for the loss on
My Ideal Self and Religion. The experimental group's lower
mean score on the concept of Sex is accompanied by the
largest variance (standard deviation of 17.10) of the entire
set of concepts. An explanation of this wide variation in
experimental subjects is not obvious but must be due to
unknown individual differences in attitudes towards sex.
The statistical variances were not a factor in the
analysis of the semantic differential for shifts of sub
jects' values toward their therapists' values. (However,
if the therapists had taken a posttest on the instrument,
variances would have also been analysed which may be a
better statistical method.) The finding of little tendency
for such a shift to occur differs considerably from general
125
opinion and from the results of other research, as cited in
Chapter II, that patients tend to adopt the values of their
therapists. Lowe (1969) states that when congruence is
lacking between the values and goals of the patient and the
therapist, they find it difficult to agree on therapeutic
goals. Yet, in this study, agreement on goals, by experi
mental subjects was close to significance on all goals and
was significant for existential goals while agreement on
values definitely was not.
In research studies of value congruency various and
differing measuring instruments have been utilized. As a
result, different definitions and interpretations of value
systems may have been made. The previously cited study by
Welkowitz et al. (1967) which found that therapists and
their own patients were closer in value than randomly
paired therapists and patients seems a comprehensive and
well-designed study. They, however, used the Morris Ways
to Live and the Strong Vocational Interest Blank as the
measures of values. These instruments varyas measures and
as definitions of values from the instrument used in this
study.
The results of this study appear to be supportive of
Lowe's statement that congruency may be neither simple nor
straightforward, but vary among subjects and therapists.
They also appear to have implication for support of Cook's
126
C1966) suggestion that his results indicate there is a
curvilinear relationship between shifts toward closer
similarity and high, low, or median original similarity.
Primarily, the conclusion from the lack of subject-
experimenter value congruency seems to be that the experi
menters neither subtly nor explicitly imposed their values
on their experimental subjects, but rather allowed and en
couraged the subjects to explore and adopt their own.
The related subject of goal orientation congruency
which was investigated revealed some interesting and signi
ficant findings. That therapists do not consistently choose
a goal for their patients which is of their espoused orien
tation supports the recent survey by Weissman et al.
C19713. They found that therapists did not necessarily
utilize in practice the techniques and approach of their
stated theoretical orientation. However, the significant
tendency found in this study for the experimenters of exis
tential and behavioristic orientation to choose like goals
while psychoanalytic experimenters chose unlike goals sug
gests that other factors may be functioning and a definite
conclusion is unwarranted.
Perhaps there is relevance here for Fielder's (1954)
belief that the ability to create "the ideal therapeutic
relationship" is the primary reason for success being pos
sible with those of differing orientations. This reasoning,
127
in view of the findings, would indicate that the "ideal
therapeutic relationship" is more easily established when
the therapist is concerned about the patient's goal, elicits
it from him, and continues his interest in the patient's
view of his progress toward it.
Although the data on subject-therapist goal congru
ency was not significant for the experimental subjects,
there was a definite trend in that direction. The focus
on goals tended to lead to congruency of subject-therapist
goals. Perhaps it would seem more probable that, if pa
tients tend to adopt the goals of their therapists, they
would be most likely to do so when only the therapist's
goal was involved, as was the case with the control sub
jects. It might seem that a conflict could occur when
each sets goals. Yet in this study it appears that the
writing of the patient's goal resulted in the therapist
most often choosing a congruent goal. He accepted the pa
tient where he was, perhaps in terms of Carkhuff and Beren-
son's (1967) core therapist dimensions of empathy, positive
regard, genuineness, and concreteness.
The statements from the experimenter-therapists re
veal that for some subjects the written cards seemed to be
more of a handicap than a help. This is certainly pos
sible, probably to be expected with some patients, and a
consideration to be met when using the process. Several
128
research, and therapeutic questions are raised in the con
sideration. First, despite the objections of the subjects,
what are the results? Is there an age or sex differential
involved in acceptance of the card method? Individual dif
ferences even among those of like age and sex play a part
here as do individual differences in therapists. Some of
the differences in patients, however, may be the person
ality problems which caused them to seek therapy, and both
acquiescence to the method and objection to it may be
characteristic of a troublesome "life style." The depen
dent, other-directed patient may more willingly accept the
method than will the independent patient. Resistance to
acceptance of self-responsibility, suppression of a "true
goal," reaction against authority, and others may be re
vealed as characteristics of a patient for therapeutic
treatment. The investigator had one subject who reported
at the second session that she had thrown the card away
because she "didn't like goals." It developed that her
fear of failure was so great that she was terrified of all
goals and all reminders of goals. The incident of throwing
away the card helped to focus therapeutic direction quickly
(though it would not have been helpful for research).
Although the results of this study confirm the ef
fectiveness of the goal-setting method, they do not directly
answer the questions and considerations presented in the
129
above paragraph. Those are tasks for the future.
One interesting and intriguing question that has
arisen from this study is the relative merit of one or more
than one original goal. It was not possible in the analy
sis of this data to separate out quantitatively, the amount
of movement toward differing numbers of selected goals. In
reading the report forms, however, it became apparent that
the goals comprised three classes: interpersonal, intra
personal, and specific behavioral. A strong impression was
obtained from the reading that most success occurred when
at least one from each of the three classes was selected.
Choosing a combination such as improving relationwhip with
spouse, children, or peers; becoming more self-aware or
self-confident; and getting a job or raising a grade in
school seemed to be conducive to more successful movement
than choosing just one of these. Reinforcement, as an in
dividual sensed achievement in any one of these, was the
motivation for further effort. Also, although again only
an impression, the written goals on cards appeared to have
a function in this, for they served as reminders of the
progress made.
What are the broader implications from the findings
that a goal or goals written out and continually referred
to by the patient functions positively in therapy, es
pecially in movement toward self-actualization and anxiety
130
reduction? In Chapter I several important aspects of such
a finding were suggested. The use of this process as an
effective means of motivation is a primary implication. Mo
tivation is an intangible unknown in therapy, yet increased
motivation may well be the reason that individuals involved
in this experimental process mademore positive movement
than other individuals in treatment with the same therapists,
who also made some positive movement. It could have been
that the individual became aware of his responsibility in
the therapeutic process more quickly. Perhaps it was that
"forgetting" of both the goals and the special impact of
each therapy session was less possible because of the daily
rereading of the goal. One therapist stated that his pa
tients reported they did not need to refer to their cards
as they knew them so well they could just "imagine" the
goal (see Addendum). These patients may have been com
plaining, but they were remembering!
It is feasible that both the question of accounta
bility and measurement of success could be better answered
through recorded evaluations of goal attainment by pa
tients and therapists. It was evident in this study that
neither a patient nor a therapist wished to evaluate a
goal as completely not attained. A broader, more clearly
defined continuum for evaluation might have been more re
vealing and would certainly be necessary for long range
131
recording and research. The List of Goal Statements could
be utilized in the beginning of therapy by patients, as
well as by therapists, to give a more standardized goal se
lection criterion. Specific sub-goals under the general
goal or goals thus selected could be explicated by the pa
tient and the rest of the process then carried out. These
goal attainment records would then give individual thera
pists, clinics, institutions, and the concerned general
public a realistic view of the efficacy of therapy.
Most of the subjects in this study did not terminate
therapy at the conclusion of the fourteen experimental ses
sions. It is possible that at termination there would be
less discrepancy between the two groups in the measuring
instruments but there would be a difference in the length
of time the subjects of each had spent in therapy treat
ment. Several experimenters mentioned in their statements
that focusing on goals early seemed a help in starting the
therapeutic process and in keeping it from wandering un
necessarily. With the ever-increasing, need for psycholo
gical professional services, a means of shortening the
time for effective therapy is needed. The goal-setting
process may give the necessary direction to therapy and mo
tivation to patients to bring about earlier successful
treatment.
The purpose of this study was to investigate
132
therapeutic movement as a function of awareness of goals.
The evidence generated showed that awareness of goals
served an effective function with this sample population
and further research is merited. The additional purposes
were to investigate the relationship of values and goal-
orientation to goals. The evidence generated for these
did not support generally held opinions and past research
about similarity of patient-therapist value systems and
revealed some new findings on patient-therapist congruency
of goal orientation.
Recommendations
The following recommendations are made on tfte basis
of the experiment:
1. The goal-setting process as outlined in this study
or with modifications, be employed by therapists interested
in exploring the effectiveness of the method and/or in
utilizing a method for recording successes and failures of
therapeutic treatment.
2. Modification of the process include for standardiza
tion purposes the use of the List of Goal Statements with
subjects at the beginning and at the end of therapy. A
recommended further inclusion is an increased number of
intervals on the evaluation or judgment scale for goal
attainment.
3. This study should be replicated with a larger total
sample and larger sub-groups not only for statistical pur
poses but also so that dichotomization of personality
types, such as dependent-independent, resistant-nonresis-
tant, anxious-hostile, etc. could be carried out.
4. For further research on patient-therapist value con
gruency a second measuring instrument would produce more
information. The Personal Orientation Inventory with all
of its scales is recommended for this use in addition to
the semantic differential. Therapists should take both the
pretest and posttest of these.
5. Further research on the goal-setting process could be
designed so that the experimental period is not limited to
a set number of sessions but length of therapy is included
as an additional dependent variable.
6. Future research could investigate the relationship
between therapeutic movement and the number and types of
goals set by patients.
7. Cards more decorative than plain white ones could be
used for the written goals which might encourage subjects,
especially teenagers, to enjoy displaying them in a con
spicuous place.
ADDENDUM
TYPICAL GOALS AND EXPERIMENTERS' REACTIONS
The primary purpose of this study was to determine
the effectiveness in therapeutic treatment of the goal-
setting process which included having the experimental sub
jects write out a personal goal or goals on a card, refer
ring to this goal habitually and being able to revise it
whenever they chose during the experimental period. Some
significant gains and many positive directional trends were
found as evidence of the effectiveness of this purpose.
It seems, therefore, not only interesting, but expe
dient, to report some examples of patient goals that illus
trate typical types and their relationship to definite move
ment, to some movement, to gain or loss on the measuring
instruments, and to therapeutic orientation of the goals.
Additionally, because the goal-setting process of
the study was an innovation to the experimenters, they were
requested, at the time each completed the research, to de
scribe in a short paragraph their reactions to the use of
this process in therapy or counseling.
The examples of goals and the written reactions of
the experimenters are presented in this chapter.
134
135
Typical Goals
Example Number 1
Judgment: Goals partly attained. Definite Movement. Gain
on POI and Anxiety scales. Congruence (existen
tial) of subject and therapist goals.
Goals: (1) To be able to say "no." Stop trying to
please the whole world.
(2) To get a job.
Requisite behavior for attainment: Confidence, self-
assurance, ability to say "no," assertiveness.
Goals were revised to include: Understand myself better
and sustain personal closeness.
Behavior which illustrates partial attainment: Increased
self-awareness and self-confidence. Less in
authentic social behavior. Obtained a satis
fying job.
Example Number 2
Judgment: Goals attained. Definite movement. Gain on POI
and Anxiety scales. Congruence (existential)
of subject and therapist goals.
The subject is a married female in her twenties.
Goals: (1) Increase in positive verbal statements made
to myself about my behavior and me.
(2) More time spent with my children.
(3) Loss of fear of being hurt in a relation
ship .
Requisite Behavior: Reduction in self-depreciatory comments
like "what a stupid ass I am."
Behavior which illustrates attainment: Able to stop self-
depreciatory comments she made to herself about
herself and increased the positive aspects of
what she was able to do and be. Began spending
three times as much time with children. Ob
tained divorce, began dating and is thinking of
the possibility of marriage.
136
Example Number 3
Judgment: Goal attained. Definite movement. Gain on POI.
Congruence (behaviorism) of sub)ect-therapist
goals. Subject is single male in twenties.
Goal: To eliminate avoidance behavior by expressing
feelings of anger more directly.
Requisite Behavior: Talking to people about strong nega
tive feelings.
Behavior which illustrates goal attainment: Talking di
rectly to people with whom he works who do
things that bother him.
Example Number 4
Judgment: Goal attained. Definite movement. Gain on POI
and Anxiety scales. Congruence (existnetial)
of subject-therapist goals.
Subject is a married female in her forties.
Goal: Want my son to be able to function, enjoy life
and use his abilities. I want to help and un
derstand him.
Requisite Behavior: No more serious problems with peers,
at home and at school.
Goal was changed to a more personal one: I want to make
up my mind about my life, to make decisions,
and to start doing things towards changing.
Behavior which illustrates goal attainment: Stopped drift,
in marriage, with self and with children. Made
a decision regarding marriage.
Example Number 5
Judgment: Goal attained. Definite movement. Gain on POI
and Anxiety scales. Congruency (existential) of
subject-therapist goals.
Subject is a single female in early twenties.
Goals: (1) To more carefully define educational goals
of long-range value and determine the
means of attaining them.
137
(2) To reinforce my belief (through behavior) that
openness and truthfulness can be achieved and
maintained in the face of the present social
climate (which emphasizes alienation), thus al
tering it and me.
(3) To find happiness with myself.
Requisite Behavior: Find an educational goal. Be truth
ful in all situations. Be optimistic.
Behavior which illustrates attainment: Accepted at a
college in a major of her choice. Has an open and
truthful relationship with mother and boy-friend.
Feels better about herself and others.
Example Number 6
Judgment: Goal partly attained. Some movement. Loss on
POI, gain on Anxiety scale. Congruence (exis
tential) of subject-therapist goals.
Subject is a seventeen year old, single female.
Goal: To be less withdrawn.
Requisite Behavior: To say "hello" to more people at
school.
Goal was revised twice: To be less self-controlled--ex
press feelings at home. To become more inde
pendent of family.
Behavior which illustrates partial attainment: Has been
able on several occasions to leave for a week
end and have somewhat better relationships with
boys.
Example Number 7
Judgment: Goal partly attained. Definite movement. Loss
on POI, no change on Anxiety scale. Congruency
(psychoanalytic) of subject-therapist goals.
Subject is divorced male in mid-twenties.
Goals: (1) Learn to express my feelings to others, or
relating to others better.
_ _ (2) Find a meaning for life.
138
(3) To accept and understand people.
(4) Find happiness in school and future
vocation.
Requisite Behavior: Stop rejecting and depreciating people
by overcoming fear of them.
Goal changed to: Face my real problem--myself.
Behavior which illustrates partial attainment: I have be
come a more accepting person and have tried to
relate with more sensitivity to people. I
have volunteered to become a "Big Brother."
Example Number 8
Judgment: Goal attained. Definite movement. Gain on POI
and anxiety scales. Incongruence (behaviorism-
existential) of subject-therapist goals.
Subject is a divorced female in early fifties.
Goal: To prove to myself I can do something other than
care for a home. I don't want to feel infer
ior to my peers.
Requisite Behavior: Take the job offer and be successful.
Appear in public without nausea.
Behavior which illustrates goal attainment: I'm working
full time and enjoying it--no more nausea. I
haven't missed a day of work.
These examples are illustrative of the type of goals
specified by the subjects of which they were habitually
aware. Other explicit goals were to control strong emo
tions (cry less frequently), to accept my mother's impend
ing death, to accept sexuality (homosexuality), to be able
to respond with openness and closeness in sex and all re
lationships, to stay in school (overcome school phobia),
and to learn my own worth to myself and others. The entire
139
group would seem to be representative of problems for
which professional help is sought.
As previously mentioned, the goals appear to comprise
three classes: interpersonal, intrapersonal and specific
behavioral. As will be noticed in the illustrations, sub
jects often chose, or made additions or revisions to in
clude,, one from each class. It is highly probable that
this allowed greater therapeutic flexibility and by habi
tually becoming aware of these, gain in one was perceived
readily and served as motivation for further therapy, lead
ing to other goal attainment.
Experimenters' Reactions to the
Goal-Setting Process
As stated above, the experimenters each sent to the
investigator, in reply to a request, a brief statement as
to their reactions to the experimental goal-setting pro
cess. The investigator felt that the statements were
thoughtful, honest and helpful. These appear below in the
order in which they were received.
Number 1
MThe use of a card to make a patient aware of thera
peutic goals produces certain definite advantages, parti
cularly during the early stages of therapy. It seems to
me that with many clients there is some confusion during
the early stages about therapeutic goals. Helping them
140
to concentrate on one, or a few, specific goals by exposing
them to regular reminders of the goal, helps them to focus
in such a way as to reduce the confusion they are feeling
about their problem. Producing some therapeutic gain in
the one goal helps the client to gain some confidence in
the therapeutic process and provides a better set for fur
ther therapy.
"Whether the constant or frequent reminder of the goal
is therapeutic per se, I have not been able to establish
since I only had two experimental subjects. Feedback from
the subjects seemed to indicate that they thought it would
have some effect on their change by helping them to focus
on something specific, whereas their tendency in the dis
turbed state was to generalize their problems to a wide
and vague range of behaviors and emotions.
"I found that during the later stages of therapy,
clients tended not to refer to the cards as often as they
did during the early stages. Their excuse was that they
knew the card so well, it was not necessary to look at it
because they could just 'imagine it.' I insisted that
they did actually look at the card, but this may not be
necessary."
Number 2
"I had no problems having the clients set goals the
first session since that is routine for the type of ther-
141
apy I do."
Number 3
"After working with four experimental subjects, I
found that two of my counselees were very enthusiastic
about writing down the goals that they wanted to attain.
One of the other subjects, however, thought that reading
his goal daily was unnecessary; he felt that this method
did not help him attain it at all. The fourth person
rarely wanted to discuss the stated goal; I do not think
that writing down a goal and placing it in a conspicuous
place helped him whatsoever either. Whether I would again
use this technique of asking a client to write and read
his goalCs) depends on the person with whom I am counsel
ing; I discovered that this method was only fifty percent
effective."
Number 4
"Setting goals has been of some help in focusing with
patients and defining clearly diagnosis and probable prog
nosis. It kept the patient (and perhaps worker) focused
on what she wanted to accomplish, what had already been
solved, and what was left to be done. It gave the patient
some implied idea of time being a factor and there not
being unending sessions. It kept one particular patient
from drifting and rambling."
142
Number 5
"1 felt the design was excellent and the experiment
well planned. I'm interested in seeing the results of
setting goals and’ their implications on therapy. I per
sonally felt the idea of setting goals is good; it allowed
my clients to be aware constantly of where therapy was
going for them. The tests were helpful to therapy, parti
cularly at the beginning and gave a definite direction to
therapy."
Number 6
"You requested a statement regarding my feeling about
using goal-setting. I find that most of the patients that
I see do not know how to focus on an area of difficulty
and consequently present an amorphous mass of complaints.
As a result of this I think therapy is critically length
ened. By using goal statements I find that my patients
begin to focus on their own feelings and behavior rather
than emphasize the complaints. I think goals are parti
cularly effective once rapport has been established."
Number 7
"My reaction to the goal-setting procedure of your
study was mixed. I generally liked the idea of working
with the patients on fairly specific goals. However, the
cards proved to be more trouble than I think they are
worth. One patient felt that the cards might have been
143
helpful, but two other patients "lost" their cards and
seemed resistant to follow up on them. I think the es
sence of goal-setting takes place between patient and
therapist in their interpersonal transaction and needs no
card to reinforce the goals with the patient. In fact,
the cards may have the effect of making the patient feel
'locked in' to his goals."
The individuality of the experimenters is evident in
these statements, but certain common themes seem to be re
current among them. First, they convey the sincerity with
which these individuals carried through on the research
and their involvement in it. Also, they seem to reiterate
agreement that the setting of goals, especially at the be
ginning of therapy, tended to serve as a focal point for
the therapy, resulting in less emphasis on complaints and
more emphasis on the function of therapy itself. The card
and the necessity for continual referral to it seemed to
cause problems in some cases, yet apparently served effec
tively in others.
None of the statements refer to age of the patients
as being either a helpful or deterring factor in the ef
fectiveness of the method. The investigator, who carried
the process through with eight patients but did not in
clude her results in the data, found that with teen-agers
of fifteen to eighteen years each step of the process was
144
more difficult, expecially at the beginning for goal-set
ting and at the end for posttesting. Furthermore, the de
pendent patient seemed to begin with enthusiasm and remain
enthusiastic even as independence was being experienced,
whereas the independent patient reacted with characteristic
reluctance to follow a set routine. Patients*reactions to
the method, therefore, seem to be a focal point for thera
peutic direction.
B I B L I O G R A P H Y
145
BIBLIOGRAPHY
1. Allport, G. W. Personality and Social Encounter.
Boston: Beacon Press, 1960.
2. Allport, G. W. "Psychological Models for Guidance."
Harvard Education Review, 1962, 32, 373-381.
3. Anderson, W, "Behavioral Counseling by John D.
Krumboltz and Carl E. Thoresen." Personnel
and Guidance Journal, 1970, 49, 78-79.
4. Atkinson, J. W., and Feather, N. T. (Eds.) A Theory
of Achievement Motivation. New York: John
Wiley, 1966,
5. Bednar, R. L. "Persuasibility and the Power of
Belief." Personnel and Guidance Journal, 1970,
48, 8, 647-653.
6. Bednar, R. L., and Parker, C. A. "Client Susceptibility
to Persuasion and Counseling Outcome." Journal
of Counseling Psychology, 1969, 16, 415-420.
7. Bergin, A. E. "The Evaluation of Therapeutic Out
comes." In A. E. Bergin and S. L. Garfield,
(Ed.) Handbook of Psychotherapy and Behavior
Change. New York': John Wil-ey , 1971, 217-270 .
8. Berlyne, D. E., and McDonnell, P. "Effects of Height
ened Arousal on Human Exploratory Behavior."
Canadian Journal of Psychology, 1963, 17, 398-410.
9. Brammer, L. M., and Shostrom, E. L. Therapeutic
Psychology: Fundamentals of Counseling and
Psychotherapy. New York: Prentice-Hall, 1960.
10. Bryan, Judith, and Locke, E. "Goal Setting as a Means
of Increasing Motivation." Journal of Applied
Psychology, 1967, 51, 274-277.
11. Bugenthal, J. F. T. The Search for Authenticity. New
York: Holt, Rinehart and Winston, 1965^
146
147
12. Buhler, Charlotte. Values in Psychotherapy. New York
The Free Press of Glencoe, 1962.
13. Buhler, Charlotte. "Psychology and the Image of Man."
Psychotherapy: Theory, Research, and Practice,
1968, 5, 89-94.
14. Butler, J. M. "Self-Ideal Congruence in Psychotherapy.”
Psychotherapy: Theory, Research and Practice,
1968, S, 13-17.------ ~ -----------------------
15. Butler, J. M., and Haigh, G. V. "Changes in the Rela
tion between Self-Concepts Consequent Upon
Client-Centered Counseling." In C. R. Rogers,
and R. Dymond (Eds.) Psychotherapy and Person
ality Change. Chicago": University of Chicago
Press, 1954.
16. Carkhuff, R. R., and Berenson, B. G. Beyond Counsel
ing and Therapy. New York: Holt, Rinehart and
Winston, 1967.
17. Cofer, C. N., and Appley, M. H. Motivation: Theory
and Research. New York: John Wiley, 1964.
18. Cook, T. E. "The Influence of Client-Counselor Value
Similarity on Change in Meaning During Brief
Counseling." Journal of Counseling Psychology,
1966, 13, 77-81.
19. Cooke, G. "The Efficacy of Two Desensitiz&tion Pro
cedures: An Analogue Study." Behavior Research
and Therapy, 1966, 4, 17-24.
20. Corey, D. Q. "The Use of a Reverse Format in Now
Psychotherapy." Psychoanalytic Review, 1966,
53, 108-126.
21. Damm, V. J. "Overall Measures of Self-Actualization
Derived from the Personal Orientation Inventory"
Educational and Psychological Measurement, 1969,
29, 977-981.
22. Dollard, J., and Miller, N. E. Personality and Psy
chotherapy . New York: McGraw-Hill, 1950.
23. Ehrle, R. A. "Performance Contracting for Human
Services." Personnel and Guidance Journal,
1970, 49, 119-122. ~~
148
24.
25.
26.
27.
28 .
29 .
30.
31.
32.
33.
Endler, N. S. "Changes in Meaning during Psychother
apy as Measured by the Semantic Differential."
Journal of Counseling Psychology, 1961, 8,
105-111.
Eysenck, H. J. "The Effects of Psychotherapy: An
Evaluation." Journal of Consulting Psychology,
1952, 16, 319-32T:
Eysenck, H. J. Handbook of Abnormal Psychology.
London: Pitman, 1960.
Fielder, F. E. "A Comparison of Therapeutic Relation
ships in Psychoanalytic, Nondirective, and Ad
lerian Therapy." Journal of Consulting Psycho- -
logy, 1950, 14, 436-445.
Fiske, D. W. , Luborsky, L., Parloff, M. B., Hunt, H.F.,
Orne, M. T., Reiser, M. F., and Tuma, A. H.
"Planning of Research on Effectiveness of Psy
chotherapy." American Psychologist, 1970, 25,
727-737.
Ford, D. H., and Urban, H. B. Systems of Psychother
apy . New York: John Wiley, 1965.
Frank, G. H., and Hiester, D. S. "Reliability of the
Ideal-Self-Concept." Journal of Counseling
Psychology, 1967, 14, 356-357.
Frankl, V. E. Man's Search for Meaning. New York:
Washington Square Press, 1963.
Friedman, C. J., and Gladden, J. W. "Objective Mea
surement of Social Role Concepts via the
Semantic Differential." In J. G. Snider and
C. E. Osgood (Eds.) Semantic Differential
Techniques. Chicago: Aldine, 1969, 484-492.
Gendlin, E. T. "Values and the Process of Experien
cing." In A. R. Mahrer (Ed.) The Goals of
Psychotherapy. New York: Appleton-Century-
Crofts, 1967, 180-205.
34. Glad, D. D. Operational Values in Psychotherapy.
New Yorlc: Oxford University Press, 1959.
149
35.
'36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
Goldberg, F. H., Lesser, S. R., and Schulman, Rena.
"A Conceptual Approach and Guide to Formulating
Goals in Child Guidance Treatment." American
Journal of Orthopsychiatry, 1966, 36, 125-133.
Heinrich, Barbara A. "Motivation and Long-Term
Memory." Psychon. Sci., 1968, 12,149-150
Hersch, C. "The Discontent Explosion in Mental
Health." American Psychologist, 1968, 23,
497-506.
Jahoda, Marie. Current Concepts of Positive Mental
Health. New York: Basic Books, 19 58.
James, W. "Attention." In P. Bakan (Ed.) Attention,
Princeton: D. Van Nostrand, 1966, 3-2T.
Kaul, T, J. and Parker, C. A. "Suggestibility and
Expectancy in a Counseling Analogue." Journal
of Counseling Psychology, 1971, 18, 536-541.
Knapp, R. R. "Relationship of a Measure of Self-
Actualization to Neuroticism and Extroversion."
Journal of Consulting Psychology, 1965, 29,
TM^TTT.------------ ^
Krathwohl, D. R., Bloom, B. S., and Masia, B. B.
"Taxonomy of Educational Objectives." Handbook
II. Affective Domain. New York: David McKay,
T96T!
Krause, M. S. "Behavioral Indexes of Motivation for
Treatment." Journal of Counseling Psychology,
1967, 14, 426-435.
Krumboltz, J. D. Stating the Goals of Counseling.
California"Counseling and Guidance Association,
1966, No. 1.
Landfield, A. W., and Nawas, M. M. "Psychotherapeu
tic Improvement as a Function of Communication
and Adoption of Therapist's Values." Journal of
Counseling Psychology, 1964, 11, 336-341.
Latscha, R. "Tests of Significance in a 2 x 2 Con
tingency Table: Extension of Finney's Table."
Biometrika, 1953, 40, 74-75.
150
47. Lazarus, A. A. ’’Group Therapy of Phobic Disorders by
Systematic Desensitization." Journal of Ab
normal Social Psychology, 1961, 63, 504-510.
48. Ligon, E. M. A Greater Generation. New York: Mac
millan, 1948 .
49. Locke, E., and Bryan, Judith. "Performance Goals as
Determinants of Level of Performance and Bore
dom." Journal of Applied Psychology, 1967,
51, 120-130.
50. Loevinger, Jane, and Ossorio, A. "Evaluation of Ther-
apy by Self-Report: A Paradox." Journal of
Abnormal Social Psychology, 1959, 58, 392-394.
51. Lowe, C. M. Value Orientations in Counseling and
Psychotherapy. San Francisco: Chandler, 1969.
52. Luria, A. R. The Mentally Retarded Child. Oxford:
Pergamon Press, r9'63.
53. Luria, Zella. "A Semantic Analysis of a Normal and
a Neurotic Therapy Group." Journal of Abnormal
Social Psychology, 1959, 58, 216-220.
54. Mahrer, A. R. (Ed.) The Goals of Psychotherapy. New
York: Appleton-Century-Crofts, 1967.
55. Maltzman, I., and Raskin, D. C. "Effects of Indivi
dual Differences in the Orienting Reflex on
Conditioning and Complex Processes." In P.
Bakan (Ed.) Attention. Princeton: D. Van
Nostrand, 1966, 95-121.
56. Maslow, A. H. Motivation and Personality. New York:
Harper, 19 54.
57. May, R. "Contributions of Existential Psychotherapy."
In R. May, E. Angel, and R. F. Ellenberger (Eds.)
Existence : A New Dimension in Psychiatry and
Psychology. New York: Basic Books, 1958,
37-92.
58. McNair, D. M., and Lorr, M. "An Analysis of Professed
Psychotherapeutic Techniques." Journal of
Consulting Psychology, 1964a, 28"^ 265-271.
151
59. McNair, D. M., and Lorr, M. "Three Kinds of Psycho
therapy Goals." Journal of Clinical Psychology,
1964b, 20, 390, 393.
60. Meltzoff, J., and Kornreich, M. Research in Psycho
therapy . New York: Atherton Press, 1970.
61. Michaux, W. W., and Lorr, M. "Psychotherapists'
Treatment Goals." Journal of Counseling Psy-
chology, 1961, 8, 250-254.
62. Nelson, R. C. "A Manual for Use with the Semantic
Differential." Unpublished Master’s Thesis,
University of Southern California, 1964.
63. Ofman, W. V. "Psychotherapy as a Humanistic Existen
tialist Encounter: Key Elements of the Basis
for Practice." Monograph. Los Angeles: Psy
chological Affiliates Press, 1970.
64. Ohlsen, M. M. Group Counseling. New York : Holt,
Rinehart, and Winston, 1970.
65. Osgood, C. E., Suci, G. J., and Tannenbaum, P. H.
The Measurement of Meaning. Urbana: University
of Illinois Press, 1957.
66. Pascal, G. F. , and Zak, M. "Psychotherapeutics: Suc
cess or Failure." Journal of Consulting Psy
chology , 1956, 20, 325-331.
67. Patterson, C. H. Theories of Counseling and Psycho
therapy . New York: Harper and Row, 1966.
68. Paul, G. L. "Insight Versus Desensitization in Psy
chotherapy Two Years After Termination."
Journal of Consulting Psychology, 1967, 31, 333-
69. Peris, F. S. Gestalt Therapy Verbatim. Lafayette,
California: Real People Press, 1969.
70. Rogers, C. R. Client-Centered Therapy. Boston:
Houghton, Mifflin, 1951.
71. Rogers, C. R. On Becoming a Person. Boston:
Houghton, Mifflin, 1951.
152
72. Rogers, C. R. "Toward a Modern Approach to Values :
The Valuing Process in the Mature Person." In
C. R. Rogers and Barry Stevens, Person to Per
son : The Problem of Being Human" Lafayette,
California: Real People Press, 1967, 13-28.
73. Rogers, C. R., and Dymond, Rosalind F. (Eds.) Psycho
therapy and Personality Change. Chicago:
University of Chicago Press, 1954.
74. Rosenbaum, M., Friedlander, Jane, and Kaplan, S. M.
"Evaluation of Results of Psychotherapy."
Psychosomatic Medicine, 1956, 18, 113-132.
75. Rosenthal, D. "Changes in Some Moral Values Following
Psychotherapy." Journal of Consulting Psychol-
ogy, 1955, 19, 431-436.
76. Ryan, V. L., and Gizynski, Martha. "Behavior Therapy
in Retrospect: Patients’ Feelings about their
Behavior Therapies." Journal of Consulting
Clinical Psychology, 1971, 37, 1-9~i
77. Schroeder, Pearl. "Client Acceptance of Responsi
bility and Difficulty of Therapy." Journal of
Consulting Psychology, 1960, 24, 467-471.
78. Seeman, J. "Ends and Means?" Contemporary Psychol
ogy , 1969, 14, 646-648.
79. Shostrom, E. L. Manual, Personal Orientation Inven
tory . San Diego, California: Educational and
Industrial Testing Service, 1968.
80. Siegel, S. Nonparametric Statistics. New York:
McGraw-Hill, 1956.
81. Skinner, B. F. Science and Human Behavior. New
York: Macmillan, 1953.
82. Singer, E. Key Concepts of Psychotherapy. New York:
Basic Books, 19 70.
83. Stefflre, B. Theories of Counseling. New York:
McGraw-Hill, 1965.
84. Strupp, H. H. "An Objective Comparison of Rogerian
and Psychoanalytic Techniques." Journal of
Consulting Psychology, 1955, 19, 1-7.
153
85. Strupp, H. H. "The Psychotherapist's Contribution to
the Treatment Process." Behavioral Science,
1958a, 3, 34-67.
86. Strupp, H. H. "The Performance of Psychoanalytic and
Client-Centered Therapists in an Initial Inter
view." Journal of Consulting Psychology, 1958a,
22, 265-274.
87. Sullivan, H. J., and O'Hare, R. (Eds.) Accountability
in Pupil Personnel Services: A Process Guide
for Development of Objectives. Fullerton, Calif
ornia: California Personnel and Guidance
Association, 1971.
88. Sundland, D. M., and Barker, E. N. "The Orientations
of Psychotherapists." Journal of Consulting
Psychology, 1962, 26, 201-212.
89. Szasz, T. S. The Myth of Mental Illness. New York :
Hoeber-Harper, 1961.
90. Tolman, E. C. Purposive Behavior in Animals and Men.
New York: Appleton-Century, 193 2.
91. Truax, C. B., and Carkhuff, R. R. Toward Effective
Counseling and Psychotherapy. Chicago: Aldine,
1967.
92. Union College Character Research Project. An Intro
duction to the Character Research Program Cur
riculum" Schenectady, New York, 1962.
93. Urban, H., and Ford, D. "Some Historical and Concep
tual Perspectives on Psychotherapy and Behavior
Change." In A. E. Bergin and S. L. Garfield
(Eds.) Handbook of Psychotherapy and Behavior
Change. New York: John Wiley, 1971, 4-47.
94. Varble, D. L., and Landfield, A. W. "Validity of the
Self-Image Discrepancy as a Criterion Measure
for Success in Psychotherapy--A Replication."
Journal of Counseling Psychology, 1969, 16,
150-156.
95. Vordenberg, W. "A Technique for Goal-Setting in
Group Work." Personnel and Guidance Journal,
1970, 48, 479-ZTW;----------------------------
154
96.
97.
98 .
99.
100.
101.
102 .
103.
104.
105.
106.
Wallach, M. S., and Strupp, H. H. "Dimensions of
Psychotherapists' Activities." Journal of
Consulting Psychology, 1964, 28, 120-125.
Webb, A. P., and Harris, J. T. "A Semantic Differ
ential Study of Counselors in an NDEA Institute.'
Personnel and Guidance Journal, 1963, 260-263.
Weissman, H., Goldschmid, M., and Stein, D. "Psycho
therapeutic Orientation and Training: Their
Relation to Practices of Clinical Psychologists.'
Journal of Consulting Clinical Psychology, 1971,
37, 31-37.
Welkowitz, Joan, Cohen, J., and Ortmeyer, D. "Value
System Similarity: Investigation of Patient-
Therapist Dyads." Journal of Consulting Psy
chology , 1967, 31, 48-55.
Wellman, F. E. "The National Study of Guidance."
Unpublished Preliminary Report. University
of Missouri, 1970.
Weston, N. B. "The Adaptability of the Webb-Harris
Word Meaning Test to Tape Recorded Administra
tion for Use with Students with Inadequate Read
ing Ability." Unpublished Master's Thesis.
Los Angeles: University of Southern California,
1966.
Winter, Sara K., Griffith, C.J., and Kobb, D. A.
"Capacity for Self-Direction." Journal of
Consulting Clinical Psychology, Iyb8, 32, 35-41.
Winter, W. D. "Values and Achievement in a Freshman
Psychology Course." Journal of Educational
Review, 1961, 54, 183-186.
Wolff, W. "Fact and Value in Psychotherapy." Ameri
can Journal of Psychotherapy, 1954, 8, 466-486.
Zaporozhets, A. V. "The Development of Conscious
Control of Movements." In L. G. Voronin (Ed.)
Orienting Reflex and Exploratory Behavior. Wash-
ington, D.C. : American Institute of Biological
Sciences, 1965, Russian Monograph Number 3.
Zuckerman, M. "The Development of an Affect Adjec
tive Check List for the Measurement of Anxiety."
Journal of Consulting Psychology, 1960, 24,
457-462.
APPENDICES
155
APPENDIX A
Subject Information Form
Experimenter (Therapist) Information Form
Manual of Instructions for the Therapist or Counselor
Acting as Experimenter for Goal Setting Research
Subject Record Form
Experimenter Goal Record Form
List of Goal Statements for Experimenter Use
List of Goal Statements
156
157
SUBJECT INFORMATION FORM
Experimental Subject______
Control Subject_______
Name_______________________________________________________
Address_______________________________________Phone________
Birthdate________Age______Race______________ M_____F______
Education (Last grade completed)______Occupation_________]
Socio-Economic Class: Lower___Lower Middle____Upper
Middle Upper______ (Or parents if subject is a teen
ager)
Marital Status: Single Married Divorced Widowed___
Common Law____
Persons in Home: Spouse Parents Siblings Children__
Other
Experimenter (Therapist)
158
EXPERIMENTER (THERAPIST) INFORMATION FORM
Name_______________________________________M______F___
Business Address_____________________________________
____________________________________ Phone____________
Education:
No. years of graduate study___________
Degrees granted: BA___BS____MA____MS____MSW____
Ed.D. Ph .D. Other_____
Are you presently working toward a degree?
Yes No Which one?_____
No. years of Experience______
Which one of the following most nearly labels your
theoretical orientation? Psychoanalytic_____
Behavioristic Non-Directive Existentialistic
159
MANUAL OF INSTRUCTIONS FOR THE THERAPIST OR
COUNSELOR ACTING AS EXPERIMENTER FOR
GOAL SETTING RESEARCH
General Information:
Subjects of each experimenter will be the first 8-10 per
sons^ Fifteen years of age and older, who have completed
any required screening process and have been accepted for
individual therapy.
1. Each experimenter is to assign these
subjects alternately to the experi
mental and control groups.
2. A married couple can be considered as
one subject. Only one member of the
pair need be tested, for the data from
one only will be used.
The Experimental Period will constitute fourteen (14) ses-
sions with each subject. Subjects who remain in therapy
eight sessions may not be dropped from the experiment. It
will be necessary for either the experimenter or the re
searcher to make special efforts to administer the post
testing and evaluation to any subject who drops therapy
between the eighth and the fourteenth session.
The Average Amount of Time which this research will in-
volve for each subject is estimated to be four hours in
the fourteen sessions. This includes two hours of test
administration. Work with Control subjects will take an
additional one hour and work with Experimental subjects
will take an additional three hours.
Materials and Forms
With the exception of the two published tests, the mater
ials and forms have been color coded for your convenience
in using. The color code:
Yellow: For use with and for information about
subj ects.
160
Blue: For your exclusive use as the experimenter.
Green:
White:
Manual of
Instructions
A special form for your use in recording
your selected goal for each subject and
the goal record of Control subjects.
4 x 6 cards for the Experimental subjects'
written goals.
Please read carefully as this will
take you through the steps of the ex
periment. Your cooperation will be
greatly appreciated. Also please con
tact me with any further questions you
may have.
Test Instruments
1 .
2 .
Personal Orien
tation Inven
tory c po i)
Multiple Affect
Check List
(MAACL)
2 forms: Today
and General
A published test designed to measure
facets of self-actualization. In
cluded in your materials are one (1)
manual, two (2) test booklets and IBM
answer sheets (two in each individual
folder for pre and posttesting and a
few extras].
A published test on which the subject
indicates first how he feels at the
present time and then how he feels
generally. Both forms should be given
at the same session. This test mea
sures feelings of anxiety, hostility,
and depression. Included in your ma
terials are one (1] manual and answer
sheets for each form (two of each in
each individual folder for pre and
posttesting and a few extras).
Note: Profile forms are available for the above two tests.
If the answer sheets are forwarded to me after the first
administration, I will score them and send the profile
to you.
3. The Semantic This test is designed to measure the
Differential meanings of certain concepts both to
you and to the subjects. It is in
booklet form, and your materials in
clude one blue booklet for your self
administration, 16 yellow booklets
101
Forms
Experimenter
(Therapist) In
formation Form
(Blue) One Only
(two in each individual folder) and
a few extras.
Your identification form and informa
tion about your education and exper
ience .
One copy of each of the following forms is in the indivi
dual subject’s folder. You will also find a few extra
copies among your materials. Please contact me if you
need more.
List of Goal
Statements for
Experimenter Use
(blue)
List of Goal
Statements
(Yellow)
Experimenter
Goal Record
Form (Green)
5. Subject Infor
mation Form
(Yellow)
6. Subject Record
Form (Yellow)
A list
goals.
of 12 general therapeutic
A list of the above 12
posed into lay wording
subj ects.
goals trans-
for use with
Form on which to record your goal
chosen for an individual subject from
No. 2 above. Also, goal movement
record for control subjects.
Identification and demographical in
formation about the subject.
Record of dates of sessions and test
administrations for use with both Ex
perimental and Control Subjects. Also
Goal Setting Record for recording
goals and evaluation of Experimental
Subj ects.
Procedures for the Experimenter
For Yourself:
Fill out the Experimenter Information Form. (Blue)
2. Administer the Semantic Differential test in the blue
booklet to yourself.
162
3. Mail or give both of the above to Marion Luenberger.
For Experimental Subjects:
1. Fill out the Subject Information Form and the first
lines of the Subject Record Form. (Yellow.)
2. Talk to the subject about the Goal Setting Process
and have him write out his goal(s) on a 4 x 6 card,
preferably on the first and not later than the third
sess ion.
(A possible way of introducing this subject is to
say, "We will be working together for some time,
and it would be helpful to know what it is you may
wish to achieve through this therapy (counseling)
--what goal or goals you will be striving to attain.
We believe there may be merit in your stating (ver
balizing) a goal and reminding yourself of it often.
I would like your cooperation in working with me
on this plan by wrriting out your goal on this card,
taking it home and putting it in a place where you
will see and read it every day. Perhaps you could
attach it to your bathroom mirror, to a closet door,
or to the refrigerator. As we go along, we will
talk about your goal. You may find that you want
to modify it later, and you may do so whenever
you wish."
Each experimenter may use his own version of the
above. Original goals, of course, will vary, but
may be similar to the following examples:
a. To get along better with my spouse (child
or parent)
b. To feel better about myself.
c. To do better in my school work.
d. To understand why I am so depressed, anxious
or worried.
e. To be more at ease in my relationships with others.
f. To handle my children better.
g. To find a meaning for my life.
3. After the goal is written out, ask the subject to state
a specific behavior which would signify to him that he
had attained this goal.
163
4. Record the subject's goal and the requisite behavior
for its attainment in the appropriate section (Goal
Setting Record) of the Subject Record Form (yellow).
5. Select one goal from the List of Goal Statements for
Experimenters (blue) which you consider most nearly
expresses your conceived goal for this subject on or
before the third session. Record it and the necessary
identifying data on the Experimenter Goal Record Form
(green). File both forms in the subject's folder.
6. Administer the POI, the Today and the General forms
of the MAACL, and the Semantic Differential (yellow)
to the subject. The POI requires about 25 minutes
and the MAACL and the Semantic Differential about 10
minutes each. Vary the order in which these are given
among subjects. The three tests must be completed
before the beginning of the third session, but other
wise scheduling may be at your convenience. Suggested
plans for the scheduling are:
a. Make pre-arrangements for the subject to re
main about 50 minutes after the first or the
second session or to come 50 minutes early for
the first, second or third session for self
administration.
b. Shorter periods of time may be scheduled be
fore or after two sessions.
c. One or more tests may be completed in your pre
sence during the therapeutic hour so that part
of sessions one, two and the beginning of three
may be used for this.
7. At each therapy session with the experimental sub
ject take a few moments to discuss his goal. Inquire
about his daily referral to it, his progress towards
its attainment and his possible desire to change it.
DO NOT suggest your ideas for these changes. Occa
sionally, a subject may have difficulty in verbalizing
a word or two which would clearly represent his
thought and you may assist him with this.
(This discussion of goals may be inappropriate during
a particular session and will necessarily be omitted.
No more than four such omissions are allowable.)
Revised goals are to be written by the subject on a
new 4 x 6 card and taken home for daily referral.
164
8. Record the revised goal on the Subject Record Form,
Page 2.
9, The experimental period will end with the fourteenth
(14th) session, and the final necessary steps will
involve a larger proportion of the session time.
a. Goal attainment information for experimental
subjects is to be recorded under the section
headed "14th Session" on the Subject Record
Form, page 2. Refer to the last goal chosen
by the subject and ask him for his evaluation
of his having attained it. Follow through
with the other four parts of the section being
especially careful to obtain specific examples
of illustrative behavior. Remember that this
record consists of subject statements and eval
uations, except for your judgment of movement.
b. Have the subject read the List of Goal State
ments (for Subjects--the yellow form) and check
the one which most nearly describes the goal he
was last or is still seeking to attain. If he
should insist that more than one of these is
pertinent to his goal, have him rank them as
"1" for most nearly, "2" for next most nearly,etc.
c. Re-administer the three tests: POI, MAACL (two
forms), and the Semantic Differential. If either
you or the subject is unable to give the time re
quired at this session, administer at least one
of them at this session and the other one or
two at the 15th session.
d. The final step is your evaluation. Refer to the
Experimenter Goal Record Form (green) which lists
your selected goal from the List of Goal State
ments for Experimenters and check your evaluation.
e. Complete all necessary information on the Subject
Record Form and place all forms pertaining to
this subject in the folder and return to Marion
Luenberger.
For Control Subjects:
1. Fill out the Subject Information Form and the first
lines of the Subject Record Form (yellow).
165
2. Administer the POI, the Today and the General forms of
the MAACL, and the Semantic Differential (yellow) to
the subject. See Step 6 for Experimental subjects for
suggestions of ways to schedule these.
3. On or before the third session, select one goal from
the List of Goal Statements for Experimenters (blue)
which you consider most nearly expresses your con
ceived goal for this subject. Record it and the iden
tifying data on the Experimenter Goal Record Form
(green). Also record on this form, specific behavior
that you would consider requisite to his attaining
your goal. File the two forms in the subject's folder.
4. Therapeutic treatment is to proceed through thirteen
sessions with no planned and as little inadvertent
reference to goals as possible. (Behaviorists who
refer to goals as a part of their regular therapy will
necessarily proceed with their customary plans.)
5. The experimental period will end with the fourteenth
(14th) session at which time it is necessary that you
carry out the following:
a. Speak to the subject about the fact that you
have been working together for some time during
which you both have presumably had an outcome
goal in mind. Ask him to read the List of Goal
Statements (for Subjects--yellow) and check the
one which most nearly describes a goal he has
been or is still seeking to attain. If he
should insist that more than one of these is
pertinent to his goal, have him rank them as
"1" for most nearly, "2" for next most nearly,etc.
b. Re-administer the three tests : POI, MAACL (two
forms), and the Semantic Differential. If either
of you is unable to give the time required at
this session, administer at least one of them
during this session aid the other one or two at
the 15th session.
c. The final step is your evaluation and judgment.
Refer to the Experimenter Goal Record Form (green)
which lists your selected goal from the List of
Goal Statements for Experimenters and check your
evaluation. Then, under the section for Control
Subjects only, describe behavior which is illus-
166
trative of goal attainment or lack of goal
attainment. Designate your judgment as to
behavioral movement toward the goal.
6. Complete all necessary information on the Subject
Record Form and place all forms pertaining to this
subject in his folder and return to Marion Luenberger.
167
SUBJECT RECORD FORM
NAME_________________________Experimental Subject No.____
Control Subject No._____
Replacement for a dropped
subject? Yes No_____
Name of Experimenter_____________________________________
Date of First Experimental Session_______________________
Dates of Experimental Sessions:__(2)______ (3)______ (4)__
(5)______(6)______ (7)______ (8)______ (9)______(10)______
(11)______(12)______ (13)______ (14)______
Pretests Administered:
POI , MAACL , Semantic Differential
(date) (date) (date)
Posttests Administered:
P O I , MAACL , Semantic Differential
(date) (date) (date)
List of Goal Statements for Subjects Checked on Last
Experimental Session Yes No . No. of the
Statement Checked______ .
Goal Setting Record
(For Experimental Subjects Only)
No. 1_____Date______
Goal(s):
Requisite Behavior which would be considered attainment
of goal:
Present Behavior:
168
Goal Setting Record (Continued)
1st Revision on Session No. (Record revised goal[s]
and any revised requisite behavior for revised goal
attainment)
2nd Revision on Session No.____
3rd Revision on Session No.____
4th Revision on Session No.____
Additional Revisions and Session No. (Use additional
page if necessary)
Fourteenth Session: Goal(s) attained? Yes No Partly__
Statement of goal as considered attained________________
Is attainment still to be sought? Yes No_____
Behavior which illustrates goal attained or not attained_
)
Behavior shows Definite Movement , Some Movement
No Movement towards the goal(s).
169
EXPERIMENTER GOAL RECORD FORM
Name of Subject_______________________________ E C_____
Name of Experimenter______________________________________
Date of First experimental session________
Goal: Select one of the goal statements for Experimenters
(blue) as your primary goal for this subject.
(Selection to be made not later than the third
session.)
Date_______________________________
No. of the Statement__________
Brief copy of the Statement________________________
Evaluation of Goal Attainment:
Above goal attained? Yes______No______Partly______
The Following Additional Information is For Use with
Control Subjects Only:
Describe specific behavior which you would consider re
quisite to goal attainment. (To be completed at the time
of goal selection).________________________________________
Describe behavior at the end of the experimental period
which illustrates goal attained or not attained________
Behavior shows Definite Movement_
No Movement______ toward the goal.
, Some Movement
170
LIST OF GOAL STATEMENTS FOR EXPERIMENTER USE
B 1. To eliminate avoidance behavior and learn new
behaviors which will be satisfying.
P 2. To bring about conscious control of uncon
scious impulses.
E 3. To become confident in "own identity," to
risk it, and to expand it in new significant
relationships with others and with the world.
P 4. To gain insight into the defense mechanisms
being utilized in order to reduce to manage
able limits the anxiety associated with a re
pressed conflict.
B 5. To develop efficient behavior in school, work,
or love.
E 6. To increase self-awareness, self-confidence,
self-respect and/or self-worth.
E 7. To become a fully functioning person who is
congruent, spontaneous and open to experience.
P 8. To re-experience earlier traumatic experiences
and thereby release and control attendant
emotions.
B 9. To alter or extinguish troublesome behaviors
(symptoms) that are related to anxiety and
are disrupting the individual's life.
P 10. To modify and restructure personality through
recapturing and expressing strong emotional
feelings.
E 11. To become an authentic, integrated person who
is actualizing his being with aware and re
sponsible choices.
B 12. To increase the frequency and duration of
behavior which is likely to be positively
reinforced.
Subject__________________________________
Experimenter
LIST OF GOAL STATEMENTS
1. To find new ways of acting which are far more
satisfying than my avoiding or withdrawing
from certain situations or people.
2. To be able to recognize and direct the im
pulses and desires of which I was formerly
unaware.
3. To be confident enough of me as me to take
chances in new relationships and situations,
knowing they may bring either fulfillment or
hurt, but always a feeling of my being alive.
4. To understand that I have been using some
"tricks" or "defenses" to avoid facing my
real problem and its anxieties.
5. To perform more efficiently in school (or
. work or play or love).
6. To gain in awareness of myself, in confi
dence and self-respect.
7. To use all of my potential and live in such a
way that my actions match my ideals, my im
pulses are to be trusted, and I am unafraid
of new experiences.
8. To discover the emotional impact that earlier
experiences had upon me and to understand that
those emotions can be redirected and need not
cause me anxiety in the present.
9. To get rid of a very bothersome habit I had
or to overcome the great fear I had of a
particular situation.
10. To undergo new emotional experiences which
will change my personality.
11. To become an honest, open person, knowing
what I want, and taking the responsibility
for achieving it.
12. To learn skills of responding more often to
others and/or situations in ways that are
rewarding to me.
B: Behaviorism, P: Psychoanalysis
E: Existentialism
APPENDIX B
i
%
172
173
ILLUSTRATIVE STATEMENTS FROM THE
PERSONAL ORIENTATION INVENTORY
1. a. When a friend does me a favor, I feel that I
must return it.
b. When a friend does me a favor, I do not feel that
I must return it.
2. a. I often make my decisions spontaneously,
b. I seldom make my decisions spontaneously.
3. a. I can put off until tomorrow what I ought to
do today.
b. I don't put off until tomorrow what I ought to
do today.
4. a. In order to grow emotionally, it is necessary to
know why I act as I do.
b. In order to grow emotionally, it is not necessary
to know why I act as I do.
5. a. I fear failure.
b. I don't fear failure.
6. a. I believe that man is essentially good and can
be trusted.
b. I believe that man is essentially evil and
cannot be trusted.
7. a. It is important that others accept my point of
view.
b. It is not necessary for others to accept my
point of view.
8. a. People should always control their anger,
b. People should express honestly-felt anger.
9. a. I like to withdraw temporarily from others.
b. I do not like to withdraw temporarily from others.
10. a. I can overcome any obstacles as long as I
believe in myself.
b. I cannot overcome every obstacle even if I
believe in myself.
APPENDIX C
MULTIPLE AFFECT ADJECTIVE CHECK LIST
174
175
active
afraid
alone
annoyed
bored
calm
contented
contrary
cruel
caring
ILLUSTRATIVE ADJECTIVES FROM THE
MULTIPLE AFFECT ADJECTIVE CHECK LIST
discouraged offended
energetic rejected
enthusiastic sad
fearful safe
fit stubborn
gloomy strong
good-natured tender
healthy tense
hostile upset
inspired young
APPENDIX D
INSTRUCTIONS FOR THE SEMANTIC DIFFERENTIAL
KEY TO INTERPRETATION OF SEMANTIC DIFFERENTIAL SCORES
176
177
THE SEMANTIC DIFFERENTIAL
Subject Name_________________________________ Pre_____Post___
Experimenter Name____________________________Date___________
INSTRUCTIONS
The purpose of this test is to measure the meanings of cer
tain things to various people by having them judge them
against a series of descriptive scales. In taking this
test, please make your judgments on the basis of what these
things mean to you. On each page of this booklet you will
find a different concept (thing) to be judged, and beneath
it a set of scales. You are to rate the concept on each of
these scales in order.
Here is how you are to use these scales :
If you feel that the concept at the tope of the page is
very closely related to one end of the scale, you should
place your check-mark as follows:
GOOD X : : : : : BAD
or
GOOD : : : : : X BAD
If you feel that the concept is quite closely related to
one or the other end of the scale (but not extremely), you
should place your check-mark as follows:
UGLY : X : : : : BEAUTIFUL
or
UGLY :
Y
• • • ♦ BEAUTIFUL
If the concept seems
opposed to the other
only slightly related
side, then you should
to one side as
check as follows
TRUE :
Y
• A • • » FALSE
or
TRUE :
Y
• • A • • FALSE
The direction toward which you check, of course, depends
upon which of the two ends of the scale seem most charac
teristic of the thing you are judging.
IMPORTANT: (1) Place your check-marks in the middle of
the spaces, not on the boundaries.
: : X : X :_____
This Not this
178
(2) Be sure you check every scale for every concept--
do not omit any.
(3) Never put more than one check-mark on a single
scale.
(4) Work quickly; it is your first impressions
that we want.
179
KEY TO INTERPRETATION OF SEMANTIC DIFFERENTIAL SCORES
14 - 20 Positive attitudes are very closely related to
the concept.
21 - 34 Positive attitudes are quite closely related to
the concept.
35 - 41 Positive attitudes are fairly closely related to
the concept.
42 - 55 Area of vacillation between slight positive
(42-48) and slight negative (49-55) relationship
between attitudes and concepts.
56 - 62 Negative attitudes are fairly closely related
to the concept.
63 - 76 Negative attitudes are quite closely related
to the concept.
77 - 84 Negative attitudes are very closely related to
the concept.
APPENDIX E
GOAL STATEMENT JUDGING
180
181
GOAL STATEMENT JUDGING
Instructions for the Judges:
PLEASE do not look at Part II until you have completed the
two steps of Part I.
PART I - List of Goal Statements for Therapists' Use with
Individual Patients.
Step 1 - The eighteen (18) statements listed on Pages 1
and 2 are goals which have been culled from the
literature and, in some cases, partially para
phrased. These represent overall goals or objec
tives of therapeutic treatment as expressed by
theorists of the following three fields of psycho
therapy :
A. Behaviorism
B. Psychoanalysis
C. Phenomenology (Existential, Gestalt,
Rogerian, Humanistic, etc.).
The object of this judging is to ascertain if
each of these statements can be identified as
a characteristic goal of one of the three fields.
Please judge each goal statement as to its
theoretical origin and mark on the line indi
cated ("Category"):
A for Behaviorism, B for Psychoanalysis
C for Phenomenology
Do not place more than eight or less than four in
any one category.
Step 2 - After you have categorized the goals, please rank
those in each category according to the importance
you believe proponents of each theory would give
to their specific goals as chosen by you. Rank
1 for most important, 2 for the next most impor
tant, etc., down to 4, 5, 6, 7, or 8 (as per num
ber you placed in a category) for least important.
PART II- List of Goal Statements for Patients' Use.
These eighteen goal statements have been reworded
in terms considered to be more applicable to the
understanding of patients. The object of this
182
judging step is to ascertain whether they do in
fact parallel the goal statements for the
therapists.
Match by placing the number of a statement in
PART I next to the statement in Part II which
seems its duplicate.
Please return the completed forms as soon as possible.
Many thanks for your help and cooperation.
(Signed) Marion Luenberger
183
PART I
LIST OF GOAL STATEMENTS FOR THERAPISTS' USE WITH
INDIVIDUAL PATIENTS
Rank Category
1. To eliminate avoidance behavior and
learn new behaviors which will be
satisfying.
2. To experience steps in the decision
making process so the individual will
ultimately be satisfied with his own
decisions.
3. To bring about conscious control of
unconscious impulses.
________ 4. To become confident in "own identity,"
to risk it, and to expand it in new
significant relationships with others
and with the world.
5. To strengthen ego integration and
functioning.
6. To learn to utilize skills of communi
cation and thus relate toothers with
less friction.
7. To gain insight into the defense mech
anisms being utilized in order to re
duce to manageable limits the anxiety
associated with a repressed conflict.
8. To increase self-awareness, self-con
fidence, self-respect, and/or self-
worth .
9. To become creative by detaching himself
from others 1 expectations.
10. To re-experience earlier traumatic ex
periences and thereby release and con
trol the attendant emotions.
11. To reason better, to see new relation
ships between antecedents and present
behavior, and to use these new under
standings in changing aspects of his
life.
12. To become a fully functioning person
who is congruent, spontaneous, and
open to experience.
184
Rank Category
13. To develop efficient behavior in
school, work, or love.
14. To become an authentic, integrated per
son who is actualizing his being with
aware and responsible choices.
15. To gain insight into relationships
with authority figures through the
therapeutic relationship.
16. To release an already existing capacity
in a potentially competent individual.
17. To modify and restructure personality
through recapturing and expressing
strong emotional feelings.
18. To alter or extinguish troublesome
behaviors (symptoms) that are related
to anxiety and are disrupting the in
dividual’s life.
Comments ?
185
PART II
LIST OF GOAL STATEMENTS FOR PATIENTS' USE
1. To use all of my potential and live in such a
way that my actions match my ideals, my im
pulses are to be trusted, and I am unafraid of
new experiences.
2. To learn better methods for making decisions.
3. To discover the emotional impact that earlier
experiences had upon me and to understand those
emotions can be redirected and need not cause
me anxiety in the present.
4. To get rid of a very bothersome habit I had or
to overcome the great fear I had of a particu
lar situation.
5. To gain in awareness of myself, in confidence,
and self-respect.
6. To become anbonest, open person, knowing what
I want, and taking the responsibility for
achieving it.
7. To undergo new emotional experiences which will
change my personality.
8. To understand the lasting emotional involvement
I have had with my father (mother) and how this
has influenced my relationships with others.
9. To learn to reason better about the particular
situations, thoughts, words, or actions of my
self or others which have caused me to act in
unsatisfactory ways and to use this reasoning
in changing these and other parts of my Ife.
10. To perform more efficiently in school (or work
or play or love).
11. To feel free enough to begin to use the abil
ities I have but did not use.
12. To be able to recognize and direct the impulses
and desires of which I was formerly unaware.
186
13. To learn better skills of communication in my
relationships with others.
14. To be able to perceive the reality of situations
in my life and cope with them logically.
15. To overcome my need to satisfy the expectations
of others and instead fulfill-my own expecta
tions for me.
16. To find new ways of acting which are far more
satisfying than my avoiding or withdrawing from
certain situations or people.
17. To understand that I have been using some
"tricks" or "defenses" to avoid facing my real
problem and its anxieties.
18. To be confident enough of me as me to take
chances in new relationships and situations,
knowing they may bring either fulfillment or
hurt, but always a feeling of my being alive.
Comments?
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
A Comparative Study Of The Marathon-Group Experience In A University Counseling Center
PDF
Stimulus and response generalization of classes of imitative and nonimitative behavior as a function of reinforcement, task, cues, and number of therapists
PDF
Differential Effects Of Leader Versus Leaderless Groups
PDF
A Comparative Study Of Marathon Placement In Short-Term Group-Counseling
PDF
Mental Imagery As A Function Of Muscular Tension And Suggestion
PDF
The Functional Aspects Of Adlerian Constructs In Understanding And Assisting Disadvantaged Children
PDF
The Differential Effects Of Marathon And Weekly Groups' Self-Perceptions And Interpersonal Feelings
PDF
Looking Back After Coming Down: Conformity And Commitment In Campus Protest
PDF
Parental Support Of Independent Free Schools. Determinant And Value Implications For Public School Counselors: An Exploratory Study
PDF
Cross-Identification And Maternal Attitudes As Factors Affecting Latency Age Male'S School Learning And Behavior
PDF
Shaping Of Verbal Behavior In An Interview Setting
PDF
Structure-Of-Intellect Factor Abilities And A Self-Concept Measure In Mathematics Relative To Performance In High School Modern Algebra
PDF
The Comparative Effects Of Tape-Led, Led, And Leaderless Groups
PDF
A Cognitive Dissonance Analysis Of Conformity Behavior, As Applied To The full Denture Patient
PDF
Tension And Anxiety In Deconditioning
PDF
The Effects Of A Specific Counseling Technique Upon The Belief In Locus-Of-Control Within The Newly Admitted Hospital Patient
PDF
Non-Specific Treatment Factors And Deconditioning In Fear Reduction
PDF
An Investigation Of Leader And Member Helpfulness In Group Counseling
PDF
A Study Of Delinquency Among Urban Mexican-American Youth
PDF
The Effect Of Group Discussion On Teachers' Attitudes Towards Risk-Taking In Educational Situations: An Investigation Of The Risky Shift Phenomenon
Asset Metadata
Creator
Luenberger, Marion Crumley (author)
Core Title
Therapeutic Movement As A Function Of Awareness Of Goals
Degree
Doctor of Philosophy
Degree Program
Education
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
education, guidance and counseling,OAI-PMH Harvest
Format
dissertations
(aat)
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Ofman, William V. (
committee chair
), Webb, Allen P. (
committee member
), Wolpin, Milton (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c18-767591
Unique identifier
UC11364260
Identifier
7300747.pdf (filename),usctheses-c18-767591 (legacy record id)
Legacy Identifier
7300747
Dmrecord
767591
Document Type
Dissertation
Format
dissertations (aat)
Rights
Luenberger, Marion Crumley
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
education, guidance and counseling