EARLY Arm LATE mmsms OF CUENLPERCEWED
THERAMST SELF-DISCLOSURE AS THEY RELAYS
T0 CONSTRUCTWE cum muse AND TO
magma m PSYCHOFHERAPI’

Thesis for the Degree of Ph. D. V
Micmem S?A‘FE UNWERSITY
WALYER L STUMP‘

1968‘

 

 

LIBRARY

Michigan State
University

 

This is to certify that the

thesis entitled

EARLY AND LATE DIMENSIONS OF CLIENT-PERCEIVED
THERAPIST SELF-DISCLOSURE AS THEY RELATE
TO CONSTRUCTIVE CLIENT CHANGE AND TO
OUTCOME IN PSYCHOTHERAPY

presented by

WALTER L. STUMP

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Counseling, Personnel
Services and Educational
Psychology

(:Lr/ff 2? Cgv4afi;brl,
/ Majofl‘ofessor

Dated/v»( fl 9:, [75 8

0-169

ABSTRACT

EARLY AND LATE DIMENSIONS OF CLIENT—PERCEIVED
THERAPIST SELF—DISCLOSURE AS THEY RELATE TO
CONSTRUCTIVE CLIENT CHANGE AND TO
OUTCOME IN PSYCHOTHERAPY

by Walter L. Stump

The general disclosure of therapists' "meanings"
and "intentions" to the client, has been considered within
Rogerian theory as a necessary aspect of successful psycho~
therapy. Recent suggestions have been made, that client-
perceived therapist self-disclosure, to be therapeutically
effective, is probably contingent on the ambiguity or
veiling of therapists' meanings and intentions during
early therapy sessions.

The major purpose of the present thesis was to test
the prediction that there would be significantly higher
frequencies of client—perceived therapist ambiguity or
veiling indices (low self—disclosure) in early measures of
(a) subjects showing higher degrees of constructive change
than §_s with low degrees of change, (b) S_s given high
success ratings than with §s with low success ratings, and
(C) §s showing more degrees of disturbance than with §s
with less degrees of disturbance. It was also predicted
that there would be higher frequencies of client-perceived
therapist clarity or meaningful (high self—disclosure)

indices in overall (or late) measures of gs showing higher

Walter L. Stump

degrees of constructive change and/or with S5 rated as more
successful clients, than with low change §S and/or with
less successful clients.

Clients used in the present study were twenty—six
self-referred Michigan State University undergraduate
females, who had been seen at the University Counseling
Center for personal counseling.

High and Low Constructive Change Groups were devel—
Oped on the basis of differences between Ss' pre- and post—
MMPI formula scores. Therapists' ratings of individual
case success was the basis of High and Low Success Group
compositions. Pre-MMPI formula scores provided a criteria
for assigning §S into More or Less Disturbed Groups.

Client's perception of therapist's self—disclosure
was defined as S's scale-checking behavior on a series of
seven—point (position) semantic differential judgments of
the concept "Counselor." Positions 1, A, and 7 (combined
and individual), the meaningless—neutral scales, were
separate indices of low therapist disclosure, and positions
2, 3, 5, and 6 (combined and individual), the most discrim—
inatory scales, were separate indices of high therapist
disclosure.

Sixteen judgments of the concept "Counselor" were
made by each of the twenty-six SS on early and/or late
administrations of the semantic differential test. Fre—

quency scores were obtained on the S's uses of the seven

Walter L. Stump

alternate positions over 16 of the seven-point scale
judgments of the concept "Counselor."

Twenty-two Mann-Whitney U, one tailed tests (a = .05)
were then used to test predicted differences between the
various groups mentioned above, on their frequency scores
for each separate measure (individual and combined positions
scores) either for l, A, and 7 or 2, 3, 5, and 6; depending
on whether the hypothesis was investigating low or high
therapist disclosure indices.

The results of the analyses supported only three of
the twenty-two eXperimental predictions of group differ-
ences. As expected, High Constructive Change clients did
use greater frequencies (significant at .016) of combined
positions 1, A, and 7 on early judgments of "Counselor"
than did Low Change clients. On this one combined measure
of positions 1, A, and 7 (low therapist disclosure), there
was support for the hypothesis (d = .05) that High Construc-
tive Change clients would perceive their therapist's
meanings less clearly during early therapy, than would
Low Constructive Change clients. The positions 1, A, and
7, however, as individual measures of low therapist
disclosure, were not significantly different between the
above groups. The latter individual measures did not
support the hypothesis that there would be greater

indices of low therapist disclosure on earlier individual

Walter L. Stump

position measures for High Change §S, than for Low
Change SS.

High and Low Success SS, as well as More and Less
Disturbed §S, showed no predicted differences in scale-
checking behavior on their early therapy judgments of
"Counselor" on positions 1, A and 7 (low therapist dis—
closure), either on individual or on combined measures
of positions 1, A, and 7. No support was gained on the
above tests for the presence of early therapist veiling in
those cases rated as successful. Also, no support was
found in the above tests (individual and combined positions
1, A, and 7) that More Disturbed clients would perceive
therapists' meanings as more veiled in early therapy than
Less Disturbed clients.

No differences were found between High and Low
Constructive Change clients on late (individual and combined)
measures of positions 2, 3, 5, and 6 (high therapist
disclosure). The latter data did not support the hypothesis
that High Constructive Change clients would have more
indices of high therapist disclosure (positions 2, 3, 5,
and 6) on overall (or late) measures, than Low Construc-
tive Change clients.

Significant differences (a = .05) were found between

High and Low Success gs on two measures of overall (or
late) indices of high therapist disclosure. Position 2

was significantly different between High and Low Success

Walter L. Stump

groups at the .007 level, and position 6 at the .025 level.
These two measures supported the hypothesis that High
Success clients would perceive their therapists more
clearly (high therapist disclosure) in overall therapy
than would Low Success clients.

No differences between the same groups above were
found, however, on the other measures of high therapist
disclosure, combined positions (frequencies) of 2, 3, 5, and
6 and on individual positions (frequencies) 3 and 5.

The results of all twenty—two of the hypotheses
tested in the present thesis, although failing to generally
support the importance of the temporal contingency of early
therapist veiling to successful therapy, did show some
encouraging trends, in predicted directions.

Although the statistical analysis did not support the
theory of the present thesis, two results of the study are
cited as reasons for failure; (a) it can be argued that the
fourth and eighth-interview semantic differentials provided
only the middle of the continuum between early and late
client perceptions of "Counselor." Consequently a much
wider continuum between "early" and "late" client percep-
tions would be needed, and (2) the subjects can be regarded
as homogeneous in terms of disturbance levels and conse—

quently a more extreme degrees of disturbance would be

needed.

EARLY AND LATE DIMENSIONS OF CLIENT—PERCEIVED

THERAPIST SELF—DISCLOSURE AS THEY RELATE TO

CONSTRUCTIVE CLIENT CHANGE AND TO

OUTCOME IN PSYCHOTHERAPY
By

-~\
\ (
Walter L: Stump

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Counseling, Personnel Services,
and Educational Psychology

College of Education

1968

DEDICATION

To my wife, Jean, for patience, help and love;

and to my son, Mike, for interest and acceptance.

ii

ACKNOWLEDGMENTS

The author is extremely grateful to the members of
his doctoral committee whose individual and collective
efforts have made the completion of this thesis possible.
Acknowledgments are made to Drs. Ralph Kron and William A.
Mann who respectfully served on the committee and provided
helpful suggestions later incorporated into the study.

A special word of thanks goes to Drs. John F. Jordan
and Norman Abeles who actively provided direction and

encouragement as Co—Chairmen of my Doctoral and Dissertation

Committees. Each of these men has given generously of his

time and interest which were invaluable in the preparation
of this manuscript.
My deep appreciation is also extended to Drs. Jose-

phine Morse and Bill Kell who have contributed immeasurably

to my personal and professional growth. Each has influenced

my life, which I trust will always reflect the integrity

and commitment that is represented in each of these persons.

In conclusion I wish to say that without the support
and time provided by the Kalamazoo Consultation Center and

the understanding of its Executive Director, K. Charles

Nesseth, as well as the proficient typing contributions
of the Center's staff secretary, Mrs. John (Judy) Parsons,

this dissertation could not have been finalized in its

present form.
iii

TABLE OF CONTENTS

ACKNOWLEDGMENTS . . . . .

LIST OF TABLES . . . . . .

LIST OF FIGURES . . . . .

LIST OF APPENDICES . . . .

Chapter

I.

II.

III.

THE PROBLEM . . . . .

Introduction . . .
Theoretical Background .

REVIEW OF THE LITERATURE .

Multi-Dimensional Studies
Uni—Dimensional Studies .
Research Conclusions .
Experimental Hypotheses
Hypothesis I . . . .
Hypothesis II . . .
Hypothesis III . . .
Hypothesis IV . .

METHODOLOGY AND PROCEDURES

Client-Perceived Therapist

Psychotherapy Outcome .

Constructive Client Change

Operational Hypotheses .
Hypothesis I . . .
Hypothesis II .
Hypothesis III .
Hypothesis IV .
Hypothesis V

The Source of Data

Selection of Cases

iv

. O O Q
C

O O O O O

Self-Disclosure

and Disturbance

o g . . . .
o o . . . .
o g . . . .
. O
O
O O O Q . . . .

Page
iii
vi
viii
ix

Chapter
IV.

RESULTS OF STUDY

Statistical Results of

Hypothesis
Hypothesis
Hypothesis
Hypothesis

I
II

III
IV

Hypothesis V .

V. SUMMARY AND CONCLUSIONS

Summary .

Conclusions
Hypothesis
Hypothesis
Hypothesis
Hypothesis
Hypothesis

Discussion

Implications for Futu

REFERENCES . . .
APPENDICES . . .

o

I
II

III

IV
V

O

0
re

the

0

Res

earch

. Q . . . .

Page

LIST OF TABLES

Page

Descriptive summary of client, therapist and
therapy characteristics . . . . . . . . A6

Pre-MMPI raw and formula scores for nineteen
Ss used in E's thesis . . . . . . . . “7

Post—MMPI raw and formula scores for eleven
§S used in E's thesis . . . . . . . . A8

MMPI change scores and signs on eleven Ss
used in Hypotheses I and III . . . . . . A9

Therapists' ratings of case outcomes for
twenty Ss used in this thesis . . . . . . 50

Positions checked (1-7) by twenty-three §S on
judgments of the concept "counselor."
Based on an early administration of the
Semantic Differential . . . . . . . . 51

Positions checked (1-7) by twenty-one Ss on
judgments of the concept "counselor."
Based on a late administration of the
Semantic Differential . . . . . . . . 52

Frequencies of neutral or meaningless positions
(1, A, 7) checked by high and low construc-
tive change groups on early Semantic
Differential judgments of the concept
"counselor" . . . . . . . . . . . . 53

Frequencies of neutral or meaningless positions
(1, A, 7) checked by high and low success
groups on early Semantic Differential
judgments of the concept "counselor" . . . 5A

Frequencies of discriminating positions
(2, 3, 5, 6) checked by high and low con—
structive change groups on late Semantic
Differential judgments of the concept

"counselor" . . . . . . . . . . .

. 55

vi

Table Page

11. Frequencies of discriminating positions
(2, 3, 5, 6) checked by high and low
success groups on late Semantic
Differential judgments of the concept
"counselor" . . . . . . . . . . . . 56

l2. Frequencies of neutral or meaningless
positions (1, A, 7) checked by high and low
disturbance groups on early Semantic
Differential judgments of the concept
"counselor" . . . . . . . . . . . . 57

13. Significance probabilities associated with all

values as small as observed values of U
in Hypotheses I-V - . . . . . . . . . 58

vii

 

LIST OF FIGURES

Figure Page

1. Percentages of neutral and meaningless
positions (1, A, and 7) checked by High—
Low Constructive Change groups, on early
judgments of the concept "Counselor" . . . 36

2. Percentages of neutral and meaningless
positions (1, A, and 7) checked by High-
Low Success groups, on early judgments
of the concept "Counselor" . . . . . . . 38

3. Percentages of discriminatory positions
(2, 3, 5, and 6) checked by High-Low
Constructive Change groups on late
judgments of the concept "Counselor" . . . A0

A. Percentages of discriminatory positions
(2, 3, 5, and 6) checked by High-Low
Success groups on all late judgments
of the concept "counselor" . . . . . . . A3

5. Percentages of neutral or meaningless positions
(1, A, and 7) checked by More-Less
Disturbance groups on all early judgments
of the concept "Counselor" . . . . . . . A5

6. Proportionate uses of alternate positions by
undifferentiated §S used in this thesis . . 73

7. PrOportionate differences between meaningless
(l, A, 7) and discriminatory (2, 3, 5, 6)
position group responses on undifferen—
tiated §S used in this thesis . . . . . . 75

viii

LIST OF APPENDICES

Appendix Page

A. The Description and Rationale for the Use
of a Semantic Differential in This Thesis . 82

B. The Semantic Differential, Form A . . . . . 91
C. The Description and Rationale for the Use

of the Minnesota Multiphasic Person-
ality Inventory in This Thesis . . . . . 96

D. Cooke's Actuarial Formula . . . . . . . 100

ix

CHAPTER I
INTRODUCTION

A growing number of studies are investigating the
complex nature of therapist—client relationships. Two basic
questions raised are: what can therapists do to increase
successful therapy outcomes and what are the critical char-
acteristics a therapist must possess or promote in himself
to effect constructive change in his client?

Rogers (1957) has listed three therapist conditions
he believed were necessary and sufficient to facilitate
change in clients and increase prospects for therapeutic
success: congruency, positive regard, and empathy. A
variety of studies (Barrett-Lennard, 1962; Demos, 196A;
Halkides, 1958; Rogers, 1965; Rogers, Gendlin, Kiesler and
Truax, 1956; Truax, Wargo, Frank, Imber, Battle, Saric,
Nash and Stone, 1966; VanderVeen, 1965) have attempted to
determine the simultaneous presence of all three of Rogers'
conditions and their effects on (a) therapeutic change,

(b) case success, and (c) expansion or explorations of
client feelings.

Much research has cited congruence or "genuineness"
as a single element of critical importance in its effect

on the dependent variables of case success and constructive

client change (Rogers, 1957; Thorne, 1950; Truax and
Carkhuff, 1965; and Whitaker and Malone, 1957). Within the
above theoretical framework the secondary but related
concepts of "transparency" (Truax and Carkhuff, 1965) and
"psychological Openness" have been Operationally defined
and studied. The concepts of congruence, genuineness,
transparency, and psychological openness all seem to relate
to the phenomenon in which the therapist's "meanings" and
"intentions" (Kell and Mueller, 1966) are revealed to the
client; in essence, a phenomenon in which therapist self-
disclosure occurs.

The traditional assumption has been that increased
amounts of therapist self-disclosure along with other
beneficial characteristics measured in a relationship would
effect improvement in therapist-client relationships and
ultimate therapy outcome. Mills and Zytowski (1967) have
called this assumption into question. They have suggested
that there may be therapist response contingencies for the
various "beneficial" characteristics of a helping relation—
ship. That is, there may be times and circumstances where
the absence of therapist self-disclosure might strengthen
the ultimate therapy relationship and results.

Several contingencies have already been cited by
past investigators. VanderVeen's study (1965) suggested
that, in part, the level of empathy and genuineness were

determined by the client. Truax, Wargo, Frank, Imber,

Battle, Saric, Nash and Stone (1966) made a re-analysis of
VanderVeen's hypothesis, using a different population of
therapists and patients, as well as a different design, to
confirm that it is the therapist who sets the level of
empathy and genuineness.

Studies by Carkhuff (1966) and Demos (196A) provide
significant evidence that therapist characteristics, i.e.,
self—disclosure, must be perceived by the clients in order
to effect therapeutic change in clients.

Rogers (1966) found evidence that perception of these
"important" therapist attitudes, i.e., self-disclosure, is
contingent on client levels of disturbance.

Kell and Mueller (1966) have pointed to another
possible contingency. After listening to many hours of
tape recorded interviews from different client—therapist
encounters, they believed that the more veiled or ambiguous
the counselor is with his intentions and meanings during
early interviews, the more opportunity there is for expan-
sion and exploration of client feelings. Kell and Mueller
assume that such early eXpansion and eXploration enhances
the therapeutic process.

To summarize the preceding remarks, researchers have
focused on the disclosure of therapist's meanings and
intentions as a necessary part of (a) successful thera-
peutic relationships, (b) constructive client change, and

(C) ultimate case outcome. Further, it has been concluded

that such therapist self-disclosure must be perceived by

the client to produce desired therapeutic effects. In
addition, it is believed that the above therapist disclosure,
to be effective, is probably contingent on other factors,
e.g., a temperal dimension, i.e., early ambiguity or veiling
of therapist’s meanings and intentions facilitates construc-
tive client change and/or successful case outcomes. The
latter assumption, on the temporal dimension, leads to the
conclusion that investigations should be conducted to
explore the effect of this early-late client-perceived
therapist self-disclosure contingency.

The major purpose of the present thesis was to assess
whether there were significantly higher frequencies of
client—perceived therapist ambiguity or veiling indices in
early measures of subjects showing higher degrees of con-
structive change and/or with subjects rated as successful
clients, than with control subjects showing little or no
constructive change and/or with subjects who were rated

as less successful clients.

Theoretical Background

 

Self-exposure has long been recognized as an impor-
tant component of psychotherapy. Early attention has been
given to those concepts describing what was hidden within
the client. Dolliver (1965) suggests that various labels

were applied; i.e., Freud's "The Unconscious," Jung's "The

Collective Unconscious" and his "Undiscovered Self,"
Rogers' "The Self Which One Truly Is," and the existen—
tialist's "Authentic Being."

Early concerns in psychotherapy dealt with the
"uncovering" process and the value of client self-disclosure.
Recent attention has shifted toward the value of therapist
self~disclosure and its relationship to the therapeutic
process. Such labels as Rogers' (1961) "congruence" or
"genuineness" Truax and Carkhuff's (1965) "transparency,"
Allen's (1967) "psychological openness," and Kell and
Mueller's (1966) therapist "unveiling" or "selective reve-
1ations," are all ways in which therapist self-disclosure
‘has been conceived.

Rogers (1951) was one of the first theorists to
develop a basis for therapist self-disclosure. His concept
of congruence or genuineness, along with acceptance or
positive regard and empathy, were cited as "necessary and
sufficient conditions" for facilitating client change.
However, this early Rogerian theory kept the client at the
center of the therapist's focus. The therapist was viewed
as attempting to get into the client's world, without
himself, as therapist, becoming obstructive or obvious in
this relationship. Dolliver (1965) indicates, however,
that recent trends in client-centered therapy encourage
the therapist to experience and express more fully his own

feelings in the relationship. Rogers (1961) states that,

 

when the therapist, in counseling, is "genuine and without
facade," openly being at that moment the feelings that are
flowing through him, he will facilitate change in the
client. This latter position was recently reaffirmed by
Rogers (1966).

The existentialist therapists (Brecht, 196A) have
held similar views to those of Rogers above. They con-
ceive psychotherapy as "a dynamic relationship between two
people in which the therapist is able to offer part of
himself for the benefit of the other person."

Wyatt (1962) quotes Fenichel, from the psychoanalytic
position, as saying that psychotherapy is "not a case of
behavior, something that has to be put on, but a delicate
interaction which has to be appropriate to its avowed pur-
pose but otherwise leaves the therapist as natural and
spontaneous as he can be."

Kell and Mueller (1966), speaking from a dynamically
oriented point of view, suggest that "Counselors help their
clients by being human." They eXplain that, "the coun-
selor has primarily himself to bring to the helping rela-
tionship." They have, as previously stated, introduced
the possibility of the temporal dimension of therapist
self-disclosure or, specifically, the importance of early
veiling of therapists' intentions and meanings.

There is a growing body of research that tends to

support the general hypothesis that therapist self—disclosure

is an integral variable in constructive therapeutic change.
The following chapter will review these latter studies in
depth. A special attempt will be made to point to the
evidence in previous literature that has a direct relation-

ship to the temporal dimension of therapist self-disclosure.

CHAPTER II
REVIEW OF LITERATURE

A number of published studies relate to the present
study. One group of investigations deals with the multi-
dimensional characteristics of therapists' variables. All
of these multi-dimensional studies focus on Rogers‘ (1957)
concepts of genuineness, regard, empathy, and/or exten-
sions of these. A brief summary of the multi-dimensional
literature and its relationship to this present thesis is
contained in this chapter.

A smaller number of studies review the single
dimension of self—disclosure, which is rooted in the
Rogerian concept of congruence or genuineness. Following
a summary of the multi-dimensional studies, an in~depth
review will be made of two single~dimensional studies. The
single—dimensional investigations have more direct impli-

cations to the present thesis.

Multi-DimensionalmStudies

 

All of the following investigations tend to support
‘Rogers' hypothesis that constructive personality change in
the client, during therapy, is dependent upon three essen—

tial attitudes in the therapist: (a) congruence or

genuineness in the relationship, (b) acceptance or prizing
of the client, and (c) an accurate understanding of the
client's phenomenal world.

Halkides (1958) demonstrated that a high degree of
therapist empathy, regard, and genuineness was signifi—
cantly related to successful case outcomes. In her study
therapists were rated on a seven-point scale by judges on
the three Rogerian attitudes above. The latter ratings
were based on a random selection of nine early client-
counselor interaction units and nine late client-counselor
units from twenty tape recorded cases. The most obvious
weakness of Halkides' study was in her instrumentation
and the use of ratings and judges.

Barrett—Lennard (1962), rather than using "objective"
observers, developed a Relationship Inventory (client and
therapist forms). The Relationship Inventory was designed
to measure the manner in which five dimensions (including
Rogers' three) of "necessary" attitudes were perceived by
both client and therapist; each in the other. Each
dimension was evaluated on a six—point scale. The results
indicated that when clients perceived more of all dimen-
sions (including Rogers"three), there was a significant
positive change in the client.

One important result of the Barrett—Lennard (1962)
study could have a possible bearing on the question raised

by the present research. The fifth dimension on the

10

Relationship Inventory attempts to measure a "Willingness
to be Known." The results revealed that a client—perceived
therapist's "Willingness to be Known," at the time of the
early interview, was not related to later success, as were
the other four dimensions. No data were available from
late interviews on the relationship between a client-
perceived therapist "Willingness to be Known" and case
success.

Concerning the use of the Barrett-Lennard Relationship
Inventory, Barrett-Lennard recognized that despite attempts
at item purification, the intercorrelation of the Inven-
tory's subtests were reported as positive and moderately
high (Barrett—Lennard, 1962). This suggests there may be
a general factor which accounts for much of the variance
for the entire instrument. Mills and Zytowski (1967)
state that a factor analysis of this Inventory pointed to
a general component which accounted for two-thirds of the
total variance.

Demos (196A) found that congruence, although judged
to be present to a greater degree in successful counselors
than unsuccessful ones, was not significantly present at
the .01 level of confidence as was empathy and regard. He
used different methods and population than the first two
studies above. Secondary school counselors' last recorded
session with "normal" high school students were rated

by trained judges on the Rogerian variables. In Demos'

11

study the early measurement of counselor characteristics
is missing.

Rogers (1966), using a different population of unmoti—
vated schizophrenics, discovered that perception of
therapist attitudes is contingent on levels of disturbance.
He concluded that a more disturbed person can less easily
perceive and trust the "positive" attitudes of the
therapist. Two significant results of Rogers' study related
to the present thesis. First, the more the patient per-
ceived of these "positive" attitudes, especially genuine—
ness, the greater the measure of therapeutic movement.

Also, Rogers found that the higher the degree of thera—
pist congruence, the higher the degree of patient inter—
action with a third person outside of therapy.

This concludes the review of important investigations
on multi—dimensional therapist variables. All of the above
studies have implications for the present research. The
following is a review of the uni-dimensional studies

focusing more specifically on therapist self-disclosure.

Uni-Dimensional Studies
Two studies (Allen, 1967; Truax and Carkhuff, 1965)
have direct implications for the present thesis. Both of
the latter studies focus entirely on one therapist dimen—
sion, i.e., "transparency" and/or "psychological Openness;"

in essence, therapist self-disclosure. Truax and

 

l2

Carkhuff (1965) e.g., assume that of all the Rogerian
facilitative elements (empathy, regard, genuinenesS),
genuineness or self-congruence is perhaps the most basic;
a cornerstone for empathy and regard.

Truax and Carkhuff, above, conducted a series of
investigations with therapy cases involving hospitalized
mental patients and institutionalized juvenile delin—
quents. They examined two major hypotheses: (a) The
greater the degree of transparancy, self-disclosure, or
self-exploration within the patient, during the thera—
peutic encounter, the greater will be the evidence of
constructive personality change in the client's total
sphere of living; and (b) an increase in the degree of
therapist transparency or self-congruence will be accom-
panied by an increase in the degrees of transparency,
self-disclosure or self-exploration within the patient.

In Truax and Carkhuff's investigation client eXplor-
ation or transparency was measured by several instruments:
(a) the Depth of Intra-Personal Exploration Scale (DX)
(Truax, 1962); (b) evaluations of "naive" lay raters on
tape samples; (c) diagnosticians' "blind" evaluations of
change based on pre-and post-test batteries; and (d)
additional objective test scores (i.e., MMPI) and patient

Q—sort measures of self—adjustment.

l3

Therapist self-disclosure in the above study was
measured by a five-point scale (Truax, 1961) from which
judges rated how the therapist "appears" to be in the
taped therapy samples. It should be noted that the latter
was designed to measure moment—to-moment therapist trans-
parency or self-disclosure throughout therapy; no
emphasis was placed on early or late differences. Their
study also failed to measure transparency as perceived
by the client.

Truax and Carkhuff, in their research (1965), found
a significant relationship between therapist transparency
and patient's level of self-disclosure. The second major
hypothesis was confirmed for hospitalized neuropsychiatric
populations; the greater the degree of therapist trans—
parency during therapy, the greater the constructive change
in clients. However,vdth.delinquents, the above authors
found that the less the transparency or self-exploration,
the greater the positive personality change. They state
that their findings may suggest that the effectiveness
of therapist self-disclosure may be contingent on whether
the illness is mental or social, or whether the disturbance
is internal or external. Another alternate rationale for
Truax and Carkhuff's findings will be discussed in the
conclusion to this review of literature, and an alternate

hypothesis will be investigated by the present thesis.

1A

Another closely related uni-dimensional study was
conducted by Allen (1967). His assumption was that
"psychological Openness," as a personality factor, is a
higher-order concept useful in predicting counselor
effectiveness. He perceived psychological Openness in
more static or continual terms than in temporal terms.
Allen prOposed that the effectiveness of student coun—
selors, from an introductory practicum in guidance, would
be related to the degree Of counselor psychological
Openness. The Rorschach Index of Repressive Style and a
Group Supervision Report Scale, developed by Allen, were
used as measures of psychological openness. As he pre-
dicted, Allen found a direct relationship existing between
the freedom with which §S reSponded to the Rorschach and
the degree of overall competence attributed to these SS
(counselors) by their supervisors. Allen, however,
raised a critical question; he wondered if his present
measures of counselor's psychological Openness were as
closely related to the movement of their clients as they
were to supervisor ratings of counselor effectiveness. This
latter question will be investigated in the present study.
An attempt will be made in the following section to tie
together all questions raised in the review of literature
in this chapter and to show how these questions will be

investigated in the present thesis.

15

Research Conclusions

 

In general, all Of the literature reviewed points to
evidence that therapist self—disclosure is an important
variable as it related to successful therapy outcomes
(Halkides, 1958), to positive client change (Barrett-
Lennard, 1962; Rogers, 1966; Truax and Carkhuff, 1965) and
to personality characteristics of successful therapists
(Allen, 1956; Demos, 196A).

The above evidence supports a general hypothesis
that successful therapy outcome is dependent on therapist
self-disclosure. But the assumption does not follow that
increased amounts of self-disclosure have a one—tO—One
relationship to client movement, counselor effectiveness,
and/or successful outcomes. But there is evidence to
support the notion that therapist self—disclosure is con—
tingent on both personal and temporal variables. Such
contingencies as client perception (Carkhuff, 1966; Demos,
196A) and the degree of client disturbance (Rogers, 1966)
are examples of contingent personal variables.

There are strong theoretical hunches (Kell and
Mueller, 1966) supported by related research evidence
(Barrett-Lennard, 1962; Demos, 196A; Truax and Carkhuff,
1965) indicating a possible contingent, temporal dimension
to effective therapist self-disclosure. Of particular
significance to the present thesis was Barrett-Lennard's

(1962) findings which show therapists' "Willingness to

16

be Known" in the early interview was not related to therapy
success. In the present thesis, the B will investigate
whether Barrett-Lennard's conclusions hold true with a
different population and with different instrumentation.

Demos (196A) discovered a greater degree of con—
gruence or self-disclosure present in cases of successful
counselors. But these latter degrees were not signif-
icant. The main weakness Of Demos' study was in the single
use of late interview measures. The present thesis
attempted to correct the latter weakness by designing
early and late measures of therapist self-disclosure in
order to assess the importance Of these early—late
temporal dimensions.

In conclusion, it should be noted that theoretical
support foriflkaimportance Of the temporal dimensions above
are indicated by Kell and Mueller's (1966) observations.
The latter authors have suggested that early counselor
"veiling" or "ambiguity" allows the client to expand his
own feelings and, of critical importance, to disclose
himself first. Kell and Mueller indicate that early
client self—disclosure is of Special importance when there
are possible client—counselor "control" struggles. "When
the counselor's controls are a function Of his inability
to tolerate the eXploration which the client can tolerate
and which is necessary to help the client to change, then

119 counselor's motivation for attempting control are

l7

suspect" (Kell and Mueller, 1966). They continue by
stating, "A client may enter a counseling relationship
because he experiences some anxiety about the effective-

ness of his controls."

The latter observation by Kell and Mueller may offer
an alternative rationale for differences found by Truax
and Carkhuff (1965) between mental patients and juvenile
delinquents. In their findings with a delinquent popu-
lation, successful outcomes were not related to therapist
self-disclosure. In fact, the Opposite was true. It
would seem, then, that the early control struggles suggested
by Kell and Mueller would especially be true with delin-
quents. This early struggle would seem to be more true
as well with "normal" than with a mentally-ill population.

Since the populations samples for the above studies
impose certain inferential limits, the present thesis
attempted to expand on the generalization possibilities.
The present use of college subjects and college counselors
will provide an additional sample expansion beyond those
Of juvenile delinquents, mentally ill, and high school
students. New instrumentation has been used in this thesis
to provide different operational definitions as well as
allowances for early and late measures of client—perceived
therapist self-disclosure. The latter, as previously
mentioned, was a weakness found in both studies by Allen

(1967) and Truax and Carkhuff (1965).

 

18

One conclusion should be made here, that Rogers'
"congruence" is a higher order concept and therefore
difficult to define. Also, secondary concepts, i.e.,
"transparency" (Truax and Carkhuff, 1965) and "psycho—
logical openness" (Allen, 1967), although related to
congruence, are also vague and illusive. It is admitted
in this thesis that the issue of therapist self-disclosure
is far from clear and that previous Operational definitions
leave much to be desired. However, the above literature
provides ample evidence that all previous investigations
discussed in this chapter, which deal with therapist self-
disclosure, support the significant effect Of this variable
on client change and successful case outcomes. Therefore,
the continued effort in this thesis to define therapist
self-disclosure and its relationship to dependent therapy
variables is clearly needed.

The review of literature does indicate several ques-
tions that the present thesis attempted to answer, at least
in part. Primarily, is there a significant difference
between early and late client—perceived therapist self-
disclosure as it relates to constructive client change and/
or to different therapy outcomes? There have been secondary
questions raised and investigated by the reviewed studies
which will be examined as well in the present thesis. First,
when using a different population and design, is there a

difference between high and low degrees of overall

l9

client-perceived therapist self-disclousre as it relates to
constructive client change and/or to different therapy
outsomes? Seond, is there a difference between high and
low degrees of client-perceived therapist self—disclosure
as it relates to different levels of client disturbance?"
These questions are developed into experimental hypotheses

in the following section.

Experimental Hypotheses

 

The theoretical vieWpoints and research evidence
previously cited, albeit at times indirectly, supports E's
position in the present thesis with regard to the importance
of the absence of high degrees Of client—perceived therapist
self-disclosure early in the therapeutic process. The

position stated in the testable form is as follows:

Hypothesis I

 

There is a significant difference between high and
low constructive change groups with respect to levels Of
client-perceived therapist self-disclosure. §S with high
degrees of positive change will, early in therapy, per-
ceive their therapist as more ambiguous or veiled than SS

with low degrees of positive change.

Hypothesis II
There is a significant difference between successful
and unsuccessful client groups with respect to levels of

client-perceived therapist self-disclosure. Successfull SS

20

will, early in therapy, perceive their therapist as more

ambiguous or veiled than unsuccessful SS.

Hypothesis III

 

There is a Significant difference between high and low
constructive change groups with respect to levels of client-
perceived therapist self-disclosure in the overall thera-
peutic process. S3 with high degrees of positive change
will on late measures perceive their therapist as being
more clear and meaningful than S8 with low degrees of posi-

tive change.

Hypothesis IV

 

There is a significant difference between successful
and unsuccessful groups with respect to levels of client-
perceived therapist self—disclosure in the overall thera-
peutic process. Successful SS will on late measures per-
ceive their therapist as being more clear and meaningful

than S5 with low degrees of positive change.

Hypothesis V

There is a significant difference between more and
less disturbed groups with respect to levels of client—
perceived therapist self-disclosure. More disturbed SS
will, early in therapy, perceive their therapist as more

ambiguous and veiled than less disturbed SS.

CHAPTER III
METHODOLOGY AND PROCEDURES

A Semantic Differentiall was used to measure the
early-late dimensions Of client—perceived therapist self-
disclosure. The concept "Counselor" was judged by each
subject over a series of Sixteen polar—Opposite adjectives,
i.e., fair—unfair, strong—weak, and active—passive. The
subjects' selection Of positions in the direction of
successive alternatives gradually eliminated uncertainty

Of "meaning" as to the stimulus concept, "Counselor."

Client-Perceived Therapist Self-Disclosure

 

The Semantic Differential used in this thesis was
made more sensitive by inserting a seven-point scale
between each pair of adjectives, so that the subject can
indicate the "intensity" as well as the "direction" of each
judgment (see Appendix A for explanations of "intensity"
and "direction"). The scale positions of 1 and 7 are
considered polar—Opposites, and position A is a neutral
"meaningless" origin. Positions 2, 3, 5, and 6 are viewed
as the discriminating positions on the scale, indicating

most clearly the direction and intensity of "meaning."

 

1Note: A complete description and rationale of the
semantic differential used in this thesis is found in
Appendices A and B.

2l

22

In the present thesis the S was not as concerned over
the "direction" or structure of "Counselor" meaning provided
by client judgments of the concept "Counselor" over the
various polar-opposite adjective scales. The S was more
interested in the "intensity" or clarity Of the subjects'
judgmentscfi‘the concept "Counselor" along the seven-
positions continuum of meaning. The perceptual clarity of
"Counselor" meanings was best Observed in the subjects' use
of scale positions over the entire series of bipolar
judgments (for instrumentation rationale see Appendix A).

Client perception of therapist's meanings, whether
those meanings were ambiguous—veiled or clear—Open, are
Operationally defined in this thesis by the subjects'
scale-checking behavior. Frequency of the S's use Of
positions 1, A, and 7 were defined as an index Of client—
perceived therapist self-veiling or ambiguity. Frequency
of the S's use of positions 2, 3, 5, and 6 were an index
Of client—perceived therapist self-disclosure or clarity.

The S assumes here that the more veiled and ambiguous
a therapist is with his intentions and meanings, the less
clear and meaningful will he be perceived by the client
during therapy. It is also assumed that the more ambiguous
or veiledeatherapist remains to the client, the less dis—
criminating will be a client's judgments of the concept

"Counselor" over a series of semantic differential scales,

like those used in this thesis.

23

The present thesis, then, tests the hypothesis that
SS using higher frequencies of less discriminating scale
positions (1, A, and 7) on early judgments of "Counselor"
will be rated as more successful clients and Show greater
degrees of constructive change than SS using lower fre-

quencies of more discriminating scales (2, 3, 5, and 6).

Psychotherapy Outcome

Counselor ratings of case outcomes will serve
as a measure of therapy success or failure. Each S
was judged by his therapist as being: (a) unsatisfactory--
"u", (b) partly unsatisfactory-~"pu", (c) partly satis-
factory--"ps", or (d) satisfactory-~"s". The rating was
made at the conclusion of therapy. Only SS rated "S" were
categorized as highly successful in the present thesis,

and SS rates "u", "pu" and "ps" were categorized as less

successful.

Constructive Client Change and Disturbance
The Minnesota Multiphasic Personality Inventory (MMPI)
was used to measure constructive client change and assess
levels of client disturbance. Cooke's (1967) methods of
actuarial diagnosis of MMPI profiles were used to obtain
a formula score on each S's MMPI profile. The MMPI and

one of Cooke's actuarial formulas are discussed in

Appendices C and D.

2A

In the present thesis the pre-MMPI formula scores
established an initial level of disturbance. Also differ-
ences between pre—and post-MMPI formula scores provided

an actuarial assessment of the degree of constructive

client change for each S.

Qperational Hypotheses

Hypothesis I

Ss with high degrees of positive change as measured
in differences between pre-and post—MMPI formula scores
will, early in therapy, perceive the therapist as being
more ambiguous or veiled and will use a higher frequency
of "neutral" or "meaningless" positions (1, A, and 7) over
an early series of sixteen, seven-point semantic differ-
ential scale judgments of the concept "Counselor," than

SS with lower degrees of positive change.

Hypothesis II
SS judged by their therapist as being successful (or

"5") clients on a four-point scale1 will, early in therapy,
perceive the therapist as being more ambiguous or veiled
and will use higher frequencies of meaningless positions
(1, A, and 7) over an early series of sixteen, seven—point

semantic differential scale judgments of the concept

lNote: "u" refers to unsuccessful; "pu" refers to
partly unsuccessful; "ps", to partly successful; and "s",
to successful.

25

"Counselor," than SS rated by their therapist as being less

successful (or "u", "pu", or "ps").

Hypothesis III

SS with high degrees of positive change as measured
in differences between pre-and post-MMPI formula scores
will, in overall therapy, perceive the therapist as being
more clear and meaningful and will use a higher frequency
of discriminating positions (2, 3, 5, and 6) over a late
series of sixteen, seven-point semantic differential scale

judgments of the concept "Counselor," than SS with lower

degrees Of positive change.

Hypothesis IV
SS judged by their therapist as being successful

clients ("5") on a four-point rating scale ("u", "pu",
"ps", and "3") will, in overall therapy, perceive the
therapist as being more clear and meaningful and will use
a higher frequency of discriminating positions (2, 3, 5,
and 6) over a late series of sixteen, seven-point semantic
differential scale judgments of the concept "Counselor,"

than SS judged to be less successful ("u", "pu", or "ps").

Hypothesis V
SS With high ranking MMPI formula scores (more

disturbed) will perceive the therapist during early inter-
views as being more ambiguous or veiled and will use a

higher frequency of "neutral" or "meaningless" positions

26

(1, A, and 7) over an early series Of sixteen, seven-point
semantic differential scale judgments of the concept
"Counselor," than SS with lower ranking MMPI formula scores

(less disturbed).

Tpe Source of the Dapg

Data relevant to the present thesis were available
as part of an ongoing research program at the Michigan
State University Counseling Center. Fifty-four late
adolescent undergraduate students volunteered to partici-
pate in the above program. They came to the Center on a
self-referral basis, seeking counseling for social and
personal concerns.

All of the above clients were first seen in an intake
interview. A decision was made, between client and coun-
selor, about the appropriateness Of psychotherapy. Most
available prospects not eXperiencing previous psychotherapy
were then asked by their screening counselor to take part
in the research project. Clients were assigned to project
therapists on the basis of staff time and competencies.
Final acceptance of clients was determined entirely by the
therapist assigned to the case. His acceptance of the
Client was a subjective decision based on client's
"readiness" for therapy. Screening notes on the intake

session were made available to each therapist.

Therapy sessions at the Counseling Center are gener-

ally conducted for one hour each week. Short-term therapy

27

is encouraged and as a rule such cases tend to consist of
ten to twenty interviews, with some exceptions extending
in either direction.

Part of the research project was to give a form of
the Semantic Differential before therapy, after every fourth
interview, and immediately after termination of the client.
SS were asked to judge twenty-one concepts on the semantic
differential, including the concept "Counselor," over
sixteen adjective bipolar opposite scales (Appendix B).
MMPI profiles were also Obtained for most of the project

SS before the first and after the last therapy sessions.

Selection of Cases

All SS and data selected for the present thesis were
drawn from the project described in the section above.
The selection of cases was based on a number of criteria
differing slightly in each hypothesis. However, (a) all
cases involve no less than four therapy sessions, (b)
cases had been terminated, and (c) all Ss were female.

Pre— and post-MMPI profiles were required on all SS
for Hypotheses I, III, and IV, to establish initial levels
of client disturbance.

Both pre— and post—MMPI profiles were needed to
assess adjustment changes in SS for Hypotheses I and III.

Therapist ratings were also required for determining

successful and unsuccessful Ss for Hypotheses II and IV.

28

Semantic differential responses after the fourth
interview were needed as an early measure Of client-
perceived therapist self-disclosure for each S in Hypoth-
eses I, II, and IV. Semantic responses which followed the
eighth interview were used to assess the overall amount of
"self—disclosure" for SS in Hypotheses III and IV.

A total of twenty-six female SS met the criteria for
part or all Of the hypotheses. Table 1 contains a summary
description and distribution Of client and therapist char-
acteristics. The all female sample selected for the
present thesis was reduced to 26 (NSS = 26) by problems of
missing data and by limits Of the sglection criteria noted
above.

Not every subject selected for this present thesis,
however, could be used in all five of the hypotheses
tested, due to missing data. Loss of data accounts then
for the variance in total numbers, as well as some differ-
ences of individual SS, listed witin the various hypotheses
tested. The latter loss Of data is reflected in the tables
listed at the conclusion of Chapter IV.

A descriptive summary of overall client, therapist,
and therapy characteristics is presented in Table 1.
Interviews for all SS used in this thesis ranged from A to
37 sessions. The mean average of interviews for all SS
was 13.8. Therapists for all Ss were 17 males and A

females. The latter therapists had a considerable range

29

or eXperience:l A senior staff (2 males, 2 females),
7 second-year interns (6 males, 1 female), 8 first—year
interns (6 males, 2 females), and A practicum students

(3 males, 1 female).

lNote: Senior Staff were clinical or counseling
psychologists with 2 to 20 years experience; Interns were
advanced doctoral candidates in clinical or counseling
psychology with an average of 2 years of intensive super-
vision in psychotherapy; and Practicum Counselors were
doctoral candidates involved in their first supervised
counseling experience.

CHAPTER IV
RESULTS OF STUDY

It has been noted in Chapter III that the hypotheses
tested in this present thesis required SS to be divided
into high and low groups within three areas: (a) disturb-
ance, (b) constructive client change, and (c) case
success.

Measures Of initial client disturbance were obtained
from an MMPI administered to nineteen Ss prior to therapy.
All scale scores were then converted into a single formula
score using Cooke's Actuarial Diagnosis Formula (Appendix
D). Nineteen pre—MMPI scale scores, and converted formula
scores for the above subjects, are listed in Table 2. The
assumption of the formula score is that higher scores
indicate increased disturbance. Cooke (1967) established
a cutting line of 550 to designate (a) nonpsychiatric
subjects with scores of 5A9 or less or (b) campus psychi-
atric male SS with scores of 550 or greater. Cooke

indicatedl

that this cutting line was an arbitrary decision
and it would probably vary with the group being studied.
As shown in Table 2, all pre—MMPI scores fell into the

nonpsychiatric category, using Cooke's cutting line Of 550.

1Personal Conversation.

30

31

In order to Obtain client change scores, a post-MMPI
was initially administered to eleven SS (Table 3) and all
scales were converted into a formula score through the
process described above. These post formula scores
listed in Table 3, revealed a total nonpsychiatric distri—
bution (all under Cooke's 550 cutting line). There was,
however, a considerably smaller range of disturbance scores
on post-MMPI'S (316.12 to 387.95 as compared to a pre range
of 268.95 to A58.73) as might be expected after therapy
termination.

Client change scores were finally derived by sub-
tracting the post from the pre formula score on eleven SS
as shown in Table A. A plus (+) Sign was assigned when
post scores were depressed more than fifteen points from
pre scores, indicating high constructive change. A
negative (-) sign was given when post scores were either
elevated or depressed less than fifteen points, indicating
low constructive change.

Table 5 lists the ratings of case outcomes assigned
by each therapist at termination of the S. The latter
provided a measure for low and high case success for
twenty SS [8 successful (S), 7 partly successful (PS),

A partly unsuccessful (PU), and l unsuccessful (U)].
High success groups (Hypotheses II and IV) consisted Of
available successful clients (those rated "8"), while low

success groups were composed from all remaining rated 85

32

(those rated "PS", "PU", AND "U"). The small percentage

of partly unsuccessful Ss Table 5), and particularly the
unexpected availability of only one unsuccessful S, did

not provide a broad distribution of case ratings needed for
a substantial difference between high and low success
groups.

All hypotheses tested in this present thesis required
measurements of client—perceived therapist self-disclosure.
It was noted in the previous section on Operational defi—
nitions that the index of such therapist "disclosure" was
obtained from the scale—checking behavior of Ss on the
early and late administrations of a semantic differential.l
Tables 6 and 7 provided a complete listing of each S's
responses to the concept "Counselor" over sixteen bipolar,
adjective scales for both early and late measures. On the
early administration (after the fourth interview) of the
semantic differential there were twenty-three SS whose
judgments provided a total of 368 responses (nl = 368; see
Table 6)° A total Of 336 responses (h2 = 336; see Table 7)
was obtained from twenty-one SS on the late differential
administered after the eighth interview. Each S responded
a total of 16 times on one or more Of the seven positions,
on both early and late differential measures.

lNote: Checks on positions 1, A, and 7 were indices
Of client—perceived therapist ambiguity or veiling; or low
self—disclosure. Checks on positions 2, 3, 5, and 6 were
indices of client-perceived therapist clarity or Openness;
or high self-disclosure.

33

The final steps taken prior to testing the hypotheses
were those of (a) obtaining scale-checking frequency scores
on each position tested for all differentiated groups1
under each hypothesis and (b) summing the latter positions
scores to derive one total frequency score on each position
as well as a combined position score for all positions
tested. These scale-checking frequency scores are listed
in Tables 8, 9, 10, 11, and 12.

In general, all five investigations (see Experimental
Hypothesis in Chapter III) in the present thesis, were
undertaken to determine if there were significant differ-
ences between various differentiated groups in prOpor-
tionate uses of combined positions (1, A, and 7 or 2, 3, 5,
and 6) as well as uses of individual positions (either 1,

A, and 7 or 2, 3, 5, and 6) on either early or late judg—
ments of the concept "Counselor." Twenty-two different
tests were conducted on the frequency-score data above,
over all investigations. The null hypothesis (H0) in

each of the twenty—two tests was that there was no differ-
ence between tested groups on frequency scores.

The data used in this present thesis involved nominal

and ordinal types of measures. Therefore, the latter

lNote: Six different groups were used in this thesis:

High Constructive Change Ss; Low Constructive Change SS;
High Success SS; Low Success Ss; More Disturbed S3; and
Less Disturbed SS.

3A

measures were apprOpriate to the application Of the Mann-
Whitney U Test (see Siegel, 1956). A total of twenty-two
Mann—Whitney U, rank order statistics (a = .05; one tailed
tests) were used over all twenty-two hypotheses.

Figures 1 through 5 Show the comparative percentage
differences of all positions on combined as well as indi-
vidual positions, between the various groups tested.

Table 13 contains all significant probabilities for values
as small as observed values for each computed Mann-Whitney

U.

Statistical Results of the Study

 

Hypothesis I

 

Null Hypothesis I (HI 1:: NO differences will be
found in the frequency scores of positions 1, A,
and 7 (combined and/or individual position scores)
between High and Low Constructive Change SS, on
their early judgments Of the concept "CouRselor,"
over a series of seven-point Semantic Differential
scales.

 

Hypothesis Ia (HIa): High Constructive Change SS
will have a higher combined frequency score on -
positions 1, A, and 7, than Low Constructive
Change Ss on early judgments Of "Counselor."

 

Hypothesis lb (HIb): High Constructive Change SS

will have a higher frequency score on position I,

than Low Constructive Change SS on early judgments
of "Counselor." _

 

 

1Note: The null hypothesis is stated only once, but
is the null for all alternate hypotheses listed; I, Ia, Ib,
Ic, and Id. This above procedure will be repeated under
each Null Hypothesis.

35

Hypothesis Ic (HIc): High Constructive Change Ss
will—have a higher frequency score on position A,
than Low Constructive Change SS on early judgments
of "Counselor."

 

Sypothesis Id (HId): High Constructive Change SS
will have a higher frequency score on position 7,
than Low Constructive Change SS on early judgments
of "Counselor."

 

A comparison of five High Constructive Change SS with
five Low Constructive Change SS, in terms Of actual fre-
quency scores on positions 1, A, and 7 for early judgments
of "Counselor," is presented in Table 8. High Constructive
SS checked the combined positions 1, A, and 7 more frequently
than Low Change Ss; 31/80 responses (or 39% Of all position
responses; see Figure l) were checked on the combined
positions of l, A, and 7 for the High Change group, as
compared to 11/80 (13%) for the Low Change group. The
latter differences between High and Low Constructive
Change S5 was significant at the .016 level (Table 13).

The null hypothesis for test Ia, of no differences, was
rejected (a = .05; one tailed), and the alternate
Hypothesis Ia was supported.

It can be noted in Figure 1, that High Constructive
Change SS had consistently higher percentage frequency
scores on each individual position (1, A, and 7) on these
early judgments. These latter individual position differ-
ences, however, were not significant (Table 13); so the
null hypothesis for tests Ib, IO and Id was accepted,

that no differences exist between High and Low Constructive

Percentages

.A0‘ .39

    

.32-

.2A‘

.16”

.08“

 

.OO

36

Combined

 

High Change

Figure l.—-Percentages
positions (1, A and 7)
tive Change groups, on
"Counselor".

Low Change

Groups

of neutral and meaningless
checked by High-Low Construc-
early judgments of the concept

37

SS on individual frequency scores for positions 1, A

and 7.

Hypothesis II

 

Null Hypothesis II (HIIO)1: NO differences will be
found in the frequency scores of positions 1, A, and
7 (combined and/or individual position scores)
between High and Low Success SS, on their early
judgments of the concept "Counselor," over a series
of seven-point Semantic Differential scales.

 

_ypothesis IIa (HIIa): High Success SS will have
a higher combined frequency score on positions 1,
A, and 7, than Low Success SS on early judgments
of "Counselor."

 

Hypothesis IIb (HIIb): High Success Ss will have
a higher frequency score on position 1, than Low
Success SS on early judgments of "Counselor."

 

_ypothesis IIc (HIIc): High Success SS will have
a higher frequency score on position A, than Low
Success SS on early judgments of "Counselor."

 

Hypothesis IId (HIId): High Success Ss will have
a higher frequency Score on position 7, than Low
Success SS on early judgments of "Counselor."

 

There were masignificant differences between High
and Low Success Ss' frequency scores (Table 13) on posi-
tions 1, A, and 7, combined and individual positions
(regardless of unexpected inconsistencies in percentage
differences noted in Figure 2 and Table 9). The Null
Hypothesis for all tests, Ila, IIb, IIc, and IId, was

accepted under Hypothesis II, of no differences between

 

lNote: Null Hypothesis is stated once for all
following alternate hypotheses.

Percentages

 

38

    

 

.56
o.A9
.A8
.A0 Combined
.35
.3A
_32 //<
/
.214 /
/” Position A
I,/
6 /
.12
A.
------------ .08
.08 -09 Position 7
.00
High Success Low Success
Groups
Figure 2.--Percentages of neutral and meaningless

positions (1, A, and 7) checked by High—Low Success
groups, on early judgments of the concept "Counselor".

39

High and Low Success SS on early uses of positions 1,

A, and 7.

Hypothesis III

Null Hypothesis III (HIIEO): No differences will
be found in the frequency scores of positions 2,
3, 5, and 6 (combined and/or individual position
scores) between High and Low Constructive Change
SS, on their late judgments of the concept
"Counselor," over a series of seven-point

Semantic Differential scales.

Hypothesis IIIa (HIIIa): High Constructive Change
Ss Will have a higher combined frequency score on
positions 2, 3, 5, and 6, than Low Constructive

Change SS on late judgments of the concept
"CounseIor."

Hypothesis IIIb (HIIIb): High Constructive Change
SS will have a higher frequency score on position 2,
than Low Constructive Change SS on early judgments

of "Counselor."

Hypothesis IIIc (HIIIc): High Constructive Change
Ss will have a higher frequency score on position
3, than Low Constructive Change SS On early

judgments of "Counselor."

Hypothesis IIId (HIIId): High Constructive Change
SS will have a higher frequency score on position
5, than Low Constructive Change SS on early
judgments of "Counselor." —

Hypothesis IIIe (HIIIe): High Constructive Change
SS will have a higher frequency score on position
6, than Low Constructive Change SS on early
judgments Of "Counselor." ’

All combined and individual position scores, except

for position 6 (Table 10 and Figure 3) were unexpectedly

higher for Low Constructive Change SS. Note that the

latter differences were significant on the combined score

(positions 2, 3, 5, and 6) and for scores on individual

positions 3 and 5 (Table 13 and Figure 3). Since all the

.90

.80

.70

O\
0

U7
O

Percentages
.1:-
O

.30

.20

.10

.00

 

A0

 

 

.93
Combined
.50
- Posjj‘inn 7 :1 ~27
'26' A.25
20’:— Position 6 / /, I ’ '23
o "'——- — 2 ’ I - ”.17
P05113109 $/’Position 5
/ ’
.0A A"
High Change Low Change
Groups

Figure 3.-—Percentages Of discriminatory positions
(2, 3, 5, and 6) checked by High-Low Constructive
Change groups on late judgments of the concept

"Counselor".

‘3

Al

differences on the above positions were, however, in the
direction Opposite to that expected (High Change > Low
Change), the null hypothesis (HIIIO) was accepted for all
tests, IIIa, IIIb, IIIc, IIId, and IIIe, that no differ-
ences would be found between High and Low Constructive
Change Ss on late judgments of the concept "Counselor"

over all positions 2, 3, 5, and 6 (combined and individ-

ual).

Hypothesis IV

Null Hypothesis IV (HIV ): No differences will be
found in frequency scor S of positions 2, 3, 5, and
6 (combined and/or individual) between High and
Low Success SS, on their late judgments of the
concept "Coufiselor," over a series of seven—point

Semantic Differential scales.

Hypothesis IVa (HIVa): High Success SS will have
a higher combined frequency score on positions 2,
5, and 6, than Low Success SS on late judgments of

the concept "Counselor."

3.

have

Hypothesis IVb (HIVb): High Success SS will
Low

a higher frequency score on position 2, than
Success SS on late judgments of the concept

"CounselOr."

have

Hypothesis IVc (HIVc): High Success Ss will
Low

a higher frequency score on position 3, than
Success SS on late judgments of the concept

"CounselOr."

Hypothesis IVd (HIVd): High Success 85 will have
a higher frequency score on position 5, than Low
Success SS on late judgments of the concept

"CounselOr."

High Success SS will have

Hypothesis IVe (HIVe):
Low

a higher frequency score on position 6, than
Success Ss on late judgments of the concept

"CounselOr."

A2

There were no significant differences between High
and Low Success SS on late judgments of combined positions
2, 3, 5, and 6, as well as on individual positions 3 and
5 (Table 13). Small percentage differences between above

groups are noted on their uses of positions 3 and 5
(Figure A). Positions 2 and 6 were significantly different

between above groups at the .007 and .025 levels, respec-

tively. The null hypothesis for tests IVa, IVc, and IVd,
of no differences in groups on this late measure, was
accepted. The alternate hypotheses, HIVb and HIVe, were
supported by significant data noted above and in Table 13,
that High Success SS would use higher frequencies Of

positions 2 and 6, than Low Success SS.

Hypothesis V

Null Hypothesis V (HVO): No differences will be
found in the frequency-Scores of positions 1, A,
and 7 (combined and/or individual position
scores) between More and Less Disturbed SS, on
their early judgments of the concept "Counselor,"
over a series Of seven-point Semantic Differen-

tial scales.

Hypothesis Va (HVa): More Disturbed SS will have
a higher combined frequency score on positions 1,
A, and 7, than Less Disturbed SS, on early
judgments of "Cousnelor."

Hypothesis Vb (HVb): More Disturbed SS will have
a higher frequency score on position I, than
Less Disturbed SS, on early judgments of
"Counselor." -

Hypothesis Vc (HVc): More Disturbed Ss will have
a higher frequency score on position A, than
Less Disturbed SS, on early jdugments of
"Counselor." _

Percentages

A3

.90.
.80- .82

Combined
.70“

.60-
.59

.50~

.AO~

'30' .27

.20- .2ITL‘= :"

     
  

Position 3 -4K 21

, - — 1 ‘A .20
Position 5

  

'10' .09

.00

 

 

High Success Low Success

Groups

Figure A.-—Percentages of discriminatory positions
(2, 3, 5, and 6) checked by High-Low Success groups
on all late judgments of the concept "Counselor".

AA

Hypothesis Vd (HVd): More Disturbed SS will have

a higher frequency score on position 7, than Less

Disturbed SS, on early judgments of "Counselor."

A comparison is made in Figure 5 of More and Less
Disturbed groups in terms of judgments on positions 1, A,
and 7. Although the More Disturbed group used considerably
larger percentages of responses on position A (28% com-
pared to 13%), no significant differences between dis-
turbance Ss were found on any frequency scores, combined
or individual, for positions 1, A, and 7, on these early
judgments of "Counselor" (Table 13). The Null Hypothesis
(HVO) was accepted for all tests, Va, Vb, Vc, and Vd, that
no differences exist between More and Less Disturbed SS
in their early use of positions 1, A, and 7 (combined and
individual) on early judgments of "Counselor."

A complete summary of significance probabilities
associated with each observed Mann-Whitney U, and a

statement of rejection or acceptance of each null hypoth-

esis tested in this chapter, are listed in Table 13.

Percentages

A5

 

 

.A8 1
.A0 3
.37 Combined
'32 -. .32
l
'2“ 1 ‘\\Position A
\
6 \ \
.l -
\ xx .13
A 10
-08 — Posit - - ‘ '
.06 - - y - .09
. A- —' " " Position 7
000 .4 O3
More Disturbed Less Disturbed

Groups

Figure 5.-—Percentages of neutral or meaningless

positions (1, A, and 7) checked by More-Less

Disturbance groups on all early judgments Of the
concept "Counselor".

A6

TABLE l.-—Descriptive summary of client, therapist and
therapy characteristics.

 

 

 

 

 

Client Therapist Therapy
Subject 8 Code Total
Number Sex* Number Sex Level** Igggiiiews

3 F 78 M 11 7
A F 35 M 12 6
5 F 33 M I2 A
6 F 65 M I1 6
7 F 79 M I1 8
9 F 7A M Il 1A
10 F A0 M s 6
11 F 75 M II 18
15 F 85 M I1 15
16 F 33 M 12 22
17 F 70 F I1 16
20 F 32 M 12 22
25 F 15 F s 23
26 F A1 F I2 18
28 F 39 M I2 11
29 F 77 M 11 7
32 F 11 M s 7
33 F 78 M 12 9
3A F 19 F s 6
A0 F 53 F 11 17
A2 F Ac M s 18
A5 F A5 M I2 20
“7 F 59 M P 12
50 F 95 M P 13
51 F 56 F P 18
5" F 98 M P 37

*H_refers to male; H refers to female.

**P refers to Practicum Student; I refers to first-

Year Infern; I refers to second-year IAtern; and S refers
to Senior Stafg. -

A7

 

mp.om: mm mm mm NO mm N: O: mm Om mm mm om mm mm mo mo Om

 

 

 

mO.OOm OO HO OO ON OH ON ON OO HO ON NN OH NH ON OO mO HO
NN.OOm OO Om NO ON mN Om Nm OH Om Om Nm ON OH OH HH OO OO
OO.OOm HO OO OO HO OH Om HO NH OO ON OH Nm OH HH OO NO NO
OO.OOO OO Om OO OH OH ON NH OH OO NH ON OH OH OH mO HO NO
HN.OOm OO mm OO Nm OH ON Hm HH mO ON ON HN HH OH OO NO OO
OO.HOm NO mm Om HN ON mm Hm OH Om ON ON ON mH NN OO NO Om
HO.OOm Hm Om mm mm ON Hm NO mH NO ON Hm ON ON OH NO OO Nm
OO.OON Om OO Om ON OH OH ON HH mm mH ON OH NH OH OO HO ON
Oe.mmm HO HO OO Om ON Hm Om mH Om ON ON Om OH OO OO NO ON
mO.OOO Om mm HO mm ON OO OO OH NO ON ON AN OH mH OH OO ON
Om.OOO Om HO OO Om HN HO HO NH Om ON ON Nm HN OH HO mO ON
ON.OOm OO Om OO mm HN ON HO OH em ON OH OH OH HH mO OO OH
NO.OON HO Om OO Om OH HN ON OO HO ON OH NN OO OO OO mO OH
O0.00m OO OO Om OO OH ON OO mH OO ON mN ON NH OH OO mO OH
HO.OOO Om HO OO Hm NN Om OO OH OO Hm Om mm ON NH OH NO HH
OO.eOm OO HO OO OO ON mm mm OH Om Hm ON Om mH OH OH OO O
Om.eNm OO OO NO ON OH ON ON HO mm ON HN ON NH NH HO OO O
NH.OOm mO OO OO ON ON mm Om OH OO mN ON NN NH OH OO OO O
mm m <
mohoo ehHoz HO O: om ON em as ON HO O mm x e H
m monasz

mmsEHom m.poohn:m

OHmom seem so monoom 3mm

 

.mfimmfip m.m CH Ummd mm COOPOCHC (MOM mthow MHSELO.H USN 3a.». Hngtmfimllom Mdm<8

A8

 

OO.HOO Om OO OO Om mN Hm Om mH OO HN mN ON OH NH OO mO OO

 

 

 

OO.mmm mO OO Om mN OH ON NN OH Om NN ON OH HH OH mO mO HO
OO.NNm NO OO mO ON OH ON ON OH Om Nm ON ON OH OH OO HO OO
NH.OHm HO OO HO Om OH ON Om OO mO NN OH ON NH HH NO mO NO
O0.0Nm NO Nm mO NH ON ON ON OH NO ON ON OH NH OH mO NO NO
O0.00m OO mO OO OO OH Hm om HH OO NN mN mN NH OH OO mo OO
OO.mOm OO HO Om om ON ON ON OO Om ON mN OH NH HN NO mO Om
OO.HOm OO OO HO Om HN Om ON OH mO ON ON HN mH OH OO OO Nm
O0.0Nm OO OO OO ON ON ON ON HH Om ON mN ON HH NH mO NO ON
Om.OOm OO OO OO Om HN ON mm OH mO ON HN OH ON mH OO HO OH
OO.Hmm NO mO OO Om OH ON mN OH OO ON HN HN OO mH OO mO OH
mm m 4

hosoom emth HO O: om ON OO NO ON OO O mm O O H honeOz
NHSEHom m.uoomnsm

mfiwom 30mm CO wmhoom 3mm

 

.mfimmflu WOQ CH Ummd mm C®>mH® LCM mmhoom MHSEQOM USN Swh Hmzzlpmomll.m mgm<9

A9

TABLE A.-—MMPI change scores and Signs on eleven Ss used in

Hypotheses I and III.

 

 

Pre—MMPI Post-MMPI
Sfifiiiit Fggggga Fggggga 323358 Sign

16 270.62 331.96 -61.3u -
17 330.28 3A6.3O -16.02 -
25 355.78 328.A8 27.30 +
32 395.A1 361.90 33.51 +
3“ 381.59 3A3.89 37.70 +
A0 356.21 355.09 1.12 -
A2 308.8A 329.05 -20.81 -
A7 389.87 316.12 73.75

50 399.22 322.90 76.32

51 3A7.93 333.85 1A.08 -
5A A2O.75 387.95 32.80 +

 

50

TABLE 5.—-Therapists' ratings of case outcomes for
twenty SS used in this thesis.

 

 

 

Subject Raging
Number Therapist
3 PU
A PS
5 S
6 U
7 PU
9 PU
lO PU
15 S
16 S
17 PS
20 PS
25 PS
28 S
29 PS
32 PS
33 3
3A S
"'70 S
A2 S
A5 PS
U represents unsatisfactory.
PU represents partly unsatisfactory.
PS represents partly Satisfactory.
S represents Satisfactory.

51

.Lmoezc m.m 0» memp wmcflommz CESHoo umpoz*

 

 

H o H m m m o m H H s s m H m m m m H s H w H mHompmuwcchmco
m m H H m m m H H H H H H H s m m m H m H m m pmmanonm
o H m H m m o m s H H s m H m m m m H H s s H m>HHoseoem-m>Hpo:chma
o o m H o m s m s m H s o m m o m m m o s s H eoouuemm
m m m H m m H H m H m o m m m m H H H m H H H m>Hmmmmnm>HHo<
H m m H m H H m H H m H m H m m m m H H H m H thoczuthm
m m m H m m m H m m m H m m m m m m m m H m H emchppmcooummem
m m m H w m o m o H s m m m o o m m H s H o H wcospmuxmmz
m m m H m m m m m m H H m H m m m m m H H H m OHSWHeHmwmauoHpmHerao
m H m H H H m m H H m H m m m m m m H m H m H HHzouapmem
H m .H H H H m m H H H H m H m m H H H H H m m ch>upmmeoz
m m m H m m m m m H H H m m H m m m H H H H H wsopmwcmanmawm
H m w H m o m H m m H H m m m m m m m m H H H oHsoHooHouzmmm
o m m H m m o H m H o s m H m w m m H s s s m HampOHoonmmmHLOHoo
H m m H m H m H m m H m H H m m H H m m H m H HOHHH-CHHH
H m m H m H m m m m m m H H m m m H m m H m H HHHEmnmmemH
Hm Hm om mH wH OH Hm mm mm mm mm mm mm om NH 0H mH OH 0 H m H *m

 

pomfinzm comm an omxomno wcoHpHmom

mmHMom

 

.Hmfipcmpmumfio oapcmEmm mnp.mocowpmpuchHE©m zapmo am so Ummmm
=.Loaomcsoo= uaoocoo mnp go mucoemosh co mm mmpzulmucmzu mp ANIHV omxoono mcofipfimomll.m mqm<e

52

.LmnEsc w.m on swamp mmcflomo: CESHoo "mpoz*

 

H o H m m m m H m H m m m m m H m H H m H memHmuwcHwemeo
H m H o m o H H H H m H m m m m H m H m m pmmmnonm
@ m m o m o m H w H m H m H m w m w H o H m>Hooseocaum>HHosgumma
o m w m H m m H H H m H m m H m o m H o H cooouemm
m m H m m m m o m H m H m m m m m m H H H m>memmnm>Hpo<
H m m m m m m H m H m H m m H m m m m m H HHmHeaupHmm
H m m H m H H m m m m m m m m m m m H H m emeHmHHmeooummcm
H m m m H H m m m H w H m o H m H m H H H weoppmnxmwz
H m m m m H m H H H m H m m H m m m H m H OHHmHeHmmmmuOprHeHoao
m H H H m m m H m H m m m m m m H H H m m HHHQIQHmem
m m H w H m m H m H m H m m m m m H H m m eHm>nummeoz.
m m m m m m m m m H m H m m H m H m H m H msonmmcmoumomm
m .m m m m m m m m H m H m m m m m m H m H HHHOHHHHQIHmmm
o m m w m m H H m H m H m o m m H m H w m HzocoHooummmHtoHoo
m m H m m m H m m w m H m m m m m m H m m HOHeenche
H m m m m m m m m m m H m m m H m m m m m HHHEmummHmH

Hm Hm HH mH oH Hm mm mm mm mm mm om HH mH mH HH 0 H o m Hm

 

Homnpsm somm Hp emxomeo mCOHHHmom

 

=.moaomc:oo: pamocoo mnp mo mpcmeozm so mm ozonmpcmzu an AHIHV onomco mcofipdmomll.H mqm¢e

53

TABLE 8.--Frequencies of neutral or meaningless positions
(1, A, 7) checked by high and low constructive change groups
on early Semantic Differential Judgments of the concept

 

 

 

 

 

 

 

"counselor." ‘

1

1‘:

High Change Group Low Change Group g
Subject POSitionS Frggzgicy Frggfizicy Positions Subject

Number 1 A 7 Score Score 1 A 7 Number ,

:5;
26 6 2 A 12 O O O 0 16
32 l A 2 7 2 O l l 17
3A 2 O l 3 5 2 3 0 A0
50 O 3 O 3 2 O 2 0 A2
5A 2 A O 6 2 O l l 51
Sums ll 13 7 31 ll 2 7 2 Sums

 

5A

TABLE 9.—-Frequencies of neutral or meaningless positions
(1, A, 7) checked by high and low success groups on early
Semantic Differential Judgments of the concept "counselor."

 

 

 

 

 

 

High Success Group Low Success Group
Positions Total Total Positions

5339):? 1 u 7 ”:23:ch “323:2” 1 u 7 5339):?
5 8 l 7 16 12 7 l A 3
15 l 2 O 3 7 A O 3 A
16 O O O O 8 l A 3 7
28 2 A A 10 10 O 10 O 9
33 O 6 O 6 O A O 10
A2 0 2 O 2 9 l 8 0 20
3A 2 O l 3 l0 O 10 O 29
£9 2 3 o _2 7 l A 2 32
Sums 15 l8 12 A5 15 0 15 0 A5
2 O l 1 l7
3 O 3 O 22

6O 13 Sums

CO

N

H
J:’

 

55

TABLE lO.--Frequencies of discriminating positions (2, 3, 5,

6) checked by high and low constructive change groups on

late Semantic Differential Judgments of the concept
"counselor."

 

 

 

 

 

 

 

 

High Change Group .Low Change Group
Positions Total Total Positionsw
gubgect Frequency Frequency fifiggggt
um er 2 3 5 6 Score Score 2 3 5 6
26 l O l 1 l5 6 5 l 3 16
32 3 2 O 2 l6 3 5 7 1 17
3A 7 o o 6 13 1A A 3 5 2 A0
A7 5 1 1 3 10 14 3 5 3 3 A2
5A 5 O l A 10 15 6 3 2 A 51
SEES 21 3 3 16 A3 7A 22 21 l8 l3 Sums

 

56

TABLE ll.--Frequencies of discriminating positions (2, 3, 5,
6) checked by high and low success groups on late Semantic
Differential Judgments of the concept "counselor."

 

 

 

 

 

 

 

 

High Success Group Low Success Group
Positions Total Total Positions
Subgect Frequency Frequency figggzgt
um er 2 3 5 6 Score Score 2 3 5 6
5 5 O 1 7 13 7 1 3 2 1 3
15 3 5 l 2 11 2 O l l O 6
16 6 5 1 3 15 13 3 A 2 A 7
28 6 2 1 A 13 11 2 2 5 2 9
33 O 7 5 O 12 16 3 5 7 l 17
3A 7 O O 6 13 3 O 2 l O 20
A0 A 3 5 2 1A 16 O 8 7 1 25
A2 3 5 3 3 1A 7 3 2 O 2 32
36;: 3A 27 17 27 105 75 12 27 25 ll Sums

 

57

TABLE 12.--Frequencies of neutral or meaningless positions
(1, A 7) checked by high and low disturbance groups on early
Semantic Differential Judgments of the concept "counselor."

 

 

 

 

 

 

Higher Disturbance Group Lower Disturbance Group
Sfiggggt Positions Frggfigicy Frggtgicy Positions fifiggzgt
l A 7 Score Score 1 A 7

9 O 10 O 10 l A 3 7
15 l 2 O 3 O O O 16
2o 1 8 o 9 0 l 1 17
25 O 3 O 3 12 6 2 A 26
32 l A 2 7 10 2 A A 28
3A 2 o 1 3 2 3 0 A0
5A 2 A O 6 O 2 0 A2
—- '— O l l 51
Sums 7 31 3 A1 __ __

A1 11 17 13 Sums

 

58

TABLE 13.--Significance probabilities associated with all
values as small as observed values of U in Hypotheses I-V.

 

 

 

Alternate Mann- Null
Hypothesis Position(s) Whitney U Hypothesis
(Directional) (PS) (No Differences)

la 1, A & 7 (combined) .016* reJected

lb 1 .075 accepted

Ic A .3A5 accepted

Id 7 .210 accepted

IIa l, A & 7 (combined) n.s. accepted

11b 1 n.s. accepted

IIc A n.s. accepted

IId n.s. accepted

Illa 2, 3, 5, & 6 (combined) .OOA+ accepted

IIIb 2 .500 accepted

1110 3 .00A+ accepted

IIId 5 .008+ accepted

IIIe 6 .3A5 accepted

IVa 2, 3, 5, & 6 (combined) .117 accepted

IVb 2 .007“ reJected

IVc 3 .A39 accepted

IVd 5 .253 accepted

IVe 6 .025* reJected

Va 1, A & 7 (combined) .232 accepted

Vb 1 .A33 accepted

Vc A .095 accepted

Vd 7 .095 accepted
Note: * refers to significance at a .05.

+ refers to significance
direction.

at d .05 in opposite

CHAPTER V
SUMMARY AND CONCLUSIONS

The general disclosure of therapists' "meanings" and
"intentions" to the client, has been considered within
Rogerian theory as a necessary aspect of successful psycho-
therapy. Recent suggestions have been made, that client-
perceived therapist self-disclosure, to be therapeutically
effective, is probably contingent on the ambiguity or
veiling <3f therapists' meanings and intentions during

early therapy sessions.

Summary

The major purpose of the present thesis was to test
the predictions that there would be significantly higher
frequencies of client-perceived therapist ambiguity or
veiling indices (low self-disclosure) in early measures
of (a) subjects showing higher degrees of constructive
change than §S with low degrees of change, (b) Ss given
high success ratings than with SS with low success ratings
and (0) SS showing more degrees of disturbance than with
§s with less degrees of disturbance. It was also predicted
that there would be higher frequencies of client-perceived
therapist clarity or meaningful (high self-disclosure)

indices in overall (or late) measures of Ss showing higher

59

60

degrees of constructive change and/or with §S rated as more
successful clients, than with low change Ss and/or with
less successful clients.

Clients used in the present study were twenty-six
self-referred Michigan State University undergraduate
females, who had been seen at the University Counseling
Center for personal counseling.

High and Low Constructive Change Groups were devel—
Oped on the basis of differences between Ss' pre-and post-
MMPI formula scores. Therapists' ratings of individual
case success was the basis of High and Low Success Group
compositions. Pre-MMPI formula scores provided a criteria
for assigning Ss into More or Less Disturbed Groups.

Client's perception of therapist's self-disclosure
was defined as S's scale-checking behavior on a series of
seven-point (position) semantic differential Judgments of
the concept "Counselor." Positions 1, A, and 7 (combined
and individual), the meaningless-neutral scales, were
separate indices of low therapist disclosure, and
positions 2, 3, 5, and 6 (combined and individual), the
most discriminatory scales, were separate indices of high
therapist disclosure.

Sixteen Judgments of the concept "Counselor" were
made by each of the twenty-six Ss on early and/or late
administrations of the semantic differential test. Fre-

quency scores were obtained on the S's uses of the seven

61

alternate positions over 16 of the seven—point scale Judg-
ments of the concept "Counselor."

Twenty-two Mann-Whitney U, one tailed tests (a = .05)
were then used to test predicted differences between the
various groups mentioned above, on their frequency scores
for each separate measure (individual and combined positions
scores) either for 1, A, and 7 or 2, 3, 5, and 6; depending
on whether the hypothesis was investigating low or high
therapist disclosure indices.

The results of the analyses supported only three of
the twenty-two eXperimental predictions of group differ—
ences. As eXpected, High Constructive Change clients did
use greater frequencies (significant at .016) of combined
positions 1, A, and 7 on early Judgments of "Counselor" than
did Low Change clients. On this one combined measure of
positions 1, A, and 7 (low therapist disclosure), there was
support for the hypothesis (d = .05) that High Constructive
Change clients would perceive their therapist's meanings
less clearly during early therapy, than would Low Construc-
tive Change clients. The positions 1, A, and 7, however,
as individual measures of low therapist disclosure, were
not significantly different between the above groups. The
latter individual measures did not support the hypothesis
that there would be greater indices of low therapist
disclosure on earlier individual position measures for

High Change SS, than for Low Change Ss.

62

High and Low Success Ss, as well as More and Less
Disturbed SS, showed no predicted differences in scale-
checking behavior on their early therapy Judgments of
"Counselor" on positions 1, A, and 7 (low therapist
disclosure), either on individual or on combined measures
of positions 1, A, and 7. No support was gained on the
above tests for the presence of early therapist veiling
in those cases rated as successful. Also, no support
was found in the above tests (individual and combined
positions 1, A, and 7) that More Disturbed clients would
perceive therapists' meanings as more veiled in early
therapy than Less Disturbed clients.

No differences were found between High and Low
Constructive Change clients on late (individual and
combined) measures of positions 2, 3, 5, and 6 (high
therapist disclosure). The latter data did not support
the hypothesis that High Constructive Change clients would
have more indices of high therapist disclosure (positions
2, 3, 5, and 6) on overall (or late) measures, than Low
Constructive Change clients.

Significant differences (a = .05) were found between
High and Low Success Ss on two measures of overall (or late)
indices of high therapist disclosure. Position 2 was sig—
nificantly different between High and Low Success groups
at the .007 level, and position 6 at the .025 level.

These two measures supported the hypothesis that High

 

 

63

Success clients would perceive their therapists more
clearly (high therapist disclosure) in overall therapy
than would Low Success clients.

No differences between the same groups above were
found, however, on the other measures of high therapist
disclosure, combined positions (frequencies) of 2, 3, 5,
and 6 and on individual positions (frequencies) 3 and 5.

The results of all twenty—two of the hypotheses
tested in the present thesis, although failing to
generally support the importance of the temporal contin—
gency of early therapist veiling to successful therapy,
did show some encouraging trends, in predicted directions.

Although the statistical analysis did not support
the theory of the present thesis, two results of the study
are cited as reasons for failure; (a) it can be argued
that the fourth and eighth-interview semantic differentials
provided only the middle of the continuum between early and
late client perceptions of "Counselor." Consequently a
much wider continuum between "early" and "late" client per—
ceptions would be needed, and (2) the subjects can be
regarded as homogeneous in terms of disturbance levels
and consequently a more extreme degrees of disturbance

would be needed.

6A

Conclusions

 

Hypothesis I

 

1. Female clients who show evidence of high construc-
tive change, as a result of therapy, are more apt to perceive
less of the therapist's meanings and intentions during early
therapy than low constructive change female clients, when
therapist's meanings are measured on combined "low thera-
pist disclosure" indices; positions 1, A, and 7, from a
series of semantic differential seven-point bipolar scale
Judgments of the concept "Counselor."

2. Female clients who evidence high constructive
change are likely to be no different from low constructive
change females on early perceptions of therapist's meanings
and intentions when these latter client perceptions are
measured on individual "low therapist disclosure" indices;
positions 1, A, and 7, from a series of semantic differ-
ential seven-point bipolar scale Judgments, of the concept

"Counselor."

Hypothesis II

No differences are apt to be found between female
clients rated as successful or unsuccessful in their early
therapy perceptions of therapist's meanings and intentions
when these latter perceptions are measured on individual
or combined "low therapist disclosure" indices; positions

1, A, and 7, from a series of semantic differential

65

seven-point bipolar scale Judgments, of the concept

"Counselor."

Hypothesis III

 

No differences are likely to be found between female
clients who show evidence of high or low constructive
change, as a result of therapy, on their overall (or late)
perceptions of therapist's meanings and intentions, when
the latter perceptions are measured on combined or indi-
vidual "high therapist disclosure" indices; positions 2,

3, 5, and 6, from a series of semantic differential seven-

point bipolar scale Judgments, of the concept "Counselor."

Hypothesis IV

1. Female clients rated as successful cases by
their therapists, are more apt to perceive more of the
therapist's meanings and intentions, during overall (or
late) therapy, than unsuccessful rated female clients,
when therapist's meanings are measured on individual "high
therapist disclosure" indices; positions 2 and 6, from a
series of semantic differential seven—point scale Judgments,
of the concept "Counselor."

2. No differences are likely to be found between
female clients rated as successful or unsuccessful, on
their overall (or late) perceptions of therapist's meanings
and intentions, when the latter perceptions are measured on

"high therapist disclosure" indices; positions 2, 3, 5, and

66

6 combined, or individual positions 2 and 6, from a series
of semantic differential seven-point bipolar scale Judg-

ments, of the concept "Counselor."

Hypothesis V

 

No differences are likely to be found between more
and less disturbed clients, on their early therapy per-
ceptions of therapist's meanings and intentions, when
the latter perceptions are measured on combined or indi-
vidual "low therapist disclosure" indices; positions 1,
A, and 7, from a series of semantic differential seven-

point bipolar scale Judgments, of the concept "Counselor."

Discussion

 

Failure to find support for nineteen of the twenty-
two analyses in this present thesis, requires further
explanation. Important differences between the present
thesis and other studies (Chapter II), which suggest a
temporal contingency to the effectiveness of client-
perceived therapist self—disclosure, may relate to sample
size and composition, instrumentation, and/or procedures.

Unlike previous studies reviewed in Chapter II, all
SS in the present thesis were college females. Sex
differences and institutional relationships of Ss in this
thesis were different from S8 in related studies by
BarrettHLennard, 1962; Demos, 196A; and Truax and Carkhuff,

1965.

67

The counselors in the present thesis were varied in
terms of sex and levels of experience (Table l). The
latter counselor differences could possibly add to the
problems of early control struggles (discussed on page 16
in this dissertation). Such counselor-client control
struggles may account for lower than predicted frequencies
of "low therapist disclosure," positions 1, A, and 7, of
the clients' semantic differential Judgments of the concept
"Counselor" in analyses of tests under Hypotheses I, II,
and V. Three of the five counselors, e.g., involved in
all investigations under Hypothesis I (Chapter IV), worked
with the Low Client Change group. Only one of the three
Senior Staff counselors worked with the Low Client
Change Ss, in studies under Hypothesis I.

One important methodological problem that may have
contributed to limitations on the present thesis, could
be related to the timing of the "early" and "late"
semantic differential administrations. The latter early
differential, e.g., was administered at the conclusion of
the fourth interview. An examination of Table 1 (Chapter
IV) of Ss involved in studies under Hypothesis I (Table
8), shows that the High Change Ss were counseled for a
total of 71 sessions (a mean average of 1A.2 sessions).
Low change Ss, however, were involved in a total of
91 sessions (a mean average of 18.2 sessions). The

latter difference in length of sessions between the groups

68

in Hypothesis I, could suggest that the fourth interview
administration of the semantic differential may not have
been "early" enough, particularly for the Low Change Ss.

Differences in length of therapy (Table l) are also
noted for Ss (Tables 10 and 11) under Hypotheses III and
IV analyses. In the latter studies, overall (or late)
measures were taken from semantic differential Judgments
which followed the eighth interview. Perhaps differences
in average length of therapy between groups, indicate
that the eighth interview is not "late" enough. High
Success SS, e.g., had a total of 80 sessions while Low
Success Ss were seen for a total of 91 session (Table 1).

Another limitation of this present thesis, relating
to length of therapy, centers on the student sample.
Therapy termination of "student" clients, may have con-
sisted of time limitations imposed by the school year, as
well as client and/or therapist decisions.

One possible source of error, specifically relating
to all tests under Hypotheses II and IV, arises from the
division between eXperimental (high success Ss) and
control (low success Ss) groups. The latter division was
made solely on the basis of therapist's ratings, which
reflect only the subJective Judgment of each therapist.
No other objective criteria was used in the above dividing

process.

69

It should be also noted here that the initial
decision in Hypotheses II and IV was to (a) assign only
those Ss rated "successful" (or those assigned an "S";
see Table 5) into the high success groups, and (b) all
other Ss, "partly successful," "partly unsuccessful,” and
"unsuccessful" into the Low Success group. This latter
decision imposed certain limitations on the studies above,
under II and IV. A closer look, e.g., at the low success
group under II tests, revealed that none of those low
success Ss were rated "unsuccessful." Only four of the

"partly unsuccessful,"

eight Low Success Group were rated
with four (one—half of the entire Low Success Group) rated
as "partly successful." The unexpected imbalance in the
ratio of "real" unsuccessful Ss to successful Ss could
well explain the lack of significant differences of groups
in these studies.

Studies under Hypothesis V also ran into some unex—
pected group-division problems, which may have been a sub-
stantial limitation on the latter investigations. Groups
were to be divided between more and less disturbed clients.
It was noted in Chapter IV, that all of the 15 Ss with
pre-MMPI formula scores, fell into the non-psychiatric
category (using Cooke's cutting line of 550). More sig—
nificant results might have been obtained if more sub-

stantial differences in group disturbance levels could have

been achieved.

70

Certain instrumentation limitations must be noted, in
regard to Cooke's formula (Appendix ID), which provided
formula scores for pre- and/or post-MMPI tests used as the
basis for group divisions in Hypotheses I, III, and V. The
formula was derived for college (freshmen) females, who were
enrolled in a nursing program. Some lower reliability
measures than those given in Appendix D (r = .91 and .80)
might be expected since the college undergraduate female
Ss in the present thesis did not represent this single
academic maJor.

All the limitations suggested so far in this discussion
are undoubtedly sources of error, and could well account for
failures to find the significant differences predicted in
this present thesis.

In addition to the three studies (Ia, IVb, and IVe;
see Table 13) in which significant, predicted results were
obtained, there were a number of other encouraging trends
(see Figures 1 through 5) which merit some attention in
this following discussion. These latter trends deal
specifically with the use of alternate semantic differ-
ential positions, as either separate or combined measures
of low or high therapist disclosure.

As suggested in Chapter III, and in Appendix A, each
position on the semantic differential was defined as a
"single" measure of differing intensity or clarity of

client-perceived therapist self—disclosure. The polar

H.”

‘-..n..H-o e

71

positions, 1 and 7, because of their extremity on the
bipolar scale, ordinarily would be considered the most
intense or clear measure of therapists' meanings. However,
as noted in Appendix A, both female and high I.Q. Ss usually
check these polar positions when under "generalized"
anxiety and when Judging an ambiguous concept. Positions
1 and 7, for the latter reasons, were assumed, in this
present thesis, to be single measures of "low therapist
disclosure," in which Judgments of "Counselor" were less
clear to the client. Positions 1 and 7 were then included
with position A, the origin position, considered to be the
most "meaningless" (gray or unclear) Judgment of the concept
"Counselor."

It was assumed in this present thesis that position
A would be the most "powerful" or "reliable" as a measure
of "low therapist disclosure," and would be chosen prOpor-
ltionately more than positions 1 and 7, when Judgments were
made of counselors' meanings as being veiled or ambiguous.
It was, however, expected that all "low therapist dis-
closure" positions (1, A, and 7) would consistently
measure differences between groups as individual and com-
bined measures of "low therapist disclosure." The latter
assumption was not substantiated throughout these studies,
and the contradictory and inconsistent results may raise
some questions about the usefulness of different alternate
positions as measure of therapist disclosure, not in terms

of levels of validity, but of reliability.

 

72

While direct differences between position frequencies,
within the neutral-meaningless position grouping (l, A,
and 7) as well as the most discriminating position grouping
(2, 3, 5, and 6) were not originally designed as studies
in this thesis, the present thesis data contain some
intriguing, post hoc descriptions, that relate to all
hypotheses investigated in the present thesis.

In Figure 6 the rank proportionate uses of alternate
positions (1-7) on total responses for all, undifferen-
tiated, female subJects, for early, late and combined
early-late prOportions is depicted. Note, that when early
and late measures were combined in Figure 6, position A
was used more frequently (19% of all combined responses)
for all undifferentiated subJects. Position 7 was used
least (8%), and position 1 was next to last (10%) in low
frequency use on all responses.

When early and late measures of undifferentiated
subjects are graphically depicted, apart from the combin—
ation early-late responses (see Figure 6), an appreciable
shift occurred on position A (2A% of all early alternate
responses and only 15% of late alternate responses as
compared to the 19% on overall alternate responses). The
latter shift on position A indicates that undifferentiated
female Ss, as might be expected, perceive the concept
"Counselor" as being considerably more meaningless in early

therapy than undifferentiated female Ss do in late therapy.

 

Percent of Alternate Responses

2A -

20 H

16 -

12 -

08 -

0A-

 

00

73

 

 

~

H
R)
LU
.1:-
U"!
axe
4

Positions

Figure 6.--Proportionate uses of alternate positions by
undifferentiated Ss used in this thesis.

Note:

refers to combined Early and Late Frequencies

~— - —— refers to Early Frequencies

—————— refers to Late Frequencies

._l.h

7A

The unexpected differences arise, however, within
position groupings, and are depicted in Figure 6; partic-
ularly significant are the descriptions on the polar
positions, 1 and 7, as well as for position 5. When early
and late responses are partialed out from total responses,
however, the polar positions (1 and 7) e.g., remain at the
same proportions as the total polar responses. This tenta-
tive, post hoc, descriptive analysis suggests, that for
this particular undifferentiated population, the polar
positions are not as useful in their individual contri-
bution to early—late differences of S's Judgments of the
concept "Counselor" as position A. Note that this is also
true for position 5 in the discriminatory group.

It is also suggested in Figure 7 that undifferentiated
Ss, however, to tend to shift from higher uses of neutral-
meaningless position in early therapy to greater uses of
discriminatory positions in later therapy. From this
latter shift it is suggested that these undifferentiated Ss
in Figure 7, characteristically tend to perceive the
concept "Counselor" in clearer terms on later Judgments
than on earlier responses.

The reliability weaknesses of the instrument are
suggested in the findings in Figures 6 and 7, particularly
on the individual use of the latter positions 1, 7, and 5,
as reliable measures of "therapist disclosure." The above

weakness in instrumentation, may account in part for the

Percentages

.70

.60

.50

.A0

.30

.20

.10

.00

 

75

 

 

' (”1.714, 7) <2, 3, 5, 6)
Combined Combined

Positions

Figure 7.——Proportionate differences between meaningless
(1, A, 7) and discriminatory (2, 3, 5, 6) position group
responses on undifferentiated Ss used in this thesis.

Note:

refers to combined Early and Late
Frequencies

—— — —— refers to Early Frequencies

------ refers to Late Frequencies

___....! _

‘A

76

failure to find statistical support on much of the data in

this present thesis.

Implications for Future Research

 

All Hypotheses analyzed in the present thesis remain
somewhat inconclusive (although three were supported and
nineteen found no support), due to limitations discussed
above. The limitations of the thesis, along with encour- ‘ J
aging "predicted" trends on twelve of the twenty—two A
measures (Tables 8-12 and Figures 1-5) invite further ' s
research into the problems investigated in the present
thesis.

Future researchers could design more rigid controls
for (a) Counselor differences, (b) early and late semantic
differential administrations, (c) a better ratio between
client-success groups, and (d) a greater difference in
levels of disturbance between disturbance groups.

Perhaps as important as the increased controls sug—
gested above, future researchers may wish to limit their
measures of "therapist self-disclosure" to those positions
which seemed to show greater reliability. It may be more
reliable in further research employing the seven—point
differential scale, to use only position A as a measure of
"low therapist disclosure" and positions 2 and/or 6 as

measures of "high therapist disclosure."

REFERENCES

77

 

REFERENCES

Allen, T. W. Effectiveness of counselor trainees as
function of psychological Openness. Journal of
Counseling Psychology, 1967, 1A (1), 35-AO.

 

 

Barrett-Lennard, G. Dimensions of therapist response as
a causal factor in therapeutic change. Psycho—
logical Monographs, 1962, 76 (A3) Whole No. 562.

 

Brecht, J. A. Existentialism and psychiatry: Principles
of existential analysis. Pennsylvania Psychiatric
Quarterly, 196A, A (2), 12—20.

 

Burros, O. K. (ed.). The fifth mental measurements
yearbook. Highland Park, New Jersey: The Gryphon
Press: A. Ellis, 1959.

Campbell, 0. T., and Stanley, J. C. Experimental and
Quasi—Experimental designs for research. Chicago:
Rand McNally, 1966.

Carkhuff, R. R. Counseling research, theory and
practice, 1965. Journal of Counseling Psychology,
1966, 13 (A), A67—A70.

Cooke, J. K. MMPI in actuarial diagnosis of psychological
disturbance among college males. Journal of
Counselipg Psychology, 1967, 1A (5), A7A-A77.

 

 

Dahlstrom, W. G., and Welsh, G. S. An MMPI handbook:
A guide to use in clinical practice and research.
Minneapolis: University of Mennesota, 1960.

Demos, G. D. The application of certain principles of
client—centered therapy to short-term vocational~
educational counseling. Journal of Counseling
Psychology, 196A, 11 (3), 280-281.

Dolliver, R. H. "Expressive" and "Instrumental" as
conceptualization of counseling. Journal of
Counseling Psychology, 1965, 12 (A), AlA-AlS.

 

Edwards, A. L. Experimental design in psychological

Eesgarch. New York: Holt, Rinehart and Winston,

78

79

Greenfield, N. S. Personality patterns of patients before
and after application for psychotherapy. Journal
of Consulting Psychology, 1958, 22, 280.

Halkides, G. An experimental study of four conditions
necessary for therapeutic change. (Unpublished
Doctoral Dissertation), Cited by C. R. Rogers,
Psychological Abstracts, 1966—1967, No. 157.

 

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

 

Kleinmuntz, B. Annotated bibliography of MMPI research
among college populations. Journal of Counseling
Psychology, 1962, 9, 373-396.

 

Maslow, A. H. Motivation andypersonalipy. New York:
Harper, 195A.

Mills, D. H., and Zytowski, D. B. Helping relation-
ships: A structural analysis. Journal of
Counseling Psychology, 1967, 1A (3), 193-196.

 

Osgood, C. E., Suci, G. J., and Tannenbaum, P. H. The
measurement of meaning. Urbana: University of
Illinois, 1957.

Parker, C. A. The predictive use of the MMPI in a college
counseling center. Journal of Counseling
Psychology, 1961, 8, l5A-l58.

 

 

Rogers, C. Client-centered therapy: Its current practice,
implications, and theory. Boston: Houghton
Mifflin, 1951.

 

Rogers, C. The necessary and sufficient conditions of
therapeutic personality change. Journal of
Consulting Psychology, 1957, 21, 95-103.

Rogers, C. On becoming a person. Boston: Houghton
Mifflin, 1961.

Rogers, C. The therapeutic relationship: Recent theory

and research. Australian Journal of Psychology,

Rogers, C. The therapeutic relationship and its impact:
A study of psychotherapy with schizophrenics.
Psychological Abstracts. 1966-1967, 157.

80

Rosen, A. Test-retest stability of MMPI scales for a
psychiatric population. Journal of Consulting
Psychology, 1953, 17, 217-221.

Siegel, S. Nonparametric statistics for the behavioral
sciences. New York: McGraw-Hill, 1956.

Thorne, F. C. The principles of personality counseling.
Journal of Clinical Psychology Press, Brandon,
Vt., 1960.

Truax, C., and Carkhuff, R. Client and therapist trans-
parency in the psychotherapeutic encounter.
Journal of Counseling Psychology, 1965, 12 (l).

 

Truax, C., Wargo, D. G., Frank, J. D., Imber, S. D.,
Battle, C. C., Hoehn—Saric, R., Nash, E. H., and
Stone, A. R. The therapist's contribution to
accurate empathy, non—possessive warmth and
genuineness in psychotherapy. Journal of
Clinical Psychology, 1966, 22 (3), 331-33A.

 

VanderVeen, F. Effects of therapist and the patent on
each other's therapeutic behavior. Journal of
Counseling Psychology, 1965, 29 (1), 19—26.

 

Whitaker, C. A., and Malone, T. P. The roots of
psychotherapy. New York: McGraw-Hill, 1953.

 

 

Wyatt, F. The self—experience of the psychotherapist.
In McGowan, J. F., and Schmidt, L. D. (Eds.),
Consulting: Readings in theory_and practice.

Neg York: Holt, Rinehart and Winston, 1962, 303-
30 .

‘5',“ ,

APPENDICES

APPENDIX A

The Description and Rationale for the Use
of 8 Semantic Differential in This Thesis

82

The Description and Rationale for the Use of
‘ a Semantic Differential in This Thesis

A Semantic Differential was develOped by Osgood and
his associates (1957) as a research tool for the measure—
ment of meaning. It is essentially a controlled asso—
ciation and scaling procedure wherein each concept to be
differentiated is rated on a series of seven—point scales
composed of polar opposite adJectives. S's task for each
scale is only to indicate the direction of his association
and its intensity on the seven-point scale.

Continuing research on Semantic Differential measure-
ment has led Osgood and his co-workers to conclude that
the three primary factors--evaluation, potency and activity—-
account for the great maJority of the factorial structure
Operating in such meaningful Judgments. To test the gener-
ality of this factorial structure, Osgood, Suci, and
Tannenbaum (1957) undertook several studies in which they
varied subJects, concepts, type of Judgmental situation
used in collecting the data, and method of factoring the
data. In each case, the same three factors emerged in
roughly the same order of magnitude.

Reliability measures of a Semantic Differential
generally are quite acceptable. Osgood, Suci, and Tannen-
baum (1957) report test—retest coefficients ranging from

.87 to .93 with a mean r of .91. Again, citing the results

83

 

8A

of their own item reliability check, the authors report a
Rho test-retest correlation of .85. Reported error
measurements of a Semantic Differential on the average
are .67 scale units-~much less than the expected deviation.

Reporting on the general validity of the instrument,
the authors cite as evidence many instances where validity
criteria of specific sorts are available; e.g., evaluative
location of the Thematic Apperception Test (TAT) pictures,
Judged by Ss against a Semantic Differential, were found
to correlate significantly with clinical Judgments of
stories told about the pictures by the same S5. A valid—
ity estimate of the Semantic Differential's factorial
structure is Obtained in the demonstration that repre—
sentation of concepts by a Semantic Differential reveals
essentially comparable structures as yielded from the
method of trials. Finally, other validity measures were
obtained by comparing a Semantic Differential with the
Thurstone and Guttman scales, which yielded correlations
of the order .90 and .78.

Osgood (1957) makes certain assumptions about what is
being measured in a Semantic Differential. Often psychol-
ogists tend to define meaning as something inherently
non-material, more akin to "idea" and "soul" than to
observable stimulus (Osgood, 1957, p. l). Osgood makes
a difference in what he calls semantic meaning, and the

measurement of an emotive reaction (Osgood, 1957, p. 320).

 

 

85

Semantic meaning is viewed as including dimensions of
"attitude" or "value" (Osgood, 1957, p. l).

Osgood (1957, pp. 1-3) specifically refers to the
use of the Differential as assessing some implicit process
or state which must be inferred from observables. It is
a mediational process which occurs in the organism when-
ever a sign is received (decoded) or produced (encoded)
(Osgood, 1957. p. 3). When a subJect decodes a given
sign, it is assumed that a complex mediating reaction
occurs, consisting of a pattern of these alternate bipolar
reactions elicited with varying intensities (Osgood,

1957, p. 319).

The problem of predicting the meaning of word
mixtures is viewed by Osgood (1957, p. 277) as somewhat
analogous to that Of predicting the color of wave lengths.
The direction of a point from the origin (position A on
a seven-point scale) of a semantic space is analogous to
wave lengths of visual stimuli; the distance from the
origin out to a point is analogous to the color metric
purity of visual stimuli. Words of Opposed meaning mixture
will cancel toward a "neutral gray." This involves what
Osgood and associates have called the "congruity prin-
Ciple" (1957, p. 277).

Osgood (1957, p. 15) specifically insists that there
is an intimate relation between perceptual and meaningful

phenomena. He feels that the advantage of the Semantic

 

86

Differential over usual perceptual methods is that avail-
ibility of habit strength of alternate meaning forms is
set within a definite pattern of meaning itself.

The rationale for the use of Semantic Differential
in the E's study_relates to the need to measure perceptions
of therapist meanings as well as clarity or purity of
those meanings; It is assumed (Osgood, 1957, p. 220) that
changes taking place during therapy are reflected by
changes hipatients' meanings of such relevant concepts as
"Counselor," which in turn should be reflected in Judg-
ments on the Semantic Differential.

It has been suggested (Osgood, 1957, p. 239) that
the Semantic Differential can be employed in a large
number of hypothetical situations. The only apparent
restriction is that some aspect of meaning be functioning
as either the dependent or independent variable. There are
several ways in which attitudes and meanings of a concept
can be measured with a Semantic Differential. One measure,
suggested by Osgood (1958, p. 2A1), is simply the location
of the concept against the coordinates of a differential.

One can also estimate the meaningfulness of
concepts to the subJect, those variables falling
near the origin being relatively meaningless
(within the aspects of the meaning sampled) and
those falling far out toward the edges of the

space being relatively "saturated" in meaning
(Osgood, 1957, p. 97).

 

87

The above indicates that the location of a concept
along the continuum between coordinates gives a measure of
"meaningfulness" or purity of conceptual stimulus to the
Judge. The nearer the center position, or origin, the
concept is viewed as relatively meaningless and impure or
"gray" (Osgood, 1957, pp. 26A-265). In this thesis the
concern is only with the deviation or polarization from the
origin, regardless of the direction of the sign; i.e.,
associative-affirmative or dissociative—negative (Osgood,
1957, pp. 122, 195, 201, 278).

It should also be noted that in early assumptions
the loading on the evaluative Judgments (Osgood, 1957,

p. 122). However, Osgood, in later studies (1957, p. 195),
states that a Semantic Differential lp 3239, i.e., eval—
uative, potency, and activity dimensions, may provide a
richer picture of this purity or saliency of meaning. The
S's investigation will sum across all three dimensions to
determine frequencies of different positions as an index

of the purity or clarity of client-perceived therapist
self-disclosure.

We cannot refer to "the" Semantic Differential as if
it were some kind of "test" having a definite set of items
and a specific score. On the contrary it is a very general
way of getting at a certain type of information. As
Osgood (1957, p. 76) states, "a highly generalizable tech-
nique of measurement which must be adapted to the require-

ment of each research problem to which it is applied."

 

88

A variety of studies have examined the differences
in scale-checking styles. Since these latter studies
relate to the design of the present research, they will
be summarized here.

It has been noted (Osgood, 1957, p. 85) that the
choice of a seven-step scale rather than five or nine is
based on the consistency with which all seven scales tend
to be used, and with roughly equal frequencies.

In l9A9 Stagner and Osgood (Osgood, 1957, p. 85)
noted that American Legion members tended to use only
three positions: 1, A, and 7 (all, nothing, or neither).
Grade school children seemed to work better with a five-
step scale, unlike Older subJects who prefer the seven-
step scale. The above indicates a possible relationship
between age, emotionality or maturity, and scale-checking
patterns.

Bopp's 1955 study (Osgood, 1957, p.85)also indicated
that schizophrenic patients used the discriminative posi-
tions (2, 3, 5, and 6) significantly less frequently than
their controls. This supported her notion that there
should be differences in the number and nature, or at
least clarity, of semantic factors.

The latter differences in the use of scale positions
were not correlated with distances (D) between profiles for
S (Stimulus) and R (Response) words. It appears from

Bopp's study that more disturbed patients are far less

89

discriminatory in their use of semantic scales. "The

reason for this difference-~whether due to intellectual
deficits or to emotional factors in the testing situation--
cannot be determined from these data" (Osgood, 1957, p. 227).

Kerrick's (Osgood, 1957, p. 227) study provided some
evidence that I.Q. score is related to "position habits" in
responding to the differential. SubJects of lower intelli-
gence tended to be more polarized in their Judgments. The
brighter subJects tended to use a higher number of the more
discriminatory intermediary positions.

Lyons and Solomon (Osgood, 1957, p. 155) studied the
relationships between extremeness of Judgment and an index
of overt reaction (Judgmental latency) time in scale
checking. Their findings supported the basic assumption
that the extremeness of Judgment on semantic differential
scales is a valid measure of the strength with which signs
are associated with the mediational process.

Studies by Osgood (1957, pp. 228—229) on response
conflict and anxiety as related to scale selection provided
several significant results. The introduction Of the
elements of anxiety reverses the usual scale-checking
patterns of high and low I.Q. subJects. The effect of
making people of lower intelligence anxious, is to make
them cautious, more wary and subsequently, more discrim-
inating in choices. Anxious, intelligent subJects become

less discriminating in Judgments, more prone to either

90

extreme "black and white" decision (positions 1, A, or 7).
Osgood (1957, p. 235) also suggests that the use of the
latter positions by intelligent subJects is like "a quick,
polite smile in an ambiguous and potentially threatening
social situation."

The above review of literature on scale checking
behavior would suggest a number of important elements
bearing on this thesis. Since college subJects are used
in the present study, we would expect them to be more dis-
criminating overall in their use of scales when the concept
Of "Counselor" is highly salient, pure or clear. Also,
college students should be less discriminating in their
Judgments when the concept "Counselor" is less salient,
impure ("gray") and/or ambiguous.

In conclusion, the design of this thesis depends on
the Semantic Differential, scale-checking patterns of the
subJects. The use of the seven positions on each scale
is an index of client—perceived therapist self—disclosure.
The assumption in the present thesis is that what takes
place in therapy is reflected in changes in client meanings
of the concept "Counselor," which in turn are reflected in
the subjects' use of discriminatory (2, 3, 5, and 6)

versus less discriminating scale positions (1, A, and 7).

APPENDIX B

The Semantic Differential, Form A

91

The Semantic Differential, Form A

 

The following Semantic Differential was used in the
present investigation. Only the concept "Counselor" has
been included here, for the purpose of illustration.
There were a total of twenty-one such concepts1 on the
original research instrument used in the Michigan State
University Counseling Center proJect. (See Source of
Data, p. 26 of this thesis.)

Measurement of Meanings--Form A

 

 

 

 

 

(1—6) Student Number
(8-25) Student Name
Last First “Middle
(26—27) Age
(28) Male (1)
Female (2)

 

(check one)
INSTRUCTIONS

The purpose of this study is to measure the meanings of
certain 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 ou. On each page of this booklet you will find a
different concept to be Judged and beneath it a set of
scales. You are to rate the concept to each of these scales
in order.

 

lNote: Other concepts used in the Michigan State
University Counseling Center research were: Woman; My depen—
dency on others for love and help; My father; Man; My feelings
of anger directed at myself; Others depending on me for love
and help; Hate; Controlling myself; Guilt; Sex; Most disliked
me; Love; My failing to accomplish something I set out to do;
My feelings of anger directed at others; Most liked me; My
mother; My mixed-up feelings about my behavior; Me; Author-
ity; and My conscience.

92

93

Here is how you are to use these scales:

If you feel that the concept at the tOp of the page is very
closely related to one end of the scale, you should place
your check-mark as follows:

 

fair X : : : : : : unfair

 

OR
fair : : : : : : X unfair

 

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:

 

strong : : : : : X : weak

 

OR
strong : X : : : : : weak

 

 

If the concept seems only slightly related to tone side as
opposed to the other side (but is not really neutral), then
you should check as follows:

 

active : : X : : : : passive

 

OR
active : : : : X : : passive

 

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're Judging.

If you consider the concept to be neutral on the scale, both
sides of the scale equally associated with the concept, or
if the scale is completely_irrelevant, unrelated to the
concept, then you should place your check-mark in the

middle space.

 

 

safe : : : X : : : dangerous

 

IMPORTANT: (1) Place your check-marks in the middle of
spaces, not on the boundaries:

: X : : : X :
this not this

(2) Be sure you check every scale for every
concept--go not omit any.

 

 

(3) Never put more than one check-mark on a
single scale.

9A

Sometimes you may feel as though you've had the same item
before on the test. This will not be the case, so g2 app
look back and forth through the items. Do not try to
remember how yOu checked similar items earlier in the test.
Make each item g separate and independent Judgment. Work
at a fairly high speed through the test. Do not worry or
puzzle over individual items. It is your first impression,
the immediate "feelings" about the items, that we want. On
the other hand, please do not be careless, because we want

your true impressions.
This study consists of ten concepts. Turn the page and
begin and continue through the conclusion of the concepts.

(The "next page" referred to will follow below, as a
matter of clarity in this present appendix.)

Remember to make your Judgments on the basis of what these
things mean So you.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Counselor
large small
thin thick
colorless colorful
easy difficult
safe dangerous
modest vain
sharp dull
optimistic pessimistic
weak strong
free constrained
fair : : : : : r_fi unfair
active . . . . ° passive
bad : : : : : : good
destructive y productive
slow fast
stable

changing

 

95

Go on immediately to the next page.1

 

1Note: In the Michigan State Counseling Center
research, the same scales as above were used under the 20
other concepts referred to at the beginning of this present

appendix. Those 20 other scales have not, however, been
included in this appendix.

1‘ .

APPENDIX C

The Description and Rationale for the Use
of the Minnesota Multiphasic Personality
Inventory in This Thesis

96

IE...-

The Description and Rationale for the Use
of the Minnesota MultiphasicOPersonality
Inventory in This Thesis

 

 

 

The Minnesota Multiphasic Personality Inventory was
developed to serve as an ObJective device for diagnosing
psychopathology (Dahlstrom and Welsh, 1960). In its most
common form the instrument consists Of at least nine
clinical scales and three validating scales. S is asked
to respond to a series of questions designed to assess
symptomatology indicative of various types of psycho-
pathology. S's responses are readily converted into
numerical scores for each of the clinical scales. In this
way an objective measure of psychopathology is obtained.
Currently, however, common practice is to interpret
profile patterns in combination with scale scores in
arriving at a diagnosis.

Cooke (1967) recently developed an actuarial formula
that was able to duplicate Judges' ratings of college male
MMPI profiles more reliably (r = .91) than the Judges were
able to duplicate their own ratings at a later time (r = .83)
or that these Judges could agree among themselves (r = .80).
Cooke has develOped a similar formula for rating college
female nurses, and this female formula was used in this
thesis to discriminate levels of psychological disturbance,
as well as degrees of constructive client change (see

Appendix D).

97

98

Commenting on the MMPI validity, Ellis (1959) cites
Calvin and McConnel who assayed eighty MMPI studies from
19A0 to 1950 and reported findings of significant discrim-
inations between different kinds of groups in seventy-one
and eighty studies. Ellis himself reviewed one hundred
and sixty MMPI studies between 19A6 and 1951 and found one
hundred and two (6A per cent) of these showed significant
between-group discriminations. Both reviews cited above
suggest that the MMPI's discriminative powers are better
than that of the average personality inventory (Ellis,
1959). Ellis concludes, however, that the question of the
MMPI's absolute validity has not been finally settled.
More recently, Kleinmuntz (1962) reviewed MMPI validity
studies with college populations under three headings--
concurrent, predictive, and construct validity. The bulk
of these studies report results favorable for MMPI
validity.

The table below lists in summary form the results of

several studies in MMPI test-retest reliability with college

 

or psychiatric populations.

Summary of Results of Studies on MMPI Test-

 

Retest Reliabilityywith College or

 

Psychiatric POpulations

 

Study

ParkeryyC. A. (1961)

65 Ss from General and Edu-
cational Psych. classes.
retest interval 1A.7 months.
Long and short form MMPI.

Greenfield, N. S. (1958)

31 college Ss. MMIP's
obtained at college admission
and at time of contact with
college health center. Test-
retest interval 11 months.

 

Dahlstrom,;W. G. and Welsh,
G. S. (1960)

COIIege Ss. Test-retest
interval one week.

Rosen, A. (1953)

A0 male psychiatric hospital
patients. Test-retest inter-
val four days.

 

 

Test—

Results

NO significant differences
between test-retest scale 1
scores except for Pd scale.

No significant differences
between test—retest scores.

Correlations reported range
between .71-.92 for
clinical scales.

Test-retest reliabilities

were between .80 to .88 for
clinical scales in common use,
except fog L, K, Mf, Pa, and
Ma scales which were

between .55 to .75.

 

l

Pd refers to psychopathic deviate.

2Note:

L refers to lie score, K to control score for

bland expressions, Mf to masculinity-feminity, Pa to

paranoia, and Ma to hypomania.

APPENDIX D

Cooke's Actuarial Formula

100

Cooke's Actuarial Formula

Cooke (1967) develOped an actuarial formula that
duplicates Judges' ratings of college male MMPI profiles
more reliably (r = .91) than the Judges were able to
duplicate their own ratings at a later time (r = .83) or
that Judges could agree among themselves (r = .80).

Cooke in an unpublished study (see letter in this
appendix), has developed another formula for rating
college female nurses. This latter formula was used in
this thesis. The entire female formula, with beta weights,
is listed as follows:

Formula for Converting MMPI Raw Scores
into Judges' Formula

The following formula takes into account the manip-
ulations to scales L and F and is the formula for repli-
cating the Judges' ratings from MMPI raw scores:

1. For L, subtract 5 from the raw score of L, then
multiply by 10, then multiply this result by
the beta weight for L (see below).

2. For F, subtract 10 from the raw score of F,
then multiply by 10, then multiply this result
by the beta weight for F (see below).

3. For all other scales, multiply the raw score
for each scale by the beta weight for each
respective scale (see below).

A. Add together the 16 scores which have been
obtained by the above steps. The psycho-

metric formula score equals the sum of these
scores

101

102

Raw Score BetaWeightings for MMPI Scales

L .21 D -.67 Pa 1.06 Si .00
F -.17 Hy 1.37 Pt 1.29 A .AA
K 2.21 Pd 1.50 Sc 3.08 R .51
Hs -.18 Mf 1.1A Ma .00 Es .57

These beta weightings were derived from three MMPI
Judges' ratings of 110 MMPI profiles of freshmen nursing
students at the Medical College of Virginia.

Correlation between the formula and the Judges'
ratings: r = .86. A letter from Cooke, concerning the
actuarial formula for female Ss described above, has been

included at the close of this present appendix.

A

when A
M New '

mm

LOCATION ADDIIII
III Rom 10‘". m
TILIHIONI HI 4-6"!

MAILING ADD!!!-
r. a. set M.
lucuuouo. Vinemls use:

 

DIPAI‘I’MINT 0'
MBNTAL HYGIENE AND HOIPITALI

VIRGINIA TREATMENT CENTER FOR CHILDREN

April 12, 1968

Mr. Walter L. Stump.
1723 Harywood Avenue
Kalamazoo, Michigan

Dear Mr. Stump:

Thank you very much for your interest in my research. I am sorry for the
delay in getting this material to you and hope that it has not held you

up too much.

I am enclosing the information about the beta weightings and the formula
for using the beta weightings for converting HIPI raw scores into the

judges' formula .

The same method was used in deriving this formula as was used in the
previous study (reprint enclosed);1 The only difference is that three
Judges were used instead of five. The three most reliable Judges from

the previous study were used.

This data was derived from a nursing student group rather than a liberal
arts group.

I hope that this information will be helpful. Please let me know if you
have any additional questions.

Again, thank you for your interest.

Sincerely,

Jane K. Cooke, Ph.D.
Director of Psychological Services

JKC
Note: Cooke, J. K. MMPI in ectusriel diagnosis of psychological

disturbance among college neles. Journal of Cbunseligg Psygholggy, 1967

14 (5), 474_s77.

ll”
“
u.
u
I“
5m
R“
H
u
N“
I
III.
Ell”
Allil
"I
H

i

illlllii

3 1293 03146 1514