University of Southern California Dissertations and Theses

The Relationship Of Dependency To Verbal Learning Without Awareness

(USC Thesis Other)

The Relationship Of Dependency To Verbal Learning Without Awareness

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE RELATIONSHIP OF DEPENDENCY TO VERBAL
LEARNING WITHOUT AWARENESS
by
Elaine Greenehaum Caruth
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Psychology)
June 1959
UNIVERSITY O F S O U T H E R N CALIFORNIA
GRADUATE SC H O O L
UNIVERSITY PARK
LO S A N G ELES 7, C ALIFOR NIA
This dissertation, written by
Elaine __Greenebay_m. Caruth ......
under the direction of h®¥....Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by the Graduate
School, in partial fulfillment of requirements
for the degree of
D O C T O R OF P H I L O S O P H Y
Dean
JUNE, 19$9
DISSERTATION COMMITTEE
l-.'L-. ^ ........
Chairman
ACKNOWLEDGMENTS
The author wishes to express her appreciation to
those who shared in the formulation and execution of her
study. She is particularly grateful to her Committee
Chairman, Dr. Alfred Jacobs, for his critical guidance,
sustained interest, and inexhaustible patience during the
unusually long preparation of this research. Dr. Mortimer
Meyer, Faculty Advisor, was of great assistance with his
sympathetic interest and helpful suggestions. Thanks is
extended to Dr. Earl F. Carnes, Faculty Advisor, for his
interest and assistance in this study at a late stage in
its execution.
The writer owes a special debt of gratitude to Dr.
Simon J. Conrad, Chief Psychiatrist, and the staff of the
Los Angeles State Mental Hygiene Clinic for the use of the
clinic patients and for their continued interest and as
sistance during the long and troublesome period of collect
ing the data.
Drs. Trent Bessent and Charles Cabeen generously
made available the facilities of the Psychology Department,
Metropolitan State Hospital. Dr. Leonard Wendland's in
terest and assistance ensured the cooperation of the per
sonnel at Rancho Los Amigos. Dr. Norman Farberow and Dr.
Jacobs made available their evening class students as sub-
H i
jects for this study. The writer Is also grateful to
Dr. James L. Way for his assistance during a difficult
period in the preparation of this report.
Finally, the writer is immeasurably indebted to
her husband for his useful criticism, and unstinted help
throughout all stages of this study.
TABLE OP CONTENTS
CHAPTER
I. THE PROBLEM ..............................
Introduction and Statement of the Problem .
Theoretical Background ...............
Purpose of the Study and Hypotheses . .
Modification of the Original Design . .
II. REVIEW OP THE LITERATURE .................
Human Learning .........................
Summary ...........................
Dependency ...........................
Summary ...........................
III. METHOD ..................................
Subjects ..............................
General criteria for all subjects
used in original design ..........
Selection of additional groups and
modification of the original design
Test Instruments ..................... ,
Dependency scale ...................
i Validation of dependency scale . . . .
Normalization of dependency scale . ,
The stimulus cards ...................
Summary of Procedure for Each Subject
PAGE
1
1
2
5
5
8
8
28
30
43
45
45
45
48
50
50
50
51
52
53
CHAPTER PAGE
Summary of Procedure for Each Group . . . 55
Experimental Design 56
IV. RESULTS 58
V. DISCUSSION Jl
VI. SUMMARY AND CONCLUSIONS 83
The Problem ..................... 83
The Method 84
Results 86
Conclusions 87
REFERENCES . .................................. 89
APPENDIX A. Statistical Data 95
APPENDIX B. Test Instruments 102
LIST OF TABLES
TABLE
1.
2.
3.
4.
5.
r
PAGE
Mean Number of First Person Pronouns Used
During Each of the Four Blocks of 20
Trials 59
Standard Deviations of Scores for Each Group
of Subjects for Each of Four Blocks of
20 Trials Each ......................... Si
Source Table for Analysis of Variance of
Repeated Measurements of the Mean Number
of First Person Pronouns Used by the Five
Groups of Subjects on Four Blocks of
Successive Trials ..................... 62
"t" Ratios of Over-all Differences between
Groups on Mean Number of First Person
Pronouns Used for the Total of 80 Trials . 63
"t" Ratios of the Mean Number of First
Person Pronouns Used between Each Block
of 20 Trials by Total Sample of Subjects . 64
Source Table for Analysis of Variance of
Repeated Measurements of the Mean Number
of First Person Pronouns Used by the Com
bined Experimental Group as Compared with
the Control Group on Four Blocks of
Successive Trials ..................... 65
TABLE
7. Mean Number of First Person Pronouns Used
During Each of the Four Blocks of 20
Trials as Reported in the Studies of
Taffel and Cohen ...................
8. Dependent Subjects: Mean Number of First
Person Pronouns Used During Four Blocks
of Trials ...........................
9. Independent Subjects: Mean Number of First
Person Pronouns Used During Four Blocks
of Trials ...........................
10. Control Subjects: Mean Number of First
Person Pronouns Used During Four Blocks
of Trials
11. Normal (I) Subjects: Mean Number of First
Person Pronouns Used During Four Blocks
of Trials ...........................
' 12. Normal (II) Subjects: Mean Number of First
Person Pronouns Used During Four Blocks
of Trials ...........................
: 13. Two Subjects Eliminated from Study: Mean
Number of First Person Pronouns Used
During Four Blocks of Trials
vii |
PAGE i
I
72
96
97
98
99
100
101
LIST OF FIGURES
FIGURE PAGE
1. Graph of the Mean Number of First Person
Pronouns Used During Each of the Four
Blocks of 20 Trials 60
CHAPTER I
THE PROBLEM |
Introduction and Statement of the Problem
Verbal behavior is one of the fundamental concerns
of clinical psychology. Individuals are studied, examined,
evaluated, diagnosed, treated, and judged through such
verbal data as test protocols, anamneses, interviews, free
associations, and recordings of therapy sessions. Psy
chology seeks methods to control and modify such behavior, i
It is therefore important to understand the variables in
volved in the elicitation and modification of the verbal
response.
In recent years, experiments have demonstrated that
|the verbal responses of human subjects may be markedly
modified despite the absence of any conscious awareness of
either verbal or non-verbal reinforcing stimuli. The re
inforcing stimuli may consist of such minimal cues on the
part of the experimenter as a murmured "mmm-hmm" or as
|
Pimple note-taking. The Implications of such findings are
! of particular interest to the clinician who uses the
clinical Interview as a means of helping the individual to
ftiodify his behavior patterns. What are the specific ways
in which the clinician's behavior may influence the behavior
2 |
of his patient? Do people differ in the extent to which j
they learn without verbal awareness? Are there personality I
correlates which help determine the degree to which such j
learning without awareness takes place? The present study !
will investigate whether the specific dimension of "de
pendency" is functionally related to this type of learning.
Theoretical Background
The dimension of "dependence-independence" has long
been considered an important aspect of behavior. The
psychoanalytic conception of dependency (11) postulates
that all human attitudes of dependence are based on the
fact that the human infant is born more helpless than other
infants and requires feeding and care by adults. Every
human being has a dim recollection that there were once
powerful, or as it may have seemed to him, omnipotent
beings whose help, comfort and protection he could depend
;on in time of need. In later life, man is said to en
counter frequently situations in which he is again as help
less as he was as a child, and in which he may long for
the protection that was his as a child. Many social situa
tions, implicitly or explicitly, promise help or protection
in exchange for obedience or conformity, much as was the
case in some infant-adult relationships.
Dynamic theory postulates that the character of
subsequent authority relationships is determined by the
quality of the earlier dependent relationships as well as
by the extent to which they resemble, or are perceived as
resembling, the earlier situation. For example, an indi
vidual whose early years were fraught with fear, mistrust,
and suspicion of a parent might be expected to be unable
to establish friendly, comfortable relationships with real
authority figures such as employers, because of having
generalized his fear and mistrust of the parent to all
subsequent authorities. As Fenichel points out, "It is
true that there are great differences between an almighty
God, or a modern employer, and a mother who feeds her
baby,* but nevertheless it is the similarity between them
that explains the psychological effectiveness of authority"
(11, p. 492). An individual who has learned as a child
that the least painful way of relating to his parents was
'through conformity, obedience, and reliance on them for
all decisions, might in later life be still relying on such
modes of behavior. Such an individual is described as
[overly dependent. He is seldom able to oppose another in
dividual. He tends to defer to the advice and opinion of
others indiscriminately. He seeks almost permanent guidance
and support from others.
If this need for guidance and support is a pervasive
and widespread characteristic of the overly dependent
person (hereafter to be referred to as the dependent per
son), it can be expected that manifestations of the de
pendency will be evident in many aspects of his behavior.
It may be predicted that he will show a generalized over
all dependence upon external cues to guide his behavior in
contrast to the independent individual's relatively greater
reliance upon such inner guides as kinesthetic cues or
previously established attitudes and preferences. Gordon
(13) has actually demonstrated that dependent individuals,
so defined by self-ratings and ratings of their physicians,
relied more heavily on external visual cues in making per
ceptual judgments than independent individuals who relied
more on inner or kinesthetic cues. It has also been
demonstrated (20) that individuals who produce more TAT
themes in which the hero can be classified as dependent,
yield more to the opinion of the incorrect majority in
^making visual discriminations than subjects not depicting
that type of hero.
Thus, be it the visual field, the opinion of the
: group, or the behavior of the experimenter, it is to be
expected that the dependent individual will rely more
iheavily on something or someone outside himself to help
guide his behavior than the independent individual.
Therefore, one should be able to predict that the dependent
individual will modify his verbal behavior under conditions
of verbal reinforcement to a greater extent than the inde
pendent Individual.
Purpose of the Study and Hypotheses
The purpose of the present study was to investigate
the general hypothesis that the susceptibility of indi
vidual patterns of verbal behavior to verbal reinforcement
is related to the specific personality dimension of
dependency-independency.
Hypothesis A: The verbal behavior of dependent
subjects will be more markedly modified by
verbal reinforcement than will the verbal
behavior of independent subjects.
Prediction 1: The mean number of positively
reinforced responses will increase signifi
cantly more in dependent subjects than in
independent subjects.
Modification of the Original Design
A basic presupposition to the present study was that
verbal behavior could be modified by verbal reinforcement.
Specifically., it was assumed that a class of reinforced
verbal responses would increase in frequency as compared
to classes of unreinforced responses. It was not deemed
necessary to test this supposition because of previous
literature reporting such findings. Consequently, the
data were not evaluated until sufficient material had been
collected to test the experimental hypotheses. At this
point it became evident that the above assumption was un
warranted. The experimenter had not succeeded in modifying
the verbal behavior of either of the experimental groups
when compared with a control group, all groups having
been drawn from a neuropsychiatric outpatient clinic
population. At this point in the study it became necessary,
therefore, to investigate the apparently deviant behavior
|of the samples used, and certain methodological hypotheses
were advanced. Additional subjects from a normal popula
tion were sought, and Hypothesis B was added to determine
if the use of a neuropsychiatric population had biased
the results. Recognizing the possible importance of even
minor differences in experimental procedure, Hypothesis C
was tested on an additional sample of normal subjects using
an experimental procedure— namely, subject and experimenter
sitting face to face and using the word "good” as the re
inforcing stimuli— which more closely duplicated one re
ported in the literature. In the actual statistical
analysis of the results all three hypotheses were tested
Simultaneously.
Hypothesis B: The verbal behavior of normal subjects
will be more markedly modified by verbal re-
inforcement than the verbal behavior of neuro-
psychiatric outpatient subjects.
Prediction 1: The mean number of positively re
inforced responses will increase significantly
more in a group of normal subjects than in
either dependent, independent, or a control
group of neuropsychiatric outpatient subjects.
Hypothesis C: The verbal behavior of normal sub
jects will be more markedly modified by verbal
reinforcement consisting of the response "good"
in a face to face situation than when reinforce
ment consists of the word "mmm-hmm" in a face
to back situation with both normal and neuro
psychiatric subjects.
Prediction 1.: The mean number of positively re
inforced responses will increase significantly
more in the final group of subjects than in
any of the other groups of subjects, experi
mental or control.
CHAPTER II
REVIEW OF THE LITERATURE
Human Learning
Investigations of human learning have traditionally
separated studies on conditioning from investigations of
more complex learning problems. Experiments on the condi
tioning of human beings have been primarily limited to
such respondent behavior as is described in the classical
conditioned reflex studies of the eyelid response, the
galvanic skin response, investigations of prenatal con
ditioning, etc. Until as recently as 19^9 (12), the
"majority of experiments In operant conditioning have been
performed with infra-human beings" (12, p. 587).
Investigations of the more complex learning problems
could roughly be divided into two types (51): 1) the
tasks presented the subjects were essentially a "lesson"
such as a series of syllables to memorize or a motor
skill to perfect; and 2) the task presented the subject
was a problem to solve. Studies of the first type com
prised the main body of literature on human learning.
;Since the introduction of the nonsense syllable by Ebbing-
:haus, practically all aspects of rote learning, memory,
recall, recognition, retention, forgetting, transfer of
training, etc., have been investigated. Studies of the
second type have included experiments on reasoning, in
sight, concept formation, etc.
In recent years, however, coincident with the
elaboration of Skinner's system of psychology, there has
been a growing interest in the extension of studies of
operant conditioning to the area of human learning. By
operant conditioning is meant the conditioning of a class
of responses which are emitted by the organism and which
operate upon the environment to generate consequences.
Skinner (42) has pointed out that the psychologist is
primarily concerned with the fut,ure behavior of an
organism, or what he has termed the "probability of action,
■That is, he is interested in predicting and bringing under
control the behavior of the individual and hence is only
interested in those response classes of behavior that are
likely to be repeated. Skinner deplores the traditional
reification of such "probabilities of action" in the
organism as are implied by the concepts of "tendencies,"
‘ "habits," "attitudes," "set," and "instincts." He regards
all such constructs as an attempt to embody within the
organism as neurological or psychological entities what
should more properly be referred to as a predicted fre
quency of occurrence. He feels that the frequency of oc
currence of a response class is, or should be, the basic
10 I
datum of scientific psychology. He points out many ad- I
i
vantages of this type of index of behavior. Traditional !
criteria of learning are concerned with qualitative changes ;
in the character of the subject's behavior, i.e., the
organism learns how to do something. Thus, once the new
task is learned, observations of the learning process
cease.
The study of operant behavior, on the other hand,
permits the observation of what happens to the behavior
(i.e., the frequency of occurrence) after the criterion of
learning has been reached but conditions of reinforcement
are still operating. Furthermore, by means of operant
conditioning techniques, it is possible to simulate the
real-life experience of varied schedules of reinforcement
and to study how minute variations in the schedules of
reinforcement may affect the subject's behavior. Other
^advantages of the operant conditioning method lie in the
orderliness of the data, and the facility with which it
may be reproduced, since the experimental variables may be
;rigidly controlled and such factors as examiner influence
eliminated. Furthermore, Skinner states that it is not
inecessary to resort to a large group of subjects since
|this "method permits a direct view of behavioral processes
Iwhich have hitherto only been inferred"(42, p. 77)- Conse
quently, we are dealing with laws concerning the individual
11
I
directly rather than with an hypothetical average indi
vidual. Also this method permits a continuous study of
many basic processes over a long time, which then provides
a substantial basis for the concept of probability of
action.
The procedure in many of the investigations has
consisted essentially of an adaptation of the original
"Skinner box" (40) to human beings, although there have
been other variations in design. Puller (12), for example,
investigated operant conditioning in a "vegetative idiot,"
using reinforcements of syringe feedings of warm milk
whenever a selected arm movement occurred. Most of the
research, however, has utilized variations of the vending
machine type of apparatus. Because such techniques re
quire little or no verbal response from, or communication
with, the subject, they have been found particularly ap
plicable to the investigation of severely disturbed
hospital patients.
Lindsley and Skinner (27) specifically investigated
the feasibility of this method for analyzing the behavior
of psychotic patients who had been hospitalized an average
of 17 years. They concluded that stable and predictable
behavior was generated by the operant conditioning tech
niques. In addition, they found a marked similarity between
the performance of these patients and of rats, pigeons,
and frogs when exposed to two different reinforcement
schedules. They speculated that the behavior of psychotic
subjects was controlled to some extent by the reinforcing
properties of the immediate situation and suggested that
the effects of different reinforcement schedules upon
their behavior should be investigated. In 1956, Lindsley
(26) reported on the results of a long term research pro
gram in chronic schizophrenia using the methods of operant
conditioning. He discussed refinements of technique which
permitted the investigation of "altruistic" behavior,
and could potentially be used to Investigate "imitation,"
"cooperation," "competition," etc.
Recently there have been an increasing number of
studies of the operant conditioning of verbal behavior
which for the most part have been designed within the
framework of learning theory (Skinner's). The dependent
variables are the subject's verbal behavior and the Inde
pendent variables are generalized conditioned reinforcers,
intended to bring the subject's verbal behavior under the
’control of the examiner without the subject's awareness.
Many of these studies have been particularly oriented
towards exploring interpersonal situations and the effect
of such generalized conditioned reinforcers as a nod, a
ismile, or a verbalized word of approval. Some investigators
:have also been concerned with the effectiveness of such
13
non-social cues as a buzzer, tone, light, etc., but such
studies, unless they also included techniques of verbal
reinforcement, lie somewhat beyond the scope of the present
investigation.
In reviewing the literature on the modification of
verbal behavior in situations where the subjects are un
aware of any specific reinforcement, the studies are most
frequently titled "learning without awareness" and "verbal
conditioning without awareness." However, such studies
include two basically different types of experimental
variables. First are those studies, including the present
one, which concern themselves with modifying verbal re
sponse patterns of subjects who do not regard the experi
mental task as a learning problem. In the second instance
are those studies where the subjects know they are faced
with a problem-solving task in which they have to find a
"correct" solution. When these subjects show "learning
without awareness" the authors mean that performance has
shown objective evidence of learning even though the sub
jects were unaware of the correct solution to the problem.
■These subjects are consciously motivated to seek the
"’ correct" solution which the experimenter had predeter
mined as right.
: Phenomenologically, the situation is entirely dif
ferent for the subject in the first type of experiment
and the subject in the latter experiment. In the former,
the subject is merely given a task to perform. In most
instances he regards this task alone as fulfilling the de
mands of the experimenter, and in such studies, which shall
hereafter be referred to as studies of verbal learning
without awareness, the subject is unaware that he is
modifying his behavior or that there is any expectation
upon the part of the experimenter that he will do so. In
the other type of design, the subject is actively seeking
to modify his behavior and is only unaware of the correct
way to do so.
Chronologically the studies of verbal learning with
out awareness start with Greenspoon's (14) investigation
Into the effect of verbal and non-verbal reinforcing
stimuli upon the frequency of occurrence of a pre
determined class of verbal responses. Greenspoon's sub
jects, drawn from psychology students at a university,
were instructed to voice nouns ad libitum throughout a
50-minute interval. He reinforced two response classes,
inamely plural and non-plural nouns with the following
methods of reinforcement: visual (light); auditory (tone);
Ipositive verbal (mmm-hmm); and negative verbal (huh-uh).
:He eliminated from the results the records of any students
;who were able to verbalize the relationship between the
iselected response class and the subsequent reinforcement.
Greenspoon found that positive verbal reinforcement re
sulted in an increase in the frequency of occurrence of
plural nouns. Visual and auditory reinforcement also
functioned to increase the frequency of occurrence of the
plural nouns. Negative reinforcement resulted in a de
crease in the frequency of occurrence of the plural nouns.
The light, tone, and positive verbal reinforcement also
served to increase the frequency of occurrence of the non
plural response class. Negative verbal reinforcement,
however, served to increase the frequency of the non
plural nouns, in contrast to the results with plural
nouns. Because of the differential effect of negative
reinforcement upon the two different response classes,
Greenspoon concluded that the nature of the response class
: was a factor in determining whether or not a particular
stimulus would be reinforcing.
Mandler and Kaplan (30) used Greenspoon's technique
and found little over-all effect of positive verbal re
inforcement upon the production of plural nouns. However,
Handler and Kaplan discovered from interviews that the
subjects formulated secondary hypotheses about the rela
tionship between their associations and the experimenter's
behavior. These hypotheses tended to fall into two dis
tinct groups; some subjects viewed the reinforcement as
"positive" encouragement, and some felt it was "negative"
16
and that they were going too fast or giving the wrong kind
of response. When the results of these two groups were
separated, the ''positive?' records showed a significant in
crease in the frequency of occurrence of the reinforced
response, while the "negative" records showed a decrease
during the same period of time.
Sidowski (39) specifically investigated the influ
ence of "awareness of reinforcement" upon verbal condition
ing. Using a blinking light as reinforcement for the
response class of plural nouns, he found no significant
difference in conditionability between those who were
aware there was a task but did not know the solution and
those "who were not aware they should be aware of some
thing." Although not significant, there was some slightly
: greater increase in the frequency of occurrence of the
reinforced response in his "Task" subjects as compared
to the uninformed "Non-Task" subjects. Subjects who were
able to verbalize correctly the relationship were
eliminated from the study. Krasner (24) mentions a study
;in which Tatz, on the other hand, found a greater rate of
conditionability in those subjects who were informed that
;there was a correct solution to be discovered. Further
more, those subjects who had been informed that there was
a task to be solved could be conditioned by either verbal
reinforcement or by the click of an electric counter.
However, although uninformed subjects could be conditioned
to some extent by verbal reinforcement they could not be
conditioned by the click of the counter.
Taffel (46) developed a variation of Greenspoon's
technique which has subsequently been replicated in several
studies, including the present one. Taffel presented his
subjects with eighty 3x5 unlined white index cards. In
the center of each card was typed a different verb in the
past tense. Below the verb were typed the pronouns
"I," "We," "He," "You," "She," and "They," with their
order randomized for each card. The subjects, who were
'drawn from a neuropsychiatric hospital patient population,
were instructed to make up sentences using the verb in
the center of the card and starting with any one of the
pronouns typed underneath. Whenever a subject started a
sentence with a first person pronoun, the response was
reinforced by the experimenter's response of "good."
Three conditions of reinforcement were investigated:
positive verbal; visual; and no-reinforcement. The dif-
jferential effect of three levels of anxiety obtained on
the Taylor Manifest Anxiety Scale upon the condltlonablllty
:of the subjects for the different methods of reinforcement
|was also studied by means of analyses of variance. Taffel
advanced the following hypotheses:
"A. Verbal responses can be operantly conditioned
B. Amount of conditioning is a function of:
1. Type of reinforcement stimuli involved
a. Verbal reinforcement will be relatively
more effective in a verbal situation.
b. Non-verbal reinforcement will be rela
tively less effective in a verbal
situation.
2. Anxiety level as measured by scores on the
Taylor Manifest Anxiety Scale" (46, p. 4).
.The subjects* responses were divided into four Blocks of
i
120 trials each. Reinforcement commenced with Block II In
!
!
|order that some initial level of use of the reinforced
i
;pronoun could be ascertained. An analysis of variance was
I performed In order to evaluate progressive response to
successive reinforcement for differing conditions of re
inforcement and differing groups of patients.
Taffel found significant reinforcement taking place
when subjects were exposed to positive verbal reinforcement
but not when the reinforcement consisted of a light.
Within his experimental groups those patients with the
highest anxiety scores showed the greatest significant in
crease in their use of the reinforced pronouns; the medium
:anxlety patients showed a smaller but still significant
(at the 5 per cent level) Increase, while the low anxiety
subjects failed to show any significant increase in the
I
19 I
I
reinforced response as compared to when the response was ;
|
not reinforced. j
Krasner (24) discusses Sarason's investigation of
the relationship of anxiety, defensiveness and compliance I
to verbal conditioning with a slightly modified experi- j
mental procedure. He presented his subjects with cards j
with the word "he" typed on top and below three verbs of
equal frequency of occurrence, including one of the response
class of "verbal activity" which he had selected to re
inforce. He found that patients rated by their therapists
as being compliant were significantly more responsive to i
reinforcement; anxiety and lack of defensiveness were also ;
associated with amenability to conditioning.
Other investigators, using Taffel's technique, have
reported both positive and negative results with different
populations and different experimental designs. Cohen
;et al. (7) sought to establish "some basic relations be
tween verbal reinforcement operations and verbal responses
as elicited in a modified version of Taffel's procedure,"
using general hospital patients. Using a slightly different
frame of reference and terminology, he spoke of investigat
ing the effect of verbal reinforcement on the acquisition
iof a preference pattern. In a second experiment he studied
I two different extinction operations following the acquisi
tion of this "preference pattern." Results of the first
20
study indicate "a progressive increase in the frequency
of the reinforced response for the experimental group and
no increase for the control group. None of his subjects
was able to verbalize the contingency between certain of
their responses and the response of the experimenter. In
his second experiment the two conditions consisted of
"no-reinforcement1 1 and "counter-conditioning." Extinction
occurred only under conditions of counter-conditioning.
In a similar investigation, Klein (22) conditioned
three groups of subjects by means of positive verbal re
inforcement. He was then successful in extinguishing the
iresponse for the following three conditions of extinction,
which are given in order of their diminishing effective
ness: 1) negative reinforcement ("no good"); 2) counter-
! conditioning to an alternative response class; 3) no re
inforcement. Out of his 100 subjects only one could
|verbalize the relationship between the reinforced response
and the reinforcement.
Daily (9), working ivith undergraduate introductory
ipsychology subjects, found somewhat different results.
Using both "mmm-hmm" and "good" as positive verbal re
inforcers, he found the same increment in the mean fre
quency of the reinforced response in his control group as
in his experimental groups. Furthermore, contrary to
Taffel*s results, he failed to find a significant
21 ;
difference between anxious and non-anxlous subjects in I
their response to verbal reinforcement. Daily attributed
the difference in results between his study and those of
Taffel and Cohen e_t al. to variations in procedure, popu- i
i
lation, and examiner status. i
Marion (31), working with college students, also !
failed to demonstrate the operation of verbal reinforcement
In a study that was concerned with the differential effect
of examiner status. He concluded that "within our experi
mental procedure, male college counsellees do not Increase
j
jtheir relative use of the criterion response class ("I"
lor "we" pronouns) unless their experimenters have relatively
high status."
Hartman (17)* working with schizophrenics and normals
(I.e.j psychiatric aides), found that the initial level of
response contributed markedly to the data, a point which
Daily had speculated about as a possible source of varia
bility. Hartman was originally Interested in differentiat
ing between the effects of different types of reinforcement
such as "good," a nod, and a shake of the head at a non- i
criterion response. However, he found such an inter
dependency of trials, type of reinforcement, diagnostic
category, and initial level of response that he felt no
generalized conclusion on types of reinforcement could be
made on the basis of his data.
22
Grossberg (15)^ using similar experimental tech
niques, investigated the effect of the reinforcement
schedule and the response class on verbal conditioning in
undergraduates. He chose the first and third person pro
nouns for his two response classes and the following
schedules of reinforcement: 0 per cent, 60 per cent,
75 per cent, and 100 per cent. He found no difference in
the effectiveness of acquisition, nor in the resistance to
extinction between any of the variables.
Kanfer (21), however, using as his response class
an estimate of distance moved in the autokinetic stimulus
situation, also investigated the effect of partial re
inforcement upon acquisition and extinction of verbal
responses, using a 100 per cent, 67 per cent, 50 per cent,
and 0 per cent reinforcement schedule with the following
results:
a. Acquisition with a partial reinforcement pro
cedure requires more trials but fewer re
inforcements than acquisition with a continu-
1 ous procedure.
b. Extinction with a continuous procedure shows
i
greater initial resistance but faster decline
than in partial reinforcement.
Spivak and PapaJohn (45) conducted a similar in
vestigation on schedules of reinforcement of a verbal
response In an autokinetlc situation. Their subjects in
the regular reinforcement group showed extinction within
10 to 35 trials, while no subject in the variable interval
schedule showed extinction within 10 to 80 trials.
Other Investigators have utilized still different
procedures in their investigations of the reinforcement
of verbal behavior. Ball (2) instructed his subjects to
tell imaginative stories that would Include one of the
three categories, "man, woman, or animal." He had the
subjects participate in a total of five experimental ses
sions of 10 minutes each. He investigated the following
conditions of .reinforcement for his five groups of
patients:
C: Control, no reinforcement
P: Positive verbal reinforcement consisting of
"mmm-hmm"
N: Negative verbal reinforcement consisting of
"uh-uh"
S: Buzzer as reinforcement
L: Light bulb as reinforcement.
;it is interesting to note that Ball sat behind his subjects
and was not visible to them. Subsequent investigators who
allowed themselves to be visible to their subjects later
realized that they were permitting the presence of other
than the selected reinforcement, such as body movements,
24
facial expressions, etc. In several studies on verbal
reinforcement it has been found that a head nod or smile
may function as positive reinforcement. Although Talbot
(47) did not find that notetaking as a reinforcer actually
functioned to increase the class of reinforced responses,
he did find that It significantly retarded extinction of
a plural response in his experimental group as compared
to his control group.
Ball found that those subjects who received positive
verbal reinforcement significantly increased their mean
number of stories containing the reinforced response cate
gory of "animal." The group that received negative re
inforcement responded bimodally, with some of the subjects
definitely increasing and some of the subjects showing a
decrement in the occurrence of the reinforced response.
The negatively reinforced group of subjects all developed
|an awareness of the procedure, whereas none of the positive
ly reinforced subjects did.
In most studies on verbal reinforcement without
awareness, those subjects were excluded from the study who,
^following the experimental procedure, showed that they had
developed an awareness of the fact that there was a re
inforced response class. Generally, the numbers excluded
Tor this reason were very small. However, this dimension
of "awareness" is a very nebulous one to define and
25
delimit. Even though subjects may have developed no In
sight into the specific experimental procedure, they may
have ended up with an awareness that something more had
been expected of them than simply fulfilling the original
task requested by the experimenter and they may have de
veloped hypotheses concerning the task which, if incorrect,
would not have been labeled by the experimenter as an in
dication of "awareness." Krasner (23), using a procedure
essentially similar to Ball's, describes one of his two
subjects who showed sudden verbalized insight half-way
through the experiment into the relationship between a
specific response on his part and a nod of the head on
the examiner's part. However, in the interview following
completion of the session, he denied noticing anything
about the examiner's behavior or noticing any change in
his stories. Furthermore, following-his verbalized in
sight, he completely discontinued all reference to the
reinforced response. Since this was a hospitalized
I schizophrenic patient his behavior might be understood
I in terms of a possible negativism, etc., but many of the
studies on verbal reinforcement are based upon work with
Just such samples of patients. In this experiment, Krasner
had two subjects tell stories with at least four characters
in each— a mother, a father, a child, and an animal. All
nouns and pronouns referring to the mother figure were
2 6
reinforced by a head nod, a smile, and a murmured ' ’ mmm-hrmn. "
The subjects were tested for five successive experimental
sessions each, alternating non-reinforcement and reinforce
ment conditions. Results showed that the reinforced class
of verbal behavior alternately increased and decreased cor
responding to the conditions of reinforcement or no re
inforcement.
An interesting variation in technique was introduced
by Verplanck (48), who used as subjects men and women whose
relationships to the experimenter were as "friend," "room
mate, " "uncle," "total stranger," and "date." He re
inforced all statements of opinion beginning with "I
think," "I believe," "It seems to me," "I feel," etc., by
agreeing with or paraphrasing their response. Verplanck
used 17 advanced psychology students who engaged their
unsuspecting subjects in an ordinary conversation over a
!half-hour period. This type of procedure probably insures
ithe maximum amount of "unawareness," since the subjects
were not even aware they were participating in anything.
I This procedure is so effective in this respect that
;Verplanck reports that he himself was used as a subject
.entirely unbeknownst to himself, in a conversation with
ia friend to whom he was describing the experiment. The
;results of the experiment showed that 21 out of 24 sub
jects decreased with respect to the frequency of all the
27 |
unreinforced response classes, while all 24 subjects in-
!
creased their rate of verbalizing the reinforced state
ments of opinion.
A number of the studies have commented in their
discussion about the possible effect of experimenter dif
ferences. Cohen., for example,, found no differences be
tween examiners. Marion, on the other hand, speculates
about the possibility of the differential effect of
clinical experience of the experimenters upon the results
in such experiments.■ Binder, McConnel, and Sjoholm (6)
specifically investigated "examiner influence" through
using two different experimenters of diametrically opposed :
characteristics. Their stimuli consisted of two sets of
iverbs, "mildly hostile" and "neutral," and the "mildly
hostile" verbs were selected as the response class. The
two experimenters were "an attractive, soft-spoken,
ireserved young lady who was 5 1/2' in height, and 90 lbs.
in weight," and a male who was "very masculine, 6,5" tall,
220 pounds in weight and had many of the unrestrained per
sonality characteristics which might be expected of a 1
former Marine Captain." Having predicted that there
iwould be a greater reinforcement effect from the male
jexaminer, they nevertheless explained their results, which
Iwere in the exact opposite direction from the predicted
■one, as due to "examiner influence" in that the female may
28 ;
have provided a less threatening environment, hence con
ducive to the flow of hostility which might have been
withheld under the more threatening environment of the ;
"Marine Captain." j
There have been other studies in this area which go
too far afield to be logically Included under a review of
strictly operant conditioning. Attempts have been made to
reinforce specific attitudes about capital punishment, edu
cation, or preferences for pictures of human beings, etc.
(24,18,8). Such studies, however, tend to gloss over such
crucial points as the original pre-experimental strength
■of the reinforced response, do not carefully compare all
possible response classes, and do not demonstrate any
evidence that they can predict from test-taking attitudes
to behavior in real-life situations. The role of the sub
ject's bias and his degree of affective involvement in the
i
^particular experimental problem or technique make such
studies highly individualistic and of little general appli
cation.
j Summary
Studies on learning without awareness have frequently
been conducted within the framework of Skinner's theoretical
system. The dependent variables are the operant behavior,
i.e., the subject's verbal responses, and the independent
variables are the generalized conditioned reinforcers such j
as "good" and "mmm-hmm." Many aspects of this process |
have been investigated with differing results. Some in- j
vestigators have been primarily concerned with the differ- I
ential response of various personalities to this procedure^ |
as was the present writer. Evidence has been presented
that more compliant people are quicker to modify their
patterns of response and that anxiety and lack of defensive
ness are also associated with amenability to this condi- '
tioning procedure. Such findings suggest that personality j
characteristics, particularly those which are mobilized in
inter-personal interaction, may play an Important role in I
determining the type and extent of the individual's
response to such procedures.
The majority of the studies report positive results
with the use of verbal reinforcers. However, there has
been a sufficient number of studies with negative results
to form a basis for advancing certain hypotheses concern
ing the reasons for the failure to obtain learning without
I ;
I awareness. The role the experimenter plays in relation to ;
the subjects is a highly important one; and may directly
affect the frequency of occurrence of the reinforced re
sponses. Furthermore, it has been suggested that the
!nature of the selected response class is extremely important
:as well as the frequency of occurrence of the selected
response during preconditioning trials.
30
Dependency
Discussions of dependency and the dependence-
independence conflict may be found in several readily
available references on clinical psychology and psychiatry
(1,11,32,36). However, there is a dearth of specific
formulations of the exact nature of the concepts used and
|of research and experimentation, except in the area of
ipsychosomatic disorders.
Various dynamic formulations about dependency tend
I to agree on the following general principles. The origins
of dependency are rooted in the initial helplessness and
dependency of the human infant upon the adult, anphasis
has been placed on both the biological and psychological
aspects of this dependence. Analytic writing stresses
ithat the biological helplessness, even in infancy, is less
I a consideration than the fact that dependency is related
to a need for love. Levy (25) emphasizes the role of
imaternal overprotection in creating excessive childhood
i
jdependence. Saul (36) points out the importance of the
jcomplete physiological reliance of the infant upon the
jmother for satisfaction of its every basic need in order
I to ensure survival. At the same time, however, an infant
does not have to conform to the limitations imposed upon
31
the older person and in this sense may be regarded as
''highly independent of his environment" (13, p. 3)- "As
a child grows older, he must give up progressively some of
his dependence and sacrifice part of his independence"
(13j p. 4). The process of maturation may be regarded as
a growing resolution of infantile needs into adult ego
needs (44). Giving up of childhood dependence also in
cludes giving up the pattern of imitating the parents or
parental surrogates as a means of mastering reality. How
ever, mature adult behavior must also include the ability
;to accept one’s own dependency needs as a part of oneself
|and to appreciate one's interdependence with those around
him. This interdependence upon others is intensified as
i
a culture grows more specialized; yet, at the same time,
’ the culture itself emphasizes the prestige attached to
independent achievement. The conflict created by the in
lability to accept one:'s own dependency needs and the long
ings created by these needs has been postulated to lead
to such widely differing symptoms as "neurotic criminality
and some types of psychosomatic disease" (1, p. 258).
Excessively strong dependency needs may be manifested
behaviorally either by an over-acceptance and acting out
iof these needs or by an extreme denial of them in cases
where there is a strong conflict over accepting them.
Although there is some disagreement as to the exact nature
of "dependency,, " whether it is a primary or secondary j
j
drive, or a character trait, and while there is some dis- !
i
■ !
agreement as to whether or not there is a bi-polar dimen- i
i
sion of dependency-independency, or two different dimen- j
isions, there exists general agreement that dependent
'people may be distinguished from independent people.
Concretely, what is meant by dependency? Maslov/ and:
:Mittelmann (32, p. 392) state that dependent tendencies
include "the need for emotional or authoritative support
in most situations, difficult in making independent deci- |
sions, and taking on responsibilities, and the dread of
lonliness." Implied in this statement is the fact that
:the dependent person perceives others as more competent
than himself, and is likely to seek and accept the guidance
land advice of others when he has to make a decision (20).
i
3ernardin and Jessor (5) state that they find considerable
iagreement about what is meant by dependency (a belief
; which seems so common that many authors dispense with
the necessity for defining the term) and define it as
i Including: "a reliance on others for approval for help
j
ior assistance, and conformity to the opinions and demands
of others." Warn (49) summarizes his discussion on de-
ipendency in the following terms: "Emotional dependency is
jthat complex of processes, which operates to arouse sub
jective cravings to so relate oneself to others that ef-
33 !
fective control and direction of one's behavior is re- !
<
linquished in exchange for restrictive monopoly of the
flow of love and affection from others. Thus., in inter- 1
personal relationships it is the dependence of one indl-
|
vidual on the support., protection and love of another
which establishes an external security system to compen
sate for the lack of internal security" (49, pp. 13-15).
Warn formulates the following working definition:
"The need for dependency is manifested objectively by at
tempts to so relate oneself to others as to have effective
ly relinquished to them the control and direction of one's
illfe." Degrees of dependency could thus be specified in
terms of the amount of help an individual accepts or seeks
:from others (49, p. 125).
Although theoretical discussions of dependency tend
to assume implicitly or explicitly, a bi-polarity or
jdependence-independence, some investigators in working with
young children have made the opposite assumption. Beller
(4) sought to investigate the relationship among several
! specific components of both dependence and independence
;in young children. He first attempted to establish whether ;
^behavioral "items" could be considered as indicators of
Idrive, and second whether dependence and independence were
exact opposites. He postulated that five specific aspects
^of the early parent-child relationships form the basis for
3^
the development of a general secondary dependence drive.
These situations are the physical contact accompanying
reduction of primary drives, the proximity to the parents
which gradually becomes a substitute for direct contact,
the paying attention to the child which is a necessary
accompaniment of attending to primary drives, the help
which is originally necessary for the satisfaction of
needs, and finally the praise or approval or recognition
which is so frequently associated with drive reduction.
The development of a general secondary independence
idrive is postulated to be associated with the following
aspects of the early relationship: the taking of initia
tive which occurs.when help is not immediate and the infant
must struggle to manipulate his environment to get help,
the overcoming of obstacles associated with explorations
of the environment, the persistence to carry actions to
jcompletion necessary to reach any goal, the desire to be
active and getting satisfaction from doing work, and the
wanting to do things by oneself. Beller tested the assump
tion that there was a negative but not completely inverse
;relationship between dependency and independency and that
lit was unwarranted to assume bi-polarity in construction
I of measures for either drive. Forty-three nursery school
! children were each rated by a pair of raters on 10 rating
|scales, five each for measurement of the above components.
Analysis of the data revealed that the children differed !
consistently from one another in their composite dependency j
scores and that the five specific components functioned
as part of a general dependency drive within each child. j
Similar results were obtained on the five specific com- j
ponents of the independency drive. The two scores ivere
moderately negatively correlated and confirmed Beller's
hypothesis of two separate drives that were not completely
inverse.
Sears et_ al. (37) conducted an intensive investiga
tion into some child rearing antecedents of dependency in
young children. Sears conceptualized dependency as an
acquired drive learned as a product of the child's inter
action with his parents, and developing its specific quality
through the exact nature of the dyadic relationship exist
ing between parent and child. This includes the behavior
;of the child towards the parent and of the parent towards
the child. The parents define the learning process through ;
providing qualitative properties of the social environment
jto which he must respond and which he can manipulate, and
through also providing the rewards, punishments, and non
rewards that reinforce his behavior patterns. The parents
provide what kinds of behavior a child can learn as well as ;
determine through their reinforcement system the strengths
of the actions that are learned. Although there can exist ;
3 6 |
|
tremendous variations between parents, basically every j
|
human child must still receive care from someone throughout j
the first years of his life. Sears states that the de- 1
pendency drive results from the pervasive presence of the I
i
mother’s, or mother surrogate’s, performance of this nur-
:turant role. The earliest and perhaps invariant factor j
in the basic rewarding situation of being picked up and
having one's physical discomforts attended to is the
|mother’s attention. As the child gets older, the mother !
;responds to more selective signals for attracting her at-
1 j
;tention; first laughing, babbling, smiling, calling, later
isuch behavior as "question asking," "being cute," holding,
ietc. Drive properties are acquired following the general
[principle "that repeated rewarding and at least occasional
ipunishing of an instrumental act produce a drive to per
form that act and secure the environmental' results that
’ normally result from it" (37, p. 180). Dependent actions
;then become the goal responses for acquired drives of
Idependency. Behavior directed towards gratification of this
jdependency drive can be instigated by any factor that has
j
jcommonly been associated with reinforcement during the
acquisition of the drive originally. Thus the original
linstigators, mother and the need for primary gratification,
which are the conditions for eliciting the need for de
pendence in early infancy may become generalized into
"people" and frustration in general as instigators for the j
now-acquired dependency drive. j
Child-rearing antecedents investigated were infancy :
experiences, with specific respect to nurturance and j
i
frustration, the objects of dependency, the current home
experiences and maternal punitiveness. Measures of the :
dependency drive were ascertained by rating the child's
behavior in school and in his doll play. Analysis of the
resulting data led Sears to conclude that the crucial
period for dependency drive development was during the :
fourth and twelfth months. The main requisite for learning i
of the drive was the dyadic relationship created by the
feeding process, which furnishes adequate reinforcement in
itself to provide a dependency drive in all children. The
actual drive strength was determined by the amount of
\
frustration experienced in connection with feeding. Sears
found that the amount of dependency behavior exhibited
towards parents, siblings, peers, and teachers was a curvi
linear function of the amount of frustration and punishment
i
i
with which the child's dependency behavior had been met.
I
In his children, the maximum amount of dependency
behavior was produced by moderate amounts of frustration
and punishment, while both greater and lesser frustration
resulted in less dependency behavior. Severe punishment
seemed to result in a generalized inhibition of all behavior
including dependency behavior. Maternal frustration and
punishment seemed to have a greater effect on girls than on
boys, and hence there was an over-all greater amount of
generalized inhibition resulting from severe punishment
found in the girls. The object choice for dependency be
havior by the time the child reaches nursery school was a
function of the strength of the dependency towards the
mother and the severity of her frustration. Girls tended
to choose the teacher and boys tended to choose their
peers more as objects.of their dependency behavior.
Gordon (13)., working with adults, attempted to
validate the hypothesis that there was a unified trait of
dependence-independence. He sought to demonstrate a con
sistency between dependency in a social situation and in
a laboratory situation. Social measures of dependence were
obtained upon military hospital patients by means of a
behavioral checklist filled out both by the subjects and
by their ward physicians. Measures of dependency in the
laboratory situation were based on the degree to which the
individual was dependent upon cues external to the self as
opposed to cues internal to the self. On the basis of his
i
^results, Gordon concluded that there was a consistent
unified trait of dependence-independence as evidenced by
i
jthe fact that those individuals who described themselves
;as socially dependent on the Behavioral Check List were
39 !
found to be dependent (i.e., on external cues) on the Rod |
I
and Frame test* those individuals who described themselves j
; j
as independent were found to be independent on the labora
tory test, and those individuals who were described by |
their physician as dependent or independent were found to i
be comparably so in the laboratory situation. Gordon '
constructed an r#!PI scale of dependency based on an item
analysis of the MMPI's of the dependent and independent
subjects. This scale consists of 85 items which were
found to be significant at the 5 per cent level or higher
in differentiating socially dependent from socially inde
pendent subjects.
Kagen and Mussen (20) were able to differentiate
dependent from independent individuals with respect to
Ithe amount of group conformity they demonstrated. They
used content of TAT themes for establishing their criteria.
^Dependent subjects were defined as those subjects whose
:heroes sought help in a problem situation or were disturbed
lover loss of sources of love and support. They hypothesized
[that the dependent person* since he perceives others as ;
jmore competent than he* would be more likely to seek and
jaccept the guidance of others when he has to make a deci-
ision* and hence would yield to the opinion of a group as
wiser and more reliable than his own. They found that
their dependent subjects actually did yield to the incor-
40 |
;
rect opinion of the group, using the procedure of making j
i
i
visual discriminations in a group as outlined by Asch, j
It has also been possible to separate dependent and I
independent groups of patients with respect to how they j
handle serious physical ills. Bard (3) was able to predictj
the occurrence of psychogenic invalidism in women following1
radical mastectomy on the basis of pre-operative interview
ratings of dependence. Warn (49) sought to clarify the
role emotional dependence played in such cases of physical
handicap as epilepsy, paraplegia, and tuberculosis. He
concluded that emotional dependency was a prominent
feature with the tuberculous and epileptic, but not with
the paraplegics. He found also that increases in emotional
dependency were closely paralleled by increases In anxiety.
Bernardin and Jessor (5) attempted a construct
validation of the Edwards Personal Preference Schedule
with respect to dependency, arbitrarily defining dependency
in terms of "deference and autonomy scores," even though
;they admit other PPS variables such as succorance might
!have been used. They stated that a review of the litera- :
I
jture indicated considerable agreement about what is meant !
| ' i
!by dependency and included the following three components:
I a) reliance on others for approval, or importance of ap
proval from others; b) reliance on others for help or
assistance; and c) conformity to opinions or demands of
4 i ;
i
other's. For the purpose of their experiment they opera- j
tionally defined dependent subjects as scoring at above j
;the seventieth percentile on deference and at or below the j
thirtieth percentile on autonomy, with a minimum separa- I
i
(
tion of 30 centile points between the two scores for each ;
subject. Three experiments were conducted, each to test •
a hypothesis derived from the above three components of
their definition of dependency. They found that dependent
persons under conditions of such negative verbal reinforce
ment as critical comments, performed less efficiently on
a finger maze than either independent subjects under the
same conditions or a control, of both independent and
dependent subjects. They also found that dependent sub
jects in a difficult problem solving situation will ask for
help more often than independent subjects when assistance
is readily available. They failed to demonstrate evidence
|that dependent subjects exhibited more conformity behavior
i
(in the Asch procedure) than independent subjects.
Livson and Mussen (28), in an unsuccessful attempt
to merge the constructs of learning theory and psycho
analysis, sought to investigate the relation of ego control
j :•
to overt aggression and dependency. Proceeding initially
from Sears postulates that dependency and aggression are |
learned drives whose strength is related to early learning
experiences, they proceed to make the point that the amount !
42
of overt aggression or dependency is a function of both I
i
strength of drive as well as strength of inhibition against ;
the drive. The authors then assume that in the American j
middle class child there is a relatively strong inhibition j
of aggressive, sexual, and dependent behavior. They reason :
that this means— on the basis of analytic thinking, that
"internalizations of relatively strong inhibitions against
. . . these impulses . . . may be attributed to the func
tioning of the ego"— that children who acquire a high degree:
of ego control ai^e likely to manifest more socialized be
havior which in middle class culture would mean less overt
aggression and dependency. Using techniques of cosatiation :
and delay of gratification as indices of ego control, they
compared scores on these measures with ratings of aggressive
and dependent behavior over a period of observation in
nursery school. They found a significant negative correla
tion between degree of overt expressed aggression and
amount of ego control as measured by the cosatiation index.
They failed to find any significant relationships with
respect to expressed dependency. They concluded from this
hot that their hypothesis had failed to receive support,
but simply that their original assumptions about inhibitions
bf dependency in a middle class group did not apply to
their sample of children. They speculated that perhaps
dependency had not been discouraged in this social class
43 |
: 1
of children, since they were from upper middle class pro- j
fessional and academic families who are likely to "be
permissive with respect to dependency" but not aggression. |
Actually, it could be more justifiable to assume that j
i
both aggression and dependency were permitted greater ex- j
pression in this group of children than in an "average"
middle class sample, both in the home and at the nursery
school, which is generally geared to handle aggressiveness
more permissively than the average home situation because
of a greater sophistication and acquaintance with prin
ciples of child development.
Summary
Dependency is most generally regarded as an emotional
: need for support or guidance from others in order that the
Individual may be helped to direct his own affairs. This
peed originates in the early infantile helplessness of the
human being and in later life may create intense conflict
When opposed by maturational and cultural pressures towards
^independence. Although generally regarded as a bi-polar
■
dimension of dependency-independency, certain research has
presented evidence of the existence of separate drives
iof dependence and independence negatively, but not neces-
I
sarily inversely, correlated.
The experimental literature includes studies of
4 4 |
i
i
dependency as an acquired drive learned as a product of j
the child's interaction with his parents. This early i
! i
; j
;relationship inevitably involves the reinforcement attendant
i i
iupon satisfaction of the child's basic physiological needs, j
; j
^Eventually the attention and help associated with the per
formance of the original nuturant role of the parents be- ;
comes a secondary goal in and of itself. The original
instigators of the drive— the parents and the physiological
needs— become generalized to other people and all frus
trations.
Experimental studies on dependency have sought to
demonstrate its universality as a general tendency to rely
on sources outside of oneself for guidance and direction
of one’s behavior, which is manifest in both laboratory
land social situations.
CHAPTER III |
: i
METHOD |
I
I
Subjects I
! |
j
Qeneral Criteria for all Subjects
Used In the Original Design
Subjects In the original design were selected from
the caseload of the Los Angeles State Mental Hygiene Clinic.
This clinic is an outpatient neuropsychiatric facility for :
residents of the State of California. Female patients were
selected for subjects because of their greater availability
for daytime appointments. All subjects had completed the
usual clinic application procedure xvhich consists of an
intake interview with a psychiatric social worker, a
diagnostic interview with a psychiatrist, and a Minnesota
Multiphasic Personality Inventory. All the subjects used
in the study had been accepted for treatment and at the
time of their appointment with the experimenter, were await-
| ;
ing treatment, or were undergoing treatment, or had
terminated their treatment some time previously. All sub
jects met the following criteria:
i 1. Native born female between ages of 18 and 50
2. No evidence of psychosis apparent in the diag
nostic interview
3. No history of hospitalization for mental disorder;
4 . "Dependency"'1 ' scores on the MMPI falling within j
i
I
the arbitrarily defined limits of "dependent" j
|
i
or "independent." j
The selection and testing of the subjects extended j
i
I
over a period of approximately six months. Subjects were j
i
selected first from the open case load and then, as addi
tional cases were needed, the closed files were consulted,
working backwards from the most recent closures. Since
■ there was no way for the open files to be inspected by 5
ithe writer, all therapists were asked to cooperate in the ;
research by referring those of their patients who met the
; above criteria to the writer. The response of the staff
was quite generally a positive one. However, it soon be
came evident that a biased sample was inevitable because
of the fact that the therapists automatically eliminated
from consideration any patients they felt would find the
|experience too anxiety provoking or disruptive to their
treatment. Thus it is probable that a "high anxiety"
; group of patients, who contributed very heavily to the
positive findings in Taffel's study, were eliminated.
Those patients who were undergoing treatment were
I asked by their therapists if they would cooperate in a
j ^The "dependency" scale will be described later in
!the chapter under "Test instruments."
47 |
routine research testing program being carried on by the i
i
; I
.Clinic. Their names were then given to the writer, who !
' ]
contacted them directly by phone to set the actual appoint- j
ment time. No estimates were available as to what propor- !
f I
tion of the patients asked by their therapists to cooperate j
did so, but all those finally contacted by the writer were '
agreeable. The files of those patients who had been ac
cepted for treatment but xvere still pending assignment
were then screened for subjects meeting the criteria, and
those patients who did so and who could be contacted by |
phone were asked to cooperate in the procedure, which was
explained to them as routine and an extension of the ap
plication procedure. Finally, the closed files were con
sulted and those patients with phones who met the above
criteria were contacted by the writer and asked to cooperate,
in what was described essentially as a follow-up program,
and appointments were made at their convenience. This
sample of patients was probably biased to some extent, due
to the automatic elimination of those former patients who
bould not be contacted because their phone numbers or
addresses were incorrect, or who had moved away, or were
unavailable for some other reason. Of those actually con-
j
tacted, all agreed to come back for the one session. The
distribution of patients with respect to their status
pending treatment and the group to which they were assigned i
48 !
j
may be seen below. It appears that the composition of the j
i
Independent Group has a higher preponderance of closed I
i
cases. However, a check of the results made during the j
; study showed no apparent difference between the responses
of patients in treatment, pending treatment, and closed.
Treatment Cases Pending
'Group Active CTosed Treatment Total
: Hep 12 1
9
22
; Ind 8
9
o
20
I Con
9 3
8 20
29 13
20 62
■Selection of Additional Groups
l and Modification of the Design
As discussed above in the statement of the problem.,
i
; midway during the research it became evident that the
Iphenomenon of verbal reinforcement was not being demon-
istrated in either of the experimental groups. The first
^question then asked was whether the selected sample.,
jneuropsychiatric outpatients, were not amenable to verbal
I
ireinforcement which had been reported in the literature to
:be effective with groups of NP hospital patients and GM&S
Ihospital patients. In order to test this supposition,
|an additional experimental group consisting of 20 normal
female subjects, Group Norm (I), was drawn from the female
49 |
personnel on the poliomyelitis ward at Rancho Los Amigos. I
■These subjects met the first three of the above criteria, |
\
as well as the additional one of having no history of
psychotherapeutic treatment, according to their statements
1 j
to the experimenter. !
One other methodological consideration was sys
tematically investigated. The technical details of the
present experimental procedure had differed slightly from
those previously reported in the literature. A second
group of normal subjects was selected from the first 20
volunteers in evening school psychology classes at the
University of Southern California and the University of
California at Los Angeles. These subjects also were
within the age range of 18 to 50, stated to the experimenter
that they had no history of hospitalization for mental dis
order and had had no psychiatric or psychological treatment,
|and showed no overt indications of psychosis in their
classroom behavior. With these subjects the experimenter
varied the procedure slightly in order to more closely
jduplicate that of Taffel.
i
! The final design included the following five groups
jof female subjects:
i Group Dep: 20 Dependent neuropsychiatric outpatients:
Group Ind: 20 Independent neuropsychiatric out
patients
Group Con: 20 Mixed Dependent and Independent
neuropsychiatric outpatients
Group Norm (I): 20 "normals"
Group Norm (II): 20 "normals" subjected to varia
tions In the experimental design.
Test Instruments
The Dependency Scale
In the present study, dependency was defined em
pirically according to scores achieved on the dependency
scale of the Minnesota Multiphasic Personality Inventory
derived by Gordon (l3)j which may be found in Appendix B.
This scale was derived by an item analysis of the MMPI
profiles of male army neuropsychiatric hospital patients
who had been found to be significantly more dependent,
according to ratings of their ward psychiatrists and
therapists, on a dependency check list (see Appendix B).
Those items finally selected in the scale were significant
at the 5 per cent and 1 per cent level in differentiating
between the dependent and Independent groups.
Validation of the Dependency Scale
j
I Further validation of the Dependency Scale upon a
female population was necessary before it could safely be
assumed to be an adequate measure of dependency for the
51
present research. Members of the Psychology Department at
Metropolitan State Hospital were asked to rate, on the same
check list used by Gordon, all female open ward patients
for whom previous MMPI's were available, and with whom
they were acquainted through individual or group psycho
therapy sessions'. An additional 11 patients from the Los
Angeles State Mental Hygiene Clinic were rated by their
therapists upon this same check list in order to add to
the 42 cases obtained from Metropolitan State Hospital.
The validity of the Dependency scale for women was demon
strated by a Pearson r of -.325* significant at the 5 per
cent level, between the sum of the weighted scores on the
check list and the raw scores on the MMPI Dependency Scale.
Although the correlation is low, it was assumed if cor
rected for attenuation, since both measures were fallible,
the coefficient of correlation might have been higher (l6).
'It was impossible actually to perform such a correction,
however, because no reliability coefficients were available
;for either the rating scale or the MMPI Dependency Scale.
•Normalization of the Dependency Scale
i
In Gordon's original study, the Dependency scale
i
consisted of 80 items derived from the long form of the
IMMPI. In the present study a shorter scale of 6l items
j
was utilized, which consisted of those items of the
52 !
I
original scale which were included in the short form of j
the MMPI, which was used at the Los Angeles State Mental \
■Hygiene Clinic. In order to empirically define dependency, j
j
: a smoothed frequency distribution of the Dependency scores j
i
of 260 former female patients was obtained. Cut off j
points were then arbitrarily assigned in the following
manner:
Dependent Subjects: 70th to 100th percentile
Independent Subjects: 20th to 50th percentile
The subjects falling within the first decile at the ' ’In
dependent" end of the scale were eliminated for the purpose
^of avoiding those subjects whose extremely "Independent"
scores might be a reflection of a "protesting too much"
attitude or reaction formation against basically intense
dependency needs.
The Stimulus Cards
The materials for the experiment consisted of eighty
3x5 white index cards. Eighty commonly used verbs were
selected at random by the experimenter (see Appendix B),
i ;
|and a different verb was typed on each card. All the
1 j
verbs were in the past tense. Six pronouns, "I," "We,"
;"He," "They," "She," and "You," were typed below each verb
;and their respective order randomized for the series of
i :
eighty cards.
53
Summary of Procedure for Each Subject
At the time of the experimental session, the sub
jects who were patients at the Clinic were again told that
;they were participating in a routine research program and
that the results would have no effect on the course of
their individual therapy. _ It was felt that such reassur
ance was necessary in order to allay unnecessary anxiety
on the part of the subjects, who had to be considered first
as patients. Hoxvever, this reassurance may have inter
fered with their attitudes towards the experiment, which
would have influenced the results, since Taffel found that
verbal reinforcement is a function of the level of anxiety.
The two normal groups of subjects who were students were
informed that they were participating in the experimenter's
doctoral research. The Clinic subjects were seen in the
experimenter's regular office. The student subjects were
seen in an office close to their respective classrooms.
In all cases the purpose of the testing was explained
^briefly to them at the beginning of the session. They
jwere then seated with their back to the experimenter, and
;the following instructions were given:
"I have a number of cards, each of which has a verb
typed in the center and a pronoun underneath. I
want you to make up any sentence that comes to
5 4 i
i
j
mind quickly, just so long as it uses the verb and j
i
starts with any one of the pronouns. The sentence I
i
l
can be long, or short and can end in any way you '
want it to. Just talk at your normal speed. Why j
don't you try this first card as a sample?" i
The subjects were informed that their responses ;
would be recorded, although the experimenter actually
recorded only the pronoun and verb. They were encouraged
■to continue with all 80 cards and reserve any questions
until they had completed the task.
Following completion of the 80 sentences, the sub
jects were asked to seat themselves more comfortably facing
the examiner and were then asked the following questions:
1. Were you aware of me? (If so) In what way?
2. Were you aware of my saying anything? (if
so) What?
3. What do you think was the purpose of this test?
■The purpose of these questions was to eliminate those sub
jects who showed any awareness that the experimenter was
reinforcing certain responses. A brief explanation of the
procedure, based primarily upon whatever the subject
herself had surmised, was then made to the subject.
Those subjects who were able to verbalize the con
ditions of the experimental procedure with respect to the
relationship between reinforcement and their responses
55
were not included in the final analysis. Two additional
subjects, both from the Norm (II) Group, were tested but
their results were not included because of their insight
during the testing into the connection between their use
of a first person pronoun and the response on the part of
the examiner. The raw scores for these two subjects are
included in Appendix A. Most of the Subjects believed
that the experiment was a personality test similar to a
free association or sentence completion test. Several
subjects believed that the purpose of the procedure was
bo investigate- such specific factors as "ego-centricity"
;or "self-centeredness." However, careful questioning of
|these particular subjects revealed that they had no
awareness at all of the contingency of the examiner's be
havior upon their responses, and a brief check of their
results indicated no difference as compared to the subjects
|who did not refer to such factors in attempting to explain
;the purpose of the experiment.
Summary of Procedure for Each Group
I Group Dep: Experimenter reinforced all sentences
j starting with either first person pronoun
i
i by her response of "mmm-hmm" after the first
| 20 cards.
Group Ind: Experimenter reinforced all sentences
starting with either first person pronoun
by her response of "mmm-hmm" after the
first 20 cards.
Group Con: Experimenter remained silent throughout.
Group Norm (I): Experimenter reinforced all
sentences starting with either first person
pronoun by her response "mmm-hmm' 1 after the
first 20 cards.
Group Norm (II): Experimenter reinforced all
: sentences starting with either first person
pronoun by her response "good" after the
i first 20 cards,
Experimental Design
The present study was primarily concerned with a
I
possible functional relationship existing between the per
sonality characteristic of "dependency" and responsiveness
;to verbal reinforcement. The results were analyzed in a
factorial design which compared a Dependent, Independent,
'Control, and two Normal Groups of subjects with respect
[to their progressive modification of their verbal responses
j
under the conditions of continuous verbal reinforcement.
An analysis of variance for repeated measures on the same
subjects was performed twice: the first time comparing
all five groups of subjects, the second time comparing a
combined experimental group with the control group.
CHAPTER IV
RESULTS
The number of first person pronouns used by indi
vidual subjects during each of the four successive blocks
of 20 trials is recorded in the Appendix (Tables 8, 9* 10,
■11, and 12). The mean number of first person pronouns used
by each of the five groups of subjects Is shown in Table 1
and Figure 1, and the standard deviations of the scores are
shown in Table 2. An analysis of variance for repeated
measures on the same subjects was then performed and the
results are shown In Table 3- Where F ratios were found to
be significant or approaching significance, the appropriate
j"t" ratios were calculated. Table 4 presents the "t" ratios
for the over-all differences in the mean number of first
person pronouns between the various groups of subjects.
i
p?able 5 presents the "t" ratios for the over-all differences
In the mean number of first person pronouns between Blocks
I
bf Trials I, IX, III, and IV. The scores for the four ex-
|
jperimental groups were then combined and compared with the
control group by means of a second analysis of variance
|
and the results are presented in Table 6.
Since the deviations appeared to be of the
same order and unlikely to account for any variance,
Hypotheses A, B, and C were tested by the trials x
59
Table 1
Mean Number of First Person Pronouns Used During Each
of the Four Blocks of 20 Trials
Group n
Block
I
Block
II
Block
III
Block
IV"
2- I f
Dep 20 6.0 5.8
6.3
6.8 6.22
Ind 20 6.1
6.7 7.4
7.0 6.80
Con 20 7.8 8.8 8.0
8.5
8.23
Norm (I) 20 6.8 6.4 7.8 8.1 7.50
2 'Iorm (II)
1
20 6.4 6.4
r * 0
5- 8
6.8
6.35
2-5
6.62 6.92 7.12 7.6
Mean Number of First Person Pronouns
60
10
■o*
I III II rv
Blocks of Trials
Dep.
Ind.
A-- --Sk— — *
Cont.
0. . .
Norm
(i)
o-- -- o - —- -a Norm H
H
Figure 1
Mean Number of First Person Pronouns Used
During Each of the Four Blocks of Trials
61
Table 2
Standard Deviation of Scores for Each Group of Subjects
for Each of Four Blocks of 20 Trials Each
Group
Blocks of Trials
I II III IV
;Dep 3.06
3.87 3.43
4.46
Ind 2.94 2.38 2.88 3.88
Con 3.67 3.17
4.16 4.22
Norm (I) 2.92 3.36 4.18 3.81
Norm (II) 5.06 6.12 4.01
4.57
62
Table 3
Source Table for Analysis of Variance of Repeated
Measurements of the Mean Number of First Person
Pronouns Used by the Five Groups of Subjects
on Four Blocks of Successive Trials
Source of
Variation
Sum of
Squares
elf
Mean
Square
F
Between Groups 273.26 4 68.32
1.97 *
Between S's in
Same Groups 3294.42
95
34.68
Total between S's
3567.69 99
Between Trials 44.87
0
D 14.99
2 ,49 **
Interaction:
Trials x Groups
51.55
12 4,26 0.71
Interaction:
Pooled S's by
Trials 1713.38
285
6.01
Total within S's 1809.50 300
^Total
!
5377.19 399
* Significant at about the 10 per cent level »
I ** Significant at about the 6 per cent level.
i
|
63 ;
j
i
Table 4 I
"t, ! Ratios of Over-all Differences Between Groups in S
Mean Number of First Person Pronouns Used j
for the Total 80 Trials
Groups
xD
SEd
t
Dep-Ind .56 .78 • 75
Dep-Con 2.06 .78 2.65 **
Dep-Norrn (I) 1.06 .78 1.36
Dep-Norm (II) 1.30 • 78 .17
Con-Norm (I) 1.0 ■ 78 1.29
Con-Norm (II)
1.93 .78 2.48 **
Norm (I)-Norm (II)
.93 .78 1.20
Ind-Con 1.48
.78 1.90
Ind-Norm (i) .48
.78 . 62
Ind-Norm (II) .54 .78 • 58
** Significant between the 5 and 1 per cent level.
("t" must equal 2.025 to be significant at the
5 per cent level; "t" must equal 2.711 to be
significant at the 1 per cent level.)
64
Table 5
"tl ! Ratios of the Mean Number of First Person Pronouns
Used Between Each Block of 20 Trials by
Total Sample of Subjects
Trials
XD D
t
H
i
H
H
.20
• 35
.56
i i - i n .44
.35 1.27
I-17 .82
.35
2.36 **
II-III .24
.35 . 69
II-17 .62
■ 35 1.79
III-17 .38
.35
1.10
j
i ** Significant between the 5 and 1 per cent level.
("t" must equal 1.972 to be significant at the
5 pe*1 cent level; "t" must equal 2.601 to be
| significant at the 1 per cent level. )
Table 6
Source Table for Analysis of Variance of Repeated
Measurements of the Mean Number of First Person
Pronouns Used by the Combined Experimental
Group as Compared with the Control Group
on Four Blocks of Successive Trials
Source of
Variation
Sum of
Squares
df
Mean
Square
F
Between Groups
222.75
1 222.75 6.42 *
Between S's in
Same Groups 3294.42
95
34.68
Between Trials
44.97 3 14.99 2,49 **
Interaction:
Trials x Groups
15-33 3 5.11 0.85
iInteraction:
Pooled S's x
Trials
1
1713.38 285 6.01
j
! * Significant at about the 10 per cent level.
** Significant at about the 6 per cent level.
66
methods interaction term, as shown in Table 3. Since this
term yielded an insignificant F, none of the hypotheses
was supported. Further analysis of the data was then made
iin order to obtain any additional information possible
from the material.
For the purpose of discussing the results, the
Dependent Experimental Group will be designated by "Group
Dep," the Independent Experimental Group by "Group Ind,"
the Control Group by "Group Con," the Normal Group by
"Group Norm (I)," and the Normal Group with whom previously
reported procedures were more closely duplicated by
|"Group Norm (II).'1
Hypothesis A. It was predicted that the mean number
of positively reinforced responses of Group Dep would in
crease significantly more than those of Group Ind. An in
significant F ratio of .71 was obtained for trials x groups
^interaction, using as an error term the pooled subjects
x trials. Thus, it was impossible to reject the null
hypothesis of no difference between Group Dep and Group
Ind in their response to verbal reinforcement, and
Hypothesis A was not supported.
Hypothesis B. It was predicted that the mean number
jof positively reinforced responses would increase signifi
cantly more in Group Norm (I) than in either Group Dep or
67
Group Ind or Group Con. The insignificant interaction F
ratio of .71 made it impossible to reject the null hypothe
sis of no difference between Group Norm (I) and Groups
iDep, Ind, or Con, and Hypothesis B was not supported.
Hypothesis C. It was predicted that the mean number
of positively reinforced responses would increase signifi
cantly more in Group Norm (II) than in either Group Dep,
Ind, Norm (I), or Con. The insignificant Interaction F
ratio of . 71 made it impossible to reject the null hypothe
sis of no difference between Group Norm (II) and Groups
Dep, Ind, Con, and Norm (I).
In order to test the present hypotheses it was as
sumed that the verbal response patterns of any subjects
exposed to verbal reinforcement would differ significantly
from the response patterns of subjects not exposed to such
reinforcement. Specifically, it was assumed that this
[difference would consist of a significant Increase In the
occurrence of the reinforced response. That is, the as
sumption was made that verbal conditioning without aware
ness was a universal phenomenon which could be evoked by
Subsequent experimenters under similar, although not
^necessarily identical, experimental procedures to those
previously reported as successful. The results of the
present study indicate that in this instance such an as
68
sumption was unwarranted. The frequency of occurrence of
;the reinforced responses did not alter in those subjects
;who were exposed to verbal reinforcement as compared to
the control subjects who were not exposed to reinforce
ment. However, the control group of subjects in the
present study was limited to a neuropsychiatric popula
tion. It is possible that a control group drawn from a
normal population might have responded differently. For
example, a normal control group might have decreased in
the frequency of their use of the first person pronouns
■with the passage of time. On the other hand, neuro
psychiatric clinic patients are oriented towards examina
tion and report of their own feelings, attitudes, interests
land behavior. In a sense, they are "first person oriented,
and this may have tended to maintain or increase their
frequency of occurrence of the selected response ir
respective of the absence of any experimenter-controlled
reinforcement. Previous experiences in treatment might
'have furnished secondary reinforcing experiences for first
Iperson responses, which in general are the primary concern
!of patient and therapist in psychotherapeutic treatment.
I
| As a matter of fact, inspection of Table 1 suggests
ithat, as a whole, the subjects in the Control Group seemed
|
;to use the first person pronouns more freely than any of
ithe other groups. Since the Between Groups F ratio of
1.97 (Table 3) approached significance at about the 10 per
cent level, the possibility arises that there may have j
: |
; existed a significant difference between Groups. In order I
; to test this hypothesis^ "t” ratios were calculated for
the over-all differences between groups with respect to j
the mean number of first person pronouns used and the
results are shown in Table 4. A "t" ratio of 2.65^ sig
nificant between the 5 per cent and 1 per cent level, was
obtained between the Control and Dependent Groups. A , ! t"
:ratio of 2.48, significant between the 5 per cent and 1 ;
' [
per cent level, was obtained between the Control and j
: Normal (II) Groups. Thus, the Control Group used slgnifl- j
cantly more first person pronouns than one of the Experi
mental Groups, as well as one of the Normal Groups.
Observation of the raw data suggests that the number :
of first person pronouns increases as trials progress. An
|f ratio of 6.49, significant at about the 6 per cent level,
Iwas obtained between trials, using as an error term the
pooled interaction for subjects x trials which is based
jupon correlated measures. "t" ratios were then computed
i
jbetween Blocks of Trials for the total 100 subjects and the
jresults are presented in Table 5. A "t" ratio of 2.36
*was obtained between Trials I and IV, which is significant !
] - ;
I '
|at the 1 per cent level. It Is possible to conclude,
! therefore, that there is a highly significant tendency for
70
subjects to use the first person pronouns with increasing
'frequency as trials progress. Since the only significant
increase occurs between the first 20 and the last 20
trials) it would be important that any future studies
include a minimum of 80 stimulus cards in view of the fact
:that this tendency apparently does not manifest itself
until after a sufficient interval of time.
Further evaluation of the data was made by means
; of a second analysis of variance which compared the Control
I Group with a combined Experimental Group consisting of the
; four separate Experimental Groups considered as a whole.
: The.results are presented in Table 6. An F ratio of 1.8
for trials x groups interaction was insignificant and
failed to reveal any evidence of a differential response
to the presence or absence of verbal reinforcement. A
^Between Groups F of S.42 was significant at about the 10
;per cent level and was interpreted as pointing to the
above discussed difference of the Control Group with
respect to their more frequent over-all use of the first
person pronouns.
CHAPTER V |
: i
|
DISCUSSION !
! j
: i
!
The present study has failed to demonstrate, despite j
:the use of a variety of techniques and subjects, operation j
:of the principle of verbal reinforcement, Such discrepant
.findings from those reported in the literature warrant
closer Inspection of previous results. The studies of
;Cohen (7) and Taffel (46), which most nearly resembled the
present research, were selected for closer study.
Table J presents the mean number of first person
pronouns used during the four blocks of 20 trials, as '
reported in their studies, for the following groups:
i
l i
! Cohen: Experimental
Control
Taffel: Combined Experimental (Irrespective of
level of anxiety)
Experimental: high anxiety
Experimental: low anxiety
i
! Control !
| Taffel and Cohen both reinforced the same class of
behavior as in the present study, namely, first person
pronouns, but used the word "good" as the verbal reinforce- ;
ment rather than "mmm-hmm" as was used by the present ex
perimenter.
Table 7 j
i
i
Mean Number of First Person Pronouns Used During Each
of the Four Blocks of 20 Trials as Reported in
the Studies of Taffel and Cohen
Group
Blocks of Trials
Mean
T II ' III IV
’ Cohen
Experimental
8.7
10.3 11.5 13.5
11.0
Control
8.5 8.5
8.4
8.3 8.3
Mean 8.6 9.4 10.0 10.9
Taffel
Experimental 8.2
11.3 10.9
12.8 14.4
Control 7-2
10.5 7.5
8.4 8.2
High anxiety
8.3
12.1
13.1 15.3 11.5
Low anxiety 7.8
9.9 7.2 9.2 8.6
Mean
7-7
11.0
9.7
11.4
73 !
|
Inspection of Tables 1 and 7 shows that the original I
operant level use of the first person pronouns of the j
various groups extended over a wide range. The mean use j
of the first person pronouns varies from a low of 6.0 |
i
found in the present study’s Dependent Group to a high j
Of 8.7 yielded by Cohen's Experimental Group. Inspection !
of the tables indicates that Cohen's results were most
markedly at variance with those of the present study. i
Cohen's Experimental Groups showed a difference of 4.8
between the mean number of first person pronouns used on
■Blocks of Trials I and IV, while there was a corresponding
difference of -.2 for the Control Group. In the present
study the greatest Increase was in the Normal (l) Group,
iwhich showed a mean increment of 1.3. Cohen reports an
F ratio of 6.6l for the methods x trials interaction
t
|(p = .01) as compared with the insignificant F ratio ob
tained in the present study.
Taffel was primarily interested In the inter
relation of anxiety to learning or verbal conditioning.
He hypothesized that the amount of verbal conditioning !
jthat occurred in an interpersonal situation would be a
jfunction of two factors. One would be the nature of the
i !
reinforcing stimuli, such as light, sound, or verbal
I i
response. The other would be the psychological character
istics of the subjects, one of which is degree of manifest ;
74
anxiety.
In Taffel's combined experimental group, there oc
curred an increase of 4.6 in the mean number of first
person pronouns used between Trials I and IV, while his
jcontrol group showed a corresponding increase of 1.2.
However, in his "low anxiety" experimental sub-group
there was an increase of only 1.4 in the mean number of
first person pronouns used. The results for this 'low
anxiety group begin to approximate the results obtained
;in the present study.
Taffel's first analysis of variance considered his
combined experimental group as a whole, irrespective of
level of anxiety. He evaluated the effect of three
methods of reinforcement on the occurrence of the re
inforced response as trials progressed. He obtained an
jF of 1.16 for trials x methods interaction, which was not
jsignificant. He did obtain an F of 5-595 (p = .01) between
groups, but this should only be interpreted as an over-all
difference between the three groups, rather than a dif-
jferentlal response to progressive reinforcement. He then
performed another analysis of variance using only the
!"high anxiety" sub-group. Here, too, he reported an in-
I
|signiflcant F ratio of 1.95 for trials x methods inter-
lac tion, but a highly significant F ratio of 6.705 (p = .01)
between groups (i.e., conditions of reinforcement). Here
75 I
: |
again he has only demonstrated an over-all difference be- j
tween the three conditions and he is not justified in j
interpreting j as he does., that these results demonstrate j
^ i
;the operation of verbal reinforcement. Analysis of vari- j
| j
ance performed on the medium anxiety sub-group yielded j
' 1
comparable results, although the F ratio between groups was ;
significant only at the 5 per cent level. Analysis of
variance performed on the low anxiety sub-group failed to
yield any significant F ratios.
Further reflections upon the various findings im-
ipresses one with the highly complex, sensitive nature of
the phenomenon under investigation. In o.ne situation
,(Greenspoon), several media of reinforcement serve to
imodify verbal behavior; in another (Taffel), only verbal
; reinforcement is successful; in the present study verbal
reinforcement is unsuccessful. It Is important to consider
i
I
ithat these conflicting experimental findings may be a re-
[sult of differences in situational variables In the various i
: experiments. Sarason has pointed out that "practically
jail clinical problems are concerned with data obtained
'from an interpersonal interaction, and that regardless of
[how interactions differ, certain variables are always
;operable and must be taken into account" (35, p. 2). These ■
[variables include the personality and psychological
orientation of the psychologist, the purpose and organiza-
:tion of the specific clinical or experimental situation, j
land finally the total personality organization of the sub- !
; |
jject or patient. j
Dynamic theory (38) stresses that in any dyadic |
i j
^interaction, the psychological properties of the members
of the dyad, i.e., the subject and the experimenter, de-
1
itermine to some degree the kind of interaction that will
occur. It is likely that the resulting harmonious or
1 discordant interaction may have unpredictable results |
: 1
; upon experimental findings. In psychology tho experimenter j
is one of the tools of observation. The interpretations
and conclusions about experiments must take into account
the effect of all its tools of observation upon the process
being studied. The field of psychodiagnostics was the
first to furnish evidence of the influence of the person
ality of the examiner upon the type of material elicited.
Lord (29), for example, was able to demonstrate that
I examiner differences were more important than differences
in administration of the Rorschach in producing variations
jin the records, although she was unable to conclude as to
what specific and measurable factors in an examiner's
personality would bring about specific changes in a Ror
schach protocol. In the field of experimental psychology,
Postman and jarrett have pointed out . . the variety
iof cues to which S's respond in a verbal learning situation ;
77
is strikingly demonstrated by the significant contribution
which differences among the E's make to the total variance
of the S’s responses. ..." "These facts point to an
;important methodological conclusion. In many experimental
'situations there is necessarily an interaction between
I the behavior of E and S. We have paid too little attention
to the contribution made by variations in the E's behavior
to the experimental results. The difficulty which many
psychologists experience in repeating the results of other
; Investigators may be due to our failure to attack sys
tematically the role of differences among E's" (33> p. 255).
IBinder, McConnell, and Sjoholm (6) found a significant
difference in the rate of learning for the subjects of two
different experimenters who differed with respect to age,
isex, and prestige. In their experiment on learning with
out awareness they found the reinforcement of "mildly
|hostile" verbs to be more effective with a soft-spoken
'young lady of 90 pounds than with a six-foot former Marine
Captain.
I The importance of the specific situation in psy
chological testing has been emphasized by Rotter and Pharee
(34). They were able to demonstrate that the preference
lvalues of a list of reinforcements differed predictably
I depending upon the situations in which the ratings were
obtained from high school students. They conclude that
i"in both clinical and experimental testing, the group and
the individual subject's behavior must be appraised in the
!light of the reinforcements provided by the physical set
ting and the examiner, as well as by test instructions.
:... It seems likely that similar effects may occur if
;the subject is a hospital patient and sees the examiner as
some one in authority or not, or if in a typical experi
ment, the subjects presume that the experimenter is an
instructor or graduate student or if they anticipate
iif the examiner is one who can provide them with help"
| (34, p. 291).
As a matter of fact, the importance of the setting
and meaning of the experiment to the subject was commented
;upon by Taffel in his attempt to explain some of the dif
ferences between his results and previous researchers.
In commenting upon the fact that both he and Ball found
| that a light stimulus failed to act as a reinforcer, in
contrast to their success with verbal reinforcement and
in contrast to Greenspoon's success with both media of
I reinforcement, he speculated upon the fact that their
|design used a clinical setting and actual patients, whereas
Greenspoon used psychology students in the setting of an
1
I experimental laboratory. Taffel postulated that there is
|no meaningful association between a visual stimulus, such
;as a flashing light, and the experimental setting in a
79
hospital, which is apparently perceived by the subject-
' patient as a means of communicating with the psychologist.
I Questioning of the subjects used in the present
study revealed that many of them believed that the experi
mental procedure was a type of personality test. It is
!quite likely that for these subjects, at least, any cue
; from the examiner such as a smile, nod, etc., might
function as a reinforcer. Talbot (47) actually did find
; that note taking served to reinforce the selected response,
using an experimental design with psychology students
i
j similar to the one used by Greenspoon. Greenspoon, whose
|original research furnished the prototype for subsequent
studies of the reinforcement of verbal behavior, was operat
!ing in an experimental psychology laboratory with subject-
|students in a first year psychology course who were all
I"primed," as it were, to regard their verbal productions
i
|as highly meaningful communications to the experimenter-
teacher. Such highly motivated and interested students
are also more likely to be "set" to perceive any change
|in their environment such as a light, sound, or buzzer
las significant and meaningful. On the other hand, it
^is probable that hospital patienfc-subjects are more clin-
jically oriented towards verbal communications or cues from
i
|the psychologist rather than from the surroundings,
j It is highly conceivable that the experience of
hospitalization itself Influences the subject-examiner j
|
relationship In such a way as to enhance the very phenomenon
under investigation, thus enabling Taffel and Cohen to j
demonstrate the operation of verbal reinforcement which
the present study was unable to do. It is probable that a j
i
i
variety of determinants influencing this phenomenon, such j
as dependency, motivation, and sensitivity, may be inten
sified in hospitalized patients. Such a hypothesis could
easily be tested by means of a research design comparing
various types of outpatients with respect to the above
'variables. It is certainly probable that the motivation
of most hospital patient-subjects to "do well," or in
other words what the experimenter-psychologlst wants— is
intensified because of a desire to speed up their ultimate
;discharge from the hospital. In Taffel's study his in
structions emphasized, in fact, that the experiment was a
:"test" to aid in their treatment. The present researcher,
ion the other hand, emphasized the routine nature of the
task and minimized the importance of the test to the pro
gress of the subject's treatment.^ Cohen's subjects were !
I |
!informed that they were participating in a research
program, but they were nonetheless hospitalized at the
! ^The importance of the experimental session was even ;
!further minimized by making no charge for the session.
81
time and perhaps predisposed to perceive the procedure as
;important to their treatment program.
The nature of the interaction between the subject
;and experimenter may well be the most influential deter
minant of whether or not a subject will modify his response
:pattern, and of course such interaction is most difficult
to specify and control. The nature of the experimental
;setting, the instructions to the subject, and the way in
;which the experimenter is identified overtly and perceived
!implicitly may all be of paramount importance.
Another possible contributing factor to the differ
ence in the present results from those reported in the
iliterature may lie in the use of all female subjects.
|It was decided to limit the subjects to women primarily
jbecause of their greater availability for daytime appoint-
!
iments. It may well be that, as a class, women are less
|amenable than men to modifying their verbal patterns of
ibehavior. Perhaps the stereotype of the "chattering
■female" who does not pay too much attention to what anyone
else has to say has, at its roots, a germ of psychological
"truth" that is reflected in the writer’s inability to
modify the verbal patterns of response of a number of
jdiffering groups of female subjects. Furthermore, the sex
i
iof the experimenter and the reaction of the subjects to
ithis may be of importance, since the male is more freely
82
accepted as an authority figure in this culture.
Ihe thesis of the present study was that individuals
vary in their responsiveness to reinforcement, reward,
suggestion, guidance, and/or approval. There now appear
to be many sources for such variation. Certainly the
experimenter-subject relationship may be one of these
sources, as may the properties or characteristics of the
Individual experimenter and the individual subject. It
is difficult to duplicate the exact properties making up
this relationship from one experiment to another, partly
because it is so difficult to objectify and make explicit
these factors.
CHAPTER VI
SUMMARY AND CONCLUSIONS
The Problem
Psychology has long been concerned with the under
standing and the modification of verbal behavior. In re
cent years investigations have demonstrated that individual
patterns of verbal responses may be modified in a pre
determined direction and without the awareness of the
subjects by means of certain techniques of verbal and non
verbal reinforcement. Having accepted that such a phe
nomenon was a demonstrable fact, the question was posed as
|to whether there were individual differences in the way
I subjects manifest such learning without awareness. The
jpresent study specifically concerned itself with the ques-
jtion of whether the personality dimension of "dependency"
jwas an important variable in the reinforcement of verbal
jbehavior. Since previous research has seemed to indicate
I that "dependent" or overly dependent individuals rely more
I
jheavily upon external cues for guiding their behavior
jrather than upon their inner resources, it was predicted
|that these dependent individuals would similarly rely more
j
'heavily upon verbal reinforcement cues for guiding their
84
response patterns, even though such reliance might be
without their conscious awareness. Specifically, this
study sought to demonstrate a differential response be
tween "dependent" and "independent" individuals with
:respect to their amenability to verbal reinforcement
without awareness.
The Method
The present study employed a technique first devised
by Taffel. The materials for the procedure consisted of
eighty 3x5 cards in the center of each of which was typed
one of 80 commonly used verbs. Six pronouns were placed
below the verb and their respective order was randomized
for the series of eighty cards. The subjects consisted
of five groups of patients selected in the following
manner:
I Group Dep: 20 female subjects who were patients at
the Los Angeles State Mental Hygiene Clinic
and who had achieved scores falling within a
| range arbitrarily defined as dependent on the
Dependency Scale of the MMPI devised by
Gordon.
j Group Ind: 20 female subjects who were patients at
the Los Angeles State Mental Hygiene Clinic
and who had achieved scores falling within a
85
range arbitrarily defined as independent on
the Gordon Dependency Scale.
Group Con: 20 female subjects who were patients at
the Los Angeles State Mental Hygiene Clinic
and who had achieved scores falling within
either the dependent or independent range
on the Dependency Scale.
Group Normal (_!): 20 female subjects who were ward
personnel at Rancho Los Amigos and who gave
no history of psychiatric or psychological
treatment.
Group Normal (II): 20 female subjects who were
; students In evening psychology classes at the
! University of Southern California and the
| University of California at Los Angeles and
who gave no history of psychiatric or
psychological treatment.
All subjects were seen individually by the examiner.
iThey were instructed to make up a sentence beginning with
;one of the pronouns and using the verb on each respective
I
[card. All subjects,, with the exception of Group Norm (II).,
iwere seated with their back to the experimenter during the
i
!
[period that they were constructing the sentences. Follow
ing completion of the cards, they were questioned closely
to ascertain if they had developed any awareness of the
8 6 j
contingency between any of their responses and those of
the examiner. For the Dependent, Independent, and Norm j
i |
(i) Groups, the examiner reinforced all sentences begin- i
i
ning with the first person pronouns, after the initial 20 |
cards, by means of positive verbal reinforcement consisting:
of "mmm-hmm." Reinforcement for Group Norm (II) consisted
of the word "good' 1 and the subjects were seated facing
the examiner. The control group received no reinforcement
throughout. ;
Results
The results of the experiment failed to confirm the
|hypothesis of a significant difference between dependent
jand independent subjects in their response to verbal re
inforcement. Furthermore, in no instance was it possible
to demonstrate the principle of verbal reinforcement. The
| experimental groups did not differ in response to the
!experimental procedure of verbal reinforcement from the
;control group who were not exposed to such reinforcement.
iThere was no demonstrable difference In the response of
i
j
jnormal and neuropsychiatric subjects. Minor variations
iin procedure did not result in any significant difference
!in the response of normal subjects to verbal conditioning.
The results of the study did show an over-all
'tendency of all subjects to Increase the number of
87
of references to the first person pronoun as trials pro
gressed. There was also shown a significant variation
ibetween groups of subjects with respect to their over-all
use of the first person pronouns. In our study, the con
trol group differed significantly from both the normal
and neuropsychiatrie subjects in their over-all tendency
to use the first person pronouns more freely.
Conclusions
; It was felt that the results of this study indi
cated the need for the publication of more exact specifi-
i
I cation of experimental procedures and design, before the
results of one experiment can be reliably repeated by
other investigators. Such specifications must include
!the nature of the experimental setting, the instructions
to the subjects, the role of the examiner in relation to
j
I the subject, sex of the subjects and examiners, and the
; properties of the individual interpersonal relationships
i
:between subject and examiner.
REFERENCES
REFERENCES
1. Alexanderj F. Our age of unreason. New York: J. B. j
Lippincott ancTTJo., 1^2. i
2. Ball, R. S. Reinforcement and conditioning of verbal j
behavior by verbal and nonverbal stimuli in a :
situation resembling a clinical interview. Unpub- j
lished doctor's dissertation, Indiana Univer., I
1952. |
3. Bard, M. The use of dependence for predicting psycho
genic invalidism following radical mastectomy. J.
nerv. ment. Pis., 1955, 122, 152-160.
4. Beller, E. E. Dependency and independence in young
children. J. genet. Psychol., 1955, 87* 25-35.
5. Bernardin, A. C., & Jessor, R. A construct validation
of the Edwards Personal Preference Schedule with
respect to dependency. J. consult. Psychol., 1957,
21, 63-67.
6. Binder, A., McConnell, D., & Sjoholm, Nancy A. Verbal
conditioning as a function of experimenter charac
teristics. J. abnorm. soc. Psychol., 1957, 55,
309-314.
7. Cohen, B. P., Kalish, H. I., Thurston, J. R., &
Cohen, E. Experimental manipulation of verbal
behavior. J. exp. Psychol., 1954, 47, 106-110.
8. Cushing, M. C. Affective components of the response
class as a factor in verbal conditioning. Disser
tation Abstr. , 1957. 17, 2313.
9. Daily, J. M. Verbal conditioning without awareness.
Dissertation Abstr., 1953, 13, 1247-1248.
10. Edwards, A. L. Experimental design In psychological
research. New York: Rinehart^ T9 "56.
11. Fenichel, 0. The psychoanalytic theory of neurosis.
New York: Norton, 1945.
12. Fuller, P. R. Operant conditioning of a vegetative
human organism. Amer. J. Psychol., 1949, 62, 587-
590.
90
13. Gordon* B. An experimental study of dependence-
independence In a social and a laboratory setting.
Unpublished doctor’s dissertation* Univer. of
Southern California* 1953.
14. Greenspoon* J. The effect of verbal and nonverbal
stimuli on the frequency of members of two verbal
response classes. Unpublished doctor’s disserta
tion* Indiana Univer.* 1951.
15- Grossberg* J. M. Hie effect of reinforcement schedule
and response class on verbal conditioning. Disser
tation Abstr.* 1956* 16* 2211.
16. Guilford* J. P. Fundamental statistics in psychology
and education. (2nd Ed.) New York: MeGraw-Hi11*'
1^50.
17. Hartman* C. H. Verbal behavior of schizophrenic and
normal subjects as a function of types of social
reinforcement. Dissertation Abstr.* 1955^ 15j 1652-
1653.
18. Hildum* D. C.* & Brown* H. 17. Verbal reinforcement
and interview bias. J. abnorm. soc. Psychol.*
1956* 53* 108-111.
19. Jones* Ernst. Pa.pers on psychoanalysis. (1st Ed.)
New York: Uoc'dand Ho .j I9T35
20. Kagen* J.* & Mussen* 0. Dependency themes on the
Thematic Apperception Test and group conformity.
J. consult. Psychol., 1956* 20, 29-32.
21. Kanfer* F. H. The effect of partial reinforcement on
acquisition and extinction of a class of verbal
responses. J. exp. Psychol. * 1954* 48* 424-432.
22. Klein* S. Conditioning extinction of operant verbal
behavior in neuropsychiatric hospital patients.
Dissertation Abstr.* 1954* IB* 2127-2128.
23. Krasner* L. A. A technique of Investigating the
relationship between behavior cues of examiner and
verbal behavior of patient. J. consult. Psychol.*
1958* 22* 364-366.
24. Krasner* L. Studies of the conditioning of verbal
behavior. Psychol. Bull.* 1958* 55* 148-170.
91
25. Levy, D. Maternal overprotection. New York: Colum
bia Univer. Press'^ 1943•
26. Lindsley, 0. R. Operant conditioning methods applied j
to research in chronic schizophrenia. Psychiat. !
Res. Rep., 1956, 118-139, 147-153.
27. Lindsley, 0. R., & Skinner, B. P. A method for the
experimental analysis of the behavior of psychotic
patients. Amer. Psychologist, 1954, 9, 419-420
(Abstract).
28. Livson, M., & Mussen, P. H. The relation of ego
control to overt aggression dependency. J. consult.
Psychol., 1957. 55, 66-71.
29. Lord, Edith. Experimentally induced variations in
Rorschach performance. Psychol. Monogr., 1950,
64, (10) No. 316.
30. Mandler, G., & Kaplan, V/. K. Subjective evaluation
and reinforcing effect of verbal stimulus. Science,
1956, 124, 582-583.
31. Marion, A. J. The influence of experimenter status
upon verbal conditioning. Unpublished doctor's
dissertation, Univer. of Calif., Los Angeles, 1956.
32. Maslow, A. H., & Mittelmann, B. Principles of ab
normal psychology. New York: Harper, 1941*7
33. Postman, P., & Garrett, T. F. An experimental
analysis of learning without awareness. Amer. J.
Psychol., 1952, 65, 244-255.
34. Rotter, J. B., & Pharee, J. E. An effect of the
situation on psychological testing. J. consult.
Psychol., 1956, 20, 291-293.
35. Sarason, S. B. The clinical interaction: with
special reference to theTl^orschachk Hew ’ York:
Harper' & Brothers,T 954.
36. Saul, L. J. Emotional maturity: the development and
dynamics of personality. Lippincott, 1947.
37. Sears, R. R., Whiting, J. W. M., Nowlis, V., & Sears,
P. S. Some child rearing antecedents of aggression
and dependency in young children. Genet. Psychol.
Monogr., 1953, 47, 135-236.
92
38. Sears, R. R. A theoretical framework for personality
and social behavior. Amer. Psychologist, 1951j 8,
476-483.
39- Sidowski, J. B. Influence of awareness of reinforce
ment on verbal conditioning. J. exp. Psychol.,
1954, 48, 355-360.
40. Skinner, B. P. The behavior of organisms. New York:
Appleton-Century, I938.
41. Skinner, B. P. Science and human behavior. New
York: Macmillan, 1953.
42. Skinner, B. P. Some contributions of an experimental
analysis of behavior to psychology as a whole.
Amer. Psychologist, 1953, 8, 69-79.
43. Skinner, B. P. Verbal behavior. New York:
Appleton-Century-Crofts, 1957.
44. Sontog, L. W., Crandall, V., & Lacy, J. I. Dynamics
of personality: resolution of infantile dependent
needs. Amer. J. Ortho., 1952, 22, 534-541.
45. Spivak, M., & Papajohn, J. The effect of the
schedule of reinforcement on verbal conditioning.
J. exp. Psychol., 1954, 48, 355-360.
46. Taffel, C. Anxiety and the conditioning of verbal
behavior. Unpublished doctor's dissertation,
Indiana Univer., 1952.
47. Talbot, E. The effect of note taking upon verbal
responses and its implication for the interview
situation. Unpublished doctor's dissertation,
Univer. of Calif., Los Angeles, 1954.
48. Verplanck, ¥. S, The control of the content of
conversation. J. abnorm. soc. Psychol., 1955* 51j
668-676. “ ------ --- ------- ~
49. Warn, L. J. A comparative Investigation of dependency
in epilepsy, paraplegia and tuberculosis. Unpub
lished doctor's dissertation, Univer. of Calif.,
Los Angeles, 1958.
93
50. Wilson, W. C., & Verplanck, W. S. Some observations
on the reinforcement of verbal operants. Amer. J.
Psychol., 69, 1956, 448-451.
51. Woodv/orth, R. S., & Schlossberg, H. Experimental
psychology. (Rev. Ed.) New York: Henry Holt,
195^
APPENDICES
APPENDIX A
STATISTICAL DATA
96
Table 8
Dependent Subjects: Mean Number of First Person Pronouns
Used During Four Blocks of Trials
Subject
Blocks of Trials
I II III IV
1 4 0 2 0
2 4 6 10
7
3
6
7 5 5
4 2 1 6
7
5
8
3 9
11
6
7
6 6 11
7 5 7 7 5
8
4 2
3
1
1 9
4
3
6 4
1 10 8 10
15
14
i
10 11
5
10
1 12
!
8
9 5
10
13
8 6 4 6
14
7 13
4 8
j 15
8 8
9
10
16
7 5 3
4
; 17
2 1
3
1
! 18
1
10
13
12
17
| 19 4
3
10
5
20
5
2 2 1
97
Table 9
Independent Subjects: Mean Number of First Person
Pronouns Used During Four Blocks of Trials
Subject
Blocks of Trials
I II III IV
1 6 6
9 7
2 4
7
11 8
3 9
10
7
8
4
3
4 4
5
5 9 5
10
19
6
5
6
5 3
7
1 4
3 3
8 1 8
13
11
9
6 8
9
8
10 10
9 9 9
11 6
7
8
5
12 6 8
5
8
13
4
5
6
9
14 10 6
9 7
15 7 5 . 7 7
16
13
12
9
4
17
4
9 9
6
18
7
1 1 0
19 6 6
9
11
20
5
8 4
3
98
Table 10
Control Subjects: Mean Number of First Person Pronouns
Used During Four Blocks of Trials
Subject
Blocks of Trials
I II III IV
1 12 11
9 7
2 10
15
18 14
3
6 10
5 5
4 12 11 12 ll
5 9
8 8
9
6 11 10 12 16
| 7 3
8 6
9
8 10 8 8 11
9 5
6 6 2
10 10 8 8
3
11 8
9 13 13
| 12
j
10 12 12 10
13
8 11 8
9
! 14 4
7
10
9
1 15
J
4
7 13 15
| 16 6 2
5 5
17
11 8 8
7
| 18 8
15
16
17
19
1 4 1 2
20 8 6
3
6
99
Table 11
Normal (i) Subjects: Mean Number of First Person
Pronouns Used During Four Blocks of Trials
Subject
Blocks of Trials
I II Hi IV
1
9
11
7
6
2
9
8 11 18
3
10
7
6 11
4 2 2 2 4
5
11
9
12
6 6
7 13
10
7
9 9
10 6
i 8
3 3
6
7
; 9
4 6 6
9
i°
9
11
19
14
ii
7
10 4 6
; 12 8
9
8 8
j 13
6 2 4 6
14 1
5
11
9
15 6 2
5
10
| 16
7
11
7 9
17
!
9 3 7
6
! 18
3
0 0 0
19
11
7
li 12
20
5
6
7
4
100
Table 12
Normal (II) Subjects: Mean Number of First Person
Pronouns Used During Four Blocks of Trials
Subjects
Blocks of Trials
I II n r tv
1
13
18
15
19
2
7
8
7
10
3 9 5
4 8
4
9
6 6 8
5 7
11 11
9
6
5 3
2
7
7
8 10 8 10
! 8 3 7 3 9
9 5
4 0 0
10 6 1 6 8
n
9 5 7 9
12
7
10 6 6
!3
2 1 0 0
i — I
4
3
2
3
15
j
4 6 11
3
16 4 8 8
5
i
17
5
6
5
11
CO
1 — 1
6
9 5
1
19
10 2
7 7
20
5
6 4
3
101
Table 13
Two Subjects Eliminated from Study: Mean Number of
First Person Pronouns Used During
Four Blocks of Trials
o„-u Blocks of Trials
Subjects j---------- n ---------jjj---------- w
9 10 6
0 16 18
APPENDIX B
TEST INSTRUMENTS
103
EIGHTY VERBS USED IN STUDY
showed furnished rode added
looted lost accused gained
charged shot beat fixed
: clashed acknowledged waved sold
resented married loved jumped
attracted walked traded donated
:made needed boasted stole
: gambled acquired gave assumed
fired
f
folded burned shoved
jfeared wanted joked fell
Ihovered saved grabbed packed
hung bathed threw pushed
waited ran doubted bought
:jabbed acted changed touched
moved had reviewed loaned
permitted consulted reduced allowed
'planted demanded closed took
Igathered dared carried adored
feasted rushed advised battered
'stated laughed hated borrowed
MMPI SCALE OF "DEPENDENCE-INDEPENDENCE"
104
Below are listed the items which were found by-
Gordon to differentiate significantly between "dependent"
and "Independent" individuals. They are presented in the
same numerical order as they appear in the group form book
let of the MMPI. The scorable direction of each response
is listed immediately opposite each item.
1. I have a good appetite. False
3. I wake up fresh and rested most mornings. False
5. I am easily awakened by noise. True
15. Once in a while I think of things too bad
to talk about. True
16. I am sure I get a raw deal from life. True
24. No one seems to understand me. True
32. I find it hard to keep my mind on a task or
j ob. True
35- If people had not had it in for me, I
would have been much more successful. True
4l. I have had periods of days., weeks or months
ivhen I couldn't take care of things because
I couldn't "get going." True
43. My sleep is fitful and disturbed. True
48. When I am with people, I am bothered by
hearing very queer things. True
49. It would be better if almost all laws were
thrown away. True
50. My soul sometimes leaves my body. True
52.
55.
57.
63.
67-
76.
79-
107.
138.
142.
154.
162.
164.
170.
171.
172.
194.
197-
I prefer to pass by school friends or
people I know but have not seen for a long
timej unless they speak to me first.
I am almost never bothered by pains over
the heart or In my chest.
I am a good mixer.
I have had no difficulty in starting or
holding my bowel movements.
I wish I could be as happy as others seem
to be.
Most of the time I feel blue.
My feelings are not easily hurt.
I am happy most of the time.
Criticism or scolding hurts me terribly.
I certainly feel useless at times.
I have never had a fit or convulsion.
I resent having anyone take me in so cleverly
that I have had to admit it was one on me.
I like to study and read about things that
I am working at.
lA/hat others think of me does not bother me.
It makes me uncomfortable to put on a stunt
at a party even when others are doing the
same sort of thing
I frequently have to fight against showing
that I am bashful.
I have had attacks in which I could not
control my movements or speech but in which
I knew what was going on around me.
Someone has been trying to rob me.
105
True
False
False
True
True
True
False
False
True
True
False
True
False
False
True
True
True
True
200.
201.
226.
236.
2 4 4 .
252.
2 5 9 .
2 6 3 .
: 265.
'267.
: 278.
292.
2 9 7 .
'299.
301.
j303.
;304.
305.
306.
There are persons who are trying to steal
ray thoughts and ideas.
I wish I were not so shy.
Some of my family have habits that bother
and annoy me very much.
I brood a great deal.
My way of doing things is apt to be mis
understood by others.
No one cares much what happens to you.
I have difficulty in starting to do things.
I sweat very easily even on cool days.
It is safer to trust nobody.
When in a group of people I have trouble
thinking of the right thing to talk about.
I have often felt that strangers were
looking at me critically.
I am likely not to speak to people until
they speak to me.
I wish I were not bothered by thoughts
about sex.
I think I feel more intensely than most
people do.
Life is a strain for me much of the time.
I am so touchy on some subjects that I
can't talk about them.
In school I found It very hard to talk
before the class.
Even when I am with people I feel lonely
much of the time.
I get all the sympathy I should.
106
True
True
True
True
True
True
True
True
True
True
True
True
True
True
True
True
True
True
False
107
309. I seem to make friends about as quickly
as others do.
312. I dislike having people about me.
318. My daily life is full of things that keep
me interested.
332. Sometimes my voice leaves me or changes
even though I have no cold.
336. I easily become impatient with people.
337- I feel anxiety about something or someone
almost all the time.
3^9. I have strange and peculiar thoughts.
O TO
361.
-
1 have no dread of going into a room by
myself where other people have already
gathered and are talking.
I am inclined to take things hard.
X feel uneasy indoors.
False
True
False
True
True
True
True
False
True
True
CHECK LIST
108
DIRECTIONS: A number of adjectives and descriptive sen
tences are listed below. Please read them over quickly
and put a check mark in front of each one you would con
sider to be descriptive of the following patient. If the
Item does not apply (i.e. items re: group therapy* or the
hospital) do not attempt to answer.
(Pa t i eri't1 s Name)
1. She has had periods of days* weeks* or months
when she couldn't take care of things because
she couldn't get going.
2. Her feelings are easily hurt,
3. She is easily downed in an argument.
4. She Is certainly lacking in self-confidence.
5. She has met problems so full of different possi
bilities that she has been unable to make up her
mind about them.
6. She is happy most of the time.
7. Criticism or scolding hurts her terribly.
8. She tires easily.
9. She wishes she were not so shy.
10. She broods a great deal.
11. She feels anxiety about something or someone almost
all the time.
12. She has several times given up doing a thing be
cause she thought too little of her ability.
13. She frequently asks people for advice.
14. People can pretty easily change her even though
she thought that her mind was already made up
on a subject.
109
15. She shrinks from facing a crisis or difficulty.
16. She is apt to pass up something she wants to do
when others feel that it isn’t worth doing.
17. She clarifies the group’s activity or feelings.
18. She attaches herself to some other patient in the
group in a dependent manner.
19. She feels she is physically sick and resists
seeing psychological factors.
20. She talks about her brothers and sisters.
21. She is very sensitive and can easily have her
feelings hurt.
22. She will usually enter the discussion only if it
is about fairly unemotional and impersonal matters.
23. She fidgets and seems uneasy in a group situation.
24. She reacts to her therapist in a rather dependent
and submissive manner.
25. She is the sort of person one would not mind
knowing socially.
26. She takes responsibility to draw others into the
conversation.
27. Her manner of discussing problems indicates she
has the ability to gain useful insights.
28. She expresses friendliness or warmth to someone
present in a fairly overt manner.
29. She considers herself completely at the mercy and
direction of others in the group.
30. She makes critical or self-derogatory comments
about herself.
31. Her behavior in the group therapy does not appear
to be like what she says her behavior is with
people away from the group.
110
32. She enjoys a lively group therapy session with
expression of hostility being quite open.
33. She expresses her feelings freely in the group.
34. The group usually accepts anything she may
express.
35. She feels called upon to assume a leadership
role in the group.
36. She doesn’t make distinctions between old and
new personnel in the hospital.
37- She chooses to talk either to the personnel or
to patients who talk sensibly.
38. She isn't backward about talking to you after she
gets acquainted.
39* She can talk sensibly if you ask her to.
40. She won't discuss many subjects.
41. She wants to see the doctor more often than others
do.
42. She gets along with the majority of patients.
43. She seems to enjoy being talked to.
44. She will always reply if you make some remark
to her.
45. She is making realistic plans for when she leaves
the hospital.
46. She occasionally needs supervision with dressing.
47. She takes pleasure in fixing her hair.
48. She is very reliable for errands.
49. She doesn't need supervision on a job.
50. acquiescent
51. adventurous
52, alert
53» anxious
54. apologetic
55- argumentative
56. assertive
57* bashful
58. broad-minded
59* capable
60, cautious
_ 6l. consistent
_62. conventional
_63. critical
_64. daring
_65. dependent
_ _66, dependab le
_67. docile
_68. eager
_69. fearful
JO. firm
_71. frank
J2. guarded
.73- guilt-ridden
74. imaginative
75. immature
76. individualistic
77. infantile
78. informal
79* inhibited
So. interest narrow
_8l. meek
_82. mischievous
83. organized
84. original
__85. outspoken
86. passive
87. practical
88. progressive
89. realistic
90. repentent
91. self-assertive
92. self-conscious
„93. s eIf-pi tyi ng
94, self-punishing
95* self-reproaching
96. sensitive
97* shy
98. sociable
99* soft-hearted
3-00* spontaneous
101. spunky
112
102. strong
103 •
subservient
104. submissive
105.
suggestible
106. super stitioui
107.
tolerant
108. unaggressive
109. unambitious
110. weak
lli. whiny

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Clinical Implications Of Verbal Learning Without Awareness: An Experimental Study

PDF

The Effects Of Group Experiences On The Aged

PDF

Conflict And Generalized Conflict In Verbal And Motor Responses

PDF

The Effects Of Sex, Assigned Therapist Or Peer Role, Topic Intimacy, And Expectations Of Partner Compatibility On Dyadic Communication Patterns

PDF

An Investigation Of The Response To Stress Of Patients Hospitalized For Anxiety State And Peptic Ulcer Patients

PDF

Psychological Test Changes In Schizophrenic Patients Under Brief Stimuluselectroconvulsive Therapy

PDF

Development And Analysis Of Some Tactual Measures Of Intelligence For Adolescent And Adult Blind

PDF

A Study Of The Relationships Between Factors Found In Cattell'S Sixteen Personality Factor Questionnaire And Factors Found In The Guilford Personality Inventories

PDF

Verbal Reports Of Emotional States And Onsets And Offsets Of Conditioned Stimuli

PDF

The Visual Fusion Threshold (Vft) Test As A Measure Of Perceptual Efficiency In Kindergarten And First Grade, And As A Possible Predictor Of Later Reading Retardation

PDF

A Study Of The Relationship Of Temperament Variables To The Ability To Make Certain Judgments Of Emotional Behavior

PDF

An Analysis Of The Selection Criteria For Assignment Of Students To Advanced Placement Classes In The Los Angeles Unified School District

PDF

Predicting Success In The Study Of Descriptive Linguistics

PDF

A Temporal Approach-Avoidance Conflict In An Academic Test Situation

PDF

Transfer Of The Partial Reinforcement Extinction Effect Across Tasks In Normal And Retarded Boys

PDF

The Concept Of Sexual Identity In Normals And Transvestites: Its Relationship To The Body-Image, Self-Concept And Parental Identification

PDF

The Relationship Of Self-Concept Discrepancies To Vocational Choice, Intelligence, School Achievement, And Socioeconomic Status

PDF

The Use Of Multiple Objective Measures In Survey Research As A Means For Predicting Respondent Future Behavior

PDF

On The Relationship Between Anxiety And Aggression In Nine-Year-Old Boys

PDF

Relationship Of Achievement Motivation To Perception Of Degree Of Task Difficulty And Estimate Of Performance

Asset Metadata

Creator Caruth, Elaine Greenbaum (author)

Core Title The Relationship Of Dependency To Verbal Learning Without Awareness

Degree Doctor of Philosophy

Degree Program Psychology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,Psychology, clinical

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Jacobs, Alfred (committee chair), Carnes, Earl F. (committee member), Meyer, Mortimer M. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-28857

Unique identifier UC11357736

Identifier 5902933.pdf (filename),usctheses-c18-28857 (legacy record id)

Legacy Identifier 5902933-0.pdf

Dmrecord 28857

Document Type Dissertation

Rights Caruth, Elaine Greenbaum

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA