University of Southern California Dissertations and Theses

Clinical Implications Of Verbal Learning Without Awareness: An Experimental Study

(USC Thesis Other)

Clinical Implications Of Verbal Learning Without Awareness: An Experimental Study

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content T h is d isser ta tio n h as b een 61-6279
m icr o film ed ex a ctly as rec eiv ed
CARUTH, C lifton J ., 1 9 2 5 -
CLINICAL IMPLICATIONS OF VERBAL LEARNING
WITHOUT AWARENESS: AN EXPERIM ENTAL
STUDY.
U n iv ersity of Southern C alifornia
P h .D ., 1961
P sy ch o lo g y , c lin ic a l
University Microfilms, Inc., Ann Arbor, Michigan
CLINICAL IMPLICATIONS OP VERBAL LEARNING
WITHOUT AWARENESS: AN EXPERIMENTAL STUDY
by
Clifton J. 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 1961
UNIVERSITY O F SO U T H E R N CA LIFO RN IA
G R A D U A T E S C H O O L
U N IV E R S IT Y PA R K
L O S A N G E L E S 7 , C A L IF O R N IA
This dissertation, written by
Clifton J, Caruth
under the direction of h.h$...Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by the Dean of
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
D ate....
DISSERTATION COMMITTEE
ACKNOWLEDGEMENTS
The author wishes to acknowledge the many people
who shared in the formulation and execution of this study.
Dr. David B. Klein, Committee Chairman, deserves a particu
lar expression of gratitude for his constructive criticism,
guidance, and encouragement throughout all stages of the
investigation. The author is also grateful for the in
terest and assistance proffered by the other members of
his Committee, Dr. William Grings and Dr. James Peterson.
The author is deeply thankful for the interest and
assistance given by Dr. Simon J. Conrad, Director, and the
staff of the Los Angeles State Mental Hygiene Clinic, par
ticularly those staff members who served as experimenters:
Dr. Merlin Dimitman, Dr. Milton Firestone, Dr. Leonard
Korot, Mr. Steven Marsh, and Mr. Howard Mass. Thanks is
also due to Dr. Walter Varnum and the Psychology Department
Staff of Los Angeles City College for permission to re
quest volunteer subjects from their classes.
To his wife the author can give only partial ex
pression in this form to his appreciation for her helpful
suggestions, ever present support, and faith during the
trying months devoted to this research.
ii
TABLE OF CONTENTS
CHAPTER PAGE
I. THE PROBLEM 1
Introduction and Statement of the Problem 1
Theoretical Background . ........... 4
The Purpose of the Study and Hypotheses . 10
II. REVIEW OF THE LITERATURE 14
Summary 51
III. METHOD 53
Experimenters 53
Subjects 54
Materials . . . . ............. 55
Procedures with Subjects .......... 55
Experimental Design...................... 59
IV. RESULTS ................................ 60
V. DISCUSSION .............................. 82
Suggestions for Future Research . . . 91
VI. SUMMARY AND CONCLUSIONS ............. 96
Summary 96
Conclusions ..................... . 101
REFERENCES 103
APPENDIX A. Tables of "I-We” Responses . . . Ill
APPENDIX B. Responses to Awareness Inquiry . . 116
APPENDIX C. Experimental Materials .... 122
iii
LIST OP TABLES
TABLE PAGE
1. Table of Means— All Groups • . 61
2. Source Table for Analysis of Variance of
Repeated Measurements of the Mean Number
of First Person Pronouns Used by the
Combined Reinforced Group as Compared with
the Non-Reinforced Group with Three Sets
of Experimenters ............ 62
3. Analysis of Variance of Trial Block Scores:
Reinforcement Treatments by Subject Class . 65
.4. t-Ratios of the Mean Number of First Person
Pronouns Used Between Each Block of 20
Trials by the Total Sample of Subjects . . 66
5. t-Ratios of the Mean Number of First Person
Pronouns Used Between Each Block of 20
Trials by the Total Reinforced Group of
, Subjects 67
6. Source Table for Analysis of Variance of
Trial Block Scores: Subject Class by
Professional Affiliation by Reinforcement
Treatments 70
iv
TABUS PAGE
7. t-Ratios for Differences Between Means of
Groups Related to Critical Factors in
This Study 73
8. Total Number of MI-We" Responses for
Reinforced Patient-Subjects Over the
Five Trial Blocks 112
9. Total Number of "i-V/e" Responses for
Reinforced Student-Subjects Over the
Five Trial Blocks 113
10. Total Number of ”I-We" Responses for Non-
Reinforced Patients Over the Five
I
Trial Blocks 114
11. Total Number of "I-We" Responses for Non-
Reinforced Students Over the Five
Trial Blocks 115
v
LIST OF FIGURES
FIGURE PAGE
1. Mean Number of First Person Pronouns for
the Total Reinforced and Total Non-
Relnforced Subjects Over the Five Trial
Blocks of Twenty Trials Each ........ 64
2. Mean Number of First Person Pronouns
Used by All Subjects Over the Five
Trial Blocks 68
3. Mean Number of First Person Pronouns Used
by the Total Patient Group and Total
Student Group Over the Five Trial
Blocks, Irrespective of Reinforcement . 72
vi
CHAPTER I
THE PROBLEM
Introduction and Statement of the Problem
The last decade has seen a surge of interest In the
analysis of the major variables Involved in what is general
ly termed the clinical Interaction. While Sarason (48)
states that the clinical interaction is but an instant of
interpersonal relationships In general, we might ncc-o uh&t
it is also an interaction of vital Importance to the per
sons Involved. Thus, it is a relationship in which as
little as possible should be left to chance and guesswork.
In a paper on emotionality and perceptual defense
McGinnies (35) gave experimental evidence in support of
responsiveness to cues without awareness. Using GSR
changes as an index of emotionality he demonstrated that
a person "actually discriminates a stimulus before he
perceives it. . . ." He thus demonstrated that affective
arousal occurs even though the Individual Is unsure of the
provocative cue or his affective reaction. Just as one
learns, at an early age, that certain words and gestures
are taboo and create lifelong feelings of anxiety, so he
also learns to perceive other words and gestures, such as
nods, smiles, etc., as rewarding and need-fulfilling.
2
It would follow, then, that under certain condi
tions if some chosen class of behavior was followed by par
ticular "unconsciously" rewarding stimuli, learning without
conscious awareness should take place. The research in
this area since 1950 has produced a variety of Investiga
tions, among which there have been a number of Important
studies examining changes in subjects' verbal behavior as
a consequence of "reinforcement" under different kinds of
controlled conditions. In many of these studies the sub
jects failed to perceive the reinforcing influences Intro
duced by the experimenter. Despite this absence of aware
ness1 their behavior was modified significantly by these
Influences.
The methods employed In these studies would appear
to be particularly suited to the study of verbal behavior
In the clinical setting. In this situation, more than in
any other Interpersonal relationship,'the clinician must
rely on verbal behavior as the basis for gaining an under
standing of his patient or client. Conversely, the
■^The concept of lack of awareness has been the sub
ject of some controversy and remains, to date, an unsettled
question (6,23,38). Whether there is partial recognition
of the instigating cues in studies of learning without
awareness is yet to be unequivocally demonstrated, but this
still remains a possibility. The present study assumes a
lack of awareness In the absence of any explicit statement
to the contrary by the subjects.
patient in the clinical relationship is even more depend
ent upon the clinician's vocalizations for gaining his
perception of what is expected of him and ultimately in
learning what his behavior means and how it might be
altered. This is especially true in clinical settings
where the clinician's technic dictates a special kind of
personal and affective neutrality in the relationship with
his patient. It is even more true in those instances
where the clinician is not visible to his patient, main
taining aB anonymous a role as possible, as in the classic
psychoanalytic interview. Vocal out-put and verbal be
havior take on dimensions of obvious importance under such
circumstances.
A number of questions arise, then,concerning the
nature of the verbal exchange in such settings. How is
the verbal behavior of the patient influenced or changed?
Do non-verbal factors in the situation contribute to such
changes as occur? Is the verbal behavior of.patients cur
rently undergoing psychotherapy changed In the same manner,
or as much as, other non-patient subjects? Do the personal
or professional characteristics of the experimenters or
clinicians significantly Influence the degree of change in
the verbal behavior of the subjects?
The present study is concerned with two clinically
important phases of verbal learning without awareness:
(a) the bearing of the experimenter's professional affilia
tion in such learning, and (b) possible differential in
fluences on patients in treatment as contrasted with
"normal” subjects.
For more precise clarification "verbal learning”
will be defined in this investigation as an increase in
the frequency of specific verbal response classes, known
as the criterion variables, following selective verbal re
inforcement of these classes by the experimenter.
Theoretical Background
The traditional goal of the psychotherapist has
been to effect constructive changes in the patients or
clients seeking his professional counsel. Classically, the
verbal interactions between himself and his patient have
constituted the therapist's chief measures of effecting
such changes. The functions and goals of* the therapist
appear similar from one school of psychotherapy to another.
However, the underlying theoretical formulations with
reference to the nature, control, and significance of the
verbal interchange between patient and therapist differ
markedly (3>8,37,42).
Thus, at one extreme, we have a proponent of the
"orthodox" psychoanalytic technic (37) proclaiming that
"it might be possible to defend the thesis that, other
factors being equal, the passive contributions of the
therapist are (alone) sufficient . . . he listens, but
makes little response except to indicate his close atten
tion. He makes no suggestion. He gives no comfort. He
makes no criticism. He gives no diagnostic opinion. He
only listens." The patient is left free to explore what
ever enters his mind, and there is minimum interference
by the analyst.
At the other extreme we find the so-called direc
tive technics of psychotherapy including such devices as
persuasion, hypnosis and hypnoanalysis, psychodrama,
bibliotherapy, and environmental manipulation in which
the therapist often assumes an outright authoritarian at
titude and is deliberately set to intervene with sugges
tions, instructions, and guidance (7*8*36,57)•
Finally, varying between these two apparent ex
tremes we find the teachings of such schools as that of
Carl Rogers' "Client-Centered Therapy" (42), and certain
neo-Freudian orientations (3,7*14,16). Rogers in his
"non-directive" therapy, also stresses the importance of
a permissive atmosphere without direction by emphasizing
the therapist's faith in the patient's capacity for self-
direction, growth, and insight. In this respect his ap
proach is like that of some analytic therapists. However,
Rogers seeks to have his therapists convey their active
interest and concern for the patient through the therapist's
frequent response to what the patient says by such tech
nics as murmuring "mm-hmm." Such therapist behavior, he
states, is without the purpose of guiding or directing
the patient's ideational output.
Despite these variations in technic, different
schools of psychotherapy report about the same percentage
of successes and failures. Thus Heine (21) finds that
patients report similar changes following therapy ir
respective of the therapist's school affiliation. He also
reports that patients who have shown improvement tend to
agree on which factors in the treatment influenced their
change. These factors have no relation to the therapist's
theoretical orientation. This is similar to the findings
by Fiedler that the therapeutic technic and results of
veteran therapists appear to be less similar to the
younger members of their own schools than to experienced
therapists of other schools (14).
The results of such studies suggest the advisabil
ity of examining the clinical relationship rather than
being exclusively concerned with the therapist's theoret
ical system and the technical maneuvers related to such
a system. We can no longer assume that behavioral change
occurs in our patients solely as the result of the employ
ment of specific technics growing out of particular
theories of therapy. There is a growing body of evidence
which demonstrates the therapeutic importance of factors
in the clinical setting which have previously been viewed
either as inconsequential or, at best, merely helpful in
establishing rapport. It now seems clear that such seem
ingly meaningless devices as a "non-directive" or neutral
"mm-hmm" or a casual nod of the head--in fact, almost any
gesture on the part of the clinician--might have an in
fluence on the patient's behavior (13*63). Studies of
testing and Interview-type situations have shown how cer
tain Incidental examiner cues can influence test results
as well as what patients say when interviewed. In each
such instance the patients or subjects failed to recognize
that their behavior had been reinforced or influenced in
any way (18,31,39*61).
Other, more subtle, aspects of the interpersonal
y
relationship may also be influential in determining a
subject's or patient's responsiveness to cues from the
therapist. In psychiatric clinics one occasionally hears
the psychiatric social worker comment that a new patient
has complained that he would rather be In treatment with
a psychologist or psychiatrist. Psychologists also report
the request of an occasional patient that he be seen by
a psychiatrist. While these professional workers express
the opinion that such comments regarding implied status
differences may be only transient, they are in agreement
that the professional affiliation of a therapist might
have some effect on what occurs in psychotherapy— at least
in the initial stages, since after the establishment of
rapport or a "working relationship" in therapy, the par
ticular therapist's profession plays little or no part in
the treatment of his patient. These observations tend to
support the suggestions of Dally (12), Cohen (ll), and the
recent finding by Marion (33)^ to the effect that role or
status differences can be an important factor in influenc
ing the effectiveness of reinforcement in verbal learning
experiments. Moreover, the importance of the degree of
rapport which an experimenter is able to gain with his
subject in a relatively short period is also noted by
Marion. In his study this ability to establish rapport
was related to the experience and "expertness" of the dif
ferent experimenters he used. Such a result also lends
support to Fiedler's findings that experienced therapists
of different schools, as contrasted with inexperienced
therapists of the same school, are very similar to one
another despite theoretical differences (l4).
In his 1957 review of the status of psychotherapy
Winder (63) ventured the opinion that verbal learning
without awareness does occur during the process of therapy,
and that it may then be influential in effecting other
behavioral changes. Dollard and Miller (13) become quite
specific in this regard when they state that n. . . re
inforcement has a direct, unconscious effect ... ."in
the therapeutic process. Moreover, they insist that this
assumption is essential to their theory of personality
and psychotherapy. This suggests that a patient in therapy
might be very sensitive to what his therapist wishes of
him, and thus highly motivated to respond in the desired
fashion. In other words, compared with non-patients, he
might be a readily conditionable subject— especially, we
might add, if his own therapist were to act as experi
menter.
On the other hand, the most frequently described
experimental technic in studies of verbal learning without
awareness employs the vocalized "mm-hmm" as a generalized
secondary reinforcement for the pronouns 1 and We. If we
wish to follow this precedent we must recognize the pos
sibility that this kind of conditioning might not occur as
readily under conditions affording a reduced sensitivity
to the kind of unconscious learning under consideration
in these studies. For example, patients in treatment
might have been so previously "conditioned to talk about
themselves" by the process of therapy that further
10
conditioning of such pronouns as 1^ and We might not be pos
sible.
It thus seems likely that "normal” subjects and
psychiatric patients in treatment will respond differently
to verbal reinforcement in a verbal learning experiment.
It is also likely that the professional affiliation of the
experimenter might be a determinant of the manner in which
subjects and patients respond to verbal conditioning.
The Purpose of the Study and Hypotheses
The present study is concerned with the manner in
which verbal learning without awareness occurs in two
groups of subjects: a group of clinic patients undergoing
psychotherapy and a group of "normal" college students.
It is hypothesized that these groups will differ from one
another In their responsiveness to the verbal reinforce
ments employed. It is further hypothesized that the pro
fessional affiliation of the experimenter will influence
the degree or manner of response to verbal reinforcement
of these two groups of subjects. In explicitly elaborated
form the more detailed implications of these views give
rise to the following hypotheses:
Hypothesis A: Patients serving as subjects, as com
pared with student subjects, will respond to
verbal reinforcement to a greater degree or In
a different manner.
Prediction 1: The means and trends of the posi
tively reinforced response classes chosen for
this study (henceforth referred to as the cri
terion variables) will be significantly greater
for the total group of patient subjects as
compared with the total group of student sub
jects.
Prediction 2: There will be a significant dif
ference between the means and trends of the
criterion variables of the patient group and
the student group irrespective of verbal
reinforcement.
Hypothesis B: Patient and student subjects will show
a difference in degree or manner of response to
verbal reinforcement depending on the professional
affiliation of the experimenter.
Prediction 1: The means and trends of the posi
tively reinforced criterion variables in the
groups of patients with experimenters of three
different professional affiliations will dif
fer significantly from one another.
Prediction 2: The means and trends of the posi
tively reinforced criterion variables in the
groups of students with experimenters of three
12
different professional affiliations will dif
fer significantly from one another.
Hypothesis C: Hie subjects of both groups will be in
fluenced by the professional affiliation of the
expermenter in their use of the criterion vari
ables.
Prediction 1: There will be significant differ
ences among the means and trends of the cri
terion variables in the combined groups of all
subjects with experimenters of three different
professional affiliations irrespective of
verbal reinforcement.
Prediction 2: There will be significant differ
ences among the means and trends of the cri
terion variables in the groups of patients
with experimenters of three different profes
sional affiliations irrespective of verbal
reinforc ement.
Prediction 3: There will be significant differ
ences among the means and trends of the cri
terion variables in the groups of students
with experimenters of three different profes
sional affiliations irrespective of verbal
reinforcement.
Hypothesis D: Assuming the universality of verbal re-
inforcement* individual differences among the ex
perimenters will have no differential effect on
the subjects’ use of the criterion variables.
Prediction 1: There will be no significant differ
ences among the means of the criterion vari
ables of individual experimenter's groups
irrespective of reinforcement.
Prediction 2: There will be no significant dif
ferences among the means of the reinforced
criterion variables of individual experi
menter' s groups.
Prediction 3• There will be no significant dif
ferences between the means of the criterion
variables of the two individual experimenter's
groups within each professional affiliation
irrespective of reinforcement.
Prediction 4: There will be no significant dif
ferences between the means of the reinforced
criterion variables of the two individual
experimenter’s groups within each professional
affiliation.
CHAPTER IX
REVIEW OF THE LITERATURE
In 1937 Thorndike and Rock (56) published one of the
earliest studies of what they described as "learning without
awareness of what is being learned or intent to learn it."
Since then there has been a growing body of literature con
cerned with the problem as it bears upon the study of inter
personal relationships. In this regard, Dollard and Miller
(13) in 1950 related the results of an early verbal learn
ing experiment, published later by Greenspoon (17)# to the
mechanics of psychotherapy. Their statement, which was in
line with other theoretical formulations at that time (35)#
led many experimenters to apply the technics of verbal
learning to Investigations of the clinical interaction. In
reviewing the history of experimental studies of verbal
learning without awareness, one can observe a general trend
toward more precise investigations of the interpersonal in
teraction both In psychological testing and in psycho
therapy .
In the Initial study by Thorndike and Rock (56)
the authors used 30 student subjects to whom they presented
a list of 320 words with the Instructions to respond with
the first word that came to mind. They positively rein
forced all responses of a sequential nature (e.g., "go—
14
15
home," "dinner— time") by saying "right" following the
response, and negatively reinforced all responses of defi
nition (e.g., "sit— recline," "eat— ingest") by saying
"wrong." They report that the sequential responses in
creased froma mean for all subjects of 30$ in the first
block of 40 words to 60$ in the last 40 words. They main
tained that they had demonstrated a lack of awareness in
their subjects because the mean learning curve for the 23
subjects who learned was gradual in slope, while the one
subject who became aware of the relationship between his
responses and the reinforcement, quickly succeeded in
supplying 100$ "positive" responses. The latter percent
age was then employed as the criterion for learning with
awareness.
Apparently there was no questioning of the subjects
regarding awareness, nor was any provision made for deter
mining partial awareness or the possibility of the sub
jects having formed what Adams (l) calls a "correlated
hypothesis" about what they were "supposed to do." In
other words, instead of evolving the more abstract notion
that they were supposed to respond with sequential words
and not with words involving meaning or definition, the
subjects (or a substantial number of them) might have de
cided that the experimenters wanted them to give as many
"personal" responses as possible. For Adams an "hypothe-
16
sis" of this kind might be "correlated" with the actual
purpose of the study in that it could well have produced
the obtained results.
At any rate, Irwin et al. (24), employing a cri
terion of "100$ correct responses," were unable to confirm
the finding of Thorndike and Rock regarding lack of aware
ness. In their repetition of the experiment the subjects
were informed of the distinction between "right" and
"wrong" responses at the outset. Despite this, the mean
learning curve obtained conformed to the one reported by
Thorndike and Rock. Since they failed to mention having
questioned their subjects regarding the subjects' under
standing of the purpose of the experiment, it is not pos
sible to know whether the subjects in these two experiments
viewed the task in the same manner. It is possible that
the subjects of Thorndike and Rook, not knowing what was
wanted of them, had to guess at the purpose of the experi
ment. Such guessing might have involved "unconscious"
learning of the particular abstract concept which the ex
perimenters were reinforcing. Moreover, if some correlated
hypotheses of their own were involved, then it might be
said they were actually conditioned without awareness.
Likewise, In the second experiment, even though the sub
jects had been informed of the nature of the "correct"
responses, they might have decided (in a manner similar
17
to subjects in an experiment to be reported below [Ball
(4-)]) not to be influenced by the reinforcement. In spite
of such a possibility, they might still have been rein
forced and responded in the predicted direction without
awareness of this fact. This is not to say that the sub
jects would not have been aware of either the reinforce
ment or their use of the "correct" response, but rather
that they might not have been aware of an increase in such
responses. Only thorough questioning regarding what they
thought about the purpose of the experiment, how many of
each type of response they felt they had given, and whether
they felt the reinforcements had had an effect on their
behavior could have answered these questions.
The factor of awareness was also investigated by
Postman and Jarret (4l). They repeated the Thorndike-Rock
experiment with certain modifications. After each block
of 20 trials half of the subjects were merely questioned
in order to learn whether they had discovered the basis
for calling some responses right and others wrong. The
remaining subjects had this explained to them at the end
of the first block of 20 trials. The majority of subjects
in the first group (51 of 60) succeeded in discovering the
key to right and wrong responses in the course of the ex
periment. Both groups showed significant learning, but to
a very slight degree before verbalization of the principle.
18
The nine subjects who failed to discover the key showed no
increase in the positively reinforced responses over chance
expectancy. These authors would thus seem to have demon
strated that some degree of awareness of goals to be
reached is necessary for progress in attainment of such
goals.
Some years later the role of awareness in verbal
learning was investigated by means of a different pro
cedure. Philbrick and Postman (40) presented their sub
jects with a sequence of words and told them in advance of
the experiment that each word had been paired with a number
which they were required to guess. After each guess the
experimenter would respond with "right" or "wrong."
Moreover, since after each block of nine trials in which
a subject had four "rights" he was asked for the prin
ciple, most subjects assumed they were being given a cue.
The principle was the number of letters in the word minus
1. The twenty subjects who had formulated the principle
at some point in the experiment showed evidence of its
operation even before its explicit formulation by supply
ing an average of 42$ correct responses in the block of
nine words preceding the one which gave rise to verbaliza
tion of the principle. (The expected chance percentage
was 11.) There is no report as to whether the subjects
were actually aware before they reported awareness two
blocks later. Nor can one rule out the possibility that
there might have been an anticipated ' ‘unconscious" or
partial awareness (which could also have been true for the
Postman-Jarret study reported above). There is also the
possibility that the subjects might first have formed some
other hypothesis concerning the principle which could have
produced these results. The 28 subjects in this experi
ment who did not verbalize awareness also showed a smaller
but significant degree of learning. According to the
authors, these subjects may have been influenced by an
hypothesis related to word-length or by some other cor
related hypothesis.
In Greenspoon's (17) experiment the right-wrong
dichotomy was less obvious than in the preceding studies.
He used 5 groups of undergraduate students as subjects.
Every member of each of these groups was instructed to
keep saying words at random for fifty minutes. For the
first twenty-five minutes the experimenter remained quiet
but counted all plural nouns emitted by the subjects. In
this way he obtained a measure of the operant level of
plural noun response. Following this period, in one group
the experimenter reinforced all plural nouns with a
positive reinforcement of “mm-hmm." In the second group
he reinforced plural nouns with a negative reinforcement
consisting of "huh-uh.1 1 The same procedure was followed
for non-plural-noun words in the third and fourth groups,
while the fifth group served as a control. As predicted,
Greenspoon found that "mm-hmm" increased and "huh-uh" de
creased the proportion of plural nouns. However, both
positive and negative reinforcers tended to increase non-
plural-noun words (such as adjectives, pronouns, verbs,
singular nouns, etc.). As a consequence Greenspoon con
cluded that a reinforcement will have either positive or
negative effect depending on the class of words selected
for reinforcement. Moreover, "mm-hmm" could be inferred
to function as a reinforcing factor since all responses
which evoked it tended to occur with greater frequency.
Ten of the subjects made spontaneous mention of
their awareness of the relationship between the response
and reinforcement: nine mentioned the association of
"huh-uh" with plural nouns, and one referred to that of
"mm-hmm" with plural nouns. Such differential awareness
of positive and negative stimuli suggests a possibility of
greater awareness of signs of social disapproval as con
trasted with signs of social approval. If this be so,
then the negatively reinforced subjects might have experi
enced greater readiness to shift mental sets in the search
for a socially approved solution. However, since there are
more classes of words which elicited the negative rein
forcement, such reinforcement of non-plural-noun words
21
(e.g., singular nouns, adjectives, adverbs, prepositions,
etc.) could have resulted in more disturbing negative in
fluences and thus the subjects would have been faced with
a more difficult search. For example, a subject might have
responded for a period with fewer adjectives but many more
verbs, at another time with fewer pronouns but more singu
lar nouns, each time finding that he was incorrect, and
finally responding with significantly more non-plural-noun
words despite their tendency to provoke an "huh-uh" from
the experimenter. In other words, it is possible that if
the experimenter had chosen another single class of words
instead of all other classes, the results might not have
supported Greenspoon's conclusion regarding the response
class as a determinant of reinforcement Influence.
Greenspoon repeated his experiment later by sub
stituting the mechanical stimulus of a red light and a
tone in place of the vocalizations employed in the earlier
experiment. These non-vocal stimuli were found to have
reinforcing effects on both plural nouns and other words
(16). Except under laboratory conditions such artificial
stimuli would not be likely to signify social approval or
disapproval. The very fact that the study was done in a
university laboratory with students as subjects renders
it possible for the latter to entertain the implicit
expectation of being required to learn or solve something.
22
In such a situation a strange light or tone may come to
have meaning and thus acquire the properties of a rein
forcing stimulus. Later studies, to be described below,
showed that such mechanical stimuli are not reinforcing
when presented in a clinical as opposed to a laboratory
setting, or when introduced in a more “meaningful" con
text than that of the random elicitation of words called
for by Greenspoon's instructions.
.In order to make a direct investigation of the
influence of awareness on verbal learning Sidowski (49)
devised a variation of the Greenspoon experiment. He
Instructed half of his experimental subjects to say
words at random with the explicit purpose of making a
light bulb blink as many times as they could. Like
Greenspoon, Sidowski chose to reinforce plural nouns.
To his other experimental group he gave no such instruc
tions. After a number of responses each subject was
asked if he was aware of the purpose of the experiment,
and the purpose of the light. Sidowski discovered that
there was only a slight (although significant) difference
In the group means between the "task" and "non-task"
experimental groups, the former being greater. For both
groups the reinforced response occurred more frequently
than in the control groups. These latter consisted of
23
"task" and "non-task" groups, neither of which was rein
forced. Incidentally, Sidowski failed to comment on the
behavior of the subjects in the "task" control group
when frustrated by futile efforts to get the light to
blink at all.
Three more studies using variations of the Green
spoon technic extended the application of this type of
experiment and gave rise to more questions about aware
ness. Wilson and Verplanck (62) reinforced nouns and
adverbs with a combination of "good," "mm-hmm,1 1 and note-
taking. They reported significant results in twenty-
nine of their thirty subjects, most of whom noticed some
part of the reinforcement, but evidenced no difference
in their degree of learning from "unaware" subjects.
These could be classified as "partially aware" subjects
as no correct relation between the reinforcements and
the criterion variables was verbalized. What other
hypotheses the subjects might have formulated is not
mentioned. There were no reported differences among
the results of the fifteen student-experimenters these
authors used.
Mandler and Kaplan (32), using the Greenspoon pro
cedure, found that their total group of twenty-eight subjects
did not reinforce. However, by separating the subjects into
24
two groups composed of those who thought "mm-hmm" was en
couragement and thoBe who thought it signified disapproval,
it was found that the "encouraged" group increased their
level of response over the operant level, while for the
"disapproved" group the level was decreased.
In a third experiment Tatz (55) required that his
subjects verbalize three-digit numbers which were rein
forced by "good" or the click of an electric counter
whenever the subject included one of two previously-
determined "correct" digits. Two groups were informed
that a correct solution was required and two other groups
were not. An interview to determine awareness followed
the experiment. Both "task" groups learned to a signifi
cant degree as did the "good" "non-task" group though
not as much. The "non-task" group which had been reinforced
by the click of the electric counter decreased their use
of the reinforced response. Questioning revealed a range
of awareness from none at all to complete recognition of
the principle involved. Where partial solutions were
formed by the subjects in either the informed or non
informed groups a greater degree of learning occurred.
Moreover, letting the subjects know there was a problem to
be solved seemed to promote the formulation of both par
tial and correct solutions.
The preceding studies were not directly concerned
25
with any clinical implications of the awareness factor.
They all employed students rather than patients as subjects.
A shift in the direction of a clinical orientation was
introduced by Cohen et al. (11). Their subjects were
ambulatory patients from a general hospital all under 55
years of age and free from a history of psychiatric or
neurological difficulty. They were divided into an ex
perimental and control group and informed that they were
participating in a research program conducted by one they
regarded as a "doctor."
Each subject sat facing the experimenter who pre
sented him, one at a time, with eighty 3 by 5 inch cards
on which was typed a common verb in the simple past tense.
The six pronouns I, We, He, She, You, They were typed in
randomized order below each verb. The subject was in
structed to make up a sentence for each card, using the
verb and beginning the sentence with one of the pronouns.
For the experimental group, after the first twenty presen-
4
tations, during which an operant level was obtained for
the criterion variables, the experimenter said "good" in
a flat, unemotional tone every time the subject began a
sentence with the first-person pronoun "I" or "we." The
experimenter made no response whatsoever during the en
tire eighty trials for the control group.
Following the experimental session the subjects
26
were questioned about their understanding of the objectives
of the study. Not one of the subjects succeeded in speci
fying the nature of these objectives. Nevertheless, a
significant trend (i.e., an increase in the criterion
variables over the four blocks of twenty trials each) is
reported for the experimental group as compared with the
control group.
In a second experiment the authors investigated
the effects of two different types of extinction as con
trasted with continuous reinforcement. The results showed
that with ordinary extinction (i.e., discontinued rein
forcement) the level of reinforced responses remained con
stant throughout the extinction period. They interpreted
this to be an indication of a "marked resistance" to ex
tinction of "I-we." Furthermore, when a competing re
sponse was reinforced during the extinction period the
level of the original ("I-we") responses dropped signifi
cantly; but when the originally reinforced response was
continuously reinforced, it continued at about the same
level as the first group with usual extinction procedures.
In this study, as in previous studies with stu
dents, despite prolonged reinforcements the initially high
level of learning soon tapered off. Furthermore, no
statistically significant differences were found to exist
among the four experimenters to whom the subjects were
randomly assigned. According to the investigators, “good"
was found to be an effective secondary reinforcement not
only with the response class "I-we," but could also be
arbitrarily shifted to another class of responses without
losing its reinforcing value. In their view this lack of
extinction in the normal extinction group might conceivably
have been related to possible reinforcing effects either
of their note-taking of the responses or to the mere
presentation of the cards. It is pertinent to add that
they also called attention to the Implications of this
study for the student of psychotherapy.
A modification of the technic just discussed was
introduced by Taffel (53) in a study of the relationship
between degree of anxiety and level of verbal learning.
He selected hospitalized psychotic and neurotic patients
divided into three groups on the basis of high, medium, or
low levels of anxiety as measured on a scale of manifest
anxiety. The procedure was identical with the previous one
except that the stimulus cards were presented in an ex
posure stand placed before the subject. The visual stimu
lus of a light proved ineffective as a reinforcement in
this experiment, but "good" produced a significant incre
ment in the number of “I-we" responses in both the medium
and high anxiety groups with a significant difference be
tween the two groups and the predicted direction. However,
28
the low anxiety group's responses were not significantly
different from those of a control group.
Subsequent interviewing indicated lack of awareness
of reinforcing factors by all of the subjects. Taffel took
his results to mean that anxiety leads to a heightened de
gree of vigilance with respect to stimulating conditions.
Because there was an initial, though Insignificant, in
crease in all groups including the control he thought this
might reflect the tendency of patients in a "clinical
situation" to talk about themselves. In line with this
notion he considered the ineffectiveness of the light to
be a consequence of the patients' failure to regard it as
a means of communication. Instead they regarded it as a
time device or a means of distraction. In his discussion,
Taffel stated that "the therapist is working on a basic
assumption that a behavioral reaction on his part will
influence or modify the behavior of the patient," and he
believed his results not only justified such an assumption
but also served to show the need for a therapist to be
mindful of all facets of his behavior, not merely those
directly related to his therapeutic objectives.
In a doctoral dissertation study using Taffel's
technic with college students, Dally (12) assumed that a
high anxiety group would condition more than a low anxiety
group, that "good" would be more reinforcing than "mm-hmm,"
and finally that there would be no sex differences between
male and female subjects. None of his predictions was
supported. Each group showed a slight but insignificant
increase in the critical response, with a highly signifi
cant sex difference favoring the male subjects. He at
tributed his failure to confirm the findings of Taffel to
three factors. First, he pretended to tape record the
responses and thus the subjects went as fast as they
wished, perhaps responding with a "time set" instead of
a "response set" with consequent lessening of attention to
the experimenter. Secondly, he used college students in
stead of patients, and this made for an important differ
ence in attitude toward the experiment. Thirdly, the
experimenter was a graduate student instead of a "doctor"
in a hospital and such a status factor might well have
been an influential variable.
Except for the question of sophistication Daily
does not discuss other possible differences between patient
subjects and student subjects, nor possible differences
due to a laboratory setting as contrasted with a clinical
setting. It is conceivable that hospitalized patients
would be highly motivated to behave in accordance with what
they believed the "doctor" expects of them and thus speed
their recovery. It may also be that the kind or content
of the anxiety experienced by the emotionally disturbed,
30
hospitalized patient will differ from that experienced by
the average college student serving as a subject even
though both patient and student make the same extreme score
on some measure of anxiety.
In a study of extinction and negative reinforce
ment, Klein (26) used a heterogeneous group of NP hospital
patients and following the Taffel technic, first reinforced
"I-we" pronouns with "good,” and then divided his subjects
into three extinction groups, all of whom showed a signifi
cant reduction in the critical response at the end of the
first eighty extinction trials. There was a significant
difference between a group with simple discontinuance of
reinforcement, and a group where he replaced "good" with
"not so good," but no significant difference between
either of these groups and another where he had merely
switched the "good" reinforcement to the other four pro
nouns. Of the eighty reinforced subjects only one indi
cated an awareness of what had elicited an indication of
approval. Nothing was reported regarding partial aware
ness, nor of the behavior of subjects negatively rein-,
forced.
In an attempt to investigate experimenter influ
ence In terms of status characteristics and to approximate
a therapy or counselling situation more closely Marlon (33)
used Taffel1s technic In an educational counselling center.
Subjects were students applying for counselling and the
experimenters were staff counselors and student counselor-
trainees. The subjects were asked to participate in a
research project at the time of application for counselling
and were randomly assigned to staff and trainees for ex
perimental and control groups. At the time of the experi
mental session each experimenter introduced himself as
either a staff counselor or trainee and at the end of the
experiment the subject was questioned at length regarding
awareness of relevant conditioning factors. Although
Marlon did hot obtain the reinforcement effect for the
total of all the experimental groups combined, his data
did show a significant difference between the two status
groups as well as significant inter-experimenter differ
ences. In his opinion these differences seemed to be a
function of the clinical experience and skill of each
experimenter. The more assured, skillful, and experienced
the experimenter, the more likely was he to differ sig
nificantly from the others in obtaining positive results
with his subjects.
The rather thorough questioning at the end of the
session revealed only two subjects who had recognized the
relationship between the reinforcement and correct re
sponse, although twenty of the reinforced subjects reported
hearing "good" occasionally. Most of these subjects did
32
not know why the experimenter had said it, though a few
felt it was an encouragement for them to keep going, go
fast, etc. None felt distracted or influenced by it.
Marion attributed the differential influence be
tween his staff counselors and his trainees to be the en
hanced status of the former. However, this may be ques
tioned; for, as he pointed out, there were inter
experimenter differences related to degree of clinical
experience rather than professional status. Since the
staff experimenters were the more experienced clinicians,
they might be expected to act as more effective sources
of reinforcement. Consequently, the question of status
effect remains somewhat unsettled.
Finally there is some question as to the applica
bility of Marion's findings to psychotherapy. He felt he
had duplicated a treatment situation since his subjects
were applicants for vocational counselling. However, being
surrounded by recording devices in an experimental room
might well have induced them to view the experience as more
a matter of laboratory research than a conventional mode
of psychotherapy.
In another study designed to investigate the ef
fects of experimenter characteristics, Binder, McConnell,
and Sjoholm (5) used two experimenters whom they described
as extreme opposites. The female experimenter was slight,
young appearing, soft-spoken; her male counterpart was big,
mature, commanding. Subjects were male and female under
graduate students. The experimenters attempted to rein
force "mildly hostile" words used by the subjects in a
sentence-constructing task by saying "good." The groups
of both experimenters increased the frequency of the re
inforced response significantly until data from those
subjects who verbalized awareness were omitted from the
analysis. Now the young lady proved to be the more ef
fective instrument of reinforcement, the male experi
menter's subjects showing no learning at allI According
to the authors, this result may possibly be attributable
to the impact of the less 1 1 threatening" personality of the
woman as compared with the more dominant personality of
the man. One is less inclined to express hostility in the
presence of an authoritarian or commanding person.
Possible differences in degree of learning as a
function of the sex of the subjects does not appear to have
been well controlled. The authors did note that the male
experimenter had (by chance) a larger number of female
subjects than the female experimenter. Thus, in order to
test for the possibility of different rates of learning
for male and female subjects, they compared the results
of all female subjects with those of all male subjects
irrespective of the experimenter. They obtained no
significant difference between the two groups. However,
an important and possibly significant analysis of the data
should have been made comparing the interaction of the
male and female experimenters with the two types of sub
jects. As reported by Daily (12), a very significant sex
difference occurred in his experiment when a male experi
menter reinforced female subjects. Moreover, his female
subjects did not increase the number of critical responses
over the operant level— a situation akin to the experiment
by Binder et al., and offering a possible explanation for
their results.
Three groups of patients in a neuropsychiatric
hospital were used by Sarason (48) in a study of verbal
learning. One group consisted of defensive, hostile
patients, another group was made up of compliant, dependent
patients, and a third group consisted of anxious patients.
Sarason predicted that the latter two groups of subjects
would condition more readily than the former. He based
this on the assumption that defensive, hostile subjects
would be resistant in the experimental situation just as
they are when involved in psychotherapy. The predictions
were confirmed by the experimental results and, in Sara
son1 s opinion, this outcome is congruent with the belief
of those who view psychotherapy as a form of verbal
learning.
35
A study concerned with the influence of dependency
on verbal learning without awareness was made by Caruth
(9). She predicted that the dependent person will condi
tion more readily than the independent person. Following
the technic described by Cohen et al. (ll), she chose de
pendent and independent subjects from psychiatric out
patient clinic patients and "normals." However, she failed
to obtain any reinforcement effects whatsoever, and con
cluded that verbal learning without awareness is a less
stable phenomenon than would be expected from published
reports. Since she had used female subjects, however, she
also concluded that, as Daily (12) had found, women are
possibly not as conditionable as men. In the absence of
a male control group with which to compare her female sub
jects this conjecture could not be tested. It is quite
possible that in both Caruth’s and Daily's experiments the
crucial factor was one related to experimenter character
istics rather than to the sex of the subjects.
By means of the Cohen technic Leventhal (29) in
vestigated the effects of reward, punishment, and a com
bination of the two upon the verbal learning of three dif
ferent groups of subjects: hospitalized schizophrenics,
neurotics, and "normals" with medical disorders. The sub
jects were divided into three experimental groups and one
control group. He reinforced one group with reward
("good"), another* with punishment ("not so good"), and a
third with both reward and punishment ("good" for "I-we"
responses, "not so good" for "he-they”). He found that
all three types of subjects learned with the combination
of reward and punishment. The normal groups also learned
with reward separately and with punishment separately. The
neurotics did not learn with punishment alone while the
schizophrenics did not learn with reward alone.
Consideration of these results could lead us to
conclude first that normal people respond in a normally
sensitive manner to cues of both reward and punishment
whether presented together or separately. Second,
neurotics seem to be very sensitive to signs of approval
and reward, but apparently selectively inattend to pain or
punishment. This is not inconsistent with such neurotic
tendencies as the repression of anxiety and the denial of
painful or anxiety-provoking stimuli. Schizophrenics, on
the other hand, are often ready to perceive the world as
a hostile, punishing place, and thus are sensitive to
cues which fit in with this perception. Furthermore, they
often refuse to acknowledge or react positively to reward,
interpreting it as a seduction likely to render them vul
nerable to further pain.
While such an interpretation of Leventhal's results
appears to be congruent with theories of psychopathology,
his findings are not in line with the earlier work of
Taffel (53) in which schizophrenics did respond to positive
reinforcement in verbal learning experiments. On the other
hand, Cohen and Cohen (10) in a recent study did find a
significant difference between the degree to which hos
pitalized neurotics and schizophrenics responded to a
positive reinforcement. In line with Leventhal's findings
their neurotic group learned significantly more than their
schizophrenics. These somewhat contradictory findings
might be attributable to a variety of factors such as dif
ferences among the psychotic subjects with respect to de
gree and kind of psychopathology, understanding of the
instructions, and willingness to cooperate. Nor should the
possible influence of personality differences among the
experimenters be overlooked as a relevant factor.
The last-mentioned factor was investigated by
Sapolski (46) who conducted two experiments using several
experimenters of different personality characteristics.
One group of subjects was chosen on the basis of test
scores which indicated that they would be "attracted to"
or "compatible with" their experimenters. A second group
was formed on the basis of "unattraction" and "Incompati
bility."
The results indicated that when the experimenter
and subject were "attracted" and "compatible" a high
38
degree of verbal learning occurred. Moreover, the level
of response remained high during an extinction period with
the experimenter absent (having been called out of the
room by a pre-planned "phone call"). However, when a
relationship of "unattraction" and "incompatibility" pre
vailed, no apparent learning took place during the rein
forcement period; but with the experimenter absent during
extinction, level of responses rose rapidly to the same
level as that of the "compatible group."
The author concluded that, although learning takes
place, the effects are delayed when the subject is in the
presence of a disliked experimenter. As applied to the
therapy situation he suggested the possibility of inter
ference with the development of insight when a negative
transference is present. He also stated that psycho
therapeutic effects might be manifested more completely
following a time lapse after termination of treatment.
This may be true, but how this follows from this study is
not altogether clear.
In a study by Verplanck (58), the influence of a
social setting on verbal learning without awareness was
investigated. To do this he had seventeen advanced psy
chology students reinforce personal opinion statements of
friends, relatives, room-mates, etc., over a half hour
period. They used agreement, smiling, or paraphrasing the
subjects' statements as positive reinforcements. The ex
periment took place in the subjects' homes, dormitory
rooms, or various places on the university campus. They
were not told that they were participating in an experiment
and note-taking was kept to a minimum. The subjects served
as their own controls. There was a significant increase
over an operant level of the selected response class while
other classes of statements decreased in frequency during
the period of the experiment. No subject verbalized any
awareness of the procedure nor anything else unusual in
the experimenter's behavior during the experiment. Indeed
the technic was so effective that Verplanck reports that
a colleague successfully used him as a subject without his
knowing itl The implications of these results in terms of
social control are obviously rather momentous.
In another investigation with a purpose quite dif
ferent from that of Verplanck, namely to exclude as many
extraneous factors of the interpersonal interaction as
possible, Hildum and Brown (22) conducted an experiment by
telephone. On the pretext that they were conducting an
educational "survey," the authors reinforced "opinion re
sponses" of their subjects with either "good" or "mm-hmm."
In this study the "good" group conditioned whereas the
"mm-hmm" group did not. The authors explained this dif
ference on the basis that theirs was a "natural interview"
40
situation, similar to a clinical interaction. They stated
that in the experimental setting the subject is searching
for clues and thus will respond to almost any stimulus, in
cluding "mm-hmm." In their "natural interview," however,
they felt that "good," being out of character for an
interviewer, produced learning, while the more neutral
"mm-hmm" did not. They concluded that interviewer bias
might influence survey results and they also suggested
that theoretical bias on the part of the therapist might
also have significant effects on psychotherapy.
If Hildum and Brown are correct in maintaining
that learning took place as a result of an "out of charac
ter" stimulus from the interviewer, it might be expected
that the subjects would be more aware of this than of "mm-
hmm." The authors report no inquiry to determine aware
ness, however, nor is the possibility discussed that the
subjects might not even have heard the telephoned "mm-hmm"
as clearly as "good." Finally, there is no discussion or
examination of the possible differences between their re
sults and what might be obtained in a face-to-face "natural
interview" situation.
A deliberate effort to simulate a clinical inter
view situation was made by Ball (4). This is clear from
the wording of his essay: "Reinforcement conditioning of
verbal behavior by verbal and non-verbal stimuli in a
41
situation resembling a clinical interview.” The experi
mental task was to make up stories including the word
classes, "man, woman, and animal.” "Animal” was reinforced
in each of four groups by means of "mm-hmm," "huh-uh," a
light, and a buzzer. A fifth group served as a control
with no reinforcement.
A significant increase in the use of the reinforced
response was found with the positive verbal reinforcement
and there was a tendency towards an increase with the
light and buzzer. There was no noticeable difference be
tween the "huh-uh" group and the control group. The author
noted that his subjects tended to reinforce rapidly and
maintain the level attained. A similar observation was
made by Greenspoon and Cohen in their experiments.
Commenting on his "clinical observations," Ball
reports that the "huh-uh" subjects tended to become "rest
less, aloof, confused, blocked, resistant, and even
belligerent" as the experiment progressed. The light,
buzzer, and control groups were all similar in behavior
to the "mm-hmm" group— that is, "cooperative, interested,
friendly," etc. The disturbed behavior of the "huh-uh"
subjects in this study lends support to our previous com
ments regarding the results of Greenspoon1s (17) experiment
with respect to "huh-uh" as a painful or unpleasant
stimulus.
42
Concerning awareness, Ball reports that no subject
in any group correctly verbalized the purpose of the ex
periment. In the "huh-uh" group, however, all subjects
recognized the relationship between the correct response
and the reinforcement though they did not understand the
experimenter's purpose in saying it. Moreover, they did
not feel that their behavior had been affected in any way.
Several subjects in the light group verbalized the correct
relationship and stated that they had tried to avoid the
use of "animal." Despite this intention they nevertheless
felt a tendency to use the word more often.
It thus seems that even with awareness of the cor
rect relationship between stimulus and response along with
a deliberate resolve to avoid using the reinforced re
sponse, such resolution is not always effective. It is
as If associations which the subject intends to avoid are
nevertheless established. In the very process of deciding
not to think of an animal when the light is flashed the
tabooed association tends to be reinforced. This is of
considerable importance for understanding the results of
the present study as well as the results of the previously
discussed study by Irwin et al. (24). Elaboration of this
point will be taken up in Chapter V.
Investigation of verbal learning by Nuthmann (39)
revealed that a significant amount of learning occurred in
the responses of subjects to the items of a personality
test of "self-acceptance" specifically devised for her
experiment. Those subjects who had made few "self-
acceptance" responses on a pre-test greatly increased
their scores in response to the positive verbal stimulus
of "good," but failed to respond significantly to a light
blink. The investigation took place in a laboratory set
ting and the subjects were even required to register their
responses by manipulating levers. The artificiality of
such a setting renders application of the findings to the
customary clinical interview situation somewhat question
able.
The effect on test results of an examiner's inci
dental behavior was studied by Wickes (6l). He had two
different examiners, working with separate groups of volun
teer student subjects, administer a test somewhat like
the Rorschach. The results showed that a combination of
"good, fine, and all right" was equally as reinforcing as
"nodding the head, smiling, and leaning forward in one's
chair." The results were substantially the same for both
examiners and hence are not attributable to chance idio
syncratic behavior of either examiner.
In another study Gross (l8) used the^ standard set
of Rorschach ink blots and reinforced human and human
detail responses by means of "good" and a nod. He used
two groups of normal adult subjects in a general hospital
setting. The "good" group conditioned while the "nod"
group was no different from a group of control subjects.
However, since the nods were not accompanied by smiles and
postural changes, as in the Wickes experiment, his results
are not necessarily at variance with those reported by
Wickes. At all events both he and Wickes agree that
minimal and seemingly unimportant incidental behavior of
an examiner may have a bearing on test results.
An investigation of the verbal learning technic
with reference to some aspects of psychotherapy was under
taken by Krasner (27). In an experiment designed to re
semble thematic apperception testing procedures he had two
hospitalized schizophrenics tell stories containing a
mother, father, child, and an animal. In two of the five
blocks of experimental sessions he reinforced the response
class "mother" with "mm-hmm" and a simultaneous nod and
smile. His subjects had been told that they were partici
pating in a research project. Moreover, a timer was
clearly visible to the subjects. It may be that this re
search or laboratory atmosphere militated against the cre
ation of what some would regard as a genuine "clinical"
setting. Be this as it may, the results showed a signifi
cant variation in the number of critical responses between
his reinforcement and non-reinforcement sessions. The
45
effect of the reinforcement appears not to have carried
over from sessions of one day to those of another. This
might be of importance in accounting for the schizo
phrenic’ s common failure to respond to psychotherapy
separated by more than 24-hour intervals. Wexler (59) has
stated that psychotherapy with schizophrenics is most ef
fective if on a daily basis, including Saturday and Sunday.
He also reports that he even takes his schizophrenic
patients with him on his vacations.
More recently experimenters using the technics of
verbal learning without awareness have attempted to create
situations closely resembling personal interviews or psy
chotherapy sessions. Two such studies are reported by.
Salzinger and Pisoni (44,45). One study involved hospital
ized schizophrenic patients and the other a group of
"normal" general hospital patients. In both studies the
same experimental procedure was followed; an interview
concerning the patient's reasons for being in the hospital,
what led to his illness, what his family thinks about it,
etc., was conducted by two experimenters, one male and one
female. After an operant level of "self-reference affect"
responses (e.g., "I was so glad . . "they made me so
angry . . .") was obtained during the first ten minutes of
the Interview, the experimenters then selectively rein
forced such statements for ten minutes with "mm-hmm," "I
see," "yeah." Both schizophrenics and normals responded
with a significantly greater number of the reinforced
statements. During an extinction period of ten minutes,
however, the schizophrenic group was much less resistant
to extinction. This is consistent with Krasner's findings
noted above. The authors believe this has implications
for psychotherapy with schizophrenic patients, explaining
that such patients will probably need many more reinforce
ments before any lasting effects of the treatment will take
place. We might add that more closely spaced "reinforce
ment" sessions might also be desirable. Psychotherapy with
schizophrenics is known to be more time consuming and to
require more active therapeutic intervention than with
other patients (j,&315*59).
In an attempt to demonstrate the reinforcing
properties of "attention" (merely looking at and attending
to the subject) as well as "assent” (i.e., "mm-hmm"),
Adams and Hoffman (2) asked college students to speak at
will for fifty minutes about educational practices. For
the experimental group the experimenter looked down stead
fastly at his notebook for the first ten minutes. At the
end of this time he reinforced all self-reference state
ments made by murmuring "mm-hmm" (assent) and looking up
at the subject (attention) for a period of ten minutes.
Following this period of reinforcement the experimenter
47
returned his gaze to his notebook for thirty minutes.
For the control group the experimenter merely kept
his gaze fixed on his notebook for the entire fifty
minutes. The experimental subjects conditioned and the
controls did not. Incidentally, there were extinction "ef
fects" following the period of reinforcement in the experi
mental group. These effects consisted of a drop-off not
only in the criterion response, but in all other responses.
The subjects also displayed an apparent feeling of emo
tional disturbance, anger, and hostility, and in some
instances the subject "left the field," by refusing to co
operate any further. Such behavior is not unlike the rest
lessness and irritability of the subjects in the negative
reinforcement ("huh-uh") group in Ball's study (4).
While the authors report that the subjects ex
pressed neither an awareness of the reinforcement nor of
any change in their behavior, they conclude that a lack of
reinforcement after having been reinforced is perceived by
the subjects as punitive and disapproving. The subject's
behavior as a result of the experimenter's silence in this
study Is reminiscent of Menninger's description of a
patient's reaction of frustration and anger to the
analyst's silence in the therapy situation (37).
That the authors have demonstrated the reinforcing
effect of attention is open to question. This variable was
48
not isolated, but instead was presented along with "assent"
and the latter, as shown by Greenspoon (17) is an effective
reinforcer.
Two recent studies dealt with^ verbal learning and
the therapeutic interaction in a more direct manner than in
previously reported experiments. One of these was con
ducted by May (34) and the other by Rogers (43). May (34)
divided introductory psychology students into two groups
and they were asked to talk at random about early memories
in a face-to-face "therapy like" interview. In one group
he reinforced any memory having to do with family events
by saying "mm-hmm." In the other group he reinforced any
non-family memories. Both groups showed a significant
increase over an operant level of the reinforced memories.
While this study certainly has elements resembling
the therapeutic interview, the investigator used students
and recording devices, and explained to his subjects that
it was his dissertation research. Consequently, it might
be that the preceding was not as "emotionally charged" and
typical of the therapy productions as the author claims.
In the second quite similar experiment Rogers (43)
reinforced self reference statements of undergraduates
over a period of several interviews with "mm-hmm" and a
nod. The subjects were divided Into a "positive self
reference" group, a "negative self-reference" group, and
a control group, and the members of each group were in
structed to "talk about themselves." The data showed an
increase in the number of reinforced responses for the
"negative self-reference" group. This was significantly
greater than their own operant level and the control
group, but this was not true for the "positive self
reference" group. The author also reported that the num
ber of positive self-reference statements decreased for
both the "negative" and the control groups. Although the
"positive" group did not increase their number of rein
forced responses significantly over their operant level,
Rogers argued that because the other two groups showed a
decrease in positive self-reference.statements, this was
indirect evidence for learning in the positive group.
A battery of tests measuring anxiety, adjustment,
and self-concept was administered both before and after the
Interview. None of these showed any change. On the basis
of the results he obtained in this "quasi-therapy" set
ting the author concludes that Winder (63) is probably
correct in his assertion that "the patient learns to talk
differently and little else." Rogers questions the gener
ality of such verbal changes to behavior beyond the
therapy interview and points to the lack of change in his
subjects as measured by his battery of personality tests
as evidence for this position.
50
On the other hand, he admits that psychotherapy
may be a helpful experience for those who seek it. He even
expresses the "heretical notion . . . that a form of psy
chotherapy might be devised which did not require the
'therapist' to acquire clinical knowledge and training but
merely to learn how to administer simple reinforcements.
11
• * •
Rogers seems to view his experimental task as al
most identical with that characteristic of a treatment
situation. Before endorsing this view one ought to check
on the motivation of those serving as subjects in the ex
periment. Motives and feelings of students participating
in "personality research" are probably quite different from
persons seeking help for personal problems. Do students
perceive their roles in such research as if they were
patients seeking treatment? Are they likely to regard the
experimenter as a psychotherapist? Is the physical setting
like that of the therapist's consultation room, or like
that of a university laboratory? We suggest that while
Rogers' experiment, as well as May's, more closely re
sembles the therapy situation than any previously reported
studies, caution must always be observed in generalizing
beyond the limits of one's results and experimental set
ting.
51
Summary
The studies reviewed in this chapter have dealt
with various aspects of examiner behavior on the outcome
of psychological test results. With specific reference .
to the clinical situation many of the studies revealed a
relationship between a psychotherapist's incidental be
havior and the patient's unconscious or subconscious re
sponsiveness to such behavior. The behavior in question
ranged from having the therapist simply murmur "mm-hmm" to
having him smile, nod, and/or lean forward in his chair
(2,4,11,18,22,27,28, etc.).
Many of the reported studies, it will be recalled,
were more like laboratory investigations rather than
duplications of genuine clinical settings (4,10,33,34,39,
43). Even when patients served as subjects in these ex
periments, they were usually hospitalized psychotics and
were informed that the study was a research project or an
experiment (26,27,29, etc.).
In those studies which attempted to approximate a
clinical setting, whether in terms of task, structuring to
the subjects, or surroundings, student subjects were used
(2,4,32,34,37,43,59), except in two instances (44,45). In
one of these, however, the subjects were "normal" general
hospital patients (45), and in the other they were psy
chotics in an NP hospital (44). In no instance were the
52
subjects currently undergoing psychotherapy. In brief,
none of the investigations dealt with typical psychiatric
out-patients in treatment. The possible implications of
verbal learning without awareness in the course of such
treatment obviously merits separate and explicit investi
gation.
The need for this sort of investigation is clearly
stated in the following quotation from an article by
Krasner (28); "Would the reinforcers or cues (used in
verbal learning experiments) be effective in a situation
labeled 'psychotherapy,' in which the S felt that there
was a therapeutic relationship involved? It would be
reasonable to assume that in such a situation the subject-
patient would be all the more alert to look for cues for
guidance as to what the examiner wants, just as the
patient does in the therapy situation."
As will be discussed in the next chapter, the
present study attempts to investigate the factors involved
In such a situation.
CHAPTER III
METHOD
Experimenters
The experimenters for the present study were all
members of the staff of the Los Angeles State Mental
Hygiene Clinic. They were selected on the basis of pro
fessional affiliation, length of clinical experience, sex,
and age. A total of six men served as experimenters:
two psychiatrists, two clinical psychologists, and two
psychiatric social workers. Each experimenter had a
minimum of three years of clinical experience, and all
were over thirty years of age, and of mature appearance.
¥
Subjects
Four groups of subjects were used. They consisted
of an experimental and control group of psychiatric out
patients, and an experimental and control group of "normal"
college students.
Patient-subjects
The patient-subjects consisted of thirty-six fe
male, non-psychotic patients who had been in treatment at
the State Mental Hygiene Clinic for not less than three
months. They were randomly selected by the author from
53
5^
the case load of the six experimenters. Each E was then
provided with a list of those of his patients who would
serve as experimental and control subjects. The mean age
of the group was 31.5 years, and the mean number of years
of education was 12.8. In terms of diagnostic classifica
tion they constituted a heterogeneous group of non-
psychotic clinic patients. Incidentally, female subjects
were used because at this clinic the proportion of male
patients is so low that there were not enough available
for the research.
Student-subj ects
The student subjects were thirty-six female under
graduates from introductory psychology courses at the Los
Angeles City College. They had volunteered to participate
in the "clinic's current effort to develop a new technic
for personality investigation." They were informed that
the "technic" would take about a half hour, and that it
would only involve "talking." This was done in order to
allay any anxiety which potential subjects might feel and
to encourage their participation.
Each experimenter submitted a list of times he
would have available for the experiment during a three
month interval. The author then contacted each subject
and arranged her appointment according to her convenience
55
with whatever particular experimenter might be available
at that time.
The mean age of the student-subjects was 20.9, and
the mean number of years of education was 12.7. The major
ity were in their first or second semester of college.
None of the subjects gave any history of emotional dis
turbance requiring psychological counselling or psychiatric
treatment.
Materials
The present study followed the procedure introduced
by Cohen et al. (ll), which approximates a clinical or
"test-like" situation. The materials consisted of one
hundred 3 by 5 inch white cards on each of which was
typed a verb in the simple past tense, such as sat, came,
saw, etc. Below the verb appeared the six pronouns, 1,
We, You, They, He, and She in random order on each card.
In order that the research interfere with the
therapy process as little as possible, each experimenter
used his own office for the experiment. All offices are
similar in size, shape, number of windows, and furnishings.
Procedures with the Subjects
Student-subjects
Each student-subject was told whom she would see
56
at the time her appointment was made. For example, "Your
appointment will be with Mr. M , a psychiatric social
worker at the clinic. Be sure to ask for him when you ar
rive at 3=30 tomorrow afternoon." Upon arrival, the sub
ject saw the receptionist at the clinic who told her that
she might be seated and that "Mr. M will be with you
shortly." The experimenter then greeted his subject in
the waiting room and took her to his office. The chair
next to his desk was turned slightly away from him, so
that the subject could see him and could take the cards
as they were handed to her, but could not see clearly what
the experimenter was writing. After the subject was
seated, the experimenter said:
I am Dr. [or Mr.J ----, a staff psychiatrist [or psy
chologist or social worker] at this clinic, and as
you were told in class, we are developing a new tech
nic for personality investigation. Now you see these
cards here have a word typed in the center of each
one. What I'd like you to do is just make up a
sentence using that word. Below the center word you
will see several other words. Take any one and use it
to start your sentence. Answer with the first sen
tence that comes into your mind, though there is no
particular hurry. If you wish to comment on anything
as we proceed, feel free to do so. Any questions you
have, however, we'll talk about when we're through.
OK? Let's go ahead.
The instructions were written in this way so as to
emphasize both the professional affiliation of the experi
menter, and the clinical nature of the task. The subject
was instructed to "comment on anything as we proceed" in
57
order to stimulate any spontaneous comments of awareness
about the true nature of the task during the experiment.
The experimenters were asked to memorize these in
structions, but they were also mimeographed and attached
to a recording sheet which was used. This sheet also
contained the inquiry which followed the experimental
session. Facsimile copies may be found in Appendix C.
After reading the instructions the experimenter
handed the cards, one at a time, to the subject. The
cards were shuffled after each subject. After the first
twenty sentences, for the reinforced group, the experi
menter murmured "mm-hmm" in a natural manner each time
the subject began a sentence with either _I or We for the
remaining eighty trials.
Following the presentation of the cards, the ex
perimenter asked the following nine questions which were
designed to elicit any awareness of the purpose or nature
of the experiment if the subject had not already "caught
„ i t
on.
1. Did you usually give the first sentence that
entered your mind?
2. How did you go about deciding which of the
pronouns to use?
3. Which do you think you used the most often?
4. Why?
58
5. Did you notice anything I did during the test?
6. Why do you think I did that?
7. Did it have any effect on what you did?
8. What was the effect?
9. What was the purpose of all this in your
opinion?
Do you have any other comments?
If the subjects then had any questions about the
experimentj they were answered in whatever way the experi
menter thought might best satisfy their curiosity.
For the non-reinforced group of student-subjects,
the procedure was identical, except for the omission of
the verbal reinforcement.
Patient-subjects
None of the patients was informed of the experiment
before the therapy session during which it occurred.
While the wording was left up to the individual therapist-
experimenter, the following statement was discussed with
the experimenters as a model which they might use in intro
ducing the experiment to their patients:
The clinic is in the process of developing a technic
for the investigation of personality and we are try
ing to get as many people as we can to participate—
and I would like you to help us out, if you will.
It's a very simple procedure which is entirely talk
ing, and it will only take us about a half hour. So
we can do it now, if you like. I think you'll find
it interesting. OK?
59
[To the E: In discussing the task, as above, be sure
to avoid any such words as "experiment(al)," or "re
search," so that the procedure will remain, in the
patient's eyes, as clinical or "test-like" as pos-
sible--the more we can keep the entire procedure
within the "therapy" framework, the better.]
If the patient then agreed to the task (and all
did), the same procedure was followed as with the student
subjects, except that the instructions began with "Now you
see these cards. ..." Following the inquiry the purpose
of the experiment was handled individually by each experi
menter with his patient in the manner he felt therapeutic
ally advisable.
Experimental Design
The experimental design follows the pattern of a
four-way classification analysis of variance. The dimen
sions of reinforcement treatments, professional affilia
tions, and subject-class form a factorial design in which
hypotheses involving means may be tested. A fourth di
mension represents the successive stages of the experi
mental sessions in five blocks of twenty trials each,
and provides data for tests of hypotheses involving the
trends (methods x trial blocks interaction), as described
by Lindquist (30).
CHAPTER IV
RESULTS
The number of first person pronouns used by each
subject during each of the five successive trial blocks of
twenty trials each is recorded in Appendix A (Tables 8,
9, 10, 1:1). The mean number of first person pronouns used
during the five trial blocks by the six experimental and
six control groups is shown in Table 1.
It was first necessary to test the assumption that
the two members of each professional affiliation did not
differ significantly from one another as sources of rein
forcement. This was done in order that they could be con
sidered members of the same groups in making further
analyses regarding differences between professional affili
ations. Thus an analysis of variance for a random repli
cation of a two-factor experiment was performed where the
different sets of experimenters of each professional af
filiation represented the replications (30). The results
of this analysis are shown in Table 2.
The different reinforcement effects of individual
experimenters and the interactions with professional af
filiation yielded insignificant P-ratios which now allowed
for the collapsing of the data of both experimenters within
each professional group for the further analysis of the
60
Table 1
61
Table of Means--All Groups
Reinforced Subjects
I II III IV V
Patients
PSW 8.66 7.00 9.83 11.50 12.83
Psych. 10.00 9.16 9.83 11.00 11.50
M.D. 8.16 10.83 11.16
13.33
12.66
Students
PSW 6.00 7.16 8.60 11.00 IO.83
Psych.
7.33
11.66 11.66 10.16 11.50
M.D. 7.66 10.16 12.00 10.83 12.16
Non-Reinforced Subjects
r II III IV V
Patients
PSW 9.50 8.83 7.83 10.33 7.83
Psych. 6.00 6.83 8.33
7.66 6.16
M.D.
8.33 . 8.83 8.83 11.16
10.33
Students
PSW 7.66 9.16 9.00
8.33
9.00
Psych.
8.33
8.66 7.66 6.83 6.83
M.D.
8.33 9.33
8.50 9.00
10.33
Table 2
Source Table for Analysis of Variance of Repeated Measurements of the Mean Number
of First Person Pronouns Used by the Combined Reinforced Group as Compared
with the Non-Reinforced Group with Three Sets of Experimenters*
Source of Variation
Sum of
Squares
df
Mean
Square
* P
Reinforcement effects
1369.39
1
1369.39
6.052 <.02
Professional affiliation effects 539.08 2 269.54 1.191
Experimenter effects 138.89 1 138.89
Interactions:
Reinforcement x professional affiliation 461.03 2 230.52 2.200
Reinforcement x experimenters
9.39
1
9.39
Professional affiliation x experimenters 745.36 2 372.68
Reinforcement x professional
affiliation x experimenters
209.53
2 104.76
Within cell
13735.33
60 228.92
Total 17208.00
71
Pooled interactions 15160.64
67
226.28
#
Random replication of a two-factor experiment
sets of E's represent the replications (Lindquist,, 30).
(A x B x R)j the different
63
data.
It had been a basic a-ssuraption of the experiment
that verbal learning does occur under conditions similar
to those in this study. This was confirmed by an F-ratio
of 5.909 for reinforcement treatments which with 1 degree
of freedom is significant at p < .02. A t-ratio of 2.37
for the difference between the means of the total rein
forced and non-reinforced groups is also significant with
71 degrees of freedom at p < .02. Figure 1 shows the mean
number of first person pronouns used by all reinforced and
non-reinforced subjects over the five trial blocks of
twenty trials each.
In order to test prediction number 1 of Hypothesis
a Lindquist type III mixed analysis of variance (30)
was then performed comparing the effects of the subject
class and reinforcement treatments over the five successive
trial blocks. The results are presented in Table 3.
Where F-ratios were found to be significant t-ratios were
calculated and are presented in Tables 4 and 5. Table 4
presents the t-ratios of the difference between mean
scores for blocks of trials I through V for the total
sample of subjects. Figure 2 shows the curve of mean
scores for the total of all subjects over the five trial
blocks. Table 5 presents the t-ratios of the difference
between the mean scores for blocks of trials I-V for the
Mean no. of first person pronouns
64
Figure 1
Mean Number of First Person Pronouns for the Total
Reinforced and Total Non-Reinforced Subjects Over
the Five Trial Blocks of Twenty Trials Each
12. Oj
11.5
11.0
10.5
10.0
9.0
8.5
V II III IV I
Trial Blocks
Relnf.
» ■ — ■■■■ — ■ » ■»
Non-reinf.
Table 3
Source Table for Analysis of Variance of Trial Block Scores:
Reinforcement Treatments by Subject Class*
Source of Variation
Sum of
Squares
df
Mean
Square
F
P
Between (subjects) 3441.60
71
Reinforcement effects 273.88 1 273.88
5.909
<.02
Subject class
7.51
1
7.51
Interactions:
Reinforcement x subject class 8.09 1 8.09
error term (between) 3152.12 68
46.35
Within (subjects) 1868.40 288
Trial blocks 230.08 4 57.52 11.061 <.001
Interactions:
Trial blocks x reinforcement
143.54
4 35.88 6.900 <.001
Trial blocks x subject class 57-96 4 14.49 2.786 <.03
Trial blocks x subject class
x reinforcement 22.94 4 5.74 1.104
error term (within) 1413.88 272 5.20
Total 5310.00
359
*Type III mixed design analysis of variance (Lindquist, 30).
vji
66
Table 4
t-Ratlos of the Mean Number of First Person Pronouns
Used Between Each Block of 20 Trials by the
Total Sample of Subjects
Trials X SE^ df t p
I-II
.97
.380
71 2.55
<.02
I-III 1.43
II it
3.76 <.01
I-IV 2.10
II it
5.52 <.01
I-V 2.17
II Tl
5.71
<.01
II-III .46
II ri
1.21 --
II-IV
1.13
W ii
2.97
<.01
II-V 1.20
1 1 it
3.15
<.01
III-IV
.67
1 1 u
1.76 —
III-V .74
n n
1.94
--
IV-V
.07
IT it
.18
— —
67
Table 5
t-Batios of the Mean Number of First Person Pronouns
Used Between Each Block of 20 Trials by the
Total Reinforced Group of Subjects
Trials X sed df t
P
I - I I 1.36 •
CD
71
1.58
—
I - I I I
2.53
tt
2.94
A
•
O
I - I V
3.33
1 1 ir
3.87
<.01
I - V
3.95
It tt
4.59 A
•
O
I I - I I I
1.17
tt it
1.36 - -
I I - I V
1.97
It H
2.29 <.05
I I - V
2.59
It it
3.01
A
•
O
I-1
I I I - I V
o
C O
•
It IT
.93
_ _
I I I - V 1.42
tr It
1.66
I V - V .62
rt It
.72
—
Mean no. of first person pronouns
68
Figure 2
Mean Number of First Person Pronouns Used by All
Subjects Over the Five Trial Blocks
10.5-
10.0
9.5.
9.0
8.5
8. C
7.
II III
Trial Blocks
IV V
69
reinforced group of subjects. Lack of significance between
trial blocks for non-reinforced subjects can be assumed by
inspection of Figure 1.
The remaining hypotheses were then tested by a
Lindquist type III four-way analysis of variance (30), com
paring the effects of subject-class, professional affilia
tion, and reinforcement during the five successive trial
blocks. The results are summarized In Table 6. A signifi
cant F-ratio was obtained for the trial blocks by subject
class interaction term, and this is shown in graphic form
in Figure 3.
Further analysis of the data was then done, and
t-ratios computed for the differences between the sums of
means of groups related to critical factors in this study.
The results are presented in Table 7*
Hypothesis A. This hypothesis led first to the
prediction that the means and trends of the positively re
inforced criterion variables would be significantly greater
for the total group of patient subjects than for the total
group of student subjects. The Interaction term for re
inforcement treatments by subject class by trial blocks of
the type III mixed analysis of variance (Table 3) yielded
insignificant F-ratios, thus not supporting prediction 1.
■While no significant difference was obtained be
tween the means or trends of the reinforced patient group
Table 6
Source Table for Analysis of Variance of Trial Block Scores:
Subject Class by Professional Affiliation by Reinforcement Treatments*
Source of Variation
Sum of
Squares
df
Mean
Square
P
P
Between (subjects) 3441.60 71
Reinforcement effects 273.88 1 273.88 5.615 <.02
Subject class
7.51
1
Professional affiliation 107.82 2 53.910
Interactions:
Reinforcement x professional
affiliation 92.20 2
Reinforcement x subject class
8.09
1
Subject class x professional
affiliation 23.10 2
Reinforcement x subject class
x professional affiliation
2.33
2
error term (between) 2926.67 60 48.778
Within (subjects) 1868.40 288
Trial blocks 230.08 4 57.520
11.343
<.001
Interactions:
Trial blocks x reinforcement 143.54 4 35.885 7.076 <.001
Table 6 (continued)
Source Table for Analysis of Variance of Trial Block Scores:
Subject Class by Professional Affiliation by Reinforcement Treatments*
Source of Variation
Sum or
Squares
df
Mean
Square
F
P
Interactions (continued):
Trial blocks x subject class 57.96 4 14.490
2.857
<.02
Trial blocks x professional
affiliation 76.85 8 9.606 1.894
Trial blocks x reinforcement
x subject class 22.94 4
5.735
Trial blocks x reinforcement
x professional affiliation
49.63
8 6.204
Trial blocks x subject class
x professional affiliation 21.68 8 2.710
Trial blocks x subject class
x professional affiliation
x reinforcement 48.72 8 6.090
error term (within) 1217.00 240
5.071
Total 5310.00
*Type III four-way analysis of variance (Lindquistj 30).
-H
Mean no. of first person pronouns
72
Figure 3
Mean Number of First Person Pronouns Used by the Total
Patient Group and Total Student Group Over the Five
Trial Blocks, Irrespective of Reinforcement
11.0
10.5 .
10.0 -
II III IV
Trial Blocks
V
Patients
Students
Table 7
t-Ratlos for Differences Between Means of Groups Related to Critical
Factors in This Study
Groups
^D
SEd df t
p
Reinforced - Non-Reinforced 8.72 3.68
71 2.37
<.02
Patients - Students
1.45
3.68
71
<1.00
Reinforced-Patients - Non-Reinforced-Patients 10.22 5.21
71 1.96 <.05
Reinforced-Students - Non-ReinforcedlStudents
7.23
5.21
71
1.38
Reinforced-Patients - Reinforced Students 2.94 5.21
71
<1.00
Non-Reinforced-Patients - Non-
Reinforced- Students
.05 5.25 71
<1.00
Reinforced-M.D. - Non-Reinforced-M.D. 8.00 6.38
71 1.25
Reinforced-Psychologists - Non-Reinforced-
Psychologists
15.25
6.38
71
2.36 <.02
Reinforced-Psychological Social Workers -
Non-Reinforced Psychological
Social Workers
2.91
6.38
71
<1.00
M.D.-Patients - M.D.-Students 2.67
6.38
71
<1.00 .
Psychologists-Patients - Psychologists-
Students 2.08 6.38
71
<1.00
Psychological Social Workers-Patients -
Psychological Social Workers-Students
3.75
6.38
71
<1.00
u>
as compared with the student group, a significant t-ratio
of I.96 (p < .05) was obtained for the difference between
the sum of the means for the reinforced and non-reinforced
patient groups. The same statistic for the reinforced and
non-reinforced student subject groups only approached sig
nificance (t = 1.38j Table 7). Patient subjects, there
fore, did condition significantly where student subjects
only tended to.
The interaction term for subject class by trial
blocks of the type III four-way analysis of variance (Tabl
6) yielded an P-ratio of 2.857 which with 240 degrees of
freedom is significant at p < .02. Prediction 2 which
stated that there would be a significant difference be
tween the means and trends of the criterion variables of
the patient group and the student group irrespective of
verbal reinforcement, was thus supported with respect to
trends.
Hypothesis A, which predicted that patients would
respond in a different manner than student subjects was
thus confirmed (see Figure 3).
Hypothesis B. It was predicted first from this
hypothesis that the means and trends of the positively re
inforced criterion variables in the groups of patients
with experimenters of three different professional affilia
tions would differ significantly from one another. The
75
type III four-way analysis of variance (Table 6) yielded
an insignificant triple interaction term for reinforcement
treatments by subject class by professional affiliation,
thus not supporting these two predictions concerning the
means. The same analysis yielded an insignificant quadruple
Interaction term for reinforcement treatments by subject
class by professional affiliation by trial blocks which
thus did not lend support to the predictions concerning the
trends.
However, a t-ratio of 2.36 computed for the dif
ference between the sums of means of the reinforced and
non-reinforced psychologist groups is significant with 71
degrees of freedom at p < .02, whereas the same statistic
for the social worker and psychiatrist groups is not
significant.^
Consequently, Hypothesis B which states that
patient and student subjects will show a difference in de
gree or manner of response to verbal reinforcement depend-
^The variances of the groups of subjects reinforced
by the different professions were tested for heterogeneity
by Hartley's test (20) and found to be consonant fp >
.10). Thus, It is possible that the significant t" for
the psychologist groups, as contrasted with the social
worker and psychiatrist groups, occurs by chance: I.e.,
as a function of the main effects (reinforcement) which,
when accounted for Inthe analysis of variance, results In
a non-significant professional affiliation by reinforce
ment interaction term. Alternative hypotheses will be
discussed in Chapter V.
76
ing on the professional affiliation of the experimenter
is not rejected.
Hypothesis C. The first prediction from this hy
pothesis was that there would he significant differences
among the means and trends of the criterion variables in
the combined groups of all subjects with experimenters of
three different professional affiliations irrespective of
verbal reinforcement. The type III four-way analysis of
variance (Table 6) yielded an insignificant F-ratio for the
effects of professional affiliation on the use of the cri
terion variables. An insignificant F-ratio for the rein
forcement treatments by professional affiliation interac
tion term was also obtained, as well as an insignificant
F-ratio for the triple interaction term for reinforcement
treatments by professional affiliation by trial blocks.
Thus prediction 1 was not supported.
Prediction 2 stated there would be significant
differences among the means and trends of the criterion
variables in the groups of patients with experimenters of
three different professional affiliations irrespective of
verbal reinforcement. Prediction 3 asserted the same with
respect to the groups of students. The triple interaction
term for reinforcement treatments by subject class by pro
fessional affiliation yielded an insignificant F-ratio,
77
thus not supporting these predictions concerning the means.
The quadruple interaction term for reinforcement treatments
by subject class by professional affiliation by trial
blocks also yielded an insignificant P-ratio which did not
support the predictions concerning the trends.
Thus, Hypothesis C which asserted that the subjects
of both groups, patients and students, would be influenced
by the professional affiliation of the experimenter in
their use of the criterion variables was rejected.
Hypothesis D. This hypothesis first led to the
prediction that there would be no significant differences
among the means of the criterion variables of individual
experimenters' groups irrespective of reinforcement. Ef
fects of individual experimenter differences in the first
analysis of variance (Table 2) yielded an Insignificant
P-ratio; thus prediction 1 was given support.
It was secondly predicted that there would be no
significant differences among means of the reinforced cri
terion variables of individual experimenters' groups. The
experimenter by reinforcement treatments interaction term
yielded an insignificant P-ratio, thus lending support to
prediction 2.
In the third place, it was predicted that there
would be no significant differences between the means of
78
the criterion variables of the two individual experi
menters’ groups within each profession, irrespective of
reinforcement. The interaction term for experimenter by
professional affiliation yielded an insignificant F-ratio,
thus supporting prediction 3-
Finally, it was predicted that there would be no
significant differences between the means of the rein
forced criterion variables of the two individual experi
menters' groups within each profession. The triple inter
action terra for experimenter by professional affiliation
by reinforcement treatments was also insignificant. This
outcome is in line with prediction 4.
Hypothesis D, which stated that individual differ
ences among experimenters would have no differential effect
on the subjects' use of the criterion variables, was thus
confirmed.
An analysis of the responses to the nine awareness
questions revealed several important differences and simi-
'larities when classified according to subject class and
reinforcement treatments. The answers for the eighteen
subjects in each group were tallied and are presented in
Appendix B.
The results of this analysis revealed first that
all subjects felt they had responded with the first sen
tence that came to mind. Recording of the responses had
79
an inhibiting effect on several subjects. Others reported
feeling inhibited by a recognition that they tended to
use the first person pronouns "too much" or by trying to
follow a story line. All the subjects felt they had com
plied with the rules as much as possible, however. More
over, the majority of subjects (eleven or more in each
group of eighteen) indicated that they had responded with
the first thing that came to mind when they were asked how
they had decided on which pronoun to use. They felt they
had made no effort to make any conscious decision.
In both the reinforced and non-reinforced groups
there was an identical feeling by the subjects that the
first person pronoun had been used most frequently. In
view of the subjects' general feeling that they had re
sponded spontaneously and with no conscious effort to
choose a particular pronoun, this would tend to indicate
a lack of correctness about hypotheses they may have
formulated concerning what was desired of them. In fact,
many felt that it was undesirable for them to use the
first person pronouns so much.
Both patients and students made essentially similar
comments about constructing sentences of self-reference
import. The patients, however, felt it was "self-
centered" or "egotistical" to use the first person pro
nouns so much, and stated that they tried to avoid it.
80
The students, on the other hand, merely commented that
they "must be pretty self-centered," and in only one case
was there any reported attempt to avoid it. Thus, we
might conclude that the students appear to be more comfort
able with or accepting of their "egotism."
Of the reinforced groups more patients noticed
the experimenter's verbal behavior, while more students re
ported an awareness of his writing. Of the non-reinforced
groups -the. students, as contrasted with the patients, were
still more axvare of the experimenter's overt behavior:
fourteen noticing his writing and three his occasional
glances at them. This is an important finding because it
suggests that patients are more sensitive to their thera
pist 's verbal behavior than non-patient subjects even though
both groups seem to be equivalent in regarding such be
havior as a means of encouraging them.
Even though most subjects were quite aware of the
experimenter and his behavior, the majority of the sub
jects denied feeling influenced in any way by his behavior,
either verbal or non-verbal. On the other hand, a number
of subjects did report feelings of "encouragement,1 1 "curi
osity," etc. Many subjects were also aware of the rein
forcement, per se, even though unable to specify just what
the experimenter intended.
The majority of the subjects felt that the experi-
mental procedure was a personality test or "for investi
gating personal relationships." The patient subjects
perceived it in far more personal terms, though, such as
a test of progress in psychotherapy, or a test of their
self-centeredness. Many of the students, on the other
hand, responded to the task with a more experimental or
academic set. They felt that it was a test- of skill and
concentration, a test of sentence structure, etc.
The possibility exists that some subjects may have
formulated a "correlated hypothesis" about using the cri
terion variables as a way of telling about themselves on
a personality test. While, theoretically considered,
this is a possibly influential factor, none of the sub
jects gave any Indication of its existence.
CHAPTER V
DISCUSSION
The results of the present study support the gen
eral hypothesis that in the clinical setting there is a
difference in manner and degree of response to verbal re
inforcement between patients currently in psychotherapy
and "normal" student subjects. The results also tend to
support the hypothesis that patients are more sensitive to
the verbal cues of their therapists than are non-patients.
Moreover, it was demonstrated that under certain circum
stances the professional affiliation of the experimenter
will have an effect on the degree to which verbal learning
without awareness occurs.
These findings are in line with the general trend
of results reported in recent studies of verbal learning
without awareness. The present results, however, give rise
to a number of questions which will be discussed in some
detail below.
First, the hypothesis suggested by Marion (33)t
Krasner (28), and by Cohen in a personal communication to
Marion, that patients will be more sensitive than non
patients in a verbal learning experiment— in fact, "will
condition easily"— is supported, but not without a sig
nificant qualification. In the present study reinforced
82
patients did not show a significant difference from rein
forced students in the obtained level of learning. How
ever, when the differences between reinforced and non-
reinforced patients and reinforced and non-reinforced stu
dents are compared, it is clear that the patients rein
forced to a significant degree whereas the students only-
approached significance. Moreover, the patient subjects
reported a greater awareness of the verbal behavior of the
experimenters while the students were more aware of the
note-taking activity.
It could be argued that the experimenters' note-
taking might have interfered with the effects of the verbal
reinforcement, but this is open to question. Talbot
using college students as subjects in an experiment on
verbal learning, was unable to demonstrate that note-taking
had any effect whatsoever. Even if it had an effect in
the present investigation, it would only serve to strength
en our hypothesis that the patients were more "set" to
respond to the experimenter’s verbal behavior rather than
to his possibly distracting note-taking.
The learning curves for the two groups--patients
and students— are significantly different in shape, thus
posing further questions regarding the manner in which the
two groups learned (see Figure 3). The curve for the stu
dent subjects, it will be noted, accelerates rapidly then
84
tends to level off. For the patients, however, a striking
difference exists in that this group shows little rein
forcement effects in the first block of reinforced trials,
and in the second block still less than the students had
shown in their first block. While the students tended to
learn quickly and to maintain their level of learning, the
patients learned more slowly, but caught up with the stu
dents by the last trial block.
Could such a difference in rate of learning be
related to the age or educational level of the subjects?
It has been reported that groups of "normal" subjects com
parable to our patients with respect to age and education
have performed in a manner similar both to the students in
the present study and to student populations in other in
vestigations (4,11,17). The results of these studies would
thus tend to rule out age and education as influential
factors in investigations of learning without awareness—
at least where age and educational differences are rela
tively small, as in the present study. Accordingly, one
is still confronted with the problem of explaining the
difference between the learning curves of our two groups
of subjects.
Taffel (53) suggested that a clinical setting en
courages one to talk about himself more than In other
situations, thus facilitating reinforcement of first
person pronouns. It may be recalled that we earlier sug
gested that, irrespective of reinforcement, the patients,
having been in therapy for some months prior to the experi
ment, might consequently be more prone than non-patients
to the use of such pronouns. However, the results seem to
be at variance with this expectation; for in the groups of
reinforced patients there was an initial tendency not to
use the criterion responses, and there was no significant
difference in the frequency with which first person pro
nouns were used by the non-reinforced patients as compared
with the non-reinforced students.
Differences in the learning curves between the two
groups might be accounted for by consideration of the
awareness factor.
As already mentioned, while both patient and stu
dent subjects reported an awareness of the predominant use
of the first person pronouns, the students appeared uncon
cerned about their "egocentricity" and made no further
comment. The patients, on the other hand, perceived the
situation in very personal terms such as an important per
sonality test, or a test to measure their progress in
psychotherapy. As a consequence they reported both a de
sire and an effort not to use the criterion variables.
They felt, as had the students, that this might indicate
"self-centeredness." In contrast to the students, however,
the patients indicated that they regarded this as an un
desirable trait which might indicate that they were not
good patients or that they were not progressing satisfac
torily in therapy. In spite of their desire not to use
the critical response, however, the patient subjects did,
by the third trial block, increase the number of these
responses to a significant degree. This Is very much like
the reaction of some of Ball's (4) subjects who also re
ported a desire to avoid the use of the critical response
in his experiment but with no success. That is, they too
conditioned significantly and without awareness of having
done so.
Furthermore, several of our experimenters reported
manifestation of irritation or distress on the part of the
patients in connection with the experimental sessions.
Such reactions ranged from "mildly resentful" to one pa
tient who declared, angrily, "Well, I guess you have to
expect this sort of thing in a public clinic." The effect
of such attitudes on the learning process has been studied
by both Sarasson (47) and Sapolski (46). Sarason found
that defensiveness and hostility, as contrasted with
compliant-dependent attitudes, inhibited verbal learning.
According to Sapolski, negative feeling toward the experi
menter may interfere with learning until a period of ex
tinction after the experimenter has left the room. In
87
fact the learning curve for his subjects is similar to the
one for our group of patients.
Unlike Sapolski1s subjects, however, our patient
group did show learning with the experimenter still present
in the room even though initially the performance was in
ferior to that of the student group. While many of the
patients were irritated or angry at having their therapy
interrupted by the experiment, it is likely that, in gen
eral, they were favorably disposed towards their thera
pists. It is also likely that they were motivated to per
form well on a "test" which they may have regarded as di
rectly or indirectly contributing to their welfare. In
fact, in later therapy sessions a number of the patient
subjects inquired about their "scores" or how well they
had done on the "test."
In summary, while our results indicated that pa
tients are more sensitive than non-patients to the verbal
cues of their therapists, the manner and speed at which
verbal learning takes place is apparently related both
to the desire to do well and to affective orientation at
the time of reinforcement. The implications of this find
ing would seem to have obvious relevance for the timing of
interpretations in psychotherapy. It is well known that
no matter how correct an interpretation may be a patient
must be "ready" for its acceptance. Otherwise it seems to
88
fall on deaf ears or else the patient openly rejects or
resists it. The present study suggests that a very im
portant element in such "readiness” is the patient's desire
to learn or respond favorably to what a liked or respected
therapist is saying.
The hypothesis concerning the differential influ
ence of professional affiliation on verbal learning
remains to be considered. As previously mentioned, there
were no significant differences in level of verbal learning
as functions of the three professional groups serving as
experimenters. Nor did they differ significantly in the
manner or degree to which they influenced the learning of
the reinforced patient and student groups separately.
Furthermore, there were no significant differences attribut
able to professional affiliation in the subjects' use of
the critical response irrespective of reinforcement (see
Table 6).
The kinds of status differences suggested by Marion
(33) and others were not manifest in the present experiment
in terms of the subjects' responses. It may be that the
alleged effects on psychotherapy of status differences re
ported in Chapter I are very transient in nature or else
they failed to become strong enough to have Influenced the
responses of our subjects.
Actually, our findings suggest that the experience
differences of Marion's experimenters rather than status
differences account for his results. In the present study,
experience was controlled, and no over-all status differ
ences appeared. In his study, Marion was unable to control
the experience factor, and since his high status staff mem
bers undoubtedly had considerably more experience than his
low status trainees, as might have been anticipated, he
obtained a difference in favor of the staff experimenters.
It should be obvious that Marion has not succeeded in
demonstrating the role of status differences. At all
events we failed to confirm his findings when the experi
ence factor was kept reasonably constant. The entire
problem calls for more study.
Related to the foregoing is the fact of a signifi
cant t-ratio for the difference between total means of re
inforced and non-reinforced subjects for the psychologist
group, while similar t's computed for the social workers
and psychiatrists were not significant (see Table 7).
This finding, as previously implied, is restricted to such
separate comparison and does not hold for concurrent
analysis of the three professional groups (see footnote,
p. 75) • There are several ways in which this apparently
contradictory finding can be explained. First, it was a
psychologist who introduced the study and asked for
volunteers from introductory psychology classes. Conse-
quently, these subjects may well have had a "set" to re
spond to a psychologist more than to members of the other
disciplines. Secondly, since the task for the experimenter
was similar to usual psychological testing procedures in
volving instructions, presentation of test-like materials,
and recording responses, the psychologists may have con
ducted the experiment in a more skilled and relaxed manner
with less distraction than the other experimenters. The
latter possibility is supported by statements made by sev
eral of the non-psychologist experimenters. One of them
complained that the non-reinforcement sessions were
"anxiety-provoking" for him and, he believed, also for his
subjects. Another experimenter was reluctant to partici
pate in the study because he thought his patients might
be upset by the change in the therapy routine. Still
another experimenter voiced his displeasure at having to
fit the student subjects into his already busy schedule.
His irritation might conceivably have been reflected in
his behavior during the sessions.
In general, it should be clear that this endeavor
to transfer laboratory methods for studying verbal learn
ing without explicit awareness to the psychotherapeutic
setting has been successful. The evidence presented indi
cates that "mm-hmm" (and probably other Incidental be
havior) plays a far more "directive" role than has hitherto
91
been suspected. Nor should one overlook the role of nega
tive reinforcements such as inattention, disparaging com
ments, or gestures of disapproval. It is well for thera
pists to be mindful of these subtle means of influencing
the kind of learning essential for successful psychotherapy.
Suggestions for Future Research
In most of the reported studies of verbal learning
without awareness reinforcement is begun after an Initial
block of trials or a predetermined period of time has
elapsed in order to establish an operant level of the
critical response. We know of no investigation dealing
with latent learning in human subjects comparable to
animal studies of this phenomenon. It has been assumed to
date that the Increased level of the critical response Is
related to the reinforcing effect of such secondary social
approval rewards as "mm-hmm," "good," a nod, etc., other
behavioral cues being disregarded by the subjects. It
would seem a worthwhile extension of psychological knowl
edge to do a study in which the experimental groups were
allowed, say, eighty to one hundred non-reinforcement
trials followed by an equivalent number of reinforced
and extinction trials.
It is currently accepted by most investigators
that the subjects form a variety of covert hypotheses
92
about the purpose of the experimental procedure in order
to arrive at an answer to what the experimenter wants of
them. We might expect that subjects in a latent learning
experiment would have entertained and rejected a sufficient
number of such hypotheses during the first eighty non
reinforcement trials that they would be particularly sensi
tive to the correct hypothesis once reinforcement was be
gun. Thus they might quickly reach the level of correct
responses attained by a control group who had been rein
forced from the beginning. That is, once reinforcement
was begun, the initial incorrect hypotheses might act as
latent cues which could be quickly dismissed in favor of
the correct "path" to the "goal" of obtaining more rein
forcement (e.g., "mm-hmm"). We would predict that such
subjects would also be far more familiar with the experi
menter's general behavior after eighty or one hundred
trials, and would thus more quickly perceive what he was
doing and why. Consequently, differences in awareness
might also be expected.
Several studies of the effects of variable rein
forcement schedules on learning without awareness have
been reported with differing results. Grossberg (19)
found that spaced reinforcement produced no differences
in learning during acquisition trials from regular rein
forcement, but led to an increased resistance to extinc-
93
tion. In two other studies Kanfer (25) and Spivak (52)
reported that spacing the reinforcement necessitated a
greater number of acquisition trials but fewer number of
reinforcements, and extinction is significantly slower than
with regular reinforcement. These studies point to the
necessity for the application of reinforcement schedules
to investigations with patient populations of various
types, and where possible in clinical settings. It Is,
after all, probably never true that a psychotherapist
always reinforces a particular response class every time
it is emitted by the patient. Rather It is more likely
that varying (and probably Inconsistent) schedules of re
inforcement are followed, perhaps quite by chance.
A logical but difficult extension of the experi
mental work suggested by the results of the present study
Is the manipulation of variables in the therapy setting
without the patient's knowledge that anything other than
therapy was In process. This would be similar to the
Verplanck experiment (58) in which friends, colleagues,
relatives, etc., acted as subjects In social settings with
out knowledge they were even participating in an experi
ment. Ethical considerations Immediately arise, but they
might be resolved at least In part in a public agency
setting by asking patients to sign a release or permit for
experimental participation at the time of application for
treatment. In this way such technics as those described
by Adams and Hoffman (2), Salzinger and Pisoni (44,45),
and Rogers (43), for conditioning self-reference or self-
reference affect statements could be employed with very
little, if any, interruption in the therapy process. Total
lack of insight into the experimental procedure might be
expected, and it is highly unlikely that any deleterious
side effects would develop in responsibly planned investi
gations .
Sex differences of both patient-subjects and
therapist-experimenters and their interaction is another
important problem which should be more thoroughly studied.
At least one leading school of psychotherapeutic thought,
traditional psychoanalysis, has claimed that the sex of
the patient and/or therapist makes no difference to the
successful termination of the analysis (15j37). This may
be true, but there is some evidence that sex differences
do make a difference in verbal learning experiments.
Since there may be a relationship between what occurs in
treatment and in verbal learning experiments, it would
seem relevant to investigate such factors more extensively.
It might also be asked if there would be a differ
ence in the degree or manner to which learning and extinc
tion would occur between a group of patients diagnosed as
"situational reactions" (short-term therapy cases) and
95
another group of "character disorders" (long-term cases)?
One might continue to devise still other research
problems in this area in the attempt to gain fuller under
standing of the clinical interaction. The foregoing sug
gestions are not to be construed as an exhaustive list of
crucial problems. They merely serve to illustrate some
whose implications extend beyond the clinical field into
that of general psychology.
CHAPTER VI
SUMMARY AND CONCLUSIONS
Summary
A thorough understanding of the psychological nature
of man can never be complete without precise knowledge of
the multiple dynamic factors which are involved in inter
personal relationships. Man never acts in a vacuum.
There are always the innumerable antecedent conditions
which have contributed both to the present behavior and
also to its meaning. In a situation with another person
there are two sets of antecedent conditions and a setting
in which behavior is determined and given meaning not only
by these antecedents, but also by the behavior of the other
person. The situation then becomes a highly complex,
dynamic interaction, the variables of which are difficult
to isolate.
Recent years have produced a number of experimental
studies which have had as their purpose the isolation and
clarification of some of the pertinent unconscious forces
contributing to the behavior observed in interpersonal
relationships. One method has been to apply the principles
of operant conditioning to the study of verbal behavior.
This has been particularly important in opening the way
96
97
for a more precise description and understanding of be
havior which occurs in one extremely significant and influ
ential segment of human life— the clinical interaction.
In this interaction., possibly more than in any
other dyadic relationship, verbal behavior is the basic
mode of communication. The measurement of this behavior
and the ultimate hope for its exact prediction and control
have stimulated a variety of studies which have been, at
times, encouraging, even sobering in the light of implica
tions which can be drawn from their results. Perhaps most
impressive has been the evidence showing that behavior can
be signlficantly--strikingly--modified without the sub
ject's knowledge either of the experimenter's purpose or
the fact that his behavior has been altered.
Reviewing the literature of studies of verbal
learning without awareness which have been executed over
the past twenty-five years, we find an emphasis first on
the development of technics of conditioning coinciding
with an attempt to explore the phenomenon of lack of aware
ness. Following this, investigators turned their attention
to a variety of subject-types, and attempted a clarifica
tion of the clinical situation, using subjects with various
diagnoses, levels of anxiety, etc. Overlapping these
studies were others attempting to investigate experimenter
differences, interaction between experimenters and
subjects, and finally, several studies more closely ap
proximating psychological testing and therapy settings.
The majority of the investigators concluded that their
results were of major significance for psychodiagnosticians
and psychotherapists. Occasionally an attempt was made to
use the results of these studies as a basis for explaining
current technics of psychotherapy as teaching the patient
to talk differently about himself and nothing more.
While many of the conclusions reported appear
valid, especially those relating to the clinical interac
tion and the power of different technics of verbal rein
forcement, the question was raised as to whether a differ
ence might exist between patients and "normals" where the
patients perceived the experiment in terms of psycho
therapy. There had never been a clear-cut test, using
psychiatric patients in treatment with their own therapists
acting as the experimenters.
The purpose of the present study was to investi
gate the application of the technics of verbal learning
without awareness to the clinical interaction in terms of
subject differences (psychiatric clinic patients versus
college students), and differences in the professional af
filiation of the experimenter-therapists. It was predicted
that patient-subjects with their own experimenter-
therapists would condition to a higher degree or in a
99
different manner than student subjects. It was further
predicted that the professional affiliation of the experi
menter would have an effect on the degree or manner to
which the two groups of subjects would condition.
The experimenters were six male staff members of
a community mental hygiene clinic: two psychiatrists, two
psychologists, and two psychiatric social workers, with
controls for age and amount of clinical experience. The
patient-subjects were thirty-six adult female, non-
psychotic patients (six from each experimenter-therapist1s
current caseload) who had been in treatment for at least
three months. The student-subjects were female volunteers
from introductory psychology classes in a nearby college.
The method consisted of the presentation of one
hundred 3' by 5 inch white cards, in the center of which
was typed a common verb in the simple past tense. Beneath
the verb were typed in random order for each card the six
pronouns, JE, We, You, He, She, and They. The subjects'
task was to make up sentences using the verb and beginning
with one of the pronouns. After the first twenty trials,
the experimenter reinforced each use of "I" or "we" with
the generalized vocal reinforcement "mm-hmm." At the end
of the experiment each subject was questioned carefully
regarding her understanding of the purpose or nature of
the experiment, as well as her awareness of the
instigating cues.
The results indicated that the patients in this
study did respond differently to reinforcement than the
students. The student group tended to learn quickly and
to maintain their level of response, much in the same man
ner as subjects in previous studies. The patient group,
however, learned more slowly but ucaught up" with the
level of the student group by the end of the experiment.
Further, the comparison between the reinforced and non
reinforced patient groups and the reinforced and non
reinforced student groups showed that the former condi
tioned significantly while the latter only approached sig
nificance. The first hypothesis concerning differences
in degree and manner of response was thus supported.
With respect to differences in degree or manner of
learning which might be related to the professional af
filiation of the experimenter, there were no apparently
significant differences until the reinforced and non
reinforced groups of all subjects were compared for each
professional affiliation. The psychologists then obtained
a significant degree of learning whereas the psychiatrists
and psychiatric social workers did not. Thus, for this
study, the second hypothesis concerning differences in
degree or manner of learning related to professional af
filiation was also confirmed.
101
Questioning revealed that the patient subjects were
more conscious of the experimenter's verbal behavior and
the students more aware of his motor activities such as
note-taking and postural changes.
An attempt was made to explain the difference in
learning curves between the patient and student groups in
terms of the patients' reaction to the experiment with an
angry or resentful feeling due to interrupting their
therapy hour with a "test." It was reasoned that such a
reaction could produce a delay in the overt expression of
their response to reinforcement. Support was found for
this interpretation in other published studies.
Questioning the experimenters following the col
lection of the data led to the conclusion that the sig
nificant degree of conditioning observed in the psycholo
gists' groups as contrasted with the social workers' and
psychiatrists' groups was probably due to the former's
greater skill and ease in following "testing1 1 procedures.
Feelings of unnaturalness and anxiety aroused in some
experimenters during the experiment were probably a
crucial factor in reducing their effectiveness as sources
of reinforcement.
Conclusions
The results of this study justify the following
102
conclusions:
1. Verbal learning without explicit awareness is
an observable phenomenon in the clinical
interaction.
2. Patients appear to be more sensitive to cues
of verbal reinforcement in the clinical
interaction than normal students serving as
subjects.
3. Differences in degree of learning related to
the professional affiliation of the experi
menter could probably be eliminated by
giving the non-psychologist therapist the
kinds of testing and investigative experi
ence clinical psychologists receive in the
course of their training.
4. The extent to which the psychological factors
investigated in this study actually operate
in the course of routine psychotherapy re
mains an unsettled problem calling for
specialized research.
REFERENCES
104
R E FER E N C E S
1. Adams, Joe K. Laboratory studies of behavior without
awareness. Psych. Bull., 1957, 54, 383-405.
2. Adams, J. S., & Hoffman, B. The frequency of self
reference statements as a function of generalized
reinforcement. J. abnorm. soc. Psychol., I960,
60, 384-90.
3. Alexander, Franz, & French, Thos. Psychoanalytic
Therapy. New York: Ronald Press, 1946.
4. Ball, R. S. Reinforcement conditioning of verbal
behavior by verbal and non-verbal stimuli in a
situation resembling a clinical interview. Unpub
lished doctor's dissertation, Indiana Univer.,
1952.
5. Binder, A., McConnell, D., & Sjoholm, Nancy. Verbal
conditioning as a function of experimenter charac
teristics. J. abnorm. soc. Psychol., 1957, 55*
309-14.
6. Bricker, P. D., & Chapanis, A. Do incorrectly-
perceived tachistoscopic stimuli convey some in
formation? Psychol. Rev., 1953, 60, 181-88.
7. Brody, E. B., & Redlich, F. C. Psychotherapy with
schizophrenics. New York: International Uni-
versiti'es Press, 1952.
8. Bychowski, G., & Despert, J. Louise. Specialized
techniques in psychotherapy. New York: Basic
Hooks, 1952.
9. Caruth, Elaine G. The relationship of dependency to
verbal learning without awareness. Unpublished
doctor's dissertation, University of Southern
California, 1959-
10. Cohen, E., and Cohen, B. D. Verbal reinforcement in
schizophrenics. J, abnorm. soc. Psychol., i960,
60, 443-46. : -------------------
11. Cohen, B. D., Kalish, H. I., Thurston, J. R., &
Cohen, E. Experimental manipulation of verbal
behavior. J. exp. Psychol., 1954, 47, 106-10.
105
12. Daily, J. M. Verbal conditioning without awareness.
Dissertation Abstr., 1953, 13, 1247-48.
13. Dollard, John, & Miller, Neal E. Personality and
psychotherapy. New York: McGraw-Hill, 1950.
14. Fiedler, Fred. E. A comparison of therapeutic rela
tionships in psychoanalytic, nondirective and
Adlerian therapy. J. consult. Psychol., 1950,
14, 434-45.
15. Fromm-Reichmann, Frieda. Principles of intensive
psychotherapy. Chicago: She University of Chicago
Press, 1950.
16. Greenspoon, J. The effect of two non-verbal stimuli
on the frequency of members of two verbal response
classes. Amer. Psychologist, 1954, 9> 384.
17. Greenspoon, J. The reinforcing effect of two spoken
sounds on the frequency of two responses. Amer.
J. Psychol., 1955, 68, 409-16.
18. Gross, L. R. Effects of verbal and nonverbal rein
forcement in the Rorschach. J. consult. Psychol.,
1959, 23, 66-68.
19. Grossberg, J. M. The effect of reinforcement schedule
and response class on verbal conditioning. Dis
sertation Abstr., 1956, 16, 2211.
20. Hartley, H. 0. The maximum F-ratio as a short cut
test for heterogeneity of variance. Biometrika,
1950, 37, 308-12.
21. Heine, Ralph ¥. A comparison of patients' reports
on the psychotherapeutic experience with psycho
analytic, non-directive, and Adlerian therapists.
Amer. J. Psychother., 1953, 7, 16-23.
22. Hlldum, D. C,, & Brown, R. ¥. Verbal reinforcement
and interviewer bias. J. abnorm. soc. Psychol.,
1956, 53, 108-11.
23. Howie, D. Perceptual defense. Psychol. Rev., 1952,
59, 308-15.
24. Irwin, F. ¥., Kaufman, K., Prior, G., & ¥eaver, H. On
"learning without awareness of what is being
learned. J. exp. Ps.vchol.. 1934, 17, 823-27.
106
25- Kanfer, F. H. The effect of partial reinforcement on
acquisition and extinction of a class or verbal re
sponses. J. exp. Psychol., 1954, 48* 4-24-32.
26. Klein* S. Conditioning and extinction of operant
verbal behavior in neuropsychiatric hospital
patients. Dissertation Abstr., 1954, 12, 2127-28.
27* Krasner, L. A. A technique for investigating the
relationship between the behavior cues of the
examiner and the verbal behavior of the patient.
J. consult. Psychol., 1958, 22* 364-66.
28. Krasner, L. A. Studies of the conditioning of verbal
behavior. Psychol. Bull., 1958, 55, 148-70.
29. Leventhal, Allan M. The effects of diagnostic cate
gory and relnforcer on learning without awareness.
J. abnorm. soc. Psychol., 1959* 59* 162-67.
30. Lindquist, E. F. Design and analysis of experiments
in psychology and education. Boston: boughton
Mifflin, 1953*
31. Lord, Edith. The influence of negative and positive
rapport conditions on Rorschach performance. Un
published doctor's dissertation. University of
Southern California, 1948.
32. Mandler, G., & Kaplan, W. K. Subjective evaluation
and reinforcing effect of a verbal stimulus.
Science, 1956* 124, 582-83.
33* Marion, A. J. The influence of experimenter status
upon verbal learning. Unpublished doctor's disser
tation, University of California at Los Angeles,
1956.
34. May, H. The effect of verbal reinforcement on the
. recall of early memories. J. abnorm. soc. Psychol.,
1959* 59* 254-58.
35* McGinnies, E. M. Emotionality and perceptual de
fense. Psychol. Rev., 1949* 56, 244-51.
36. McKinney, Fred. "Directive Techniques," in Penning
ton, L., & Berg, I. A. (eds.) An introduction to
clinical psychology. New York: The Ronald Press,
1948*
107
37* Menninger, Karl. Theory of psychoanalytic technique.
New York: Basic ’ ’ Books', Inc., 1958•
38. Murdock, B. B. Perceptual defenses and threshold
measurements. J. Personality, 1954, 22, 565-71.
39. Nuthmann, A. Conditioning of a response class on a
personality test. J. abnorm. soc. Psychol,, 1957,
54, 19-24.
40. Philbrick, E. B., & Postman, L. A further analysis
of "learning without awareness." Amer. J. Psychol.,
1955, 68, 417-24.
41. Postman, L., & Jarrett, R. An experimental analysis
of learning without awareness. Amer. J. Psychol.,
1952, 65, 244-55.
42. Rogers, Carl. Client-centered therapy. Boston:
Houghton Mifflin Company, 1951•
43. Rogers, J. M. Operant conditioning In a quasi
therapy setting. J. abnorm. soc. Psychol., i960,
60, 247-53.
44. Salzinger, K., & Pisoni, Stephanie. Reinforcement
of affect responses of schizophrenics during the
clinical Interview. J. abnorm. soc. Psychol.,
1958, 57, 84-91.
45. Salzinger, K., & Pisoni, Stephanie. Reinforcement
of verbal affect responses of normal subjects dur
ing the clinical interview. J. abnorm. soc.
Psychol., I960, 60, 127-31.
46. Sapolski, A. Effect of interpersonal relations upon
verbal conditioning. J. abnorm. soc. Psychol.,
i960, 60, 241-47.
47. Sarason, I. G. Interrelations among individual dif
ference variables, behavior in psychotherapy, and
verbal conditioning. Paper read at the Western
Psychological Assoc., Eugene, Oregon, May, 1957
(cited by Krasner (25)).
48. Sarason, S. B. The clinical interaction. New York:
Harper and Brothers, 1954.
108
49. Sidowski, J. B. Influence of awareness of reinforce
ment on verbal conditioning. J. exp. Psychol.,
1954, 48, 355-60.
50. Skinner, B. P. Science and human behavior. New
York: The Macmillan 'Company, 1953.
51. Skinner, B. F. Some contributions of an experimental
analysis of behavior to psychology as a whole.
Amer. Psychologist, 1953, 8, 69-79.
52. Spivak, M., and Papajohn, J. The effect of the
schedule of reinforcement on verbal conditioning.
J. exp. Psychol., 1954, 48, 355-60.
53. Taffel, C. Anxiety and the conditioning of verbal
behavior. J. abnorm. soc. Psychol., 1955, 51,
496-501.
54. Talbot, E. The effect of note-taking upon verbal
responses and its implication for the interview
situation. Unpublished doctor's dissertation, Uni
versity of California at Los Angeles, 1954.
55. Tatz, S. J. Symbolic mediation In learning without
awareness. Paper read at The Eastern Psychological
Association, Atlantic City, March, 1956 (cited in
Krasner (25)).
56. Thorndike, E. L., & Rock. R. T. Learning without
awareness of what is being learned or intent to
learn it. J. exp. Psychol., 1934, 17, 1-19.
57. Thorne, F. C. "Directive and Eclectic Personality
Counseling," in McCarey, J. L., & Sheer, D. E.
(eds.) Six approaches to psychotherapy. New
York: The Dryden PreT3s,' 1955.
58. Verplanck, W. S. The control of the content of con
versation: reinforcement of statements of opinion.
J. abnorm. soc. Psychol., 1955, 51, 668-676.
59. Wexler, Milton. Psychotherapy with schizophrenics
and borderline states. Lecture given at Los
Angeles Institute for Psychoanalysis, Los Angeles,
December, i960.
60. White, R. W. The abnormal personality. (2nd Ed.)
New York: The Ronald Press, 1956.
109
61. Wickes, Thos. A., Jr. Examiner influence in a test
ing situation. J. consult. Psychol., 1956, 20,
23-27.
62. Wilson, W. C., & Verplanck, W, S. Some observations
on the reinforcement of verbal operants. Amer. J.
Psychol., 1956,' 69, 448-51.
63. Winder, C. L. "Psychotherapy," in Farnsworth, P. R.,
& McNemar, Quinn (eds.) Annual review of psy
chology . Palo Alto: Annual Reviews, 1$5?•
APPENDICES
A P P E N D IX A
TA BLES OF MI - w e " R E SPO N SE S
112
Table 8
Total Number of "I-We" Responses for Reinforced
Patlent-Subjects Over the Five Trial Blocks
Patient
Number
I II III IV V
1 8
7
12
9
11
2
7
2 8 11
9
3 3 5 7
16 18
4 14
15 15 19 15
5
6
7 7 7
11
6 14 6 10
7 13
7
14 6 12 14
13
8
5 9
8
9 15
9
12 6 8
9
12
10 4
7
6
5 7
11 16 20 20 20
15
12
9 7 5 9 7
13
14
15
16 16 16
14 6 10
9
12
15
15
8 12 12
13 13
16
5
6
9
11 10
17 7
12 10
15
12
18 9 .
10 11
13
10
113
Table 9
Total Number of "I-We" Responses for Reinforced
Student-Subjeets Over the Five Trial Blocks
Student
Number
I II III IV V
1
5 3
11 10
9
2 8
13 15
12
13
3 7 7
6 12 8
4
7 5
1 8 8
5 5
8
7
12 14
6 4
7
11 12
13
7
6
13 17 9
12
8 10 12 12 12 14
9
8
13 9
12 11
10 6 8
9
6 8
11
7 13 13 ' 9 13
12
7
11 10
13 11
13
12 14
13
12 14
14
7 7
11
5
10
15 7
8 12 12 12
16
5 9
10 11 8
17 7
10 12 12 15
18 8
13 14 13
14
Table 10
Total Number of "I-We" Responses Tor Non-Reinforced
Patients Over the Five Trial Blocks
Patient
Number
I II III IV V
1 6
5 3 7 5
2 6
5 7
6 4
3
8 8
9
16 11
4 16
15
12 14
13
5 9
10
9
8 6
6 12 10
7
11 8
7 7
6 8
7
8
8
3
2 8
7
8
9
8 12 10
9
1
10
9
12 14 14 10
11
3
2 2
3 3
12 6
7
8 6
7
13 5
4 12 10 11
14 8 8 8 10
9
15
14
15
16 16
15
16 11
7
4 12
7
17
2 8 1 6
7
18
10
11 12
13 13
s'ssa y i—
115
Table 11
Total Number of "I-We" Responses for Non-Reinforced
Students Over the Five Trial Blocks
Student
Number
I II III IV- V
1
5 9 3 7
6
2 11 . 12
13
9 16
3 7
10 11 8 8
4 6 4 6 6
5
8
9
12 11
9
6
9
11 8
9 9
7
10
7
10 14 11
. 8 4 4
3
2 2
9
10
13
10 10 10
10 11
i _
10
5
4
5
11 10 10 10 6 6
12
5
8 8
5 7
13
11 12 10 11 12
14
13 19
20
19
20
15
4 4 2 6 4
16 6 10
9 7
11
17
12
3
6 2 6
18
4 8 4
9 9
A P PE N D IX B
R E SPO N SE S TO AWARENESS IN Q U IR Y
117
Patient-subjects, reinforced:
1. Did you usually give the first sentence that entered
your mind?
18 yes.
2. How did you go about deciding which of the pronouns
to use?
11 first thing that came to mind.
3 first one I saw.
2 tried to think of friends or family.
1 tried to vary my choice randomly.
1 didn1t know.
3. Which do you think you used most often?
13 I or we.
5 various.
1 definitely not I.
4. Why?
7 thinking of myself so used I, but 5 Ss tried to
avoid it as "self-centered" or "egotistical."
4 first thing came to mind.
7 didn't know.
5. Did you notice anything I did during the test?
11 noticed "mm-hmm."
5 noticed writing.
4 noticed nothing E did.
2 commented at this point on E's patience.
6. Why do you think I did that?
7 felt E was trying to encourage her to "keep going"
or make her feel comfortable by saying "mm-hmm."
1 felt "mm-hmm" meant E was ready for the next
sentence.
2 saw the use of "mm-hmm" as "part of the test," but
with no other insight as to its purpose.
1 thought it a habit of E.
7 didn't know.
7. and 8. Did it have any effect on what you did? What
was the effect?
118
15 felt none of E's behavior affected them in any way.
3 felt "mm-hmm" made them more relaxed and encouraged
to "move along."
9. What was the purpose of all this in your opinion?
14 personality test investigating personal problems,
complexes, personal relations, etc.
4 test of self-centeredness.
2 test of progress in psychotherapy.
1 some kind of attitude test.
Patient-subjects, non-reinforced:
1. 18 yes.
2. 11 first thing that came to mind.
2 whatever felt best with the noun.
1 related sentences to self.
3 tried to vary choice randomly.
1 didn't know.
3. 12 I or we.
6 various.
4. 7 thinking of myself so used I, but all 7 reported
trying to avoid it as "self-centered."
7 didn't know,
4 miscellaneous.
5. 8 noticed writing.
8 noticed nothing E did.
2 commented on E's handing her the cards.
6. 6 felt E's writing merely part of recording responses.
12 felt no significance attached to any of E's be
havior.
7. and 8.
” 16 felt none of E's behavior affected them in any way.
1 felt she must not "go too fast" because of E's
writing.
1 felt nervous about E's recording what she said.
9. 15 personality test investigating personal problems
and personal relationships.
1 test of what a person is thinking.
1 test of progress in psychotherapy.
1 didn't know.
119
Other comments: Most of the patient-subjects expressed an
interest in the "results of the test," and many spoke of
feeling anxiety about how the "test" affected their status
in treatment.
Student-subjects, reinforced
1. 18 yes.
2. 6 first thing that came to mind.
4 first one I saw.
1 related sentences to self.
1 tried to vary choice randomly.
1 whatever felt right.
2 tried to make it like a story.
2 didn’t know.
3. 12 I or we.
6 various.
4. 8 thinking of self is easier so use "I" a lot. 4 com
mented on this being "self-centered," but did not
say they tried to avoid use of "I," except for one
S.
2 first thing that came to mind.
1 tried to "make it impersonal."
7 didn't know.
5. 9 noticed "mm-hmm."
11 noticed writing.
6. 6 felt "mm-hmm" meant E was trying to be encouraging.
1 felt "mm-hmm" was to distract.
3 felt writing merely part of recording responses.
8 didn't know.
7. and 8.
14 felt none of E's behavior affected them in any way.
3 felt "mm-hmm" encouraging and relaxing.
1 felt curious about both mm-hmm" and writing, but
didn't feel it affected her behavior or responses.
9. 13 personality test investigating personal relation-
ships.
1 test of altruism.
1 eye test.
3 didn1t know.
Student-subjects, non-reinforced:
1. 18 yes.
2. 11 first thing that came to mind.
1 first one I saw.
1 related sentences to myself.
2 tried to vary my choice randomly.
2 whatever felt right.
1 didn't know.
3. 11 I or we.
7 various.
4. 5 thinking of self is easier so used "I" a lot; only
1 S avoided it as "self-centered.1 1
1 first thing that came to mind.
3 thinking of males and boyfriends.
1 tried to be impersonal.
8 didn't know.
5. 14 noticed writing.
3 noticed E looking at her.
4 noticed nothing E did.
6. 4 noticed writing as recording responses.
1 felt E wanted to see her expression.
1 felt E did not want to influence her in any way.
10 didn’t know.
7. and 8.
15 felt none of E's behavior affected them in any way.
1 test of difference between males and females.
1 test of how you feel about pronouns'.
1 test of skill and concentration.
1 test of sentence structure.
2 didn't know.
Other comments: The majority of student subjects were
curious as to the nature of the "test" and expressed an
interest in "how I did." Several were critical of the
validity of "investigating personality with pronouns," and
a number of the students displayed some anxiety and a per
sonal interest in psychotherapy, the clinic, and what a
therapist does. Several spoke of having personal problems.
A P P E N D IX C
EX PER IM EN T A L M A TERIA LS
123
R E PEA T ED
HEARD
SAVED
SHOWED
REM INDED
TURNED
SHARED
RAN
TA STED
G U ESSED
R E ST E D
RUSHED
UNDERSTOOD
THOUGHT
L IK E D
D E C ID E D
WORE
WANTED
MET
VOLUNTEERED
PU T
COMPARED
BEGAN
ROAMED
P R A C T IC E D
V E R B S U SED I N T H IS STUDY
S P E N T
D ID
WENT
CHANGED
TOLD
WALKED
SA T
N O T IC E D
MADE
STOOD
L IV E D
CO NTIN U ED
RELAXED
CALLED
D IS C U S S E D
REMEMBERED
ASKED
KNEW
WORKED
R E C E IV E D
S A ID
D R ESSE D
SAW
RETURNED
S T E P P E D
LOOKED
SM IL E D
L IS T E N E D
CHOSE
PR O M ISED
B E L IE V E D
STA RTED
P A ID
LEARNED
EN TERED
IN TEN D ED
RODE
READ
WONDERED
U SED
REACHED
OPENED
F IN IS H E D
R E P L IE D
S T U D IE D
SPO K E
FOUND
S L E P T
E X PEC T ED
WATCHED
C O N SID E R E D
S E N T
CAME
GAVE
W ISH ED
REM A IN ED
H E LPED
SEARCHED
V I S I T E D
T R IE D
TALKED
ATE
LAUGHED
THANKED
STAYED
PR EPA R ED
NEEDED
PLANNED
DREAMT
WROTE
W AITED
SERV ED
MOVED
IN V IT E D
EN JO Y ED
124
INSTRUCTIONS FOR SUBJECTS
I AM DR. [OR MR. ] _____, A STAFF PSYCHIATRIST [OR PSY
CHOLOGIST OR SOCIAL WORKER] AT THIS CLINIC, AND AS YOU
WERE TOLD IN CLASS, WE ARE DEVELOPING A NEW TECHNIC FOR
PERSONALITY INVESTIGATION. [Begin here for patient sub
jects] NOW YOU SEE THESE CARDS HERE HAVE A WORD TYPED
IN THE CENTER OF EACH ONE. WHAT I'D LIKE- YOU TO DO IS
JUST MAKE UP A SENTENCE USING THAT WORD. BELOW THE
CENTER WORD YOU WILL SEE SEVERAL OTHER WORDS. TAKE ANY
ONE AND USE IT TO START YOUR SENTENCE. ANSWER WITH THE
FIRST SENTENCE THAT COMES INTO YOUR MIND, THOUGH THERE
IS NO PARTICULAR HURRY. IF YOU WISH TO COMMENT ON ANY
THING AS WE PROCEED, FEEL FREE TO DO SO. ANY QUESTIONS
YOU HAVE, HOWEVER, WE'LL TALK ABOUT WHEN WE'RE THROUGH.
OK? LET'S GO AHEAD, THEN.
125
Name of E:_____________ Age of S:______ Student S:___
Patient S:
Initials of S: Education: No. months in Rx:
Dx:
Verb . PN Verb PN Verb PN Verb PN Verb PN
Spontaneous comments:
126
Questions:
1. Did you usually give the first sentence that entered
your mind?
2.. How did you go about deciding which of the pronouns
to use?
3. Which do you think you used the most often?
4. Why?
5. Did you notice anything I did during the test?
6. Why do you think I did that?
7. Did it have any effect on what you did?
8. What was the effect?
9. What was the purpose of all this in your opinion?
Other comments:

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Wechsler Scores Vs. Piaget Levels: A Study Of The Cognitive Efficiency Ofinstitutionalized Retardates

PDF

The Relationship Of Dependency To Verbal Learning Without Awareness

PDF

Anxiety Level And The Repression-Sensitization Dimension In Desensitization Therapies

PDF

The Modification Of Partial Reinforcement Effect As A Consequence Of Electrical Stimulation Of The Caudate Nucleus In Cats

PDF

Some Psychological Correlates Of The Process-Reactive Concept Of Schizophrenia

PDF

The Effects Of Cognitive Motivation To Avoid Awareness On Measures Of Performance And Awareness In Verbal Conditioning

PDF

An Effect Of Verbal Conditioning On Nonverbal Behavior

PDF

Reward Expectancy Strength As Related To The Magnitude Of Frustration In Children

PDF

The Effects Of Reinforcement Pattern, Intertrial Interval Regularity, And adaptation On The Orienting Response And Classical Gsr Discrimination Conditioning

PDF

Intellect After Lobotomy In Schizophrenia: A Factor-Analytic Study

PDF

The Effect Of Discriminability On The Partial Reinforcement Effect In Human Gsr Conditioning

PDF

A Study Of The Effects Of Generalized Expectancies Upon Accuracy Of Interpersonal Perception

PDF

Sex-Role Identity, Self-Concept, And Vocational Interests Of Adolescent Hemophiliacs

PDF

The Effects Of Ether And Electroconvulsive Shock On One Trial Appetitive And Adversive Learning

PDF

Inhibitory-excitatory interaction of conditioned autonomic responses through interoceptive and exteroceptive conditioning

PDF

The Effect Of Motor Ability Loss On Cognition And Emotion

PDF

The Effects Of A Self Shock Procedure On Hallucinatory Activity In Hospitalized Schizophrenics

PDF

Human Gsr Classical Conditioning And Awareness Of The Cs-Ucs Relation

PDF

Differential effects of epinephrine and propranolol on shuttle box avoidance learning in rats of different ages

PDF

Classical Discrimination Conditioning As A Function Of Probability Of Reinforcement

Asset Metadata

Creator Caruth, Clifton J. (author)

Core Title Clinical Implications Of Verbal Learning Without Awareness: An Experimental Study

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 Klein, David Ballin (committee chair), Grings, William W. (committee member), Peterson, James A. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-232950

Unique identifier UC11358079

Identifier 6106279.pdf (filename),usctheses-c18-232950 (legacy record id)

Legacy Identifier 6106279.pdf

Dmrecord 232950

Document Type Dissertation

Rights Caruth, Clifton J.

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