University of Southern California Dissertations and Theses

An Experimental Study Of The Effect Of Punishment Of The Expectancy To Stutter On The Frequency Of Subsequent Expectancies And Stuttering

(USC Thesis Other)

An Experimental Study Of The Effect Of Punishment Of The Expectancy To Stutter On The Frequency Of Subsequent Expectancies And Stuttering

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content This dissertation has been
microfilmed exactly as received 67— 1 3 ,7 4 0
CURLEE, Richard Frederick, 1935-
AN EXPERIMENTAL STUDY OF THE EFFECT OF
PUNISHMENT OF THE EXPECTANCY TO STUTTER
ON THE FREQUENCY OF SUBSEQUENT EXPECT
ANCIES AND STUTTERING.
U niversity of Southern California, Ph.D., 1967
Speech Pathology
University Microfilms, Inc., Ann Arbor, Michigan
AN EXPERIMENTAL STUDY OF THE EFFECT OF PUNISHMENT
THE EXPECTANCY TO STUTTER ON THE FREQUENCY OF
SUBSEQUENT EXPECTANCIES AND STUTTERING
fey
Richard Frederick Curlee
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
(Speech)
June 1967
UNIVERSITY O F S O U TH ERN CALIFORNIA
T H E 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 , C A L IF O R N IA 9 0 0 0 7
This dissertation, written by
.................. .Ri c ha r d _ _ F r e d _ e r i c k. C _ y ._ r 1 e.g................
under the direction of his Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by the Graduate
School, in partial fulfillment of requirements
for the degree of
D O C T O R OF P H I L O S O P H Y
d & Z a g * e . ......
Dean
D ate June*.. 1. 9.67
TABLE OP CONTENTS
Page;
LIST OP TABLES iii j
Chapter j
I INTRODUCTION 1 ■
Statement of the Problem
Questions and Hypotheses ;
Importance of the Study I
I
II REVIEW OP LITERATURE 6 j
j
I
Expectancy j
Punishment j
III METHODS AND PROCEDURES.............' 19 j
I i
Overview of Design i
Subjects
Materials and .Apparatus
Procedures
IV RESULTS AND DISCUSSION......................... 30
Expectancy
Stuttering
Supplementary Findings
V SUMMARY, CONCLUSIONS AND IMPLICATIONS ......... 45
Summary
Conclusions
Implications
APPENDIXES................................................ 56
LIST OP REFERENCES.......... ' 77 :
LIST OP TABLES
Table Page
; 1. Expectancies to Stutter During Baseline and
| Extinction of the Experimental Group (E),
Yoked-control Group (S) and Time-control
l Group (T)......................................... 31
; 2. Summary of the Analysis of Variance of
the Difference Scores of Table 1 ................ 33
3. Moments of Stuttering during Baseline
and Extinction of the Experimental Group
(E), Yoked-control Group (S)a and
Time-control Group (T) 36
4. Sentences Stuttered During Baseline and
Extinction of the Experimental Group (E),
Yoked-control Group (S)3 and Time-control
Group ( T) . ....................................... 37
! 5. Summary of the Analysis of Variance of
the Difference Scores of Table 3 .............. 39
; 6. Summary of the Analysis of Variance of
the Difference Scores of Table 4 .............. 41
7. Summary of the Analysis of Variance of
the Adjusted Difference Scores of
Signaled Expectancies to Stutter .............. 43
iii
CHAPTER I
INTRODUCTION
The hypothesis that stuttering is an anticipatory
struggle reaction would appear to dominate the thinking of
contemporary workers in the area of Speech Pathology. This
hypothesis has been formulated in various ways; with or
without the assumption that a predisposition of some nature
is essential to the learning of the anticipatory avoidance
behavior (Johnson and Knott, 1936; Van Riper, 1937; Wischner,
1952; Bloodstein, 1958; Eisenson, 1958; Sheehan, 1958;
Mysak, i960). A central construct among these formulations
is the expectancy to stutter which is posited as a causal
factor in the precipitation of moments of stuttering.
Essentially, it is hypothesized that stuttering is an avoid
ance reaction that results from the stutterer's efforts not
to stutter subsequent to his expectancy to stutter. That is
to say, the speech behavior which the stutterer manifests in
his attempts to avoid an anticipated interruption of his
speech is the phenomenon called stuttering. Consequently,
the less frequently a stutterer expects to stutter, the
less he stutters.
Statement of the Problem
The use of expectancy, as an explanatory concept
1
iwithin a given theoretical formulation, must be separated ;
i i
ifrom the expectancy phenomenon per se. The expectancy phe- ;
Inomenon is an empirically established relationship between
Itwo classes of behavioral responses performed by a stutter- j
ing subject to a stimulus word (Johnson and Sinn, 1937;
Johnson and Solomon, 1937; Knott, Johnson and Webster, 1937;
Milisen, 1938). In general^ these studies have operationally
defined the expectancy phenomenon by having a group of stut
tering subjects indicate whether they think they will stutter
on a given word before they read it aloud. The subjects'
expectations of stuttering have been compared to their sub
sequent performances of reading the word aloud. Without
exception, these studies have found a substantial relation
ship between expectancy to stutter and subsequent speech
performance. It should be noted, however, that the nature
of that substantial relationship was not demonstrated. Thus,
the positing of expectancy as a causal factor in the precipi
tation of moments of stuttering is a rather tentative
assumption inferred from the expectancy phenomenon per se.
Although it has been shown that the disfluencies of
stutterers can be increased or decreased by making various
-types and schedules of reinforcement and/or punishment con
tingent upon the occurrence of the disfluencies (Flanagan,
Goldiamond and Azrin, 1958; Goldiamond, 1965; Martin and
Siegel, 1966a; Martin and Siegel, 1966b), there is no evi
dence in the literature that an attempt has been made to
decrease the frequency of the expectancy to stutter by j
i J
joperant techniques or to explore the effect of a conditioned:
decrease in the frequency of expectancy on the frequency of j
jsubsequent moments of stuttering. ;
Questions and Hypotheses
The general purpose of this study, then, is to
[investigate the effect of punishment of the expectancy to
stutter on the frequency of expectancy and stuttering sub
sequent to punishment. Specifically, it is designed to
answer the following questions:
1. Will the frequency of the expectancy to stutter
decrease following its punishment by electro
shock?
2. Will the frequency of stuttering decrease
following the punishment of the expectancy to
stutter by electro-shock?
The hypotheses generated from these two questions
; were:
1. The mean decrease in the frequency of signaled
expectancies to stutter will differ significantly
(p < .05) among three groups of stutterers who
(a) have been presented shock contingent on the
signaled expectancies (Group E); (b) have been
presented shock not contingent on the signaled
expectancies (Group S); (c) have not been
i shocked (Group T).
4 ;
la. The mean decrease In the frequency of j
signaled expectancies to stutter will be J
significantly greater (p < .05) in Group E •
than in Group S.
i
lb. The mean decrease in the frequency of
signaled expectancies to stutter will be
significantly greater (p < .05) in Group E
than in Group T.
The mean frequency of signaled expectancies to
stutter prior to punishment will be significantly
greater (p < .05) than after punishment in
Group E.
The mean decrease in the frequency of stuttering,
will differ significantly (p < .05) among
Groups E, S, and T.
3a. The mean decrease in the frequency of
stuttering will be significantly greater
(p < .05) in Group E than in Group S.
3b'. The mean decrease in the frequency of
stuttering will be significantly greater
(p < .05) in Group E than in Group T.
The mean frequency of stuttering prior to punish
ment will be significantly greater (p < .05)
than after punishment in Group E.
Importance of the Study
The abundance of various theories of stuttering
which are found in the current literature and the relatively;
;poor therapeutic results, particularly with adult stutterers;,
f |
^attest to the inadequacy of contemporary theories as explana
tory systems and as generators of effective rehabilitative
procedures. The significance of this study lies in the need
to provide empirical data that confirm or refute that
expectancy to stutter is a causal factor in precipitating
moments of stuttering and in the exploration of the efficacy
of utilizing the conditioning of the signaled expectancies
to stutter as a rehabilitative procedure.
CHAPTER II
REVIEW OF LITERATURE
The scope of this survey is limited to a critical
review of the concept of expectancy to stutter and an
exploration of the use of punishment as a methodological
procedure to manipulate behavior. Thus, this chapter has
been divided into two major sections: Expectancy and
Punishment. Within each section appropriate sub-categories
have been utilized to provide a more meaningful framework
for their discussion in relation to the experimental problem
of the current study.
Expectancy
The concept of expectancy is generally differentiat
ed into a generalized and a specific component. In 1937*
Van Riper hypothesized that there was a generalized expec
tancy elicited in the stutterer when he is confronted by a
situation in which he has previously stuttered, and a speci
fic expectancy which is aroused by specific cues associated
with various words. Working within the framework of an in
strumental avoidance paradigm, Wischner (1952) referred to
these components as situational anxiety and specific word
anxiety. Since this differentiation of expectancy is formu
lated in terms of the stimuli which elicit expectancy, it
constitutes an explanatory convenience. That is to say, it
expands the range of phenomena that can be explained by
expectancy rather than describing a qualitative or quantita
tive difference in the presumed internal event of expectancy
within the stutterer.
Generalized Expectancy
There is only one study that has demonstrated a re
lationship between expectancy to stutter and a given situa
tion (Porter, 1939)- Subjects read different 500-word pas
sages in eight different situations judged to be of varying
difficulty for stutterers. Estimates of the percentage of
words on which they expected to stutter were made by the sub
jects prior to each reading. The estimated frequencies var
ied in the same direction as the frequency of stuttering
among the situations; however, there were relatively large
differences between the estimated and obtained frequencies
in a given situation. It should be noted that although such
writers as Wischner (1952) and Wingate (1966) have cited var
ious studies that manipulated the difficulty of reading sit
uations and that measured stuttering as supporting the gen
eralized expectancy concept, even the existence of expectancy
was inferred since expectancy was not a variable in that
research.
Specific Expectancy
The relationship between a signal of expectancy to
stutter and the subsequent occurrence of a moment of
I 8 i
i !
j
[stuttering was first demonstrated by Knott, Johnson and |
[Webster (1937). In this study subjects read a 500-word :
passage one word at a time immediately after they had sig
naled their expectancy. It was found that stuttering
i
followed a signaled expectancy 9^ to 96 per cent of the time,
while no expectancy of stuttering was followed by stuttering
: 0.4 to 3.0 per cent of the time.
The results of two subsequent investigations corrob
orated the findings of the original study in addition to
attempting to control the effect of signaling an expectancy
■to stutter (Johnson and Sinn, 1937; Milisen, 1938). Johnson
and Sinn’s subjects read a 1000-word passage aloud. On a
subsequent day the subjects re-read the same passage, but
were instructed to omit reading every word that was expected
to be stuttered. When compared to the initial reading of
the passage, it was found that during the "control" reading
conditions: (1) stuttering was reduced by 98 per cent, and
(2) both expectancy to stutter and unexpected moments of
stuttering tended to occur on words stuttered during the
initial reading. Essentially, Milisen’s study differs from
Johnson and Sinn’s in that an additional reading condition
was employed In which the subjects read the word aloud after
Signaling their expectancy to stutter. While his data
generally supported the results of Johnson and Sinn, he also
■reported that: (1) 97 per cent of his subjects did not pre
dict every moment of stuttering; (2) the median subject
I 9 |
(predicted but 6l per cent of his moments of stuttering; (3) j
I i
stuttering is increased when it is preceded by signals of
'expectancy to stutter; and (4) most subjects do not stutter i
‘ following every signal of expectancy. Neither of these
studies effectively controlled several variables that may
have confounded the results. For example, neither study
Separated the effects of instructions from the effects of
the tasks nor the effects of the signaling of expectancy
from the effects of the inferred internal event of expectancy
between conditions. From the results of these studies, an
Inference that expectancy to stutter is a causal factor in
the precipitation of moments of stuttering would, at best,
be speculative.
The question of whether stutterers’ signaled
expectancies exhibit an adaptation effect is still unanswer
ed. Wischner (1952) reported that the frequency of signals
of expectancy to stutter decrease, or adapt in much the same
manner as stuttering across five consecutive markings of the
same passage. Peins (1961a), however, was not able to
demonstrate an expectancy adaptation curve. Interestingly
enough, her subjects were also much less accurate in pre
dicting moments of stuttering than were subjects in the
previously reported studies. In noting this feature of her
data, Peins suggested that since her subjects were experi
mentally naive while those of the earlier studies were
sophisticated, subject variation may have accounted for the ;
l O i
i
i
I
difference in results. |
Although they are of little relevance to the present!
'study, it should be noted that several investigations of
;other aspects of expectancy to stutter have been performed. ;
Johnson and Solomon (1937) found a substantial relationship
between signaled expectancies to stutter and the occurrence
I
'of moments of stuttering even though the readings were per
formed up to seven days after expectancies were recorded.
IVan Riper and Milisen (1939) reported that stutterers can
predict not only moments of stuttering but also whether its ;
duration will be short, medium or long. Johnson and ;
iAinsworth (1938) and Peins (1961b) found that signaled
expectancies to stutter on a given passage remain relatively
■consistent for as long as six weeks. Finally, Van Riper
(1936) reported that the inspiration-expiration duration
ratio measured immediately after the presentation of a word
:stimulus could differentiate between words that were
lexpected to be stuttered and those not expected to be
'stuttered.
Summary
Overwhelming support exists in the literature of a
^substantial relationship between signaled expectancies to
'stutter and subsequent occurrences of moments of stuttering.
The operations of signaling expectancies to stutter would
;also appear to be associated with an increase in the
11
frequency of moments of stuttering. In addition to having
it least one physiological correlate, expectancy would
jappear to be a relatively consistent and enduring phenomenon!
I :
tThe data also tend to support the inference that signaled
i |
Expectancies are based on some type of internal event ;
t ;
experienced by stutterers prior to stuttering. That expec- ;
i
jfcancy to stutter is a causative factor in the precipitation :
i
i
pf moments of stuttering is supported primarily by evidence ;
it hat: (1) it precedes stuttering in time; (2) it is associ- '
ated with stuttering to a greater extent than can be
accounted for by chance; and (3) its absence is associated
with no stuttering more often than can be accounted for by
chance. It would appear, therefore, that an experimental
manipulation of the frequency of signaled expectancies would!
constitute a more appropriate test of the assumption of
causality than has heretofore been attempted.
Punishment
i
; Prom its first formulation in the literature in 19131
! :
jby Thorndike, the concept of punishment has received exten
sive theoretical elaboration and empirical investigation.
Although considerable uncertainty currently exists in the
idevelopment of an adequate theoretical explanation of some
i ;
iof the varied effects of punishment (Church, 1963; Azrin and:
:Holz, 1966), there is little doubt that punishment can be a
most effective procedure In the elimination or suppression j
i 12l
i j
|of behavior (Franks, 1961; Church, 1963; Holz, Azrin and ,
I I
! >
lAyllon, 1963; Azrin and Holz, 1966; Kushner and Sandler, j
• i
|l966). Following a discussion of the definition of punish- j
Iment, the remainder of this survey is devoted to a relatively
| ;
ibrief review of variables relevant to the present study that,
have been found to enhance the effectiveness of punishment.
^ Definition of Punishment
As noted above, the first definition of punishment
was presented by Thorndike (1913) as a decrease in the
strength of a connection between a situation and a response
’that is accompanied or followed by an "annoying state of
:affairs.” This subjective type of definition was subse- ;
iquently modified by a type of definition that infers the
occurrence of an internal event, perhaps best exemplified
;by Dollard and Miller’s (1950) positing of punishment as a
drive variable. A third type of definition defines punish
ment in terms of observable variables. Skinner (1953) and
■Dinsmoor (195^) defined punishment as the presentation of
an aversive stimulus following a response. An aversive
stimulus was then defined as a stimulus that increases the
; probability of a response which terminates that stimulus.
Thus, this definition requires the demonstration that a
given stimulus serves as a negative reinforcer of escape
behavior before it can be utilized as a punishing stimulus.
;A much more simple and practical definition, especially for
[research, has been offered by Azrin and Holz (1966) wherein j
I j
jthey define punishment as a reduction in the future proba- ;
toility of a given response as the result of an immediate ,
presentation of a stimulus contingent on that response.
Thus, the contingent stimulus is referred to as the punish- ;
ing stimulus, while the entire procedure is called punish
ment. It is this definition that is utilized in the present
investigation.
Type and Intensity of Punishment
Various types of stimuli have been employed in
punishment procedures, although electric shock is undoubtedly
the most widely used in human research (Solomon and Brush,
1956). The primary advantages of shock are that it has
minimal adaptation properties, provides maximal control, and
has produced definitive changes in behavior. It should be
noted, however, that the selection of a punishing stimulus
is influenced, if not determined, by other related factors.
For example, the subject should not be able to
utilize some form of escape behavior to eliminate or mini
mize the effect of the punishing stimulus (Dinsmoor and
Campbell, 1956; Azrin and Holz, 1966). That is to say, a
subject should not be able to shut his eyes if the punishing
stimulus is a bright flash of light or pull away from the
electrodes of a shock device.
Another important variable is the intensity of the
punishing stimulus. It has been found that the greater the
l4i
intensity of the punishing stimulus, the greater the reduc-
i I
jtion in the frequency of the response being punished (Estes,j
; I
119^ 4 ; Dinsmoor, 1952 ; Karsh, 1962 ; Azrin, Holz and Hake, i
;1963)- The implications of these findings pose serious
•problems for human research. For example, in two studies
which punished human subjects with noise (Azrin, 1958 ;
Flanagan, Goldiamond and Azrin, 1958 ) , complete response
suppression was not achieved even though intensities as
high as 135 decibels were utilized.
Immediacy of Punishment
The importance of the temporal relationship between
the occurrence of the response and the presentation of the
punishing stimulus has been confirmed on several occasions
(Mowrer and Uliman, 19 ^ 5 ; Lichtenstein, 1950 ; Azrin, 1956 ;
Azrin,1958 ; Kamin, 1959 ) . In general, It can be said that the'
longer the interval between the emission of the response
and the onset of the punishing stimulus, the less effective
the punishment procedure for the reduction of the frequency
of the response.
Schedule of Punishment
The effects of partial and continuous reinforcement
on the acquisition and/or extinction of behavior have been
well documented (Ferster and Skinner, 1957). The effects
of partial and continuous punishment are not so clear cut.
Continuous punishment produces more response suppression
I 15;
! '
!than does intermittent punishment; however, intermittent
ipunishment is followed by a more gradual recovery of
responses than is continuous punishment after the termina-
ition of the punishing stimulus (Estes, 1944; Azrin, Holz
and Hake, 1963; Melvin, 1964). Thus, continuous punishment
would appear to have a more extensive suppression effect
initially, while the effect of intermittent punishment is
more durable. Church (1963) has suggested that, with human
subjects, continuous punishment should be presented initi
ally, followed by an intermittent schedule of punishment.
■ Alternative Response and
Punishment
In a two-choice situation, there will be a greater
reduction in the frequency of the punished response than in
a one-choice situation (Church, 1963; Herman and Azrin,
1964; Azrin and Holz, 1966). That is to say, when a subject
has an alternative response available, to him that is not
punished, the punished response will evidence greater sup
pression than if the only response he can make in a given
situation is punished.
Punishment and Reinforcement
Investigations with both humans and animals have
indicated that a punishing stimulus, after it has been
paired with a positive reinforcer, may produce an increase ■
:in subsequent responses (Holz and Azrin, 1961; Sandler,
1964; Ayllon and Azrin, 1966). Although this apparently
16
paradoxical effect of punishment may be interpreted as an
example of a punishing stimulus acquiring conditioned rein
forcement properties, or becoming a discriminative stimulus,
the researcher who is interested in reducing the frequency
of a response must be careful that unplanned reinforcement
does not occur contiguously with the punishing stimulus.
Immediacy and Permanence of
Suppression and Punishment
Virtually all studies agree that any reduction in
the frequency of the occurrence of a response is immediate
if the punishing stimulus is to be effective (Estes, 1944;
Dinsmore, 1952; Azrin, I960). Thus, the experimenter is
able to ascertain in a relatively brief period of time if
the procedure he is employing is to be effective in sup
pressing the desired response.
Although the permanence of the suppression of a
response is directly affected by such factors as the inten
sity and schedule of the punishing stimulus, the permanence
of the suppression also appears to be related to the degree
of suppression. Once the frequency of the response has been
reduced to zero over an extended period of time, it is
virtually an irreversible effect (Masserman, 1946;
Lichtenstein, 1950; Azrin, I960). Again, this finding
would appear to pose a serious problem for human research;
for it would be unwise, at best, to permanently eliminate a
\ t
response of any human subject until the possible effects of j
' i
: i
jthat elimination on other behavior could be determined. I
Undesirable Aspects of Punishment ;
• )
After an extensive review of the literature on
punishment, Azrin and Holz (1966) presented an extended and,
at times, somewhat speculative discussion of some of the
undesirable effects that punishment could produce on human
;behavior. They concluded that punishment does not appear to
produce any enduring, undesirable emotional states. It may,
ihowever, produce three types of disruptive social behavior: ,
!
(1) it reinforces tendencies of the subject to escape from
:the situation or interpersonal relationship in which the
punishment is administered; (2) aggression by the subject
may be directed toward the punishing agent; and (3) aggres
sion may be directed toward 'any convenient organism or
individual. These factors are particularly important with
humans, since the effective functioning, if not the survival,
of a human being is quite dependent on the maintenance of
harmonious social relationships.
Summary
The literature on punishment suggests several guide
lines in order to maximize the effectiveness of punish
ment. The intensity of the punishing stimulus should be as
high as possible with no escape responses available to the
subject. The response to be suppressed should be followed
iimmediately by the punishing stimulus, and the contiguity j
I ’
I !
jbetween response and punishing stimulus should be continuous!
^initially, after which a partial schedule of punishment can ■
; i
!be used. An alternative response to the punished response
^should be available to the subject so he can avoid punish
ment, and the presentation of the punishing stimulus should
not be contiguous with positive reinforcement. Although
the effects of punishment are immediate and can be relatively
permanent, there are certain undesirable forms of social
behavior which may result from punishment that should be
considered.
! CHAPTER III ;
METHODS AND PROCEDURES j
j :
Overview of Design
; This was an experimental study which was designed to
'investigate the effect of punishment of signaled expectan
cies to stutter on the frequencies of subsequent signals of
‘ expectancy and of stuttering. Thus, punishment of signaled
expectancies to stutter was the independent variable, while
;the frequencies of the signaled expectancies and of stutter-
ling were the dependent variables.
Originally, it had been planned to have the subjects-
indicate their expectancies to stutter prior to saying a
stimulus word from the Thorndike-Lorge (1944) word list.
This procedure was attempted with a pilot group of four
stutterers, and it was found that none of the subjects
^signaled an expectancy to stutter or stuttered after five
Iminutes of performing this task. Since this design required
la relatively consistent frequency of signaled expectancies
to be emitted for approximately two hours, the original task
was modified to one in which the subjects utilized the
stimulus word in composing a five to six word sentence and
then indicated their expectancies to stutter on the sentence
prior to their saying it. The efficacy of this procedural
19
jchange was confirmed with two of the four pilot subjects, j
i i
land the modification was incorporated into the final design.;
In its final form, this study employed three groups
jof stuttering subjects who indicated if they expected to
jstutter before they attempted to say a sentence which con
tained a stimulus word from the Thorndike-Lorge (19^4) word
list. After a baseline of the frequencies of the groups’
signaled expectancies and stuttering had been obtained, the
experimental group was punished with electric shock immedi
ately after an indication of an expectancy to stutter.
Subjects in the yoked control group received an equal number;
;of shocks, but the shocks were not contiguous to the signal
ing of expectancy. In neither group was the administration
of shock contiguous to moments of stuttering. Subjects in
the time control group were not shocked. The three groups
then underwent extinction procedures which replicated those
of the baseline period. The frequencies of signals of
expectancy and of stuttering obtained during the baseline
and extinction procedures were computed; individual and group
results reported; and the mean differences between the
frequencies of signaled expectancies and of stuttering
obtained during the baseline and extinction periods were
compared within and among the groups.
Subj ects
Twelve male subjects were selected in the following
manner from an available population of stutterers who had
211
i
i
japplied for therapy at the Center for the Study of Speech j
| |
land Hearing. Originally, thirty subjects were contacted by i
1
^telephone or in person by the experimenter and informed of
|the general nature of the study. Twenty-two of the thirty
|volunteered to participate. All subjects who volunteered
underwent the Practice and Baseline procedures, discussed in
^greater detail in a later section. At the end of the Base
line procedure, the experimenter counted the frequency of
signaled expectancies which had occurred. If the subject
had signaled that he expected to stutter ten or more times,
the experiment continued as planned. Otherwise, the pro
cedures were stopped, and the subject was thanked for his
participation. As a consequence of this screening procedure,
ten of the original twenty-two volunteers were eliminated
from the study. Of the remaining twelve, the first four
who passed the screening constituted the experimental group,
while the remainder were randomly assigned to yoked control
and time control groups of four subjects each. Since a
sample of twelve stutterers was not immediately available
and had to be accumulated over a relatively long period of
time (nine months) random assignment of subjects to the
experimental group was not employed in order to insure the
accumulation of an adequate number of subjects. The final
sample ranged in age from seventeen to thirty-six with a
mean age of 27.75 years in the experimental group and . 25.50
years in the two control groups.
Materials and Apparatus
Prom the list of the 1,000 most frequently used
words (Thorndike-Lorge, 1944), 600 were chosen at random,
using a table of random numbers, as stimulus words. Eighty
of these 600 were then randomly assigned to two groups of
forty words each, Experimental Words I and II. (See
Appendix A.) The forty words within each of the Experimen
tal Words groups were placed in random orders and each group
further divided into four groups of ten words each. The
order of presentation of these four groups of words within
each Experimental Words groups was assigned to the four
subjects within each group according to a latin squares
design (Edwards, 1963). In addition, Experimental Words I
and Experimental Words II were alternately assigned between
the Baseline and Extinction procedures among the subjects
in all groups so that any systematic effect of the words was
counterbalanced. From the remaining 520 of the original
600 stimulus words, 120 were selected at random as Practice
Words for use during the Practice procedure. The remaining
400 constituted the Training words and were utilized during
the Conditioning procedure. The order of presentation of
both the Practice Words and the Training Words was assigned
in a random manner. Finally, transparencies of each of the
600 words were individually mounted in 2" x 2” slide holders.
23
The experiment took place at the Center for the
Study of Speech and Hearing. A specially constructed elec
tro-shock device provided the shock. The shock apparatus
consisted of two small wooden boxes (3" x 8") with formica
tops which contained ten, 9-volt D.C. transistor batteries
soldered in series. One end of the series of batteries was
connected to a resistor which was controlled by a small knob
protruding from one of the boxes. By turning this knoh the
amount of D.C. current could be increased or decreased.
Leading from the resistor were two strands of shielded
copper wire connected to two copper pennies which served as
electrodes. The two copper pennies, in turn, were attached
to an adjustable rubber strap. At the other end of the
series of batteries, two wires were soldered to the termi
nals of a red button which, when depressed, closed the
circuit and produced the D.C. current to the electrodes. In
addition, the experimental room contained a Sony tape
recorder which recorded each session; a table on which was
mounted a red and a green light; a desk; three chairs; two
electronic signaling devices; and two projectors. The
experimenter and an assistant were present during the ses
sions .
Procedures
Each subject entered the experimental room and was
seated at the table. After his right forearm was connected
to the electro-shock device, the intensity of the shock was
gradually increased until the subject reported that it was
j 241
i j
jpainful. He was then told that he might be shocked at
jvarious times throughout the experiment. Each subject was I
jinstructed that he would see a number of different words, ;
I ‘
ipresented one at a time, on the wall in front of him; that
he was to compose a five to six word sentence which contain-;
ed that word; and that if he expected to stutter on the
sentence, he was to press the button in his right hand
(expectancy button) as soon as the red light came on. Con
versely, he was to press the button in his left hand (no
expectancy button) if he did not expect to stutter. He was
told that speed was of the utmost importance and that he
would have only three seconds to compose a sentence and
.decide if he expected to stutter if he were to say the
isentence. In order to prevent the association of shock with
^verbalizations, it was then emphasized that he was to say
the sentence only if the green light came on and that he was
not to say the sentence if the green light did not come on.
At the conclusion of these instructions, the tape recorder
was turned on to record the session, and the experimental
procedures began.
Practice
Stimulus words from the Practice Words list were
;projected on the wall in front of the subject for ten min
utes. Three seconds after the presentation of each word
the red light came on. Immediately after the subject
pignaled if he did or did not expect to stutter, the green
i !
light was turned on approximately 75 per cent of the time,
put remained off the remainder of the time. These procedures
i
hot only permitted the subjects to practice the tasks he had;
i
been instructed to perform, but it also provided the experi
menter with the opportunity to observe that each subject was
performing his task adequately before beginning the proce
dures from which the data were gathered. In addition, since
practice continued for ten minutes, the relatively complete
adaptation of stuttering should have occurred (Cohen, 1955).
Baseline
Each subject was presented one of the groups of 40
Experimental Words. Immediately after he pressed his expec
tancy or no expectancy button following the onset of the
red light, the green light was turned on, and he said the
sentence he had composed.
Conditioning
The 400 Training Words were then presented in their
entirety or until two hours had passed, whichever came
first. The latter alternative was utilized for two subjects
in order to eliminate excessive fatigue factors. Subjects
in the experimental group were shocked immediately after
they pressed their expectancy button according to the
following Variable-Ratio schedule:
1st 50 words .
^5.
O
o
i —i
•
punishment
2nd 50 words . . 40$ punishment
3rd 50 words . . 20$ punishment
4th 50 words . . 60$ punishment
5th
50
words . . 10$ punishment
6th 50 words . . 50$ punishment
7th 50 words . . 80$ punishment
8th 50 words . . 100$ punishment
Whenever a subject was shocked, however, the green light did
not come on, and he did not say the sentence. Whenever a
subject pressed his no-expectancy button, the green light
came on, and he said the sentence.
Each subject in the yoked-control group was paired
;in a random manner with a subject in the experimental group
and received the same number of shocks within each 50 word
grouping as the subject with which he was paired. The
shock, however, was administered randomly at least five
seconds after he had said a sentence and five seconds before
the presentation of a stimulus word. In addition, the pre
sentation of the green light to subjects in the control
groups was determined by the performance of the subjects
with which they were paired in the experimental group. That
lis to say, following those stimulus words that a given
experimental subject did not verbalize a sentence (since he
was shocked), his paired control subjects also did not
27;
i
I
s
jverbalize a sentence. In effect, this tended to equalize !
jthe amount of verbalizations among the groups. !
Subjects in the time control group underwent
Identical procedures as the subjects in the yoked-control
group with the exception that they were not shocked. It
should be emphasized at this point that the presentation of
; shock was not contiguous with moments of stuttering in any
of the groups.
l Extinction
The Extinction procedure differed from the Baseline
procedure described above only in that each subject was
presented the group of forty Experimental Words which were
not presented to him during the Baseline procedure.
Following the Extinction procedure, the experimenter
disconnected the subject from the electro-shock device,
thanked him for his participation, and left the room. The
assistant then asked the subjects in the experimental group
the questions contained in Appendix B. The purpose of
these questions was to ascertain if subjects in the experi
mental group avoided pressing the expectancy button to
escape the shock even though they still expected to stutter.
Responses
Each subject's signals of expectancy to stutter
were recorded on a specially constructed score sheet by the
experimenter during the experimental procedures. i
I I
jDefinitions of stuttering were obtained from judgments made j
jof the tape recorded sentences of the subject. Three i
'^graduate student clinicians at the Center for the Study of
;Speech and Hearing, who have had extensive clinical experi- '
;ence with stutterers, served as judges. Following the
administration of instructions (see Appendix C), the judg
ements were performed in two 3-hour sessions. Agreement of
two of the three judges on the number of moments of stut
tering that occurred during a given sentence constituted
the definitions of moments of stuttering. From these defini
tions, two measurements of the frequency of stuttering were :
^obtained: (a) the number of moments of stuttering that
occurred during the Baseline and Extinction procedures; and ;
■(b) the number of sentences that contained stuttering during
the same procedures.
Evaluation
The analysis of the data followed the recommendation
of Edwards (1963). The frequency of signaled expectancies
to stutter that were recorded during Baseline and Extinction
was computed, and the differences between these frequencies
were obtained. Bartlett’s test was performed to test the
assumption that the samples were drawn from populations with
equal variances. An analysis of variance was then performed
.to evaluate the differences among the three groups, while
j
t tests were utilized for specific comparisons between and
within groups. The frequency of stuttering measures were j
I
evaluated in the same manner. The .05 level was accepted asj
! 1
significant in all of the analyses. Although the compari-
isons among the three groups constituted two-tail tests, a
one-tail test was utilized for the between and within group
comparisons.
i CHAPTER IV
I RESULTS AND DISCUSSION j
Expectancy
The subjects' frequencies of signaled expectancies
to stutter obtained during the Baseline and Extinction pro
cedures, difference scores and group totals and means are
presented in Table 1. It should be noted that the differ
ence scores were obtained by subtracting Extinction fre
quencies from Baseline frequencies. Thus, a positive number
in the difference column indicates that a decrease occurred
from Baseline to Extinction. Conversely, a negative number
indicates an increase in frequency between the two proce
dures. In order to be assured that the separate group
variances of the difference scores did not violate the
assumption of homogeneity, Bartlett's test for homogeneity
|of variance was employed before subjecting the data to the
analysis of variance (Edwards, 1963). A summary of the
computations employed is presented in Appendix D. The
obtained x2 °f 2.6055 with 2 degrees of freedom was well
below the tabled x2 value of 5.991 at the .05 level of
confidence. Consequently, the data offer no significant
evidence against the assumption that the within group
variances were drawn from populations with eq.ual variances.
30
TABLE 1
I
; Signaled Expectancies to Stutter During Baseline and
! Extinction of the Experimental Group (E),
; Yoked-control Group (S), and
| Time<-control Group (T)
I ______________________________________________________________________
:Sub j ect s Baseline Extinction Difference:
i El 12
5 7
E2 14 0 14
E3 39 0 39
E4 30 0 30
Total 95.00 5.00 90.00
Me an
23.75 1.25 22.50
SI 18 10 8 !
S2
23 30 -7
S3 13 18
-5
S4 22 24 -2
Total 76.00 82.00 -6.00
Mean 19.00 ro
o
o
-1.50
T1 40 40 0
T2
15 3
12
T3 17 20
-3
T4
.39 . 40 -1
Total 111.00 103.00 8.00
Mean 27.75 25-75
2.00
32
The summary of the analysis of variance of the
difference scores of signaled expectancies to stutter is
shown in Table 2. The obtained F ratio of 6.6t with 2 and
9 degrees of freedom was significant at the .025 level of
confidence. Assuming that the groups' variances are homo
geneous, an P ratio this large is indicative that the mean
difference scores among the three groups differ signifi
cantly. Thus, Hypothesis 1 was supported.
The results of t tests performed after an analysis
of variance should not be interpreted as if no test had
preceded them (Guilford, 1956)., Since the null hypothesis
that the means do not differ significantly at the .05 level
of confidence has already been rejected, all interpretations
of subsequent tests must be considered from this frame of
reference. Consequently, Guilford proposes the use of the
within groups variance in the computation of the standard
error of the difference between means for t_ tests performed
subsequent to an analysis of variance. In essence, these
modifications provide a more stable estimate of the popula
tion variance and increase the power of the statistical test
for detecting differences between the means being compared.
Hence, with appropriate reservations, t_ tests were employed
to compare the means between these groups which specifically
relate to the hypotheses stated in Chapter I. First, the
mean difference score of the experimental group (E) was
compared with that of the yoked-control group (S).
35,
TABLE 2
Summary of the Analysis of Variance
Difference Scores of Table 1
of the
Source SS d.f. MS F
Between groups 1344.67 2 672.34 6.64*
Within groups 912.00 9 101.33
Total 2256.67 11
:* p < .025 > .01
...... 34!
i
Utilizing a one-tail test with 9 degrees of freedom, the ;
obtained t value of 3-37 was significant at the .005 level !
!of confidence. In like manner, the mean of the experimental;
group (E) was then compared to that of the time-control
■group (T). This comparison yielded a t of 2.88 which was
significant at the .01 level. The results of these specific1
t tests, therefore, would appear only to confirm the results
of the analysis of variance. Thus, both Hypotheses la and
lb were supported.
One final comparison was performed within the
experimental group. This comparison, of course, was not
involved in the original analysis of variance and should not
be interpreted as if a test had preceded it. Since the mean
'frequencies of signaled expectancies to stutter during
Baseline and Extinction should be expected to be correlated,
a convenient way of conceptualizing this relationship is as
paired observations (Hays, 1963). In analyzing the mean
difference of paired observations, a significant t_ ratio
should be interpreted as indicating that a significant
difference exists between the paired means. The obtained t_
of 3.07 with 3 degrees of freedom was significant at the .05
ilevel of confidence. Thus, Hypothesis 2 was supported.
Stuttering
In order to provide a more comprehensive evaluation
;of the frequency of stuttering two measures were obtained
35!
from the operational definition: (a) the number of moments I
i
of stuttering that occurred during the Baseline and Extinc- '
■tion procedures; and (b) the number of sentences that con
tained moments of stuttering during the same procedures.
Summaries of the data obtained by these two measures are
presented in Table 3 and Table 4, respectively. It had
been anticipated that the frequency of sentences stuttered
measure would be more stable and that the moments of stutter
ing measure would be more sensitive to any changes which
might occur. Again, it should be noted that a positive
number in the difference columns is indicative that a de
crease in the frequency of stuttering occurred between the
Baseline and Extinction procedures. As with the expectancy
idata, the difference scores of both measures were subjected
to Bartlett's test for homogeneity of variance (Edwards,
1963). Summaries of these computations are presented in
Appendix D. For the moments of stuttering measure a x2
of 2.998 was obtained, while the sentences stuttered measure
yielded a x2 if 2.162. Thus,both of the obtained values of
X25 with 2 degrees of freedom each, were comfortably below
the tabled value of 5-991 associated with the .05 level of
confidence. Consequently, the assumption that the separate
!group variances of the difference scores were drawn from
populations of equal variances was supported. The results
Iof the statistical analyses of moments of stuttering will
;be presented first, followed by those of sentences stuttered!.
TABLE 3
Moments of Stuttering During Baseline and
Extinction of the Experimental Group (E)3
Yoked-control Group (S) and
Time-control Group (T)
Subjects Baseline Extinction Difference
El
35" 19
16
E2 24
9 15
E3 127 69
58
E4 32 10 22
Total 218.00 107 .00 111.00
Mean
54.50 26.75 27.75
SI 18 10 8
S2
55
52
3
S3
30 20 10
S4 40
49 -9
Total 143.00 131.00 12.00
Me an
35.75 32.75
3.00
T1 126 116 10
T2 64
73 -9
T3
4 1
3
T4 34
25 9
Total 228.00 215.00 13.00
Mean 57.00
53.75 3.25
37!
TABLE 4
Sentences Stuttered During Baseline and
Extinction of the Experimental Group (E),
Yoked-control Group (S)s and
Time-control Group (T)
|Sub j ects Baseline Extinction Difference
El
E2
E3
E4
Total
Mean
24
17
40
24
105.00
26.25
16
8
33
10
67. 00
16.75
8
9
7
14
38.00
9.50
51
52
53
54
Total
Mean
15
31
23
24
,93.00
23.25
10
33
16
29
88.00
22.00
5
-2
7
■5
5.00
1.25
T1
T2
T3
T4
40
35
4
27
40
36
1
23
0
-1
3
4
Total
Mean
106.00
26.50
100.00
25.00
6.00
1.50
38!
I
i A summary of the analysis of variance of the differ-!
| i
; ence scores of the frequencies of moments of stuttering is
presented in Table 5. The obtained F ratio of 4.29 with
1 :
2 and 9 degrees of freedom was significant at the .05 level .
of confidence. Since Bartlett’s test for homogeneity of
variance found no significant differences among the within
group variances, an F ratio of this size is indicative that
the means among the three groups are significantly different.
Thus, Hypothesis 3 was supported.
The subsequent use of t_ tests with these data is
subject to the same problems encountered with the expectancy'
data; therefore Guilford's (1956) modified computational
procedures were employed again. The mean difference score
of group E was compared with that of group S. Employing a
one-tail test with 9 degrees of freedom for each comparison,;
the obtained t value of 2.55 was significant at the .025
level of confidence. The comparison of the mean differences
between groups E and T produced a t_ of 2.52, which was
significant at the .025 level. Thus, Hypotheses 3a and 3b
Were supported.
The within group E comparison employed Hays (1963)
;t test for paired observations. With 3 degrees of freedom;
and employing a one-tail test, the obtained t_ of 2.72 was
significant at the .05 level. Thus, Hypothesis 4 was
supported.
39 i
1
f
!
I
TABLE 5
Summary of the Analysis
Difference Scores
of Variance
of Table 3
of the
...
Source . SS d.f. . MS . . . P.
Between groups
1617.17
2
808.59 4.29*
Within groups 1695.50
_9
188.39
Total
3312.67
11
* p < .05 > .025
'.... 4o]
j
The summary of the analysis of variance of the j
difference scores of frequency of sentences stuttered is \
jpresented in Table 6. This analysis produced an F of 5-55 •
jwhich, with 2 and 9 degrees of freedom, was significant at
the .05 level of confidence. Again, Hypothesis 3 was sup
ported.
Since the results of the analysis of variance were
indicative of a significant difference among the groups1
means, Guilford's (1956) modified computational procedure
was employed for the comparisons of interest. In each
instance a one-tail test with 9 degrees of freedom was
utilized. The—comparison between group E and group S pro
duced a t_ of 2.95 which was significant at the .01 level;
■while that between groups E and T yielded a t of 2.86 which
was significant at the .01 level. Consequently, Hypotheses
3a and 3b were again supported.
As before, Hays' (1963) test of paired observations
was used to evaluate the mean difference within group E.
The obtained t value of 6.13 with 3 degrees of freedom pro
vided a significant one-tail test at the .005 level of
confidence. Thus, Hypothesis 4 received additional support.
! Supplementary Findings
At the conclusion of the experimental procedures,
each subject in group E was administered a questionnaire.
'(See Appendix B.) The verbatim responses of the subjects
4 T i
i
TABLE 6
; Summary of the Analysis of Variance
Difference Scores of Table 4
of the
Source SS d.f. MS F
Between groups 176.17 2 88.09 5.55* :
jWithin groups
142.75 9
15.86
Total 318.92 11
* p ■ < .05 > •025
M2
are presented in Appendix B. Of primary interest to this
study were the responses to question number 3. Two subjects
reported that the presentation of shock did not prevent them
from making honest reports of their expectancies to stutter.
The other two subjects, however, estimated that they had ex
pected to stutter three and six times respectively during
the presentation of the final MO stimulus words. Conse
quently, these estimates were added to the obtained frequen
cies of signaled expectancies to stutter and subjected to
the same analyses as the original data.
A summary of the analysis of variance of the adjust
ed difference scores of signaled expectancies to stutter is
presented in Table 7. The obtained F ratio of 5-^7 with 2
and 9 degrees of freedom was significant at the .05 level of
confidence. The specific comparisons between the means of
groups E and S yielded a t of 3.08 which, with 9 degrees of
freedom was significant at the .01 level. Evaluated in the
same manner, the mean difference between groups E and T pro
duced a t_ value of 2.58 which was significant at the .025
level. Within group E, the comparison of the mean differ
ence of the paired observations yielded a t_ of 2.80 with 3
degrees of freedom which was significant at the .05 level of
confidence. Thus, the statistical analyses of the adjusted
difference scores of signaled expectancies to stutter
supported Hypotheses 1, la, lb, and 2.
Returning briefly to the responses of the subjects
to the questionnaire, it is difficult to understand why none
43!
TABLE 7
Summary of the Analysi.3 of Variance of the
Adjusted Difference Scores of Signaled
Expectancies to Stutter
! Source
1
SS d.f. MS
F .
Between groups
1091.17
2
545.59 5.47*;
Within groups
897.75 _9 99.75
Total 1988.92 11
p < .05 > .025
of the subjects in the experimental group verbalized the I
! contingency between signaled expectancies to stutter and ■
’ the presentation of shock. Although this is of no concern |
jto the problem being investigated, it does constitute a
jfootnote of interest. That one of the four subjects indi
cated he would not be willing to be shocked even if it
alleviated his stuttering, is of significance to the impli
cations of this research in terms of the development of
rehabilitative procedures and will be discussed in a later
section.
In summary, the results of this investigation of the
relationship between signaled expectancies to stutter and
subsequent moments of stuttering supported all of the
^hypotheses stated in Chapter I.
CHAPTER V
SUMMARY, CONCLUSIONS AND IMPLICATIONS
Summary
The Problem
1 The purpose of this study was to investigate the
effect of punishment of the expectancy to stutter on the
frequency of expectancy and stuttering subsequent to punish
ment. Arising from the assumption of the anticipatory-
struggle theories of stuttering that expectancy to stutter
is a causal factor in the precipitation of moments of stut
tering were the following questions: (1) Will the frequency
of expectancy to stutter decrease following its punishment
by electro-shock? (2) Will the frequency of stuttering
decrease following the punishment of the expectancy to stut
ter by electro-shock? The principal hypotheses generated
from these questions were: (1) The mean decrease in the
frequency of signaled expectancies to stutter will differ
significantly (p < .05) among three groups of stutterers who
have (a) been shocked contingent to signaled expectancies,
(b) been shocked not contingent to signaled expectancies,
(c) not been shocked. (2) The mean frequency of signaled
expectancies to stutter prior to punishment will be signi
ficantly greater (p < .05) than after punishment. (3) The
45
| 46;
i j
jmean decrease in the frequency of stuttering will differ
j |
Isignif icantly (p < .05) among the same three groups. (4) !
The mean frequency of stuttering prior to punishment will be;
isignificantly greater (p < .05) than after punishment.
Method
Three groups of four stutterers each signaled their
'expectancies to stutter immediately prior to verbalizing a
self-formulated sentence which contained a stimulus word
ifrom the Thorndike-Lorge (1944) word list. After a baseline
of the frequencies of the groups' signaled expectancies and
'stuttering had been obtained, the experimental group was
shocked immediately after a signaled expectancy. Subjects
in the yoked-control group received an equal number of
shocks, but the shocks were not contiguous to their signaled
expectancies. In neither group was shock contiguous to
moments of stuttering. Subjects in the time-control group
were not shocked. The three groups then underwent extinc
tion procedures which replicated those of the baseline. The
frequencies of signaled expectancies and of stuttering
'before and after punishment were computed and compared with
in and between groups. The statistical evaluations utilized
the analysis of variance and t_ tests.
Results
Significant F and ;t ratios at or beyond the .05
:level characterized all of the statistical comparisons of
47
decreases in the frequency of signaled expectancies. Thus,
! the mean decrease in frequency of signaled expectancies was j
l ;
igreater in the experimental, or shock-contingent, group than;
;in the two control groups. In addition, the mean frequency ;
!of signalled expectancies within the experimental group was .
significantly lower subsequent to punishment. The mean
’ decrease in the measures of frequency of stuttering also
yielded significant F and t_ ratios among the groups; and
;again it was found that the experimental group's frequency
of stuttering was significantly attenuated when compared to .
the control groups. Furthermore, a statistically signifi
cant difference in the frequency of stuttering within the
experimental group subsequent to punishment was obtained.
Conclusions
The results of this investigation support the
following conclusions:
1. The mean decrease in the frequency of signaled
expectancies to stutter differs significantly
among three groups of stutterers to whom (a)
shock is contingent to signaled expectancies,
(b) shock is not contingent to signaled
expectancies, (c) are not shocked. The mean
decrease in the punishment (shock-contingent)
group is significantly greater than In the
I
! other groups.
. ........ ...“ ......... '...... * 48:
I
2. The mean frequency of signaled expectancies to ;
' stutter prior to punishment is significantly '
greater than after punishment. ;
3. The mean decrease in the frequency of stutter-
ing differs significantly among the same three
groups. The mean decrease in the punishment
(shock-contingent) group is significantly
greater than in the other groups.
4. The mean frequency of stuttering prior to
punishment is significantly greater than after
punishment.
Implications
Within the limitations imposed by the nature of the
itheory tested, the type of experimental design employed, the
number and selection of subjects and the lack of differentia
tion between specific and generalized expectancy to stutter;
;the results of this study provide additional support for
ithe assumption that expectancy to stutter is at least one
causative factor in the precipitation of moments of stutter
ing. Although the present investigation would appear to
offer a more definitive demonstration of a causal relation
ship between expectancy and stuttering than is currently
reported in the literature, this demonstration is far from
conclusive. Thus, the status of expectancy remains some
what problematical. In.order for the reader to be better
49;
able to assess the degree of support provided the expectancy
concept by this research, a brief discussion of its primary :
ilimitations follows.
Any attempt to test, experimentally, a theory that
iUtilizes a hypothesized internal event, or response, as an '
explanation of a subsequent response necessitates that some
; set of operations be performed on the subject from which it
;can be inferred that the hypothesized response was manipulat
ed. A basic methodological problem of this type of experi
mental design is the possibility of confounding by an un
controlled subject variable. This is to say, seldom, if
ever, can it be conclusively demonstrated that the opera
tions utilized to manipulate the response defined as the
independent variable did, in fact, manipulate only that
variable. Therefore, variations in the response defined
as the dependent variable which are associated with the
inferred manipulation of the response defined as the inde
pendent variable could possibly be produced by an unobserved
subject variable. This possibility can not be refuted
unless the experimenter can demonstrate that each of the
;theoretically infinite number of possible responses of the
^subject was observed and that the only response to systemati
cally precede the dependent variable-was the independent
variable. With reference to the present study, for example,
it could be argued that both the decreases in expectancy
and stuttering were produced by the incidental
5o:
conditioning of some variable not amenable to observation; |
jor that expectancy is but one of several independent causal |
Ifactors; or that expectancy is part of a dynamic system of '
‘ factors which must form some minimal combination in order j
1
to precipitate a moment of stuttering.
The latter speculation, to digress for a moment,
would provide a convenient explanation of why the frequency
of signaled expectancies to stutter can be reduced to zero
without the same degree of reduction occurring in the
frequency of stuttering. In addition, it could also account
ifor the findings that stutterers sometimes signal an expec- :
itancy to stutter that is not followed by stuttering and ;
■that stutterers do not always signal an expectancy to stutter
iprior to stuttering (Johnson and Sinn, 1937; Johnson and
iSolomon, 1937; Knott, Johnson and Webster, 1937; Milisen,
1938). In any event, to return to the case in point, the
manipulation of a given response that is followed by con
comitant variations of another response does not demonstrate
a causal relationship as unequivocably as when the variable
manipulated was Independent of the subject.
As will be remembered from a previous discussion
;(Chapter III), ten potential subjects did not signal a
:sufficient number of expectancies to stutter and were eli
minated from the study. In addition, only adult, male
■ stutterers who were seeking therapy were employed In the
51
extremely small sample. The question may be raised, there
fore, to what extent can the results of this study be gener
alized to the population of stutters? There is no published
evidence that would indicate that any of the above factors
would be correlated with differential reactions to the
experimental procedures employed. Furthermore, all subjects
within the experimental group responded in the same direc
tion. Although any attempt to answer the above question
would be speculative, a tenable assumption would appear to
be that the results of this study should apply to approxi
mately 50 per cent of the adult stuttering population, if
it can be assumed that the 20 subjects screened constituted
a representative sample of the stuttering population.
Although the hypothesized distinction between
generalized and specific components of expectancy to stutter
(Van Riper, 1937) is of little relevance to the problem
explored by this investigation, it should be noted that the
design of this study did not permit differentiation between
them. That is to say, when a subject signaled that he ex
pected to stutter, it can not be determined if the referent
of his signal was the general speaking situation; a specific
word within the sentence; several words within the sentence;
or some combination of them. Consequently, it can not be
specified which components of expectancy to stutter were
associated with the presentation of the punishing stimulus.
! Little merit can be seen in further attempts to test!
' !
; j
jthe assumption that expectancy is a causal factor in the i
'occurrence of moments of stuttering at the present time.
i ,
;Since the nature of the assumption itself requires the utili,-
ization of an experimental design in which the effective eli
mination of the possibility of confounding by uncontrolled
subject variables would appear to be extremely unlikely, no
conclusive test is possible. Consequently, some of the more
appropriate research strategies would seem to be:
1. Explorations of variables that may prove to
correlate with signals of expectancy or that
systematically precede a significant number of
moments of stuttering. For example, the
pupillary response associated with the presenta
tion of pleasant and unpleasant stimuli might
provide fruitful results in this area (Hess,
Seltzer, and Shlien, 1965).
2. Factor analysis of those variables found to
correlate with expectancy and/or that systemati
cally precede a significant number of moments
of stuttering.
In general, the suggested emphasis for future research is
the delineation of more variables for which there are reli
able data. This emphasis should provide a better under
standing of the significant variables involved and should
lead to the more effective formulation and testing of
i ' ..........................................’........ " ....... ” ..... ’................................. 53
! !
jhypotheses. i
i In addition to its primary purpose of testing the !
.'assumption that expectancy is a causal factor in the occur- :
Irence of stuttering, this study constituted an initial
lexploration of the efficacy of conditioning signaled
expectancies to stutter as a rehabilitative procedure.
Viewed from this frame of reference, the results of this
investigation are suggestive of considerable rehabilitative
significance in that substantial decreases in stuttering
occurred following a relatively brief period of conditioning
in every subject. Further research, however, is definitely
indicated in order to verify the results obtained in this
study. Replications should include a larger number of
:subjects of both sexes who exhibit lower frequencies of
signaled expectancies than were included in the present
investigation. Assuming that future replications would
essentially confirm the present results, the following
questions are suggestive of some of the appropriate problems
to be studied:
1. How permanent is the decrease in the frequency
of stuttering following the punishment of
expectancy to stutter?
2. To what extent does the decrease in frequency of
stuttering generalize from the situation in
which the expectancy to stutter is punished?
3. What is the effect of increasing the intensity i
i
I
I
of the punishing stimulus on the degree and j
I permanence of decreases of stuttering subsequent;
• to punishment of expectancy to stutter? ;
4. What other stimuli can be utilized as a punish- ;
ing stimulus to decrease expectancies to stutter
and subsequent moments of stuttering?
5. What is the effect of spacing the punishment of ;
i expectancy to stutter across several relatively
brief sessions of punishment as compared to one ;
extended session?
6. Is reward of no expectancy to stutter as effec
tive as punishment of expectancy to stutter in
decreasing subsequent moments of stuttering?
7* Is the combination of reward of no expectancy
and punishment of expectancy to stutter more
effective in decreasing subsequent moments of
stuttering than either procedure alone?
8. What is the effect on the permanence, degree and
generalization of the decrease In stuttering
when conditioning of expectancy to stutter is
combined with the conditioning of stuttering
itself?
It should be noted in passing that several of the above
problems are amenable to single subject research. In fact,
this may be the most judicious research methodology to
j
jemploy until the possibility of undesirable ; side-effects
'of the procedures can be better evaluated.
In summary, then, this investigation provided
jadditional support for the assumption that expectancy to
stutter is at least one causal factor in the precipitation
iof moments of stuttering. It was suggested that explora-
i
tions of variables that correlate with expectancy and/or
that systematically precede stuttering constitute the most
appropriate research strategy at this time. The more
significant implications arising from this study were re
lated to the possible development of rehabilitative pro
cedures in which the conditioning of expectancy would be
utilized to reduce stuttering. Finally, the necessity of
Ifurther research was emphasized, followed by suggestions of
some of the appropriate problems to be studied.
APPENDIXES
56
APPENDIX A
STIMULUS WORDS
57
EXPERIMENTAL WORDS I
PINE
RICH
STORM
BLOW
AFFAIR
DATE
SHOW
HOT
KIND
INDUSTRY
TRIP
BOOK
WORTH
MINUTE
GROUND
HAPPY
DEMAND
SOLDIER
FARM
NATION
SAFE
BAG
MYSELF
MASTER
WOOD
LAND
GRASS
NORTH
WEIGHT
LIE
BALL
WISE
STORY
THIN
STATE
ANOTHER
BEAUTY
RULE
WILD
COURSE
EXPERIMENTAL WORDS II
WATCH
ANIMAL
MUSIC
BLOOD
LOVE
FAMILY
RAIN
WORK
AFRAID
SERVE
INTEREST
BOAT
PROBLEM
ENOUGH
NO
RED
SHAPE
DOCTOR
MONEY
DESIRE
LIFE
HAT
WISH
SALT
FELLOW
LABOR
INCREASE
FAMOUS
HURT
NOTE
BLUE
PAGE
TOMORROW
MILK
BODY
MAKE
GRANT
HIM
ONCE
SHADE
PRACTICE WORDS
jWINTER AIR LOST FLY ;
jYELLOW NOTICE CAR FOREST
' feet SWEET SUPPLY PERIOD
| amount PRINCE ARM ART
| lead STOP BACK AMERICA ;
BIT CLOTHES FREE EAST ■
FEEL PRESIDENT HEAT PLEASE
fact OLD HAIR IRON
brother
!
FIELD KITCHEN FINGER
either SEA FISH LAW
:SENSE WINDOW CHARGE BEST
TRUTH CHIEF RIVER PASS
I CENTURY RETURN COLD START
' left FORM DEGREE HUMAN
jCOUNTRY WALK MONTH NECESSARY
|BAY RECORD SISTER POUND
IKNIGHT MILE GLASS LOT
; RIGHT NEWS FAIR DINNER
FIND CHANCE AFTERNOON BRIDGE
OTHER YOUR STICK HOUR
; PAPER PLAN CARE HARD
' fire SMILE CROSS METHOD
60.
SUIT
GENTLEMEN
i
KEPT
;SECOND
FUTURE
ITSELF
NATIONAL
‘ queen
RACE
DRINK
FATHER
CUP
REAL
INCH
SEASON
SPIRIT
ICE
GATE
YOUNG
CHURCH
CONDITION
MIGHT
NONE
HOPE
BORN
CLOUD
TOUCH
SINGLE
WAY
REPORT
NAME
LADY
TRAINING WORDS
!MR.
BUY DOLLAR HUSBAND
•HISTORY STAR
SUCCESS ROSE
GUARD SHIP SOUL STEP
I baby LIFT POOR
DAUGHTER
BUSINESS SPEAK SUN END
iCIRCLE REACH COMMAND BOARD
CALL
TROUBLE • BILL NEW
iGAIN
DIFFERENT LAY WEATHER
; COVER COAL
PIECE BETTER
THESE DOG
REST HOLE
1 YOURSELF DEAL OFFER THINK
|CHAIR
ENEMY # DRIVE GOD
;DIE HORSE WHO FRIEND
LAKE JUST TEAR STORE
' STRONG SHE COAT WATER
MOTHER SICK SPECIAL SOIL
:advance STONE NATURE MOST
:TRAVEL LOSS EGG YOU
EDGE REPLY
GREAT HEAD
!CROWD BEING
OFFICER PUBLIC
BIG WAIT
MANY TASTE
• CASE MEN
POSSIBLE SHARE
6 2
63
DUTY THOSE EARTH IMPORTANT
ARTICLE MARRY CHANGE STRENGTH
GIRL SHOP MORNING SUGAR
nothing HIGH STAND SKY
length LAUGH FIGURE HILL
LAST NEITHER TRUST DEATH
ACCOUNT FULL SNOW PAST
SORT BROWN PLAY CAPTAIN
COMPANY SOIL MATERIAL KISS
EXPERIENCE FLOWER SIDE ANYTHING
SHOUT PERSON NECK BOX
WEST EFFORT MODERN AGE
NEIGHBOR PROMISE MANNER SOFT
SHOUSE MOVE MARCH SHOULDER
WIND GUESS SCENE EAR
fruit UNCLE USE MEASURE
letter ROCK SIGN COLOR
SERVICE BANK DIFFERENCE MAN
FEAR EYE MEETING SYSTEM
FINISH ACTION OFFICE SONG
FALL LITTLE PAIN POINT
bear STRANGE KING SEVERAL
BATTLE TRAIN HELP TREE
EVERYTHING SHOE EACH GARDEN
SOME ME STREET SON
SOUND LORD VOICE
REASON CORNER CENTER
PEOPLE OPINION TABLE
SCHOOL TALK LIGHT
CONTROL DREAM GUIDE
PRODUCE SPOT OUTSIDE
DISTANCE MEAT SUBJECT
FRONT GROUP VALUE
LEG BRIGHT MOUNTAIN
TOP GOVERNMENT ILL
BED HALL FAVOR
SIR STREAM EXPRESS
CLUB LINE BIRD
WRONG BEAT WHITE
REGARD POWER THOUGHT
'GOLD MEMBER LEAST
NOSE QUESTION GONE
FOOT MIND COURT
'DEAR SMALL DAY
CRY NEED RESULT
FEW VALLEY SPREAD
HOME KILL PRESS
HEART COST MIDDLE
TRUE HONOR JUDGE
WORD ROUND FLOOR
HEIGHT
PEACE
PRICE
TOWN
GREEN
MOMENT
BOY
DRESS
SQUARE
SHOT
MOVEMENT
KNEE
GAME
WONDER
I
DROP
SPRING
OIL
LIP
YEAR
IT
EVENING
POOD
FORCE
STUDY
SILVER
j
OUR COMMON FACE
' fool HEAR CATCH RUSH
TURN ANSWER WOMEN TIME
DARK SUFFER CUT FOREIGN
CAUSE WANT THING ESCAPE
MISS PAY DIRECTION ACT
PAIR FIGHT WAVE WEEK
VILLAGE ROLL SKIN GENERAL
THEM POSITION SAME VIEW
attempt CHARACTER NIGHT PICTURE
BAND PRESENT ORDER PLANT
RIDE SUMMER SMOKE WORLD
SHORE PRACTICE OBJECT NUMBER
tire BEHIND WONDERFUL PURPOSE
well BUILDING OCEAN STOCK
SLEEP YARD MARKET RING
COOK POST MATTER PARTY
CITY ROOM WAR DOUBT
WIFE WHOLE MARK ISLAND
PLEASURE CHAIN SHORT PLACE
MOUTH WING WALL ISSUE
TIE DARE WARM VISIT
army GOOD SURPRISE ROAD
: STATION CAN SOUTH QUARTER
SPACE HALF HEAVEN CHILD
66
PAINT
HER
BRANCH
DOOR
JOY
ALL
TRADE
DANCE
CLASS
IDEA
COLLEGE
JOB
I
i
I
I
I
j
APPENDIX B
EXPERIMENTAL GROUP QUESTIONNAIRE
AND RESPONSES
67
EXPERIMENTAL GROUP QUESTIONNAIRE
What do you think was the nature of this experiment?
Did the shock affect your speech in any way?
Did the shock affect your pressing of the buttons in
any way?
Did you ever avoid pressing the button which signaled
that you expected to stutter even though you did expect
to stutter? (If yes) How often do you think you might
have done this during the last''40 sentences?
Would you be willing to be shocked as you were today
if it would help your stuttering?
68
RESPONSES OP EXPERIMENTAL GROUP
TO QUESTIONNAIRE
Subject, E 1
I’m very confused. I don't know, maybe that’s the
object.
I don’t know.
No. I’m so confused. I’m not even sure I understand
the question.
No.
Yes .
Subject, E 2
Well, I felt that if I knew if I was going to have a
shock I wouldn’t stutter. To see if having a shock
hanging over your head so to speak would help you speak
more fluently and normally.
It made me concentrate more on speaking fluently.
Sure it did.
Yes. Two or three, maybe.
No, because that isn’t the answer. I hope it isn't
the answer. It's a very uncomfortable answer if it is.
69
Subject, E 3
Well, I thought the point of it is that if you stuttered
you would get a shock. I think the purpose is training
somebody that if they stutter there is fear of pain so
they don't stutter.
Not exactly. It did at first, but then I got used to it
Yes, I didn’t press the button that I got shocked on as
much as the other one.
No. Changed sentence to one in which I didn't think I
would stutter.
Well, it got a little painful at the end, but if it
would help I would do almost anything.
Subject, E 4
Beats me. I haven't given any thought to it.
Don't think it affected my speech.
Yes .
Yes. Half a dozen.
, APPENDIX C
JUDGES’ INSTRUCTIONS
71
INSTRUCTIONS TO THE JUDGES
; Today you will hear a tape recording of twelve
^different people. All of them stutter. You will hear them
say a series of self-formulated sentences. While you listen
|to each sentence, you are to make a mark on the paper you
have been supplied every time you hear a moment of stutter
ing. At the conclusion of each sentence, I will ask each
;of you how many moments of stuttering you heard. At least
itwo of you must agree upon the number of moments of stutter-
:ing which occurred. The sentence will be replayed until
:that agreement has been obtained.
In performing these definitions, each of you is to
use your personal criteria as a basis for determining
moments of stuttering with the following exception: Count as
one moment of stuttering all attempts made by a subject to
:say a given syllable in the sentence. In other words, no
matter how many attempts a subject makes to say a specific
syllable In the sentence, all of those attempts are to be
considered as a single moment of stuttering.
Are there any questions before we begin?
72
APPENDIX D
TESTS OF HOMOGENEITY
OF
VARIANCE
Summary of Bartlett’s Test of Homogeneity of
Variance of the Difference Scores of the
Frequency of Signaled Expectancies to
Stutter
Treatment d.f. ES,
k Sk
log Sk
E 3
641.00 213.67
2.3298
S 3
133.00 44. 33 1.6467
T 3
138.00 46.00 1.6628
304.00 5.6393
Computations:
1.
ES
k _ 304 _
ES
k
k
= = 101.33; log = 2.0005
ES
2. k log = (3) (2.0055) = 6.0165
2
3.
ES
Diff. = k log — j j r 2 log Sk2 = 6.0165
- 5.6393 = .3772
4. x2 = (2.3026) (n-1) (diff.)
= (2.3026) (3) (0.3772) = 2.6055*
# 5-991 significant at .05 level with
3 d.f.
74
Summary of Bartlett's Test of Homogeneity of
Variance of the Difference Scores of the
j Frequency of Moments of
i Stuttering
Treatment d.f. z^k^ ^k^
: E 3 1248.75 416.25 2.6194
S 3 243.00 81.00 1.9085
: T 3 228.75 76.25 1.8823
; 573.50 6.4102
Computations:
2 2
ES, j.. ES
1. = -1 ^ ^ = 191.17; log = 2.2814
ES 2
2. k log — y ~ = ( 3) (2.2814) = 6. 8442
ES 2
3. Diff. = k log — y slog Sk2 = 6 .8442
- 6.4102 = .4340
4. x2 = (2.3026) (n-l) (Diff.)
= (2.3026) (3) (.4340) = 2.998*
* 5.991 significant at .05 level with
3 d.f.
Summary of Bartlett's Test of Homogeneity of
The Difference Scores of the Frequency
of Sentences Stuttered
[Treatment
i
I
! E
J
! s
T
d.f.
3
3
3
ESk
39.00
96.75
17.00
k
13.00
32.25
5.67
50.92
Computations:
1
es
k
zs,
k
= 5°;92 = 16.97; log — = 1.2279
ES
2. k log
k
k
= (3) (1.2297) = 3.6891
3.
ES, p
Diff. = k log — y £log sk = 3.6891
- 3.3761 = .3130
:4. X2 = (2.3026) (n-l) (Diff.)
= (2.3026) (3) (.3130) = 2.162*
log Sk2
1.1139
1.5086
.7536
3.3761
* 5.991 significant at .05 level with
3 d.f.
LIST OF R E F E R E N C E S
77,
LIST OP REFERENCES
jAyllon, T. and Azrin, N. H. , Punishment as a discriminative .
stimulus and conditioned reinforcer with humans.
! J. Ex p. Anal. Behav., 9, 411-419 (1966).
i
Azrin, N. H., Effects of two intermittent schedules of
immediate and nonimmediate punishment. J. Psychol.,
| 42, 3-21 (1956).
I
' , Some effects of noise on human behavior. J. Exp.
j- Anal. Behav.. 1, 183-200 (1958).
t
; ________, Effects of punishment intensity during variable-
interval reinforcement. J. Exp. Anal. Behav., _3,
123-142 (I960).
_________, and Holz, W. C., Punishment. In W. K. Honig (Ed.).
Operant Behavior: Areas of Research and Application.
New York: Appleton-Century-Crofts (1966).
, and Hake, D., Pixed-ratio punishment. J. Exp.
Anal. Behav.. 6, 141-148 (1963).
iBloodstein, 0., Stuttering as an anticipatory struggle
reaction. In J. Eisenson (Ed.), Stuttering: A
Symposium. New York: Harper and Brothers (1958).
Church, R. M., The varied effects of punishment on behavior.
Psychol. Rev., 70, 369-402 (1963).
Cohen, E., A comparison of oral reading and spontaneous
speech with special reference to the adaptation and
consistency effect. Speech Monogr., 20, 144 (1955).
Dinsmoor, J. A., A discrimination based on punishment.
Quart. J. Exp. Psychology, 4_, 27-45 (1952).
_________, Punishment: I. The avoidance hypothesis.
Psychol. Rev., 6l, 34-46 (1954).
_________, and Campbell, S. L., Escape-from-shock training
following exposure to inescapable shock. Psychol.
Rep., 2, 43-49 (1956).
I z >
Dollard, J. and Miller, N. E., Personality and Psycho-
j therapy: An Analysis in Terms of Learning, Thinking
| and Culture. New York: McGraw-Hill (1950).
! i
Edwards, A. L., Experimental Design in Psychological ;
Research. New York: Hold, Rinehart and Winston i
j (1963). j
jEisenson, J., A perseverative theory of stuttering. In J.
i Eisenson (Ed.), Stuttering: A Symposium. New York:
Harper and Brothers (1958).
Estes, W. K., An experimental study of punishment. Psychol.
Monogr., 57, whole No. 263 (1944).
I !
Ferster, C. B. and Skinner, B. P., Schedules of Reinforce-
; ment. New York: Appleton-Century-Crofts (1957).
'Flanagan, B., Goldiamond, I., and Azrin, N. H., Operant
stuttering: the control of stuttering behavior
through response contingent consequences. J. Exp.
Anal. Behav.. 1, 173-177 (1958).
^Franks, C. M., Conditioning and abnormal behavior. In H. J.
Eysenck (Ed.), Handbook of Abnormal Psychology.
New York: Basic Books (1961).
Goldiamond, I., Stuttering and fluency as manipulable
; operant response classes. In L. Krasner and L. P.
Ullman (Eds.), Research in Behavior Modification.
New York: Hold, Rinehart and Winston (1965).
Guilford, J. P., Fundamental Statistics in Psychology and
Education- ! 3rd ed. New York: McGraw-Hill (1956) .
Hays, W. L., Statistics for Psychologists. New York: Holt,
Rinehart and Winston (1963).
Herman, R. L. and Azrin, N. H., Punishment by noise in an
alternative response situation. J. Exp. Anal.
Behav., 7, 185-188 (1964).
Hess, E., Seltzer, A., and Shlien, J., Pupil response of
hetero-and homosexual males to pictures of men and
women: a pilot study. J. Abnorm. Psychol., JO,
165-168 (1965).
Holz, W. C. and Azrin, N. H., Discriminative properties of
punishment. J. Exp. Anal. Behav., 4, 225-232 (1961).
; 8o
j
(Johnson, W. and Ainsworth, S., Studies in the psychology of
| stuttering: X. Constancy of loci of expectancy of
| stuttering. J. Speech Pis. , 3, 101-10*1 (1938).
Johnson, W. and Knott, J. R., The moment of stuttering.
J. Genet. Psychol., 4_8, 475-479 (1963).
Johnson, W. and Sinn, A., Studies in the psychology of
• stuttering: V. Frequency of stuttering with
expectation of stuttering controlled. J. Speech
Pis., 2, 98-100 (1937).
Johnson, W. and Solomon, A., Studies in the psychology of
1 stuttering: IV. A quantitative study of expecta
tion of stuttering as a process involving a low
degree of consciousness. J. Speech Pis., 2, 95-97
! (1937).
Kamin, L. J., The delay-of-punishment gradient. J. Comp.
Physiol. Psychol., 52, 434-437 (1959).
Karsh, E., Effects of number of rewarded trials and
' intensity of punishment on running speed. J. Comp.
Physiol. Psychol. 55, 44-51 (1962).
Knott, J. R., Johnson, W., and Webster, M. J., Studies in
the psychology of stuttering: II. A quantitative
evaluation of expectation of stuttering In relation
to the occurrence of stuttering. J. Speech Pis.,
2, 20-22 (1937).
Kushner, M. and Sandler, J., Aversion therapy and the
concept of punishment. Behav. Res, and Therapy,
iL, 179-186 (1966).
Lichtenstein, P. C., Studies of anxiety: the production of
a feeding inhibition in dogs. J. Comp. Physiol.
Psychol., 43, 16-29 (1950).
Martin, R. and Siegel, G. M., The effects of response
contingent shock on stuttering. J. Speech Hearing
Res., _2, 340-352 (1966a).
■ _______ , The effects of simultaneously punishing stuttering
and rewarding fluency. J. Speech Hearing Res., 9,
466-475 (1966b).
Masserman, J. H., Principles of Pynamic Psychiatry.
Philadelphia: Saunders (1946).
Melvin, K. B., Escape learning and vicious circle behavior
as a function of percentage of reinforcement. J.
Comp. Physiol. Psychol., 58, 248-251 (1964).
Milisen, R., Frequency of stuttering with anticipation of
stuttering controlled. J. Speech Dis., 3, 207-214
j (1938).
i
i
jMowrer, 0. H. and Ullman, A. D., Time as a determinant in
I integrative learning. Psychol Rev. , 5.2, 61-90
! (1945).
i
jMysak, E. D., Servo theory and stuttering. J. Speech
i Hearing Pis., 25, 188-195 (I960).
Peins, M., Adaptation effect and spontaneous recovery in
stuttering expectancy. J. Speech Hearing Res., 4_,
91-99 (1961a).
Consistency effect in stuttering expectancy.
J. Speech Hearing Res., 4_, 397-398 (1961b).
Porter, H. V. , Studies in the Psychology of Stuttering:
XIV. Stuttering phenomena in relation to size and
personnel of audience. J. Speech Dis., _4, 323-333
(1939).
Sandler, J., Some aspects of self-aversive stimulation in
the hooded rat. J. Exp. Anal. Behav., 7, 409-414
(1964).
Sheehan, J., Conflict theory of stuttering. In J. Eisenson
(Ed.), Stuttering: A Symposium. New York: Harper
and Brothers (1958).
Skinner, B. F., Science and Human Behavior. New York:
Macmillan (1953).
Solomon, R. L. and Brush, E. S., Experimentally derived
concepts of anxiety and aversion. In M. R. Jones
(Ed.), Nebraska Symposium on Motivation. Lincoln,
Nebraska: University of Nebraska Press (1956).
iThorndike, E. L., Educational Psychology, Vol. II. New
York: Teachers College, Columbia University (1913).
_________, and Lorge, I., The Teacher's Word Book of
30,000 Words. New York: Teachers College,
Columbia University (1944).
Van Riper, C., A study of the thoracic breathing of
stutterers during expectancy and occurrence of
stuttering spasm. J. Speech Dis., 1, 61-72
(1936).
, The preparatory set in stuttering. J. Speech
Dis. , 2, 149-154 (1937).
_______, and Milisen, R., A study of the predicted duration
of the stutterer’s blocks as related to their
actual duration. J. Speech Dis. , 4, 339-345 (1939).
■Wingate, M. E., Stuttering adaptation and learning: II.
The adequacy of learning principles in the inter-
i pretation of stuttering. J. Speech Hearing Dis.,
31, 211-218 (1966) .
1
Wischner, G. J., An experimental approach to expectancy
and anxiety in stuttering behavior. J. Speech
Hearing Dis., 17, 139-154 (1952).

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

An Experimental Study Of The Effect On Babbling Of Visual Interaction Between Mother And Infant

PDF

Effect Of Types Of Rehearsal On Frequency Of Stuttering

PDF

Frequency Of Stuttering As A Function Of Connotative Word Meaning

PDF

An Investigation Of Attitude Differences In Parents Of Stutterers And Parents Of Non-Stutterers

PDF

An Experimental Investigation Of The Relationship Between Frequency Of Stuttering And Open Expression Of Aggression

PDF

An Investigation Of Physiological Measures In Relation To The Moment Of Stuttering In A Group Of Adult Males

PDF

An Investigation Of Pupillary Response Preceding Expectancy And Stuttering

PDF

The Effects Of Stuttering On Systolic Blood Pressure

PDF

Word Familiarity And Frequency Of Stuttering

PDF

An Empirical Study Of Sex-Role Identification And Sex-Role Preference In A Selected Group Of Stuttering Male Children

PDF

A Factor Analysis Of The Responses Of Brain-Damaged Adults To The Sklar Aphasia Evaluation Summary

PDF

Autonomic correlates of stuttering, fluency, and threat-of-shock

PDF

Changes In The Intensity Of An Overt Response Before, During, And After Instances Of Stuttering

PDF

Listener Identification Of 'Stuttering' And 'Stutterer' As A Function Of Variation In Speech Patterns

PDF

Stuttering and time perspective

PDF

An Experimental Study Of The Effects Of Fluency Feedback And Stuttering Feedback On The Subsequent Frequency Of Stuttering Utterance Duration, And Latency Of Response

PDF

An Experimental Study Of Auditory Thresholds Of Adults For Warble Tone, Pure Tone, And Recorded Speech

PDF

The Effects Of Separating Production From Perceptual Judgment Of Articulation In Children With Articulatory Defects

PDF

Chronological Age And Grammatical Development As Determinants Of The Proportions Of Disfluencies On Lexical And Function Words In Preschool Children

PDF

A Normative Study Of Oral Sensation And Perception: Two-Point Discrimination, Form Identification, Tactile Pattern Recognition, And Mandibular Kinesthesia

Asset Metadata

Creator Curlee, Richard Frederick (author)

Core Title An Experimental Study Of The Effect Of Punishment Of The Expectancy To Stutter On The Frequency Of Subsequent Expectancies And Stuttering

Degree Doctor of Philosophy

Degree Program Speech

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

Tag health sciences, speech pathology,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Perkins, William H. (committee chair), Haney, Russell (committee member), Harvey, Herman M. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-144620

Unique identifier UC11360863

Identifier 6713740.pdf (filename),usctheses-c18-144620 (legacy record id)

Legacy Identifier 6713740.pdf

Dmrecord 144620

Document Type Dissertation

Rights Curlee, Richard Frederick

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