THE INFLUENCE OF INCREASED
DURATIONS OF SPEECH STIMULI ON THE
RECOGNITION AND RETENTION OF

WORDS? BY APHASIC ADULTS

Thesis for “we Deg-m of pit. D.
MICHIGAN STATE UNIVERSITY
William R. Dopheide

1968

w s! wmg,“;z {may-r... fr

LIBRARY ”
Michigan Scam
Ilaivcnfigy

v—‘w

  
 

4,)

'01 '

This is to certify that the ~

. .‘.
v “

s '5, ,
thesis entitled ’

THE INFLUENCE OF INCREASEb
DURATIONS OF SPEECH STIMULI on THE
RECOGNITION’AND RETENTION OF

wonns BY APHASIC'ADUL$S*“ “

presented By
WILLIAM R. DOPh’EIDE

has been accepted towards fulfillment
of the requirements for

3—0.9;— degree in

AUDIOLmY AND SPEECH SCIENCES

2a: my my

M... m» (Jo‘s-sear”
5" /é " 4 5’

Date

0-169

 

 

 

ABSTRACT

THE INFLUENCE OF INCREASED DURATIONS OF
SPEECH STIMULI ON THE RECOGNITION AND
RETENTION OF WORDS BY APHASIC ADULTS

by William R. Dopheide

Recognizing the importance of auditory language
reception in the total rehabilitative effort directed toward
the language problems of the aphasic adult, Speech patholo-
gists have sought to develop modifications of the auditory
language stimulus that will make the stimulus more adequate
for the aphasic adult. One recommended modification that
appears to have rather widespread clinical application
is increasing the duration of the words spoken to the aphasic
adult.

The purpose of this investigation was to assess the
effectiveness of temporal expansions in the phonemic seg-
ments of words on the aphasic adult's performance in recog-
nizing and retaining words. In addition, two independent
(control) variables were employed: meaningfulness of the
stimulus word and subjects' adequacy in auditory language
function.

The dependent variables were performance on a word
recognition task and performance on a retention Span task,
with stimuli being presented auditorily on both tasks.

These tasks were selected since they seemed highly related

William R. Dopheide

to significant characteristics of the aphasic language
deficit, according to reports in the literature.

For the word recognition task, the stimulus words,
derived from the Peabody Picture Vocabulary Test, were
arranged in a manner that provided four equivalent lists
of 75 words each. Each list was divided arbitrarily into
three sets of 25 words each that were regarded as reflect-
ing three levels of meaningfulness. The stimuli for the
retention span task varied in complexity from two word
series to five word series, with all complexity levels
equally represented at all stimulus durations. The varia—
tions from the normal durations of the stimulus words were
produced by an electromechanical process to the niminal
levels of 125%, 150%, and 175% expansions of the original
recorded duration (100%); hence, four stimulus duration
conditions were available. Subjects were assigned to a
High auditory language disturbance group or a Low auditory
language disturbance group on the basis of their scores
on the Auditory Disturbance section of the Minnesota Test
for Differential Diagnosis of Aphasia.

The data were treated by analysis of variance:
three factor (repeated measures) for Meaningfulness, Groups,
and Durations on the word recognition task; and two factor
(repeated measures) for Groups and Durations on the reten-
tion Span task. An inter-task comparison was made employing
a Spearman rank correlation at each stimulus duration con-

dition.

William R. DOpheide

No significant main effect for Duration was observed
on either task; however, Groups diSplayed some interaction
with Durations on the word recognition task (but not on the
retention task). The interaction appeared to be generated
by the High auditory language disturbance group. The Mean-
ingfulness interaction with Durations, which was of interest,
was not statistically significant. Of importance to the
design of the experiment, the main effects for Groups and
Meaningfulness levels were statistically significant and
manifested no interaction. The inter-task comparison re-
vealed statistically significant correlations in the High
auditory language disturbance group at three of the four
duration conditions. This finding might be interpreted as
meaning less independence of available vocabulary and
retention Span in aphasic adults with poor auditory lan-
guage function.

On the basis of this investigation, increased
stimulus durations, as produced for use in this study, do
not have a general positive influence on word recognition
and retention span. Increasing the duration of the stimulus
may have a positive effect on word recognition for subjects
with poor auditory language function, but further research

is necessary.

THE INFLUENCE OF INCREASED DURATIONS OF SPEECH STIMULI
ON THE RECOGNITION AND RETENTION OF WORDS

BY APHASIC ADULTS

BY

William R. DOpheide

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Audiology and Speech Sciences

1968

Gfi/‘Vé/

ACKNOWLEDGMENTS

The writer is indebted to a number of people whose
guidance, assistance, and cooperation made the completion
of this study possible. Their interest in this study is
gratefully acknowledged.

To Drs. Herbert J. Oyer, Leo V. Deal, and Bradley
Lashbrook, staff members of the Department of Audiology
and Speech Sciences at Michigan State University, for their
guidance.

To Miss Josephine Simonson (University of Michigan),
Mr. William Locke (American Legion Hospital, Battle Creek),
Mr. Pat Sacco (Constance Brown Hearing and Speech Center),
Mr. Charles Hillary (VA HOSpital, Battle Creek), and
Dr. Clyde Willis (Western Michigan University) for their
assistance in obtaining subjects.

To the aphasic adults for their willing participa-
tion in the tasks of this investigation.

To my family, Marjorie, Kathleen, Marcia, and Mark
for bearing an indescribable responsibility for the com-

pletion of this study.

ii

TABLE OF CONTENTS

ACKNOWLEDGMENTS. . . . . . . . . . . . . . . . .
LIST OF TABLES . . . . . . . . . . . . . . . . .

LIST OF FI
LIST OF AP

Chapter
I. INT

GURES. . . . . . . . . . . . . . . . .
PENDICES . . . . . . . . . . . . . . .

RODUCTION. . . . . . . . . . . . . . .

Purpose of the Study . . . . . . . . .
Importance of the Study. . . . . . . .
Definitions. . . . . . . . . . . . . .
Organization of the Report . . . . . .

II. REVIEW OF LITERATURE. . . . . . . . . . .

III. EXP

IV. RES

Development of Concepts of Aphasia . .
Neurophysiological Rationale . . . . .
Clinical Basis of Investigation. . . .
Experimental Tasks and Associated
Variables. . . . . . . . . . . . . .
Summary. . . . . . . . . . . . . . . .

ERIMENTAL PROCEDURES . . . . . . . . .

Selection and Preparation of Stimulus

Material . . . . . . . . . . . . . .
Equipment and Materials. . . . . . . .
Recording of Auditory Stimuli. . . . .
Selection and Grouping of Subjects . .
Presentation of Stimuli to Subjects. .

ULTS AND DISCUSSION. . . . . . . . . .

Results from Word Recognition Task . .
Results from Retention Span Task . . .
Results from Inter-task Comparison . .
Discussion . . . . . . . . . . . . . .

V. SUMMARY AND CONCLUSIONS . . . . . . . . .

BIBLIOGRAP
APPENDICES

Summary. . . . . . . . . . . . . . . .
Conclusions. . . . . . . . . . . . . .
SuggeStions for Future Research. . . .

HY O O O O O O O O O O O O O O O O O 0

iii

Page
ii

iv

vi

Nahum t—'

14
24
33

37
48

51

51
58
59
67
76

84

89
102
105
107

113

113
116
118
121

127

LIST OF TABLES

1%
1. Separation of PPVT Forms into Four Lists and
Three Hearing Vocabulary Levels . . . . . . . . .
2. Comparison of Auditory Language Disturbance
(ALD) Error Scores of Two Subject Groups. . . . .
3. Comparison of Age, Educational Level, and Elapsed
Time since Onset of Aphasia of Two Subject Groups
4. Sample Arrangement of Lists with Block Index
Numbers . . . . . . . . . . . . . . . . . . . . .
5. Summary of Analysis of Variance Comparing the
Effects of Hearing Vocabulary Levels, Subject
Groups, and Stimulus Duration on the Word
Recognition Task. . . . . . . . . . . . . . . . .
6. Hearing Vocabulary Level Means, and Differences
between Means for Each Auditory Language Dis-
turbance Group. . . . . . . . . . . . . . . . . .
7. Mean Performance for Each Auditory Language
Disturbance Group at Four Stimulus Durations,
and Within Group Mean Differences between
Duration Conditions . . . . . . . . . . . . . . .
8. Summary of Analysis of Variance Comparing the
Effects of Subject Groups and Stimulus Durations
on the Retention Span Task. . . . . . . . . . . .
9. Spearman Rank Correlations of Performance on

the Word Recognition Task and Retention Span
Task for the Two Subject Groups . . . . . . . . .

iv

Page

52

72

74

80

91

93

100

104

106

Figure
1. Graph of Group Means and Total Group Mean
at Three HVL . . . . . . . .
2. Graph of Means for Hearing Vocabulary Levels
of Groups at Four Durations.
3. Schematic of Speech Expansion Device .
4. Expansion Process:

LIST OF FIGURES

vs. Percentage of Expansion.

Tape Transport Velocity

Page

97

99

128

132

LIST OF APPENDICES

Appendix Page
A. Further Description of Process by which
Stimulus Durations Were Increased . . . . . . 128

B. Span of Retention Response Plates and
Associated Stimuli. . . . . . . . . . . . . . 133

C. Data Used to Establish Acceptance Intervals
for Durations of Words in Original Recording. 139

D. Calibration Check of Expanded Tapes . . . . . 144

E. Scores of Subjects on Recognition and
Retention Tasks . . . . . . . . . . . . . . . 146

vi

CHAPTER I

INTRODUCTION

Man can experience no more abrupt loSS of his
uniqueness than to lose his ability to communicate. For
as Gibran has written

When you meet your friend, . . . let the Spirit

in you move your lips and direct your tongue.

Let the voice within your voice speak to the ear

of his ear;

For his soul will keep the truth of your heart

as the taste of the wine is remembered
When the colour is forgotten and the vessel is
no more.1
To experience disability of the "voice within your voice"
is to suffer deprivation in perhaps the most essential
and sustaining of human behaviors. As Goldstein has stated
The different views regarding the nature of man
reflect themselves in the discussions of man's out-
standing capacity: language.2

The abrupt disruption of this capacity is aphasia.

It is a language disorder associated with cerebral insult

and as such is very frequently accompanied by other phy-

sical and behavioral disabilities.

 

1Kahil Gibran, The Prophet (1923) (New York: Alfred
A. Knopf, Inc., 1964), p. 61.

2Kurt Goldstein, Language and Language Disturbances
(New York: Grune and Stratton, 1948), p. 345.

 

 

Aphasia has been recorded in the literature over
the centuries.1 However, as man's nature has lead him to
increase his life expectancy, to improve his means of deal-
ing with disease and disability, and to invent more traumatic
ways of doing harm, the problems of the aphasic adult have
drawn increased attention. Within the last 100 years, over
30,000 books and articles have dealt with the problems of
the aphasic adult.2 In this time, Esti Freud has seen the
investigations as taking three distinct approaches: direct
anatomical-physiological investigations; psychological
observations of the total behavior of the patient from
which central nervous system dysfunction might be inferred;
and structural, linguistic analyses of the language behavior
itself.3

The responsibilities of Speech pathology in this
interdisciplinary approach have been centered, of course,
on the language behavior itself. The increased demands

for language rehabilitation services during and after World

War II4 seem to have intensified the search for better

 

1Arthur L. Benton and Robert J. Joynt, "Early
Descriptions of Aphasia," AMA Archives of Neurology, III
(1960), pp. 205-22.

2Esti D. Freud, "Recent Trends in Aphasia," American
Journal of Psychiatry, CX (1953), p. 186.

31bid.. pp. 186-93.

4Joseph M. Wepman, Recovery from Aphasia (New York:
Ronald Press Company, 1951), p. 4.

diagnostic and treatment procedures with which to assist
the aphasic adult. It was from this clinical viewpoint

that the present experimental investigation was undertaken.

Purpose of the Study

 

The purpose of this study was to examine the effec-
tiveness of a treatment procedure used by clinicians in
language therapy with aphasic adults. The study was directed
at investigating, experimentally, the influence of increased
durations of selected Speech stimuli on the receptive lan-
guage performance of aphasic adults. “Six: questions con-
cerning aphasic adults were posed for investigation.

1. What influence does increasing the duration

of Speech stimuli have on the performance of
an auditory word recognition task?

2. Is there a relationship between the meaning-
fulness of the Speech stimuli and the influence
of increases in stimulus duration on performance
of an auditory word recognition task?

3. Does the influence of increased durations of
the speech stimuli vary as a function of the
severity of the subject's disturbance in
auditory language reception on tasks requiring
the auditory recognition of words?

4. What influence does increasing the duration

of the speech stimuli have on the performance

of a retention span task in which words are
presented auditorily and in series?

5. Does the influence of increased durations of
the Speech stimuli vary in relation to the
severity of the subject's disturbance in audi-
tory language reception on a task requiring
the retention of stimuli presented auditorily?

6. What relationship exists between performance
on an auditory word recognition task and an
auditory retention span task at several levels

of increased stimulus duration?

Importance of the Study

It has been said ". . . in the investigation of
problems as complex as aphasia almost any question one
could raise would be worth study."1 This subcommittee of
the American Speech and Hearing Association, studying the
research needs in aphasia, was indicating the strong need
for investigating any and all aspects of the clinical
management of the language problems of the aphasic adult.
This subcommittee did indicate the Significance of questions

related to the differential effectiveness of therapeutic

 

lWalter Amster, g£_§l,, "V. Report of Subcommittee
on Problems of Aphasia," in Research Needs in Speech Pathol-
ogy and Audiology (Washington: American Speech and Hearing
Association, 1959), p. 42.

processes with different aphasic categories.l

These different categories are not the classical
categories of the last century which are being perpetuated
today by psychological and neurological textbooks.2 Major
investigations within the last thirty years have denied
strongly the classical dichotomy between sensory and motor
language deficits. With increasing emphasis on the objective
assessment of the language behavior itself, concern has
focused on the general nature of the aphasic language deficit
and the definition of its dimensions which have diagnostic
and treatment significance.3

Two facts have made the definition of aphasia dif-
ficult. The disabilities of the aphasic adult, stemming
as they do from some cerebral insult, require the profes-
sional attention of several disciplines.4 Secondly, attempt-
ing to understand disordered language function, masked by
other disabling physical and psychological problems, is
extremely difficult, eSpecially when one considers that

normal language processes are only now becoming clearer

 

lIbid.

2James J. Jenkins and Hildred Schuell, "Changing
Concepts of Aphasia," Perceptual and Motor Skills, XIII
(1961). p. 270.

3Charles E. Osgood and Murray S. Miron, Approaches
to the Study of Aphasia (Urbana: University of Illinois
Press, 1963), p. 126.

 

41bid., p. 3.

through scientific investigation.1 The definition of aphasia
adopted for use in this study was one derived from extended
empirical research by Schuell and Jenkins.2 This definition,
which recognizes a general dimension of language, has re-
ceived a good deal of clinical interest since it operation-I
alizes the concept of aphasia in terms which are quite
amenable to clinical management in language therapy. The

two major characteristics of aphasia given in this defini-
tion suggested clear-cut tasks upon which the experimental
variable could be tested.

Recognizing the coexistence of some degree of both
receptive and expressive disability in all patients, there
appears to be rather substantial support for the Opinion
that auditory language problems are generally an aspect of
the total language dysfunction of the aphasic adult. As
Schuell has stated

Probably all aphasic patients Show some impairment

of auditory processes, because language, learned by
ear, remains dependent upon discrimination, recognition,

and recall of learned auditory patterns, and upon audi-
tory feedback processes.

 

lIbid.

2Hildred Schuell and James J. Jenkins, "The Nature
of the Language Deficit in Aphasia," Psychological Review,
LXVI (1959), pp. 50-51. James J. Jenkins and Hildred Schuell,
"Further Work on Language Deficit in Aphasia,“ Psychological
Review, LXXI (1964), pp. 87—93.

3Hildred Schuell, Differential Diagnosis of Aphasia
with the Minnesota Test (Minneapolis: university of
Minnesota Press, 1965), p. 5.

It appeared that any investigation of techniques for facili—
tating auditory language reception would have relevance to
the broad range of problems in aphasia.

Explaining improved language function in aphasia
on the basis of treatment procedures is difficult Since
there is insufficient knowledge of cerebral function in
such complex functions as language.1 However, there have
been several theories of cerebral function advanced which
give reason to believe certain modifications of an auditory
language stimulus may elicit improved recognition or reten—
tion of the linguistic significance of the stimulus,2 while
other current theories would suggest no improvement from
such modifications of the stimulus itself.3

The particular modification of the auditory lan-
guage stimuli to be studied is increased duration. Clinical
practice has indicated that giving greater temporal value

to the phonemes of a word is a means of improving its

 

lOsgood and Miron, Approaches to Aphasia, pp. 33-37.

 

2Aleksandor Romanovich Luria, Higher Cortical Func-
tions in Man, trans. by Basil Haigh (New YOrk: Basic Books,
Inc., 1966), pp. 97-103. Wilder Penfield and Lamar Roberts,
Speech and Brain Mechanisms (Princeton, N.J.: Princeton
University Press, 1959), pp. 21-22. Karl Pribram, "Neo-
cortical Function in Behavior," in Biological and Biochemical
Bases of Behavior, eds. Harry F. Harlow and Clinton N.
Woolsey (Madison: university of Wisconsin Press, 1958),
pp. 167-168.

 

 

 

3Norman Geschwind, "Disconnexion Syndromes in Ani-
mals and Man, Parts I and II," Brain, LXXXVIII (1965), pp.
237-293 and 585-644.

recognition by aphasic adults.l Other variables which have
appeared to influence the adequacy of the stimulus for the
patient were recognized and were considered in evaluating
the influence of the experimental variable. The availabil-
ity of a device capable of producing these expansions in
word duration2 offered promise of providing the speech
pathologist with another clinical tool. It seemed worth-
while to make a test of its potential effectiveness in the

language rehabilitation of the aphasic adult.

Definitions

Aphasia.--This nonfunctional language disturbance
is characterized by a reduction or loss of a centralized
ability to cope with the arbitrary meanings, associations,
and relationships which exist in an individual's system
of communication. This basically unidimensional behavior
may be further impaired by independent but related trans-
missive dysfunctions, variously referred to as agnosias,

apraxias, or sensorimotor deficits.

 

lHildred Schuell, James J. Jenkins and Edward
Jimenez-Pabon, Aphasia in Adults: Diagnosis, Prognosis,
and Treatment (New York: Harper and Row, Publishers, 1964),
pp. 340-41.

2Hugh 8. Allen, Jr., "The Eltro Information Rate
Changer - Mark II: Simple Quality Speech Compression"
(paper presented at Conference on Compressed Speech, Uhi-
versity of Louisville, October 19, 1966).

 

While this definition was drawn from several the-
oretical positions regarding the nature of the disorder,
it is admittedly biased in the direction of the Schuell
and Jenkins definition which characterizes aphasia as a
reduction of available vocabulary, impaired verbal retention
span, and impaired perception and production of messages.
The impaired perception and production of messages, they
have stated, may be secondary to the impairment of avail-
able vocabulary and verbal retention Span.

The definition recognized also, the contributions
of Wepman, §£_§l,, who have viewed aphasia as a disruption
in the central nervous system's capacity to integrate lan—
guage stimuli once such stimuli are free of their input
modalities, with the disruption having the potential of
falling at any point in the process; in the arousal of mean-
ing, in the semantic word selection, or in the syntactic
process.

Subjects in this study were identified as "aphasic"
on the basis of competent clinical diagnosis employing

one or both of the following recognized diagnostic tools:

the Language Modalities Test for Aphasia3 and The Minnesota

 

lSchuell, Jenkins and Jimenez-Pabon, Aphasia in
Adults, p. 113.

2Joseph M. Wepman, §£_§l,, "Studies in Aphasia:
Background and Theoretical Formulations," Journal of Speech
and Hearing Disorders (JSHD), XXV (1960), p. 328.

3Joseph M. Wepman and Lyle V. Jones, The Langugge
Modalities Test for Aphasia (Chicago: Education-Industry
Service, 1961).

-10-

Test for Differential Diagnosis of Aphasia.l Hence, "aphasia"
was Operationalized on the basis of performance on one or

both of these tests.

Auditory language disturbance.—-Deficits in ability

 

to deal with receptive language, auditorily, was defined
on the basis of a subject's score on seven subtests of the
Auditory Disturbances Section of the Minnesota Test.2 The
subtests used from this section of the test were as follows:
recognizing common words; discriminating between paired
words; recognizing letters; identifying items named serially;
following directions; repeating digits; and repeating sen-
tences. All of the stimulus material for these tasks were
presented auditorily.

It was reported, on the basis of a factor analysis,
that all these tests loaded heavily on a general language
factor, indicating the importance of auditory processes in

3
language.

Meaningfulness: levels of hearingyvocabulary.
The meaningfulness of a word was determined by the develop-

mental period in which the word was expected to have become

 

lHildred Schuell, Minnesota Test for Differential
Diagnosis of Aphasia: Administrative Manual (Minneapolis:
University of Minnesota, 1965).

21bid.. pp. 5-8.

3Hildred Schuell, Differential Diagnosis of Aphasia,
p. 25.

-11..

a part of subject's hearing vocabulary, i.e., his auditory
receptive vocabulary. Those words from the earliest develop-
mental period were considered more meaningful than words
from the later developmental period.

The word stimuli used were those appearing in the
Peabody Picture Vocabulary Test.1 This test provided an
empirically derived set of words, scaled in terms of the
chronological develOpment of hearing vocabulary through at
least the first eighteen years of life.2 On the basis of
age group data, the stimulus words of the test were arranged
in the empirically—determined order of difficulty. The
order of difficulty was based solely on the number of correct
responses rendered by successively older age groups.

A subject's hearing vocabulary was assessed on the
basis of his word recognition score obtained on the Peabody
Test; it was assumed the subject recognized the stimulus
word if he correctly identified the response picture asso-
ciated with the stimulus word. Levels of hearing vocabulary
were defined by arbitrarily dividing the scaled set of
words into three groups. The division points were at words
indexing mean expectations for five year old subjects, and

for subjects approximately 15 years-6 months of age. This

 

lLloyd Dunn, Peabody Picture Vocabulary Test:
Expanded Manual (Minneapolis: American Guidance Service,
Inc., 1965).

2

 

Ibid., p. 25.

-12-

division assigned test items 1—50 to Level I, items 51-100
to Level II, and items 101-150 to Level III. By determining
the number of correct responses made by the subject within
each of the three levels, his available hearing vocabulary
for single words was analyzed in terms of words associated
with the initial, intermediate, and late phases of the early

development of hearing vocabulary development.

Increased stimulus duration.--As used in this study,

 

increased stimulus duration referred to increasing the
temporal value of phonemic segments of the words. By a
process to be described, the increases were produced as a
percentage of the duration of the word as originally recorded.
Three nominal levels of increase were employed: 125%, 150%,
and 175% increase. Hence, if a word (as originally recorded)
had a duration of 500 msec., it would approximate 750 msec.

at the 150% level of increased duration. This processing

is frequently referred to as expanded Speech.

Organization of the Report

In Chapter I, the direction in which concepts of
aphasia have moved in recent years has been indicated,
and the Special concerns and interests of Speech patholo-
gists have been identified. The questions posed by this
study were presented and related to the needs of the

aphasic adult in his efforts to regain better language

-13-

function. In addition, terms of particular importance to
this study are defined.

Chapter II consists of a comprehensive review of
literature pertinent to: (1) the development of current
concepts of aphasia, (2) the neurophysiological basis for
the possible outcomes from the experimental treatment,

(3) clinical observations which suggested the selection

of the experimental variable, and (4) matters related to
the selection, design and development of the language tasks
and the preparation of the stimulus materials.

The following tOpics are discussed in Chapter III:
procedures employed in develOping materials for the experi-
mental tasks: selection and grouping of subjects; and pro-
cedures followed in presenting the tasks to the subjects.
Included also is a discussion of procedures used in verify-
ing the expanded durations of the stimulus material, plus
a description of the characteristics of the two groups of
subjects.

In Chapter IV the results of the statistical analyses
are discussed in terms of the null hypotheses generated
from the six questions stated in Chapter I. The results
are discussed in terms of relevant theoretical notions,
clinical observations and research findings.

Chapter V summarizes the present study and presents
conclusions that can be drawn within the limitations imposed.

Recommendations for future research are suggested.

CHAPTER II

REVIEW OF THE LITERATURE

This chapter presents a review of the theoretical,
clinical and research literature relevant to this inves-
tigation. Beginning with a consideration of the develop-
ment of understandings of aphasia, the importance of auditory
language reception problems in aphasia is indicated. The
discussion focuses on theories of neurophysiological func-
tion that might account for the possible outcomes of the
experimental treatment employed in this investigation.

This review reports the clinical observations that prompted
the investigation, as well as factors reported to influence
auditory language reception in adult aphasics. The chapter
concludes with a consideration of reports concerning the
process employed to produce increases in the duration of the

stimuli.

DevelOpment of Concepts of Aphasia

The history of the study of aphasia has been marked
by a great deal of variation in the concepts concerning the
nature of the language disorder. This is not at all unusual

when one considers the comment of Osgood and Miron, who

-14-

-15-

observe that
. . . the problems of aphasia represented a natural
intersection of the interests of neurologists, clin-
ical psychologists, linguists, Speech pathologists,
and experimental psychologists specializing in language.
Even a simplified definition of aphasia as "disturbance
of language behavior" gives balanced weight to these
fields: "disturbance" (neurology and clinical psychology),
"language" (linguistics and speech pathology), and
"behavior" (experimental psychology).1
While the early reports appear to have been largely descrip—
tive medical accounts, later efforts tried to explain the
language deficit in terms of an underlying neurological
dysfunction. In more recent years, emphasis appears to
have Shifted to a more systematic investigation, aimed
at describing the actual language deficits in ways meaning-
ful to the treatment of the language disorder itself.
In the brief review that follows, the intention
is to indicate the direction in which the concepts of aphasia
have moved in a little over 100 years. The review reflects
the development of the realization that any concept of the

language disturbances in aphasia must include considera-

tion of both receptive and expressive language functions.

Early develOpments.—-A ccnsiderable amount of de-
scriptive clinical material had been accumulated up to the

beginning of the nineteenth century. This was the conclusion

 

1Charles E. Osgood and Murray S. Miron, Approaches

to the Study of Aphasia (Urbana: University of Illinois
Press, 1963), p. 3.

 

~16-

reached by Benton and Joyntl after their reasonably extensive
review of literature dating from Hippocratic writing to
1800. They felt that the writers of this period (largely
physicians) had recorded a rather broad range of aphasic
disturbances, with one notable exception. Other than one
reference, Benton and Joynt found no mention of specific
impairment in speech comprehension as an aSpect of aphasic
disturbances. Benton and Joynt state

. . . sensory aphasia remained virtually unrecognized

as a Specific language deficit for another three-

quarters of a century, until the appearance of Wernicke's

monograph in 1874.2

Wernicke regarded the disturbance discussed by

Broca3 as a motor aphasia. Wernicke proposed another type
of language disturbance which he labelled sensory aphasia.
He reasoned that, while the two language deficits (sensory
and motor aphasia) could exist as separate aSpects of an
aphasic disturbance, he was not convinced that in pure motor
aphasia the ability to understand speech always remained
unimpaired. The coexistence of the two deficits he seemed

to ascribe to another concept of conduction aphasia, about

which he had some uncertainty. Wernicke defined sensory

 

1Arthur L. Benton and Robert J. Joynt, "Early
Descriptions of Aphasia," AMA Archives of Neurology, III
(1960). pp. 205-22.

21bid., p. 220.

3Jules Kann, "A Translation of Broca's Original
Article on the Location of the Speech Center," Journal of
Speech and Hearing Disorders (JSHD), XV (1950), pp. 16-20.

-17-

aphasia as loss of understanding of Speech, with hearing
and articulation intact, and writing severely disturbed.
He stated that sensory aphasia resulted from temporal lobe
lesions in the area of the auditory center of the first
temporal convolution--"Wernicke's area" in classical neur-
ology.l

A little over a decade later in 1887, Kussmaul
had formulated a definition of aphasia which seemed to
accommodate more comfortably the sensory and motor aspects
of language. To understand aphasia, he wrote, it was
important to understand the development of Speech in its
linguistic and utilitarian contexts. He defined aphasia as
"the impairment of execution, expression or understanding
of any symbols by which one man communicates his ideas or
feelings to another."2 With this broad view of the problem,
it is not surprising to find that Kussmaul did not support
a strict neurological localization point of view, as did

Broca and Wernicke.

Hughlings Jackson, initiating some thirty years

 

lCarl Wernicke, "Symptom-Complex of Aphasia (German-
1874)," in Diseases of the Nervous System, ed. by A. Church
(New York: Appleton, 1908), pp. 265-324, cited by Hildred
Schuell, James J. Jenkins and Edward Jimenez-Pabon, Aphasia
in Adults: Diagnosisy Prognosisy and Treatment (New York:
Harper & Row, Publishers, 1964), p. 17.

2Adolph Kussmaul, "Disturbances of Speech: An
Attempt in the Pathology of Speech," in Cyclopedia of the
Practice of Medicine, ed. by H. vonZiemssen (New York:
Frank Wood, 1887), pp. 581—870.

-18-

of careful clinical observation at about this same time,
was also unable to accept popular theories of localization.
While his work was not widely recognized until the early
twentieth century, he made some very significant contribu-
tions to the study of this disorder.1 Jackson's study of
aphasia caused him to search for a dynamic concept of
cerebral function which involved the dependency of lower
functional centers upon higher levels of functional organ-
ization. He was aware of the distinct receptive language
deficit manifested by aphasic patients, although he con-
centrated more attention on deficits in expressive language.
However, Jackson certainly perceived the reciprocal rela-
tionship between the receptive and the expressive aspects
in aphasic disorders. He discussed the apparent word loss
of the aphasic as more a matter of words not being avail-
able for prOpositional expression than an absolute loss of

the words.2

Recent developmentS.--It appeared that the first

 

recognition (in the United States) that aphasia was neither

purely receptive nor purely expressive was in the 1935

 

lTheodore Weisenberg and Katherine E. McBride,
Aphasia: A Clinical and Psyghological Study (originally
published 1935), (New York: Hafner Publishing Company,
Inc., 1964), p. 11.

2J. Hughlings Jackson, The Selected Writings (2
vols.; New YOrk: Basic Books, Inc., 1958), II, pp. 365—392.

-19-

work of Weisenberg and McBride.l In introducing their
fourfold division of aphasic disorders, they state
. . . it must be emphasized that none of the terms
is other than descriptive and that none does more than
point out the most marked characteristics of the case.
Neither patients of the eXpressive nor those of the
receptive group are handicapped solely in expressive
functions or solely in receptive. The disorders are

predominantly expressive or receptive; the other pro-
cesses in either case are always more or less affected.

2
The four classifications of expressive, receptive, expressive-
receptive, and amnesic were derived from an empirical study
of sixty aphasic patients.

While there was a good deal to criticize in this
study,3 the explicit statement concerning the coexistence
of receptive and expressive deficits is noteworthy. It
seemed Weisenberg and McBride were more explicit in this
regard than was Head.4 Head also avoided the sensory-motor
dichotomy in his classifications of verbal, syntactical,
and nominal. It will be noted that Head's observations
lead him to a linguistically oriented classification.

Closer to the present time, one finds a nomencla-

ture for aphasic language disturbances in Eisenson's work

 

lWeisenberg and McBride, Aphasia, pp. 145-149.

2Ibid., p. 144.

3Hildred Schuell, James J. Jenkins and Edward
Jimenez-Pabon, Aphasia in Adults: Diagnosisy Prognosisy
and Treatment (New York: Harper and Row, Publishers,
1964). pp. 41-42.

4Henry Head, Aphasia and Kindred Disorders of
Speech (2 vols.; New York: Macmillan, 1926), I, pp. 221-68.

 

-20-

which is apparently a modification of the Weisenberg and
McBride work. Eisensonl avoided the sensory-motor dichot-
omy in proposing the following four classes: predominantly
evaluative (receptive or sensory); predominantly productive
(expressive); subsymbolic disturbances on evaluative side
(agnosias); and subsymbolic disturbances on the productive
side (apraxias and dysarthrias). Pointing out that these
categories were the predominant area of disturbance, not
pure disturbances, he suggested that this classification
system points to the area in which the major treatment
emphasis Should be focused. It is worth noting that Eisenson
allowed, in his nomenclature, for separate behavioral defi-
cits of agnosia, apraxia, and dysarthria. This recognition
of nonlanguage disabilities, coexisting with aphasia, has

been reflected in other recent theoretical positions.

Present status.--In 1960, Wepman, 3p 31.,2 develOped
their formulation of aphasia as a disruption in the central
nervous system's capacity to integrate language stimuli
once such stimuli are free of their input modalities, with

the disruption having the potential of falling at any point

 

lJon Eisenson, Examining for Aphasia (rev. ed.;
New York: Psychological Corporation, 1954), pp. 12-19.

2Joseph M. Wepman, §p_gl,, "Studies in Aphasia:
Background and Theoretical Formulations," Journal of Speech
and Hearing Disorders, XXV (1960), pp. 323-32.

 

 

-21-

in the process. They have suggested such points as being
in the arousal of meaning, in the semantic word selection
process, or in the syntactic process. They regarded agnosias
and apraxias as nonsymbolic transmissive disorders related
to receptive and expressive language modalities, while the
aphasias were viewed as disruptions in the symbolic language
process.

In setting out to research this concept of aphasia,
Wepman and Jones develOped the Language Modalities Test for
Aphasia (LMTA)1. This test was designed to explore not
only the stimulus-response relationship along common input
and output modalities but also three types of linguistic
involvement commonly found in aphasic patients. Their
research indicated, to them, that one way to differentiate
among symbolic disorders associated with central nervous
system impairment was to classify them in terms of the
loss or dysfunction in one or more of three processes.
The three processes that they defined and labeled were
pragmatic aphasia, semantic aphasia and syntatic aphasia.
They added a fourth category (jargon aphasia) for patients
generally unintelligible; and a fifth category (global
aphasia) for patients unable to produce a real verbal

response to stimuli. From a study by Spiegel, Jones and

 

1Joseph M. Wepman and Lyle V. Jones, The Language
Modalities Test for Aphasia (Chicago: Education-Industry
Service, 1961).

-22-

Wepman,l it was concluded that aphasic patients evaluated
with the LMTA were distributed along a continuum of general
language disability. In addition, the patients also dis-
tributed themselves along other dimensions indicative of
Specific losses in more specialized language abilities
related to input and output modalities.

The theoretical position that most strongly supports
the concept of a general language deficit in aphasia is
expressed by Schuell, Jenkins, and Jimenez-Pabon.2 They
have viewed aphasia as a general language deficit crossing
all language modalities and as a deficit that may or may
not be further complicated by impairment of auditory, visual
and sensorimotor processes. In fifteen years of empirical,
systematic observations of some 1000 aphasic patients,
these investigators saw the language deficit itself as
characterized by a reduction of available vocabulary, im-
paired verbal retention Span, and impaired perception and
production of messages. The impaired perception and pro-
duction of messages, they have stated, may be secondary to
the impairment of available vocabulary and verbal retention

Span.

 

lDouglas K. Spiegel, Lyle V. Jones, Joseph M. Wepman,
"Test Responses as Predictors of Free-Speech Characteristics
in Aphasic Patients," Journal of Speech and Hearing Research
(JSHR). VIII (1965), pp. 349-362.

Schuell, Jenkins, and Jimenez-Pabon, Aphasia in
Adults. PP. 113-133.

 

-23-

A major outcome of the work of Schuell and her
associates was the Minnesota Test for Differential Diagnosis
of Aphasia,l a test intended to provide an operational
description of aphasia through observation of the level at
which language performance breaks down in each of the prin-
cipal language modalities. This, Schuell stated,2 is essen-
tially what there is to observe in aphasia. A section of
the Minnesota Test for Differential Diagnosis of Aphasia
was used in this study.

This review of several major contributions supported
the position that an investigation of an aspect of the
receptive auditory language ability of adult aphasics was
an investigation having relevance to general problems of
aphasia. Receptive language and expressive language cannot
be dichotomized, and the auditory modality is certainly a
principal receptive language modality.

Among the more recent descriptive studies of aphasia,
the empirically derived concepts presented by Schuell and
her associates provided the basis for this investigation
of the influence of increased durations of Speech stimuli

on the receptive language performance of aphasics. Their

 

lHildred Schuell, Differential Diagnosis of Aphasia
with the Minnesota Test (Minneapolis: University of Minn-
esota Press, 1965).

2Schuell, Jenkins, and Jimenez-Pabon, Aphasia in
Adults, p. 170.

 

-24-

two principal characteristics of the language deficit in
aphasia, a reduction in available vocabulary and an impair-
ment in verbal retention Span, suggested the two tasks on

which the experimental variable will be tested.

Neurophysiological Rationale

 

While this study was primarily concerned with the
language behavior of aphasics, an attempt was made to find
some neurOphysiological basis for the possible influences
of the increased durations of the auditory stimuli. Several
current theories of the functional organization of the cere-
brum were explored. One of these seemed to predict no
positive effect, and three which appeared to offer some
suggestion of a positive effect from increasing the dura-
tion of the stimuli. At the outset it was realized the
task could not be completed with a high degree of clarity,
since the perceptual and cognitive aSpectS of auditory
language reception were still cloaked in too much uncertainty,
ambiguity and conflicting Opinion. Yet the undertaking

seemed a necessary preliminary step in this investigation.

Localization position.--A present day reflection

 

of more classical neurophysiological concepts appeared in

the work of Geschwind.l Geschwind has presented a rather

 

lNorman Geschwind, "Disconnexion Syndromes in Ani-
mals and Man, Parts I and II," Brain, LXXXVIII (1965), pp.
277-291, and pp. 604-44.

-25-

extensive discussion of disconnection syndromes and their
influence on language behavior. In addition to dealing with
disruptions in connections between primary and secondary

(or associative) areas of the cortex, he placed importance
on the inferior parietal lobule (roughly Brodmann's areas

39 and 40) as a key site in cortical integration. This

area is uniquely develOped in man, Geschwind points out.

Not essentially concentric with any of the primary sensory
or motor projection areas, this area--"the association area
of association areas"--may be capable of powerful intermodal
associations. The loss or reduction of intermodal associa-
tions between audition, vision and some kinesthetic sensa-
tions would undoubtedly lead to impaired language function.
He discussed the disconnections in terms of both intra-
hemispheric and inter-hemiSpheric fibers. From Geschwind's
position, it appeared that once the connections are dis-
rupted the resulting loss or attenuation of function is
permanent and irreversible. The mediating influence of the
"association area of association areas" over the transactions
of the sensory and motor areas is no longer as effective

or efficient as it was prior to the cerebral insult.

It seemed that, from the vieWpoint of Geschwind's
transcortical model, the experimental treatment of the
speech stimuli in this study would have no positive influ-
ence on the aphasic subject's ability to recognize or retain

the stimuli.

-26-

Moderate localizationyposition.--An orientation
toward cerebral function that appeared to hold more promise
in terms of the experimental treatment is the Pavlovian
reflex theory of sensation and his concept of cortical
"analyzers." These views have been investigated and set

forth by Luria, a very thorough presentation appearing in

1962.1 According to Luria, the reflex theory of sensation

. . . is always an active reflex process associated
with the selection of the essential (Signal) compon-
ents of the stimuli and the inhibition of the nonessen-
tial, subsidiary components. It always incorporates
effector mechanisms leading to the tuning of the peri-
pheral receptor apparatus and reSponSible for carrying
out the selective reactions to determine the signal
components of the stimulus. It envisages a continuous
process of increased excitability in respect to some
components of the stimulus and of decreased excitabil-
ity in respect to others . . . In other words sensation
incorporates the_process of analysis and synthesis of
signals while they are still in the first stages of
arrival . . . The units of any sensory process (includ-
ing hearing) are not only acts of reception of individual
signals, measurable in terms of thresholds of sensation,
but also acts of complex analysis and integration of
signals, measurable in units of comparison and discrim-
ination.

 

 

The cortical region having to do with the most precise
differentiation and the most complex integration of audi-
tory stimuli was part of the "nuclear zone“ for audition

located in the superior temporal subregion. The "nuclear

 

lAleksandor Romanovick Luria, Higher Cortical Func-
tions in Man, trans. by Basil Haigh, (New York: Basic
Books, Inc., 1966).

2Ibid., p. 97.

 

-27-
zone" was defined as Brodmann's Areas 41, 42 and 22.1
Injuries to Area 41, the primary area of the analyzer, have
produced marked decreases in detection and discrimination
of auditory stimuli. However, lesions in the secondary
areas (Brodmann's 42 and 22) have resulted in quite differ-
ent auditory dysfunction, according to Luria. These lesions,
in the left hemisphere, have lead to disorganization in
the ability to cope with complex groups of stimuli and the
relationships among component parts of complex stimuli.
This dysfunction in the "analytic-synthetic" capa-
city is said to lead to difficulty in the identification of
the phonemic signs in the acoustic patterns of the language
system.2 Because the "auditory analyzer" is closely asso-
ciated with the cortical apparatuses of kinesthetic (arti-
culatory) analysis, the efferent link for the perception
of the sounds of Speech was not sufficiently activated.
These secondary areas possess wider systemic connections
with adjacent cortical areas than does the primary auditory
area. Direct stimulation of the secondary areas has given
rise to potentials detected in the inferior segments of
the premotor frontal portions of the cortex--Brodmann's
Areas 44, 46 and 10.

There was reason to suspect that the experimental

 

lIbid., p. 44.

2Ibid., pp. 102-03.

-28-

treatment might activate sufficiently the post-cerebral
insult capability of these secondary areas. Luria has cited
studies that indicate the induction of potentials in these
secondary auditory areas requires more intensive stimula-
tion of the peripheral receptor than is required to induce
potentials in the primary auditory area.1 It seemed the
increased durations of the Speech stimuli might provide just
such an intensification of the phonemic elements. Through
the neurophysiological principles of facilitation and sum-
mation, this intensification may have improved perception
and enabled better recognition and retention of the word
stimuli. It was important to recognize that just a moderate
increase in the sensory aspect of perception might be suffi-
cient to trigger the efferent (motor) aspect of the percep-
tual action, hence sharpening the subject's perception of
the stimulus to the point where linguistic meaning was

aroused.

Subcortical integration.--A second basis for hypo-

 

thesizing a positive influence from the increased durations
of the stimuli on the tasks of this investigation seemed

to be found in theories of cerebral function which deempha-
sized transcortical pathways. Reducing reliance on the

extent of post-injury intactness of cortical pathways seemed

 

lIbid., p. 99.

-29-

to place less demand on the residual, functional adequacy
of both the primary cortical areas and the associative cor-
tical areas. It appeared that the intensification or am-
plification of the phonemic elements would have a greater
likelihood of positively influencing performance if the
residual adequacy of both primary and associative areas
could be maximally integrated without depending heavily
on the extent of the intactness of transcortical pathways.
Two theories of functional integration of the cere-
bral cortex seemed to offer such a basis. They were the
theories advanced by Penfield and Roberts and by Pribram.
Penfield and Roberts1 emphasized cortical-subcortical
relationships in the matter of cortical integration. They
relegated transcortical connections to a secondary role.
They have drawn, as support for this position, evidence
derived from the functional adequacy of patients who have
experienced radical excision of cortical tissue bordering
on primary sensory and primary motor areas. They have
named this system the "centrencephalic system" and defined
it roughly as a ganglionic area of the higher brain which
acts as the central integrating mechanism--the point at

which integration of neuronal transactions reaches its peak.2

 

lWilder Penfield and Lamar Roberts, Speech and
Brain Mechanisms (Princeton, N.J.: Princeton university
Press, 1959).

 

 

2Ibid.. pp. 21-22.

-30-

This postulated interdependency of subcortical
centers and cortical centers, they suggested, is a complex
one. Impulses triggered in subcortical areas take on the
aspects of cortically bound efferent impulses. As efferent
impulses aimed at the cortex, they activate one or more
cortical areas. The activated cortical areas provide a
mediating influence over the impulses having subcortical
origin. In this mediating function, certain modifications
may be made before the patterned (or repatterned) impulse
complex travels along the post-cortical leg of the pathways
to the peripheral effector mechanism.

In the view of Penfield and Roberts, subcortical
integration functionally unites the two cortical areas
they have indicated enable the elaboration of language
through speech. The most important of the two areas for
language function is the Posterior Speech Cortex, which
encompasses Brodmann's Areas 39 and 40. This is the area
to which Geschwind attached great significance in terms of
intermodal associations. The Posterior Speech Cortex also
encompasses most of the classical Wernicke's area. Of
secondary importance to language function was the Anterior
Speech Cortex Situated in the frontal lobe, encompassing
Broca's area, Area 44.1 It will be recalled that Luria

reports radiations from secondary acoustic fields reach

 

lIbid.. pp. 192-204.

-31-

Area 44. While these points suggested areas of correspon-
dence in the three rationales, a very important and dis-
tinguishing difference rested in the reliance Penfield and
Roberts placed on subcortical integration.

Pribraml has proposed an alternative to transcor-
tical models of cerebral organization in cognitive processes.
He cited evidence from several sources which, he believed,
made transcortical models unsatisfactory. One type of
evidence indicated the classical dichotomy between motor
(efferent) and sensory (afferent) cortical areas was not
valid.2 Secondly, while the first issue might be solved
by designating the areas "primary motor (or sensory) pro-
jection areas," the functional relationship of these primary
projection areas to their association areas was not sub-
stantiated by various surgical intervention techniques
applied to the cortex. Again, the transcortical model
becomes less useful when it was observed that the behavioral
changes following such removal, isolation or cross-hatching
of the cortex were not those that would be predicted by the
transcortical models.3

Pribram suggested an alternative model based on

1Karl H. Pribram, "Neocortical Function in Behavior,"
in Biological and Biochemical Phases of Behavior (Madison:
University of Wisconsin Press, 1958), pp. 151-72.

2Ibid., pp. 155—158.

3Ibid.. pp. 159-166.

-32-

the thalamocortical connection that the cortical area has
with the thalamus. If the thalamic nucleus from which the
thalamocortical connection arose received inputs from outside
(extrinsic) the thalamus, the cortical area was regarded as
an extrinsic sector. Intrinsic sectors of the cortex are
those sectors having thalamocortical connections originat-
ing in thalamic nuclei which received inputs from within
(intrinsic) the thalamus.

Pribram reported the extrinsic sectors correSpond
to the motor-sensory projection areas, whereas the intrinsic
sectors were the classical association areas. With direct
functional connections between projection areas and asso-
ciations being doubtful, the functional Significance of
the subdivisions of the intrinsic cortex was believed to
be the convergence of their output on a common subcortical
mechanism. This mechanism also receives output from the
extrinsic sectors. The output from the intrinsic system was
seen as influencing the output of the extrinsic sectors
through this convergence.l Without denying the specificity
of function associated with certain primary and association
areas, a model of cortical function resembling Penfield
and Roberts' "centrencephalic system was developed."

In both the Penfield-Roberts and Pribram concep-

tions, a hypothesized system was seen which seems more

 

lIbid.. pp. 167-168.

-33-

resilient and capable of compensating for injury. Working
from these systems, there appeared to be some reason to
suppose that Speech stimuli, in which phonemic durations
were increased, would have some positive consequences on
the auditory functions of aphasic subjects. Luria's theor-
etical views might encompass such an outcome, but a posi-
tive influence from the intended stimuli would be somewhat
less probable. From Geschwind's point of view, the pre-
diction of no significant effect from the expanded speech

stimuli appeared to be the most logical.

Clinical Basis of Investigation

The short history of language therapy for the
aphasic dates back to the period immediately after World
War I and seemed to have made its most significant advance
since World War II in response to the needs of those who
sustained brain injury in that war.1

From his experiences in treating aphasics during
World War II, Wepman reported the major receptive disability
of the aphasic adult was in the comprehension of the Spoken
word. The most common deficit in this modality, which he

labeled "auditory verbal agnosia," resembled the sensory

aphasia described by Wernicke. However, Wepman reported

 

lJoseph Wepman, Recovery from Aphasia (New York:
Ronald Press Company, 1951), p. 4.

-34-

. . . patients showing this problem have all grada-
tions between comparatively normal Speech production
and a total lack of that facility, between a total

lack of comprehension and a partial lack of the ability
to understand spoken words, and between a retained
ability to imitate and an apparent total lack of imi-
tative ability.l

After necessarily ruling out the influence of impaired
hearing acuity and general intellectual impairment, the
comprehension loss remains, indicating the cortical disabil-

ity of integrating aural stimuli, he stated.

Facilitating auditory language function.--The
treatment approach recommended by Wepman, and endorsed by
others such as Longerich and Bordeaux,2 is to employ other
sensory inputs in combination with auditory stimulation.

Wepman notes

In this area of disturbance more than any other
must the total available resources of the patient be
utilized. Where auditory stimuli are meaningless, the
patient must be approached from the tactile-auditory,
the visual-auditory, the kinetic-auditory and the com-
bined visuo-tactile-kinetic-auditory. It should be
noted that in each of the proposed approaches the audi-
tory function is an integrated part of the training.
The end product is not an ability to develOp a substi-
tute for hearing, nor by circuitous methods to hope
for a transfer of training, but to redevelop in the
patient his use of the disturbed auditory function.3

 

lIbid., p. 199.

2Mary C. Longerich and Jean Bordeaux, Aphasia Therapeu-

tics (New York: Macmillan Co., 1954), pp. 17-18.

3Wepman, Recovery from Aphasia, p. 199.

-35-

In terms of this objective, increasing the duration of the
auditory stimuli was viewed as an attempt to accelerate
the redevelopment of the disturbed auditory function by
making the auditory stimuli more immediately usable to the
patient.

Schuell has maintained that auditory stimulation
is the foundation of all aphasia therapy and is essential
for all patients.1 She continues to stress the importance
of auditory stimulation, stating in 1965

All evidence suggests that auditory stimulation is
crucial in control of language processes. However,
since feedback from more than one sensory modality may
contribute to behavior, there is no reason for using
this mode exclusively. This suggests that the first
principle of treatment for aphasia should be the use

of intensive auditory stimulation, although not neces-
sarily stimulation through auditory channels alone.2

Nature of aphasic auditory language dysfunction.--
In addition to intensifying the stimulus appropriately,
Schuell has discussed making the stimulus as adequate as
possible for the patient. Many clinicians have heard patients
make statements like those which Schuell referred to when

she wrote

 

lHildred Schuell, "Auditory Impairment in Aphasia:
Significance and Retraining Techniques," JSHD, XVIII (1953),
pp. 14-21. Hildred Schuell, Virginia Carroll and Barbara
Stansell, "Clinical Treatment of Aphasia," JSHD, XX, (1955),
pp. 43-53.

2Schuell, Jenkins, Jimenez-Pabon, Aphasia in Adults,
p. 338.

-36-

People talk too fast and say too much at a time
for the aphasic patient to follow. Patients sometimes
tell you that people do not seem to be talking right.
To some patients it often seems they (peOple) are not
even talking a language he knows.

Luria has made a comment which seemed relevant to
this type of aphasic experience. Discussing the role of
the “auditory analyzer" in language and speech comprehension,
he has written

It is clear, therefore, that the perception of speech
by hearing requires not merely delicate but systematized,
hearing. When this selection of the essential, phonemic
signs is no longer possible, phonemic hearing is dis-
turbed. This is why the boundary between hearing speech
and understanding it loses its Sharp distinction. A
person ignorant of a foreign language not only does

not understand it, but does not even hear it, i.e.,

he does not distinguish from the flow of sounds the
articulated elements of the language and does not sys-
tematize the sounds of speech according to its laws.

The unfamiliar language is thus perceived by that per-
son as a stream of unarticulated sounds, not only im-
possible to understand, but inaccessible for accurate
auditory analysis.2

To the clinician aware of the problems confronting the

aphasic patient, this statement must have a familiar ring.
Schuell discussed ways of manipulating the speech

stimuli themselves so as to intensify or improve their ade-

quacy for the patient. The objective was to enhance the

patient's linguistic perception of the language units.

The suggested variables were as follows: the loudness of

the stimulus; the meaningfulness of the language units;

 

lIbid., p. 339.

2Luria, Higher Cortical Functions, p. 102.

-37-

the length of the stimulus; and the duration of the stim-
ulus.l

Basing the tasks of this investigation on the
theoretical notions of Schuell and her associates, it seemed
logical to consider these variables in the design of the

investigation. The four variables are given close considera-

tion in the following section.

Experimental Tasks and Associated Variables

In this section the discussion centers on the
rationale for selecting the two language tasks employed in
this study and the variables considered in structuring the

tasks.

Selecting experimental taskS.--The tasks of word
recognition and serial retention were employed in this
study as procedures for operationalizing the two primary
characteristics of aphasia identified by Schuell and
Jenkins.2 They believed the language deficit in aphasia
is characterized primarily by a reduction in available
vocabulary and an impairment of verbal retention span.

Impaired perception and impaired production of messages

 

lSchuell, Jenkins, Jimenez-Pabon, Aphasia in Adults,
pp. 339-40.

2Ibid., p. 113.

-38-

are additional characteristics which they felt may be
secondary to the two primary characteristics. These char-
acteristics of the language deficit are to be found in all
language modalities to some extent and may be further im-
paired by primary disturbances in auditory, visual and
sensorimotor processes.

To assess available vocabulary in the auditory
modality, the word recognition task used seemed appropriate.
Expecting the patient to point to a picture having the
most meaningful association with a word spoken by the examiner
is a task involving auditory recognition and the gross motor
performance of pointing. Auditory recognition for single
words might be considered a very low level test of a sub-
ject's available vocabulary. Successful performance on
this task indicates a capacity to link the auditory repre—
sentation of the referent with a visual representation of
the referent. The recognition of stimulus equivalence
attested to the availability of the word meaning in the
subject's receptive vocabulary.l

In the second task it was necessary for the subject
to retain a series of auditorily presented words, and then
(maintaining the order of presentation) identify the pic-
tures which had the most meaningful association with the

words that had been presented. The period of retention

 

lIbid.. pp. 149-53.

-39-

being on the order of from two to five seconds before the
subject could begin his response, this task constituted

a minimal test of auditory retention span. However, ex-
pecting the subject's reproduction (by pointing) to reflect
the order of the auditory stimuli made this a more complex
task than it would have been were order not required. It
also reduced the likelihood of a subject's guessing a cor-
rect reSponse.

The question of the relationship of performance on
these tasks to language behavior was answered from the
statistical analysis of the tasks comprising the Minnesota
Test for Differential Diagnosis of Aphasia. The Minnesota
Test included tasks Similar to the two tasks employed in
this study. In the factor analysis of the battery of tasks,
the word recognition task and the serial retention task
loaded high on the factor identified as a general dimen-
sion of language behavior.1 The retention task, loading
highest on this factor, had its second highest loading on
a factor which appeared to represent "visual discrimination,
and ensuing recognition and recall of learned visual pat-
terns."2 The factor on which the word recognition task had

its highest loading was a factor involving recognition of

 

lIbid.. pp. 137-142.

2Ibid., p. 143.

-4o-

stimulus equivalence.l Aspects of the factor Of general
language dimension were identified as: vocabulary, audi-
tory retention span, and the ability to deal with language
formulations at various levels of complexity.2 Therefore,
it seemed the two tasks employed in this study were sampling
language as it has been identified by these investigators.
The linguistic perception of auditory stimuli by
adult aphasics is said to be influenced by four factors.3

The following is a consideration of these factors.

Loudness of the stimulus.--In addition to being

 

alert to the incidence of moderate to severe hearing losses
in aphasic patients,4 it has been suggested there may be
an Optimal loudness level for maximum auditory language
reception. Except in cases of clearcut perceptual prob-
lems, most patients prefer listening at ordinary conver-

sational levels in a free-field.5

Meaningfulness of the stimuluS.--It is well estab-

 

lished that meaningful units of language are recognized

 

lIbid., p. 151.

2Ibid., p. 153.
3Ibid., pp. 338—43.

4Barbara Stansell Street, "Hearing Loss in Aphasia,"
JSHD, XXII (1957). pp. 60-67.

5Schuell, Jenkins, Jimenez-Pabon, Aphasia in Adults,
p. 340.

-41-

more easily than non-meaningful patterns. In the sense
that all stimuli used in this study were words in the native
language of the subjects, all the stimuli were meaningful.

On the word recognition task, an attempt was made
to vary systematically the factor of meaningfulness in the
stimulus words Of the task. A measure of meaningfulness
used in several studies of both normal and aphasic language
performance is the frequency with which a given word is
used.1 Aphasic subjects have tended to follow the normal
usage patterns but at reduced levels. This indicated that
such a measure of meaningfulness was appropriate for the
language performance of aphasic adults. The index of usage
has generally been determined from word counts of written
language,2 not spoken language.

Another index of meaningfulness was suggested by
the Peabody Picture Vocabulary Test.3 As empirically de-

rived, the test presented a set of words scaled in terms of

 

lJames Deese, "Form Class and the Determinants of
Association," Journal of Verbal Learning and Verbal Behavior,
I (1962), pp. 70-84. Joseph M. Wepman, pt 31,, "Psycholin-
guistic Study of Aphasia: A Revision of the Concept of
Anomia," JSHD, XXI (1956), pp. 468-77. Hildred Schuell,
James Jenkins and Lydia Landis, "Relationship between Audi-
tory Comprehension and Word Frequency in Aphasia," JSHR,
IV (1961). pp. 30-36.

2Edward L. Thorndike and I. Lorge, Teachers Word
Book of 30,000 Words (New York: Bureau of Publications,
Columbia University, 1944).

3Lloyd Dunn, Peabody Picture Vocabulary Test:
EXpanded Manual (Minneapolis: American Guidance Service,
Inc., 1966), pp. 25-33.

-42-

the chronological develOpment of hearing vocabulary. Since,
on this task of auditory recognition of words, the factor
of meaningfulness was to be varied systematically, this
scaled set of words seemed uniquely suited to assessing
auditory receptive vocabulary in relation to the other
independent variables of the study.
An inspection of the stimulus words of the Peabody

Test, in terms of their frequency of usage in written
English, revealed those words associated with the period
of early auditory language development were very common or
highly frequent words in written English. Words associated
with later periods of auditory language development were
less and less frequent, becoming highly infrequent at the
upper levels of the test. As Schuell, Jenkins, and Jimenez-
Pabon have stated

Both from a linguistic and neurophysiological viewpoint,

we would expect that more would happen in the brain in

response to frequently used words . . . than in response

to rarely used words. . . .1

On a retention task, since the objective would be

to equate the stimuli in terms of meaningfulness, only very
common or highly frequent words would be used (e.g., the
AA and A words of the Thorndike-Lorge General Count). In

addition, on such a task, only one class or form of word

might be used, perhaps the most frequent class of words.

 

lSchuell, Jenkins, Jimenez-Pabon, Aphasia in Adults,
p. 399.

-43-

Recognizing the semantic-syntactic problems connected with
classifying words, it appears that nouns are the most fre-

quently occurring class Of words in the language.

Length of the stimulus.--Controlling this variable
is, reportedly, related to the reduced auditory retention
span always found in aphasia.2 Control has been suggested
by varying the number of syllables per word and the number
of words in the stimulus.

In a retention task, the number Of words in the
stimuli could be varied systematically to create a range
of complexity in the task. The discrete words from which
the stimuli were created could be controlled in terms of

the number of syllables per word.

Duration of the stimulus.--In differentiating this
factor from the factor of stimulus length, duration of the
stimulus refers to the amount of time given to the utter-
ance of elements in the stimulus, whereas stimulus length
refers to the number of elements in the stimulus. Schuell,
Jenkins and Jimenez-Pabon have found

. . . patients with perceptual problems are often able
to respond more adequately when a word or phrase is

spoken a little more slowly than in ordinary conver-
sational speech. However, inflection Should be natural,

 

lRoger Brown, Words and Things (Glencoe, Ill.: Free
Press, 1959), pp. 53-55.

2Schuell, Jenkins, Jimenez-Pabon, Aphasia in Adults,
p. 340.

 

-44-

and slowing should not fragment or distort the language
unit. Like increasing loudness, slowing down the rate
seems to help severely impaired patients initially,
then to make no difference.

The above statement implied that the effectiveness
of increasing the duration of the stimulus is related to
the severity of the auditory language disturbance. Hence,
any test of the effectiveness Of durational increase Should
control for this aspect of the language disorder. Control
over the severity of auditory language deficit might be
achieved through use of a section of the Minnesota Test
for Differential Diagnosis of Aphasia.2

One study has been found that applied this variable
of stimulus duration (as defined here) to the comprehen-
sion of language. This investigation, by Luterman, Welsh
and Melrose,3 was designed to determine the effect Of
stimulus duration on the performance of subjects in respond-
ing to recorded phonetically balanced words. The experi-
mental group consisted Of aged subjects with sensorineural

hearing losses. The experimental conditions included con-

ditions in which the stimulus words were increased in duration

 

lIbid., pp. 340-41.

2Hildred Schuell, Minnesota Test for Differential
Diagnosis of Aphasia: Administrative Manual (Minneapolis:
University of Minnesota, 1965).

3David M. Luterman, Oliver L. Welsh and Jay.Melrose,
"Response Of Aged Males to Time-Altered Speech Stimuli,"
JSHR. Ix (1966). pp. 226-30.

-45-

by a factor of 110% and by a factor of 120%. Other condi-
tions were 10% and 20% time compression of the word dura-
tions, as well as the original duration Of the stimulus
words. The experimental group did not Show improved scores
under the conditions of increased stimulus duration. The
investigators found this outcome surprising, since they
had observed, clinically, some advantage in Speaking slowly
and clearly with aged subjects while engaged with them in
Speech audiometry.

While investigations of the relationship between
rate Of speaking (words per minute) and intelligibility
or listener preference have been conducted, the study cited
above was the only relevant investigation employing the
type of increase in stimulus duration used in this study.
The relationship between Speech compression (decreasing
stimulus durations) and comprehension Of messages has been

investigated with normal subjects and with blind subjects.1

Increasing the duration of the auditory stimuli.2--

Descriptions of two processes for accomplishing the kind

 

1David B. Orr, Herbert L. Friedman and Jane C. C.
Williams, "Trainability Of Listening Comprehension of Speeded
Discourse," Journal of Educational Psychology. LVI (1965),
pp. 148-156. Emerson Foulke, gpupl., "The Comprehension
Of Rapid Speech by the Blind," Journal Of Exceptional Child-
;pp, xxrx (1962). pp. 134-141.

2Since increases in the duration of the stimuli
was the experimental variable of interest in this study,
and because the process by which the durational changes

-46-

of durational changes Of interest in this study have been

published. A computer process for compressing and expanding

the durational aSpects of speech has been reported, but

the use of this process was not economically feasible at

this time.1 In 1954, Fairbanks, Everitt and Jaeger2 pub-

lished a description of an electromechanical apparatus

capable of producing expanded or compressed versions Of

tape recorded materials. It was an electromechanical device

of this nature that was employed in the compressed speech

and expanded Speech studies previously cited. This type

of device was used in this study.3
The device operates on a sampling principle. In

expanding a previously recorded signal, the device increases

the temporal value Of segments of the acoustic patterns of

the original signal and records a second time some portions

 

were produced has had such very limited application, it
seems appropriate to provide a general description of the
process at this point and a more detailed description which
appears in Appendix A.

1Emerson Foulke and Thomas Sticht, "A Review Of
Research on Time Compressed Speech," (from the Proceedings
of the Louisville Conference on Time Compressed Speech,
University of Louisville, 1967).

2Grant Fairbanks, W. L. Everitt and R. P. Jaeger,
"Method for Time on Frequency Compression-Expansion Of
Speech" (1954), in Grant Fairbanks, Experimental Phonetics:
Selected Writings (Urbana: University of Illinois Press,
1966). Pp. 37-42.

3The device used is sold under the name of Eltro
Information Rate Changer-Mark II by Infotronic Systems,
Inc., 2 West 46 Street, N.Y. 10036.

-47-

of the signal. These extra portions (or second samples)
appear in the expanded reproduction of the original signal
immediately following and adjacent to the first reproduction
of the segment. With these "extra samples" of the initial
signal added and the temporal value Of all samples increased,
the expanded reproduction of the signal will have a greater
(expanded) duration than the original recording of the
signal. When that original signal is speech, the expansion
process increases durations of the phonemes in the word,

as it reproduces some portions of the initial recording for
a second time. This durational increase is accomplished
without changing the frequency characteristics Of the ori-
ginal signal in any perceptually significant way. This
expanded reproduction can be recorded in the standard manner.
Depending on the level Of expansion, the word, when expanded,
will be perceived auditorily as being “stretched out" in
comparison to the original version Of the word, free Of the
pitch distortions associated with "slow played" speech

which severely reduces its intelligibility.l At the upper
limits Of the device there is a distracting type Of dis-
tortion in the expanded reproduction which is a function

of the temporal values of the repeated segments.2 The

 

1William R. Tiffany and Delmond Bennett, "Intel-
ligibility Of Slow Played Speech," JSHD, IV (1961), pp.
248-258.

2Fairbanks, Everitt, Jaeger, "Compression-Expansion
of Speech," p. 41.

-43-

upper limit of the device used in this study is 200% expan-
sion (a reproduction having twice the duration of the ori-
ginal signal). There is some indication this distortion
has little influence on intelligibility with normal (non-

aphasic) listeners.l

Summary

This review has indicated that as the study of
aphasia has develOped during the last one hundred years,
it became apparent that the language dysfunction experienced
by the aphasic adult involved both the expressive and recep-
tive modalities. It was possible to discern a trend toward
conceptualizing the language dysfunction as a unidimensional
characteristic, that was operationalized most meaningfully
on the basis of observable language performance. It appeared
these performances could be analyzed from a linguistic
frame of reference. Also, it was possible to take a des-
criptive approach, based on Observations of the aphasic
adult's performance of language tasks.

There appeared to be rather authoritative agreement
that disturbances in auditory language reception are gener-
ally a significant aspect of the total aphasic problem,

and are, therefore, very important in the treatment of the

 

1George H. Kurtzrock, "The Effects of Time and
Frequency Distortion upon Word Intelligibility," (Ph.D.
Dissertation, University of Illinois, 1956).

-49-

1anguage dysfunction. In View of this importance, it
appeared worthwhile to investigate a means of facilitating
auditory language function in aphasic adults. The proce-
dure selected for investigation was increasing the duration
of the auditory stimulus.

There seemed to be some neurophysiological support
for investigating this dimension of a linguistic stimulus.
Several current theories of cerebral function offered a
basis for assuming increased stimulus duration would posi-
tively influence the performance of aphasic adults on cer-
tain language tasks.

In the clinical application of this procedure,
several variables have been described that bear on the
auditory language reception of aphasic adults. Two of
these variables were the subjects' level of auditory lan-
guage function and the meaningfulness of the stimulus material.
These factors were considered in the design of the present
experimental investigation.

Two tasks that appeared to have a high degree Of
relevance to a general dimension Of language dysfunction
in aphasia were word recognition and retention span. There-
fore, auditory tasks involving these capabilities seemed
apprOpriate tasks upon which to test the influence of in-
creased stimulus durations. An electromechanical means
of producing temporal expansions in recorded words was

available and appeared adequate for an initial experimental

-5o-

investigation of increased stimulus durations on the lan-

guage receptivity of aphasic adults.

CHAPTER III

EXPERIMENTAL PROCEDURES

This chapter includes a description of the devel-
Opment of the stimulus materials for the word recognition
task and the serial retention span task, as well as a dis-
cussion of the recording procedures and the manner in which
the durational increases were affected in these recordings.
The chapter also deals with the selection and grouping of
subjects and the procedures followed in presenting the
tasks to the subjects. A description of the scoring pro—

cedure is given.
Selection and Preparation of Stimulus Materials

Word recognition task materials.--The two forms
Of the Peabody Picture Vocabulary Test (PPVT)l were used
as the source of stimulus words and reSponse pictures for
the word recognition task. Each form of the test, consist-
ing of 150 words each, was divided into two lists of seventy-

five words each. The division was made by assigning the

 

1Lloyd Dunn, Peabody Picture Vocabulary Test:

Expanded Manual, (Minneapolis: American Guidance Service,
Inc., 1965).

-51-

-52-

Odd numbered items Of the form to list number 1 and the

even numbered items to list number 2.

Dividing both Form A

and Form B of the test in this manner yielded four lists

Of seventy-five words each,

A

2'

B

1' and B .

2

(See Table 1)

that were designated as A

1’

TABLE 1.--Separation of PPVT Forms into Four Lists and

Three Hearing Vocabulary Levels.

 

PPVT - Form A

PPVT - Form B

 

HEARING VOCABULARY LEVELS

II

III

List
A1

List

A2

List
Bl

List
B2

 

odd items
1 - 50

1

even items
1 - 50

2

odd items
1 - 50

3

even items
1 - 50

4

 

Odd items
51 - 100

5

even items
51 - 100

6

odd items
51 - 100

7

even items
51 - 100

8

 

 

odd items
101 - 150

9

 

even items
101 - 150

10

odd items
101 - 150

11

 

even items
101 - 150

12

 

In terms Of hearing vocabulary,

as defined earlier,

it seemed reasonable to assume the overall difficulty of

the four lists would be equivalent.

With the two forms Of

the PPVT being generally equivalent and dividing each form

as described, this assumption seemed justified.

Perry and

Boswell report a correlation of .92 between the two forms

-53-

when used with aphasic adults.l

Each form of the test was regarded as a scaled set
of words reflecting the chronological development of hearing
vocabulary through the first eighteen tO twenty years of
life. It is the scaled aspect of the words in each form of
the test that permitted the segmentation of each 75 word
list into the three Hearing Vocabulary Levels indicated
in Table l. The arbitrary division Of each list into three
levels provided a means of looking at the available hearing
vocabulary of adult aphasics in terms of three developmental
levels of hearing vocabulary under the four experimental
conditions.

For the purposes of this study, the twenty-five
words at Level I of each list were regarded as reflecting
the initial phase of the early development of hearing
vocabulary. The twenty-five words in the Level II, and
Level III segments reflected intermediate and late phases
Of hearing vocabulary development.

For each of the stimulus words of the PPVT, there
was a response plate consisting Of four pictures. There
were 150 response plates, the same set of 150 plates being
used with Form A of the test as were used with Form B.

While the stimulus words differ for a given numbered item

 

1Peter S. Perry and Nathalie S. Boswell, "Adult
Aphasic Performance on the Peabody Picture Vocabulary Test,"
(paper presented at American Speech and Hearing Association
Convention, November 3, 1967).

-54-

on Form A and Form B, the response plate was the same. One

of the four discrete pictures of each plate was the correct

picture to associate with a stimulus word on Form A of

the test, whereas another picture was the correct picture

to associate with the same numbered stimulus word from Form

B of the PPVT. The two other pictures were foils.
Thirty-five millimeter slides were prepared Of

the entire set of 150 response plates. These slides were

used to project the response plates in front of the sub-

jects, as will be described later.

Serial retention span materials.--Two basic con-

 

siderations entered into the develOpment Of materials for
this task. They were building into the task an adequate
range Of difficulty and maintaining equivalency of the
discrete units of each stimulus.

An adequate range of difficulty was built into
this task in order to promote a reasonable range of scores.
Only with such a range of scores would it be possible to
observe the influence of the experimental treatment. For
this study, the series contained two, three, four and five
words.

Secondly, it was important to keep the words used
in the construction of all series equivalent on certain
relevant criteria in order to stay as close to the defined
purpose Of the task. The three equivalence criteria used

in this study were grammatical class, frequency of usage

-55-

(written English), and hearing vocabulary level.

One grammatical class of words was used. This
class was nouns.

Frequency of usage in written English was deter-

1 Schuell,

mined from the Thorndike-Lorge General Count.
Jenkins and Landis2 demonstrated the Significance of this
variable on the ability of aphasic subjects to recognize
words presented auditorily. All of the words used in the
development of materials for this task had a frequency of
usage equal to or greater than 100 times per million.

This placed the words among the most frequently used
written words of English.

To establish equivalence in terms of hearing vocab-
ulary, the empirically derived scaled word lists of the
PPVT were again employed. For this retention span task,
only objects pictured in the first fifty plates Of the
PPVT were used, providing they met the frequency of usage
criteria. By empirical determination, the names of the

Objects pictured in the first fifty plates were within

the mean expectation for individuals with a chronological

 

1Edward L. Thorndike, and I. Lorge. The Teacher's
Word Book of 30,000 Words (New York: Bureau of Publications,
Teachers College-Columbia university, 1944).

2Hildred Schuell, James J. Jenkins, and Lydia Landis,
"Relationship Between Auditory Comprehension and Word Fre-
quency in Aphasia," Journal of Speech and Hearing Research
IV (1961), pp. 30-36.

-56-

age Of approximately five years.1 This provided some addi-
tional assurance that the words used were a long standing
part Of the subject's hearing vocabulary.

It was determined that thirty words met these three
criteria. Hence, there were thirty stimulus words and
their associated pictures (from the PPVT) available for the
develOpment of materials for this task. The thirty picture
names are given in Appendix B. The arrangement of these
pictures into response plates and the selection of the
appropriate number of stimulus words for each plate was
accomplished in the following manner.

For this task, thirty-two response plates were
required, one for each of the thirty—two stimuli. A series
of discrete words was defined as a stimulus. The four
duration conditions and the four complexity levels (two,
three, four and five word series) created sixteen condi-
tions. Providing two stimuli for each condition resulted
in the need for thirty-two stimuli and associated response
plates. The pictures for each of the eight-picture response
plates were randomly selected, with replacement, from the
pOOl of thirty available pictures. The thirty-two arrange-
ments were photographed on thirty-five millimeter film
and processed into slides. Two response plates were then

assigned randomly to each of the Sixteen conditions. A

 

lLloyd Dunn, Peabody Picture Vocabulary Test, p. 12.

-57-

plate was designated as either plate A or plate B for a
particular cell, indicating the order in which they would
be presented to subjects.

With the assignment of plates to conditions com-
pleted, the appropriate number of words for each stimulus
was selected randomly from each response plate. The num-
ber of words selected depended on the complexity level
to which the response plate had been assigned. The auditory
stimulus consisted of either two, three, four or five picture
names with the word "and" inserted before the final word
of the series. The inclusion of this conjunction made the
stimulus a more complete unit, and enhanced its resemblance
to a language unit.

In Appendix B are found the names of the pictures
included in each response plate and the stimulus associated
with each response plate. The duration conditions to which
the materials were assigned are also indicated in this
appendix.

The particular stimulus duration-stimulus complexity
condition to which the materials were assigned was the same
for all subjects. The discrete units of the stimuli were
equivalent, and the different random order of conditions
was employed for each subject. This randomization proce-

dure will be elaborated in a later section.

-58-

Equipment and Materials

The equipment and materials reported below were
used in the phases of this experiment to be described in

following sections of this chapter.

Equipment.

 

Tape recorders:
Ampex, Model 350-2
Wollensak, Model T-1500
Sony, Model TC-211TS
Microphone:
Electro-Voice, Model 654
Spectographs:

Voiceprint Laboratories
Kay Sonograph, Model 606l-A

Slide projector, rear projection:
SonOSCOpe
Amplifier-Speaker:
Ampex 620
Power Level recorder:
Bruel-Kjaer Level Recorder, Type 2305
Speech compresser-expander:

Eltro Information Rate Changer-Mark II

Materials.

Kodacolor II, 35 mm.

Scotch Brand Magnetic Recording Tape-111

_59-

Recording of Auditory Stimuli

 

One adult male recorded all the stimulus words
for both the word recognition task and the serial retention
Span task. The use of a Single voice eliminated the pos-
sible influence of differences in the fundamental pitch of
voices and variations in voice quality which would be asso—

ciated with the use of more than one speaker.

Equipment.--The speaker recorded the material in

 

the sound isolated environment of a prefabricated sound
treated booth, using a table mounted micrOphone. The
recorder used was Operated at a tape Speed of fifteen inches
per second. The speaker was able to monitor visually the
loudness of his voice by means of the VU meter in the re-
corder. The speaker was successful in keeping the intensity
of the voice signal delivered to the recorder within plus

or minus three decibels of the zero decibel reference on

the VU meter.

Intelligibility.-—An intelligibility check on the

 

recording was made. The recording of the 300 single words
and the thirty-two serial retention Span items was played

to six persons who were not familiar with the Peabody Pic-
ture Vocabulary Test. Each person listened to the recording
by himself, responding to the recorded words by repeating

each word heard. The responses of the six listeners

~60-

revealed that all but five of the words in the original
recording were correctly identified. These five words were
re-recorded into the lists. The segments of the lists
containing these words were then replayed to the six lis-
teners. On the second presentation the five words were
correctly identified, thus making it possible to say that

the recorded words were 100% intelligible.

Controlling duration of original stimuli.--It was
important to exercise some degree of control over the dura-
tion Of the words as they were uttered by the Speaker in
the original recording of the stimulus material. It was
felt the durations of the words in the original recording
should conform to the expected durations of the words in
a connected speech context. In the absence of any rele-
vant previous research, this control was sought in the
following manner.

A random sample of one-syllable, two-syllable,
three-syllable, four-syllable, and five-syllable words
was drawn from the 300 words Of the PPVT. The words were
drawn so that there were ten words in the first three cate-
gories (one, two and three syllable words) and five words
in the latter two categories (four and five syllable words).
These forty words were then imbedded in syntactically cor-
rect paragraph contexts. The Speaker rehearsed reading

the several paragraphs until the rate of oral reading

-6l—

approximated the rate Of 160 words per minute. His oral
reading of these paragraphs, at this rate, was then recorded.
This rate Of 160 words per minute was an overall
rate which includes pauses between sentences. This seemed
to be an appropriate rate to use after examining the analysis
of measures of speaking and oral reading rate presented
by Johnson, Barley and Spriestersbach.l
The duration of each Of these forty words was deter-
mined from spectrograms Obtained for each of the words.
In converting the linear measurements from the spectrograms
to durations in milliseconds, one millimeter was given the
value of 7.874 msec. The mean duration in milliseconds
and the standard deviation was determined for each group
of words, the one, two, three, four and five syllable word
groups.
Using the Obtained sample means (M) from each of
the several groups Of words, an estimated range Of word
durations was determined, such that 95% Of the words in
that group would be expected to fall within this range.
Based on the standard deviation (S.d.) of each sample,

the formula used to derive the interval was

9 [M - 1.96 (S.d.) 3x 3M + 1.96 (S.d.)] = .95

 

lWendell Johnson, Frederic L. Barley, and Duane C.
Spriestersbach, Diagnostic Methods in Speech Pathology,
(New York: Harper and Row, Publishers, 1963), pp. 202-04.

-62-

Assuming the durations of words within each of these groups
is nearly normally distributed, it seemed reasonable to

say that additional Observations (x) of word durations,
which fell within the interval for the word's particular
group, were quite probably of a duration to be found in
connected speech delivered at a rate of 160 words per minute.
A word having a duration which fell outside the interval,

for its appropriate group (one syllable, two syllable, etc.),
would be regarded as an extremely unlikely event. The
results Of this analysis of the forty words from context

and the intervals computed for each group Of words is given
in Appendix C. Treating the Obtained intervals as accep-
tance intervals, it was possible to evaluate the recorded
single word utterances and the words of the serial retention
span stimuli in terms Of the extent to which their durations
conformed to the durations one would expect in connected
Speech.

Rather than evaluate all of the recorded words, a
random sample of words was drawn from both sets of recorded
stimulus materials: thirty words from the 300 single word
stimuli and six words from the thirty-two serial retention
span stimuli. An g_priori decision rule was made. This
rule was as follows: if more than 90% of the word dura-
tions in the sample (thirty-six words) were within the
acceptance intervals of their syllable groups, the entire

set of recorded stimuli (single words and serial retention

-53-

items) would be accepted as having durations that would
quite likely be found in connected Speech uttered at the
rate of 160 words per minute. If 10% or more of the words
in the sample (four or more words) had durations exceeding
the limits of the appropriate interval, the entire set of
words was to have been examined spectrographically, in
terms Of their durations. Recorded stimuli having dura-
tions falling outside the established acceptance intervals
were to be re-recorded and made to conform to the derived
intervals.

The durational measures made on the randomly drawn
sample of thirty-six words are given in Appendix C. It
will be observed that only two of the thirty-six words had
durations falling outside the acceptance interval for their
class. The words were: transportation and illumination.
They exceed the upper limits of their respective intervals
by 261.58 msec. and 65.18 msec.

Based upon the p priori decision rule, the dura-
tions of all the recorded words were accepted as being
within the expected durational limits for words in connected
Speech uttered at 160 words per minute. In the sample Of
thirty-six words, over 91% of the words had durations fall-

ing within their appropriate acceptance interval.

Calibration check tone.--TO have a means Of mea-

 

suring the precision of the temporal expansions to be made

-64-

of the stimulus words, a puretone was recorded into the
several lists of words. It was felt that a puretone pro-
vided a well defined unit upon which to base these measure-
ments. Its onset and termination were better defined than
the beginning and ending of a word. Comparing expanded
durations of the puretone with the original duration of
the tone provided a measure of the durational changes pro-
duced in the recorded words by the expansion process.

A 1000 Hz tone was placed at the beginning and
end of each of the four lists of seventy-five words, lists

Al, A2, B and B2. The tone was also placed at the

1.
beginning of the eight stimuli associated with each Of the
duration conditions for the serial retention span task.
The tone had a duration of approximately two seconds.

The outcome of the calibration check is reported

in the discussion which follows regarding the expansion

of the recorded materials.

Production of increased duration stimuli.--The
increased durations of the auditory stimuli - single words
and word series - were produced by an electromechanical
device, sometimes referred to as a speech compresser-
expander. From the original tape recordings of these
stimuli, this electromechanical device sampled the stimuli
with a degree Of temporal expansion that approximated the

desired level of durational increase. The product of this

-65-

processing was reproduced on a second tape at 7.5 inches
per second.

As has been previously explained, given the normal
duration of the signal or message, the device was able to
increase the duration of the signal by a factor Of from
1.00 to 2.00. In other words, if the signal at the initial
recorded duration (normal) was expanded by a factor of 1.50,
it had a duration 50% longer than its initial duration.

The signal was now 150% of its original time or duration.

The fixed levels Of increased duration selected
for this investigation were 125%, 150% and 175%. Adding
the normal duration (duration as initially recorded or the
100% level) to the three increased durations provided four
levels of the experimental treatment. These levels were
selected as representative levels within the expansion
capabilities of the electromechanical device.

As previously mentioned, a calibration check tone
was used to evaluate the precision of the expansion process.
A power level recorder was used to obtain a read-out of
the duration of the two second tone (1000 Hz) at the normal
duration and at the three expanded durations. Appendix D
contains the measurements Obtained from this calibration
check.

It will be Observed in Appendix D that each of the
B

four listS--A l and B2--were expanded to the three

1! A20

nominal duration levels. The calibration check tone appeared

-66-

at the beginning and end of each list. Hence, it was pos-
sible to examine the precision of the processing of each
list and the constancy of the processing throughout each
list.

The measurement of the reference tone (two seconds)
was 61 mm., Obtained from the read-out of the level recorder.
The measured durations (in milliseconds) for the expanded
versions of the 1000 Hz tone were compared to this 61 mm.
reference. The ratio Of the expanded duration to the refer-
ence duration was expressed as a percentage Of the reference
tone. The measurements and percentages are reported in
Appendix D.

All obtained expanSions were within five percent
Of the nominal levels of durational increase specified.

On the basis of an p_priori decision, and using the obtained
measurements of the calibration check tone, the recorded
word stimuli were accepted as being increased in duration

to each of the three nominal levels. Before processing,

it was decided that deviations of up to five percent from
the specified levels of expansion would be accepted under
that nominal level.

Each of the four lists--Al, A2, 31' and B2--was
now available at each of the four durations of 100%, 125%,
150% and 175%. 'Within each version of the four lists,
groups of twenty-five words, representing each Of the three

Hearing Vocabulary Levels were re-recorded, at 3.75 cps

-67-

on a separate tape. Separately storing each group of twenty-
five words at each of the durational levels enabled the
type of random presentation procedure to be described in a

following section.

Selection and Grouping of Subjects

This section describes the manner in which subjects
were identified and selected for participation in the study,
as well as the procedure followed in separating the twenty-
four subjects into two groups. The subjects were Obtained
from five treatment centers in Michigan: Constance Brown
Hearing and Speech Center, Kalamazoo; Veterans Administration
HOSpital, Battle Creek; American Legion HOSpital, Battle
Creek; Rehabilitation Medical Center, Lansing; and Speech

Clinic-University of Michigan, Ann Arbor.

DiagnOSiS.--Each of the subjects in this study had

 

been diagnosed as having some degree of language deficit
consequent to a cortical insult of some nature. For all
subjects, the aphasia was exhibited at least ninety days
following the cortical insult. Delaying the diagnosis until
ninety days after the cortical insult generally avoids the
period of greatest physiological instability and avoids
transient language disturbances that are frequently asso-

ciated with the immediate post cerebral insult period.

-68-

Schuell, Jenkins, and Jimenez-Pabonl have indicated the
consensus is that most patients reach this stability in
about ninety days.

The diagnoses of aphasia were made by several pro-
fessional Speech pathologists. Several examiners were
involved as a consequence of having Obtained the subjects
from several treatment centers. All of the examiners held
at least a master's degree in Speech pathology and had three
or more years of professional experience working with aphasic
adults. The speech pathologists based their diagnoses,
at least in part, on generally used diagnostic tests such
as The Minnesota Test for Differential Diagnosis of Aphasia
(MTDDA)2 and The Language Modalities Test for Aphasia.3

All Of the subjects had been accepted for treatment
in language rehabilitation. This fact is cited as indicat-
ing a degree Of physiological and psychological stability
sufficient to coping with tasks such as those undertaken

in this investigation.

 

lSchuell, Jenkins, and Jimenez-Pabon, Aphasia in
Adults, p. 160.

2Hildred Schuell, Minnesota Test for Differential
Diagnosis ongphasia: Administrative Manual (Minneapolis:
University of Minnesota Press, 1965).

3Joseph M. Wepman and Lyle Y. Jones, The Language
Modalities Test for Aphasia (Chicago: Education - Industry
Service, 1961).

—69-

Auditory and visual screening.--It was desired that
the auditory sensitivity and visual processes of all sub-
jects be reasonably intact. Primary sensory deficits in
these modalities could affect the subjects' performances
on the experimental tasks in ways quite independent from
the language deficit.

Information was obtained on the auditory sensitivity
of all subjects. This information was taken from recent
threshold tests or from a screening assessment made at the
time the subject was considered for participation in the
experiment. The pure tone average, for 500, 1000, and
2000 Hz, by air conduction was no greater than 25 dB (ISO,
1964 Standard) for the better ear of each person aCcepted
as a subject.

In terms of visual function, the medical and clin-
ical records of all potential subjects were examined for
indications Of visual field problems or uncorrected visual
acuity problems which would make the use of the response
plates (in experimental tasks) difficult or impossible.
Failing to find such evidence, the performance of each
person on the~first two subtests of the Auditory Disturbance
section of the MTDDA was scrutinized for suggestions of
interfering visual disturbance. These subtests employ
pictured objects in the response plates, as did the tasks
Of the experiment. All of the persons accepted as subjects

made correct responses on 74% or more Of the forty-two

_70-

items on these two subtests. In fact, the median percentage
correct for all subjects on these two tasks was 99%. For
subjects accustomed to wearing glasses, it was determined

if the glasses enabled him to View the reSponse plates
better.

On the basis of these outcomes, it seemed reason-
able to assume that all subjects were free from primary
auditory sensitivity problems and primary visual problems
which would impair their performance on the experimental

tasks.

Grouping subjects.--After determining that the

 

adult aphasics had met the previously mentioned criteria,
five additional subtests of the Auditory Disturbance
Section Of the MTDDA were administered to them.

These five subtests plus the two previously men-
tioned subtests constituted the diagnostic tests Of this
section and are intended to assess impairment in auditory
discrimination, reduction of auditory comprehension of
vocabulary, and reduction of auditory retention Span.l
The two nondiagnostic subtests omitted from this section

of the MTDDA were the test of understanding sentences and

the test of understanding a paragraph. The subtests were

lHildred Schuell, Differential Diagnosis Of Aphasia
ggith the Minnesota Test (Minneapolis: University of Minne-
sota Press, 1965), pp. 26-36.

_71-

administered as directed in the manual,1 with one slight
modification. The reSponse plates for three subtests were
presented to the subjects by means of the Slide projector
to be described. This was done to condition the subject
for the tasks of the experiment. This modification was
made for the following subtests: recognizing common words;
recognizing letters; and identifying items named serially.
NO modification was made in the administration of the other
four subtests concerned with discriminating between paired
words, following directions, repeating digits, and repeating
sentences.

An error score was determined for each subject on
the basis of the seven subtests, with the possible range
of scores being zero to ninety-six. This score was regarded
as a global measure of auditory language disturbance.

Auditory language disturbance was a factor Of inter-
est in analyzing the outcomes of this study, as indicated
by the hypotheses stated earlier. To control and examine
the influence of this variable on the performance of the
experimental tasks by the subjects, the sample of twenty-
four subjects was separated into two groups on the basis
of these error scores.

No attempt was made to control the range of audi-

tory language disturbance scores by selectively accepting

 

lHildred Schuell, Minnesota Test, pp. 5-8.

 

-72_

subjects. The division of the total sample was made after
all the subjects had been tested.

Subjects with error scores in the lower 50% of
the distribution of scores were designated as belonging to
the group displaying low auditory language disturbance.
Those subjects with error scores in the upper 50% of the
distribution were designated as belonging to the high audi-
tory language disturbance group. For the twenty-four sub-
jects, the range of error scores was from four to sixty-
five, with the median value being 16.5.

As indicated in Table 2, the distribution of error
scores for the low auditory language disturbance group
differed from the distribution of scores for the high audi-
tory language disturbance group. The statistical signifi-

cance of this difference was tested by the Mann-Whitney

TABLE 2. Comparison of Auditory
Language Disturbance (ALD)
Error Scores of Two Subject

 

Groups.
M
Mann-
Group Median Range Whitney
U
Low-ALD 10.5 4-16
0.00*

High—ALD 32.0 17-65

— I
- t

*Probability (one-tail) < 0.00

-73-

U Test,1 as the distributions of scores could not be con-
sidered normal. The obtained U was 0. This U had a prob-
ability Of approximately .00 in testing the hypothesis of
no difference between groups against the directional hypoth-
esis that the error scores of the high auditory language
disturbance group were in fact higher than the error scores
of the low group. Hence, the null hypothesis was rejected
in favor of the alternative hypothesis which stated the
groups were different in terms of auditory language distur-
bance.

Other characteristics of the two groups that were
examined and tested are given in Table 3. It would have
been desirable to have found the two groups similar in
terms of these characteristics. The characteristics were
quantified in the following manner: age in years to nearest
birthday; education in years; and time since onset of aphasia
in months. Because the distributions of these measures
could not be assumed to be normal, the Mann-Whitney U Test2
was used to test whether the two groups had been drawn from
the same population. Since differences in either direction
between the two groups was of interest, a two-tailed test

for statistical significance was made, setting the confidence

level at .95.

 

lSidney Seigel, Nonparametric Statistics for the
Behavioral Sciences (New YOrk: McGraw-Hill Book Co., 1956),
pp. 111-116.

21bid.

-74-

TABLE 3. Comparison Of Age, Educational Level, and
Elapsed Time since Onset of Aphasia of Two
Subject Groups (Low-ALD and High-ALD).

 

 

Low-ALD High-ALD Mann-
Characteristic Group Group Whitney U
Age (years)
Median 41 56 28.0*
Range 21-60 37-72
Education (years)
Median 13 12 42.0
Range 12-24 7-17
Onset (months)
Median 14 23.5 53.5
Range 3-33 2-50

 

*Probability (two-tailed) < .05

The hypothesis of no difference between groups was
rejected for the age characteristic, with °< = .05. Interest
in the age of the subjects in the two groups stems from
the question of the influence of aging on intelligence,
more Specifically on vocabulary ability. As Wechslerl has
indicated, vocabulary is less vulnerable to the influences
associated with aging than the other capacities assessed
by the Wechsler Adult Intelligence Scale. Wechsler indicated
little change in the vocabulary scores over the sixteen

year period represented in the median age difference between

 

lDavid Wechsler, The Measurement and Appraisal of
Adult Intelligence (Baltimore: Williams & Wilkins Co.,
1958). p. 85.

-75-

the two groups in this investigation. The vocabulary test
of the WAIS is more difficult than the PPVT, in that, on
the WAIS the subject must define the word verbally. While
the desirability Of having greater similarity in the ages
of the two groups cannot be denied, the difference was not
believed to be critical in terms of the experimental tasks
of this investigation.

Educational level showed no statistically signifi-
cant difference between groups, and only one year Of real
difference. The median educational level for the low audi-
tory disturbance group was one year beyond high school,
while the median level for the other group was high school
graduation. This Similarity between groups was a most
important factor. Lacking information concerning the pre-
morbid intelligence of the subjects, it indicated some
degree of equality between groups in terms of premorbid
academic achievement. From this similarity, some degree
of equality in premorbid vocabulary development was inferred.1

The time elapsed since onset Of the aphasia was
the other group characteristic reported in Table III. On
the basis of the statistical test, it was assumed the
Observed difference was not significant. This factor was
seen as reflecting on the degree of motivation subjects

had toward participation in the experimental tasks.

 

lIbid., p. 84.

-75-
Subjects with long standing aphasic conditions might be
less motivated to participate in the experimental tasks
than an equally impaired subject who had experienced the
aphasia more recently.

Presentation of Stimuli to Subjects

Apparatus.--The tapes bearing the recorded audi-

 

tory stimuli were played on a tape recorder that had a
built-in photo-synchronizer circuit, enabling one to syn-
chronize the audio output of the tape recorder with visual
stimuli presented on a slide projector. The synchronizing
signal from the recorder was delivered to a rear-projection
type Of slide projection device. The image was projected
on a glass screen at the front of the unit. Since the
projector itself was behind the screen, touching the screen
did not disturb the projected image on the screen. The
projector was equipped to handle thirty-five millimeter
Slides.

The projected image from the slides of the Peabody
Test response plates was practically the same size as that
of the response plates themselves (6.25 inches x 8.75 inches
vs. 6.25 inches x 8.50 inches). The image of the specially
constructed response plates for the serial retention span
task measured 9.25 inches x 6 inches.

The inaudible signal, which activated the slide

-77_

changing mechanism of the projector, was placed ahead of
the stimuli on the tapes. The signal preceded the stimuli
by a period of time sufficient to allow the slide changing
mechanism to get the slide into position before its asso-
ciated word was played. The slide was in position anywhere
from two to three seconds before the auditory stimulus was
played.

The auditory stimuli were played through an amplifier-
speaker rather than through the internal Speaker of the
tape recorder. The amplifier-speaker was used to provide
an auditory stimulus having better fidelity than would have
been Obtained from the internal Speaker of the tape recorder.
The amplifier-speaker was placed directly above the pro-
jection screen. Hence, the auditory stimuli and the response
plate were aligned in the same vertical plane. The output
of the speaker was adjusted to a comfortable loudness level,
as determined by questioning the subject about the loudness
of the signal. In general the level of the auditory stimuli
was on the order of 60 to 70 dB SPL (re: .0002 dynes cmz)

at the ear of the subject.

Test conditions.--The conditions under which the
testing was done varied as a result of having secured sub-
jects from several different treatment centers. Each sub-
ject was tested in the room (or a room quite similar to the

room) in which he had previously received therapy. The

-78-

rooms were reasonably free of visual and auditory distrac-
tions and were physically comfortable in terms Of temperature
and ventilation. Also, these conditions were essentially
constant for each subject for each Of the three test ses-
sions required for the completion of the tasks in the
experiment.

Each of the three test sessions lasted from thirty
to sixty minutes. The first session approached sixty minutes
for all subjects since the Auditory Disturbance Section
of the Minnesota Test for Differential Diagnosis of Aphasia
was administered during this session. Testing of a sub-
ject was completed within a three day period, with the
elapsed time between test sessions being twenty-four hours.

The subjects were seated in front of the projection
screen and were assisted in positioning themselves at a
comfortable visual distance and angle in relation to the
screen. For no subject did this position preclude his
being able to touch easily and comfortably the screen of
the projector. If the subject was accustomed to wearing
glasses, it was determined whether or not his glasses enabled
him better to view the visual image on the screen of the
projector.

The experimenter was seated to the left of the
subject. From this position, the experimenter was able
to observe the subjects' pointing responses and still con-
trol the tape recorder and record the subjects' responses

to the auditory stimuli.

-79-

Randomization of stimuli.--The two sets Of stimuli--
single words and serial items--were randomized independently,
with the randomization of the stimuli for each of the tasks
being different for each subject. The randomization pro-
cedure for the word recognition task is described first,
followed by a description of the procedures employed in
randomizing the serial retention items.

The twenty-four randomizations, used in presenting
the stimuli for the word recognition task to the subjects,
were develOped by first assigning each Of the four lists
of seventy-five words, lists Al’ A2, B1' and B2, to one of
the four stimulus duration conditions. The assignment was
made so as to employ all twenty-four permutations Of the
four lists. In each of these twenty-four arrangements,
the first list was assigned to the 100% stimulus duration;
the second list to the 125%.duration; the third tO the 150%
duration; and the fourth list to the 175% stimulus duration.
Within each of these lists, the three Hearing Vocabulary
Levels were represented. At each level there were twenty-
five stimulus words. One such matrix is given in Table 4.
Each block of the matrix represents twenty-five stimulus
words.

In Appendix E is found the arrangement of lists for
each subject. For each of the twenty-four arrangements Of
the lists, the blocks of twenty-five words were indexed in

the same manner. Based on this system of indexing, the

-80-

TABLE 4.--Sample Arrangement of Lists
with Block Index Numbers.

Lists

 

 

H
H
no
u:
.p

 

 

 

 

 

 

 

 

Hearing
Vocabulary Level

II 5 6 7 8
III 9 10 11 12
1.00 1.25 1.50 1.75

Stimulus Duration Conditions

 

order of presentation for the twelve groups of twenty-five
words was randomized for each subject. There were twenty-
four such randomly selected block sequences.

The serial retention span stimuli were randomized
in a manner similar to the randomization of blocks for the
word recognition stimuli. For the serial retention stimuli,
each of the four stimulus durations were coupled with four
levels of stimulus complexity (two, three, four, and five
word series). This resulted in a sixteen block matrix.
The auditory stimuli and their associated response plates
were not varied across durations for this task, as they
were for the word recognition task. Each block contained
two serial retention items, the number of words in the

auditory stimulus being determined by the complexity level

-81-

at which the items were placed. The blocks were indexed
(in the manner indicated for the word recognition stimuli)
with the numbers one through sixteen. One randomization
of the sixteen blocks was generated for each subject, for
a total of twenty-four block sequences.

One randomization for the word recognition task
and for the serial retention Span task were paired, and
assigned a subject number. The subject was exposed to both
of the experimental tasks at each of the three test ses-
sions: four sets of single words with each set containing
twenty-five words; and five or six pairs of serial items.
The order in which the tasks were presented to a subject
varied systematically from test session to test session.
If word recognition was the first task of session one, it
was the second task Of session two, and the first task of
session three. The word recognition task was the first
task Of session one for half of the subjects, while the
other half Of the subjects began session one with the serial
retention span task. This arrangement of tasks within
sessions Offset possible order effects of one task upon

the other.

Instructions to supjects.-—Following the administra-

 

tion of the Auditory disturbance section of the MTDDA,
either the word recognition task or the serial retention
Span task was administered to the subjects. At the first

test session, the following instructions were given.

-82—

For the word recognition task, the instructions
were as follows:

I'm going to Show you some pictures like these (point-
ing to screen with PPVT Example A diSplayed). You

see there are four pictures (pointing to each picture).
You will hear a word, then I want you to point to the
picture which best tells the meaning of the word.
Let's try one. Point to the picture which best tells
the meaning of crib. (Get correct response before
proceeding.) Fine. (Show PPVT Example B on screen.)
Show me IEB: (Get correct response.) Good. (Show
PPVT Example C on Screen.) Butterfly. Good. Now
each time you hear a word point to the picture which
best tells the meaning of the word. You may not be
sure of the meaning of some words. But look carefully
at all the pictures anyway and point to the one you
think is right. DO you have any questions? Here's
the first one.

 

Instructions for the serial retention Span task
were as follows:

You are going to see pictures like these (pointing to
screen on which Sample eight-picture plate is dis-
played). You see there are eight pictures (pointing
to each picture). YOu will hear several words. There
may be two words, or three words, or four words, or
five words. The words will name things pictured here.
I want you to point to the pictures named. Let's try
one. Show me p£p_and knife. (Get correct response
emphasizing order.) Good. Show me duck and‘ppn.

(Get correct response emphasizing order.) Fine. Show
me duck,‘pyg, and broom. Now each time you hear the
words point to the pictures named in the same way you
heard the words. Sometimes you may not be sure of all
the words. But look carefully at all the pictures
anyway and point to the pictures you think are right.
Do you understand? Do you have any questions? Here's
the first one.

AS it was necessary, subjects were encouraged and
directed to make the pointing response even though they had
to make "a wild guess." A response was obtained from each
subject to every stimulus. The subject was given as much

time as he took to make his response. The subjects were

-83-

reminded of the nature of the tasks at the beginning of the

second and third sessions.

Scoring responses.--The experimenter recorded the

 

subject's responses by writing down the number or numbers
of the picture (or pictures) to which the subject had
pointed.
The response record sheet for the word recognition
task contained the ordered set of stimulus words as well
as the number of the correct picture for each stimulus word.
The subject's score was the number of correct responses.
The response record sheets for the serial retention
span task provided a space in which the experimenter wrote
the sequence of picture numbers pointed to by the subject.
A response was scored as correct if, and only if, the sub-
ject pointed to all the pictures named in the same order
in which they were named. The subject's score on this
task was weighted, depending upon the number of words in
the stimulus. The scoring procedure for each correct
response was as follows: two points for a two word series;
three points for a three word series; four points for a

four word series; and five points for a five word series.

CHAPTER IV

RESULTS AND DISCUSSION

This chapter presents the results obtained from
the procedures described in Chapter III. Twenty-four
aphasic adults participated in a word recognition task
and a retention span task, in which the stimuli were pre-
sented auditorily, under four experimental conditions of
stimulus duration. On the word recognition task, the
stimuli were associated with one of three levels of hearing
vocabulary, as an index of the meaningfulness of the stim-
uli. On both tasks, subjects were assigned to either the
low auditory language disturbance group (hereafter referred
to as LALD Group) or the high auditory language disturbance
group (hereafter referred to as HALD Group) on the basis
of their performance on the Auditory Disturbance Section
of the Minnesota Test for the Differential Diagnosis of
Aphasia.1

Subject performance was measured by the number of
correct picture identifications made in response to the

auditory presentation of either a single word (recognition

 

lHildred Schuell, Minnesota Test for Differential
Diagnosis of Aphasia: Administrative Manual (Minneapolis:
University of Minnesota, 1965), pp. 5-8.

-84-

-85-

task) or a series of from two to five words (retention
task). The number of correct identifications per condition
constituted the subject's score and was considered the de-
pendent variable.

For each Of the tasks, there were independent var-
iables that were of interest in terms Of assessing the
design Of the experiment. For the word recognition task,
these independent variables were subject groups and levels
of hearing vocabulary. The independent (control) variable
for the retention task was subject groups. These variables
are considered first in the following report of results.

Of major interest, of course, was the influence of
stimulus durations, the third independent variable. In
addition to examining the independent influence of this
variable on performance in both tasks, the interaction of
stimulus durations with the other independent variables

was of equal interest. .

Assumptions underlying statistical tests.--Before
proceeding with the several statistical analyses, it was
necessary to consider the extent to which the assumptions
underlying the use of the statistics had been met. Analysis
of variance and a nonparametric correlation were the sta-
tistics employed.

The assumptions underlying the use of the analyses

of variance were (1) the performance measure was normally

-86-

distributed in the aphasic adult pOpulation, (2) the popula-
tions had variances which were homogeneous over the Observa-
tions to be made, and (3) the performance made by each
subject was independent of the performance of every other
subject.

Available studies of the performance of aphasic
adults on tasks such as those employed in the present study
indicated that the distribution of scores was moderately
to markedly skewed.l The degree of skewedness appeared to
depend largely on the range of aphasic disturbance repre-
sented in the particular sample of the general pOpulation.
Allowing that the characteristics may be moderately skewed
in the aphasic adult population, it was recognized that the
F-test is quite insensitive to the shape of the distribu-
tion. Even when a good approximation to a normal distribu-
tion fails, Lindquist has stated

. . . the F-distribution seems so insensitive to the
form of the distribution of criterion measures that

it hardly seems worthwhile to apply any statistical
test to the data to detect non-normality, even though
such tests are available. Unless the departure from
normality is so extreme that it can easily be detected
by mere inspection Of the data, the departure from

normality will probably have no appreciable effect
on the validity of the F-test. . .2

 

1Peter S. Perry, Personal communication, November,
1967. Hildred Schuell, Differential Diagnosis of Aphasia
with the Minnesota Test (Minneapolis: university of Minn-
esota Press, 1965), pp. 27-29, 31.

2E. F. Lindquist, Design and Analysis Of Experiments
in Psychology and Education (Boston: Houghton Mifflin
Company, 1953), p. 87.

 

-87-

The above is conditional upon the distribution being homo-
geneous in both form and variance for the treatment pOpula-
tions.

Regarding the assumption Of equal variances across
the treatment groups, it is generally recognized that so
long as the sample sizes are equal for all conditions of
the analysis, violations Of the homogeneity assumptions
have little effect upon the F-distribution.1 Sample sizes
for all conditions of both analyses were equal.

It seems, therefore, the assumption of normal dis-
tribution and homogeneity of variance can be satisfied for
this study.

The third major assumption pertains tO the inde-
pendence of the various observations. Only on the auditory
language disturbance factor was independence true. The
factors of stimulus durations and hearing vocabulary levels
were not independent. Each of the twenty-four subjects
was Observed under all twelve combinations of stimulus
durations and hearing vocabulary levels. It was for this
reason the three-factor analysis of variance for repeated
measures on two factors was selected as the most appropriate
statistic to test the null hypotheses derived from the

questions concerning word recognition. A two-factor analysis

 

lWilliam L. Hays, Statistics for Psychologists,
(New York: Holt, Rinehart and Winston, 1964), p. 408.
Lindquist, Design and Analysis, pp. 33-34.

 

 

-88-

with repeated measures on one factor was selected for the
retention task.

In addition to the assumptions specified above,
Winerl has listed two additional assumptions that are
necessary to repeated measures designs. They are as fol-
lows: the order in which the repeated measures are made
must be randomized independently for each subject, and the
p subjects in a group must be a random sample from a Spe-
cified population. The randomization procedures described
in the previous chapter clearly indicate that the first
of these Special assumptions was satisfied. The second
assumption was met in a satisfactory manner by drawing
subjects from several clinics in Michigan.

The Spearman rank correlation coefficient was used
to assess the degree of association between a subject's
performance on the word recognition task and his perform-
ance on the retention Span task. The use Of this statistic
requires that both variables me measured in at least an
ordinal scale so that the individuals can be ranked in two
ordered series.2 The ordinal level of measurement was

achieved on both variables of interest in this comparison:

 

1B. J. Winer, Statistical Principles in Experimental
Design (New York: McGraw-Hill Book Company, 1962), pp.
298-301.

2Sidney Siegel, Nonparametric Statistics for the
Behavioral Sciences (New York: McGraw-Hill Company, Inc.,
1956). P. 202.

-89-

word recognition scores and weighted retention Span scores.
NO further assumptions need be made in the use of this
statistic.

The presentation of the results of the statistical
treatments is organized on the basis of the two tasks in
which the subjects participated, followed by presentation
of the inter-task comparisons. A discussion of the Obtained

results concludes this chapter.

Results from Word Recognition Task

 

The null hypotheses generated from the three ques-
tions concerning the performance of the subjects on the
word recognition task were tested by means of three factor
analysis of variance for repeated measures.1 The repeated
measures entered the analysis as the factors of stimulus
durations (Durations) and hearing vocabulary levels (HVL),
the experimental variable and a control variable, respec-
tively. The second control variable was subject groups
(Groups): the group of subjects with low auditory language
disturbance (LALD); and the group with high auditory lan-
guage disturbance (HALD). The three factors of the analysis
were arranged as a three dimensional table of a 4 x 3 x 2

design. The four columns represented the four Durations,

 

lWiner, Statistical Principles in Experimental
Design, pp. 319-35.

-90-

while the three rows represented the three HVL. The third
dimension was conceptualized as two slices of the 4 x 3
matrix (Durations x HVL), the slices representing the two
subject groups--LALD Group and the HALD Group. The depen-
dent variable (word recognition score) was entered, for
each subject, in the appropriate cell of the twelve cell
matrix associated with his subject Group.

This analysis of variance was conducted on a Con-
trol Data Corporation 3600 Digital Computer (as were all

1 The F-ratio was

statistical computations in this study).
used in testing the statistical significance of the variance
attributable to the three main factors and the several
interactions. The .05 level of confidence was established.
The summary of this analysis is given in Table 5.

Before considering the influence of the experimen-
tal variable, an examination was made of the control var-
iables, auditory language disturbance (Groups) and hearing
vocabulary levels (HVL).

The summary table shows the Obtained F-statistic
for both Groups and HVL is statistically significant.

For the Groups, the probability of the F-statistic was
approximately 0.002 while for the HVL factor, the prob-

ability was less than 0.0005.2 These outcomes provided

 

1Agricultural Experiment Station, "STAT Series
Description NO. 14: Analysis of Variance with Equal Fre-
quency in Each Cell," (East Lansing, Michigan State Univer-
sity, 1968).

21bid.

-91-

Hm>ma mo. map Ocommn HGMUNMHGmHm#

 

 

nmm www.mmmm- Hmpop

mmo.m mm- omm.om¢ omm + OONN u N

#nem.m omm.HH o «mm.as O x m x N

ooh. moo.~ o mmm.m- m x a

Hmm.6m mm Nev.mmam so + O u N

#Nmm.-H w-o.¢¢- H s-o.S¢- -o- masons

Hm-.v mo mmo.N-N mm + Oom u N

#nmo.m vom.NH m vam.nm O x m

o-m.o mm-.m m nem.-H Am- co-pmusa

mm¢.mm 66 om~.-NHH ON + OON n N

emo.m omw.mm m osm.wo- O x 4

#nmm.mo- N-A.wom¢ m ¢~6.moom AN- .>mq .pmoo> .pmmm
oaumflumum mumswm mmumsvm

.m Gmmz .mQ MO saw OUGMHHM> MO OOH-DOW

coauflcmoomm OHOB on» so mGOHumusn mSHSEflum Ugm

.xmma

.mmsouw pomnnsm .mH0>0A mum
Iasnmoo> mafiummm mo muommmm mnu mcflummEoo mosmflum> mo mama-NS4 m0 mumEESmll.m mqmde

_92-

general support for the design Of the experiment. Both
factors were expected to exert an independent influence on
performance of the task, and it was desired to control for
their effect in examining the influence of stimulus dura-
tions on word recognition.

While the F-statistic indicated the two groups of
subjects performed differently, it was noted that the LALD
Group made the higher scores. The mean difference, averaged
across all twelve duration x hearing vocabulary conditions,
was 3.99. In other words, the group with better auditory
language function made, on the average, about four more
correct responses per treatment condition than did the
group of subjects with poorer auditory language function.

Since the F-statistic did not reveal where the
significant differences in hearing vocabulary levels lay,
a critical difference test1 was employed to determine if,
in fact, the expected differences in hearing vocabulary
levels did exist. On the basis of the rationale for struc-
turing these levels, the scores on Level I should have been
significantly greater than those at Level II, and Level
II scores should have been greater than those Of Level
III.

In Table 6, the means at each of the hearing vocab-

ulary levels for the two Subject Groups are given, as well

 

lLindquist, Design and Analysis, p. 93.

-93_

TABLE 6.--Hearing Vocabulary Level Means, and
Differences between Means for Each
Auditory Language Disturbance Group

'1
I

 

 

HALD LALD
Group Group
Means
Level I 21.292 24.542
Level II 16.146 21.833
Level III 8.354 11.375
Difference
between Means
Levels I-II 5.146* 2.709
Levels II-III 7.792* 10.458*
Levels I-III 12.938* 13.375*

 

*critical difference, .05 level

as the differences between means within groups. It was
determined that a difference of more than 4.254 would be
significant at the .05 level for differences within the
subject groups. As seen in Table 6, one of the differences
of interest was not statistically significant: the Level

I - Level II differences for the LALD Group. With the ex-
ception of this difference, all other differences were
statistically significant. It may also be noted, in the
analysis of variance summary table, that the interaction

of Groups x HVL was not significant, statistically, reflect-
ing the strength of the independent influence of these two
factors on the performance of the word recognition task.

The adequacy of the experimental design, with regard

-94-

to the control exercised over auditory language disturbance,
appeared to be substantial. From this same vieWpOint, the
control of hearing vocabulary appeared to have failed at

only one point.

Results of experimental treatment.--Turning to the
results of major interest, the null hypotheses derived
from the first three questions stated in Chapter I (p. 3)
are stated and then evaluated in terms of the outcomes of
the statistical analysis.

The general question of the influence of stimulus
durations on word recognition lead to the following null
hypothesis:

There is no difference in the mean improve-

ment Of auditory recognition under several

conditions of stimulus duration on a task

involving auditory recognition of Single

words.

In the summary table (Table 5, page 91) the test of this
null hypothesis was found in the main effect for Duration.
This main effect was not statistically significant, making
it impossible to reject the null hypothesis of no differ-
ence from the variations produced in the duration of the
stimuli. On the basis of this experiment, increasing the
duration Of words did not appreciably affect the performance

of aphasic adults in recognizing single words presented

_95-

auditorily. It was noted that the influence of Duration
was tested here on the performance of all subjects, ignoring
potential differences that may exist between subjects mani-
festing different degrees of aphasic involvement, an inter-
action to be considered later.

The question of the interaction of stimulus dura-
tion and hearing vocabulary levels generated the following
null hypothesis:

The duration of the stimulus words will

not affect mean performance in recognizing

single words presented auditorily at several

levels of hearing vocabulary.

The Duration x HVL interaction provided the means for test—
ing this null hypothesis. The F-statistic resulted in a
ratio that was not significant statistically. This indicated
there was a systematic relationship between these two var-
iables. This null hypothesis was not rejected, on the basis
of this analysis.

Of the two remaining interactions, both were statis-
tically significant: Duration x Groups, and the triple
interaction of HVL x Durations x Groups. The two inter-
actions were of interest to the third question posed in this
study. A first approach to an understanding of these inter-

actions was a graphic approach.l

 

lWiner, Statistical Principles in Experimental
Design, pp. 178-84.

-95-

In utilizing this graphic approach, it seemed
helpful to examine first an interaction found to be not
significant statistically. The Groups x HVL interaction
was re-examined for this purpose. The plot of this inter-
action is given in Figure 1.

In inspecting such plots of data (in this case,
means are plotted), it was helpful to remember that inter-
actions not statistically significant reflect a high degree
of parallelism between related sets of observations. The
plot of the Group x HVL interaction certainly reflected
this characteristic. The plot of means for both groups
(LALD Group and HALD Group) were quite parallel to each
other and parallel to the plot of means for all subjects
combined (Total Group). The relationship was quite system-
atic across all levels of Hearing Vocabulary. As stated
previously, this outcome added strength to the experimental
design, attesting to the important independent influence
of these two variables.

To test the null hypothesis developed from the
third question posed in this investigation, the statistical
significance of the Groups x Duration interaction was con-
sidered. The null hypothesis was:

Under several conditions of stimulus

duration, there will be no difference in

the mean performance between subjects

identified as having a low degree of audi-

tory language disturbance and those

-97-

FIGURE l.--Graph of Group Means
and Total Group Mean

Number of Correct Responses - Word Recognition

at Three HVL
25::

24 --
23 .r
22 -.
21-- K

20 -- \

 

19" \ ..
18 u \
l7d- \
1.-. x
15,, \
l4-m \

13 +- \

12 u \
11 1F
10-.

9 “

8-p

 

 

 

/L
01’

61—— LALD Group

 

‘ Total Group

\‘ ~HALD Group

 

I II

III

Hearing Vocabulary Levels

-98-

subjects identified as having a high degree

of auditory language disturbance, on a task

of auditory word recognition.
The real significance of the Groups x Durations interaction
was examined in Figure 2, at the heavy lines labelled A,
B, and C. At line A, it was seen that the means, at each
of the four durations, for the LALD Group arranged themselves
in a manner approximating a straight line. The means of
the HALD Group (B), on the other hand, did not make as good
an approximation to a straight line. This departure of
the HALD Group Means from the configuration of the LALD
Group Means was sufficient to cause the Groups within Dura-
tions interaction to reach statistical significance. Further-
more, since the plot of means for the Total Group (C) re—
flected a relatively straight line function, it was in the
HALD Group (B) that departure from parallelism was to be
found.

It appeared that this departure was produced by
the observed difference between means at 125% and 175%
Stimulus Duration. When, in fact, the differences between
means within each of the two groups was tested, the differ-
ence between means at 125% and 175% in the HALD Group was
the only difference found to be statistically significant
at the 5% level of confidence. The test used was the

critical difference procedure suggested by Lindquist.1 The

 

lLindquist, Design and Analysis, pp. 164—66.

25

24

23

22

21

20

19

18

17

16

15

14

13

12

11

10

-99-

FIGURE 2.--Graph of Means for Hearing
Vocabulary Levels of Groups
at Four Durations.

 

 

\
\

@U‘J

II

100% 125% 150% 175%

Stimulus Durations

:sIeAeq

TE"I

l

dnoxo

:sIaAaq

duels 01v1-_.

dnoxs 01VH===

IEQOJ, 000 0

V

8

suoTQEJna qe sueew IeAaq KJeInqeooA BUIIEBH

-lOO-

differences between means within each of the two groups of

subjects are reported in Table 7.

TABLE 7.--Mean performance for Each Auditory Lan-
guage Disturbance Group at Four Stimulus
Durations, and Within Group Mean Differ—
ences between Duration Conditions

W

 

 

HALD LALD
Group Group

Duration Means
100% 15.139 19.472
125% 16.028 19.000
150% 15.389 19.111
175% 14.500 19.417

Difference

between Means
125%. 100% 0.889 -0.472
150% 100% 0.250 -0.361
175% 100% -O.639 -0.055
150% 125% -O.639 0.111
175% 125% -1.528* 0.417
175% 15G% -0.889 0.417

 

*critical difference,

.05 level.

The difference in means at 175%P125%.in the HALD
Group (the only statistically significant difference) was
a negative difference. Performance was better at the dura-
tion more closely approximating the "normal" duration of
100%, i.e., the 125% duration, than it was at the 175%
duration. No real significance was attached to this finding
since the performance at 175% duration was not critically

different from the performance at 100% duration. The

-101-

statistical significance of this interaction might lead
one to reject the null hypothesis under discussion. However,
the interpretation made of this interaction lead to the
decision to postpone judgment on this hypothesis.

This decision was influenced by an inspection of
the statistically significant triple interaction of HVL x
Duration x Groups. This interaction was vividly displayed
in Figure 2. While the influence of Durations was negli-
gible, the control variables of Groups and HVL appeared
to interact in a highly nonsystematic manner. At both the
125% and 150% Durations, it was possible to observe the
between group interaction at IIfI, and IIfII. While con-
tributing to the statistical significance of the triple
interaction, these observations were expected, considering
the basis upon which the subjects were grouped. That basis
was degree of auditory language disturbance.

Examining the within group differences by levels
of hearing vocabulary at the four durations, it was seen
that the LALD Group plot of means, at the three HVL, bore
a strong resemblance to the LALD Group Means (A). Such
was not the case when a similar examination was made within
the HALD Group. It was easy to see the dissimilarity in
the plot of means for the HALD Group at HVL II and HVL III.
The means at these hearing vocabulary levels not only dif-
fered from those at HVL I and HALD Group Means (B), but

HVL II differed in its configuration from HVL III.

-102-

It was undoubtedly in both the between groups
interaction and within the HALD Group itself that the sta—
tistical significance of the triple interaction was gener-
ated. The observed trends within the HALD Group at the
three HVL, while not of statistical significance, presented
an interesting pattern that is discussed in a later section
of this chapter. The interest in this pattern was the
basis for postponing a decision on the null hypothesis
concerning the influence of the durations on the two groups

of subjects.

Results from Retention Span Task

Two of the questions prompting this study concerned
the performance of aphasic adults on a retention Span task
in which the stimuli were presented auditorily at several
levels of complexity. The null hypotheses developed from
these questions were:

(1) Increasing the duration of the speech stimuli

will have no influence on the mean performance
of a retention span task in which the words

are presented auditorily and in series.

(2) There is no difference in the mean performance
of two groups of subjects, identified as having
different amounts of auditory language distur-

bance, at several levels of stimulus duration

-103-

on a task requiring the retention of words

presented auditorily and in series.

The performance of the subjects was analyzed by a
two factor analysis of variance, with repeated measures on
one factor.1 The repeated measure was on stimulus duration
(Duration). The second factor was Groups, defined as for
the word recognition task. The grouping of subjects was
the same as for the word recognition task. The two factors
were arranged in a two dimensional table of a 4 x 2 design.
The four columns represented stimulus durations, while the
two rows represented subject groups (Groups): as before,
LALD Group and HALD Group. The dependent variable, weighted
retention span score, was entered for each subject in the
appropriate cell of the eight cell matrix.

The F-ratio was used in testing the statistical
significance of the variance attributable to the two main
effects and the two way interaction. The .05 level of
confidence was established. The summary of this analysis
of variance is given in Table 8.

In this summary table, the only statistically sig-
nificant F-test obtained was that for Groups. The measured
difference in performance between the two groups was 13.61;

the mean for the LALD Group was 18.06, while the HALD Group

 

lWiner, Statistical Principles in Experimental
Design, pp. 302-09.

.Hm>mH mo. may onoamn unmoflmflcmum#

 

 

W mm mmv.oomn Hmuoa
1.
_ mmm.>n mm mum.vaha um + o n N
#mnm.om oop.a¢¢¢ H oms.H¢¢¢ Ame msouw
moa.ma we 5mm.omm 04 + 0mm n N
omw.o omm.m m mmo.oa m x 4
hm>.o mmo.mH m «Ha.mm Adv aofiumusn
oHumHumum mumsvm mmumsqm
m coo: ma mo 85m oUQMHHm> mo monsom

l
i

.Mmma smmm coaucmumm may no mGOHumusn msasfiflum Ucm mmsouw
uomflnsm wo muomwmm mnu mcflnmmeoo mocmflum> mo mammamcd mo mumsasmll.m mamas

-105-

had a mean of 4.45. This difference was in the expected
direction.

Since the main effect for Duration was not statis-
tically significant, there was no basis for rejecting the
first null hypothesis concerned with performance on the
retention Span task.

The second null hypothesis could not be rejected
since the Duration x Groups interaction was not statistically
significant.

Under the conditions of this experiment, it must
be reported that the duration of the stimulus did not have
any important influence on the performance of this reten-
tion span task by aphasic adults. Furthermore, the two
groups of subjects did not perform differently at any of

the four stimulus duration conditions.

Results of Inter-Task Comparison

The final question posed in this investigation
dealt with the relationship of subjects' performances on
the two tasks employed in the experiment. The question
led to the following null hypothesis:

There is no relationship between performance

on a word recognition task and on a retention

Span task in which the stimuli are presented

auditorily.

Since both analyses of variance had indicated that the

-lO6-

established grouping of subjects was a significant factor
in the performance of both tasks, the measure of associa-
tion between tasks was applied within groups. The Spearman
rank correlation was the measure of association used.1
The significance of the correlations, as they might reflect
on the aphasic adult population, was determined. If, in
fact, the two performances were correlated in the population,
it seemed reasonable to assume the correlation was positive.
This reasoning was based on the previously cited theoretical

2 The .05 level of confidence

notions of Jenkins and Schuell.
was established. The obtained correlations are presented
in Table 9. The only statistically significant correla—

tions were obtained in the group having the poorest auditory

language function, the HALD Group.

TABLE 9.--Spearman Rank Correlations of Performance on
Word Recognition Task and Retention Span Task
for the Two Subject Groups.

W

 

Stimulus Durations
Group_ 100% 125% 150% 175%
LALD -0.456 -0.l44 -0.287 -0.101
HALD 0.599* 0.670* 0.359 0.506*

 

*significant at .05 level (one-tail)

 

lSiegel, Nonparametric Statistics for Behavioral
Sciences, p. 202.

2Hildred Schuell and James J. Jenkins, "The Nature
of the Language Deficit in Aphasia," Psychological Review
LXVI (1959), pp. 50-51.

-107-

It was not possible to reject the null hypothesis
at any of the duration conditions for this LALD Group.
For the HALD Group, at the three durations, it was possible
to reject the hypothesis of no association in favor of a
hypothesis stating a positive relationship between perform-
ance on the word recognition task and the retention span
task. The durations were 100%, 125%, and 175% stimulus

duration.

Discussion

 

Within its limitations, this investigation has
demonstrated that increasing the duration of auditory
stimuli does not significantly alter the performance of
aphasic adults on two tasks which appeared to be highly
relevant to the language dysfunction in aphasia. The recog-
nition of words presented auditorily seemed to be related
to the reduction in available vocabulary, while the reten-
tion span task (for words presented auditorily in series)
seemed to be clearly related to the degree of impairment
in verbal retention span.

It was entirely possible the increases in stimulus
duration, employed in this study, were not great enough
to reach an effective level of improvement for the aphasic

subjects. This point deserved serious consideration. If

 

Ibid.

-108-

the clinical modifications made in the duration of auditory
language stimuli are, in and of themselves, an effective
means of facilitating language reception, then the levels
of durational increase may exceed the levels employed in
this study. However, this possibility was not fully clari-
fied by this study.

It should be noted that this study was unable to
provide information about the possible influence of the
"processing noise" resulting from the expansion process.
While the lack of statistical significance from the main
effect of Duration might lead one to believe the "noise"
had no negative influence (as well as no positive influence),
it was possible, however, that the "noise" masked some real
positive influence due to durational increase.

It was recognized that in both tasks the vocabulary
being sampled was more likely a lexical vocabulary, not a
linguistic vocabulary. It could be argued that, for this
reason, the tasks were assessing only the original voca—
bulary development of the subjects and not his "available
vocabulary." When the general familiarity or meaningfulness
of the words within the first two levels of hearing vocabu-
lary are examined, this argument seemed weak. The tasks,
as presented, may have been more difficult than typical
auditory language reception because of the absence of a
linguistic context from which to draw additional meaningful

cues.

-109-

The hearing vocabulary levels, as defined in this
study, certainly did account for a highly significant amount
of the variability in the performance of the word recog-
nition task, as was anticipated on the basis of previous
research.1 As indicated by the absence of statistical
significance in the Duration x HVL interaction, the influ-
ence of Duration was not sufficient to disturb the strong
systematic influence of hearing vocabulary levels or the
meaningfulness of the stimuli. In fact, the levels of
hearing vocabulary, as construed in this experiment, may
have been so pervasive as to mask the influence of the
experimental treatment, especially in the second and third
levels. Both of these levels encompassed an extremely
broad range of meaningfulness, in terms of the frequency
with which the words within the levels are used. Had levels
been defined on more homogeneous groups of words (re: fre—
quency of usage) perhaps the interaction of durations within
hearing vocabulary levels would have been significant in
more than just statistical sense.

Nevertheless, at the durational increases employed,
no significant positive influence on any reduction in avail-

able vocabulary or retention span was observed. It was

 

1Joseph M. Wepman,I§£.§l., "Psycholinguistic Study
of Aphasia: A Revision of the Concept of Anomia," JSHD,
XXI (1956), pp. 468-77. Hildred Schuell, James Jenkins
and Lydia Landis, "Relationship Between Auditory Compre-
hension and Word Frequency in Aphasia," JSHR, IV (1961),
pp. 30-36.

-110-

not surprising that durational increases did not affect

all subjects,1 since the level of auditory language function
was thought to be a variable affecting any possible posi-
tive effect from duration increases. Scores on the Auditory
Disturbance Section of the MTDDA did predict those subjects
who would do well on the tasks of this study and those who
would not. This was indicated in the significant main effects
of Groups in both analyses of variance. Of course, test
performance and task performance were not totally indepen-
dent. The MTDDA section used includes tests that reportedly
measure the same dimensions of the language deficit as the
tasks of this experiment.2 However, pattern of performance
within groups deserved close examination.

Several very interesting observations were made
within the group of subjects with high auditory language
disturbance at the several durations. There seemed to be
little question that their behavior on the tasks was dif—
ferent from the behavior of the other group of subjects.

The difference was not just in level of performance but
in the kind of performance.

This latter type of difference is eSpecially true

of their performance on the word recognition task. It was

 

lIbid.

2Hildred Schuell, James J. Jenkins, Edward Jimenez-
Pabon, Aphasia in Adults: Diagnosis, Prggposis, and Treat-
ment (New YOrk: Harper and Row Publishers, 1964), pp. 340-41.

~111-

tfluis HALD Group that appeared to generate the statistically
sjugnificant interactions that were obtained. It was diffi—
CLilt to attach meaning to the interaction of groups within
dharations. However, a visual inspection of performances of
tflue HALD Group at each of the hearing vocabulary levels across
1ihe four durations revealed interesting patterns, even
“though the magnitude of the variations were small. -At
IEVL II, the positive change associated with the 125% dura-
‘tion was contrasted with the relatively equivalent mean
:scores the group achieved at HVL I. For less meaningful
vvords a slight increase in duration effected a slight im-
Iprovement in performance, after which the group performance
vvorsened at the higher levels of duration. In the plot of
Dneans for the least meaningful words, HVL III, a slow posi-
t:ive change was observed that reached its peak at 150%
diuration. Again, for the least meaningful words, a still
Egreater increase in duration was required to make the lin-
Sguistic significance of the stimulus available to the sub—
:ject with poor auditory language function. Perhaps these
(abservations do not deserve to be called trends, but at
ILeast they were patterns which attracted the attention of
‘the investigator.

When the inter-task relationships were considered,
.it was again this HALD Group that attracted first attention.
.All of the significant correlations fell in this group.

'Ihis significant degree of association may reflect a

-112-

general low level of language function. The apparent lack
of significant association between task performance in the
other group of subjects, the LALD Group, provided some
support for the notion that reduced available vocabulary and
impaired retention span are two basic characteristics of

the aphasic language deficit which tend to improve or oper—
ate with some degree of independence as auditory language

function improves.

CHAPTER V

SUMMARY AND CONCLUSIONS

This study was concerned with an investigation
of a clinical procedure used to facilitate the auditory
language reception of aphasic adults. The influence of
increased stimulus durations was assessed in relation to
the meaningfulness of the stimulus and the level of audi-
tory language disturbance manifested by the patient on
tasks involving word recognition and retention span for

words, with auditory stimuli being presented on both tasks.

Summa y

Recognizing the importance of auditory language
reception problems in the total rehabilitative effort with
the aphasic adult, Speech pathologists have sought to
develop modifications of the auditory language stimulus
that will make the stimulus more meaningful and more ade-
quate for the aphasic adult. .One modification that has been
recommended, and appeared to have rather widespread clin-
ical application, was increasing the duration of the words
Spoken to the aphasic adult by increasing the temporal
value of the phonemes within the word. Clinical observa-

tion has indicated that some aphasic adults profit from

-113-

-ll4-

this procedure at certain points in the recovery process.

The purpose of this investigation was to assess
the effectiveness of such temporal expansions in relation
to other variables known to influence the reception of
auditory language by aphasic adults. Two language related
tasks were selected and developed for the purpose of test-
ing the independent influence of the experimental variable
and its interaction with the selected control variables.

Control was exerted over the range of auditory
language disturbance represented in the sample of twenty-
four aphasic adults by separating the subjects into two
groups on the basis of their performance on the subtests
of the MTDDA that are intended to assess auditory disturb-
ances in language. One group of twelve subjects was then
regarded as having a low degree of auditory language diS-
turbance, while the other group of subjects was regarded
as manifesting a high degree of auditory language distur-
bance. The groups were similar in educational achievement
and elapsed time Since onset of aphasia but did Show a
statistically Significant difference in age, with the high
disturbance group having the higher median age.

A standardized picture vocabulary test (PPVT) pro-
vided a means of controlling a second variable known to
influence auditory language reception, in addition to pro—
viding one of the two language tasks used in the study.

The word stimuli within this test were divided in a manner

-115-

providing four lists of seventy—five words each that were
regarded as equivalent in terms of the chronological de-
ve10pment of early hearing vocabulary. Within each list,
three levels of hearing vocabulary were defined arbitrarily
and were regarded as reflecting decreasing levels of mean-
ingfulness. Hence, the independent influence of the mean-
ingfulness of the stimulus word was assessed in its inter-
action with the two groups of subjects and with the experi-
mental variable of increased stimulus duration on this word
recognition task.

AS a second task upon which to test the influence
of the experimental variable, a retention Span task was
develOped. As this task was structured, a series of dis-
crete words were presented auditorily, and the subjects
responded by pointing to pictures of objects named. The
pictured objects were equated in terms of meaningfulness
and grammatical class (nouns). The auditory stimuli varied
in complexity as the stimuli varied in length from two to
five words, with all complexity levels represented equally
at all four levels of stimulus duration.

The tape recorded auditory stimuli associated with
both tasks were assessed to determine if their recorded
durations approximated a duration that would be expected
in connected Speech and to determine if they were intelli-
gible to listeners. Satisfying these criteria, the tape

recorded stimuli were submitted to an electromechanical

-ll6—

processing that increased their duration to specified
nominal levels. The nominal levels of increased duration
were 125%, 150%, and 175%. These temporally expanded re-
corded words, plus the recorded words of normal duration
(100%), provided four levels of the experimental variable—-
stimulus duration.

On a subject by subject basis, the stimuli for
the word recognition task and the retention span task were
systematically randomized across the four stimulus dura-
tion conditions, and presented to the subjects under con-
ditions resembling an average clinical environment.

The dependent variable was the number of correct
identifications made by the subjects on each of the two
tasks. The data from the word recognition task were treated
by a three factor analysis of variance in which the factors
were (1) hearing vocabulary levels, (2) subject groups,
and (3) stimulus duration. A two factor analysis of var-
iance was used to evaluate the data from the retention
Span task, the factors being (1) subject groups, and (2)
stimulus duration. An inter-task performance comparison

was made using a Spearman rank correlation coefficient.

Conclusions

Within the limitations of this study, the inves-
tigator believes there are certain conclusions that can

be drawn tentatively from the hypotheses tested in this

-117-

investigation and from observations made of other outcomes
of this research.

1. In general, increasing the duration of spoken
Single words does not improve their recognition or reten-
tion by aphasic adults.

2. However, the possibility exists that aphasic
adults with poor auditory language function might profit
from increases in the duration of words in terms of being
better able to recognize the meaning of the words.

3. Within groups of words having considerable
variation in their meaningfulness (re: development of
auditory receptive vocabulary or frequency of usage), in-
creasing the duration of the words is not likely to cause
any general improvement in the aphasic adult's ability to
recognize the words.

4. Word recognition performance and retention Span
performance appear to function with a greater degree of
independence as the auditory language ability of the aphasic
adult improves. These performances seem to reflect on
the constructs of reduced available vocabulary and impaired
verbal retention span, two disabilities reported to be
primary characteristics of the aphasic language deficit.

5. The language deficit in aphasia manifests a
very systematic reduction in word recognition performance
(available vocabulary) as observations are made across

groups of subjects that manifest different degrees of

-118-

auditory language disturbance. A major factor in the sys—
tematic nature of the reduction is the meaningfulness of
the stimulus.

6. In investigations such as the present study,
it is imperative that the investigator quantify and con-
trol for the expected variations in language deficit to
be found within any sample of aphasic adults. Research
that ignores such variations, especially variations in
the language modalities being studied, will fall far Short
of providing the better understandings of aphasic language

problems that need to be attained.

Suggestions for Future Research

From the observed performance patterns of aphasic
adults on the tasks of this study under conditions of
increased stimulus duration, several lines of investiga-
tion can be suggested.

It appears that a study quite Similar to the pre-
sent study could be meaningfully replicated with a group
of aphasic adults manifesting moderate to severe reductions
in auditory language function. Selecting more homogeneous
groups of subjects within the lower range of auditory lan-
guage function would provide a clearer analysis of the
influence of increased stimulus duration on auditory lan—
guage reception.

In addition to employing more homogeneous samples

-ll9-

of aphasic subjects, any repetition of a study like this
should narrow the range of meaningfulness within the arbi-
trarily defined levels of meaningfulness.

A more linguistically oriented approach employing
increased stimulus durations or expanded Speech might be
considered. It would be interesting to study the influence
of temporally expanded Speech on the performance of aphasic
adults in comprehending linguistic constructions of varied
complexity.

It would be reasonable, if it is established that
SXpanded Speech does have a positive influence on some
aphasic adults, to examine the influence of expanded Speech
input on the Speech output of aphasic adults. The working
hypothesis in an investigation of this type would be that
improved Speech comprehension Should lead to improved Speech
expression. Perhaps, in this area, a means could be devel—
Oped to test the effectiveness of expanded Speech as a
procedure for improving the self-monitoring of Speech output
in aphasic adults.

A word of caution is in order regarding the use of
the expansion process employed in this study. It is strongly
recommended that investigators check the actual level of
expansion achieved by the processing. The nature of the
process is going to produce variations in the levels of
expansion actually achieved. It is important to know what
this variance iS with reference to the nominal levels of

expansion being sought.

-120-

It iS this investigator's opinion that this line
of research is one that needs, and is deserving of, further
investigation. At best, this study has taken but a glimpse
of an important aspect of treatment for the aphasic adult.
Additional experimental efforts are needed to discover the
clinical Significance of changes in stimulus duration on

the auditory reception of language by aphasic adults.

BIBLIOGRAPHY

-12l-

BIBLIOGRAPHY

Books

Brown, Roger. Words and Things. Glencoe, Ill.: Free
Press, 1959.

Dunn, Lloyd. Peabody Picture Vocabulary Test: Expanded
Manual. Minneapolis: American Guidance Service,
Inc., 1966.

Eisenson, Jon. Examining for Aphasia. Rev. Ed. New York:
Psychological Corporation, 1954.

Fairbanks, Grant. Experimental Phonetics: Selected Writ-
ings. Urbana: University of Illinois Press, 1966.

Gibran, Kahil. The Prpphet. New York: Alfred A. Knopf,
Inc., 1964.

Goldstein, Kurt. Language and Language Disturbances.
New York: Grune & Stratton, 1948.

Hays, William L. Statistics for Psychologists. New York:
Holt, Rinehart, and Winston, 1964.

Harlow, Harry F. and Woolsey, Clinton N., eds. Biological
and Biochemical Bases of Behavior. Madison: Uni-
versity of Wisconsin Press, 1958.

Head, Henry. Aphasia and Kindred Disorders of Speech,
Vol. 1. New YOrk: Macmillan, 1926.

Jackson, J. Hughlings. The Selected Writing_. 2 vols.
reprinted New York: Basic Books, Inc., 1958.

 

Johnson, Wendell, Darley, Frederic L. and Spriestersbach.
Diagnostic Methods in Speech Pathology. New YOrk:
Harper & Row, Publishers, 1963.

Lindquist, E. F. Design and Analysis of Experiments in
Psychology and Education, Boston: Houghton Mifflin
Company, 1953.

Longerich, Mary C. and Bordeau, Jean. Aphasia Therapeutics.
New York: Macmillan Co., 1954.

-122-

-123-

Luria, Aleksandor Romanovich. Higher Cortical Functions
in Man. trans. by Basil Haigh. New York: Basic
Books, Inc., 1966.

 

Osgood, Charles E. and Miron, Murray S. Approaches to the
Study of Aphasia. Urbana: University of Illinois
Press, 1963.

Penfield, Wilder and Roberts, Lamar. _§peech and Brain
Mechanisms. Princeton: Princeton University Press,
1959.

 

 

Schuell, Hildred. Minnesota Test for Differential Diagnosis
of Aphasia: Administrative Manual. Minneapolis:
University of Minnesota Press, 1965.

. Differential Diagnosis of Aphasia with the Minne-
sota Test. Minneapolis: university of Minnesota
Press, 1965.

 

Schuell, Hildred, Jenkins, James J. and Jimenez-Pabon,
Edward. Aphasia in Adults: Diagnosis, PrognosisL
and Treatment. New York: Harper & Row, Publishers,
1964.

 

Siegel, Sidney. Nonparametric Statistics for Behavioral
Sciences. New York: McGraw-Hill Book Company,
1956.

 

Thorndike, Edward L. and Lorge, I. The Teachers Word Book
of 30,000 Words. New York: Bureau of Publications,
Columbia University, 1944.

Wechsler, David. The Measurement and Appraisal of Adult
Intelligence. Baltimore: Williams & Wilkins Co.,
1958.

Weisenberg, Theodore and McBride, Katherine E. Aphasia:
A Clinical and Psychological Study. reprinted
New YOrk: Hafner Publishing Company.

 

Wepman, Joseph M. Recovery from Aphasia. New York: Ronald
Press Company, 1951.

Wepman, Joseph M. and Jones, Lyle V. The Language Modalities
Test for Aphasia. Chicago: Education-Industry
Service, 1961.

Von Ziemssen, Hans, ed. (Cyclopedia of the Practice of
Medicine. New YOrk: Frank Wood, 1887.

Winer, B. J. Statistical Principles in Experimental Design.
New York: McGraw—Hill Book Company, 1962.

-124—

Articles and Periodicals

Benton, Arthur L. and Joynt, Robert J. "Early Descriptions
of Aphasia," AMA Archives of Neurology. III (1960),
pp. 205-22.

Deese, James. "Form Class and the Determinants of Associa-
tion," Journal of Verbal Learning and Verbal Behavior,
I (1962), pp. 70-84.

Foulke, Emerson, Amster, Clarence H., Nolan, Carson Y. and
Bixler, Ray H. "The Comprehension of Rapid Speech
by the Blind," Journal of Exceptional Children,
XXIX (1962), pp. 134-41.

Freud, Esti D. "Recent Trends in Aphasia," American Journal
of Psychiatry, CX (1953), pp. 186-93.

Geschwind, Norman. "Disconnexion Syndromes in Animals and
Man, Parts I and II," Brain, LXXXVIII (1965), pp.
237-93 and 585—644.

Jenkins, James J. and Schuell, Hildred. "Changing Concepts
of Aphasia," Perceptual and Motor Skills, XIII
(1961). P. 270.

"Further Work on the Language Deficit in Aphasia,"
Psycholpgical Review, LXXI (1964), pp. 87-93.

 

Jones, Lyle V. and Wepman, Joseph M. "Dimensions of Lan-
guage Performance in Aphasia," Journal of Speech
and Hearing Research, IV (1961), pp. 220-32.

Kann, Jules. "A Translation of Broca's Original Article
on the Location of the Speech Center," Journal of
Speech and Hearing Disorders, XV (1950), pp. 16-20.

Luterman, David M., Welsh, Oliver L. and Melrose, Jay.
"Responses of Aged Males to Time-Altered Speech
Stimuli," Journal of Speech and Hearing Research,
IX (1966), pp. 226-30.

Orr, David B., Friedmann, Herbert L. and Williams, Jane C.C.
"Trainability of Listening Comprehension of Speeded
Discourse," Journal of Educational Psychology, LVI
(1965). PP- 148-56.

Research Needs in Speech Pathology and Audiology: A Special
Report by The Committee on Research, American Speech
and Hearing Association. Steer, Max D., Chrm.
Washington, D.C.: American Speech and Hearing
Association, 1959, pp. 40-43.

-125-

Schuell, Hildred. "Auditory Impairment in Aphasia: Sig-
nificance and Retraining Techniques," Journal of
Speech and Hearing Disorders, XVIII (1953), pp.
14-21.

 

Schuell, Hildred and Jenkins, James J. "The Nature of the
Language Deficit in Aphasia," Psychological Review,
LXVI (1959). pp. 45-67.

. "Reduction of Vocabulary in Aphasia," Brain,
LXXXIV (1961), pp. 253-61.

Schuell, Hildred, Carroll, Virginia and Stansell, Barbara.
"Clinical Treatment of Aphasia," Journal of Speech
and Hearing Disorders, XX (1955), pp. 45—53.

Schuell, Hildred, Jenkins, James J. and Landis, Lydia.
"Relationship between Auditory Comprehension and
Word Frequency in Aphasia," Journal of Speech and
Hearing Research, IV (1961), pp. 30-36.

Spiegel, Douglas K., Jones, Lyle V. and Wepman, Joseph M.
"Test ReSponseS as Predictors of Free-Speech Char-
acteristics in Aphasic Patients," Journal of Speech
and Hearing Research, VIII (1965), pp. 349-62.

 

Street, Barbara Stansell. "Hearing Loss in Aphasia," Journal
of Speech and Hearing Disorders, XXII (1957), pp.
60—67.

Tiffany, William R. and Bennett, Delmond. "Intelligibility
of Slow Played Speech," Journal oijpeech and Hearigg
Research, IV (1961), pp. 248-58.

Wepman, Joseph M., Jones, Lyle V., Bock, R. Darrell and Van
Pelt, Doris. "Studies in Aphasia: Background and
Theoretical Formulations," Journal of Speech and
Hearing Disorders, XXV (1960), pp. 323—32.

Wepman, Joseph M., Bock, R. Darrell, Jones, Lyle V., and
Van Pelt, Doris. "Psycholinguistic Study of Aphasia:
A Revision of the Concept of Anomia," Journal of
Speech and HearinngisorderS, XXI (1956), pp. 468-77.

Unpublished Materials

Allen, Hugh 8.. Jr. "The Eltro Information Rate Changer
Mark II: Simple Quality Speech Compression," Paper
presented at the Conference on Compressed Speech,
University of Louisville, October 19, 1966 (Mimeo-
graphed).

-126-

Foulke, Emerson and Sticht, Thomas. "A Review of Research
on Time Compressed Speech." From the Proceedings
of the Louisville Conference on Time Compressed
Speech, University of Louisville, 1967 (Mimeographed).

Kurtzrock, George H. "The Effects of Time and Frequency
Distortion upon Word Intelligibility." Unpublished
Ph.D. Dissertation, University of Illinois, 1956.

Michigan State University: Agricultural Experiment Station.
"STAT Series Description No. 14: Analysis of Variance
with Equal Frequency in Each Cell." E. Lansing,
Michigan State University, 1968.

Perry, Peter S. and Boswell, Nathalie 8. "Adult Aphasia
Performance of the Peabody Picture Vocabulary Test."
Paper presented at the American Speech and Hearing
Association Convention, November 3, 1967.

Perry, Peter S. Personal communication, November, 1967.

APPENDICES

-127-

APPENDIX A

FURTHER DESCRIPTION OF PROCESS BY
WHICH STIMULUS DURATIONS WERE INCREASED

This description will begin with a general discus-
sion of the mechanical process, followed by a discussion
of the electromechanical aspects of the process that account

for the temporal expansions. In Figure 3 is a schematic of
the device.

FIGURE 3.--Schematic of Speech
Expansion Device.

 

 

 

 

 

 

 

 

 

 

. 0)
2 l

633;. j

33.: —’ 0

C(35 u dTfib

msaCho a? r

%8Ҥ8

naasoch’

 

 

 

The tape is directed to the surface of the cylinder
(A) in such a manner as to make contact with one-quarter of
the cylinder. The tape is driven by the roller and capstan

(B). The tape supply and take-up mechanism, an independent
unit, are represented at (S) and (T).

When the device is activated for some level of ex-
pansion, the tape is advanced in the direction indicated.

-128-

-129-

The cylinder rotates in a clockwise direction. The relative
velocity of the tape transport and the effective linear
velocity of the head rotation is maintained at 15 ips--the
Speed at which the Signal on the tape had been recorded.

AS the tape tranSport Speed is reduced for greater levels

of expansion, the rotational Speed of the cylinder is in—
creased. Hence the speed of the tape relative to the surface
of the cylinder (and reproduce heads) is held constant at

15 ips.

From the schematic diagram, it can be imagined that
each of the heads, in turn, makes contact with the tape and
that only one head is in contact with the tape at any moment.
A Single control knob regulates the tape Speed and head Speed
synchronously to maintain the relative Speed of 15 ips. The
calibration of the control knob is in percentage of original
time expansion or compression. The frequency response is
reported to be essentially flat to 15000 Hz at all four
reproduce heads.

The increases in duration can be produced over a
range extending up to approximately 200% of the duration
of the original signal. If a word had an initial duration
of 500 msec., expanded version of the word could be produced
which would have durations ranging up to 1000 msec. In
the sampling process associated with the production of the
expansions, several aspects of the samples themselves need
to be understood.

First, the temporal value of the acoustic patterns
within each sampled segment of the original Signal or word
is increased as the percentage of desired expansion is
increased. If a segment of the original recording had a
duration in real time of 30 msec., and the Signal was being
processed for 150% expansion, this segment would have a
duration of 45 msec. in the expanded reproduction. Its
temporal value now equals 150% of its temporal value in the
original recording. The increase in the temporal value of
the acoustic patterns in each segment is a function of the
reduction in the tape transport velocity. It is necessary
to make such reductions as greater expansion levels are
sought. Where one inch of the recorded word represented
1/15 of a second at its recorded velocity of 15 ipS, one
inch of the tape recorded Signal transported at a reproduc-
tion velocity of 7.5 ips now represents 1/7.5 of a second,
or just twice its temporal value in the‘original recording.
This reduction in transport velocity during expansion pro-
cessing would result in the frequency distortion and intel—
ligibility loss (associated with "slow played" recordings
of Speech) were it not for the compensatory effect of the
rotating reproduce heads. This compensatory effect of the
rotating reproduce heads restores the reproduction velocity
to the level of the recording velocity.

-130—

This restoration comes from develOping a relative
velocity between tape surface and the gap of each reproduce
head which is equal to the 15 ips at which the word was
originally recorded. Since the tape is being transported
in a direction opposite to the linear direction of the head
rotation, the relative velocity of the reproduction process
(rV) is the sum of the transport velocity (VT) and the
effective linear velocity of the rotating heads (VR), i.e.
rV = VT + VR. In the example given, reducing transport
velocity from 15 ips to 7.5 ips would require a compensatory
velocity of head rotation equivalent to 7.5 ips. Achieving
this relative velocity would give to the expanded reproduc-
tion the frequency characteristics of the original Signal.
It is necessary to observe, however, that as rotational
velocity increases, the number of samples taken per unit of
time increases. Each of the rotating heads is scanning a
segment of the recorded tape in less time, hence more samples
are taken per unit of time. This results in portions of
the original Signal being sampled a second time as the samp-
ling produced by a given head overlaps a portion of the tape
already sampled. But, because the relative velocity with
which this scanning occurs is equivalent to the velocity of
the original recording, each segment of tape sampled is
“treated" as though it were being transported across a sta-
tionary reproduce head at a velocity of 15 ips, when in fact
it is not. In the example being used, the segment being
scanned represents 1/30 of a second of the original recording,
Since the transport velocity has been reduced by 50%.from
15 ips to 7.5 ips. However, with the relative velocity of
the reproduction process functioning at 15 ips, that seg-
ment is given a temporal value of 1/15 of a second. In the
expanded reproduction, the acoustical patterns in that seg-
ment of the original recording now have a temporal value
which is twice that of their original temporal value. The
temporal value of the segment has been increased by 200%.

Repeating this sampling process throughout the ori-
ginal duration of a word would result in an increase in the
duration of the word equivalent to 200% of its original dura-
tion. It should be recognized that with this type of expan-
sion processing it is highly unlikely that any two expansions
of a recorded word will result in exactly the same expansion
of the acoustical patterns of that word. This variability
results from the differences that will arise in the defini-
tion of segments from successive processing of the same
utterance. The linear value of each segment is invariant
and is determined by the distance (on the circumference of
the cylinder containing the reproduce heads) between the
reproduce heads. This linear distance on the Eltro Infor-
mation Rate Changer - Mark II is 13. However, in successive
expansions of the same recorded utterance, it is highly
unlikely the beginning of that utterance will fall at

—13l—

exactly the same point in the Space between any two of the
four reproduce heads. Hence, segments will be mapped on
the word differently from one reproduction to t? next.
The influence of such variations on perception are unknown
at present.

To illustrate expansion processing, the example used
has been that of 200% expansion because it provides the
clearest case. The assumption has been that the relation-
ship between reductions in transport velocity and effective
velocity of head rotation is relatively linear through the
expansion capability of the Eltro device, the three expansion
levels used in this investigation could be analyzed as the
200% expansion level has been analyzed. However, the rela-
tionship is not exactly linear as indicated in Figure 4.
Calculations based on the assumption of linearity from such
calculations will not be precisely the expansion levels
sought.

From the experience in this study, it is necessary
to refer to the obtained levels as nominal levels and Specify
the observed variation from the stated value of the level
which has been allowed. AS reported, in this study, a 5%
variation from the stated nomial value was allowed. Without
knowing the exact values for the transport velocity and the
rotational velocity at the three expansion levels employed
in this study, the calculations could not be meaningfully
made.

-l32-

FIGURE 4.--EXpanSion Process: Tape
TranSport Velocity vs.
Percentage of Expansion.*

 

Percentage Expansion

 

 

 

 

 

 

 

 

 

 

 

 

 

 

200%
17 5% \\
\
150% \\
1.25% \
o m C) m d
o h u» N o
m m o: H o
:4 ox' F} or; u;
F1 H H

Tape Transport Velocity

*from Allen,

p.

9.

"Eltro Rate Changer,"

amugum

 

 

 

cmmsv won monm 3mm
Adv
cmupaflno a mmuu.monm Hmmn USmHmH Smupaflno monu
mocom mso 3mm mmmam
Ams
moo G 3mm Hmflvaom GOHUHHQU cflmucsofi mop
EmUHum
cmo mmuu ummn own
Ass
own» w coupaflso sflmummo mcflu cmupaflno Hmflpaom
o.H ”SOHumusa
.
B msasaflum MMMflM
1..
_ Amvs ocmamfl Isms mmn Ame moo
laws cumummo Amas mocmu Ame Hams
Ammo :mmss ISAC mags Ass Huum
Ammv umHUHOm Aoav ummn fies mop
Ammv nonommu Amv :mo Amv wwuu
Ammo 3mm Ams Hams Ame mason
Isms “mm: Amy cmucaflno Ame snap
Aomv mmmHm Anv umon Amv 300
Ammo campuses 16v Hmmcfim Ass magma
Ammo nm>flu Ame worm Ads umo

.Ucdom mm3 mus»

loam nofl£3 SH oumam B>mm mmumoflvcfl mflmwSDCwumm CH Honesz .mUHOB :50: on mcofluooamm mGHuHEHH
pcm.mfl mo mmcflumu mmuoqlmxflosuona Um: mmEmc mm0£3 mmusuoflm haco mcflms mmumam om umuflm umme
mnmasnmoo> muzuoflm wvonmmm map Eoum COCHmqu mmusuoflm om mcfl3oHHom mnu Eoum Ummoam>mp mmumam

mDQDEHBm QMBdHUOmmd 02¢ mafifiqm mmZOmmmm ZOHBZWBMM m0 dem
m NHQmem¢

Gmmgw w uwmc

QSO Q HmU
smmHOQSHHMqummg

Cmo w HmO
.mmHO£.Hm£0mmu.HHmfl

-l34_

cmmsv w
3mm.mmmam.ummc

Q90 w
mmMHmsngmW~CHmUmmu

magma w mmmam.um£ommu.

 

msassflum

mso

mmHOS

mmmHm

HHmQ

mmHOQ

mGHH

UCMHmH

pmmc

DMOQ

GHMDQMO

maflmu

Hamn

 

 

 

ummfl mogm
mmmam umHUHOm
umHUHOm moo
ummc Huflm
Unmamfl Honommu
Cmo moo
HMOQ COOSU
mmmHm Hamn
cmuoaflno mocmw
moo GHmaGSOE
mmmam mcflu
Hmflvaom Hmmn
M3.

 

Smuflum
Gmwsv
Ass
camucsofi
Hmo
Amy
me03
Smufium
Hamn
fies
Hmo
Hmo
Amy
3mm
EWUHIV
Hmmcflm
Acs
mmmam
cmo
Amv
Honommp

A.ucoov m XHQmem¢

-135—

own» a ummcaw
.Huam.mso.mucmm

OHQMD
o mmuu.3mm.mmu0£

cflmummo
w 3oo.no>fln.mon

GHmuGSOE
w mmn.ummn

manmu
w mocwm.cmmsv

cmucaflgu w magma

 

msassflum

mso

GHmuGSOE

mso

mmuon

Honommu

own

3mm

magma

cflmucdoe

maflmu

manna

mmn0£

 

 

 

Huwm mocmm
mono wmuu
mmuu 300

Sam umon
Ho>HH cmmsv
Hmmn 300
aflmucsoE won
umon Hm>flu
mocmm Hamn
mso monm

mcflu umon
aflmucsofi ummCflm

mumam

amuuum

mmnoz

Amy

Hmmcflm

um>HH

Amy
manna

Emufluw

ocmamfl

A¢v

cflmummo

cflmummo

Amy

Hmmn
Ewufilm
cmmsv

Aév

pcmamfl

mow
Amv

amuwugnu

A.ucoos m xHozmmm<

mmuu a
aflmummo.300.mocmm

OOSm
OW meonsaHmmGHN

CHM#CSOE

fi smmHogsmCHH

-l36_

COOSU fi OOQ

Hmmfl Q HO>HH

monm w wmuon
~00C0W~UCMHWH~HMQQ

 

OmHOQ

300

OmHOS

QOQ

NQMQ

mOU

OOHU

mCHH

Hm>HH

umflpaom

mocmm

300

 

 

 

cflmummo moo
mmuu cmmsv
Hwflomwu Hmmcflw
magmu cmo
women HOHUHOm
cflmucsoa Samummo
uMOQ 30m
xflmn umo
:mo mafia
ummcflm Honommu
mmmam panama
wmuoz own
man...

 

HHOQ
14s
mocmw
OOSm
Amy
umOQ
OOSm
Adv
OGHH
cmmsv
Amv
OOQ
HMOQ
Ass
umofl
omoH "QOHHMHSQ
003m
Am:
HMOQ

I “zoos m xHozwm
me

HHOQ
w mUcmM~QHMUGSOE

OOQm % BOO

mmuoz w bcmamfl

-l37-

mafia w Unmamfl
.mop.mocmm.cmm5v

.300.Hamn.umon

Hams w
HHOQ.HHHm.GmU

a

Hmmcflm

mmHOQ

GHMDCSOE

OOQm

Hmmcflm

umflpaom

TUGGM

@OU

mmmam

ummcflm

damn

Hamn

 

 

 

HHmQ mmmam
mGHH TOGO“
300 Emu
OHQMH Hmwfl
Hufim pcmamfl
003m mmHO£
mafia mmuon
UGmamH Gmmsv
HHOQ HMOQ
COODU HHMQ
mso mmmam
Huflm Cmo
an.

emUHlm
:Hmummo

Ass

GHmuGDOE

HO>HH

Ams
HHMQ

emUHIN

HO>HH

Amy
UMOQ

 

mb.H "GOHHMHDQ

GOHUHHQU
Amv
Hamn
emgfllm
300

Ame

HO>HH

OUEmw

Amv
UMOQ

-l38-

magma O mmHOQ
samQCsmMMHmscmwsv

~®O£m~mUcmM~mCHH

OHQMU Q

Hmmfl fi
smGHHshgmggummC

mmnm OW mGHH smog

a

wanmu

mmmam

QUGQM

monu

coupaflno

Hugo

NQMQ

umoc

mCHH

ummc

 

 

 

mow mmHO£
among ummc
umnummu mso
mozm GHmuQmu
mow cflmucsofi
manmu won
mGHH HMOQ
Hmmn wmnu
umflommu UGmHmH
mop Hm>HM
a

Hams
:3
won
Smuflum
mCHH

A<V
Huflm

HO>HH

Ev
HHOQ

EQUHl¢

OOQw

:3

mmmam

manmu
Am:

own

A.ucoov m XHQZMQQ<

-l39-

Hm>umucH w>onm aflnufl3 mpHOB mamcflm Had

 

.Avv.m¢o .mm.vav «muHEHH Hm>umucH
m.hmm no sumo
H.mmm an sumo:
moHHm 00 *HMQQ
m.>a¢ mm qu
«.mmv mo Edna
m.nmm no ammoo
¢.mmm mo pom
v.mmm mo ummc
©.Ho¢ Hm mama
N.mm¢ mo Mcsuu
A.ommEv A.EEV

COHHNHSQ SHUGUQ UHOS
cousmmmz

.mEmUH xmmu cowucoumu Eoum Umamfimm mvuoz mmDMOHUcH«

 

 

 

mUHOB mdmcflm mo coflumuzo

OZHQZOUWM deHOHmO ZH mQMOB m0 mZOHB¢MDQ mom
mA<>mmBZH m0249mm00¢ mqumdfimm OB QmmD ¢B¢Q

vm.voa u.Q.m ma.mm¢ "smmz
m.mom em swan
m.ovv om Hugo
m.om¢ Ho umo
¢.©vm wv mmma
n.mmm om com
N.>nm om ammoo
m.mmm mg G30Ho
n.mam mo mmnmn
m.o~m mm Hams
H.05m mg ummc
A.UmmEv A.EEV
Goaumusn numcmq puoz
UmuSmmoz

 

UXOHCOU EOHM UQCHMUQO mCOflUMHSQ

mouoz magmaawm mac

0 XHQmem¢

-140—

.mEmuH xmmu coaucmumu Eoum Umamfimm mUHOB mmumoflchs

Hm>umDSH m>onm canufl3 mpuo3 mamcflm HH4

 

 

 

 

.Amm.mon .o¢.omms "muflsfla Hm>umucH mn.oma ".o.m m¢.¢me cams
v.Ho© wm sflmummo
¢.mmm on «Hmmcfim
m.mom «m «mnmn
k.m¢o mm um>muo h.sm~ 4m mammm
m.¢ao ms Macmmmn >.mon om wsmcflums
m.mmo om Mooomwm o.mmo hm Hmucsoo
G.Hov Hm snag m.mmm om xomSmx
«.mmv mo msnmm 6.46v mm manmm
~.mm¢ mo mcfluuflg m.mm¢ mo mcflamu
o.mmm hm uSmEmmmo H.mm¢ mm umgofl3
N.Hmm on ucmvou m.¢mm mg mcflusom
m.mmo om uouuos s.mam mo snag
N.Hmm on magmas“ s.mmm om maucflnm
A.ommEV A.EEV A.OmmEv A.EEV
coaumnsn numcmq UHOB COHumusn fiumcmq who:
Umusmmwz Umusmmmz
mUHo3 mamsflm mo coflumusn uxmucoo Eoum pmcfimuno mcoflumusn

meuos magmaasm 039

A.uGOUv U NHQZHAQ<

Hm>umucfl m>onm cflnuw3 mUHOB wamcflm Had

 

 

 

 

.Amv.hom .no.mmmv "muHEHH Hm>umucH HH.0h u.n.m mo.om© "cam:
¢.nmn OOH quEmHUGH h.mo> om mum>moxm
o.~oo on mcflum>os m.oo> mm ucmpauom
o.mmo hm mmcmEEH m.mmm om usmEm>mmqu
h m.m>© mm coauumo m.mmo mm maHHMEQSm
4 ¢.vme mm msmommsmm m.mvm boa nouomfloum
4 m.mss mm Smflcoum m.oom ms magnumuflu
o.ahn mm ooummcmx m.nmm Hm conuusuummu
o.ann mm maflpasmmmm H.mmm me mcflaflmcoo
m.oon mm mum>oamu m.mmm mm mmon>am
m.mmh mm uwumficsmn n.mvo mm ooummsmx
A.ommEv A.EEV A.oomEv A.EEV
SOHpmHSQ numcmq UHOB :oHumHDO Qumcmq UHOB
UmHSmmmz pwHSmmoz
mouoz mamdflm mo cofiumusa uxmucoo Eoum pocflmuno mcofiumuso I

mpuoz maanHSm mmune

A.ucoos o xHozmmm¢

SOHDMHSG SH Hm>uwucfl may Umpmmoxm =coflumuuommcmuu:

 

 

 

 

.Ime.mmm .om.memv "muusfla Hm>umch mm.om u.n.m Sm.moe "cams
_
2
M m.m¢m boa Gowumunommamuu
_ 0.5mm wad coaumHSQEm
m.mom mm mumuwuflano
m.smn moa Suoumcwum e.mvo mm msoumuflcoo
0.5HHH oma coaumuuommsmuu ©.th mm Umumswmum
A.oomEV A.EEV A.ommEv A.EEV
coaumusn numcmq UHOZ COHumudQ numcmq UHOB
pmHSmmmE UmHSmmmz
mUHOB mamcflm mo coaumusa uxmucoo Eoum Umcflmuno mcoflumuso. (Ln

mpuoz mannaamm usom

A.uaoos o xHozmmma

COHDMHSU ca Hm>umusfl map Umpomoxm =GOfipMGHESHHH=

 

 

 

 

 

 

.Amh.amoa .nm.¢mhv "muHEHH Hm>umucH ha.m© ".Q.m ma.mmm "cmmz
% H.oom OHH cofluommmflummmflo
M 0.3m m3 coflmfimflumua
_ «.05m «NH msoflumumamp
6.06m NNH umflmoHomamm
o.mmoa mma GOHumGHESHHH m.©mm moa coflumgflfizaafl
A.UmmEv A.EEV A.ommEV A.EEV
COHumusa numcwq UHOB GOHDMHSQ numan UHOB
pmusmmmz UmHSmmmz
mUH03 mammflm mo cofiumusn uxwucoo Eoum Umcflmuno mcoflumusn

mpuoz mHQmHaxm m>Hm

A.uCOUv U XHQmemm

APPENDIX D

CALIBRATION CHECK OF EXPANDED TAPES

The 1000 Hz tone at the beginning (I) and the end
(F) of each tape (Al, A2, B1' and B2) had a measured length
of 61 mm. in the original tape. The tone was of the same
length on the Retention Span tape (RT), but the tone ap-
peared only at the beginning of tape.

Hence, the length of the 1000 Hz tone as reproduced
in the expanded version of each tape provided an estimate
of the extent to which the stimulus material on that tape
had been expanded. The percentage of expansion was given

by the formula

Expanded mm.

Expansion = x 100%

 

61 mm.

The data below are the measured lengths of the 1000
Hz tone for each of the tapes at the nominal levels of
expansion. Data obtained from read-out of power level

recorder operating at paper Speed of 30 mm./sec.

-144-

—145-

APPENDIX D (cont.)

-__ fl:

 

m

E X P A N S I O N S

 

125% 150% 175%
Tapes
mm. expns. mm. expns. mm. expns.
I 75 122.9% 90 147.5% 108 177.0%
A1
F 75 122.9% 89 145.9% 108 177.0%
I 75 122.9% 90 147.5% 108 177.0%
A2
F 74 121.3% 89 145.9% 108 177.0%
I 74 121.3% 90 147.5% 108 177.0%
B1
F 74 121.3% 90 147.5% 108 177.0%
1 75 122.9% 90 147.5% 109 178.7%
82
F 75 122.9% 90 147.5% 109 178.7%
RT 75 122.9% 90 147.5% 109 178.7%

 

-146-

 

 

Ho Ho Ho Ho so Mo Mo mo Mo HH mo HH sH Mo oH No HMNMNMHM NH
No Ho No Ho sH NH MH MH oN HN Mo MH HN mo MH MH NMHMHMNM HH
Mo No Ho Ho oH MH NN oH NH sN mo oN NN Mo MH MN HMNMNMHM oH
No No Mo so mo HN sN MH MN NN MH sN sN MH MH sN HMHMNMNM M
Ho so Mo Mo Mo NH sN Mo MH MH No MH HN mo MH MN NMHMNMHM M
Ho No so Ho so MH sN mo HN HN Mo MH NN No MH oN HMNMNMHM M
mo mo sH mo mo MH MN Mo HH NN mo HN MN oH oN MN NMHMNMHM M
Mo Mo No Mo Mo NH MN No MH sN Mo HN NN No MH sN HMHMNMNM M
so so No so Mo Mo MH Mo Mo HN Mo oH HN so oH HN NMNMHMHM s
Mo Mo Mo so mo NH MN oH HN MN No oN MN so HN MN HMHMNMNM M
oH sH NH Mo Mo MH MN oo oN MN sH sN MN mo MN sN HMNMNMHM N
so No so so oH NH MN No Mo NN Ho MH HN Mo sH HN NMHMNMHM H

HHMM
HH so MH sH NN MN MN HN MN MN HN NN MN oN MN MN HMNMHMNM NH
MN MN sH MH MH MN sN NH sN sN MH MN sN sH MN MN HMNMHMNM HH
MH sH MN sN HH NN MN Mo NN MN Mo MH MN Mo HN MN NMHMHMNM oH
MN MN MN MN mo NN MN oH NN MN NH NN MN NH MN MN NMNMHMHM M
oH mo oH Mo MH MN MN MH sN MN MH sN MN NH sN MN HMNMHMNM M
MN MN MN oH sH NN MN MH MN sN MH sN sN HH sN MN HMHMNMNM N
MN HH MH sN Mo MH MN Mo MH sN Mo MH sN Mo NN MN HMNMHMNM M
No MH HH sH No MH NN Mo MH NN Mo NH HN Mo HH sN NMHMHMNM M
MN MN MH MH oH sN MN oH NN sN MH MN sN mo MN MN NMHMHMNM s
sH MN sH MH No NN MN Mo HN MN oH MH MN mo oN sN NMHMNMHM M
sH sH sH sH MH MN MN MH MN MN MH sN MN NN MN MN NMNMHMHM N
MN MN MN sH Mo MN MN No MN MN Mo MN sN NH MN MN NMNMHMHM H

HHMH
MNH oMH MNH ooH HHH HH H HHH HH H HHH HH H HHH HH H MonHMMoH MM

Ass MonHMMoH NMNH NoMH NMNH sooH BM MHMHH
HMooM onHZMHMM M M M o o M z o H a H z o o o M M o M o z

 

 

 

 

 

 

 

 

mMde

ZOHBZMBmm QZ< ZOHBHZOOUmm ZO mfiumhmbm m0 mmmoom

m NHQmemfi