A STUGY OF VGWEL {NTENSE‘TY AS RELATED
TO THE LOMBARD £FFECT

Thesis {or the Degree 0‘ M. A.
MECHEGKN STATE L’HEVEKSF’EY
Carolyn Jean Livingstone
1963

 
 

LIBRAR Y

Michigan Sfitc
University

 
 
   

 

 

_,._ .—--

A STUDY OF VOWEL INTENSITY AS RELATED
TO THE LOMBARD EFFECT
By

Carolyn Jean Livingstone

An Abstract

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

Master of Arts

College of Communication Arts, Department of Speech

1963 ,
fl/mo/ ,

/

   

ABSTRACT

A STUDY OF VOWEL INTENSITY AS RELATED
TO THE LOMBARD EFFECT

by Carolyn Jean Livingstone

The purpose of this research study was to determine
the relationship between the Lombard effect and vowel in-
tensity with respect to specific questions. The questions
prOposed were: (1) How intense is the voice reflex for
each isolated vowel sound when Specific levels of white
noise (masking) are presented to the right and left ears?
(2) How intense is the voice reflex for each isolated vowel
sound when Specific levels of white noise (masking) are
presented binaurally? (3) Is there a significant difference
in the intensity of the voice reflex on vowel sounds between
the binaural and monaural masking? (4) Is there a system-
atic intensity pattern that characterizes the eleven
spoken vowel sounds?

Twelve, normal hearing, graduate students in Speech
and Hearing Science, Michigan State University were used
as subjects in the experiment. Equipment used consisted
of: (1) an audiometer, (Allison Model 20); (2) headsets,
(Telephonics TDH-—39); (3) micrOphone,(Electro-voice 654);
(A) tape recorder,(Ampex Model 60l);(5) tape, (Scotch Tenzar

Backing Magnetic Tape 311); and (7) a modified electric

‘Carolyn Jean Livingstone

timer, (GraLab Mocrotimer). The twelve subjects produced
a series of eleven randomized vowels at which time white
noise was fed into their ears. binaurally, right and left
at 70, 80, 90, and 100 db (re: 0.0002 microbar). ‘

The findings indicate the mean intensity of
utterances corresponding to the four intensitiies of white
noise differ significantly in binaural and monaural (right)
presentation. There appears to be no significant difference
in mean intensities in monaural (left) presentation. With
all three modes of presentation there is a significant
difference in the mean intensities of utterances corres-
ponding to the eleven vowels and there is significant
interaction between vowels and intensity of white noise.
The mean intensity for the three different modes of pre-
sentation differ significantly from one another. All
vowel sounds demonstrated a greater mean intensity when

the white noise presentation was binaural.

A STUDY OF VOWEL INTENSITY AS RELATED
TO THE LOMBARD EFFECT

By

Carolyn Jean Livingstone

A THESIS

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

MASTER OF ARTS

College of Communication Arts, Department of Speech

1963

TABLE OF CONTENTS

LIST OF TABLE 0 0 0 0 O O 0 O O 0 O I O I O O O 0
LIST OF ILLUSTRATIONS . . . . . . . . . . . . .
Chapter
I. STATEMENT OF THE PROBLEM . . . .
Introduction
Hypothesis
Distinctiveness of Study
Definition of Terms
II. REVIEW OF LITERATURE . . . . . . . . . . .

Vowel Intensity Studies
Lombard Effect Studies

III. SUBJECTS, EQUIPMENT, AND PROCEDURES. . . .
Introduction
Subjects
Equipment
Procedures
IV. RESULTS AND DISCUSSION . . . . . . .

Results
Discussion

V. SUMMARY AND CONCLUSIONS . . . . . . . .
Summary
Conclusions
Implications for Further Research
APPENDICES. . . . . . . . . . . . . . . . .
BIBLIOGRAPHY. . . . . . . . . . . . . . . . . .

ii

22

27

54

58
86

Table

10.

ll.

12.

13.

14.

LIST OF TABLES

Computational Formulas for Two Way
Analysis of Variance. . . . . . . .

Binaural Mode of Presentation
Two Way Analysis of Variance. . . . .

Monaural (Right) Mode of Presentation
Two Way Analysis of Variance. . . . .

Monaural (Left) Mode of Presentation
Two Way Analysis of Variance. .

Binaural Mode of Presentation
Difference of Means (Columns) . .

Monaural (Right) Mode of Presentation
Difference of Means (Columns) . . . .

Monaural (Left) Mode of Presentation
Difference of Means (Columns)

Binaural Mode of Presentation
Difference of Means (Rows). . .

Monaural (Right) Mode of Presentation
Difference of Means (Rows). . . .

Computational Formula for One Way
Analysis of Variance.

Modes of Presentation for Vowel [i ]
One Way Analysis of Variance.

Modes of Presentation for Vowel [I]
One Way Analysis of Variance. . . . .

Modes of Presentation for Vowel [8]
One Way Analysis of Variance.

Modes of Presentation for Vowel [31
thest. . . . . . . . . . .

iii

Page

28

28

29

. 29

31

32

.‘33

34

34

35

36

36

37

37

Table

15.

16.

17' .

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

Modes of Presentation for Vowel[£ ]
One Way Analysis of Variance. . . .

Modes of Presentation for Vowel[ C]
E-TESD. o o o o o o o o o o o o o 0

Modes of Presentation for Vowel[ae]
One Way Analysis of Variance. . . .

Modes of Presentation for .Vowel[33]
t- Test. . . . . . . . . . . . .

Modes of Presentation for Vowel[i4]
One Way Analysis of Variance. . . .

Modes of Presentation for Vowel[ u]
t MTeSt O O 0 O O O I I O O C 0

Modes of Presentation for Vowel[/\]
One Way Analysis of Variance. . . .

Modes of Presentation for .Vowel[ A]
-t_ -TeStO I O O I O I O O C C O 0

Modes of Presentation for Vowel[ Q]
One Way Analysis of Variance. . . .

Modes of Presentation for Vowel[ a]
t -Test. . . . . . . . . .

Modes of Presentation for Vowe1[ a]
One Way Analysis of Variance.

Modes of Presentation for Vowe1[.>]
t- Test. . . . . . . . . .

Modes of Presentation for Vowe1[ 0]
One Way Analysis of Variance.

Modes of Presentation for Vowe1[ o]

E-TeSt o o o o o o o o o o o o o o 0

Modes of Presentation for Vowe1[\f]
One Way Analysis of Variance. . .

iv

Page

38

38

39

39

4O

4O

41

41

42

42

43

43

44

44

45

Table

Page
30. Modes of Presentation for Vowe1[-u]
E-TeSt o o o o o o o o o o o o o o o o “'5
31. Modes of Presentation for All Vowels
Combined One Way Analysis of Variance . . . . 46

Figure

\OCDNQU‘I

10.
11.
12.
13.
14.
15.

LIST OF ILLUSTRATIONS

Intensity of Vowels--Binaural Presentation

Varied Intensities White Noise. .

Intensity of Vowels-~Monaural (Right)
Presentation-—Varied Intensities White

Noise

Intensity of Vowels--Monaural (Left)
Presentation-—Varied Intensities White

Combined Vowel S0unds--Modes of

Noise
Presentation.
vowel [ I Modes

Vowel [I Modes

Vowel[ 5 Modes
'Vowel[£ Modes

Vowel[ as. Modes

]
J
]
]
]
wae1.[ u ]Modes
Vowel [ A ]Modes
Vowel [ q ]Modes
Vowel [ J ]Modes
Vowel [0 ]Modes
]Modes

Vowel[v

of
of
of
of
of
of
of
of
of
of

of

Presentation.
Presentation.
Presentation.
Presentation.
Presentation.
Presentation.
Presentation.

Presentation.

Presentation.
Presentation.

Presentation.

vi

Page

49

49

50

50
51
51
51
51
52
52
52
52
53
53
53

CHAPTER I

STATEMENT OF THE PROBLEM

Introduction

 

The research reported in this thesis has as its
purpose to determine the relationship between the Lombard
effect and vowel intensity with respect to Specific questions.
The presentation of material includes first; the
statement of the problem which includes a hypothesis, a
definition of terms and Specific questions to be answered
about the Lombard effect as it is related to vowel intensity.
This is followed by the Justification of the thesis tOpic
through a review of pertinent literature in the area of
vowel intensity studies and Lombard effect studies. The
procedures for the study are reported; including (1) the
test facilities, equipment, and materials, (2) the description
of subjects, and (3) the procedures used in conducting
the experiment. A statistical anlaysis of data collected
and an interpretation of the data constitute the results
of the research project. The Significance of the data is
discussed in the summary and conclusions.
It is h0ped the results of this research study may
‘contribute information which will stimulate further studies

1

of (1) the Lombard effect, (2) vowel intensity, and (3)

communication in noise.

Hypothesis

 

It was noted some years ago that a Speaker with
normal hearing will unconsciously adjust the level of
his own voice to maintain it over the surrounding noise
level. This phenomenon is known as the "voice reflex" or
the "Lombard effect" and was the basis for the develop-
ment of the Lombard Voice Reflex Test for the detection
of malingering in the testing of hearing.

This phenomenon, the Lombard effect, has been in-
corporated in the Lombard Test and in addition has stimulated
research in the area of (1) tests for malingering and (2)
communication in noise. The pertinent literature has been
reviewed and will be presented in Chapter II of this thesis.

The literature examined by the writer motivated
prOposal of the questions to be answered within this
research study.

In the research involving the Lombard effect the
studies have dealt primarily with (l) intelligibility,

(2) duration, (3) rate, (4) frequency, and (5) intensity
with words and continuous speech in noise or communication
in noise. This study pr0poses to place emphasis on the
measurement of vowel intensity as related to the Lombard

effect. The intensity of the vowel sound in isolation

will be measured in contrast to numerous studies in which
vowel intensity has been measured with consonantal
influence present. The Specific vowel sounds to be used-
for measurement of intensity are eleven American vowels.
This particular study pr0poses to answer the following
questions:
1. How intense is the voice reflex for each

isolated vowel sound when Specific levels

 

of white noise (masking) are presented monaurally?
2. How intense is the voice reflex for each
isolated vowel sound when Specific levels

of white noise (masking) are presented binaurally?

 

3. Is there a significant difference in the
intensity of the voice reflex on vowel sounds
between the binaural and monaural masking?

4. Is there a systematic intensity pattern that
characterizes the eleven spoken vowel sounds?

These four questions define the problems to be considered

in the present research study.

Distinctiveness of the Study

 

It is evident from the review of literature presented
in Chapter II that there have been numerous experimental
studies concerned with the measurement of vowel intensity
as well as studies utilizing the voice reflex or Lombard
effect. In all the vowel intensity studies reviewed, the

emphasis has been the vowel in a consonant setting and

therefore consonantal influence has been present. In the
research involving the Lombard effect the studies have not
been limited to intensity of an isolated sound but have
dealt in terms of duration, rate, frequency, and intensity
with words and continuous Speech in noise or communication
in noise. The distinctiveness then of this research
project is:

l. the measurement of isolated vowel intensity
without consonantal influence and in an
environment of noise, and

2. a study of the Lombard effect measuring the
intensity of isolated vowel sounds.

The results of this experimental study may contribute

significant information in the area of (1) measurement of
vowel intensities, (2) the Lombard effect, and (3) communi-

cation in noise.

Definition of Terms

 

It is desirable at this point to Specifically define
the terms Lombard effect and the Lombard Voice Reflex Test.
Newby states:

This test is based on the fact that we monitor our

own voices through the sensation of hearing. If

we are Speaking in a noisy environment, we unconsciously
increase the intensity of our voice to compensate for
the masking effect of the noise.

 

lHayes A. Newby, Audiology Principles and Practice
(New York: Appleton-Century-Crofts, Inc., 1958), pp. 156-
157.

Korn defines as ”psychological feedback" the same phenomenon.
It is evident that every person will tend to
adjust the level of his own voice to maintain
it over the ambient noise level.
Taylor says:
That one unwittingly increases the loudneSs of his
voice when he is unable to hear it because of deafness
or because of the masking effect of the environment 1
noise is the principle on which this test is based.
These definitions present the basis on which the Lombard
Voice Reflex Test was developed.

Grove relates effectively the causes of the Simulation
of deafness or malingering that create the need for audio-
metric evaluation procedures such as the Lombard Test.

When the simulation of deafness is a conscious act it is
usually an attempt to defraud. However, the Simulation

of deafness may also be an unconscious act at which time it
is indicative of an abnormal state of mind. The otologist
is responsible for detecting malingering whether it is
conscious or unconscious. Conscious simulation or
malingering is found most frequently in times of economic
stress and war. It may be used to avoid military

responsibilities or ”to extract a pecuniary award to which

the perpetrator is not entitled."3 The feigning of

 

1T. s. Korn, ”Effect of Psychological Feedback on
Conversational Noise Reduction in Rooms," Journal Accoustical
Society of America, XXVI (September, 1954): 793.

2Glenn J. Taylor, "An Experimental Study of Tests for
the Detection of Auditory Malingering ” Journal of Speech
and Hearing Disorders, XIV (March, 1949), 119,130.

3

 

W. E. Grove, "Simulation of Deafness," Annals of

 

deafness may be unilateral or bilateral and it may be
either partial or complete. Bilateral simulation is
fortunately more uncommon as it is very difficult to
detect because the patient has prepared himself to be on
guard constantly.
Grove continues with a description of the process
of administering the Lombard Voice Reflex Test.
A Barany noise ap aratus is alternately sounded in
the patients "bad' and good ears so that he becomes
acquainted with its characteristics and will not be
easily startled thereby.
The patient is given material to read and instructed to read
orally and to continue no matter what happens. The noise
is introduced to the good ear while the patient is reading.
In the case of a patient whose alternate ear is deaf he will
"automatically raise the intensity of his voice or even
shout as he continues to read."1 However, the malinger
maintains an even tone or in some cases a slightly raised

2 The basic assumption here is that the

tone will occur.
malinger is capable of monitoring the loudness of his

voice with his unmasked "bad” ear.

 

Otology, Rhinology, and Laryngology, LII (September, 1943),
573-580-

lIbid.

Ibid.

“

 

2

Newby gives a different description of the Lombard

Voice Reflex Test and places a different significance on
the results. The process of evaluation in the Lombard Test
is based on the patient reading material while masking is
presented into earphones which he wears. The increases
and decreases in the patients voice are noted by the
tester as the level of the masking noise is fluctuated.

The result of the test is positive if the patients

voice does become more intense when the masking

is increased.
The tester notes at what level of masking the patient's
voice becomes more intense and this is compared with the
degree of hearing loss the patient is supposed to have.
When the patient‘s voice is affected by the level of masking
being less than the particular degree of the supposed
hearing loss, it is an indication that the patient is
really hearing at lower levels than those to which he
has admitted.

The result of the Lombard Test is negative if the

patient's voice remains at the same intensity

regardless of the fluctuations in the level of the

masking noise within the limits of the supposed

hearing loss.
The Lombard Test is used predominantly in determining
bilateral functional loss, however, it can be used monaurally.
This is accomplished by presenting a constant masking
noise in the good ear as a variable noise is placed in the

l
ear that is supposed to be impaired.

l
Newby, loc. cit.

 

Heller, et al., resolve the conflict in this manner.

While the patient reads aloud, masking is first
introduced into the "deaf" ear. If there is an
increase in voice volume, we give credence to the
thought that the ear does hear. Even if the
voice does not change, we then mask the better
ear. If there is an increase in the voice volume,
we suspect that the "deafened" ear is impaired.
If there is no change in the voice volume when
the good ear is masked, we suspect that the
alledgedly deafened ear is permitting the patient
to monitor his own Speech volume.

Despite the variation in interpretation and method of
administration of the Lombard Voice Reflex Test the

basis for the test, as previously defined still remains

the same.

 

lMorriS F. Heller, M. Anderman, and Ellis E. Singer,
Functional Otology: The Practice of Audiology (New York:
Springer Publishing Co., Inc., 1955), p. 90.

CHAPTER II

REVIEW OF LITERATURE

Due to the nature of this research study, literature
related to vowel intensity studies will first be reviewed

followed by the presentation of the Lombard effect studies.

Vowel Intensity Studies

 

Vowel intensity has been the subject of previous
research studies, but not in terms of measurement in an
environment of noise.

In 1926, Sacia and Beck published the results of
a study in which 16 subjects produced eleven monosyllabic
words. The words used were: team, tool, took, tone,
talk, t0p, tap, ten, tape, tip, ton. The initial consonant
[t]remained constant but there were five different final
consonants. In the analysis, sounds were considered
individually on the basis of instantaneous and mean power
and were represented in terms of microwatts. The vowel
sounds in particular were compared with regard to conver—
sation and normal values. There was also a comparison of

peak and mean values.

 

l
C. F. Sacia, and C. J. Beck, "The Power of Funda-
mental Speech Sounds,” Bell System Technical Journal, V

(July, 1926) 393—403.

 

10

Black, in 1949, reports the results of a research
project which sought to determine intensity, duration, and
frequency from the same sample of vowels and investigate
possible interrelationships among the three characteristics
by the use of analysis of variance. In the experiment
forty-two males read eleven words into a micr0phone: [-bfP],
[tarp], Item. [tsp]. [1:80P]. [top], [top], [twp]. [top].
[tljp], FtLjp]. The subjects each (1) practice the words under
direction, (2) produced the words while in a standing
position in a sound treated room; the distance between the
lips and the micr0phone was standardized at 10 inches, and
(3) read the words at five-second intervals. There were
four sets of measurements made; one each for intensity and
frequency and two for duration. They were obtained in the
following manner.

(1)The peak vertical displacement of the graphic

recorder stylus was used as an indication of the

peak intensity of the vowel in the word. (2) The
duration of the vowel was also calculated from the

record. . . The length or horizontal dimension of
the tracing of the deflected stylus was assumed to
indicate the duration of the word. . . (3) The

magnetic prOpertieS of tape recordings were utilized

in determining the frequency of the vowels. Each

sound was Spotted on the tape, and the portion of

the tape that contained the word was immersed in

iron filings. The filings arranged themselves in
striations, crosswise of the tape, sufficiently

plainly to be counted. The speed of the tape through
the recordin head was determined. The total striations
(sound waves in the vowel could be counted and the
distance they occupied on the tape measured. Frequency
was computed from these measures. (4) The length of

magnetized portion of the tape was measured
for another determination of duration.

11

No quantitative conclusions were made with regard to
the interrelationships among frequency, duration, and
intensity. However, independent analysis of variance
demonstrated significant differences among some vowels in
their natural frequency, duration and intensity. It was
determined that frequency of the vowel varied with the
degree of Openness of the vowel. Also, duration varied
directly with the Openness of the vowel. In addition,
significant differences were found among the over-all
intensities of vowels.l
In an experimental study of vowel intensities,
Fairbanks, House, and Stevens were concerned with two
specific questions:
Are there significant differences between the
intensities of vowels that may be attributed to the
vowels as such? Does a given vowel vary Significantly
in intensity in relation to its consonantal environ-
ment?

The subjects usedby the researchers were 10 young adult

male college students. The materials used were 110 words,
10 for each of the eleven vowels. The vowels were: [ i],
[I]. [C]. [5], [cl-3!. [A], [O]. [31.[O], [ULHM-
All the words were meaningful consonant-vowel-consonant

monosyllables. Eight different voiceless consonants were

 

1John W. Black, "Natural Frequency, Duration, and
Intensity of Vowels in Reading," Journal of Speech and
Hearing Disorder, XIV (September, 1949), 216-221.

 

12

used. Each of the subjects Spoke 110 words while positioned
10 inches from a micr0phone. The recording.apparatus was
arranged conventionally in two rooms and recording was
direct. The graphic record of a word demonstrated a peak-
like curve with a readily identifiable maximum. A Single
value was obtained for each word. (The level of the
maximum in db above a common arbitrary reference.)

The results of the statistical analysis of the
collected data indicated (1) that most of the common
American vowels, when Spoken in isolated words, are signifi-
cantly different in mean intensity and (2) when the same
vowel is Spoken in different isolated words its intensity
may vary significantly in part due to consonantal
environment. According to the researchers the differences
in mean intensities and their relationship to physiological
differences between the vowels merits further investigation.1

Steven and House outline a contemporary acoustical
theory of vowel production in a 1961 research publication.
The relative intensities of vowels are considered to a

limited extent. The authors report that it has long been

recognized that different vowels "generated" with the same

 

lG. Fairbanks, A. S. House, and L. Steven, "An
Experimental Study of Vowel Intensities," Journal Acoustical
Society of America, XXII (July, 1950), 457-459.

l3

vocal effort have different over-all levels. The range of
over-all levels for the common vowels of American English
is approximately 4 to 5 db, with[ i],and[LJ],having the

lowest levels andIai], [C2], and[.3] the highest levels.1

Lombard Effect Studies

 

Taylor studied the effectiveness of certain principles
on which numerous tests of malingering have been based.
In addition, the study was designed to evaluate if results
of the tests could be reported in objective terms that
would adequately describe the qualitative and quantitative
aspects of a supposed unilateral loss. The Lombard Test
was used as one of the tests because it has appeared
throughout the literature and has been used extensively.
The subjects used in the experiment were 32 normal hearing
persons, 13 hard—of—hearing persons, and 12 normal hearing
graduate students in Speech correction. The author requested
that the s0phisticated group of Speech therapists "attempt
to beat the test." Taylor concluded that the Lombard Test
was useful for the detection of malingering in unSOphisti-
cated subjects, but was invalidated by trained subjects.
Also it was determined that it is not possible with the
Lombard Test to make qualitative or quantitative evaluations

of hearing in a supposed deaf ear.

 

1Kenneth N. Stevens and Arthur 8. House, "An Acousti-
cal Theory of Vowel Production and Some of its Implications,"
Journal of Speech and Hearing Research, IV (December, 1961),
303-320.

2Taylor, 100. cit.

 

14

Communication in noise is a problem which in the past
has been most prevalent in war time. Several pertinent
research studies follow which have as their purpose effective
communication in noise.

In 1946, a Series of articles which were the result
of war-time research related to Speech training were published

in Speech Monographs with John W. Black as editor of the

 

particular series. In 1943, the army requested that
investigations be undertaken for the purpose Of providing a
training program for increasing the effectiveness of inter-
phone and radio communication concerned with aircraft. The
aim of the training program was functional. That is, it
had as its purpose to improve voice efficiency in order

to get messages through to a listener. This aspect of
Speech was studied under the limiting circumstances of high
level noise, Such as that characteristic of the interior of
military airplanes, and with the voice in rise over army
intercommunication and radio sets. The majority of this
work was done at the Voice Communication Laboratory, Waco
Army Field, Texas. Service personnel were used as
experimental subjects. The work continued for two years,
until mid 1945. One of the most pertinent sections of

the series entitled "Intelligibility Related to Loudness"

indicated that for Operating both aircraft radios and

l5

interphones, loudness of voice is an important factor
and the voice should be loud--just under shouting.

Henley and Steer conducted a research study in 1949,
which was designed to measure the effects of masking upon
(1) speaking rate, (2) mean syllable duration, and (3)
mean Speech intensity level. Forty-eight students were
directed to read a seventy-three-word passage into a
micrOphone while airplane type noise was fed into the
headsets at predetermined levels. In measuring mean
intensity level the tape was used in a High Speed Power
Level Recorder. The tape was divided into 5 db line
markers. Those peak values were summed, then divided by
the number of peaks to give a mean peak value for each
reading for each subject. The following conclusions were
made. Subjects who have not been trained to communicate
in noise are confronted with a communication problem in the
presence of noise. They appear to react in what has been
termed a desirable manner. That is "subjects reduce rate of
speaking, prolong syllables, and speak with greater inten-
sity as noise increases.” It is interesting to note in the
results that there is a stronger response given to an
increase in a noise barrier at lower levels. The reverse

occurs in the upper levels. The "response" becomes

 

1John W. Black (ed.),"Studies in Speech Intelligi-
bility,” Speech Monographs, XIII (Number 2, 1946), 1-64.

 

16

less intense as the subject's limits of reSponse
potentialities are approached.1

Korn studied the effect of room-conversation-type
noise upon the vocal output of those Speaking in competi-
tion with it. The author presented white noise to the
Speaking-listening environment of 50 subjects. The noise
level was increased in steps of 10 db,from 40 db to 90
db, and then decreased in the same manner.

It was demonstrated that low noise levels (below 45
db or so) do not seriously influence speech power.
For higher noise levels, however, over 55 db, the
slope of.k 0f the Speech-versus-noise curve becomes
steeper and can be estimated as constant and equal
to about 0.38 db/db.

That is to say, that with the increase of one
decibel in the room noise there was an increase of 0.38
decibels in the Speech level. This occurred at levels
above 55 decibels.2

Dreher and O'Neill hypothesized that possibly Speech
produced in noisy surroundings might include compensatory
factors, which within a certain range might be more
intelligible for a listener under poor reception conditions

to understand.

therefore, the plan of the study was to get
the reaction of several speakers under several

 

1T. D. Hanley and M. Steer, "Effect of Level of
Distracting Noise upon Speaking Rate, Duration, and Intensity,‘

lgnIngé__i_Spee_h_and_Eearias__isgrde£§. XIV (December, 1949),
3 3 3

Korn, 10c. cit.

17

levels of masking at the ear, comparing intelligibility
of the Speech s0 produced against a constant back-
ground Of masking noise with a large number of
listeners.
The authors tested the hypothesis in the following manner.
Fifteen subjects were employed to read five sets Of Spondee
words and five air traffic control sentences. This was
first done with dead phones and then with 70, 80, 90 and 100
decibels of white noise fed into their headsets. Later an
overlay of wideband random noise was recorded on the test
tapes, at 4 decibels lower than the peak values of the
speech signals. These recordings were then played to a
listener panel and intelligibility scores were computed
for words and sentences. It was determined through analysis
that the level of 70 decibels allowed for 35 per cent
better word intelligibility scores and 27 per cent better
sentence intelligibility scores than the recordings done
at the ''quiet" level. It was found that there was not an
increase in intelligibility at 80 and 100 decibels (90 db
was not available due to an error in administration).
Although intensity of the voice did not contribute
to the intelligibility results, the original recordings
could be measured to estimate the effect Of speaker-
masking on his production. . .The largest intensity
increase comes, both for words and sentences, when the
first level of masking is employed. After that,
succeeding increments of 10 db masking produced an
average vocal increase Of approximately 1 db.

The conclusions of Dreher's and O'Neill‘s research

study of Speaker intelligibility in noise are as follows:

18

(1) There is a measurable and important increase in

the intelligibility of both words and sentences pro-
duced by Speakers operating with a broadband random
noise mask in their headsets. (2) The intelligibility
values Of Speech produced under 70 db of masking were
statistically equivalent to those produced under 80

and 100 db of masking. (3) Plots of both mean dur-
ation of Speech elements and vocal output Show a

re ular increase with increase in Speaker masking.

(4 The range of masking between "Quiet” and 70 db

of masking needs further exploration to determine
additional points on both the intelligibility and

vocal output curves, and it is suggested that live

voice tests for hearing evaluation consider the use

Of sufficient Speaker masking to effect the vocal
stability observed in this study. (5) It is recommended
that the observation of increased intelligibility be

put to Operational test in aircraft circuits. (6) It

is recommended that the aforementioned test, if success—
ful, should result in the engineering of a device for
emergency use in air-ground communications for Speakers
in relatively quiet Speaking environments.1

The most recent study, to the author's knowledge, on
the Lombard effect was done by Waldron in 1960. The purpose
of the study was to obtain the answers to seven questions.
These questions are as follows:

(1) Will the unsuspecting naive subject demonstrate

a voice reflex when one of his two normal ears is
masked? (2) Will the SOphisticated examinee be able
to control his voice reflex when stimulated monaurally?
(3) Is there a marked difference between the extent
of voice reflex resulting from monaural noise presen-
tation and that resulting from binaural presentation?
(4) Can the SOphisticated subject control his voice
reflex when noise is introduced into both ears simul-
taneously? (5) Will gradual or instantaneous presen—
tation of the masking noise result in the greater
voice reflex for naive subjects? (6) Which one of the

 

1John J. Dreher and John J. O'Neill,"Effect of '
Ambient Noise on Speaker Intelligibility of Words and Phrases,
Laryngoscope, LXVIII, Part I (1958), 539—548.

 

19

two modes of presentation will bring about a greater
reflex by the SOphisticated test subjects? (7) Will
any of the test conditions yield a significant oral
reading rate change?
In order to answer these Specific questions equipment was
used which:

1. gradually introduced a complex masking noise
into one or both ears,

2. allowed by-passing the gradual introduction
mechanism so that noise could be presented
instantaneously,

3. provided a pick-up micrOphone for the reader's
voice,

4. traced the relative voice intensity levels, and

5. allowed measurement Of the words-per-minute
reading rate for each subject.

There were four groups of twenty subjects each. Group I knew
nothing about the test and had the noise presented gradually.
Group II were informed of the principles of the test and
requested to attempt to "beat it,” by controlling the
tendency to change reading rate or vocal intensity. Group

I also received a gradual presentation. Group III had the
masking noise presented instantaneously and did not know of
the purpose of the test. Group IV, as Group II, were told

to "beat the test” and received the noise instantaneously.

The intensity changes and words-per-minute rates were

recorded under both monaural and binaural masking conditions.

20

The results of the research study were as follows:

1. Monaural stimulation elicited statistically
Significant changes in vocal intensity for all
groups.

2. The SOphisticated subject groups recorded
vocal reflex scores which were significantly
smaller than those of the uniformed groups
under the binaural presentations; the monaural
stimulation scores of the uninformed and
SOphisticated groups did not differ significantly.

3. The binaural stimulation resulted in a signifi-
cantly greater voice reflex than the monaural
stimulation, for all groups.

4. The rate of presentation of the masking noise
did not seem to have any affect on the extent

or control of the voice reflex.

U‘l

None of the test conditions resulted in
significant changes in the oral reading rates

of the subjects:L

Summary
It is evident from the review of literature that there

have been numberous studies conducted on the measurement

 

1D. L. Waldron, ”The Lombard Voice Reflex Test: An
Experimental Study” (unpublished Doctoral dissertation,
Department of Speech, Stanford University, 1960).

21

of vowel intensity, Sacia and Beck, Black, Fairbanks, and
Stevens and House; as well as studies utilizing the voice
reflex or Lombard effect by Taylor, Black, Hanley and Steer,
Korn, Dreher and O'Neill, and Waldron. In all the vowel
intensity studies reviewed the emphasis has been the vowel
in a consonant setting and therefore consonantal influence
has been present. The intensity of isolated vowel sounds
has not been studied. Also, there are not any studies of
vowel intensity in an environment of noise. In the
research involving the Lombard effect the studies have not
been limited to intensity of an isolated sound but have
dealt in terms of duration, rate, frequency, and intensity
with words and continuous Speech in noise or communication
in noise. The uniqueness then of the present research
project is:

l. the measurement of isolated vowel intensity
without consonantal influence and in an
environment of noise, and

2. a study of the Lombard effect as measured by

the intensity of isolated vowel sounds.

CHAPTER III

SUBJECTS, EQUIPMENT, AND PROCEDURES

Introduction

 

The experiment described in this chapter was
accomplished in the sound-treated test room in the Speech
and Hearing Science Area of the Department of Speech,

Michigan State University, East Lansing, Michigan.

Description of the Subjects

 

The twelve subjects who participated in this research
study were graduate students in Speech and Hearing Science.
The subjects were selected on the basis of an expressed
interest in research and their availability. The age range
of the subjects was 21 years to 46 years, with a mean age
of 26.5 years. All subjects had received hearing evalua-
tions and were considered to have normal hearing. None of

the subjects knew the purpose of the experiment.

Equipment

 

The Specific equipment employed was:
Audiometer (Allison Audiometer Model 20)
Headsets (Telephonics TDH-39)

MicrOphone (Electro-voice 654)

FUUIDH

Tape recorder (Ampex Model 601)

22

23

5. Magnetic recording tape (Scotch Tenzar Backing
Magnetic Tape 311)

6. Sound level recorder (Gruel and Kjaer Model 2305)

7. Electric timer modified to time at 3 seconds
(GraLab Microtimer).

The sounds employed in the study were eleven common
American vowels. Included were: [ I ], [I 1: [C], [5]:
[:38], [LI],[/\], [Q],[O], [O], [L‘]. The vowels were
printed phonetically in black on individual white 5 x 8
cards. A one-syllable word was printed beneath each
vowel as a clue for those less SOphisticated in phonetics
(See Appendix).

It has been domonstrated that low noiselmasking levels
(below 45 db) do not seriously influence speech power.
Therefore, the Specific decibel levels of white noise
analyzed in this study were 70, 80, 90, and 100 db ( re:
0.0002 microbar).

A form was developed for recording the collected
data. The form provided three charts: (1) binaural
presentation, (2) monaural presentation (right), and (3)
monaural presentation (left). Each chart allowed for
recording the intensity of each of the eleven vowels
Spoken by subjects, at each of the given intensities of

white noise (see Appendix).

 

lKorn, loc. cit.

24

Procedures for Research

 

The subjects were seated at a table in the sound-
proof testing room four feet from the two way window which
is located in the wall between the soundproof testing room
and the test equipment room. The headset (Telephonics TDH--
39) was placed On the subject's ears and he was positioned
with his lips seven inches from the micrOphone (Electro-
voice 654). The micrOphone was connected to the tape
recorder (Ampex Model 601). The tape recorder was calibrated
prior to the testing of subjects. Also located in the test
equipment room was the Allison Model 20 Audiometer. An
assistant Operated the controls Of the Allison Audiometer
unit while the researcher located in the test equipment
room also, presented the vowel cards to the subjects. The
vowel cards along with the microtimer (GraLab) appeared
in the glass window for viewing by the subjects. The
microtimer was modified to time electrically at 3 seconds,
with the red light remaining on for 3 seconds and Off for
three seconds. The vowel sound was produced for the length
of time the light remained on, which was 3 seconds.

All subjects were tested individually. The subject
was first seated at the table, four feet from the two-way
window in the soundproof testing room and given the follow-
ing instructions.

(1) You are to produce the designated vowel sound

for the length Of time the red light remains on.

The red light will remain on for 3 seconds and then

be off for 3 seconds. The vowel card will appear in
the window next to the light. You will have the

25

Opportunity to view the next vowel sound for 3 seconds
before the light goes on and you are to produce the
sound. As you are producing these vowel sounds
noise will be fed into your ears. (2) We will review
all eleven vowel cards before the actual experi-
mentation rocedures are begun. (Vowel cards then
practiced. (3) The first thing you are to do
upon receiving a signal from the control room is to
give your subject number and name.
Immediately following these instructions the earphones
were placed on the subject and his lips placed seven
inches from the micrOphone. The researcher then re-
turned to the test equipment room and began the
testing procedure with the aid of the assistant.

Each subject produced each of the eleven vowels,
twelve times, at which time they were recorded by a
tape recorder (Ampex Model 601). During this procedure
white noise was being presented to the ears at 70, 8O
90, and 100 db (re: 0.0002 microbar).

Four subjects received the different levels of
white noise (1) monaurally-right ear, (2) binaurally,
(3) monaurally-left ear; four more received the noise

(1) monaurally—left ear, (2) monaurally—right ear, (3)
binaurally; and four more received the noise (l)binaurally
(2) monaurally-left ear, (3) monaurally-right ear. The
order of vowel presentation for each subject was randomized
by the use of a random numbers table prior to the experiment
time. For example, Subject 1 produced the series of eleven

vowels first with the white noise in his right ear and

26

with an intensity of 70, 80, 90, and 100 db ( re: 0.0002
microbar). Then he produced the series with the white
noise being fed into his ears binaurally with intensity of
70, 80, 90, and 100 db (re: 0.0002 microbar). Finally,.
he produced the vowel series with white noise in his
left ear at the intensity of 70, 80, 90, and 100 db
(re:0.0002 microbar). Each time the vowels were
presented in a different random order. This meant 12
different orders for 12 subjects with a total of 144
different random orders for the presentation of the vowel
sounds.

As the final step in the procedure of the experiment
the tapes recorded were employed with the Bruel and Kjaer
Model 2305 Sound Level Recorder to obtain the vowel

intensity readings in decibels.

CHAPTER IV
RESULTS AND DISCUSSION

Results

The test results were tabulated and tested
statistically by use of (l) a two-way analysis of variance
(2) a one-way analysis of variance, and (3) t-tests in
cases where the F from the analysis of variance was signifi-
cant. It was the purpose of the researcher to analyze the
change in vocal intensity of the isolated vowel sounds.

Variation of vowel intensities as a result of degree
of masking and mode of presentation.--Two—way analysis of
variance was employed to determine if there was significant
variance in the isolated vowel intensities as a result of
masking and mode of presentation. The computational
formulas used in the three two-way analyses of variance
are reported in Table 1.

The results Of the analyses of vowel intensities as
a result of masking and mode of presentation are presented
in Tables 2, 3, and 4.

The estimated values of F (v needed for signifi-

1 V2)
cance were made from Table A—7C, "Percentiles Of the F

(v1 v2) Distributions," Dixon and Massey.1

 

1

Wilfrid J. Dixon and Frank J. Massey, Jr., Intro-
duction to Statistical Analysis (New York: McGraw—Hill Book
Company, Inc., 1957), 390—403.

27

TABLE 1

COMPUTATIONAL FORMULAS FOR TWO WAY-ANALYSIS OF VARIANCE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sum Of Mean Square
Source Squares df is estimate of
2 2
Row Means E‘ T-J- " T... r-l = 3 02 + cno'2
132 528 r
Column MeanSEELTi--2 ‘ T°'°2 c-l = 10 02 +rno§
48 528
Interaction Subtotal - Rows
R x C and Comumns (c-l)(r-l) 02 +n02
Sum Of Squares = 30 I
2
Subtotal :ég'TiJ- ' T---2 rc-l = 43
12 528
Withi Total S. S. - rc (n-l) = 02
n Subtotal 8. s. 484
rjk ijk 528
TABLE 2
BINAURAL MODE OF PRESENTATION
TWO WAY ANALYSIS OF VARIANCE
Source Sum of Squares df M. S. F
White noise
effect 449.09 3 1A9,69 9.47*%
Vowel effect 1420.85 10 142.09 8.99**
N x V Interaction 3712.54 30 123.75 7.83**
Subtotal 4304.48 43
Within 7652.58 484 15.81
Total 11057.06 527
*Significant at the .05 level of significance.
**Significant at the .01 level of significance.

29

TABLE 3

MONAURAL (RIGHT) MODE OF PRESENTATION
------- -TWO WAY ANALYSIS-OF VARIANCE~ ._. --~----

 

Source. Sum of Squares df M. S. F

 

White noise

 

 

effect 235.6 3 78.5 5.89**
Vowel effect 402.1 10 40.2 3.02*%
N x V Interac- -
tion 658.6 30 21.9 1.64

Subtotal 1296.3 43

Within 6446.7 484 13.32

Total 7743.0 527

 

*Significant at the .05 level of Significance.
**Significant at the .01 level of Significance.

TABLE 4

MONAURAL (LEFT) MODE OF PRESENTATION
TWO WAY ANALYSIS OF VARIANCE

W

 

 

 

Source Sum of Squares df M. S. F

 

White noise

 

 

effect 22.2 3 7.7 .720
Vowel effect 667.7 10 66.8 6.24 **
N x V Inter- .
action 970.6 30 32.4 3.03 **
Subtotal 1660.5 43 ‘
Within 5195.7 484 10.73
Total 6856.2 527

 

*Significant at .05 level of significance.
**Significant at .01 level of significance.

The tables which follow represent the results Of
tftests which were used whenever the F tests in the three
two way analyses of variance and twelve one way analyses of
variance proved to be significant. These tests are
represented in matrix form. In the first matrix, for example,
the mean intensity for the vowel in a given row has been
subtracted from the mean intensity for the vowel in the
corresponding column. In order to determine whether or not
these differences are significant, the following computa-
tions were made:

—' ——”—-*-—' whenever the variances

 

1 1
N1 + N2 O'and 02 are equal.

The variances must be assumed to be equal, since this same
assumption is made in the analysis of variance. The best
estimate of this variance, furthermore, is the within
variance is shown in the particular analyses of variance
table under consideration. This variance as actually the
pooled estimate for all rows and columns. By algebraic

manipulation the formula for 3 becomes:

£=X ~X2

 

’\/—82

.oocmOAMchHm mo Ho>oa Ho. 96 pcmodmacmfim

**

 

 

 

 

 

 

 

 

 

 

 

 

.oocmOAMcham mo Ho>oa mo. um pamOfimacwHw *
a~.6u u-
~6.H- 6m.on u-
*
ewes- lhm.a-.l~m.6- ..
*
mm.au mm.6- s:.6 as.6 ..
.6m.m mm.m mm.m. 6H.: 6:.6, ..
** ** ** ** **
.nm.6- w:.6 ::.H HS.H 66.H ms.m- ..
* **
66.6- 66.6 mo.H mm.H mm.6 6m.m- 6:.6- I- .1
**
63.6 SH.H mH.m 6s.m 66.H m~.H- m6.6 HH.H ..
** ** * .x J
6H.H Hw.a .s~.m :6.mw mm.m ma.H- mm.a .ns.a :6. --
.x. ** ** ** * J
ss.m m:.: ::.m HS.mlw 66.m mm.H 66.: m:.: Hm.m S6.m .:
** ** ** ** ** ** ** ** **
HHHL HAVL HAUL H OH H <L m.a_ .htL HHwL HHOL HRH“ a _L

m mamas

Amzzsqoov mzamz so mozmmmmmHm
oneaezmmmmm so who: hamsasz

H.3H
HAUL
an,“
a UL
H <L
His.
amt;
H.w_
HMUL
Hung
H .L

32

 

 

 

 

 

 

 

 

 

 

 

 

 

.ooEdoHeHsmam so Ho>oH H6. one an osnoasacwam as
.oostcasacman no Ho>oH 66. one on Samoasacmam *

sa.a- --

mm.6- 6H.6 - --

HS.6- .ms.6 smwo -

66.6- 66.6 ms.6 HH.6- -

.M~.6 Hm.H ms.H 6H.H .mm.H -

66.6 1mswa 66¢H 66.H SH.H NH.6- -

66.6 6st wme ms.6 :6.6 6:.6- mm.6- -

6m.6 mm.H Mm.H mmuo .m.6 .mm.6- sm.6- 66.6 -

ms.6 sme 66.H SH.H mm.H 36.6- 66.6 H:.6 6m.6 -

Sc.m [,Hmmm1. 66mm 6:.6 66.6 6m.H 66.H 6S.H m6.H am.H -

ms ** *s ** As * a *

HHHL Haw. Hn.H HMO“ H <H HHDL Ham; HHwL Hme H H; H 1L

Hmzzsqoov 62262 m6 mozmmmmmHm

oneaezmmmmm so 6662 HemuHmv qgmpazoz

. m mam/5H.

HHHH
Ham.
Hn,H
HHUL
H/\_
H.LH
Hot“
H.mw
H“6H
H HH

1.:

33

.oocmOHmacwam mo Ho>oa Ho. one 66 undefimficwam **
.ooeeoa6acman 6o Ho>oH 66. 626 66 62662666626 *

 

 

 

 

 

 

 

 

 

 

- H2:
16646. - , , . HnVH
66.6- 62.6- - Hn,H
66.6- x66.6 62.6 - HHOL
66.6- 66.6 66.6 62.6 - . H,\H
-6w.2 22.2 66.6 .66.6 62.6 - HHJL
Ho-H 62.6 26.6 62.6 22.6 66.6- - H662
26.6 62.6 26.6 62.6 26.6 .66.H- 66.6- - HHmL
:6 66.2. 2.2.2 66.2 66.6 66.6- 666- 66.6 - H02
66.6 -Mm.H 66.2 66.2 66.6 62.6- 62.6- 26.6 6H.6 - HHHH

 

Od.m $21.20 OMAN mm.m ma .M m-MJ mo.m mm.m mm.m Hm.m :-

r--1
‘Q.
L——J

 

HHHH HAVH Hn.H _HGH_ H <4 H.6H H662 Hnwg Hmug HHHH H .2

H62266666 62262 66 6626666666
266629266666 66 6662 H6666v 62262262

N. 393.

34

TABLE 8

BINAURAL MODE 0F PRESENTATION
DIFFERENCE OF MEANS (Rows)

 

 

 

 

 

 

 

70 80 90 100
70 —- 0.33 0.58 1.18*
80 -- 0.25 0.60
90 -— No.60
100 --
TABLE 9

MONAURAL (RIGHT) MODE OF PRESENTATION
DIFFERENCES OF MEANS (ROWS)

 

 

 

 

 

7O 80 90 100
70 -— 0.50 1.22** 1.51**
80 -- 0.72 1.01*
90 -- 0.29

 

100 _-

 

*Significant at the .05 level Of significance.
**Significant at the .01 level of significance.

35

Variation among intenSity levels 0f the eleven -
common American vowels asfia result of mode 0f presentation.
One way analyses of variance were used to determine if there
was a significant variance of intensity within reSponses of
subjects to each vowel sound as a result of mode of pres-
entation. When the F's were significant the E-test, as
described previously, was used. The computational for-

mulas used are presented in Table 10.

TABLE 10
COMPUTATIONAL FORMULA FOR ONE WAY ANALYSIS OF VARIANCE

fiean Square is

 

 

 

 

Source Sum of Squares df Estimate of:
Means 2 Ti2 — T..2 K - l 02 + no?n
j ni N

w th x 2 T 2 2
1 in ff:— ij '2 1° N - K 0’

J “1
Total 2.2x 2 - T 2

iJ 1;] .._._°_'__. N - l
N

 

The results of these analyses and Eftests are reported

in Tables 11-31.

36

TABLE 11

MODES 0F PRESENTATION FOR VOWEL [i ]
ONE WAY ANALYSIS OF VARIANCE

m

Source Sum Of Squares df M. S. F

 

White noise

 

 

 

effect 234.37 2 117.19 0,944
Within 117495.37 141 124.09
Total 17729.75 143

TABLE 12

MODES OF PRESENTATION FOR VOWEL [IE]
ONE WAY ANALYSIS OF VARIANCE

 

 

 

 

Source Sum Of Squares df M. S. F
White noise

effect 321.78 2 160.89 1.2137
Within 18690.86 141 132.56
Total 19012.64 143

 

*Significant at .05 level of Significance.
**Significant at .01 level of significance.

37

TABLE 13

MODES OF PRESENTATION FOR VOWEL [6]
ONE WAY ANAIYSIS OF VARIANCE

W

Source Sum of Squares df M. S. F

 

White noise

 

 

 

 

 

 

 

effect 389.77 2 194.88 21.81**
Within 128.79 141 v8.94 ' '
Total 1676.56 143

TABLE 14

MODES OF PRESENTATION FOR VOWEL [EL]
thEST
Binaural Monaural (R) Monaural (L)

Binaural -- -2.89 -3.87*
Monaural (R) -- - [98-

 

Monaural (L) --

 

*Significant at the .05 level of significance.
**Significant at the .01 level of Significance.

38

TABLE 15

MODES OF PRESENTATION FOR VOWEL [5,]
ONE WAY ANALYSIS OF VARIANCE

_—_.__-—_—_—__._—___———____w.—_—_———_—__ _ .._._ _-_ __

 
 

     

Source Sum of Squares df M. S. F

 

White noise

 

 

 

 

 

 

effect 547.91 2 213.95 19.3;::
Within _ 2001.03 141* 14.19
Total 2548.94 143
TABLE 16
MODES OF PRESENTATION FOR VOWEL [g ]
thEST
W
B M (R) M (L)
B -— -3.93** -4.13f*‘
M (R) —- -0.3
M (L) --

 

*Significant at the .05 level of significance.
**Significant at the .01 level of significance.

39

TABLE 17
MODES 0F PRESENTATION FOR VOWEL [23].
»ONE WAY ANALYSIS OF VARIANCE

 

 

 

Source Sum of Squares df M. S. F
White noise 634.68 2 317.34 18.76**
effect
Within 2384.87 141 16.92
Total 3019.55 143

TABLE 18

MODES OF PRESENTATION FOR VOWEL [a]
t—TEST

 

 

 

 

B M (R) M (L)
B __ 3.94** -4.84**
M (R) —~ —0.90
M (L) --

 

*Significant at the .05 level Of Significance.
**Significant at the .01 level Of significance.

40

TABLE 19

MODES OF PRESENTATION FOR VOWEL [U]
ONE WAY ANALYSIS OF VARIANCE

 

 

Source Sum Of Squares df M. S. F

 

 

White noise

 

 

 

 

 

 

effect 177.73 2 88.83 7.61**
Within 1645.58 141 11.67
Total 1823.31 143
TABLE 20
MODES OF PRESENTATION FOR VOWEL [t4]
thEST
B M (R) M (L)
B —- —1.59* -2.71
M (R) -— -1.12
M (L) -—

 

*Significant at the .05 level Of Significance.
**Significant at the .01 level of Significance.

41

TABLE 21

MODES OF PRESENTATION FOR VOWEL [A ]
ONE WAY ANALYSIS OF VARIANCE

m

 

Source Sum Of Squares df M. S. F».

 

White noise

 

 

 

 

 

 

effect 600.04 2 300.3 l3.72**
Within 3091.40 141 21.93
Total 3691.44 143
TABLE 22
MODES OF PRESENTATION FOR VOWEL [A ]
It—TEST
B M (R) M (L)
B —- —3.77** -4.72**
M (R) -- -O.95
M (L) --

 

*Significant at the .05 level Of significance.
**Significant at the .01 level of significance.

42

TABLE 23

MODES OF PRESENTATION FOR VOWEL [C1]
~-ONE WAY ANALYSIS OF VARIANCE

W

 

Source Sum Of Squares df M. S. F

 

White noise

 

 

 

 

 

effect 784.23 2 392.12 12.63**
Within 4377.10 141 31.04
Total 5161.33 143
TABLE 24
MODES OF PRESENTATION FOR VOWEL-[a ]
» thEST
B M (R) M (L)
B -— -4.58** —5.25**
M (R) —— —0.67
M (L) __

 

*Significant at the .05 level of Significance.
**Significant at the .01 level of significance.

43

TABLE 25

MODES OF PRESENTATION FOR VOWEL LJ.]
ONE WAY ANALYSIS OF VARIANCE
W

Source Sum Of Squares df M. S. F

 

White noise

 

 

 

 

 

effect 570.60 2 285.30 12.64**
Within 3181.62 141 22.57 .-
Total 3752.22 143
TABLE 26
MODES 0F PRESENTATION FOR VOWEL[33:]
. .._... .E-TEST
B M (R) M (L)
B -- -3.77** -4.56**
M (R) __ 30.74
M (L) "

 

*Significant at the .05 level of Significance.
**Significant at the .01 level of significance.

44

TABLE 27

MODES OF PRESENTATION FOR VOWEL [o 1
ONE WAY ANALYSIS OF VARIANCE

 

Source Sum of Squares df M. S. F

 

 

White noise

 

 

 

 

 

 

effect 425.39 2 212.69 18.39**
Within 1629.92 141 11.56
Total 2055.31 143
TABLE 28
MODES OF PRESENTATION FOR VOWEL [o]
.t-TEST
B M (R) M (L)
B —— -2.79** -4.12**
M (R) -- -1.23
M (L) --

 

*Significant at the .05 level of significance.
**Significant at the .01 level of significance.

45

TABLE 29

MODES OF PRESENTATION FOR VOWEL [1!]
ONE WAY ANALYSIS OF VARIANCE

W

 

Source Sum of Squares df M. S. F

 

White noise

 

 

 

 

 

effect 366.05 2 183.02 11.63**
Within 2219.42 141 15.74
Total 2585.47 143
TABLE 30
MODES OF PRESENTATION FOR VOWEL [1;]
thEST
B M (R) M (L)
B —— -3.10** -3.61**
M (R) -- ~o.51
M (L) ~-

fi—

*Significant at the .05 level Of significance.
**Significant at the .01 level of significance.

. 'Ffln‘d‘...
,.

46

TABLE 31
MODES OF PRESENTATION FOR ALL VOWELS COMBINED

 

 

 

 

Source Sum Of Squares . df .M. S....F
White noise
effect 14861 2 7431.5 l8l9.22**
Within 6417 1571
Total 21378 1573
Discussion

 

The results of the two way analysis Of variance for
binaural mode of presentation and monaural (right) mode of
presentation demonstrate: (l) the mean intensity Of
utterances correSponding t0 the four intensities Of white
noise differ significantly at the .01 level of signifi—
cance; (2) the mean intensity Of utterances corresponding
to the eleven vowels differ significantly at the .01
level of significance; and (3) there is interaction be—
tween vowels Of intensity Of white noise at the .01
level of significance. Tables 5 and 6 represent the
results Of the_t-tests and show the significance Of
differences among the specific vowel sounds. It is
worthwhile to note the contrast in number of Signifi-
cant differences in means at both the .05 level and .01

level of significance illustrated by Tables 5 and 6.

47

The two way analysis of variance used for monaural
(left) mode of presentation yields, in part, different
results. It was determined the mean intensity of
utterances correSponding t0 the four intensities of white
noise do not differ Significantly at the .05 level of
significance. The remainder Of the results correspond
with those found in the binaural and monaural (right)
modes of presentation. The findings through use Of
the pftest are reported in Table 7. ,It is interesting
to note, there are again fewer Significant differences
in the mean intensities of the eleven vowel sounds at
both the .05 and .01 levels of significance than
found in the binaural mode Of presentation.

Tables 8 and 9 illustrate the differences of means
in terms of white noise, of binaural mode of presentation
and monaural (right) mode Of presentation. In the
binaural made Of presentation there was a Significant
difference between the levels of 70db and 100db, at the
.05 level Of significance. There were more and greater
differences in the monaural (right) mode of presentation.
It was found that there were Significant differences
between: (1) 70 db and 90 db at the .01 level of
significance; (2) 70 db and 100 db at the .01 level of
significance; and (3) 80 db and 100 db at the .05 level

of significance.

48

One way analyses Of variance were used with each Of
the eleven vowels to determine whether significant
differences resulted from binaural, right-and left
presentations. The results Of these analyses and
correSponding t—tests, in cases Of significance, are
represented in Tables 11-30.

Vowels [i ] and [1'] demonstrated no significant
change in intensity as a result Of the varied modes of
presentation. The other vowel sounds, [e] , [L] ,

[6E] . [U] . [A] . [0] . [O] . [O] , and [1!].
showed Significant differences at the .01 level of
significance. The significant and non-significant
differences between modes Of presentation for the nine
sugnificant difference vowels are presented in table
form along with the analyses of variance results.

Table 31 illustrates the results of the analysis
variance of the eleven vowels c0mbined with regard to
mode of presentation. The mean intensities for the
three different modes of presentation differ significantly
from one another.

To further illustrate the data discussed, graphic

representations Of the results follow (Figures 1—15).

 

 

 

49

 

 

 

 

 

 

 

 

 

 

 

 

29
8 .
U) 67
26
5
§ 64
H 63
O 62
>. 61
3 60
g .
.§ 70db 80db 90db 100db
H

Intensity of White Noise

Figure l-elntensity of Vowe1S-—Binaural Presentation
Varied Intensities White Noise

Figure 1 shows the increase of mean intensity (voice
reflex) of the vowel sounds as a result Of masking
presented binaurally.

 

 

 

 

 

 

 

 

 

 

 

 

69
28
7
3 26
m 5
6 64
4- 2 ____
O 61
5° 60
'53
C1 ,
E 70db 80db 90db 100db
H

Intensity Of White Noise

Figure 2-—Intensity Of V0wels——Monaura1 (Right)
Presentation-—Varied Intensities White Noise

Figure 2 shows the increase of mean intensity (voice
reflex) 0f the vowel sounds as a result Of masking
being presented to the right ear.

Intensity of Vowels

Intensity of Vowels

69
28

7
66
65
64
63

62
61
60

50

 

 

 

 

 

 

 

 

 

 

 

70db 80db 90db lOOdb
Intensity of White Noise

Figure 3-—Intensity of Vowels-~Monaural (Left)
Presentation Varied Intensities White Noise

Figure 3 illustrates the irregular pattern of the
increase and decrease of mean intensity (voice
reflex) of the vowel sounds as a result of masking
being presented to the left ear.

 

 

 

 

 

 

 

 

 

B M (R) M (L)
Mode of Presentation
Figure 4-—Combined Vowel Sounds--Modes of Presentation

Figure 4 demonstrates that the mean intensity of
the vowel sounds was greatest during binaural
presentation. The figure also shows the vowels
were greater during monaural (right) presentation
than during monaural (left) presentation.

Intensity of Vowels

Intensity of Vowels

69
68
67
66
65
64
63
. 62
61
6O

Mode

69
68
67
66
65
64
63
62
61
6O

 

 

 

 

51

 

 

 

 

U)
r—!
o
3
o
>.
¢+
0
>3
4.)
H
2%
8
1—l C:
H
B M (R) M (L)
of Presentation
Figure 5--Vowe1[] ] Modes
in Presentation
0).
H
0)
z
o
>
CH
0
>3
.p
.H
U)
c
Q)
.p
c
H

 

 

 

 

 

 

 

 

 

B M(R)

M (L)

Mode of Presentation

Figure 7--Vowel[ 5 ].'Modes

in Presentation

69
68

67.

66
65
64
63
62
61
6O

69
68
67
66
65
64
53
62
61
6O

 

B M (R) M (L)
“Mode of Presentation

 

 

 

 

 

 

Figure 6-—V0wel[ I 1
Modes in
Presentation

 

7‘1

 

 

 

 

 

H

B M (R) M (L)
Mode of Presentation

 

Figure 8——Vowel[é: ]
Modes in
Presentation

Intensity of Vowels

Intensity of Vowels

69
68
67
66
65

63
62
61
6O

52

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3 69
g 68
> 67
a g6
o 5 'r-'“
i, 64
t: 63
g 62
f3 61
c 60 l l
_ , H _ i
B M(R) M (L) B M (R) M (L)
Mode of Presentation Mode of Presentation
Figure 9-—Vowe1[ l ] Modes Figure lO-—Vowel [U ]
in Presentation Modes in
Presentation

 

 

 

 

Intensity of Vowels
.oumcnomnowmcnown
owmwtmmflmo

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

B M(R) M(L) B

Mode of Presentation

M (R) M (L)

Mode of Presentation

Figure ll-—Vowe1[/\ ]Modes Figure l2--Vowel [a]

in Presentation

Modes in
Presentation

Intensity of Vowels

Intensity of Vowels

69
68
67
66
65
64
63
62
61
6O

53

 

T—

 

 

 

 

 

 

 

 

B M (R) M (L)

Mode of Presentation

Figure 13—-Vowel [ 3] Modes
in Presentation

 

 

 

 

 

 

 

 

 

 

B M (R) M (L)
Mode of Presentation

Figure 15--Vowel[v] Modes
of Presentation

Intensity of Vowels

 

 

 

 

 

 

 

 

I
69 '7—
68
2g
65
2“ ‘ “"‘
3
62
61
60
B M (R) M (L)
Mode of Presentation
Figure 14--Vowe1[
Modes in
Presentation

Figures 5-15 illus-
trate the effect of
binaural masking in
contrast to monaural
(right and left)
presentations.

CHAPTER V

SUMMARY AND CONCLUSIONS

Summarx

It was the purpose of this research study to

determine the relationship between the Lombard effect and

vowel intensity with reSpect to specific questions. The

questions prOposed were:

1.

How intense is the voice reflex for each
isolated vowel sound when specific levels
of white noise (masking) are presented to
the right and left ears?

How intense is the voice reflex for each
isolated vowel sound when specific levels
of white noise (masking) are presented
binaurally?

Is there a significant difference in the
intensity of the voice reflex on vowel
sounds between the binaural and monaural
masking?

Is there a systematic intensity pattern
that characterizes the eleven Spoken

vowel sounds?

The review of literature included previous vowel

intensity studies and Lombard effect studies. It was

54

55

evident from this review that numerous studies had-been
conducted on the measurement of vowel intensity as well as
research based on the Lombard effect. In all the vowel
intensity studies the emphasis had been on the vowel in a
consonant setting. The research concerned with the Lombard
effect were not limited to intensity of an isolated sound
but dealt in terms of duration, rate, frequency, and
intensity with words and continuous speech in noise or
communication in noise.

Twelve, normal hearing, graduate students in Speech
and Hearing Science, Michigan State University were used as
subjects in the experiment. Equipment used consisted of
(1) an audiometer, (Allison Model 20); (2) headset, (Telephon-
ics TDH-39); (3) a microphone, (4) tape recorder, (Ampex
Model 601); (5) recording tape, (Scotch Tenzar Backing
Magnetic Tape 311); (6) sound level recorder, (Bruel and
Kjaer Model 2305); and (7) a modified electric timer,
(GraLab Microtimer). The twelve subjects produced a
series of eleven randomized vowels at which time white
noise was fed into their ears, binaurally, right and left
at 70, 80, 90, and 100 db (re: 0 0002 microbar).

The results of the research study suggest that the
intensity of spoken vowels may be significantly effected by
intensity of masking being presented to the ears and by the

mode in which this making is presented.

56

Cenclusions

 

Within the experimental arrangements of this study
the following conclusions appear to be in order.

1. There is a significant difference in the mean
intensity of the Spoken vowels corresponding to the four
intensities of white noise when the white noise is pre—
sented binaurally or to the right ear. There is not a
significant difference in the mean intensity of the
spoken vowels corresponding to the four intensities
of white noise when the white noise is presented to the
left ear.

2. The mean intensity of the eleven Spoken vowels
differ significantly when the white noise is presented
binaurally, and to the right and left ears.

3. There is an interaction between the eleven Spoken
vowels and the intensity of white noise presented to the
ears. This interaction is significant when the white noise
is presented binaurally, and to the right and left ears.

4. The Spoken vowels [ f ] and[I'] did not Show a
Significant change in intensity as a result of binaural,
or right and left presentation of the white noise.

5. The spoken vowels [& ],[ 5] ,[83] ,[ Ll ],
[A ],[ Q ],[ Q ],[ O ], and [ 'U ]Showed Significant
differences in intensity as a result a binaural, or right

and left presentation of the white noise.

57

6. There is a significant difference in the mean
intensities for the binaural, left and right modes of
presentation.

7. The eleven Spoken vowels demonstrated a greater
mean intensity when the white noise was presented

binaurally in contrast to right and left ear presentation.

Implications for Further Research

The following questions might well be considered for
further research.

1. What would be the effect on the voice reflex
if the order of intensities of white noise presentations
were changed for each subject?

2. Is there an effect on the intensity of vowels
with regard to placement in the series of vowels for
production?

3. Is fatigue a factor in the degree of voice
reflex exhibited, and if so, to what extent?

4. What Specific characteristics in production
of vowel sounds causes a significant or nonsignificant

reSponse to masking?

APPENDIX A

APPENDIX A

59

Sample of Eleven Vowel Cards for Presentation

 

 

 

 

feet

 

 

fit

[6]

case

 

 

 

 

 

 

 

 

 

[8]

set

 

 

[66]

cat

 

 

soup

 

 

 

 

 

 

 

cup

 

 

[O]

cot

 

 

[J]
caught

 

 

 

 

 

 

[0]

coat

 

cook

 

 

 

 

 

 

 

APPENDIX B

60

61

Tue“
m _L
m/\_
”neg

m 5H
fl 0;
HAUH
HWLH

an,H
Hhmu
m .H
Hue“

.2:
.2:
42
.L .;

.Hari
.H«L_
.Hnmg
.Hm I

.Hewg
.H.og
.HDH
.Hnug

.HAIH
.Hiuw
._ 0_
.m«o_

.H.u_
.fiarL
.H lg
.HHHL

.Ci
.R:
.LHL
.72

.fiwng
.amu_
.HHHL
.Herg

.flfiuw
.flmmg
.HLUH
.H,\_

.8 0_
.finug
.finmw
.mmrg

.wumg
._.uw
.a <H
.mmm_

.HHHL
.H,\_
.flnni
.finug

.HwLH
.HmMH
.Hnmu
.Humw

.HO O
.HHH_
.HXH_
.fin,g

.HMLH
.fi NH
.HM.I
.Hdwg

.H <_
.HOL
.Hmvg
.HHHH

._ <H
.fian
.544“
.H,\H

.7:
.2:
.2;
.62

.flsrL
.Hnna
.HKHH
.HHwH

.r:
3;
.2:
.Bi

.flmtg
.fimug
.Homg
.fij I

.~.mg
.H,\H
.mj I
.H,\_

.finvi
.H.w_
.HMVL
43

Lfl<u
.Hn,u
.Hmww
.HnVH

 

COfipmpcomoLm pom mesmo

Hozo> go sopho Eoocmm

.quu .H HI
.mew .fiisn
.Ln,I .L ii
4H6 i .fin,L
.HLUL ._me
.quI .L/\L
.anI .LAUL
.hiui .H_m_
.5 mi .fl_si
.L.mL .HLHL
.fl,\_ .H 0L
.fli I .Ln,L
d m m

OOH
om
cm
on a q

OOH
om
ow
ow : m

OOH

om

om

ON I m

Adv somnnsm

62

:0:
T:
:0:
5m:

H:
5m:
:0:
2:

:0:
:<:
::
:H:

40:
42.:
AP:
40:

.HuH:
AU:
.HO:
Am:

43.:
.7:
.T:
4N:

.HH:
40:
Au:
AH:

.HO:
4H:
.33
.HU:

AN:
.H 6:
HO:
.2:

.83
.23
.8:
.2.

4w:
.::
.26:

.2:
.3:
HR:
.H<:

.51
.26:
.7:
42

AU:
.HU:
Ad’s:
.Hmu:

4.:
40:
.fi:
.3:

.3;
.23
.f;
.2:

i:
.7;
.2:
.2:

Am:
.32
.2:
.71.:

.2:
40:
.H<:
.HH:

.2:
.2;
so:
.2:

An):
.HH:
AH:
.2:

so:
.L .L
.6:
.7:

Aw:
.Hm:
.2:
.24:

.3:
.::
AG:
40:

.i C
.2:
5.:

An:

 

QOHQMpcomosm LOM mpsmo

Hozo> we Looso Eoocmm

.HLH: .Hflu:
.2: Am:
.H: An:
:0: Am:
.2: .HG:
4w: .fi<:
.73.: .3:
.Z: . R:
.3: £0:
.32 .::
.Hw: 40:
Am: .HG:
mmq

OOH
om
om
ON a m

OOH

om

ow

ON I m

OOH
om
ow
ON I H

Am: pednnsm

63

any:
fl ::
.HU:
m m:

f:
t:
I:
f;

2:
2:
T;

.finw:
.Hw :
.mmw:
.H _:

.flP:
.HMU:
.fl _:

.mb.:

.Hnu
.h _
.5
.5 <

l—ll—JL—JI—d

.Hnu:
.HAU:
.H.H:
.wa:

.Hfiv:
.msfi:
.HKH:

.finv:

F4 ’7 <? no

.6:
.2:
.2:
.2:

.H.::
.an:
.Hn,:
.H_4:

.Hfi4:
.qu:
.mnu:
.HHU:

.HH.:
.HMH:
.Hum:
.HHO:

.Hnfi:
.H:4:
.th:
.Hnr:

.HhH:
.HH.:
.flmm:
.HLw:

.Hmm:
.mmw:
.Hnu:
.Hmm:

.fi/\:
.Hu”:
.Huu:
._~u:

.EMM:
.Hdw:
.fl~U:
.fi,\:

.finv:
.Hn,:
.HMH:
.Hmv:

.muw:
.9 w:
.mfiw:
.an:

.5:
.2:
.5:
f;

.m/\:
.Hnl:

.HHH:
.Hun:

.H.v:
.finq:

 

COHpmpcmmmpm pom mcpmo

szo> mo LmULO Soccmm

.23 42
i: f:
.2: .52
42 .52
52 42
.2: .2:
.3; f:
.2; .2;
f: .2:
f: i:
.Z: g2
.i: .3:
mqm

OOH
om
ow
ow u m

OOH
om
om
ow : q

OOH

om

ow

ON. I m

Amv pommnsm

64

33
::
T:
2:

S2
22
I:
:2

2:
RU:
Ham:
:2

.2:

4.:
.72.:

.7:
.T:
.2:
4H:

.7:
.7:
.22
.52

.2:
f:
.2:
.20:

.j:
.2:
.3:
.HH:

.H:
.8:
.T:
QM:

.23
.2:
.2:

.2:

£0:
Hm:
Aw:
.H<:

.2:
AH:
A:
An:

.2 .2
.52
.3;
.2 .:

.Zu:
.6:

.:\:
.2:

.2:
.HO:
4”:
.Z:

.2:

2.:
gm:
.3:
.32

An:
£02
.53
.2:

 

COprpcmmmpm pom mvpmo

Hm30> mo mmULO Eowcmm

2-: .22
.22 .7:
.62 .22
22 .Z:
.52 .32
.62 .2 .:
.22 .3:
.32 f:
.52 .2:
.Z: 22
.22 .32
.22 .8;
H m m

OOH
om
ow
ow : q

OOH
om
om
ON I m

OOH
om
ow
ON a m

Adv pommnsm

55

202
2.22
202
202

222
2:2
202
2:

2.:
23
2.;
2.:

.202
.202
.22.;
.2”:

.22:
.202
.22.:
.202

.202
.202
.22;
.202

.202
.202
.23
.2 22

.202
.2:
.202
.2<2

.2<2
.2:
.32
.2<2

.25
.23
2.:
.23

.23
.2:
.22
.22

4.2.2
.2”:
.2H2
.201

.2”:
.202
.23;
.202

A”:
.22.:
.202
.2<2

.32
.202
.2 .2
.22

.262
.2<2
.202
.202

.202
.22;
.202
.202

.232
.23
.202
.202

.22.:
.2“:
.291
.202

.2:
.222
.2:
.2 22

.262
.2<2
.22
.2H2

.203
.2”:
.2<2
.202

.2w2
.2m2
.202
.202

.2“:
.202
.262
.232

.202
.32
.202
.2”:

.232
.202
.2w2
.31

 

COHpMpcmmmpm pom mvpmo

Hmzo> O0 mmULO Eowcmm

.2: .22
.23 .23
.22.: .22
.22 2.:
.22 .23
.23 .2 .2
.22 .23
.22: .22:
.202 .52
.22 .22
2.: .23
.23 .22
mmq

OOH

om

ow

ON. I m

OOH
om
ow
ON 1 m

OOH
om
ow
ON I A

Amv pomnpsm

66

2®2
202
2<2
2<2

23
2.:
2.:
23

262
2H2
202
292

.2u2
.232

.2H2
£02

.2:
2.:
.23
.23

gfll
.202
A22
402

2»;
.292
.2:
.2m2

.202
£22
£02
.2w2

.202
.202
.2<2
422

.23
.2 .2
.28
.32

.2H2
4&1
.2H2
£22

.202
£32
.202
.202

.202
.2<2

2»;
.202

4»;
.2<2
.202
£02

.2 .2
.82
2.:
.32

.2H2
.2w2
.202
.232

.202
.202
.262
.2<2

42;
.2w2
A»;
.202

.202
.262
£02
.202

AD;

.2m2
4M3
.202

.2w2
.202
.232
.2H2

2.:
.23
.22:
2.:

.H<2
.202
.2w2
222

.2<2
.2P2
.262
.2<2

432
.202
.2<2
.2H2

.2H2
.2<2
.2H2
Am2

 

COHpmpcmmmhm LOO mnhmo

Hm30> mo pmcpo Eowcmm

.22: .29
.22 .22
.202 .202
.22 .23
.202 .22:
.22 .232
.23 .202
.22 .32
.202 .23
2.: .202
.22: .202
.23 .23
mam

OOH
om
om
ON I m

OOH
om
ow
ON I A

OOH
om
ow
ON I m

AwV pomnnsm

67

7N2
2<2
202

2H2

20m2
H2u2
2 22
2.22

2222
22w2
2222
HAVH

.2/\2
.22fi2
.20_2
.2222

2HG2
JflwH
.2n,2
.22u2

.2n._
.2n.2
2Hw2
2H<2

.22
2 2
.2 .2
2.2

.Hn.2
.22w2
.22w2
.2/\2

.2nV2
.mmv2
.2Mw2
.2mw2

202
.222
2.2
.2 .2

.2Mu2
.202
.202
.2n,2

.2 .2
2S
.2 E
202

.2H22
.2 H2
.2Mv2
.Hnu2

.2uq2
.2mm2
.2H.2
.2wH2

2.2
2.:
2.2
2.2

.2 .2
.202
.2 22
.22

.2/\2
.2 22
.2mm2
2HG2

.20m2
.2H22
.2um2
2Hm2

.HnV2
.Hn.2
.HMMH
.20m2

.2._2
.2222
.26 2
.2mm2

.22H2
2H<2
.HnV2
.2HH2

.22r2
.2: 2
.2m.2
.22u2

.2Hr2
2H<2
.20H2
.2w22

.2m22
.2nV2
.2H~2
.Hn.2

.2HH2
.2AV2
.20 2
.2< 2

.23 2
.2242

2H<2
.2mw2

.222
2.2

.22u2
.23 2

 

COHQMucwmmpm pom mvpmo

Hm30> Oo pmcuo Eoccmm

2m: 23
.22 .23
22 2E
28 2.2
.22 202
.202 .22
.222 2.2
2.2 .22
.2/2 .23
.62 2.2
22 2E
.22 2.2
H m m

OOH
om
om
ON I q

OOH
om
ow
ON I m

OOH
om
ow
on I m

25v pommnsm

68

2<2
222.2
2.22
2.222

202
22.2
2H2
2<2

2<2
202
262
2H2

.202
.2 .22
.22
.282

.222
2.2
.2 .2
23

.232
.232
.202
.202

 

.232 .2 .22 .202 .292 .282 .2H2 .2m2
.2v2 .2H2 .2uw2 .202 .202 .202 .2<2
.2m2 22.2.2 .202 .202 .202 .282 .202
.232 .2P2 .202 .2m2 .202 .222 .202
.202 .232 .202 .202 .2<2 .2w2 .2H2
.202 .202 .202 .2<2 .2w2 .232 .202
.202 .202 .202 .202 .292 .282 .232
.202 .23 .22 2&2 2.2 .22.: 2.12
.202 222 .202 .222 2.2 .202 .282
.2 .22 .2<2 .232 .202 .202 .2w2 .2H2
.2.22 .292 2.4.2 .232 .2<2 .202 .202
.202 .2<2 .232 .202 .222 .202 .2m2

COHpmpcwmmhm pom mUhmo Hm30> mo gmcpo Eovcmm

.53 .202
.22: .23
2,: 2:2
2<2 .202
2.2 .222
22.2 .202
.222 .22
.23 .202
.22: .22
202 .23
.282 .23
.82 2:2
mmq

OOH
om
ow
on u m

OOH

om

ow

ON. I m

OOH
om
om
ON I q

2mv pomnnsm

69

202
202
202
292

2.2
23
22:
22

2.2
23
22
202

2w2
2»:
.202
.2H2

232
.202
.202
.2w2

2<2
222
2:2
2H2

.22
2,:
.202
2.2

2M2
202
.2w2

.2<2

.232
2M2
A»:

202
.2H2
222
2%:

.232
2.2
.2 .2
.201

.202
2&2
.232
.2P2

222
.262

.262
.202

.2<2
.2<2
.2H2
.2P2

.2w2
.202
.2<2
2m:

.292
.2P2
A»:
.202

.23
.22:
.23
2.2

.202
.202
.202
.202

2<2
.2w2
.2p2
.202

202
2m:
2»:
.202

.202
.202

.232
.232

.202
.202

.202
.202

202
.2w2
.2H2
2m2

.292
.2H2
.202
.232

 

COHmecmmmpm pom mcpmo

Hmzo> we pmwho Eowcmm

.282 .232
22: .222
.22 .22
.22 .22
.22 .202
.2 .2 .202
2:2 2.:
22.2 .232
.23 .292
.2<2 .22
.23 .202
.2<2 .22
mHm

OOH
om
ow
ON I m

OOH
om
ow
ON I A

OOH
om
ow
cm 1 m

20V pommnsm

7O

2H2
202

2M2
202

202
632
232
262

2.82
2H:
2w2
2:2

2%;
.Tw2
.2H2
2u2

.23
2...;
.2 .2
2.2

.202
.232
.202
.202

.202
.2H2
.202
.202

2.2
.2<2
.202
.2 .2

.202
.202
222
2m2

.2w2
.2m2
.2m2
.2m2

.2H2
202
2mg
.202

.2H2
.2<2
.2<2
.202

.202
.2m2
.202
.292

232
2w2
.202
2w2

.202
.292
.292
.292

222
.202
222
.202

.202
.292
.232
.2H2

.232
.2H2
2G2
.202

.202
2:2

.202
232

.22
.2 .2
22
.23

.23
.22
2.2.2
.2 .2

.232
.2<2
.2<2
222

2R2
2:2
202
.2<2

.2<2
.202
.202
.202

.2<2
222
232
.2<2

.202
.2H2
.2m2
.202

.2w2
.262

.232
2n:

2<2
.202

2.2
2.2
.22
.22

2 m m

 

COHpmpcmmmpm LOO mwpmo 2®30> mo hmwpo Eovcmm

232
2M2
.292
202

202
.262
.202
.202

.2b2
.2w2
.2H2
.2<2

OOH
om
om
ON I 2

OOH
om
om
ON I m

2:
om
ow
om I m

2022 pomnnsm

71

202
202
202
2 <2

22r2
22~2

2H2
202

202
202
202
2:2

.2 .2
2:2
.282
.202

.222
.2 .2
.2<2
.62

.232
.22H2
.222
.2<2

22.2
.5H2
.2H2
.2H2

.262
.232
.202
.232

.2H2
.2w2
.2H2
.2w2

.2<2
.202
.23
.202

.202
.202
.232
.2P2

.202
.202
.232
.2M2

.2w2
.202
.202
.262

.202
.202
.202
.2<2

.23
.2 .2
2.2
.202

.202
.2m2
.2U2
.232

.2H2
.SM2
.202
.2H2

.OU2
.2<2
.dm2
.232

2.2.2
2H~2
.232
.232

.2»:
.2w2
.202
.2w2

.2 <2
23
2&2
.2 .2

.202
2nw2
.2w2
.2w2

.2w2
.262
.2w2
.202

2.2
.222

..202

.2fl2

.282
.2<2
2.. 2
.2 2

.222
.2H2
.2M2
.222

2mm2
2nm2
.202
.2H2

.232
.2 .2

.292
.202

.2<2
.202
.222

.fiN2

.2w2
.232
.2w2
.202

m m 2

 

COHumwcmmmpm pom mcpmo Hm30> we hmcpo Eoucmm

.262
.202
.2<2
2um2

.202
.2<2
.292
.202

.2P2
.202

.2<2.

.202

092
om
cm
0» I m

OOH
om
ow
cm I m

OOH
om
ow
ON I 2

2222 pomfinsm

72

202
202
2.2
2<2

:H2
22.2
2H2
202

202
202
202

292

2.22
.232
2&2
.202

.222
.222
.2<2
23

.22:
2.:
2.2
.2<2

2:2
.2H2
.202
.2H2

202
232
2U2
.232

.2H2
2w2
2H2
.2w2

2<2
.262
.2H2
202

.202
.202
.232
.292

.202
.202
.232
.2m2

.2w2
.202
202
.262

.262
.202
.202
.2<2

.22:
2.2
2.2
.202

202
.2w2
.202
.232

.2H2
.202
.202
.2H2

202
.2<2
.202
.232

.2H2
.232
.232
.2P2

2M2
.2w2
.202
.2w2

.2<2
.22:
.282
.2 .2

.2v2
2M2
2w2
.2m2

2m2
2M:
.2w2
.202

2.2
.22:
.202
.61

.201
.2<2
.202
2.22

.2 .2
.22:
.282
.2 .2

.202
2»:
.202
.2H2

 

COHQMpcmmmhm pom mwpmo

Hmzo> mo pmcgo Eoccmm

.232
222
.292
.202

.2<2
.22
2.2
2.2

.2w2
.232
.2m2
.202

$1.2m

.202
.202
.2<2
2N2

.222
.202
.2<2
.202

OOH

om

om

ON I m

OOH
om
ow
ON I 2

OOH

om

ow

ON. I m

2m2v pomwnsm

APPENDIX C

73

74

Subject No. 1

Binaural Presentation

[i ] [1?] [CE] [El] [as] [L4] [/\J [63] [CD] [CD] [If]

       

 

70 db 64 66 66 68 7O 66 72 72 73 7O 7O
80 db 62 64 62 62 66 62 68 68 68 64 65
90 db 6O 62 62 62 64 62 66 66 66 66 64
100 db 62 62 62 63 65 62 64 63 63 67 64

Intensities

 

Monaural Presentation (Right)

[ i] [.I] [<3] [5;] [at] [L4] [/I] [CI] [:>] [CD] [If]
70 db 62 64 66 66 64 66 67 68 66 67 68
80 db 64 64 66 64 65 66 66 70 7o 66 7o
90 db 66 68 68 70 7o 68. 73 72 74 72 70

100 db 64 66 68 68 7O 66 71 71 72 7O 72

-.—.._—.

       

Intensities

 

Monaural Presentation (Left)

 

[ I] [.I] [It] [15] [ab] [L4] [A 1 [CI] [5)] [CD] [If]

 

70 db 60 60 6o 60 6o 62 62 66 6o 62 62
80 db 6O 6O 6O 62 62 62 64 66 62 64 64
90 db 6o 62 62 63 66 66 66 66 7o 68 68
100 db 6o 60 62 64 64 64 64 68 68 67 66

Intensities

 

Binaural Presentation

75

Subject No.

2

 

[i J [I] [C] [5] [ac] [u] [A] [a] [J] [O] [‘U]

‘4

 

 

 

 

 

 

 

 

 

U)
§7o db 62 62 66 67 64 64 64 7o 64 66 64
.H
2 80 db 66 66 68 66 66 66 69 68 66 7o 69
Q)
E 90 db 66 66 66 68 7o 66 70 7o 70 68 68
100 db 68 7O 66 7O 71 69 76 74 74 66 72
Monaural Presentation (Right)
[i][I][C][8][ac][Ll][/\][GI][J][o][‘U]
CD
3370 db 6O 62 64 62 62 62 64 62 64 64 64
.24
8 86 db 64 64 64 64 62 63 64 63 64 64 65
(D
E 90 db 62 64 66 66 64 68 66 64 66 68 66
100 db 64 66 66 64 63 66 72 68 68 66 64
Monaural Presentation (Left)
II‘IIIIIdIIEJIacIIuII/IIIcII[a]IonI
2?, 70 db 66 68 69 7o 68 68 6o 62 68 7o 68
.21
E80 db 6O 62 66 66 6O 64 64 64 64 66 65
E90 db 62 64 65 66 63 62 62 62 68 66 66
H
100 db 62 66 67 66 62 66 64 64 66 68 67

 

Intensities Intensities

Intensities

76

Subject No. 3

Binaural Presentation

 

H] [r] [e] [a] [as] [u] [A] [a] [o] [o] [v]

 

70 db 62 7O 68 7O 69 64 69 72 72 66 68
80 db 61 64 7O 69 68 66 72 74 72 7O 68
90 db 6O 66 68 69 7O 64 74 76 7O 7O 70
100 db 62 66 65 62 67 67 72 69 68 7O 7O

 

Monaural Presentation (Right)

 

[f] [I] [C] [5] [as] [U] [A] [Q] [J] [O] ['U']

 

70 db 6o 60 62 60 6o 60 62 62 62 62 62
80 db 68 64 64 68 66 61 66 64 66 66 65
90 db 64 64 64 64 62 64 66 66 64 66 66
100 db 62 64 64 62 64 66 66 66 66 65 66

 

Monaural Presentation (Left)

-—‘
‘___

[Milne] [81 [281w] [/\][a][D] to] [U]

 

70 db 58 62 62 6O 59 6O 64 62 62 68 62
80 db 58 62 62 62 6O 62 64 62 63 62 64
90 db 58 62 62 62 62 64 64 66 64 6O 64
100 db 58 62 62 61 6O 62 62 62 62 64 64

 

77

Binaural Presentation

Subject No. 4

 

[I ] [I] [6] [5.] [a5]

[44] [A] [O] [o] [O] [V]

 

 

 

 

 

 

 

 

U)
E 70 db 6o 62 62 64 64 6o 66 66 66 64 64
..-4
g 86 db 62 64 62 64 66 58 7o 67 66 64 66
(D
E 90 db 58 62 64 64 68 6o 68 66 66 64 64
100 db 58 64 62 65 66 6o 66 66 68 66 64
Monaural Presentation (Right)
[[1 [I] [e] [a] [as] [u] [A] [a] [a] to] [U]
U) ,
Q)
E 70 db 62 59 60 6o 58 62 6o 58 6o 62 58
”5:: 80 db 58 58 58 58 56 58 6o 60 62 6o 60
E 90 db 62 59 60 6o 60 6o 64 63 64 62 60
100 db 64 62 61 6o 62 62 66 66 66 62 6o
Monaural Presentation (Left)
[i] [I] [6] [£2] [ac] [H] [A] [a] [D] [O] [1!]
C0
0)
+3 70 db 56 58 58 59 6o 56 6o 60 58 56 60
E?) 80 db 56 60 6o 58 57 58 62 58 64 58 58
4.)
.E 90 db 56 58 58 58 56 58 62 58 6o 60 60
100 db 58 60 6o 60 58 57 62 6o 64 58 6o

 

Intensities Intensities

Intensities

78

Binaural Presentation

Subject No.

5

 

 

[i] [I] [e] [61 [ac] [u] [A] [a] [a] [0] [U1

 

 

 

 

 

 

 

 

 

 

 

70 db 6O 64 64 66 64 66 66 64 64 66 72
80 db 62 66 66 72 68 66 7O 7O 68 69 68
90 db 62 66 66 7O 68 66 7O 68 7O 68 70
100 db 64 66 66 72 68 7O 68 7O 7O 68 68
Monaural Presentation (Right)
[1'] [I] [e] [a] [as] [u] [A] [a] [a] [o] [v]
70 db 58 6O 62 64 62 62 62 62 62 62 62
80 db 58 62 6O 64 62 6O 62 62 64 6O 62
90 db 6O 64 62 66 62 6O 64 64 64 62 64
100 db 6O 62 64 64 64 64 62 64 65 62 66
Monaural Presentation (Left)
[[ ] [l7] [5'] [5'] hi5] [Li] [A ] [0 ] [C3] [CD] [If]
70 db 6O 64 64 66 62 6O 66 62 64 63 64
80 db 6O 65 64 64 62 6O 64 64 63 64 62
90 db 6O 64 62 64 62 62 64 62 63 62 64
100 db 6O 62 62 66 62 6O 63 64 64 62 62

 

79

Subject No. 6

Binaural Presentation

 

[I] [I] [e] [a] [as] [u] [A] [a] [o] [o] [v]

70 db 6o 62 69 72 64 65 72 66 68 62 62
80 db 6o 58 68 7o 56 58 6o 60 6o 58 56

Intensities

90 db 54 66 56 58 56 56 56 66 53 62 58
100 db 54 56 54 56 58 56 56 58 64 54 55

 

Monaural Presentation (Right)

 

 

[]’] [I] [e] [6] [as] [u] [A] [a] [o] [o] [v]

 

70 db 54 62 56 56 54 56 56 58 58 58 56
80 db 58 54 51 58 58 58 54 6o 58 56 56
90 db 56 56 56 56 56 58 58 56 54 58 58

Intensities

100 db 57 58 56 56 58 5 58 58 56 6o 59 56

 

Monaural Presentation (Left)

 

 

[i] [I] [ens] [as] [u][/\][a][o][o1 [v]

 

70 db 54 56 54 58 54 58 54 56 56 58 58
80 db 56 56 54 54 56 56 52 56 56 56 56
90 db 54 55 54 52 54 58 56 54 54 54 56
100 db 54 54 56 56 58 59 56 56 56 54 54

Intensities

 

Intensities

80

Subject No. 7

Binaural Presentation

 

 

[i] [I] [e] [8] [86] [u] [A] [a] [o] [o] [v]

 

70 db 66 7o 68 7o 70 66 68 68 71 7o 67
80 db 68 70 7o 70 72 70 7o 70 74 74 7o
90 db 68 72 72 74 72 7o 75 74 76 74 74
100 db 68 72 72 76 76 7O 76 76 76 76 76

 

Monaural Presentation (Right)

 

 

Intensities

Intensities

0

[I ] [1:] [CB] [5,] hit] [Li] [A ] [Cl] [5)] [C7] [11]

 

70 db 66 68 66 68 7O 68 66 68 7O 7O 68
80 db 64 68 68 68 67 68 66 68 7O 69 68
90 db 68 68 68 68 72 66 7O 68 72 72 68
100 db 68 68 68 68 68 66 68 7O 7O 7O 68

 

Monaural Presentation (Left)

 

 

[ [J [Li] [€31 [E5] [35] [Ll] [/\] [O ] [D ] [C3] [Li]

 

70 db 62 64 66 64 64 64 66 62 64 66 64
80 db 62 64 64 62 62 6O 63 64 66 64 64
90 db 62 64 64 64 64 62 64 66 66 64 66
100 db 64 65 66 66 65 65 66 68 68 66 64

 

Intensities

Intensities

81

Subject No. 8

Binaural Presentation

 

 

[i] [I] [e] [e] [as] [U] [A] [a] [o] [0] [U]

 

70 db ' 6o 62 66 64 66 62 66 68 68 68 68
80 db 66 68 68 72 68 64 66 72 72 7O 68
90 db 68 74 72 72 74 68 7O 76 74 72 70
100 db 74 76 72 76 76 74 76 76 78 76 76

 

Monaural Presentation (Right)

 

[['] [I] [e] [a] [as] [u] [A] [a] [a] [c] [v]

 

70 db 58 62 64 62 62 6O 66 68 66 66 6O
80 db 62 62 68 64 66 62 66 64 67~ 68 66
90 db 6O 64 64 66 65 6O 64 66 66 66 64
100 db 62 64 68 68 64 62 66 66 66 68 64

 

Monaural Presentation (Left)

 

 

[f] [I] [e] [a] [as] [u] [A] [a] [o] [o] [v]

 

Intensities

70 db 62 66 66 66 7O 66 66 68 68 68 68
80 db 64 66 64 68 68 68 66 66 66 7O 68
90 db 59 64 65 66 66 61 64 64 66 68 60
100 db 6O 64 64 66 66 64 66 61 66 66 64

 

Intensities

Intensities

Intensities

82

Binaural Presentation

Subject No. 9

[i] [I] [8] [81m] [u] [A] [a] [o] [c] [U]

 

 

 

 

 

 

 

 

 

 

70 db 64 68 66 72 72 66 72 7O 74 7O 7O
80 db 64 68 7O 7O 68 64 72 72 68 7O 68
90 db 62 68 68 68 68 62 68 68 68 68 68
100 db 64 65 65 66 66 6O 7O 64 68 66 66
Monaural Presentation (Right)
[[1 [I] [C] [E] [a] [U] [A] [a] [J] [0] [7U]
70 db 62 66 62 66 65 64 64 66 64 68 62
80 db 64 66 66 6O 66 6O 67 64 68 66 66
90 db 62 64 64 66 66 62 68 64 64 68 66
100 db 59 66 62 66 62 68 62 6O 6O 68 62
Monaural Presentation (Left)
[[1 [I] [o] [a] [as] [u] [A] [a] [o] [c] [‘U‘]
70 db 6O 6O 6O 64 61 58 62 6O 66 62 6O
80 db 6O 64 63 62 64 6O 64 66 66 66 64
90 db 6O 66 64 66 62 6O 66 68 6O 64 63
100 db 6O 64 62 64 61 58 62 64 64 64 62

 

83
Subject No. 10

Binaural Presentation

 

 

[1'] [I] [e] [a] [de] [[4] [A] [o][o] [01w]

 

 

 

 

 

 

 

 

 

 

U)
3 70 db 58 6o 60 60 6o 61 60 6o 62 62 60
.p
"3* 80 db 60 6o 60 6o 60 62 60 6o 62 64 6o
{‘2’ 90 db 58 58 6o 60 63 61 60 6o 60 6o 60
H

100 db 6O 6O 64 61 6O 64 58 62 58 62 62

Monaural Presentation (Right)
[f] [I] [e][e][ae] [U] [A] [mm] [mm]

m 2
£1 70 db 60 6o 60 60 6o 62 58 56 58 6o 58
E 80 db 58 6O 58 56 56 60 6O 58 56 58 57
Q)
E 90 db 58 6o 58 58 55 62 58 56 58 60 60

100 db 6O 6O 58 58 58 62 58 58 58 6O 6O

Monaural Presentation (Left)
[1'] [I] [e] [2] [as] [u] [A] [a] [o] [c] ['u]

m __
fi’ 70 db 59 6O 59 59 60 60 6o 58 60 6o 59
H
g 80 db 57 62 6o 58 58 6o 60 58 6o 56 60
E 90 db 58 56 55 56 56 58 56 56 58 6o 57

100 db 56 6O 6O 58 6O 60 6o 58 6O 58 6O

 

84

Subject No. 11

Binaural Presentation

 

 

[i] [I] [6?] [E] [80] [U] [A] [CI] [0] [O] ['U']

 

70 db 62 62 64 65 66 6o 66 68 68 64 64
80 db 62 62 64 66 66 62 67 68 66 66 64
90 db 6o 66 66 66 66 63 67 66 68 66 66
100 db 64 66 67 68 68 62 68 70 7o 68 66

Intensities

 

Monaural Presentation (Right)

 

 

[1'] [I] [e] [6] [as] [u] [A] [a] [a] [c] [1r]

 

70 db 6O 62 62 64 64 6O 64 66 64 68 64
80 db 58 62 62 66 64 6O 66 66 65 64 64
90 db 6O 62 64 64 64 6O 62 66 66 64 62
100 db 6O 62 64 64 64 62 66 64 66 64 64

Intensities

 

Monaural Presentation (Left)

 

 

[i] [I] [e] [6] [ae] [u] [A] [a] [a] [0] [U]

 

70 db 62 66 66 68 7O 64 68 72 72 68 66
80 db 58 62 64 66 66 64 66 67 66 66 66
90 db 62 66 64 66 66 62 68 68 66 66 66
100 db 62 64 64 66 64 62 66 66 66 64 68

Intensities

 

Intensities Intensities

Intensities

85

Subject No. 12

Binaural Presentation

 

[i] [I] [e] [2] [as] [u] [A] [a] [a] [o] [1!]

 

70 db 64 68 7O 64 63 66 64 7O 68 68 66
80 db 66 68 7O 68 67 68 66 7O 7O 7O 66
90 db 68 66 74 72 68 68 7O 72 72 72 68
100 db 68 7O 74 72 73 72 72 76 72 74 7O

 

Monaural Presentation (Right)

 

 

[1'] [I] [e] [a] [as] [u] [A] [a] [o] [0] [v]

 

70 db 62 62 64 6O 62 64 62 64 62 64 64
80 db 62 62 66 62 62 66 64 66 66 66 64
90 db 62 64 65 64 64 64 62 64 64 68 64
100 db 64 62 64 66 64 66 68 64 66 . 68 64

 

Monaural Presentation (Left)

 

 

[i] [I] [e] [a] [as] [u] [A] [a] [.3] [o] [v]

 

70 db 64 62 64 66 62 62 62 64 64 64 64
80 db 66 62 64 66 63 64 64 64 64 66 64
90 db 62 62 64 64 64 64 64 64 64 66 64
100 db 64 62 68 64 62 64 64 64 68 64 64

 

BIBLIOGRAPHY
Books

Dixon, Wilfrid J., and Massey, Frank J., Jr. Introduction
to Statistical Analysis. New York: McGraw-Hill
Book Company, Inc., 1958.

 

Heller, Morris F., Anderman, J., and Singer, Ellis E.
Functional Otology: The Practice of Audiology.
New York: Springer Publishing Co., Inc., 1955.

Newby, Hayes A. Audiology Principles and Practice.~ New
York: Appleton—Century—Crofts, Inc., 1958.
Articles and Periodicals
Black, John W. "Natural Frequency, Duration, and Intensity

of Vowels in Reading," Journal of Speech and Hearing
Disorders, XIV (September, 1949).

 

Black, John W. (ed.). "Studies in Speech Intelligibility,’I
Speech Monographs, XIII (Number 2, 1946).

 

Dreher, John J. and O'Neill, John J. "Effects of Ambient
Noise on Speaker Intelligibility of Words and Phrases,"
Laryngoscoge, LXVIII, Part 1 (1958).

 

Fairbanks, G., House, A. S., and Steven, L. ”An Experimental
Study of Vowel Intensities," Journal Acoustical Society
of America, XXII (July, 1950).

 

Grove, W. E. "Simulation of Deafness," Annals of Otology,
Rhinology, and Laryngology, LII (September, 1943).

Hanley, T. D., and Steer, M. ”Effect of Level of Distracting
Noise Upon Speaking Rate, Duration, and Intensity,"
Journal of Speech and Hearing Disorders, XIV (December,

1949)-
Korn, T. S. ”Effect of Psychological Feedback on Conversa-

tional Noise Reduction in Rooms,” Journal Acoustical
Society of America, XXVI (September, 1954).

86

87

Sacia, C. F., and Beck, C. J. "The Power of Fundamental

Speech Sounds," Bell System Technical Journal,
V (July, 1926).

Stevens, Kenneth N., and House, Arthur S. "An Acoustical
Theory of Vowel Production and Some of Its Implica-

tions,” Journal of Speech and Hearing Research, IV
(December, 1961).

Taylor, Glenn J. "An Experimenta1.Study of Tests for the
Detection of Auditory Malingering," Journal of
Speech and Hearing Disorders, XIV (March, 1949).

 

Unpublished material

Waldron, D. L. ”The Lombard Voice Reflex Test: An Experi-
mental Study." Unpublished Doctoral dissertation,
Department of Speech, Stanford University, 1960.

III
III