LIPREADING PERFORMANCE AS A
FUNCTION OF LIGHT LEVELS

Thesis {or the Degree of M. A.
MICHEGAN STATE UNIVERSITY

Sharon Thomas

1962

‘91“ fifmc 1":.- *

 

ABSTRACT

LIPREADING PERFORMANCE AS A FUNCTION
OF LIGHT LEVELS

by Sharon Thomas

The purpose of this study was to measure the differ-
ence in the lipreading performance of trained subjects
during varying intensities of room illumination in an
effort to determine the effect of various room lighting
conditions on the lipreading efficiency of these subjects.

The viewer subjects, ten men and ten women, were
screened for visual acuity at 20/30. One additional female
subject was present at five of the six group experiments.
Her scores were used for comparative purposes. One female
speaker presented the test sentences at all training and
experimental groups. The subjects were trained to lipread
a list of thirty test sentences. The sentences were
designed to contain speech sounds in an approximate ratio
to sounds in their frequency of occurrence in American con-
versation. Those individuals who could not correctly
identify 95 per cent of the sentences after the training
session, were eliminated from the study. During training,
the room lighting conditions were held constant at thirty
foot-candles. White noise, which measured from 84 to 86
decibels, and inactive headsets worn as muffs, were used

to mask auditory clues.

Sharon Thomas

During the actual experiment, the speaker presented the
thirty sentences during each of six varying room lighting
conditions. The six light levels were thirty foot-candles,
sixteen foot-candles, eight foot-candles, four foot—candles,
and one-half foot—candle, as measured by a light meter at the
speaker's face. Inactive head sets and white noise were,
as in the training sessions, used for the purpose of masking
auditory clues from the speaker. The subjects responded in
written form to the test sentences.

The findings.of this study indicate that lipreading
efficiency of trained subjects tends to decrease, but not
significantly, as the intensity of room illumination decreases.

The conclusions suggest that the lipreading efficiency
of individuals who are highly familiar with the Content of
the message decreases only slightly, and not significantly,
as the intensity of room illumination decreases. They
further suggest that lipreaders, who are highly familiar
with the content of a message, when sufficiently motivated
will lipread with an adequate degree of efficiency under

the most adverse lighting conditions.

\ :."
Approved 13517 LUL“ I/fiA LU“-
/ ‘DirectoiJV' )
a
(

LIPREADING PERFORMANCE AS A FUNCTION

OF LIGHT LEVELS

By

!
1

Sharon Thomas

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

1962

TABLE OF CONTENTS

Page
LIST OF TABLES . . . . . . . . . . . . . iii
Chapter
I. STATEMENT OF THE PROBLEM. . . . . . . 1
Introduction. 1
Statement of Problem and Purpose of
Study . 3
Questions Posed at Onset 3
Importance of Study 5
Definition of Terms 5
Organization of the Thesis 6
II. REVIEW OF THE LITERATURE. 8
Factors Influencing Lipreading
Efficiency. . . . 8
The Need for Investigations of. the
Relationship Between Light and
Lipreading. . . . . . . . . . 10
III. SUBJECTS, EQUIPMENT, MATERIALS, AND
PROCEDURE. . . . . . . . l2
Viewer Subjects. . . . . . . . . l2
Speaker Subject. . . . . . . . . 12
Equipment. . . . . . . . . . . 12
Materials. . . . . . . . . . . 13
Procedure. . . . . . . . . . . 13
IV. RESULTS AND DISCUSSION . . . . . . . 18
Results . . . . . . . . . . . 18
Discussion . . . . . . . . . . 21
V. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS . 25
Summary . . . . . . . . . . . 25
Conclusions . . . . . . . . . . 27
Recommendations. . . . . . . . . 27
BIBLIOGRAPHY. . . . . . . . . . . . . . 29
APPENDICES . . . . . . . . . . . . . . 30

ii

LIST OF TABLES

Table Page
1. Mean Per Cent of Recognition of Sentences
During Six Light Levels of Decreasing
Illumination Intensity . . . . . l8
2. Results of t_Test for Differences Between
Total Number of Sentence Identifications
for Six Light Levels. . . . . . . . 22

iii

CHAPTER I

STATEMENT OF THE PROBLEM

Introduction

 

It has been suggested that the first handicap of

l Man's need for communi—

deafness lies in communication.
cation with his fellow man is perhaps his greatest need

and the fulfillment of his other needs is greatly depen-
dent upon his ability to satisfy this basic one.2 Research
indicates that the average American spends 70 per cent of
his active hours communicating verbally-—listening, speaking,
reading, and writing, in that order.3 The individual with

a severe hearing loss is at a serious disadvantage. He is
unfortunately handicapped in the most frequently occurring
facet of communication-—listening.

Lipreading is a valuable substitute for hearing as

far as spoken language is concerned, but like other

 

lHallowell Davis and s. R. Silverman, Hearing and
Deafness (New York: Holt, Rinehart, and Winston, Inc.,
19505, p. xi.

 

2Ibid.

3David Berlo, The Process of Communication (New York:
Holt, Rinehart, and Winston, Inc., l9EO), p. l.

 

1

2

substitutes, it has its limitations.1 The rapid and fre-
quently obscure movements of speech itself presents a major
problem. Often unfavorable conditions of extraneous nature—-
uncertainties of light, position, movement, and distance of
the speaker-wmake lipreading unpredictable.2

The importance of one of these variables, light, and
the effect that it has on lipreading efficiency, cannot be
overestimated. The lipreader is perhaps more aware of the
effect of, and more dependent on the existing light condi-
tions, than any other environmental factors. The individ-
ual with normal hearing, when present in a situation where
his vision is obscured by shadows or darkness, can function
and communicate effectively by depending on auditory clues.
The hard of hearing person who must, in fact, ”hear with
his eyes" has, under the same circumstances, lost or
limited his only means for successful interaction. This
perplexing situation has been but generally commented on in
the literature. There is, however, no available research
on the effect that various conditions of light have on lip-
reading efficiency. Knowledge in this area is important in"
in the complete assessment of the handicap which a hearing
loss imposes and in the measurement of the extent to which

lipreading can substitute for hearing.

 

lElizabeth Nitchie, New Lessons in Lip Reading (New
York: J. B. Lippencott Co., 1950), pp. 25-26.

2Irving Fusfeld, ”Factors in Lipreading as Determined
by the Lipreader," American Annals of the Deaf, 102 (March,
1958), p. 240.

3

Statement of Problem and Purpose of Study

 

The problem from which this study arose is that of

determining the effect of various room lighting conditions

on the lipreading performance of trained subjects. The pur-

pose of this study is to measure the differences, if any,

in the lipreading performance of trained subjects during

varying intensities of room illumination.

Questions Posed at Onset

 

Questions posed at the onset of this study were:

1. Is there a decrease in over—all lipreading performance

as a function of decreased light level?

2. Is there a difference

obtained at the first

in performance between the score

light level (criterion score)

and the score obtained at the second light level?

3. Is there a difference
score obtained at the
obtained at the third

4. Is there a difference

score obtained at the

in performance between the
first light level and the score
light level?

in performance between the

first light level and the score

Obtained at the fourth light level?

5. Is there a difference
score obtained at the
obtained at the fifth

6. Is there a difference

obtained at the first

in performance between the

first light level and the score
light level?

in performance between the score

light level and the score ob-

tained at the sixth light level?

10.

ll.

12.

13.

14.

15.

Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained

obtained

4
a difference in performance between the
at the second light level and the score
at the third light level?
a difference in performance between the
at the second light level and the score
at the fourth light level?
a difference in performance between the
at the second light level and the score
at the fifth light level?
a difference in performance between the
at the second light level and the score
at the sixth light level?
a difference in performance between the
at the third light level and the score
at the fourth light level?
a difference in performance between the
at the third light level and the score
at the fifth light level?
a difference in performance between the
at the third light level and the score
at the sixth light level?
a difference in performance between the
at the fourth light level and the score
at the fifth light level?
a difference in performance between the

at the fourth light level and the score

at the sixth light level?

SCOPE

SCOPE

SCOTS

SCOPE

score

SCOPG

score

SCOPE

5

16. Is there a difference in performance between the score
obtained at the fifth light level and the score

obtained at the sixth light level?

Importance of Study

 

This study is considered basic research and the writer
does not seek immediate practical application either in the
training or the counseling of lipreaders. It is hoped that
this study will provide some basis for future investigations
of the effect of various environmental factors on lipreading
performance, and thus enable a more complete assessment of

the total limitations imposed by a hearing loss.

Definition of Terms

 

For the purpose of this study, the terms used are

defined in the following manner:

Snellen Vision Test.--A chart consisting of selected

 

letters of the alphabet that are pictured in varying sizes.
The size of the letter that the individual is able to read
at a certain distance from the chart enables the examiner
to obtain a measure of the individuals' visual acuity for

distance only.

Illumination.--The amount of luminous flux falling on

1

 

an object emitting no light of its own.

 

1H. A. E. Keitz, Light Calculations and Measurements
(London: Cleaver Hume Press, Ltd., 1955), p. 72.

6

Foot—candle.--One foot-candle is the illumination pro—

 

duced by one lumen falling on a surface one foot distance

away from the source.

Light meter.--The light meter is a direct measure of

 

luminance. It is used to determine the intensity of light

which it measures in foot-candles.

Light levels.—-Certain measures of illumination

 

intensity.

Lipreading.--The art of reading speech when the reader

 

is deprived of the acoustic component of the oral symbol.

Organization of the Thesis

 

Chapter I has contained the statement of the problem
which led to this study and the purpose for which it has
been conducted. It has included an introduction to the
topic, and stated the questions to be discussed. The im-
portance of the study has been discussed and the terms
which will be used have been defined.

‘Chapter II will contain a review of the literature
which pertains to this topic.

Chapter III will consist of a discussion of the
subjects, equipment, materials, and testing procedure
employed in this study.

Chapter IV will consist of a discussion of the results

of the study.

7

Chapter V will contain a summary and the conclusions

of the study.

CHAPTER II

REVIEW OF THE LITERATURE

, Factors Influencing Lipreading Efficiency

 

No one factor fully accounts for lipreading efficiency,
but rather a combination of them—-natural aptitude, easy
command of the English language, acquaintance with the
vagaries present in speech, and a large functioning vocab—
ulary.l These factors are variables relative to the individ—
ual lipreader and, in part, determine the degree of success
he may expect to achieve in oral communication. The
specific intellectual assets and the personality of the
lipreader are sometimes considered to be variables. The
research, however, reveals that there appears to be no
significant relationship between lipreading skills and
general intelligence.2 The relationship between lipreading
ability and certain behavioral or personality factors has
been investigated and supported by research.3 The sociable

person who takes an active interest in the people and

 

lFusfeld, op. cit., p. 236.

2John J. O'Neill and Herbert J. Oyer, Visual Communi—
cation for the Hard of Hearing: History, Research, and
Methods (Englewood Cliffs, New Jersey: Prentice-Hall, Inc.,
lgEI): p' 39. E '

31bid., p. do.

9
events around him usually makes more rapid progress in lip-
reading than does the shy, reserved, unimaginative individ-
ual.1 The alertness of the lipreader, his flexibility,
and his attitude, are likely factors that affect lipreading
ability.

The influence of the speaker and his message are im-
portant determinents of lipreading efficiency. The lip-
reader‘s acquaintance with the speaker, his knowledge of
the subject materials on which the conversation is based,
and the manner of address of the speaker contribute to the
over-all effectiveness of the verbal exchange.2 Other
contributing factors include the amount of animation with
which the talker is speaking, his speech mannerisms, the
manner in which he articulates, and the degree of under—
standing and consideration he possesses.

That the amount of the speaker's face which is visible
in addition to his lips is related to how efficiently and
individual can lipread is evidenced in the results of a
recent investigation. The results showed that there were
definite improvements in lipreading scores as more of the

face was visible.3 Certain factors in the environment

 

1Davis, 0 . cit., p. 365.

2Ibid.

3Edgar Lowell, "New Insight Into Lipreading,” Rehabili-
tation Record, II (July—-August, 1961), p. 4.

 

10
are assumed to be partly or wholly responsible for limiting
the visibility of the speaker's face. More specifically,
the unsatisfactory conditions of a poorly lighted room that
throws deep shadows on the speaker's face, or a seating
arrangement that forces the lipreader to look into a bright
light.1 Dr. Max Nelson of Michigan State University, in
reference to these specific environmental factors, states
that "the literature has constantly and continually
advocated that lipreading be done under very clear lighting
conditions. The reasons are the assumed negative influ-
ences of poor lighting conditions on the ability to lip—

read.”2

The Need for Investigations of the
Relationship Between Light
and Lipreading
Although the literature does assume that the intensity
and direction of the light in the immediate environment has
its effects on lipreading efficiency, there is no published
research in this area.3 Dr. Herbert Oyer of Michigan State
University states that "for all practical purposes one

would want a well lighted room as it would seem that shadows

and dim light would have a deleterious effect on lipreading,

 

1Davis, op. cit., p. 364.

2Interview with Dr. Max Nelson, associate professor,
Michigan State University, July 10, 1962.

3O'Neill and Oyer, op. cit., p. 44.

ll
but the importance of a well lighted room is open to ques-
tion and experimentation."1 In one instance in the liter-
ature it has been urged that controlled studies should be

made of the effects of lighting on lipreading performance.2

 

1Interview with Dr. Herbert J. Oyer, Director of Hearing
and Speech Clinic, Michigan State University, July 10, 1962.

2
O'Neill and Oyer, op. cit.

CHAPTER III
SUBJECTS, EQUIPMENT, MATERIALS, AND PROCEDURE

Viewer Subjects
Ten male and ten female adults were subjects in this
study. One additional female subject was used for compar-
ative purposes. The subjects' educational background
varied from one to six years of college. No previous lip-
reading experience was required of these subjects, although

the potential ability to lipread was necessary.

Speaker Subject
One female adult presented the test sentences to all

training and experimental groups.

Equipment

 

Amplifier (Ampex Amplifier—Speaker, Model 620)
Light fixture (Eight mounted, lOO-watt bulbs)

Light meter (Spectra Gold Seal Exposure Meter,
Model 8-500)

Miscellaneous head sets
Reostat (Powerstat, Model 116)

Sound pressure level meter (M. S. A. Soundscope,
Model B)

White noise generator (Grason-Stadler White Noise
Generator, Model A55b)

12

13
Vision test (Green Test Cabinet, Model 20)
Tape recorders (Telectro-Tape, Model 5953, Wollensack,
Model T1500)

Materials

 

Thirty test sentences were devised by the examiner
for use in this study. Each sentence contained six
syllables. A total of A84 speech sounds was contained in
these thirty sentences. The examiner intentionally
attempted to create a comparable ratio of the frequency of
individual speech sounds occurring in these sentences with
the relative frequency of occurrence of speech sounds com-
iled by Dewey.l These sentences may be found in Appendix
B, page 33. The comparison of the frequency of occurrence

of the speech sounds is reported in Appendix C, page 34-

Procedure

 

1. Vision screening.--Each subject was given a

 

Snellen Vision Test for both eyes in order to test their
visual acuity as far as distance was concerned. This test
~was given to each subject individually prior to training.
Any individual who was unable to pass the vision test at
20/30 was eliminated from the study.

2. Orientation to the test situation.--Each subject
was presented with a list of the thirty test sentences

several days prior to the initial training sessions. They

 

lHarvey Fletcher, Speech and Hearing in Communication
(Princeton: D. Van Nostrund’Inc., 1961), p. 9II

14
were instructed to familiarize themselves with the content
of each sentence and, if possible, practice lip reading
them.

3. Training of subjects.--The training sessions took

 

place in a room in which no exterior light was present.

The room lighting conditions for training were held con-
stant at the face of the speaker at thirty foot-candles as
measured by the light meter. White noise was present at

an intensity deemed sufficient to mask auditory clues from
the speaker. The intensity of the white noise measured
from 84 to 86 decibels at the subjects' seats. Unconnected
headsets were worn by the subjects as further insurance
against auditory clues.

During the training session, the sentences were
presented to the subjects in random order by the speaker.
The subjects were instructed to repeat the sentences orally
in a soft voice. The speaker was thus able to quickly
identify the subject or subjects who had had difficulty with
a particular sentence. The sentence was then repeated until
the subject or subjects felt familiar with the visual clues
that the sentence afforded. The speaker kept a record of
the number of incorrect identifications made by each subject.

The presentation of the complete list of test sentences
was the final exercise in the training session. These
sentences were written by the subjects and immediately scored

by the speaker. Those subjects who successfully identified

15
95 per cent (28) of the sentences, were deemed eligible to
participate in the experiment. The training sessions never
exceeded one hour and fifteen minutes.

4. Testing procedure.--The experiment was conducted

 

in the same room in which the subjects had been trained
within 48 hours after the training session. There was no
source of light present in the room other than that provided
by the light fixture designed for the experiment. This
light fixture was suspended 26 inches above the speaker's
head. The subjects were seated in a straight row in front
of the speaker. Since the experiment was conducted on six
separate occasions, there were never more than five subjects,
including the female control subject, present in any one
group. The distance between the speaker and the subjects
varied from seven feet eight inches to eight feet two
inches. The speaker stood behind a rostrum during the ex—
periment. The subjects were seated in the same seating
arrangement in which they had been trained.

White noise was present in the room at the same inten-
sity as was used for the training session (84 to 86 decibels).
The inactive headsets were worn by all subjects.

The subjects were presented the thirty test sentences
for review purposes preceeding the actual test situation.
Immediately preceeding the actual experiment, the subjects
were furnished with paper and pencils and were instructed
to record every sentence that they were able to lipread

during the succeeding light levels.

16

During the experiment, the thirty test sentences were
presented randomly at each light level. Six light levels
were employed: 30 foot-candles, l6 foot-candles, 8 foot-
candles, 4 foot-candles, 2 foot—candles, and 1/2 foot-
candle. There was a five minute break between light levels
during which the subjects experienced the next light level
at which they were to be tested. The succeeding light levels
gradually decreased the amount of room illumination.

The same speaker presented the test sentences to all
training and experimental groups. She wore identical
clothing and makeup for each group. An undecorated medium
grey blouse was worn by the speaker in order to prevent
distraction or possible glare that might arise from cloth
of a brighter color.

The speaker attempted to control, to the best of her
ability, unnecessary ”mouthing” of words and/or unnecessary
facial expressions.

In order to offer some insurance that the speaker's
behavior did not vary significantly from group to group,
one female subject voluntarily participated in five of the
six experimental groups. Her scores are reported in Appen-
dix, page 32.

5. Recording and scoring.—-Each subject's list of

 

sentences for each light level was graded and recorded
separately. A sentence was considered correct if four of

the six syllables corresponded with the original sentence.

17
This two-syllable error allowance was designed to give some
leeway for common errors such as confusion of articles or

word endings.

CHAPTER IV

RESULTS AND DISCUSSION

Results

The number of correct sentences identified by each
viewer subject at each light level was tabulated and
recorded. The per cent of correct identifications at each
of the six light levels was analyzed and tabulated. Table
1 presents the data obtained in response to question one
posed in Chapter I, namely:
1. Is there a decrease in over-all lipreading performance

as a function of decreased light level?

TABLE 1

MEAN PER CENT OF RECOGNITION OF SENTENCES
DURING SIX LIGHT LEVELS OF DECREASING
ILLUMINATION INTENSITY

I —
—.—.— .—‘

Light level 1 2 3 4 5 6
Illumination
intensity 30fc* l6fc 8fc 4fc 2fc 1/2fc

Per cent of

recognition 96.83 96.16 94.66 94.66 94.50 88.83

 

*fc = foot-candles.

18

19

A comparison was made of the differences between the

total number of identifications made at each of the several

light levels in response to questions two through sixteen

posed in Chapter I.

2.

The questions were as follows:

Is there a difference in performance between the score

obtained at the first light level (criterion score)

and the score obtained at the second light level?

Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained

obtained

a difference in performance between the

at the first light level and the score
at the third light level?
a difference in performance between the
at the first light level and the score
at the fourth light level?

a difference in performance between the
at the first light level and the score
at the fifth light level?

a difference in performance between the
at the first light level and the score
at the sixth light level?

a difference in performance between the
at the second light level and the score
at the third light level?

a difference in performance between the

at the second light level and the score

at the fourth light level?

SCOPE

SCOPE

SCOPE

SCOPE

SCOPE

SCOPE

lo.

11.

12.

l3.

14.

15.

16.

Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained
obtained
Is there
obtained

obtained

20
a difference in performance between the score
at the second light level and the score
at the fifth light level?
a difference in performance between the score
at the second light level and the score
at the sixth light level?
a difference in performance between the score
at the third light level and the score
at the fourth light level?
a difference in performance between the score
at the third light level and the score
at the fifth light level?
a difference in performance between the score
at the third light level and the score
at the sixth light level?
a difference in performance between the score
at the fourth light level and the score
at the fifth light level?
a difference in performance between the score
at the fourth light level and the score
at the sixth light level?
a difference in performance between the score
at the fifth light level and the score

at the sixth light level?

To analyze the data obtained in response to questions

21

two through sixteen the following formula was employed:1

Ml - M2

t =
'— y/r .zld2
N (N-l)

 

 

The results of this analysis are presented in Table 2.

Discussion

 

The per cent of recognition does not differ greatly}\
among the six light levels; however, it does appear to de-
crease as a function of decreased light level, thus indi-
cating a tendency for lipreading efficiency to decrease as
illumination intensity decreaSes. Table 1 shows the mean
per cent of correctly identified sentences for each light
level.

The subjects were able to correctly identify 96.83
per cent of the sentences during the first light level and
96.16 per cent of the sentences during the second light
level. The comparison of the differences between the scores
for these two light levels indicated no significant differ-
ence. The intensity of illumination had been decreased from
thirty foot—candles (first light level) to 16 foot-candles
(second light level).

A comparison of the difference between the first and
third light levels revealed no significant difference. The

per cent of recognition for the light level was 96.83 per

 

1E. F. Lindquist, Statistical Analysis in Educational
Research (New York: Houghton Mifflin Co., 1940), p. 59.

22

TABLE 2

RESULTS OF £_TEST FOR DIFFERENCES BETWEEN
TOTAL NUMBER OF SENTENCE IDENTIFICATIONS

 

 

FOR SIX LIGHT LEVELS

Light Light Level of
Level Mean Level Mean p score Confidence
LLl 29.05 LL2 28.85 .222 Non sig. at .05
LLl 29.05 LL3 28.40 .221 Non sig. at .05
LLl 29.05 LLn 28.40 .211 Non sig. at .05
LLl 29.05 LL5 28.35 .227 Non Sig. at .05
LLl 29.05 LL6 26.65 .229 Non sig. at .05
LL2 28.85 LL3 28.40 .227 Non sig. at .05
LL2 28.85 LLu 28.40 .227 Non sig. at .05
LL2 28.85 LL5 28.35 .226 Non sig. at .05
LL2 28.85 LL6 26.65 .296 Non sig. at .05
LL3 28.40 LL“ 28.40 .000 Non sig. at .05
LL3 28.40 LL5 28.35 .398 Non Sig. at .05
LL3 28.40 LL6 26.65 .229 Non sig. at .05
LL“ 28.40 LL5 28.35 .398 Non sig. at .05
LL” 28.40 LL6 26.65 .229 Non sig. at .05
LL5 28.35 LL6 26.65 .229 Non sig. at .05

 

23

cent and the per cent of recognition for the third light
level was 94.66 per cent. The illumination intensity was
decreased from thirty foot-candles to eight foot-candles.
These results are applicable to the difference between the
first light level and the fourth light as the per cent of
recognition was identical for both the third and fourth
light levels. The intensity of illumination had been de-
creased to four foot-candles at the fourth light level.

At the fifth light level, 94.5 per cent of the
sentences were correctly identified as compared to 96.83
per cent recognition at the first light level. The p_test
revealed no significant difference between these two light
levels. The intensity of illumination during the fifth
light level was two foot-candles.

No significant difference was found between light
levels one and six. Illumination was one—half foot-candle
at the sixth light level; 88.83 per cent of the sentences
were identified by the subjects at this light level. It
was reported by the subjects that they experienced difficulty-
in writing the sentences at this light level as it was ”too
dark to see the paper.” The speaker had difficulty locating
and identifying the sentences to be read.

No significant differences were found when any other
combination of light levels were analyzed.

One must remember in studying the above data that

trained subjects were used. They were highly familiar with

24

the content of each sentence and with the specific visual
clues that each sentence afforded. The subjects remarked
that they felt highly motivated to identify the sentences
presented at the two light levels of lowest illumination
intensity. It was observed that the subjects made a more
active, conscious effort to attend to the visual clues
during these last two light levels.

The results of this study indicate that there is some
question as to whether or not lipreading efficiency notice-
ably decreases as a function of decreased room illumination
intensity when the lipreader is highly familiar with the
content of the message.

Questions concerning reasons for these results can
only be answered hypothetically at this time. The manner
in which this study was conducted seemed to present a
challenge to the subjects. Some subjects explained that
they were motivated by a desire to ”beat the test." Some

inter—group competition was observed.

CHAPTER V

SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS

Summary

Lipreading, as a substitute for hearing spoken lang-
uage, has its limitations. The evaluation of the effects
that various personal and environmental factors have on
lipreading efficiency is important in the measurement of
the extent to which lipreading can substitute for hearing.
An important variable affecting the lipreader's performance
is the direction and intensity of the environmental lighting
conditions. Although it has been conceded that lighting
conditions are an ever-present variable and can cause the
lipreader much difficulty, the actual effects of lighting
conditions has not been previously investigated.

The purpose of this study has been to measure the dif-
ferences in the lipreading performance of trained subjects
during varying intensities of room illumination in an effort
to determine the effect of various room lighting conditions
on the lipreading efficiency of these subjects.

This study does not have immediate practical appli-
cation, but may provide some basis for future investigations
of the effects of various environmental factors on lipreading

performance.

25

26

A review of the literature concerning the variables
affecting lipreading performance indicates that, while poor
lighting conditions are generally acknowledged as having a
negative influence on lipreading efficiency, there has been
no reported research on this topic.

The viewer subjects for this study were ten male and
ten female adults. One additional female viewer subject
was present at five of the six group experiments. Her scores
were used for comparative purposes. One female speaker sub-
ject presented the test sentences at all training and experi-
mental groups. The Snellen Vision Test was employed as a
screening device for subject selection. The subjects were
trained to lipread a list of thirty test sentences. The
sentences used were designed to contain speech sounds in an
approximate ratio to sounds in their frequency of occurrence
in the American language. During training the room lighting
conditions were held constant at thirty foot-candles. White
noise which measured from 84 to 86 decibels and inactive
head sets worn as muffs were used to mask auditory clues.
Those individuals who could not correctly identify 95 per
cent of the sentences after the training session, were
eliminated from study.

During the actual experiment the speaker presented
the thirty sentences during each of six varying room
lighting conditions. The six light levels were thirty foot-

candles, sixteen foot-candles, eight foot-candles, four

27

foot-candles, two foot-candles, and one—half foot-candle,
as measured by the light meter at the speaker's face. In-
active head sets were again worn by the subjects, and white
noise was present for the purpose of masking auditory clues
from the speaker. The subjects responded in written form
to the test sentences.

The findings of this study indicate that lipreading I:
efficiency of trained subjects tends to decrease, but not

significantly, as the intensity of room illumination decreases.

Conclusions

 

l. The lipreading efficiency of individuals who are
highly familiar with the content of the message decreases
only slightly, and not significantly, as the intensity of
room illumination decreases.

2. Lipreaders, who are highly familiar with the con-
tent of a message, when sufficiently motivated will lipread
with an adequate degree of efficiency under the most adverse

lighting conditions.

Recommendations

 

1. An investigation of the recognizability of word
lists by trained and/or untrained subjects under varying
room lighting conditions.

2. An investigation of the lipreading efficiency of
subjects involved in conversation during varying intensities

of room illumination.

28

3. An investigation of the lipreading performance of
untrained subjects identifying sentences at various room
light levels.

4. An investigation or the recognizability of speech
'sounds during varying intensities of room illumination.

5. An investigation of the lipreading performance
of trained or untrained subjects identifying sentences or
words under light levels measuring from one foot-candle to

zero foot-candles.

APPENDICES

3O

RAW SCORES OF CORRECTLY IDENTIFIED SENTENCES
DURING SIX LIGHT LEVELS

APPENDIX A

 

 

 

 

Subject
Number LLl (30 fc) LL2 (16 fc) LL3 (8 fc)
l 29 29 26
2 30 28 29
3 28 29 28
4 29 30 27
5 28 26 24
6 28 27 25
7 3O 30 30
8 30 30 30
9 29 30 30
10 30 29 29
ll 28 27 26
12 28 28 28
13 28 29 29
14 30 29 30
15 30 30 30
16 30 30 30
17 3O 30 29
18 28 29 29
19 30 28 30
20 28 29 29
LL4 (4 fc) LL5 (2 fc) LL6 (1/2 fc)
1 2g 30 28
2 2 30 29
3 29 30 28
4 28 28 22
5 25 24 23
6 26 24 20
7 30 3O 3O
8 29 28 25
9 29 29 25
10 29 29 26
ll 28 26 26
12 26 26 26
13 28 28 29
’14 29 29 29
15 30 30 30
16 30 30 30
17 29 26 25
18 29 30 26
19 30 30 30
20 29 30 26

 

31

32

RAW SCORES OF SUBJECT ATTENDING
FIVE EXPERIMENTAL GROUPS

 

 

 

LL1 LL LL LL 4 LL LL
29 29 22 26 24 25
30 3O 30 30 29 28
29 30 29 30 3O 29
3O 3O 30 3O 30 29
30 30 30 30 3O 30

 

"__T'E

 

!:_

O O O O O O

\OCD\H%U1#XUR)H

APPENDIX B

TEST SENTENCES

His last look was costly.
The log is very hot.

We baked no bread today.
Our room was filled with noise.
I played a game of gin.

Your vision is not good.

The day was much too short.
The card fell from the pack.
He has a cute grandson.

In spring we play outside.
Is this the road around?
Your voice seemed very loud.
Have you seen this movie?

A cat must have nine lives.
This is a thin penny.

They just ate red jello.

Do get some mint candies.

I just sat on a tack.

It is found in that nook.
This has been a pleasure.
Let's go for a quick swim.
The price is one nickle.

A cough broke the stillness.
She put on both the lights.
Fred's mom dropped an armchair.
Their lunch is being served.
Fran's glad she went to Flint.
Yes I just saw your dog.

I will stop very soon.

Start every gong ringing.

33

 

A COMPARISON OF THE FREQUENCY OF OCCURRENCE OF

APPENDIX C

SPEECH SOUNDS OCCURRING IN TEST SENTENCES

WITH THEIR RELATIVE FREQUENCY OF
OCCURRENCE IN AMERICAN
CONVERSATION

 

SpeechYSound

Test Sentences

Conversational

 

u (tool)
6' tone;
0 talk
a top)
a' tape)
2' eat)
u took)
9 ton
a tap

e ten

i tip
1' dike)
ou our)
oi oil)
w

y

h

l

r

m

n
ne

v

2

th (then)
zh

f

s

th (thin)
sh

b

d

J

S

p

t

ch

k

\1

HmmHNOmmmmwmrwmHommmwrommmppwwmmmw
H

H

UL)

ewe

{$30mel—‘I—J-P’F—‘OOU'IHOWWMHOWMQWHHMOOHCDUUWQMHH
i—'

O O O O O O O O

H

O O O I O O

LOOKOHl—‘O-F-‘OOO-P‘h)OMMHHCDWO-PMMWOOMCNMMOHMI—‘mI—‘l—‘ID
ONUJCDNIUTUOONCBNNOOOUWmml—‘Hml—JONMH‘QIUCh-P'UONI—‘OOHQHUUU'IO

'fi

 

LU
:-

p- 29 (”"9511“)! ”It 9 “ “TC 5" “Cd, Hr) j...‘2r'f}k'~/

 

__.

. l
Ci}. _.I, '__.