ABSTRACT
THE PERFORMANCE OF AUDITORY PERCEPTUALLY HANDICAPPED
CHILDREN ON TEMPORALLY DISTORTED SENTENTIAL
APPROXIMATIONS AND NORMAL SENTENCES
By

Gail Ann Overholt

The effect of temporal characteristics on the intel-
ligibility of the signal has been of interest to researchers
and clinicians in recent years. Several investigators have
found that distortion of these temporal characteristics
increases the load placed on short-term memory resulting
in decreased accurate perception of the stimulus. One such
method of investigation is to control stimulus duration and
vary the silent interstimulus interval using sentential
stimuli controlled for length and syntactic and semantic
contstraints.

Current theories of perceptual abilities in children
with auditory perceptual handicaps suggest that these chil-
dren do not process auditory stimuli in the same manner as
normal children. In turn, children'with perceptual prob-
lems may demonstrate perceptual strategies different from
those used by normal children on a task designed to over-
load auditory short-term memory and affect information pro-

cessing strategies.

Gail Ann Overholt

The purpose of this investigation was to study the
effect of sentence length, order of sentential approxima-
tion, and modified interstimulus interval upon linguistic
processing by children with auditory perceptual handicaps.
The results were compared to data obtained with the same
task on normal children.

In order to increase syntactic and semantic complexity,
ten three—word and ten five-word first order sentential
approximations, ten three-word and ten five-word second or-
der sentential approximations, and ten three-word and ten
five-word normal sentences were constructed. The sentential
approximations and normal sentences were read by a male
speaker from a Second Level standardized reading list of
monosyllabic words. These forty sentential approximations
and twenty normal sentences were placed into three experi-
mental conditions, each condition with a specific silent in-
terstimulus interval size. Word duration remained constant
(normal speaking rate), but three interstimulus interval
sizes (a normal interstimulus interval, 200 msec, and "

#00 msec) were used.

Twelve elementary school students enrolled in the Re-
medial Reading Center at Michigan State University served
as subjects. All children had normal hearing as assessed
by audiological evaluation. Each subject was presented
with twenty sentences of each interstimulus interval, twen-
ty sentences of each order of sentential approximation, i.e.

synthetic sentences (first order, second order, and normal

Gail Ann Overholt
sentences), and thirty sentences of each sentence length
(three-word and five-word), for a total of 60 sentences per
subject. The tape recorded material was presented under
earphones to each subject in a two—room, double-walled,
sound treated booth.

Examination of responses consisted of the percent cor-
rect words recalled. Words substituted for, omitted, or
repeated in the wrong order were considered errors. The
results of this investigation demonstrated that auditory
perceptually impaired children depended upon sentence length,
order of sentential approximation, and interstimulus inter-
val size in processing Speech and language.

Recall accuracy scores decreased as sentence length
increased from three words to five words, as order of sen-
tential approximation decreased from normal sentences to
second order and first order approximations, and as the size
of the interstimulus interval increased from a normal con-
dition to 200 and #00 msec, respectively.

Interactions of the experimental conditions revealed
that for three—word sequences, recall accuracy varied only
slightly with increasing sentential approximations and in-
terstimulus intervals. However, there were substantial
differences in recall scores for five-word sequences as sen-
tential approximation and interstimulus interval increased.
Also, for normal sentences, temporal alteration in either

form of increasing sentence length or increasing

Gail Ann Overholt

interstimulus interval size did not affect recall accuracy,
which was near 100 percent.

These findings were discussed in terms of theories of
speech and language perception in normal and auditory per-
ceptually impaired children. Particularly important were
implications to auditory diagnosis of and language deve10p-

ment in auditory perceptually handicapped children.

THE PERFORMANCE OF AUDITORY PERCEPTUALLY HANDICAPPED
CHILDREN ON TEMPORALLY DISTORTED SENTENTIAL
APPROXIMATIONS AND NORMAL SENTENCES

By

Gail Ann Overholt

A THESIS

' Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

MASTER OF ARTS

Department of Audiology and Speech Sciences

1976

ACKNOWLEDGMENTS

I would like to thank the members of my committee who
contributed to this thesis with their knowledge and exper-
ience in their various areas of specialty: Ms. Janis For—
bord in diagnostic audiology, Dr. Frank Brfino in learning
disabilities, and Dr. Daniel S. Beasley, committee chairman,
in speech perception. My sincerest appreciation goes to Dr.
Beasley, who gave me the impetus to begin this study and the
guidance to finish it, for the confidence and assuredness he
has instilled in me during my education at Michigan State.

I would also like to express my thanks to Dr. Byron van
Roekel, director of the Remedial Reading Center, for his
cooperation and interest in obtaining subjects; and to the
children who served as subjects and their parents, who took
the time from their busy schedules to participate in this
study.

Ifd also like to thank my good friend Dr. Barry Freeman
for the time and effort he took in assisting me in this work
and in my career.

Finally, I'd like to thank my mother for the long hours

she Spent typing and trying to understand this work.

11

TABLE OF CONTENTS

LIST OF TABLESOOOOO0.0.000...OOOOOOOOOOOOOOOOOOOOOO

LIST OF FIGUESOIOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

CHAPTER
I.

II.

III.

INTRODUCTIONOOOOOOOOOOO00......O.000......

A Model of Short-Term Memory.............
Sentential Stimuli.......................
Distorted Speech Stimuli.................
Auditory Perception in Children..........
Learning Disabled Children...............
Statement of the Problem.................

EXPERIMENTAL PROCEDURES.OOOOOOOOOOOOOOOCOO

subJeCtSOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOIOO
Stimuli..................................
Recording and Stimulus Generation
Procedures..........................
Counter-Balancing and Randomization
Procedures..........................
Presentation Procedures..................
Analysis.................................

mSULTS..0...00......OOOOOOOOOOOOOOOOOOOOO

Effect of Sentence Length................

Effect of Order of Sentential
Approximation.......................

Effect of Interstimulus Interval.........

DISCUSSIONOO00......OOOOOOOOOOOCOOOOOOOOOO

Comparison of Trends to Previous
Investigations......................
Comparison of Interactions to Previous
Investigations......................
Types of Errors..........................
Theoretical Implications.................

Implications for Audiology...............‘

Implications for Language Development....
Implications for Future Research.........

111

A0
A2

A6
#7
#8
A9

 

V
, «AI—5'

CHAPTER
V. SUMMARY AND CONCLUSIONS...................
APPENDICES
APPENDIX A
Criteria for Normal Hearing..............
APPENDIX B

Revised Word List Basal vocabulary
On We Go Houghton Mifflin (1966)....

 

APPENDIX C

Standardized Verbal Instructions.........
APPENDIX D

Response Forms...........................
APPENDIX E

Mean Percent Correct Scores..............
APPENDIX F

Ranges of Subjects' Percent
correCt scoreSOOOOOOOOOOQOOOO0......

LIST OF REFEmNCESOOOOOOOOOOOOOOOOOOOOOOOO0.0.0....

iv

50

53

SH

55

56

62

6A
55

TABLE

LIST OF TABLES

Mean percent correct scores of normal
reading (NR-BAP) and reading impaired
children (RI—BAF) in the Freeman and
Beasley (1976) study and of subjects
in this study (RI-GAO) for the main
effects of sentence length (three-
word (3) and five-word (5)) and sen-
tential approximation (first order
(1), second order (2), and normal
sentences (N))...........................

Mean percent correct scores for inter-
stimulus interval by order of senten-
tial approximation and sentence length...

Mean percent correct scores for senten-
tial approximation by interstimulus
interval and sentence length.............

Ranges of subjects' correct scores for
sentential approximation by inter-
stimulus interval and sentence length....

Ah

FIGURE
1.
2.

3A.

3B.

in.

AB.

SA.

SB.

LIST OF FIGURES

Experimental deSignO000......00000000000000

Main effects of sentence length, sen-
tential approximation, and interstim-
ulus intervalOOOOOOOOOOOOOOOOOOOOOOOOOOO

Mean percent correct scores per sentence
length and sentential approximation
(3- first order;[§- second order; E]
- normal sentence)...00.000000000000000

Mean percent correct scores per sentence
length and interstimulus interval
(3- normal interval;[§- 200 msec;[]
- 400 msec)............................

Mean percent correct scores per senten-
tial approximation and sentence length
«3- three words;[j- five words).........

Mean percent correct scores per senten-
tial approximation and interstimulus
interval «3- normal interval;

A. 200 msec;[]- uoo msec)...............

Mean percent correct scores per inter-
stimulus interval and sentence length
(3- three wordsij- five words)..........

Mean percent correct scores per inter-
stimulus interval and sentential
approximation «3— first order;A-
second orderij- normal sentence)........

Mean percent correct scores per inter-
stimulus interval, sentence length
(open symbols - three words; filled
symbols - five words), and sentential
approximation (circles — first order;
triangles - second order; squares -
normal sentence)........................

vi

23

32

.33

33

35

35

37

37

39

CHAPTER I
.INTRODUCTION

Several investigators have shown that the auditory per-
ception of Speech is at least in part dependent upon the tem-
poral characteristics of the stimuli (Aaronson, 1967; Beasley
and Shriner, 1973; Hirsh, 1959; Shriner and Daniloff, 1970;
Speaks and Jerger, 1965). These temporal characteristics in-
clude rate and duration, and interact with stimulus frequency
and intensity, as well as with the listener's perceptual stra-
tegies and memory storage capacities and abilities. The in-
teraction of these several factors affect the intelligibility
of the signal, which is necessary before the message can be
perceived and comprehended adequately.

It has been found that the perceptual abilities of child-
ren are different from those of adults, and that these abil-
ities are develOpmental in nature (Beasley and Flaherty-Rin-
telmann, 1976; Beasley, Maki, and Orchik, 1976; King and Wes-
ton, 197“; Thompson, 1973). Further, there are some children
whose auditory perception is not like that of the majority of
children (Katz, 1971; Manning, Johnston, and Beasley, 1975;
Lerner, 1971; Rosenthal,‘l97u). The children in this latter
catagory have been labeled auditory perceptually handicapped
or specific learning disabled. It has been hypothesized that

2 .
the difficulties in auditory perception these children en—
counter may be due to problems of short-term memory storage
(Bryan, 1972; Manning, Johnston, and Beasley, 1975; Senf and
Fruendl, 1971). If this is the case, these children would
show results that varied from normative data on a test pur-
ported to assess short-term memory function. The purpose of
this study was to investigate the performance of learning
disabled children characterized by auditory perceptual prob-
lems on a measure purported to reflect upon the functioning
of short-term memory, in which sentence length, word sequence,
and interstimulus interval were covaried. This data was
compared to data obtained with the same measure on a group
of normal children.

A Model of Short-Term Memory

 

In order for a message (stimulus) to be perceived, it
must first enter into a memory storage system of some type.
Most current theories of short-term memory discuss the idea
that perception occurs in at least two stages (Aaronson,1967;
Broadbent, 1958). The first stage, the sensory stage, is a
.parallel processor, whereby more than one stimulus can be
dealt with at one time. Characteristic of this stage is its
parallel processing capabilities, large capacity, and rapid
storage decay. The second stage, the perceptual stage, is
characterized by series processing (that is, stimulus items
can enter this stage only one at a time), small capacity,
and long storage decay time. Broadbent (1958) hypothesized

that stimulus items first enter into the sensory system at a

3.
rapid rate, and then are transferred to the perceptual system
at a slower rate. This second stage is believed to occur at
higher perceptual levels. Based on this theory, it may be
hypothesized that problems in perception arise if items are
delayed in the first stage (sensory system) for too long a
period of time, since this is a rapid decay system, or if
items arrive at the second stage (perceptual system) at too
fast a rate to be processed accurately.

Aaronson (1967) reviewed literature which suggested that
the temporal aSpects of the message have an important effect
in short-term memory tasks. From this review she revised
Broadbent's short-term memory model. It appeared that vary-
ing the stimulus duration alone resulted in fewer number of
items being correctly recalled in a listening task. However,
when the rate of the message was changed by keeping stimulus
duration constant and varying only the interstimulus interval,
order of the stimulus items was less correctly perceived for
shorter interstimulus intervals. Aaronson, therefore, sug-
gested that the physical qualities of the stimulus were pro-
-cessed in the first stage of short-term memory, simultaneous
with stimulus presentation; however, the order in which the
stimuli are recalled were processed in the second stage.

This second stage was believed to be dependent in part upon
the length of the interstimulus interval.

Aaronson, Markowitz, and Shapiro (1971) studied this
phenomenon. They kept presentation rate constant, but com-

pressed the stimulus duration by 33%, thereby increasing the

a

length of the interstimulus interval. Aaronson et al.(l971),
found that the number of item errors for the normal stimulus
duration condition was similar to the number of item errors
in the compressed stimulus duration condition. However, for
the normal condition, which had a smaller interstimulus in-
terval, the number of order errors was significantly larger.

Miscik, Smith, Hamm, Deffenbacher, and Brown (1972)
also found that an increased interstimulus interval elicited
increased recall in a short-term memory task for digits. In
another experiment, Miscik et al. again studied the effect of
interstimulus interval upon perceptual processing, as well as
the type of encoding process used by the subject. They found'
that when subjects were asked to combine digits into digit
pairs and triplets, recall accuracy increased. The authors
hypothesized that combining digits in such a manner decreased
the number of areas in the central auditory nervous system
where this information was stored, and that retrieval from
this reduced number of areas was easier than from a larger
number of areas.

Some limitations on the capacity of short-term memory
were discussed by Miller (1956). He reported that human sub-
jects were able to accurately recall approximately seven pre-
sented items or units. When more items were presented, re-
call accuracy declined. However, even though short-term
memory limits could not be extended beyond these seven units,
Miller discovered that memory for items could be increased
by combining and increasing the item information per unit.

He labeled this perceptual strategy "chunking". Miscik

5
et al. (1972) stated, "Error in short—term recall may be due
not only to forgetting caused by inadequate rehersal, for
example, but also to inadequacies in the encoding process
(p. 151)."
Sentential Stimuli

 

Assessment of auditory processing abilities using con-
ventional methods has not resulted in adequate measurement of
the functioning of the central auditory system. The use of
pure tones, clicks, digits, and undistorted word lists ap-
pears to be too simple a task for the detection of lesions
in the patient with a central auditory problem (Bocca and
Calearo, 1963; Calearo and Lazzaroni, 1957; Jerger, 1960;
Willeford, 1967). These nonsentential stimuli principally
reflect peripheral mechanism functioning, while leaving the.
central auditory mechanism essentially unassessed. Bocca
and Calearo (1963) and Willeford (1967) concluded that pure
tones, even when used in a difficult listening situation,
were not adequate for the purpose of localization of the site
of lesion in a central auditory disorder. They concluded
that speech signal tests were more efficient in central audi-
tory functioning assessment.

In identifying the type of stimuli that will adequately
assess the central auditory system's functioning, Jerger
(1960) discussed two principles. The first is the "sublety
principle," that is, as the lesion becomes more centrally
located in the central auditory system, the stimuli needed to

detect this lesion must necessarily become more complex. The

Ali lllillI.

6.
second principle eSpoused by Jerger was the "bottleneck
principle." He suggested that speech signals encounter a
bottleneck in the region of the VIIIth nerve and the lower
brain stem. Consequently, if the lesion occurs in this
location, or more centrally, the patient will tend to show
normal responses to pure tones and clicks, but his ability
to understand speech will be dramatically limited.

Intrinsic redundancy refers to the continual crossing
and interaction of neuronal networks in the auditory path-
way. In this way, a lesion of the central auditory system
would have to be quite severe before any noticable problem
in communication would occur. Indeed, it has been stated
that an entire temporal lobe could be removed or destroyed,
and the effect that this would have on perception would be
unnoticed unless a complex auditory task was used to detect
the lesion (Willeford, 1967).

Extrinsic redundancy refers to the redundancy of the
Speech message per se. Harris (1960) defined this as the
syntactic and contextual cues of the message, the "....func-
tions of language organization (p. 228)." For example, three
or four words may not be heard, but the listener will compre-
hend the message because he knows the syntactic and semantic
constraints of the language. Related to extrinsic redun-
dancy is the concept of multiple-cueing.

Multiple-cueing relates to the physical characteristics
of the message. In a complex message, such as speech, there

are several characteristics present, including frequency,

7.
intensity, and duration. These characteristics, individual-
ly, have a broad range of values that may be present or ab-
sent from the Speech signal. It is these cues and the in-
teractions of these cues relative to the whole message that
are processed by the central auditory system. Distortion or
disruption of the physical characteristics of the speech
signal would not significantly reduce the intelligibility of
the message unless the signal was greatly degenerated (Harris,
1960). When extrinsic redundancy and the effects of multi-
ple-cueing are combined, it becomes possible to process a less
than perfect message if the central auditory system is intact.

Distorted Speech Stimuli

 

The patient with a central auditory lesion (reduced
intrinsic redundancy) is able to process speech because of
extrinsic redundancy and multiple-cueing. However, when
extrinsic redundancy and multiple-cueing are decreased in the
speech stimuli, these subjects may Show reduced message com-
prehension.

Investigators have found several ways to reduce re-
dundancy and multiple-cueing. Bocca and Calearo (1963) stated
that one method used to reduce redundancy (multiple-cueing)
was to filter the Speech with high-pass or low-pass filters.
Matzker (1962) devised a test in which a high-pass filtered
message was presented to one ear and the same message, low-
pass filtered, was presented to the Opposite ear. The person
with an intact brain stem could integrate and understand the

message, whereas the patient with a brain stem lesion could

not.

Another method of reducing redundancy is via time-com-
pressed Speech. Calearo and Lazzaroni (1957) described
three techniques for altering the temporal characteristics
of language: \

(l) have a Speaker talk faster.

(2) play a magnetic tape recording back at a higher

Speed than was used in the recording process.

(3) use a Special device that accelerates Speech with-

out changing frequency.
Calearo and Lazzaroni discussed the ability of patients with
confirmed temporal lobe desions to comprehend these three
types of time—compressed Speech. They hypothesized these
patients would Show an abnormal breakdown in intelligibility
of the signal in the ear contralateral to the lesion because
the decreased number of interneuronal connections would be
insufficient to process information in the message in which
redundancy was reduced. Indeed, this effect has been shown
to occur (Kurdziel and Noffsinger, 1973).

Speaks and Jerger (1965) discussed the importance of
temporal characteristics in the assessment of auditory func-
tioning. They suggested that it was important to use test
stimuli of sufficient duration to permit alteration of the
temporal characteristics of the speech signal. They conclud-
ed that single words were of too short a duration to permit
these changes. Speaks and Jerger controlled for sentence

length and informational content by devising synthetic

5V0

9
sentences, i.e., sentential approximations. In this way,
extrinsic redundancy (as defined by Harris, 1960) was re-
duced. Syntactic and sematic cues were eliminated because
each word in the sequence was chosen with varying degrees of
probability of following the previous word. The authors
found that as the synthetic sentences approached normal
sentences, subjects were more able to accurately recall the
approximations. .

Beasley and Shriner (1973) reduced extrinsic redundan-
cy by using first and second order sentential approximations.
They also varied the amount of time allowed for perception
of the stimuli by independently varying the stimulus (word)
duration and the length of the interstimulus interval.

Using normal adult subjects, they found that the effect of
word duration, interstimulus interval size, and order of sen-
tential approximation were independently Significant. As the
word duration was increased from 200 msec to ”00 msec in 100
msec intervals, recall accuracy increased. As interstimulus
interval increased in 100 msec steps from 100 msec to AOO
msec, recall accuracy increased. Recall accuracy also in-
creased as the order of sentential approximations approached
normal sentences (i.e., first order and second order senten-
tial approximations). The authors also found a significant
interaction between word duration and interstimulus interval
length, and between word duration and order of approximation.
The greatest difference between first and second order sen-

tential approximations was in the 300 msec word duration

10
condition, suggesting that, beyond #00 msec and 200 msec,
increases or decreases in word duration, respectively, would
result in only minimal changes in recall accuracy. The com-
plex interaction between word duration and interstimulus
interval was significant, suggesting that the important
factor in speech perception was not either the stimulus du-
ration or interstimulus interval, but rather, the overall~
rate of Speech. Beasley and Shriner found no significant
interaction between the length of the interstimulus interval
and the order of sentential approximation.

Other investigators have studied the effect of vary-
ing interstimulus intervals in sentential stimuli and have
obtained results that are consistent with that of Beasley
and Shriner. Kuhl and Speaks (1972) compared two kinds of
distorted Speech stimuli. In comparing time-compressed
Speech with interrupted Speech, the authors found that, at
rates of less than 2 words per second, recall accuracy was
similar for both conditions.. However,at rates of 2 words
or more per second, compressed Speech was more difficult to
understand than interrupted speech. One of Kuhl and Speaks'
hypotheses as to why this might occur was that the silent
intervals in the interrupted stimuli provided time for the
eliminated portions of the signal to be processed, whereas
the compressed stimuli provided no interval for processing
the deleted portions of the message.

It appears that the length of interstimulus interval

influences the accuracy of recall. Aaronson et a1. (1971)

 

\l..ul (I) .
a I» :

ll
and.Miscik et a1. (1972) found increased recall for digits
with increased interstimulus interval. Beasley and Shriner
(1973) and Kuhl and Speaks (1972) found this same effect
with sentential stimuli. Apparently the interstimulus in-
terval enhanced the possibility for accurate perceptual
processing (Broadbent's second stage of Short-term memory)
to occur. That is, the increased interstimulus interval
provided more time for the stimuli to be perceived, result-
ing in greater recall accuracy.

Auditory Perception in Children

 

Several investigations have been designed to study aud-
itory perception in children. Beasley, Maki, and Orchik .
(1976) studied children's perception of time-compressed
speech using two measures of speech discrimination: The
WOrd Intelligibility by Picture Identification (WIPI) and
the Phonetically Balanced Kindergarten lists (PBK's). Their
results were compared with data obtained by Beasley, Schwim—
mer and Rintelmann (1972) and Beasley, Fbrman and Rintel-
mann (1972) for adults using time-compressed Northwestern
University Auditory Test Number Six monosyllabic word lists
(NU#6'S). Beasley et al.(l976) found that children tended
to Show decreased performance at lower levels of time-com-
pression eSpecially when an Open message set response task,
such as the PBK'S was used. Children showed a major break-
down in scores at 60% time-compression, whereas adults did
not show a significant decline until the point at which 70%

time-compression was employed. When a closed message set

12

was used (the WIPI), there was also a breakdown at 60%, how-
ever, it was not as dramatic as with the PBK lists. Hodgson
(197“) found that the closed message set reSponse task of
the WIPI makes up, in part, for the limited linguistic capa-
bilities of children, thereby explaining the WIPI-PBK dis-
crepancy that Beasley et al. found.

Other investigators have studied short—term memory
processing of children using sentential stimuli. The effects
of time-compressed and time-eXpanded sentences on sentence
discrimination using a closed message set were studied by
Thompson (1973). She found that by decreasing presentation
rate, the recall accuracy of children increased. However,
as age increased, the effect of presentation rate became
less significant. She found that younger children tended
to perform better under conditions of time-expansion, in-
dicating that younger children need more time to process
linguistic information. .

Another approach to investigating perception in child-
ren has been to study the effect of varying interstimulus
interval while keeping stimulus duration constantu Shriner
and Daniloff (1970) varied the amount of time between phon-
emes arranged in meaningful and meaningless CVC units.‘ Per-
formance by children in both first- and third—grades on mean-
ingless stimuli was poorer than their performance on meaning-
ful stimuli. While the children of both grades performed
Similarly on the meaningless stimuli, for the meaningful

stimuli, third-grade children tended to perform better at all

\I'

13
interstimulus intervals. The authors hypothesized this
effect was due to increased linguistic capabilities pos-
sessed by the older children. Shriner and Daniloff found an
interstimulus interval of about 200 msec to be a pivot point.
That is, resynthesis of meaningless stimuli remained con-
stant for increasing intervals until 200 msec, at which
point, increasing the interstimulus interval brought about
a decrease in correct reSponses. For meaningful stimuli,
percentage correct responses declined as the interstimulus
interval increased to 200 msec, after which scores remained
at the same level for longer interstimulus intervals.

The results of Shriner and Daniloff may be applied to‘
Miller's "chunking" hypothesis (1956), as discussed by Beas-
ley and Beasley (1973). Beasley and Beasley postulated that
’stimuli with an interphonemic interval of less than 200 msec
was processed as a single word. With the slight increase in
scores observed at the A00 msec interphonemic interval condi—
tion, Beasley and Beasley hypothesized the CVC units were
processed as three separate phonetic units. At interphon-
emic intervals of intermediate duration (greater than 100
msec but less than A00 msec), it was difficult for the child-
ren to decide whether the stimuli should be processed as one
unit or as three separate units; therefore, scores were
depressed.

Beasley and Beasley used this same measure to compare
short-term memory functioning of Black and White children

from an inner-city environment. They obtained results

1“.
similar to those of Shriner and Daniloff (1970); that is,
third-grade children performed better than first-grade child-
ren on the meaningful stimuli, although there was no grade
difference on the meaningless stimuli. Beasley and Beasley
also found a major breakdown in the 200 msec interphonemic
interval condition. Since the results of Beasley and Beasé
ley did not differ from those of Shriner and Daniloff, the
former authors concluded that short-term memory functioning
was not culturally based.

Beasley and Flaherty-Rintelmann (1976) assessed short-
term memory functioning in children by varing sentence
length (three-word and five—word), order of sentential ap-
proximation (first order, second order, and normal sen-
tences), interstimulus interval (an unaltered interval, and
200 msec, and A00 msec), and grade level (second-grade and
fourth-grade). Using a recall task, results indicated that
fourth-grade children were better able to recall the sen-
tences than second-grade children. As sentence length in-
creased from three words to five words, the number of errors
increased, except in the normal condition, where accuracy in
both three- and five-word recall was near 100 percent over
all interstimulus interval and sentence length conditions.
As a general trend, recall accuracy increased as order of
sentential approximation increased, whereby first order sen-
tential approximations showed the greatest number of recall
errors and normal sentences showed the least number of re-

call errors. Greatest recall accuracy was obtained with an

15.

unaltered interstimulus interval, which was determined to be
the shortest interstimulus interval. The A00 msec condition
produced the largest number of recall errors. The interme-
diate condition of 200 msec was Shown to have an effect on
recall accuracy somewhere between the unaltered condition
and the A00 msec condition; Beasley and Flaherty-Rintelmann
stated that temporal characteristics such as the interstim-
ulus interval might be important in the processing of audi-
tory stimuli at lower, more peripheral neural centers. As
the stimuli became more linguistically meaningful, it was
analyzed at higher levels of the central auditory system and
temporal alteration of the stimuli did not affect recall
accuracy.

All of these studies dealing with auditory processing
capabilities of children have tended to show two main char-
acteristics of the children's ability to process information:

(1) auditory perception in children tends to be differ-

ent from that of adults. In general, they are not
able to process temporally distorted stimuli as
well as adults.

(2) auditory perception appears to be developmental

in nature. Ability to use linguistic strategies
and to chunk stimulus items increases with age,
thereby increasing short-term memory recall

accuracy.

16
Learning Disabled Children
Senf and Freundl (1971) investigated short-term memory
function in learning disabled children. Three visual-audi-
tory digit pairs were presented in a simultaneous condition
and in two alternating conditions, the visual digit being
presented first in both conditions. Subjects, eighteen
learning disabled males matched for age and intelligence
quotient with eighteen children who did not possess learning
problems, were directed to recall the digits in one of two
manners:
(l) directed pair: subjects were to recall the
digits in their order of arrival, beginning
with the first visual digit.
(2) directed modality: subjects were to recall
the digits by modality; that is, the three
visual digits first, followed by the three
auditory digits.
Both gross number of errors and order errors were analyzed.
For the directed pair conditions, there was no difference
among conditions (simultaneous or alternating presentation).
The learning disabled children differed from the control
group of children, Showing poorer recall scores in the au-
ditory error analysis, while performing Similarly to the
control group with the visually presented digits. For the.
directed modality condition (recall by modality), learn-
ing disabled children performed significantly lower than

the control group on all conditions (visual and auditory

17
modality, Simultaneous and alternating conditions, and gross
number and order errors). Presentation conditions showed a
significant difference only with visual gross errors. The
authors concluded that these results were not due to prob-
lems in cross-modal integration (from visual to auditory
stimuli and vice versa), but rather, to inadequate higher~
order (short-term memory) functioning in the learning dis-
abled children.

Rosenthal (197A) proposed a theory based on his re-
search of auditory perceptual difficulties in children with
language disorders. In his theory, Rosenthal stated that
the Specific perceptual disorder, whether it be auditory or
visual, interferes with accurate processing of incoming
linguistic stimuli. If the disorder is an auditory percep-
tual impairment, then the incoming signal is speech, and,
because the Speech signal is not perceived accurately, the
child is not able to internalize language. The consequence
is delayed language develOpment..

McCrosky and Thompson (1973) studied the effect of
time-compression, time-expansion, and a normal presentation
rate using children from the ages of 5 through 17 years who
demonstrated difficulty in understanding Spoken messages,
but whose auditory sensitivity was within normal limits.

The authors time-compressed and time—eXpanded simple declar-
ative sentences and presented them in a discrimination task.
Significant differences were obtained only for the younger

grade-school-age children. What was found was that as

18
presentation rate decreased (that is, as the Speech message
was expanded) the accuracy of performance of these children
increased, suggesting that those children with auditory per-
ceptual handicaps needed more time to process incoming in-
formation. The older children showed no differences in
perception of the varied presentation rates. This may be
because their knowledge of linguistic rules was such that
they were able to obtain enough information from the syntac-
tic and semantic constraints, ignoring the temporal factors
involved.

The inability of auditory perceptually handicapped
children to adequately process time-compressed stimuli was
also studied by Manning, Johnston, and Beasley (1975). They
found that these children tended to Show decreased accuracy
scores at a lower sensation level and percent level of time-
compresSion than children without auditory processing prob-
1ems (Beasley et al., 1976). Using time-compressed PBK'S,

a Speech discrimination measure, Manning et al. found that
scores of the impaired children agreed with those of the
normal children at 0% and at 60%; however, at 30% time-
compression, the impaired children showed significantly
lower scores. This would again suggest that auditory
perceptually impaired children process Speech at a slower
fate than normal children.

Freeman, Beasley, and Overholt (1975) found this same
effect with a learning disabled population using the WIPI

speech discrimination test. Again, the learning disabled

19
children performed more poorly at lower sensation levels
and lower levels of time-compression than normal children.
As well, Freeman et a1. studied the performance of learning
disabled children in recalling sentential stimuli of vary;
ing word length, order of sentential approximation, and in-
terstimulus interval size.. Results indicated that learning
disabled children had more difficulty than normal children
(Beasley and Flaherty-Rintelmann, 1976) in processing
temporally altered first and second order sentential approx-
imations. Also, those learning disabled children with audi-
tory perceptual problems did Significantly worse than learn-
ing disabled children without auditory perceptual problems,
whose performance agreed with the normative data.

As Broadbent's theory of Short-term memory is stated,
by decreasing presentation rate, more time is available to
process the stimuli, thereby increasing recall accuracy.
Research results have shown that in children with learning
disabilities short-term memory functioning is inferior to
that of normal children. Since these children cannot ade-
quately process linguistic stimuli until these stimuli are
presented at a Slower rate, they may be having problems
transferring information from Broadbent's sensory processing
stage of short-term memory to the perceptual stage.

Statement of the Problem

 

It has been found that the temporal characteristics of
the Speech signal play an important role in auditory percep-

tion of speech. Normative data has been obtained on a task

20

that varies interstimulus interval, as well as controlling
for grammaticalness, for both adults (Beasley and Shriner,
1973) and children (Beasley and Flaherty-Rintelmann, 1967).

Theories as to how children who are labeled Specific
learning disabled process speech stimuli have.been postulat-
ed. Although there is a paucity of research in this area,
it is generally believed that these children demonstrate a
breakdown in the Speech processing mechanisms of the central
auditory pathways (Willeford, 197A). If this is true, per-
formance by these children on a Speech processing task that
purports to measure a major characteristic of the function-
ing of the central auditory system should Show a significant
difference from responses of normal children. This differ-
ence may be important in the detection and subsequent
remediation of children with auditory perceptual problems.

The purpose of this investigation was to determine the
perceptual processing strategies of auditory perceptually
handicapped children using a task of sentential stimuli,
controlling for syntactic and sematic constraints, and
sentence length. The effects of modifying the sentence
length, order of sentential approximation, and interstimu-
lus interval size on auditory perception of children with
auditory perceptual problems were investigated.

Specifically, the following questions were asked:

(1) what effect would variations in sentence length

have on recall accuracy of auditory perceptually

handicapped children?

(2)

(3)

(A)

(5)

21
what effect would various orders of sentential
approximation and normal sentences have on recall
accuracy of these children?
what effect would varying interstimulus intervals
of a normal condition, 200 msec, and A00 msec
have on recall accuracy of the stimuli by these
children?
what would be the effect of interactions of the
above factors on recall accuracy of auditory per-
ceptually handicapped children?
how would the performance of auditory perceptually
handicapped children compare with the performance'

of children without such handicaps?

CHAPTER II

EXPERIMENTAL PROCEDURES

Twelve reading impaired children were presented with
a temporally altered linguistic test. There were four sub-
jects per condition of sentence length by sentential ap-
proximation by interstimulus interval, for a total of 36
subjects for each sentence length (three-word and five-word),
2A for each order of sentential approximation (first order,
second order, and normal sentences), and 2A for each inter-
stimulus interval (a normal condition, 200 msec, and
A00 msec). (See Figure 1.)
Subjects

Twelve children with an age range of 8.5 years to 11.7
years and a mean age of 10.1 years who were enrolled in the
Remedial Reading Center at Michigan State University served
as subjects. These children were chosen as subjects because
theories have been postulated (Norman, 1969; Mattingly,
1972) and research has shown (Birch and Belmont, 1965;.
McGrady and Olsen, 1970) that visual stimuli are processed
auditorilly. Therefore, children who indicate reading dif—
ficulties would likely also manifest auditory perceptual
problems (de Hirsch, 1973; Freeman and Beasley, 1976; Stark,

1975). None of the children evidenced any other known

22

23

 

 

 

 

 

 

 

 

 

 

.m._. seesaw .aecez
fiestas.

=e_~.e_...aa¢ _._u=..=om

400

200

Interval

Interstimulus

(msec)

Experimental design.

~Figure 1.

2A.

handicaps. All children received a thorough audiologic
evaluation and exhibited normal hearing. (See Appendix A.)

Each subject was presented 20 first order sentential
approximations (ten three-word and ten five-word), 20 second
order sentential approximations (ten three-word and ten five-
word), and 20 normal sentences (ten three-word and ten five-
word). Also, each subject received 20 sentential stimuli
(ten three-word and ten five-word) with an unaltered inter-
stimulus interval (an interstimulus interval of normal Speak-
ing rate), 20 sentential stimuli (ten three-word and ten
five-word) with a 200 msec interstimulus interval, and 20
sentenial stimuli (ten three-word and ten five-word) with a
A00 msec interstimulus interval.‘ The two experimental con-
ditions of sentential approximation and interstimulus
interval were counterbalanced to minimize possible inter—
actions and randomized to reduce any effect that might be
due to learning.
Stimuli

The stimulus materials used here were taken directly
from those used by Beasley and Flaherty-Rintelmann (1976) in
a previous investigation of Short-term memory function in
second- and fourth-grade children. The stimuli consisted of
three separate experimental conditions:

(1) order of approximation: 20 first order sentential

approximations (ten three-word and ten five-word),
20 second order sentential approximations (ten

three—word and ten five-word), and twenty normal

25
sentences (ten three-word and ten five—word).

(2) interstimulus interval: all sixty sentential
approximations and sentences produced with a
200 msec interstimulus interval, a A00 msec inter-
stimulus interval, and an unaltered interstimulus
interval (normal Speaking rate).

(3) sentence length: ten first order sentential
approximations, ten second order sentential approx-
imations, and ten normal sentences, each consist-
ing of three words, as well as ten first order
sentential approximations, ten second order
sentential approximations, and ten normal sen—
tences of five-word length.

The sentential approximations to full grammaticality
were constructed in a manner similar to a procedure described
by Speaks and Jerger (1965), based on the probability of one
word following another in a normal sentence. The words used
in the sentential approximations for this investigation were
taken from the Basal Vocabulary of the primary reader work-
book, ON WE GO (Teacher Edition) (Second Level) (Houghton
Mifflin, 1966). One hundred monosyllabic words were chosen
from the ON WE GO list and randomized into a new list. (See
Appendix B.)

For the first order sentential approximations, words
were chosen randomly from the revised list and constructed

into ten three-word and ten five-word sentential stimuli.

26

Construction of the second order sentential approxi-
mations was achieved by randomly choosing the first word
from the revised list. The second word was then chosen
from the revised list such that it might follow the first
word in a normal sentence. The third word was chosen, with-
out the knowledge of word one, to follow the second word in
a similar manner. This procedure was continued until ten
three-word and ten five-word second order sentential approx-
imations were constructed.

Recording and Stimulus Generation Procedures

 

For the experimental tape recordings, sixty sentential
stimuli (ten three-word and ten five-word first order approx-
imations, ten three-word and ten five-word second order ap—
proximations, and ten three-word and ten five-word normal
sentences) were recorded onto a master tape recording by a
male Speaker who Spoke general American English and was
trained in phonetics. The Speaker used an Electro-voice
635A micrOphone and an Ampex AG AAO-B tape deck. The re-
cordings were made at conversational pitch and effort level
with minimum inflection in a sound treated recording suite.
Minimum inflection was used to reduce prosodic cueing ef-
fects that might influence the stimuli. The speaker moni-
tered his vocal intensity on a VU meter. Peaking occured
between words in order to facilitate the location of the
initiation and termination of each word for splicing

purposes.

27.

Two COpies of the master tape recording were made at
7% ips (inches per second) using Ampex 601 and Ampex 600
tape recorders. Each of these copies was prepared as exper-
imental stimuli by varying the interstimulus interval size
(200 msec and A00 msec, respectively). To prepare the
200 msec condition, the acoustic initiation and termination
points of each word were determined aurally by using a play-
back record head coupled to a pre-amplifier and an
amplifier/speaker. These points were marked and then
manually severed. The existing interval between words was
discarded and a length of magnetic recording tape ascertain-
ed to be 200 msec in length (1.5 inches of tape) was Spliced
between each word. This same procedure was followed in the
preparation of the A00 msec condition stimuli. For this
condition, a length ascertained to be A00 msec (3 inches of
tape) was used.

These experimental tapes were played through a Bruel
and Kjaer Type 2305 high speed graphic level recorder (paper
Speed=30 mm/sec, writing speed=250 mm/sec, 50 dB, 20 Hz).
The silent interstimulus intervals were hand measured in
millimeters and converted to millisecondsi Any silent in-
terstimulus interval in error by more than approximately
:30 milliseconds was reprocessed.

These stimulus tapes (used by Beasley and Flaherty-
Rintelmann, 1976) were rerecorded and reordered for counter-
balancing and randomization procedures used in the present

investigation. The stimuli were c0pied from an Ampex AG

28
AAO-B tape deck to an Ampex AG 601 tape recorder, and the
carrier phrase, "Number'___A (the number of the sentence
for a given condition) was inserted in front of each senten-.
tial stimuli.

Counter-Balancing and Randomization Procedures

 

The condition of order of sentential approximation
(first order, second order, and normal sentences was com-
bined with the interstimulus interval condition (an inter-
val associated with normal speaking rate, 200 msec and A00
msec) and sentence length (three-word and five-word). This
resulted in eighteen experimental conditions. (See Figure 1.)

These experimental conditions were arranged such that
each subject would receive two conditions of each order of
approximation (first order, second order, and normal sen-
tence), two conditions of each interstimulus interval (200
msec, A00 msec, and an unaltered interstimulus interval),
and three conditions of each sentence length (three-word and
five-word). Among subjects, these conditions were counter-
balanced to offset any effect that might be due to the
interaction between sentence length,sentential approximation,
and length of interstimulus interval. All of the counter-
balanced conditions were randomized to minimize any effect
due to practice that might occur.

Presentation Procedures

 

Each child was presented with 30 sentences of each
sentence length (three-word and five-word), 20 sentences

of each order of sentential approximation (first order,

29
second order, and normal sentences), and 20 sentences of
each interstimulus interval (normal, 200 msec, A00 msec)
for a total of 60 sentences per subject.

The listeners were seated in a chair in a two-room,
double-walled IAC 1200 series sound-treated testing suite.
The pure-tone air- and bone-conduction tests were presented
via a Beltone 150 clinical audiometer. The clinical tape
recordings for speech reception threshold (SRT) (Central
Institute for the Deaf - CID W-l) and Speech discrimination
(Northwestern University Auditory Test Number Six - NU#6)
and the experimental tapes were generated through an Ampex
600-2 tape recorder to a Grason-Stadler 162 speech audio-
meter located in the control room. The experimental
stimuli were presented at 7% ips binaurally, through TDH—39
earphones housed in MX—Al/AR cusions. The intensity level
was set at 50 dB HL (re: ANSI, 1969).

After a brief greeting period, the hearing evaluation
was performed. For the experimental portion of the testing,
standard verbal instructions (Appendix C) were given to each
subject. If there were any questions, they were answered.
Each subject was tested individually.

The oral responses of the subjects were recorded by
the examiner on the reSponse forms (Appendix D).

Analysis
The data were hand scored. The percent of words cor—

rectly recalled was the score for each subject. Three types

30.

of errors were pOSSible:
(1) discrimination errors consisted of words inaccu-
rately recalled.
(2) omiSSion errors consisted of any word not recall-
ed within the sequence.
(3) any words that were repeated in the wrong order

were considered to be order errors.

CHAPTER III
RESULTS

The results of this study supported the contention
that Speech perception is dependent, at least in part, on
the syntactic, semantic, and temporal constraints imposed
upon sentential stimuli. Overall results indicated that sen-
tence length, order of sentential approximation, and inter-
stimulus interval all affect recall accuracy of auditory per-
ceptually handicapped children. In general, recall accuracy
decreased as sentence length increased. As the order of
sentential approximations rose to full grammaticality, re-
call accuracy scores increased. Recall accuracy decreased
as the silent interstimulus interval was increased. (These
effects, discussed below, can be seen graphically in Figure
2 and in Tables 2 and 3 in Appendix E.)
Effect of Sentence Length

 

Figure 2 Shows the main effect of sentence length;
that is, as sentence length increased from three words to
five words, recall accuracy was reduced.

The interaction between sentence length and order of
sentential approximation is illustrated in Figure 3A. As
can be seen from the graph, the three-word stimuli provided
for better recall accuracy. This is true across order of

sentential approximation. However, as the order of

31

32

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

lOOlr 1
F.
F.
F?
390" F
5 F F‘
3 lst 2nd n h 260 460
Sentence Sentential Interstimulus
length Approxnmation Interval
(words) (oldet) (msec)

Figure 2. Main effects of sentence length, sentential

approximation, and interstimulus interval.

 

33,

 

 

 

100 t ENE] 100 I
2
C)
3 so «- so a-
3 A
g A
S’.
[J
C)
80 t t i t
3 S 80 3 5
Sentence Sentence
length length
Figure 3A Figure 3B
Figure 3A. Mean percent correct scores per sentence
length and sentential approximation ((3 -
first order;A- second order; D- normal sen-
tence).
Figure 3B.

Mean percent correct scores per sentence
length and interstimulus interval ( 0 - normal
interval;[§r200 msec;[j-A00 msec).

3A.
sentential approximation increased from first through second
order to full grammaticality, the effect of sentence length
decreased, so that at the normal sentence condition, there
was only a negligible difference between three-word and
five—word length.

Sentence length also interacted with interstimulus in-
terval, as seen in Figure 38. It appears that recall accu-
racy decreased as sentence length increased; and, further,
that this occurred over all interstimulus intervals. Also,
as the interstimulus interval condition increased from a
normal speaking rate to 200 msec and A00 msec, the differ-
ence between recall accuracy for the three-word and five-
word sentence lengths increased. (This is also illustrated
in Figure 5A.)

Effect of Order of Sentential Approximation

 

As can be seen in Figure 2, there was a main effect
for order of sentential approximation. In general, as the
order of approximation increased from first order to second
order to full grammaticality, the scores in recall accuracy
increased.

There was an interaction between sentential approxima—
tion and sentence length (Figure AA). While the effect of
increased recall accuracy with increasing order of senten-
tial approximation occurred over both conditions of sentence
length (three-word and five-word), the effect was much

‘greater for the five-word stimuli. Also, for normal sen-

_ tences, recall accuracy scores of three- and five-word

 

 

 

 

loo 1' too I
O”MIMO
A
8 so r so '-
.. CI
5:. u
.. A
:1
C!
D I
3° lst 2nd N lst 2nd u
Sentential Sentential
Approximation Approximation
Figure AA Figure AB
Figure AA.‘ Mean percent correct scores per sentential
approximation and sentence length (C) - three
words; [J - five words).
Mean percent correct scores per sentential

Figure AB.

approximation and interstimulus interval ((3
normal interval; A- 200 msec;['_‘) — A00 msec).

36
stimuli were essentially equal and approached 100 percent
accuracy.

Order of sentential approximation also interacted with
interstimulus interval (Figure AB). Again, recall accuracy
scores can be seen to increase as the sentential approxima-
tions approach full grammaticality. As well, at the lower
orders of approximation (first and second order), the inter-
stimulus interval had a greater effect on recall scores than
at the normal sentence condition. With the normal sentence
stimuli, variations in recall due to changes in interstim—
ulus interval were negligible.

Effect of Interstimulus Interval

 

The main effect of interstimulus interval is shown in
Figure 2. The graph shows that as the silent interstimulus
interval (181) increased from a normal ISI (determined to be
the Shortest ISI--approximately 100 msec in length) to 200
msec and A00 msec, scores in recall accuracy decreased. I

The length of the interstimulus interval interacted
with sentence length to produce an effect on recall accuracy
(Figure 5A). Recall accuracy decreased as the interstimulus
interval increased for both three-word and five-word stim-
uli. However, this effect was greater for the five-word
stimuli than for the three-word Stimuli.

Figure 5B shows the interaction of interstimulus inter-
val and sentential approximation. For first and second
order sentential approximations, as the interstimulus

interval increased, the main effect of decreased recall

 

37_

 

 

 

too-=- 100 r C] D
o A
E] C)
A
c3 90-"- 90 ..
g A
3 C] c)
.:
C)
30 l t 4 80 i 1 . :__
N 200 400 N 200 400
Interstimulus Interstimulus
Interval Interval
Figure 5A Figure 5B
Figure 5A. Mean percent correct scores per interstimulus
interval and sentence length ((3 — three words;
[3:— five words).
Figure SE. Mean percent correct scores per interstimulus

interval and sentential approximation ((3 -

first order;[§

tence).

- second order;[] - normal sen-

38

accuracy scores was observed. However, for the normal sen-
tence condition, the length of the interstimulus interval
appeared to have a negligible effect.

A three-way interaction is illustrated in Figure 6.
This graph Shows the interaction of sentence length, sen-
tential approximation, and interstimulus interval. Both
decreased sentential approximation and increased interstim-
ulus interval had a slight effect on the three-word stimuli,
resulting in decreased recall accuracy; however, this effect
was more easily seen for five-word stimuli. The normal sen-
tences were not affected greatly by sentence length or in-
terstimulus interval. Recall for all normal sentences was
near 100 percent. It can be seen from Figure 6 that recall
accuracy decreased as sentence length increased, as order of
sentential approximation decreased, and as interstimulus

interval size increased.

39

.Aoocopcom Adamo:

I cannon .noomo psooom I oawnoanu .noono pmhfiu I moaondov coapmeonnqoo
Heapsopnom pew “mono: o>am I mHonE»m ooaaak .monoz oops» I mHonemm cooov

newcoa ooeopcon .Hm>aopca moanedpmnopsfi nod monoom poohnoo psoonoa can:

.0 onswfim
0
A . jso
4
s
n
“V“
0 m V
o .30 .l .I
2 t e
s t
. r n
e |
in
I!" I

 

 

up

100

Because

80 u
70

«goose;

CHAPTER IV

DISCUSSION

The results of this study indicated that recall accu-
racy in auditory perceptually handicapped children is af-
fected by the conditions of sentence length, order of sen-
tential approximation, and length of Silent interstimulus
interval independently. These three conditions also inter-
acted at various levels. In certain instances Specific var-
iables took precedence over others in affecting the recall
accuracy. These findings, along with the results of pre-
vious investigations, support implications which may be
important in both discussing the theoretical aspects of
auditory processing and language aquisition in children and

in diagnosing auditory perceptual impairments.

Comparison of Trends to Previous Investigations

 

In reviewing the main effects of the three experimental
conditions in this study, the results support the findings
of previous investigators.

AS sentence length increased, recall accuracy decreased.
Previous studies on sentence length and short-term memory
have all Shown this effect (Giattino, 1973; Pantalos,
Schuckers, and Hipskind, 1972; Schuckers, Shriner, and Dani-

loff, 1973; Smith and Beasley, 1973). As the number of

A0

oi

items in a string approaches and exceeds the limits of
short-term memory, the subject is not able to process all
the items correctly, and consequently makes more errors in
recall. Miller (1956) found this number of items to be
seven, plus or minus two. From the results of this and
other studies, it appears that, for children, three items
is well within the limits of Short-term memory, while five
items approaches and, in some cases, exceeds short-term
memory limits.

Results of the present study were consistent with those
of earlier studies in showing that increased sentential ap—
proximation resulted in increased recall accuracy (Beasley
and Shriner, 1972; Carrow and Mauldin, 1973; Speaks and
Jerger, 1965; Tejirian, 1968). As the sentential approxi-
mations increased from first order to second order to full
grammaticality, the syntactic and semantic constraints on
the stimuli increased, thereby abetting perceptual proces-
sing by multiple-cueing and extrinsic redundancy (Harris,
1960). These effects aided in increasing the amount of in-
formation per chunk (Miller, 1956), resulting in more com-
plete auditory processing and increased recall accuracy.

The trend for recall accuracy to decrease with increas-
ing interstimulus interval size supports other studies uti-
lizing sentential stimuli with children (Beasley and Flaher-
ty—Rintelmann, 1976). While other investigators found that
increased interstimulus interval size resulted in increased

recall for digits (Aaronson et al., 1971; Miscik et al.,

A2
1972) , Beasley and Shriner (1973) discussed results which
indicated that when linguistic stimuli were used, the rate
of speech was more important than the length of the inter-

stimulus interval itself for auditory processing.

Comparison of Interactions to Previous Investigations

 

While Beasley and Flaherty-Rintelmann (1976) obtained
normative data using the same stimuli as used in the pre-
sent study, a direct comparison cannot be made due to dif—
ferences in the experimental procedures. In that study,
children were tested in the field; but in the present inves-
tigation, children were tested in a sound-treated testing
suite. However, for both normal second- and fourth-grade
children and for auditory perceptually handicapped children
(as defined by reading impairment), similar trends can be
seen for all experimental conditions. In both investiga-
tions, recall accuracy decreased as sentence length in-
creased, as order of sentential approximation decreased, and
as interstimulus interval increased. Consequently, Similar
interactions of experimental conditions can be seen in both
investigations. Three—word stimuli were very nearly 100
percent correctly recalled in both instances. The five-
word stimuli were found to be more sensitive than the three-
word stimuli to the Short-term memory limits of both normal
and auditory perceptually handicapped children. Normal sen-
tences were recalled with greater accuracy than first order
approximations (which were found to be least correctly re-

called) and second-order approximations, and were less

“3.

affected by temporal alterations than either order of
sentential approximation.

Freeman and Beasley (1976) employed reading impaired
subjects, sentential stimuli, and experimental procedures
similar to those used in the present investigation. Results
of these two studies compared favorably. Both investiga-
tions were concerned with auditory processing via temporal
alteration of the speech Signal in reading impaired chil-
dren. Freeman and Beasley time-compressed the stimulih .
whereas, in the present investigation, the interstimulus
interval was altered without varying word duration. The
condition of 0% time-compression in the Freeman and Beasley.
study was comparible to the normal interstimulus interval
condition used in the present study. Table 1 shows a break—
down of scores for normal readers and reading impaired chil—
dren in the Freeman and Beasley study, as well as the Scores
of the subjects in this study. The Significant main effects
of sentence length and order of sentential approximation
can be seen. The table also shows that as syntactic and
semantic constraints were reduced on the stimuli (i.e., as
sentential order decreased from normal sentences to first
order approximations), the normal reading children were not
affected greatly, while the reading impaired children in
both investigations showed a decline in recall accuracy.
Also, the normal readers in the Freeman and Beasley study
were not as greatly affected by temporal alteration (time-

compression) of the stimuli as the reading impaired subjects.

on

Table 1. Mean percent correct scores of normal reading
(NR—BAF) and reading impaired children (RI-BAF)
in the Freeman and Beasley (1976) study and of
subjects in this study (RI-GAO) for the main
effects of sentence length (three-word (3) and
five-word (5)) and sentential approximation
(first order (1), second order (2), and normal
sentences (N)).

 

 

 

Subjects Sentence Length Sentential Approximation
3 5 N 1 2 N
NR-BAF 97.8 95.5 91.5 99.5 100.0
RI-BAF 97.3 89.5 87.3 9A.o 99.0
Total- BAF 97.6 92.5 89.2 96.8 99.5
RI-GAO 97.8 89.5 88.8 92.8 99.6

 

Further, time—compression did not greatly affect the recall
scores for either normal readers of reading impaired chil-
dren when the sentential stimuli were normal sentences.
Based on the Freeman and Beasley data that the normal
reading children did not make as many errors as the reading
impaired children on temporally altered sentential stimuli,
it can be speculated that, using similar stimuli and exper-
imental design as was used in this study, children without
auditory perceptual impairments (or normal reading children)
would Show greater recall accuracy scores than children with
auditory perceptual (or reading) problems (i.e., the chil-
dren who served as subjects in this study). However, due to

differences in the experimental procedures between the

A5.
present study and that of Beasley and Flaherty-Rintelmann

(1976), the above statement can only be hypothesized.

Type of Errors.

 

An error analysis performed on the data showed that
32 percent of the errors in recall were errors in intelligi-
bility, while 36 percent were memory errors (ommissions),
and 32 percent of the errors were order errors. This break-
down of error types was due primarily to the five-word se-
quences. For three-word sequences the errors were 93 per-
cent intelligibility errors and only 7 percent memory
errors. Errors made on the five-word sentences were 26 perv
cent intelligibility, 39 percent memory, and 35 percent
order errors. These trends compare favorably with the Free-
man and Beasley data, which found A5 percent intelligibil-
ity errors and A0 percent memory errors in the reading impair-
ed population, while the normal readers showed 75 percent in-
telligibility errors and only 10 percent memory errors. For
first and second order sentential approximations and for all
interstimulus interval sizes, the distribution of error
types was equal over the three catagories; that is approxi-
mately 1/3 of the errors were intelligibility, approximately
1/3 were memory (ommission),and the final 1/3 were order
errors. For normal sentences, there were 50 percent intel-
ligibility and 50 percent order errors. This finding tends
to suggest that the children used the syntactic and semantic

cues to effectively chunk information (Miller, 1956),

A6,

thereby avoiding omitting words. The order errors in the
normal sentence condition tended to consist of changing a

sentence to a question or vice versa. (I.e., you do like

 

the snow was repeated as do you like the snow.)

 

Theoretical Implications

 

In three investigations using sentential stimuli with
normal and reading impaired (or auditory perceptually handi-_
capped) children (Beasley and Flaherty—Rintelmann, 1976;
Freeman and Beasley, 1976; and the present investigation),
similar trends have been found. It appears that as the sen-
tential stimuli approach full grammaticality (i.e., normal
sentences), temporal alterations of the Signal did not
affect auditory perception to a great extent. In this study,
normal sentences were recalled with near 100 percent accu-
racy regardless of sentence length or interstimulus inter-
val. Freeman and Beasley found that normal sentences were
correctly recalled by both normal reading and reading im-
paired children in both the 0% and the 60% time-compression
conditions. These findings suggest that once the stimuli
has reached SOphisticated levels of syntactic and semantic
constraints, temporal alterations has little effect on re-
call accuracy. This may indicate that highly constrained
stimuli are analyzed at a very high level of auditory pro-
cessing, and that temporal alteration affects only stimuli
processed in the lower, more peripheral, neural centers.

Another interesting trend of this investigation can be

seen in Figure 6. Normal sentences were recalled with 100

AT
percent accuracy in the normal interstimulus interval (ISI)
condition and in the A00 msec ISI condition. However, a
slight drop in recall accuracy can be seen in the 200 msec
ISI condition for both three-word and five-word sentences.
Beasley and Beasley (1973) hypothesized that a 200 msec in-
terphonemic interval might be a pivot point for auditory
processing. The trend in the present study, along with the
findings of Beasley and Beasley, suggest that children may
not be able to effectively chunk information in the 200 msec
ISI condition. At the normal ISI condition, children may be
processing the whole sentence as one unit. At the A00 msec
ISI condition they may be processing five separate words.
However, the 200 msec condition may be too Slow for the
children to process the sentence as a whole and too rapid
for the words to be processed individually, thus resulting

in a Slight reduction in recall accuracy scores.

Implications for Audiology

 

It appears that conventional auditory tests are not
sensitive enough to evaluate auditory functioning in audi-
tory perceptually handicapped children. These tests eval-
uate the status of the peripheral mechanism (outer and mid-
dle ear, chochlea, and the VIIIth cranial nerve), but lack
the complexity necessary to tax the central auditory system.
Since this system is reSponsible, hypothetically, for tem-
poral integration at the lower neurological levels and struc-
tural organization at the higher, more cortical, levels, it

would seem reasonable to assume that temporally altered

1‘8.

Speech, if not highly syntactically and semantically organ-
ized, will aid in diagnosing the presence or absence of a
lesion in the lower central auditory nervous system. In
order to adequately assess this area of the auditory system,
the stimuli must be of sufficient length to permit temporal
alteration; however, the stimuli must simultaneously be con-
trolled for structural organization. Sentential approxi-
mations have been found to effectively reduce the syntactic
and semantic constraints of speech (structural organization)
and yet maintain the temporality needed to assess central

auditory system functioning.

Implications for Language Development

 

Normal language acquision and development has been
hypothesized to depend upon the auditory processing of
Speech, and such processing, in turn, is dependent, at least
partially, upon the temporal patterns and the organization
of the speech signal. As the speech stimuli become more
complex, auditory perception becomes more difficult. Chil-
dren exhibiting Specific learning disabilities, or auditory
perceptual impairments, are unable to temporally integrate
and organize stimuli in such a manner as to adequately pro-
cess the stimuli. Therefore, auditory perceptually impaired
children are unable to internalize the temporal and organi-
zational characteristics of Speech. Consequently, if the
child does not possess these characteristics, he will be
unable to effectively use them in his expressive speech:

The result of this inability to internalize the temporal

A9.
and organizational characteristics of Speech is delayed
language development, Since language acquisition and develop-
ment is indeed dependent upon these characteristics. In au-
ditory perceptually handicapped children, the central audi-
tory pathways do not exhibit enough intrinsic redundancy to
overcome the reduced extrinsic redundancy and multiple-
cueing effects of time—altered and unorganized speech. In
order for this child to prOperly perceive Speech, the sig-
nal must be highly organized and/or of short enough duration

that the child will process the entire message.

Implications for Future Research

 

While findings indicate, at least tentatively, that
temporally altered sentential stimuli may be of significance
in diagnosing auditory-based perceptual handicaps, more
information should be obtained. Specifically, a normative
population should be evaluated in a clinical setting for
direct comparison to these results. Also, more information
may be obtained by varying and increasing the number of ex-'
perimental conditions (sentence length, sentential approxi-
mation, and interstimulus interval size). Information as
to what takes place during auditory misperception might be
hypothesized from more complete item, error, and subject

analyses than was possible in the present study.

CHAPTER v
SUMMARY AND CONCLUSIONS “

Temporal characteristics of the Speech signal have
been found to play an important role in auditory processing.
In the past, investigators used digits and word lists to
study short-term memory and auditory processing. Only very
recently have linguistic stimuli been used to assess central
auditory system functioning. These stimuli, however, are
not subtle enough to be useful unless extrinsic redundancy
and the effects of multiple-cueing are reduced. First and
second order sentential approximations have been found to
accomplish this task. Temporal alteration of these approxi—
mations would then be subtle enough for central auditory
testing.

It has been hypothesized that children with auditory
perceptual disorders manifest problems of reduced intrinsic
redundancy (i.e., an inability to integrate distorted stim—
uli). Therefore, these children would not be able to ade-
quately process Speech stimuli from which extrinsic redun-
dancy has been reduced and temporality has been altered.

The purpose of this study was to examine auditory per—
ceptual processing in children with Specific learning dis-

abilities (auditory perceptual impairment) by controlling

50

51 '

for sentence length, order of sentential approximation, and

interstimulus interval size.

Specifically, the following questions were investigated:

(1)

(2)

(3)

(A)

(5)

what would be the recall accuracy of auditory per-
ceptually impaired children on sentential stimuli
of various sentence lengths?

what would be the recall accuracy of these chil—
dren on variations in order of sentential approxi-
mation?

what would be the recall accuracy of these children
on interstimulus interval Sizes of normal duration,
200 msec, and A00 msec?

what would be the effects of interactions of these
above factors on recall accuracy of auditory per-
ceptually impaired children?

what would be the recall accuracy of auditory per-
ceptually impaired children compared to normal

children?

Overall results indicated that the main conditions of

sentence length, order of sentential approximation, and

length of silent interstimulus interval all affected recall

accuracy.

Recall accuracy was near 100 percent for the three-

word stimuli, but as sentence length increased to five words,

recall accuracy decreased. AS order of sentential approxi-

mation rose from first order to second order to full gram-

maticality (normal sentences), recall accuracy increased.

‘Recall for normal sentences was very nearly 100 percent

52.
correct over all other conditions. Increased interstim-
ulus interval (ISI) produced decreased recall acturacy;
that is, the normal ISI condition (considered to be the
shortest ISI--about 100 msec) produced the greatest recall
accuracy, while the longest ISI condition, A00 msec, pro-
duced the most number of recall errors. The intermediate
ISI condition of 200 msec produced recall accuracy scores
somewhere between those scores associated with the normal
condition and the A00 msec condition.-

These results provided information that may be impor-
tant in discussing theories of Speech and language percep-
tion in normal and pathologic (auditory perceptually im-
paired) children. Implications for audiology and the diag-
nosis of these impairments through central auditory system
testing were discussed. AS well, theories as to how faulty
auditory perception can affect language develOpment and
implications for language therapy were presented. However,
more information is needed, especially in terms of more
controlled normative data, more and varied pathologic p0pu-
1ations, eXpanded experimental conditions, and various

types of analyses, before solid conclusions can be made.

I APPENDICES

APPENDIX A

53
APPENDIX A

CRITERIA FOR NORMAL HEARING

l. Pure-tone air-conduction thresholds at 15 dB HL (re:
ANSI, 1969) or better for octave frequencies 250 Hz -
A000 Hz.

2.. Pure—tone bone-conduction thresholds within :5 dB of
air-conduction thresholds.*

3. Speech reception threshold (SRT) of 15 dB HL (regANSI, /
1969) or better.

A. Speech discrimination scores of 9A% or better at A0 dB

SL or 50 dB HL, whichever intensity was lower.

*If air—conduction and bone-conduction thresholds varied by
more than :5 dB at two frequencies, impedance audiometry
was done to determine middle ear functioning. Normal
middle ear functioning was defined as:

1. Middle ear pressure within :50 mm air pressure.

2. Static compliance within .27-1.5 cc equivalent

volume.

APPENDIX B

girl
cry
all
boy
but
had

her

dot
man
out
on
put
tell
see
sun
two
of

here

BASAL VOCABULARY ON WE GO
(1966)

box
big
car
five
day
red
zoo
with
ten
set
is
in
to
we
POP
take
will
me
hot

the

5A

APPENDIX B

REVISED WORD LIST

HOUGHTON MIFFLIN

eat
do
call
cat
him
up
us
you
next
store
pan
high
eye
dog
feet
like
my
not
milk

that

toy
wish
sat
good
it
ride
play
wood
yes
no
name
end
door
run

one

sand-

stOp
was
late

jar

come
six
ten
ring
said
they
lock
see
snow
food
dish
go
egg
know
work
word
tree
this
time

would

APPENDIX C

55
APPENDIX c

STANDARIZED VERBAL INSTRUCTIONS

You are going to hear a man say some sentences. Some
of the sentences will make sense (they will be real
sentences), some of them will not. I want you to listen
to what the man says and say it back to me. Do you have

any questions? (Answer questions.) Okay, let's begin.

APPENDIX D

First Order

56.
APPENDIX D

RESPONSE FORM

Subject

 

Condition

 

 

Three Word
1. go
2. take
3. food
A. jar
5. red
6. wish
7. egg
8. up
9. nest

10. name

Comments

girl zoo
hot girl
jar wish
sand man
toy us
p0p milk
call ten

jar said

ring sun

jar will

First Order

 

 

Five Word
1. men
2. red
3. snow
A. name
5. time
6. snow
7. p0p
8. tree
9. me

10. dOg

Comments

 

ZOO

play

take

milk

all

P0P

ZOO

UP

play

cat

57.
APPENDIX D
RESPONSE FORM

-Subject

 

Condition

p0p nest is
late I call
play_ big man
good girl sun
sun p0p tree
dish is name
cry food milk
yes me word
wish milk nest
high man hot

Second Order

 

ThreeEWord

 

1. men will
2. that man
3. snow time
A. d0g nest
5. wish with
6. good girl
7. like food
8. up end
9. man zoo
10. do hot

Comments

58.
APPENDIX D

RESPONSE FORM

Subject

 

Condition

do

of

will

girl

her

work

one

food

wish

59_
APPENDIX D

RESPONSE FORM

Subject

Condition

Second Order

 

Five Word

 

1. I put up to like

2. here all of stOp tree
3. stOp egg run with good
A. feet is big ten good
5. nest milk one sun said
6. man sun sat with milk
7. late time to wish you
8. is said five feet pOp
9. I milk p0p come zoo
10. eat egg take up ten

Comments

Normal Sentences

 

Three Word

 

1. I will

2. the cat

3. I know

A. stOp the

5. you —work

6. I am big

7. IV eat

8. I like

9. I like

10. lock the

Comments

 

60
APPENDIX D

RESPONSE FORM

Subject
Condition

played

Normal Sentences

 

 

Five Word
1. we went
2. that girl

3. she ran
A. we play
5. you do
6. that boy
7. I was

8. that boy
9. five men
take

10. you

Comments

 

61
APPENDIX D
RESPONSE FORM

Subject

 

Condition

to the zoo

is not good
to the store
in the snow
like the snow

is with me

good boy
will play here
came to work

that cat out

APPENDIX E

 

 

62
APPENDIX E

MEAN PERCENT CORRECT SCORES

 

~.mm m.om H.mm m.~m Hmqu cacao
m.mm o.ooa m.wm 0.00H Houoe
m.mm o.ooa m.mm 0.00H oposIo>Hm
~.mm o.ooa m.mm o.ooa opozIooaee monopcom Hmanoz
m.mm «.mm o.mm m.mm Hmpoa
m.wm m.mw m.~m o.mm oaozIo>Hm
m.bm m.mm =.wm m.>m onozloonea noono ocooom
m.mw m.mm m.nm m.mm Hmuoa
~.om o.m~ m.~> m.am opozIo>Hm
pm :.mm m.>m o.ooa opozIoosnB hoono umufim
Hmpoa 00: com Hospos newcoq oocopcom coameaxondd<
Hm>noch measefipmpopCH Hmfipcopsom

 

 

.epwcoa monopcon can coaumsaxouedm Homecoueom

no nacho he Hm>nopca m5HSEHpnpousd non monoon poonnoo unwound coo: .m manna

63
APPENDIX E

MEAN PERCENT CORRECT SCORES

 

 

b.mm m.mm o.mm m.mm Hmpop cacao
m.om o.ooH m.mw «.mm Hmpos
m.mm o.ooa m.mw o.m> onozIo>Hm
=.mm o.ooa m.mm =.mm opoSIoohne come 00:
m.mm 9mm m.mm mém Hence
m.~m m.mm m.hm m.- onozIo>Hm
=.mm m.mm :.mm m.~m onozloonea some oom
:.~m o.ooH m.mm w.mm Hmuoe
m.mm o.ooa o.mm m.Hm opoSIo>am
w.mm o.ooa m.~m o.ooa opoSIooueB Hm>uopcfi amenoz
Hospoc oem uma Hm>nopeH
Hmpoa soameonnoa¢ Hmapcopcom camcoq mocopeom mSHSEaumuoneH

 

 

 

.npwcoa monopcom one Hm>nopsa meassfiumnmpca
an coapwefixondam Hmfiucopcom no nooho you nonoom poonuoo pcoonoa coo: .m canoe

 

 

___

APPENDIX F

6A»
APPENDIX F

PERCENT CORRECT SCORES

RANGES OF SUBJECTS'

 

 

00H manna smnww onozucsem
00H omIQOH w.mwls.mm opoztoouea oo:
mmIOOH thmm mmImm onozIo>Hm
>.mmIooa h.mmIooa m.mmIooa onozloonea oom
ooa NmIOOH mewm onozIo>am
can omIooH ooa onozIoopnB amenoz
Hmsnoz pew umH Hm>poch
cofipmefixopaa< HmfiuCopcom sumsoq oocopcom meassfiumnoch

 

 

.euwcoa mocopsom cam Hm>uopca mSHSEaumpopnfi
an sodumsdxosdeo_amapcopsom mom nouoom poonnoo .muoonnSm ho mowcmm .z oHnt.\

LIST OF REFERENCES

“to

.‘u'l._~. '..

[Ir-m;- ‘ 5-1“?5; I Tarn: ‘3‘ ‘11.

LIST OF REFERENCES

Aaronson, D., Temporal factors in perception and short-term
memory, Psychol. Bull., 67, 130- 1AA (1967).

 

Aaronson, D., Markowitz, N., and Shapiro, H., Perception
and immediate recall of normal and "compressed" audi-
tory sequences, Percept. and Psychophys., 2, 338-3AA
(1971).

American National Standard Specification for Audiometers
(ANSI 83.6-1969; revision 0? 229.5-1951, 225.12-1952,
and 22A.l3-1953). New York: American National Stane
dards Institute (1970).

 

Beasley, D., and Beasley, D., Auditory reassembley abilities
of Black and white first- and third-grade children,
J. Speech Hearing Res., 16, 213-221 (1973).

 

Beasley, D., and Flaherty-Rintelmann, A., Children's per-
ception of temporally distorted sentential approxima-
tions of varying length, Audiology: J. Aud. Communic.,
In press (1976).

 

Beasley, D., Fbrman, B., and Rintelmann, W., Perception of
time-compressed CNC monosyllables, J. aud. Res., 12,
71-75 (1972).

Beasley, D., Maki, J., and Orchik, D., Children's perception
of time-compressed speech on two measures of Speech
discrimination, J. Speech Hearing DiS., In press (1976).

 

 

Beasley, D., Schwimer, S., and Rintelmann, W., Intelligi—
ibility of time- compressed CNC monosyllables, J. Speech
Hearing Res., 15, 3A0— 350 (1972).

 

 

Beasley, D., and Shriner, T., Auditory analysis of temporal-
1y distorted sentential approximations, Audiology: J.
Aud. Communic., 12, 262- 271 (1973).

 

 

Birch, H., and Belmont, L. , Auditory-visual integration,
intelligence and reading ability in school children,
Percept_. Mot. ,Skills, 20, 295- 305 (1965).

 

65

 

 

.
m_

 

66.

Bocca, E., and Calearo, C., Central hearing processes. In
J. Jerger (Ed.), Modern Deve10pments in Audiology.
New York: Academic (1963)

 

Broadbent, D., Perception and Communication. London: Per-
magon (1958).

 

Bryan, T., The effect of forced mediation upon short-term
memory of children with learning disabilities, J. Learn.
Disabil., l9: 605-609 (1972).

Calearo, C., and Lazzaroni, A., Speech intelligibility in
relation to the Speed of the message, Laryngosc0pe,

_6__7__, 1710-1719 (1957).

 

Carrow, E., and Mauldin, M., Children's recall of approxi-
mations to English, J. Speech Hearing Res., 16, 201-
212 (1973).

deHirsch, K., Concepts related to normal reading processes
and their application to reading pathology. In S. Sapir
and A. Nitzburg (Eds.), Children with Learning Prob-
lems. New York: Brunner7Mazel (1973).

 

 

Freeman, B., and Beasley, D., Performance of reading impair-
ed and normal reading children on temporally altered
monosyllables and Sentential stimuli. Unpublished
manuscript, Michigan State university (1976).

Freeman, B., Beasley, D., and Overholt, 6., Performance of
learning disabled children on temporally altered speech
stimuli, Unpublished manuscript, Michigan State Uni-
versity (1975).

Giattino, J., Analysis of children's elicited imitation of
sentences controlled for length and complexity. Paper
presented at Annual Convention of the American Speech
and Hearing Association, Detroit (1973).

Harris, J., Combinations of distortions in speech: The
twenty-five per cent safety factor by multiple cueing,
Arch. otolaryng., 12, 227-232 (1960).

 

Hirsh, I., Auditory perception of temporal order, J. Acous.
Soc. Ameri., _;, 759-767 (1959).

Hodgson, W., A comparison of WIPI and PB-K test scores.
Paper presented at Annual Convention of the American
Speech and Hearing Association, Las Vegas (197“).

 

Jerger, J., Audiological manifestations of lesions in the
audétory nervous system, Laryngosc0pe, 70, “17—h25
19 0). ’—

 

 

67

Katz, J., Auditory perception training for children with
learning disabilities, Menorah Medical J., g, 18-29

(1971).

King, K., and Weston, P., The effect of percentage of time-
compression, sentence length, and age on children's
performance of well-formed sentences. Paper presented
at Annual Convention of the American Speech and Hear-
ing Association, Las Vegas (197A).

 

Kuhl, P., and Speaks, C., Temporal processing of speech.
Paper presented at Annual Convention of the Acoustical
Society of America (1972).

Kurdziel, S., and Noffsinger, D. Performance of cortical
lesion patients on “0% and 60% time-compressed speech
materials. Paper presented at Annual Convention of the
American Speech and Hearing Association, Detroit (1973).

Lerner, J., Children with Learning Disabilities: Theories,
DiagnosIS, and Teaching Strategies. Boston: Houghton

 

 

Manning, W., Johnston, K., and Beasley, D., The performance
of children with auditory perceptual disorders on a
time-compressed speech discrimination measure. Paper
presented at Annual Convention of the American Speech
and Hearing Association, Washington (1975).

Mattingly, I., Reading, the linguistic process, and linguis-
tic awareness. In J. Kavanagh and I. Mattingly (Eds.),
Language by_Ear and by Eye: The Relationships between
Speech and Reading. Cambridge: MIT (I972).

 

 

Matzker, J., The binaural test, Internat. Audiology, 1,
209-211 (1962).

 

McCrosky, B., and Thompson, N., Camparison of rate controlled
Speech by children with specific learning disabilities,
J. learn. Disabil., 6, 621-627 (1973).

 

McGrady, H., and Olson, D., Visual and auditory learning
processes in normal children and children with specif-
ic learning disabilities, Except. Child., 36, 581-589

(1970).

’McKee, P., et al., Basal Vocabulary from ON WE GO, Teacher's
Guide. Boston: Houghton Mifflin (1966).

 

 

Miller, G., The magical number seven, plus or minus two:
Some limits on our capacity for processing information,
Psychol. Rev., 63, 81-97 (1956).

 

68

Miscik, J., Smith, J., Hamm, N., Deffenbacher, J., and
Brown, E., Short-term retention of auditory sequences
as a.function of stimulus duration, interstimulus in-
terval, and encoding technique, J. exp. Psychol., 96,
1&7-151 (1972).

 

Norman, D., Memory and Attention. New York: John Wiley
and Sons (I969).

 

Pantalos, J., Schuckers, G., and Hipskind, N., Sentence
length-duration relationships in an auditory assembly
task, J. Comm. Dis., 8, 61-7” (1975).

 

Rosenthal, w., The role of perception in child language dis-
orders: A theory based on faulty signal detection
strategies. Paper presented at Annual Convention of
the American Speech and Hearing Association, Las
Vegas (1974).

Schuckers, G., Shriner, T., and Daniloff, B., Auditory as-
sembly of segmented sentences by Children, J. Speech
Hearing Res., ‘1g, 116-127 (1973).

 

 

Senf, G., and Freundl, P., Memory and attention factors in
specific learning disabilities, J. learn. Disabil.,
5. 99—106 (1971).

 

Shriner, T., and Daniloff, B., Reassembly of segmented CVC
syllables by children, J. Speech Hearing Res., 13,
537-547 (1970).

 

Smith, A., and Beasley, D., Short-term memory and imitation:
Effects of sentence length, sentence type, stress, and
grade. Paper presented at Annual Convention of the
American Speech and Hearing Association, Detroit (1973).

Speaks, C., and Jerger, J., Method for measurement of speech
idegtification, J. Speech Hearing Res., 8, 185-19“
19 5). —

Stark J. Reading failure: A language based problem, Asha,
$32—$3u (1975).

 

Thompson, N., Comprehension of rate controlled speech of
varying linguistic complexity by normal children.
Paper presented at Annual Convention of the American
Speech and Hearing Association, Detroit (1973).

Tejirian, B., Syntactic and semantic structure in the recall
of orders of approximations to English, J. verb. Learn.
verb. Behav., 1, 1010-1015 (1968).

 

 

69

Willeford, J., Audiological evaluation of central auditory
disorders, Maico Audiol. Libr. Series, 6 (1,2) (1967).

 

Willeford, J., Central auditory function in children with
learning disabilities. Paper presented at Annual Con-
vention of the American Speech and Hearing Association,
Las Vegas (197M).

 

 

”7171177717771? 77!!!! [77171171777777