ERROR MIGRATION: FACTOR FOR CONSIDERATION
WHEN DEVELOPING NORMATIVE PHONOLOGICAL DATA
An Honors Thesis (ID 499)
by
SARA MCDERMOTT
ADVISOR:
DR. CHARLES W. MARTIN
/
BALL STATE UNIVERSITY
MUNCIE, INDIANA
August, 1987
Expected Graduation:
8-15-87
TABLE OF CONTENTS
LIST OF CHARTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
iii
LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
iv
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
REVIEW OF THE LITERATURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5
METHOD . . . . . . . . . . . . . . . . . . . . . • . . . . . . . . . . . . . . • • . . • . . . . . . . . . . . .
17
Selection of Subjects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
17
Materia,ls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
17
Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
17
FINDINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
19
SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
21
.. .... . . . ... .. ..... ............. ..... ..... .
21
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
21
Suggestions for Further Research . . . . . . . . . . • . . . . . • . . . . .
22
...........................................
23
APPENDIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
24
BIBLIOGRAPH'p' . . . . • . . . . . . • . • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
28
Summary ....
Conclusions
LIST OF CHARTS
1.
Haelsig & Madison Normative Data on
Phonological Processes
iii
Page
28
LIST OF TABLES
PAGE
TABLES
1.
Data Results Form the Goldman-Fristoe Test
of Articulation:
Normal Subjects
26
2.
Summary of Mean Percentage of Occurrence on the
Three Articulation Tests used
26
3.
Individual error migration Patterns of Subjects
27
iv
I~TRODUCTION
Articulation tests have
t~o
determine the nature of the articulation
is
to
determine
how
extensive
Information obtained from an
tHO purposes
is used
the problem is
articulation test
(~o~rer,
concerning these
are available to
and final
positions of
Some tests employ single-Ford
tested.
are
include procedures
Hhile others
elicitation tasks,
1982).
each generally related in many Hays,
initial, medial,
especially in that the
Fords
the second
to develop a therapy program, provided one
speech-language clinicians,
in
problem and
Many types of articulation tests
is necessary.
phonemes
The first is to
basic purposes.
t,o obtain a
spontaneous connected speech sample.
The method
most often Gsed requires
when testing
sounds Fithin single Fords that is
of eliciting
young children
(Mm-Ter,
procedure, the client is asked to
represented
in
the
picture.
target phoneme and within
Standard
American
in pictures, especially
naming objects
1982).
say
the
] S
the Photo
Selmar,
the
~,my
and
Soder
test, each
The target
the cl ient pronounces them.
Articulation
Test
(1969)
of
the object
phoneme in
represented in each of the three
basic positions (initial, medial, final).
evaluated by
name
Each word represented contains a
the particular
English
In tests using the
developed
"consists
by
of
sounds are
As an example,
Pendergast, Dickey,
seventy-two
photographs of objects familiar to most children"
(Darley,
color
1979).
2
As
the
child
evaluates
names
how
Examples of
the
object
accurately
tests that
in the picture, the clinician
target
the
involve single-word
Developmental Articulation Test, developed
the Compton-Hutton
is
sound
pronounced.
elicitation are the
by
Hejna
(1963) and
Phonological Assessment, developed by Compton
and Hutton (1978).
Several
methods
have
been
used
spontaneous, connected speech sample.
engaging a conversation
questions
about
with
his/her
the
in
order
obtain
a
One method involves simply
Asking
client.
hobbies
to
and
interest
the client
can
elicit
a
spontaneous sample that is representative of the client's speech.
Problems
arise
if
1980).
(Weiss,
the
client is uncooperative or unresponsive
Another
method
of
obtaining
a spontaneous,
connective speech sample involves the use of pictures to elicit a
story from the client.
Articulation,
For example, the
developed
by
Goldman-Fristoe Test of
Goldman and Fristoe (1969), employs
pictures to be shown to the client, who
pictures.
Each picture
in turn
usually elicits
tells about the
more than one phoneme;
therefore, the time to administer the test is reduced.
A growing body of
research
literature
has
questioned the
validity of testing articulation with the traditional single
elicitation procedure.
single
word
articulation
psychometric criteria.
articulation
basis of 10
Research
and
tests
McCauley
language
psychometric
has
indicated
are
not
and
tests
criteria.
that available
valid
Swisher
~ord
in
(1984)
terms
of
tested 30
for preschool children on the
Results
of
the analysis
3
criteria.
Healy and Madison (1984) compared the results of single word
articulation tests and connected speech samples in speech delayed
children.
They discovered
difference
in
the
They
procedures.
both a
quantitative and qualitative
recorded
errors
questioned
whether
from
the
single
two
testing
word articulation
test accurately measure articulation abilities.
Although
pathologist
testing
standardized
with
information
standardized procedures
provides
the
error
patterns,
regarding
speech
such
may result in an unrepresentative speech
Since a representative sample of speech
sample (Schwartz, 1983).
is necessary for effective remediation of articulatory disorders,
the
contemporary
beneficial to
compare frequency
including error
pathologist
speech-language
migration
may
find
it
and type of articulation error,
between
single
word
production and
connected speech samples (Healy and Madison, 1983).
Considering the
clinical significance of the often observed
discrepancy
between
single word
stimuli and connected speech samples, there has been
limited
articulation
description
increased frequency
of
it
(Healy
of articulation
performance
and
Faircloth,
migration
have
Faircloth, 1970;
1970).
shown
Current
significant
DuBois and
Madison,
response
1983).
to
The
errors as a function of the
type of test given has been termed "error
and
in
migration"
research
finding
variability
(Faircloth
on
(Faircloth
error
and
Bernthal, 1978; Johnson, Winney and
Pederson, 1980; Healy and Madison, 1983).
4
The present study seeks to further
Faircloth
and
Johnson et. al.
(1983).
frequency
Faircloth
(1980)
The focus
and
and,
in this
type
migration between
(1970),
of
in
DuBois
and
particular,
study will
articulation
single word
the line
of research of
Bernthal (1978),
Healy
and Madison
be a comparison of both
error,
including
error
productions and continuous speech
samples when vocabulary is held constant.
5
REVIEW OF THE LITERATURE
Research has been done
test
results
and
comparing
connected
1)
a comparison
been
on
single
word
divided
into
For
two major
of normal children's and delayed
children's phonological error patterns
errors
word articulation
speech articulation results.
this study, previous research has
categories:
single
a
and 2)
comparison of
articulation tests and connected speech
articulation tests.
NORMAL VS. DELAYED CHILDREN
Some researchers disagree on
in
articulation
error
patterns
whether there
is a difference
between children who differ in
patterns
are incorrect
production due to internalized speech sound rules.
Holloway and
overall
Daniloff
language
(1969)
normal speaking
compared
articulatory
children in
30 children in grades
and 30
Error
performance.
the area of syntax.
1-3 who
normal children.
defective
had severe
Each
children
to
The study used
articulation problems
child had an individual language
sample recorded and analyzed for syntax and sentence length.
The
results revealed a significant difference between the two groups.
The articulation problem children
and
produced
utterances
than the
normal
language
delay
that
children's
is
related
had
smaller
were less grammatically complete
utterances.
to
sentence lengths
They
articulation
concluded that
problems.
Often,
6
articulation
defective
children
differ
from
normal
children
because of language acquisition problems.
(1973)
Panagros
made
Holloway, and Daniloff
an
hypothesis based on the Shriner,
study.
(1969)
He
knew
that children
usually learn to say words with one consonant and one vowel first
(open
His
syllable).
have
children
hypothesis
articulation
the
omitted
that
sound.
though
He
they
they
are
believed
No
acquisition and articulation interact.
language delayed
because
problems
utterances to open syllables, even
producing
was
reduce
capable of
that
data was
language
provided to
support this position empirically.
Oller (1973)
found that children with articulation problems
usually have similar error patterns. These patterns are different
than the ones used by normally developing children.
Oller used 5
subjects with articulation problems.
He
of the
He discovered consistencies
subject's language
samples.
carefully analyzed each
in the types of errors made, not obvious through other methods of
analysis.
He concluded
that children
do formulate sound rules
and diagnostics should determine what those rules are.
Only then
may the teacher correct the erroneous rules.
In the
study done
by Schwartz, Leonard, Folger, and Wilcox
(1980) no
substantial
patterns
between
children.
This study differed in findings from the one
(1973).
differences
normal
were
speaking
and
found
in phonological
language
disordered
of Oller
The subjects of the two groups were matched on the basis
of mean utterance length,
sex, and
cognitive development. Three
7
of the children were enrolled in remedial therapy programs at the
The
time of the study.
seven
months
determined
to
children ranged
three
that
the
linguistically at
seven
years,
three
other
a normal
rate.
in age
from two years,
months.
The researchers
were
children
developing
This group ranged in age from
one year, seven months
to one
children had
100 non-imitative utterances recorded and
at least
year, nine
months.
A phonological
analyzed on the basis of phonological processes.
process is
a procedure
attempting to
Kresheck,
used by children to simplifY speech when
produce
adult
The
1983).
disordered
the
revealed
speech
children,
with
normal spealring children evidenced
while none
of the
Also, the groups
children,
used
deletion
process.
occurrence for
group that
The
similar
of
final
the
evidenced each
normal
speaking and
process
Two of the
of deaspiration
children used this process.
in
that
for
five
of
the six
consonant was the most frequently
differences
each process
of
that the processes
one exception.
language delayed
were
Hurryman, and
(Nicolosi,
words
results
evidenced were the same in
language
Each of the
included
and the
process.
the
frequency
of
amount of children in each
However, these differences
were not considered to be substantial.
Schwartz,
et.
al.
suggest
in
conclusion
there may be a
significant relationship between the acquisition of phonology and
the development
of other
aspects of
linguistic behavior.
would account for the difference of findings in
did not control for language delay.
This
the studies that
8
Hodson and
Paden (1981) analyzed articulatory errors of 60
"essentially unintelligible" children between 3 and
These
children
had
been
described
as
having
misarticulations. They were each being seen at the
Illinois
for
diagnostics
articulation
analyzed
"intelligible" 4
individually
60
of
year olds for comparison.
to
the
multiple
University of
The researchers also
therapy.
errors
responded
Processes.
and
8 years old.
normally
developing
Each of the subjects
Assessment
of
Phonological
Verified transcriptions were then transferred to the
analysis form of The Assessment of Phonological Processes.
The results of the study revealed
evidenced
phonological
evidenced
liquid
processes.
deviation,
that all
of the children
The unintelligible children
cluster
reduction,
deletion, stopping and assimilation (see Appendix A).
of
the
intelligible
processes.
The
evidenced velar
children
most
severely
deviation,
syllable reduction,
demonstrated
each
unintelligible
backing,
prevocalic voicing
final
stridency
Only a few
of
subjects
these
also
consonant deletion,
and glottal replacement.
Many of the intelligible children evidenced examples of devoicing
of word-final
obstruents.
Also, some of them
strident phonemes to replace
non-strident
error closely
approximated adult
did occur did
not
affect
produced anterior
interdentals.
These
models and the deviations that
intelligibility.
Hodson
and Paden
concluded there are specific error patterns which are predictable
in unintelligible speech.
patterns
differ
from
the
Further, they stated that
ones
used
by
normally
these error
developing
9
children in both the number and type.
Nelson and Kamhi (1984) compared the relationship of syntax,
semantics and
phonology in
Each subject
children.
During each
utterances of four normal 2 year old
was seen
for at
least eight sessions.
session the clinician gave a sentence imitation task
with nonsense
The sentence
words, and
recorded a
imitation task
spontaneous speech sample.
was controlled
for sentence length
and complexity and phonological complexity.
The researchers analyzed the sentence imitation task for the
relationship
sentence
complexity
phonological
accuracy
between
and
complexity
of
between
of
consonant
grammatical
nonsense
words.
Consonants Correct in target nonsense words
consonant production
production
accuracy (Shriberg
Grammatical errors were measured by the
errors
and
and
the
The Percentage of
was used
to measure
and Kwiatkowski, 1982).
percentage of omissions,
substitutions, additions and transpositions.
The
researchers
consonant production
sentences
found
errors as
increased.
complex nonsense
that
words
The
all
four
subjects made more
the complexity
productions
improved
over
of
time,
of the imitative
low phonologically
whereas
phonologically complex words remained in error.
made
more
grammatical
errors
as
the high
Only one subject
phonological
complexity
revealed
consistent
increased.
The
spontaneous
speech
samples
relationship between sentence length
production.
The researchers
and
accuracy
no
of consonant
used the number of propositions to
10
measure
semantic
It
complexity.
was
found
that
consonant
production was not related to semantic complexity.
The
suggested
authors
in
accuracy and sentence complexity
when
child
the
are
going
is
conclusion
the
most
be
length of
correlated
with
phonological
closely related
transitional
through
(reorganizing internalized speech production
words, the
that
periods
In other
rules).
the utterance can sometimes but not always
correct
consonant
in
production
normal
children.
A
study
was
done
to analyze the relationship between the
phonological processes of final consonant deletion, final cluster
reduction and deletion and grammatical morpheme use for bound slz
(Richardson, Dunn, Davis,
morphemes and BE verbs
and Daly,
1984).
Goleski, Lopez
A 30-minute connected speech sample was taken
from each of ten children enrolled in articulation
children ranged
in age
form 3
years, 8
therapy.
months to
The
5 years, 10
months.
The researchers
form content
found that
words also
found that children who
deleted
"is"
speech.
However, when
(final
analyzed
and
consonant
in
significant
those children
who deleted slz
deleted slz morphemes.
The researchers
deleted slz
"was"
the
more
correlation
to
slz
and
final
bound
were
phonological errors can cause
phonological processes
cluster
morphemes
found.
words usually
forms of BE) in connected
generalized
deletion
relation
from content
and
reduction) were
Be
verbs,
no
The authors suggest that
grammatical errors
where accurate
11
This is the only case the researchers
articulation is necessary.
found
in
which
and
language
errors
articulation
interact
significantly in speech defective children.
Some researchers found differences in error patterns between
articulation problem children
and some
and
normally
developing children
did not find significant differences.
needed to determine
valid
differences
or
More research is
similarities between
these two groups.
SINGLE WORD VS. CONNECTED SPEECH
In comparing
single word
and connected speech articulation
testing, some researchers found
Others found
that connected
no
difference
in
the results.
speech results were more indicative
of the child's articulation problem than single word articulation
test
results.
phonemes
in
DuBois
the
different speech
same
and
20
Bernthal
stimulus
(1978)
words
tested 10 target
elicited
demonstrated
at
screening procedure.
instructed the
pictures.
least
three
Eighteen children ranging in
sampling methods.
age from four to six years old served as subjects.
had
by
four
articulation
Each
errors
In the continuous speech task,
of them
in
the
the testers
children to tell stories when shown multi-concept
In the modelled continuous speech task, responses were
elicited through
delayed imitation.
Single word responses were
elicited through the spontaneous picture naming task.
The researchers counted each
error
in
the
data analysis.
12
Results
showed
there
were
significantly
more
errors
in the
Also, more
continuous speech task than in the other two
tasks.
errors occurred
speech task than in
in the
the spontaneous picture
limited in
modelled continuous
naming
task.
interpretation because
The
though,
study,
only the
is
frequency of errors
was considered.
Johnson, Winney, and Pederson
(1980) did
a study comparing
both the frequency and type of articulation errors made under two
testing conditions. They tested
age
from
3.7
to
9.5
thirty-five children
years.
The children were identified as
articulation impaired through at least
connected speech
Fristoe
Test
Sentences to
screening.
of
one
phoneme
The researchers
Articulation
elicit single
ranging in
Sounds-in
error
in a
used the Goldman-
Words
and Sounds-in-
word and connected speech responses,
respectively.
Johnson, et. al. considered each error in the analysis.
results
confirmed
were less errors
speech.
The
the
in
DuBois
single
results
and Bernthal conclusions.
word
also
responses
showed
than
there
The
There
in connected
were
significant
differences in the type and pattern of errors in the two sampling
methods.
For
example,
continuous speech.
There
there
were
were more
more
omission
errors in
substitution than omission
errors in single word responses.
Pollock and
of one male
Schwartz (1983)
speech
samples were used.
disordered
analyzed the speech production
child.
Five
connected speech
These had been collected over the period from
13
3
age 3 years, 5 months to
7
examined
complexity
the
form
and
years,
The researcher
months.
of
syllable
The distribution
monosyllabic, disyllabic and trisyllabic words.
of
phonetic
in
segments
positions was determined.
syllable
initial
structures in
and syllable final
The analysis focused on the sounds the
child was capable of producing rather than whether it was correct
by adult standard targets.
Initially, The child had
many internal
rules that
allow sounds to occur in certain positions.
that
these
constraints
were
consonants, vowel-initial
This happened first in
lifted
Later samples showed
to
allow
then
words,
important
part
in disyllabic
It was shown that
words and then partially in trisyllabic words.
For a
syllable final
syllables and syllable final clusters.
monosyllabic
syllable position was an
did not
of
sound acquisition.
period of time, the child used Idl and Inl consistently to
begin syllables and did not learn to produce other sounds in that
position.
The
are
results
indicate
important
Available
that syllable structure and position
considerations
single
word
in
articulation
articulation
tests
do
assessment.
not assess this
aspect of speech production.
A study was done comparing the results of one of five single
word
articulation
tests
to
the
results
of
a
50
utterance
connected speech sample (Traweek, Aitken, Daiy, Fomby, Perot, and
Sheehan, 1983).
years were
used
Twenty-five children between the ages of 3 and 6
as
subjects.
The
researchers
analyzed the
14
results
of
the
single
word
The
phonological processes.
speech
sample
and
and
connected speech samples for
results showed
the
single
word
that the connected
responses
phonological processes equally if the child's
identified
speech exhibited a
moderate to severe disorder.
Healy and Madison (1984) compared both frequency and type of
articulation error between single word production
and continuous
speech samples when vocabulary is held constant.
used twenty children with ages between 5.4 and
Each of
them had
conversational
demonstrated at
12.8 as subjects.
least three phoneme errors in
Single
speech.
The researchers
word
responses
were elicited
through the Weiss Comprehensive Articulation Test (WCAT:
Form).
when
The researchers instructed the
shown
samples.
a
multi-concept
picture
for
tell a story
the connected speech
The same phonemes were elicited for each task.
Each single
Results
children to
Picture
error
confirmed
that
connected speech than
classified
as
was
in
a
considered
greater
single
substitutions,
researchers found that connected
category than
in
the
number
of
words.
omissions,
speech
single word samples.
data analysis.
errors occur in
Errors
and
distortion.
revealed
percentage
errors.
of
more
The
of each
The major difference between
these findings and that of Johnson, et. al. is the
types of
were further
proportion of
For example, Johnson, et. al. showed a higher
omission
errors
and
a
lower
percentage
of
distortion errors than Madison and Healy did.
Kenney, Prather,
Mooney and Jeruzal (1984) used 30 children
15
aged 4.4 to 4.8.
normal
rate.
pictures for
Each
of these
Eight
each of
children were
specific
phonemes
the three
developing at a
were elicited through
articulation tests.
The tests
consisted of a nonsense test (multisyllabic nonsense utterances),
a single word test, and a story-retell test.
The results showed
similarly in
the number
word and story tests.
test,
that
though
the
the
and type
children performed
of articulation errors in the
They did make more errors
on the nonsense
difference was not statistically significant.
The researchers also found that
fewer omissions
preschool
boys
produced
However,
than girls.
more
errors but
this is not the focus of
the present study.
Paden and Moss (1985) found no difference in
single word
and connected
speech articulation tests.
subjects were aged 4.11, 6.11, and 7.6.
for
phonological
the results of
remediation.
They were
Single
word
The three
each enrolled
responses
elicited through the Assessment of Phonological Processes
The researchers
obtained a
sample of
Natural Process Analysis (NPA).
Both
were
(APP).
continuous speech for the
of
these
responses were
evaluated separately for the Procedures for Phonological Analysis
of Children's
decided that
Language
a process
(PPACL).
The
that occurred
researchers arbitrarily
at least
50% of the time
should be considered a remediation target.
The
results
identified through
revealed
any of
no
difference
the procedures.
the APP may be the most efficient test
in
the
processes
They suggested that
tool analyzed
because it
16
took the
least time
However, these results are
to administer.
inconclusive because only three subjects were used.
Some of the researchers found that
tests
are
accurate
authors prefer
because the
in
assessing
single word
single word
single word articulation
articulation
tests over
tests are
errors.
These
connected speech samples
more time
efficient.
Other
researchers found that connected speech samples are more accurate
in
than
assessment
word
single
articulation
literature also reveals contradictory findings in
tests.
The
the comparison
of error patterns between children who differ in overall language
Some
development.
researchers
found
differences
in
error
patterns between the normally developing and the language delayed
children
and
(Hodson
Paden,
1981;
Oller,
1973).
Other
researchers found similar error patterns between these two groups
(Schwartz, Leonard, Folger, and
Wilcox, 1980;
Nelson and Kamhi,
1984) .
The present
study was
discussed, especially on the
(1984) and
Kenney, et.
a!.
designed to
studies done
(1984).
expand upon the studies
by Healy
and Madison
A comparison of the results
found in two single word tests and one connected speech test will
be reported.
children.
The subjects will be normally developing preschool
It
is
difference between
hypothesized
the results
connected speech test.
there
will
be
a significant
on the single word tests and the
It is further hypothesized
there will be
differences in the results on the two single word tests.
17
METHOD
SUBJECTS
The
(mean age
school.
subjects
in
= 3:6).
this
They
Children were
study
were
were each
ten preschool children
enrolled in
excluded if they exhibited by history or
observation any systematic disorder such as
cerebral palsy,
not used.
or cleft
Each child
a private pre-
palate.
passed
a
developmental delay,
Also bilingual children were
standard
audiometric screening
test.
MATERIALS
Each of
the subjects
of Articulation.
responded to
Half
responded to the Goldman-Fristoe Test
of
two additional
the
subjects,
articulation tests.
Weiss Comprehensive Articulation Test
Test of
Articulation
were recorded
standard score
at random,
These were the
and the Forced-Alternative
(Martin, 1984).
on the
selected
Each subject's responses
sheet for
each test using
broad phonetic transcription and were audio recorded.
PROCEDURE
Each
subject's
responses
assisted data
analysis program,
phonological
analysis
were
analyzed using a computer-
Linguist II.
identifies
21
general
The computerized
simplification
processes, t;heir total opportunities of occurrence, total
18
instances of occurrence, and percent of occurrence.
In addition,
the computer-assisted diagnostic program performs consonant error
analysis for 25 consonants of English.
The
total
total
opportunities
for
occurrence,
correct, frequency of omissions,
each initial,
medial and
and specific
and
component
program
The
correct,
percent
substitutions for
final position consonant singleton and
each initial position twotested.
program reveals the
three-element
also
consonant cluster
provides
quantitative
measures representing contrastivity, variability and accuracy for
initial, medial and final position consonant singletons.
19
FINDINGS
Chart I
provides a
graphic representation of the normative
data developed by Haelsig and Madison (1986) for the phonological
processes characteristic
of children
3-5 years of age.
Results
of the current study differed significantly from those norms (see
Table I).
For example, in the current study, gliding occurred at
a mean of 6%
subjects.
of
the
opportunities
In Haelsig
of the total
The
in increasing
Value of 1.3-2.97:
for
all
of the
and Madison's norms, gliding occurred 52%
opportunities.
processes, listed
available
following
five phonological
order of significance, had a Z-
Labial Assimilation
(La), Cluster Reduction
(CR), Vocalization (Vo), Syllable Deletion (SD), Gliding (GI).
Mean
percentage
of
error
for
processes which could be compared to
of occurrence
(see
Table II).
each
of
the phonological
the norms
ranged from 0-8%
This is significantly different
from the norms of 24-52% as provided by Haelsig and
Table I).
It can also be seen from Table II that the three tests
differed in mean
percentage
phonological processes.
occurrence of
8%
for
of
occurrence
Articulation Test)
the
two
single
and 5%
for the
Articulatior~,
of
3%
several
of the
word
on
the
elicitation tests
We iss
Compr e hens i ve
conversational speech sample
(Forced-Alternative Test of Articulation).
occurrence
on
For example, depalatalization had a mean
( =Gc..::o:....::l::...:d:::::m=a:.=.:n'--....::F::...;r"'--"'i;..::s:....:t::...:o:::...e"'""--_--'T::....e=s...;;t'--_=o..::;f_--'A=r...;;t:....::i::...:c"'-u=lc..::a:....:t:::...1::....·o=n ,
an
Madison (see
Vocalization (Vo) had
Forced-Alternative
Test
of
4% on the Goldman-Fristoe Test of Articulation, and
20
5% on the Weiss Comprehensive Articulation Test.
Table I I I provides the percentage of
each
each of the
subject that responded to all
phonological
processes
three tests.
Error migration, as exhibited by the change
percent
occurrence
of
for
error for
of
demonstrated by the data.
occurrence
on
velar
the
For
phonological
(VF)
for
articulation tests and a 16% occurrence
articulation
test.
vocalization (Vo) on
~,
0%
a
This
the
subject
(FA),
Articulation
Test
and
had an
two
on the
had
an
11%
occurrence of
the
Alternative
a
6%
(WCAT).
on
the
Subject
on
the
occurrence of
FA,
a 0% of
connected speech
on
Forced
was
single word
Test
gliding (GI) on the GF, 2%
Subject #10
the
#5 had
Goldman-Fristoe
occurrence
Articulation
processes,
example, subject
fronting
in the
of Articulation
Weiss
Test
of
Comprehensive
#7 had a 0% occurrence of
and
10%
on
the WCAT.
18% on Affricate assimilation
(Aa) on the GF, 7% on the FA, and 14% on the
WCAT.
was the
the three tests for
biggest occurrence
subject #8.
(La), and
This
difference
nasalization (Na).
difference in occurrence on
labial assimilation.
difference from
occurred
in
However, 3%
labial assimilation
For subject #9, 5% was the biggest
the three
tests.
This occurred in
21
SUMMARY AND CONCLUSIONS
SUMMARY
Ten subjects
three and
selected for this study.
Goldman-Fristoe
Test
1969).
them also
Half of
Articulation Test
four years
Each of the
of
old (mean age 3:6) were
subjects responded
to the
Articulation.
(Goldman and Fristoe,
responded to
the Weiss Comprehensive
(Weiss, 1980)
and the Forced-Alternative Test
of Articulation (Martin, 1984).
The
subjects'
responses were
analyzed with a computer-assisted data analysis program, Linguist
II.
The results
normative data
of
analysis
developed by
year old children.
these norms
the
were
then
Haelsig and
There was
compared
Madison (1986) for 3-5
a significant
difference between
and the results of the present study.
Percentage of
error for each of the
phonological
processes
for
that
three
was
computed.
was
given
all
tests
to the
also
each subject
Error
migration was demonstrated by the data.
DISCUSSION
The results of this study support the findings
Madison (1984),
Johnson, Winney
and Bernthal (1978).
Error
and Pederson
migration
was
of Healy and
(1980) and DuBois
shown
to
occur for
several of the phonological processes between the two single word
and
the
connected
speech
articulation
tests.
More
errors
occurred in connected speech responses.
The
results
differed
from
the findings of Paden and Moss
22
(1985) and Kenney, Prather, Mooney
studies
found
no
difference
and
Jeruzal
(1984).
between errors identified between
single word and connected speech articulation tests.
the present
study shows
The data in
that some subjects demonstrated a large
difference of percentage of occurrence on
between the
These
phonological processes
two types of tests, while some subjects demonstrated
This seems to indicate
a small difference.
that error migration
only occurs in some subjects.
This would explain the findings of
those studies
no observed
that indicated
difference in results
between single word and connected speech articulation tests.
This
(1984) .
the
study
furthered
the
findings
It was discovered there was a
responses
provided
by
of
the
Haelsig
differences
in
the
articulation tests.
percent of
present
and
results
Madison and Healy
significant difference in
study's
Madison
of
subjects and the norms
(1986).
between
Also,
the
Specifically, there were
two
there were
single
word
differences in the
occurrence of glides, labiodentals, deletion of final
consonants,
and
lingua-alveolars
detected
in
the
two tests.
These differences support the conclusions of McCauley and Swisher
(1984) that results will
vary
across
standardized articulation
tests due to a lack of validity.
SUGGESTIONS FOR FURTHER RESEARCH
The findings of this study could be further verified using a
larger number of subjects.
will
exhibit
error
This may also indicate which children
migration
and
which will not.
Also, more
23
standardized
single
word
study
could
tests
could
used
be
for
further
comparison.
Another
processes of a
speech.
large
The
be
done
number
results
of
may
analyzing
normal
alter
the
the phonological
children
in connected
presently
established
normative data.
CONCLUSIONS
This study found a significant difference between the normal
subjects used and the presently available normative data (Haelsig
Error migration was also found across single
and Madison, 1986).
word
and
connected
speech
frequently observed problem for
occurred a
word
few times.
tests
migration
used
in
may
be
a
normative data
on
the
Error
samples.
some
migration
subjects;
for
was a
others, it
It was also found across the two single
the
study.
These
data
indicate error
factor for consideration when establishing
development
of
phonological processes.
Further, these data suggest that the establishment of appropriate
phonological processes for remediation can be
measurement devise
used.
been indicated
further
standardized
to
tests.
a function
of the
Suggestions for further research have
understand
Further
research
the
differences between
could
also refine the
available normative data on phonological processes.
24
APPENDIX A:
DEFINITION OF PHONOLOGICAL PROCESSES - NICOLOSI,
HARRYMAN, & KRESHECK (1983).
1.
Cluster Reduction (CR):
The consonant cluster is produced as
a single consonant; e.g., spoon=/p~n/.
2.
Initial consonant deletion (ICD):
Reduction of consonantvowel (CV) or consonant-vowel-consonant (CVC) to a vowel or to a
CV; e.g., pig=/:g/.
3.
Final consonant deletion (FCD):
reduction of consonantvowel-consonant (CVC) to a consonant-vowel (CV) form; e.g.,
pig=/pl / .
the omission of a syllable in a
4.
Syllable deletion (SD):
baby=/be/.
multi-syllable word; e.g.:
5.
Velar fronting (VF):
The tendency to replace velar
consonants with alveolar ones; e.g., shoe=/su/.
6.
Depalatalization (Dp):
replacing a palatal consonant with a
non-palatal consonant; e.g., shoe=/mu/.
7.
Backing (Bk):
The substitution of /k/, /g/, /h/ and glottal
stops for non-back phonemes; e.g., tub=/t~g/.
8.
Stopping (St):
Replacement of fricatives, and occasionally
other sounds, with a stop consonant; e.g.:
seat=/tit/.
Replacement of a non-palatal consonant
9.
Palatalization (Pa):
with a palatal consonant; e. g.:
man='",,'n/.
10.
Gliding (GI):
Substitution of /w/ or /j/ (glides) for
another sound; e.g., soap=/jop/.
11.
Vocalization (Vo):
a tendency of consonants to be voiced
when preceding a vowel; e.g., pen=/b~n/.
Substitution of a labial consonant for
12.
Labialization (Lb):
e.g., top=/bap/.
a non-labial consonant;
13.
Devoicing of final consonants (DFC):
Process whereby a
child anticipates the silence following a word and substitutes a
voiceless consonant for the final consonant; e.g.,
bed=/bct/.
14.
Velar Assimilation (Va):
Tendency of apical consonants to
assimilate to a following velar consonant;
e.g.:
duck=/g:k/.
15.
Alveolar Assimilation (Aa):
substitution of an alveolar
consonant for a non-alveolar consonant to make production similar
to another labial consonant in a word;
e.g., mutt=/t t/.
16.
Nasal Assimilation (Na):
Tendency of sound to take on the
nasality of a following nasal consonant; e.g., friend=/ntnd/.
25
17.
Labial Assimilation (La):
Substitution of a labial
consonant for a non-labial consonant to make production similar
to another labial consonant in a word; e.g., top= /bop/.
18.
Affrication (Af):
The replacement of a fricative with an
affricate; e.g., sun=/tsAn/.
19.
Denasalization (Dn):
The replacement of a nasal consonant
with an oral one; e.g. no=/do/.
20.
Deaffrication (Da):
The replacement of an affricate with a
non-affricate consonant; e. g., church= 6sY .
21.
Voicing of initial consonants (VIC):
Substitutes a voiced
consonant for a voiceless consonant for the first consonant in a
word; e.g., peg=/bEg/.
)
TfB....E I
CRTA 1<I25l..l.:rs FI<D'1 n-£ €O.LJ1'N-FRISTCE TEST CF FRTIO..l.RTICN: N:f<I'FL SI..BJECTS
FIE
9.BJEJ:T
IN
2N
.....
3'1
9i
6N
7N
eN
9N
1CN
~
3.5
4.1
3.9
3.11
3.11
3.8
3.6
3.8
3.1
4.3
3.63
~
H.SI6 & I'FIJlSO
2~=
CR
a!
a!
a!
a!
(F-
a!
a!
a!
ICD
a!
3-!
a!
a!
a!
a::
a::
Fro
w
SO
EF.
1~
~
~
(F(F(F-
a!
1~
1~
S-!
a!
a!
a!
a!
a::
(F-
(F-
~
a!
Q:>
......
11~
ll~
1~
1~
13-!
~
~
11~
~
a!
a!
1~
a!
7"..!
a!
a!
a!
1~
(F-
~
(F(F-
~
a!
a!
a!
a!
a!
a!
11~
~
~
a!
a!
a!
a!
a!
a!
1~
I~
~
~
(F-
~
S-!
IS-!
la::
a::
a::
a::
a::
a::
a!
a::
a::
a::
a::
a::
a::
a::
a::
~
a!
IS-!
a::
a::
a::
a::
1~
(F-
11~
~
~
a!
S-!
S-!
a::
a::
a::
a!
a!
<t:P.
~
(F-
a!
~
l~
1~
~
~
~
~
l~
~
1~
~
1(p'
lEF.
1~
(F-
1~
3-!
13-!
~
S-!
~
~
~
2. ':17
2.47
Lb
a::
a::
a::
a::
1~
~
Reo
I,Ib
~
~
1.7'3
V...
61
~
St:.
~
a!
a!
a!
(F-
I:FC
Pa
~
Be
5(:P.
3-!
(F(F(F-
a::
(F-
a!
(F-
(F~
a::
a::
a::
a::
~
1~
1~
a!
......
IS-!
EF.
1~
S-!
EF.
3-!
1~
(F-
l~
1~
~
a!
a!
lEF.
~
~
~
~
a!
~
1~
~
~
~
~
1~
~
1~
a!
a::
AF
a!
a!
a!
a!
a::
a::
a::
(Fa::
a::
a::
a::
Lh
a!
a!
a!
0..
a!
a!
VIC
EF.
(F-
~
(F(F(F(F-
a!
a!
a!
(F-
(F-
(F-
a!
(F-
a!
a!
a!
(F-
a!
a!
a!
(F(F-
~
~
(F1~
a!
(F-
1~
22?
~
1.3
2.02
TRBLE II
SUHHRRY OF HER" PERCEHTRGE OF OCCURREHCE
,"CR
"_an weat.
OZ
OZ
OZ
0::1:
FCD
2:0:
12
12
grand/".an
OZ
O?
lZ
"_an
".081"
9f"
fa
'" = ITEHS
oor
ICD
oor
TESTED BY HRELSIG & HRDISOH
OZ
VF
OZ
OZ
lz
Dp
8Z
Sz
8Z
O?-
OZ
7Z
-SO
lor
O?
OZ
lZ
St.
2z
2Z
3Z
Pa
OZ
OZ
OZ
,"Gl
E>:Ie
E>Z
8Z
1Z
20r
OZ
7Z
Bk
lor
10r
3Z
50r
Lb
... :0:
3Z
...::1:
... or
3:0:
,"Vo
22
3Z
O?
lZ
Ra
1Z
2Z
2Z
2?-
oor
3Z
DFC
lor
Va
OZ
Ha
1z
OZ
OZ
,"La
30r
32
5z
Rt"
OZ
OZ
0::1:
On
OZ
OZ
OZ
Da
OZ
3Z
3::1:
VIC
OZ
OZ
O?-
02
... ?-
OZ
O?-
2Z
oz
)
------ ------ ------ -----SUBJECT
TEST
"lH
"IN
"IN
GF
FA
NCAT
5H
SF
5M
5H
7H
7H
7"
CR
FCD
SO
OZ
02
1Z
S?
lz
OZ
3;.:
OZ
"I;:
(J;(
OZ
OZ
OZ
OZ
OZ
FA
JfCAT
OZ
OZ
OZ
OZ
3?3:>:
2:>:
12
SF
OZ
0:1:
OZ
OZ
2:1:
0:>1:
12
OZ
OZ
OZ
oz
I:>:
12
2:>:
2:>:
3:>:
OZ
0:>:
0:>1:
52
3z
1%
OZ
52
02
02
oz
FA
NCAT
02
ICD
------ ------
8"
8M
8N
SF
OZ
FA
JfCAT
OZ
9M
SF
9H
OZ
OZ
OZ
9H
FA
NCAT
10M
SF
loti
loti
NCAr
FA
OZ
OZ
OZ
oz
2%
1:>1:
12
Ill:
32
OZ
2%
12
OZ
OZ
2Z
oz
i"Af;t
t T1
~
------ ------ ------ ------ ------ ------ ------ ------ ------ ------ ------ ------ - +---- ------ --
-- ----
------
0",
VIC
INDIVIDUAL ERROR HIGRATIOH PArTERNS OF SUBJECTS
VF
Dp
Bk
P ...
Gl
Vo
Sz
2:::
1_
oz
10%
0:>:
0;':
0;.:
2'1
0;.:
0;.:
o?
oz
OZ
.. :>:
Oz
0:::
02
0:>:
IG?0;':
10:::
Oz
G;':
I:>:
52
2?
"I:>:
8Z
W-
7;':
0;.:
IS:::
S?
10Z
0:::
0:::
0:::
0:::
0;.:
OZ
0:>:
50'!:
12
2Z
30'!:
3:>1:
0;.:
02
02
0:::
10:::
0;.:
02
0:>:
0:::
o:>!
0:::
02
2:>:
0:::
3:>:
1:>1:
OZ
OZ
0:>1:
OZ
0:>:
0'1
0:>:
OZ
0'1
0;.:
W-
0:>1:
0;(
0:>1:
OZ
0'1
0:>1:
0:>:
2;(
OZ
OZ
OZ
10Z
352
13;(
12
7%
1:1:
OZ
0%
IS?
112
OZ
32
1;.:
3:>:
Sl
lZ
1;(
OZ
2Z
"12
OZ
"Z
8:::
oz
2Z
2Z
32
Lb
DFC
7?
I"'
7Z
0;.:
11:::
'3:::
0:>:
G:>:
V ..
A..
H..
On
5;:
G?7;.:
o='!
.. z
0;.:
o?O?0;.:
8:>:
0;.:
2;':
O?-
0:::
0:::
oz
0;':
2?0;.:
0'1
0;.:
o?-
0:::
O:>!
o:>!
0:>:
0:>:
0;.:
0:>:
0:>:
o:>!
00'!:
2Z
2Z
00'!:
2;':
O?
0:"":
OZ
2;.:
0;.:
"I:::
1;.:
2:>!
OZ
"I:>:
5:::
7;.:
0:>:
0:>:
'3;':
0:>:
52
62
I:>:
7?
1?
0:::
"I:>:
13;.:
OZ
0:>:
O?
0;.:
1;':
1;.:
0;.:
1;.:
R;.:
..;.:
o?OZ
3;.:
1;.:
80'!:
o?-
0:>:
02
5:>:
0:>:
3:>:
1?
0;.:
OZ
O?
02
0;.:
0;.:
OO'!:
0;(
O;'t
0;.:
0;.:
2;.:
o:>!
0;':
0;.:
o:>!
0;.:
8Z
5;.:
0;.:
0:>:
0;.:
0:1:
I?
1;(
I?
OZ
0'1
0:>:
0:::
O;'t
0;(
0:::
1;'t
GO'!:
O?
O?
2:::
OZ
O?0:::
O?
O?
0:::
0;':
0:>:
0;.:
OZ
lZ
2Z
OZ
182
2Z
7Z
lOl
oz
3?
5?1:::
OZ
OZ
22
OZ
OZ
OZ
2Z
OZ
0;(
OZ
02
IS?
02
02
02
6;r
0:>1:
0;(
2Z
OZ
102
132
32
12
12
-
0:>:
1;(
O?-
1'1:1:
0;':
0;.:
0;.:
At"
O:>!
0:::
0;':
O:!
o?
La
1;.:
0;':
3;..:
'3;':
7?
2Z
02
OZ
----------_.
)
RE\,TISI()N i:-\N'D EXTI~NSIC~N H-,-~ELS]C;' & ~IAI)IS()l~
N 0 F~1\'1.A T I,rl~ IJ A 1'1:\ () N F)lI()N ()J~()C~I C~l'-IJ P]~{) (~ESSES
()
f~
.-----.-..----..-
-----,
~u
W
C)
--;:>'"
4--
h (-)
' - ' '--
LJ.J
,"1
•.... -
( ... J
Et"·,J [J
'7.1
___
"--'5
L
4'
. ~'I=:J
~'-
~5
O~
Ct.':
=~J
LE(~
'l' -',
.
I
"'- .
I
I . •'
('-~I
""--
·::L.
l..J.J
._, ,",
-'c_
- =11,. ,I
u._
<=J
l.JJ
<_~I
<1
:=:, I:~)
t--=:
l..JJ
~~., "
eJ
CL
10
,,-,~
.......
[
.
,~:::- ['I
(" F;'
C;L
\"/('1
C~R
Lo
--'
'--' ,
, L'-"I LO '-_"':' I(~.
F' F,J (-'I C" E'~~-::-- 1.:::- E ":' B"('
__ . t", t.I..
'. ___ -' ,', L
J
•A {-.., E
F) H l'J
,
~
.(-J"
_
'. , _
_.'
_-, __ I
_
I
I
'-_:I
-,
I, '.
'\ 1
\1 \1 •
t',,1I,,u,,r~:,,
~ ,', ["~,' r r"
. . r-, -7
I j ' ,::"1 i
1
BIBLIOGRAPHY
Compton, A.J. and Hutton, S.
Phonological Assessment.
Darley,
F.
(1978).
Compton and Hutton
San Francisco:
Carousel House.
(1979).
Evaluation of Appraisal Techniques in
~S~p~e~e~c~h~_~a~n~d~~L~a~n~g~u~a~g~e~~P~a~t~h~o~l~o~g~y.
Reading, Massachusetts:
Addison-Wesley.
DuBois, E.
and Bernthal, J.
(1978).
A comparison of three
methods for obtaining articulatory responses.
Journal of
Speech and Hearing Disorders, ~, 295-305.
Garn-Nunn, P.G.
(1986).
Phonological Processes and Conventional
Articulation Tests:
Considerations for Analysis.
Language, Speech,
and Hearing Services in Schools, 11,
244-252.
Goldman-Fristoe Test
Goldman, R. and Fristoe, M.C.
(1969).
American
of Articulation.
Circle Pines,
Minnesota:
Guidance Service, Inc.
Haelsig, P.C.
and Madison,
C.L.
(1986).
A Study of
Phonological Processes exhibited by 3-, 4- and 5-year-old
children.
Language, Speech, and Hearing Services in
Schools, 11, 107-114.
Healy, T. and Madison, C.
(1984).
Articulation Error
Migration:
A Comparison of Single Word and Connected Speech
Samples (unpublished paper).
Hejna, R.
(1963).
Hejna Developmental Articulation.
Michigan:
Speech Materials.
Ann Arbor,
Hodson, B.W. and Paden, E.P.
(1981).
Phonological Processes
which characterize Unintelligible and Intelligible speech in
Early Childhood.
Journal of Speech and Hearing Disorders,
46, 369-373.
Johnson, J., Winney, B., and Pederson, O.
(1980).
Single Word
vs. Connected Speech Articulation Testing.
Language,
Speech and Hearing Services in Schools, 11, 175-179.
Kenney, K., Prather, E., Mooney, M., and Jeruzal, N.
(1984).
Comparison Among Three Articulation Sampling Procedures with
Preschool Children.
Journal of Speech and Hearing Research,
~, 226-231.
Martin, C.W.
(1984).
A Forced-Alternative Test of Articulation
(unpublished paper)
National Convention of American
Speech-Language-Hearing Association, San Francisco.
McCauley, R. and Swisher, L.
(1984).
Psychometric Review of
La~guage and Articulation Tests of Preschool Children.
Journal of Speech and Hearing Disorders, 49, 39-42.
McReynolds, L. and Elbert, M.
(1981).
Criteria For
Phonological Analysis.
Journal of Speech and Hearing
Disorders, ~, 197-204.
Mower, D.E.
(1982).
Methods for Modifying Speech Behaviors:
Learning Theory in Speech Pathology.
Second Edition,
Columbus:
Charles E. Merrill Publishing.
Nelson, L. and Kamhi, A.
(1984).
Syntactic, Semantic and
Phonological Trade-offs,
(unpublished paper).
National
Convention of the American Speech-Language-Hearing
Association, San Francisco.
Nicolosi, L., Harryman, E. and Kresheck, J.
(1983).
Terminology
of Communication Disorders, second edition,
Baltimore:
Williams and Wilkins.
Oller, D.K.
(1973).
Regularities in Abnormal Child Phonology.
Journal of Speech and Hearing Disorders, 38, 36-47.
Paden, E.P., and Moss, S.A.
(1985).
Comparison of Three
Phonological Analysis Procedures.
Language, Speech, and
Hearing Services in Schools, li, 103-109.
Panagos, J.M.
(1973).
Persistence of the Open Syllable
Reinterpreted as a Symptom of Language Disorder.
Journal of
Speech and Hearing Disorders, ~, 23-31.
Pendergast, K., Dickey, S., Selmar, J. and Soder, A.
(1969).
Photo Articulation Test.
Danville, Illinois:
The
Interstate Printers and Publishers, Inc.
Pollock, K. and Schwartz, R.
(1983).
Syllabic Aspects of
Phonological Acquisition in a disordered Child,
(unpublished paper).
American Speech-Language-Hearing
Association Convention, Cincinnati.
Richardson, A., Dunn, C., Davis, B"
Goeski, L, Lopez, A. and
Da]y, J.
(1983).
Interaction of Phonology and Morphology
in Disordered Child Language
(unpublished paper).
American
Speech-·Language-Hearing Association Convention.
Ritterman, S., Zook-Herman, S.L., Carlson, R.L., and Kinde, S.
(1982).
The Pass/Fail Disparity Among Three Commonly
Employed Articulatory Screening Tests.
Journal of Speech
and Hearing Disorders, .il, 429-433.
Schwartz, R.G., Leonard, L.B. Folger, M.K. and Wilcox, M.J.
(1980).
Early Phonological Behavior in Normal-Speaking and
Language Disordered Children: Evidence for A Synergistic
View of Linguistic Disorders.
Journal of Speech and Hearing
Disorders, 45, 357-377.
Schwartz, R.G.
(1983).
Diagnosis of Speech Sound Disorders in
In Diagnosis, In Speech Language Pathology.
Children.
Baltimore: University Park.
Shriberg, L. and Kwiatkowski, J.
(1983).
Phonological Disorders
III,
A Procedure for assessing severity of involvement.
Journal of Speech and Hearing Research,
il, 256-270.
Shriberg, L. and Kwiatkowski, J.
(1985).
Continuous Speech
Sampling for Phonologic Analyses of Speech-Delayed Children.
Journal of Speech and Hearing Disorders, 50, 323-334.
Shriner, T.R., Holloway, M.S., and Daniloff, R.G.
(1969).
The
Relationship Between Articulatory Deficits and Syntax in
Speech Defective Children, Journal of Speech and Hearing
Research, ~, 319-325.
Traweek, S., Aitken, P., Daiy, K., Fomby, S., Perot, S., and
Sheehan, K.
(1983).
Look What We Found in Traditional
Articulation Testing.
(unpublished paper)
National
Convention of American Speech-Language-Hearing Association,
Cincinnati.
Weiss, C.
(1980).
Weiss Comprehensive Articulation Test,
Hingham, Massachusetts:
Teaching Resources Corporation.