Dr. M.

V SHETTY COLLEGE OF SPEECH AND HEARING

DEPARTMENT OF SPEECH DIAGNOSTICS

STUTTERING PREDICTION INSTRUMENT(SPI)

BACKGROUND:
Practicing clinicians working with children who stutter face a variety of
complex decisions; should this child be placed in the case load? Should the parents
receive counseling and; If so, what are the main goals? Does the child need
modification of attitudes about fluency? Does the disfluency behavior need
modification? The need for an instrument to help answer these and similar questions
are evident. The problems of quantifying severity and predicting chronicity are not
simple and any instrument will provide only partial satisfactions. With these complexities
in mind, a set of six criteria was devised to guide development of the “Stuttering
Prediction Instrument” (SPI).

1. It should be simple enough for routine clinical use by trained observers.

2. It should quantify behaviors which are consistently identified as abnormal
disfluencies and assign weights to them compatible with experienced clinical
judgment and, where possible, research findings.
3. It should have reliability and validity characteristics to meet American
Psychological Association standards.
4. It should be standardized on a large representative sample of children who
stutter, and a control group.
5. It should yield data useful in developing and monitoring treatment goals.
6. It should have apparent validity as a predictive instrument to justify longitudinal
studies.

Criteria 1 and 5 can be considered together as they both deal with the
practical aspects of the instrument. Every effort was made to describe each item
on the test in clinical terms rather than in terms requiring knowledge of
spectrographs, fiberoptics, or other research instruments, For eg: abruptly
separated syllable turned out to be more meaningful than air-flow breaks to
practicing clinicians. Also, judgments of duration of prolongations were made
more reliably than judgments of degree of tension. In both cases, the simpler
terms were used. Duration was estimated rather than timed by stop watch.
Researchers may want to use more sophisticated timing methods, but clinicians
can use estimated duration and still achieve adequate reliability. Another
important concept which makes this instrument useful to the clinician is that the
intuitive definition of stuttering behaviors is retained. The fact that untrained
listeners do as well as trained listeners in judgment of presence of stuttering in
an utterance (Schluvetti, 1975; Culatta, 1977; McDonald and Martin, 1973;
Huliti,1978) supports this approach. Some of the relations of each item of the
instrument to diagnostic and treatment decisions are explained in the text.

Criterion 2 deals with selecting and quantifying behaviors that define

abnormal disfluencies. In recent years, descriptions of disfluency behaviors have
been characterized by disfluency types. Two types of disfluency- part word
repetitions and prolongations- are closely related to perceived stuttering and are
found significantly more often in children diagnosed as stutterers than in children
not so diagnosed. The repetition of sounds and syllables including part-word
repetitions and sometimes the repetition of one syllable words, was found to be
significantly related to 1978; Wingate, 1977; Conture, 1975; Sheehan, 1974;
Silverman, 1974). Prolongation affected various aspects of the syllable including
initiation of a consonant , initiation of a vowel, an abnormal duration of voice
production. All of the studies reviewed found some forms of prolongation
significantly related to stuttering. Previously van riper in 1971 performed an
extensive review of research related to normal and abnormal disfluencies.
Repetition and prolongations including vowel prolongation including phonatory
errors, articulatory posturing were among the disfluencies types associated with
abnormal disfluencies.
The degree and the type of reactions to disfluencies by the child and by the
parents have also been considered important in the development and
maintenance of abnormal disfluency (Van Riper, 1971; Coper, 1972; Kasprisin-
Brusilli, Egolf and Shames, 1972; Yari, 1973; Crowe and Cooper, 1977).
Based on the review of previous studies, the following definitions derived:

Stuttering occurs when sounds or syllables are repeated, vowels are prolonged,
or the air-voice stream is arrested at the laryngeal level or impeded by abnormal
articulatrory postures. It is aggravated buy the child’s reactions to the disfluencies
and by attempts to avoid sounds, words, situations, or audiences associated with
the disfluencies.

Criteria 3 and 4 with the reliability, validity, and standardiazation of the

instrument. Normative data are provided based on eighty five children who
stutter, eleven children whose disfluencies ere considered high risk but who did
not develop stuttering within two years of the time they were examined, and six
children who are being monitored because of potential stuttering. Adequate
reliability and validity studies have been conducted. Clearly, more extensive
normative data are desirable, and larger longitudinal studies are sorely needed.
These available data were considered adequate to make the instrument available
to any interested professionals while extensive research is being conducted.
Criterion 6, predictive usefulness of the instrument, was not met in a definitive
manner. The small sample (n=11) used in the longitudinal study did yield useful
results which indicate a reasonable probability of quantifying the process of
determining chronicity. Also, some qualitative data emerged. Taken together,
these data should provide useful guidelines for case load selection until more
definitive studies are completed.

Preliminary study
A review of attempted predictive testing of young children who stutter and the
conclusions of clinicians who had experience with the large populations of these
children resulted in a list of possible behaviors. These behavior were synthesized
and edited into six parameters and numerical values were assigned:
I. Frequency
II. Repetitions of sounds and syllables
III. Vowel prolongations
IV. Phonatory arrest
V. Articulatory posturing
Reactions to stuttering behaviors
The resulting experimental instrument, the children’s stuttering severity
instrument (Riley and Riley, 1978), and the original stuttering severity instrument
(Riley, 1972) were administered to thirty-none children, ages three through eight,
who stuttered, and to seven children who were seen for clinical evaluation for
their disfluencies but for whom treatment was not recommended. None of these
seven had developed stuttering one year after their evaluations.
Table 1 shows the ranges, means and standard deviations for the two
groups of children. The range for the stuttering group was 10-45 with a mean of
21; for the children who did not develop stuttering, the range was 4-8 with a
mean of 5.6. Inspection of these data indicates a wide difference between the
group means and ranges that do not overlap.
Table II shows the relation of each behavior on the experimental CSSI
to its total and to the parameters and total of the SSI. Based on these data, all of
the parameter except for frequency were revised.
Repetition (II) was revised by omitting co-articulation errors because
they were not correlated with the CSSI total.
Vowel prolongations (III) were revised by omitting the tension judgment
because it was not correlated with the total.
 Phonatory arrest (IV) and Articulatory Posturing (V) were revised . For
these two parameters, the amount of tension judgment were omitted because
they correlated highly (r=88) with the total duration judgment.
For Reactions (VI), numerical value was increased because it had the
highest correlation (r=78) with the total CSSI score.
The resulting form became the Stuttering Prediction Instrument For
Young Children (SPI). It was simple, clinically relevant, and showed some
promise of predicting chronicity of stuttering.
The most highly correlated parameters were assigned task score
values to reflect the part-whole r (see Table II). These parameters were
organized into four sections with possible task score ranges as follows:
Reactions 0-12
Part word repetition 0-07
Prolongations 0-12
Frequency 0-09
TOTAL 0-40
 TABLE I
CSSI scores of children who outgrow disfluencies compared to children who did not
outgrow their disfluencies within a year of examination.

Children who outgrow Children who did not

disfluencies outgrow disflencies
Mean Range Mean Range
I. Frequency 3.67 2-7 6.2 2-9
II. PWR Number 1.30 1-2 2.2 1-3
Severity 0.28 0-1 3.7 0-8
III. Vowel 0 0-0 2.1 0-8
Prolongation
IV. Phonatory Arrest 0 0-0 2.0 0-8
V. Articulatory 0 0-0 2.2 0-8
Posture
VI. Reactions 0 0-0 2.4 0-8
TOTAL SCORE: 5.6 4-8 21.0 7-45
 TABLE II
Relation of CSSI variables to its total, and to the SSI based on 40 children ages 3
through 0.

CSSI Variable Pearson Correlations with

CSSI Total SSI Total
I. Frequency 0.54 0.83
II. Part word repetition
a. Number 0.27 0.29
b. Severity 0.52 0.54
III. Vowel prolongation 0.41 0.30
IV. Phonatory Arrest 0.66 0.43
V. Articulatory Posturing 0.59 0.42
VI. Reaction 0.78 0.45
TOTAL SCORE: XXX 0.70

Confidence Levels: 0.05 = 0.22

0.01= 0.30

0.001= 0.37

0.0001= 0.41
Description:
The SPI Test Form is divided into five major sections:

Section I: History
Section II: Reactions
Section III: Part-Word Repetition
Section IV: Prolongation
Section V: Frequency

The stimulus materials, which are found in the back of this manual, can be used to elicit
conversation from the child. Included are a list of possible topics for conversation, and
five composite illustrations.

Plate I: Fire fighting

Plate II: Washing Dishes

Plate III: The garage

Plate IV: Bath time

Plate V; The Restaurant

Administration Procedures:
Administration o the SPI involves three procedures: (!) A parent
interview (Sections I and II), (2) observation and tape recording of the child’s speech,
and (3) analysis of the tape recording (Sections III,IV, and V).

Section I: History:
This section is divide into two parts: background information and
family history of stuttering. The items in Section I are important elements in the
evaluation process, but they are not scored; therefore they become part of the
examiner’s qualitative analysis but do not add to the quantitative analysis. Item 1 asks,
“when did your child first exhibit disfluencies? What were the related
circumstances?”Item 2 asks,” Is the severity of the stuttering increasing? Decreasing?
Item 3 attempts to find out if “today’s speech is more or less disfluent than usual, or
about average.” Some stuttering is so episodic that it cannot be evaluated without the
answer to that question.

Any family history of stuttering is reported in item 4 (see figure 1).

Familial stuttering is more likely to be chronic (Neaves, 1970; Cooper, 1972;
Porfert,1978;K. Kidd, Records, and J. Kidd,1978; Kidd,1980).

Section II: Reactions

The information recorded in this section is also taken from the parent
interview (see figure 2). Item 5 deals with the parent’s reaction and is scored.

0 = unconcerned

1 = concerned

2 = very concerned

Items 6 through 10 deal with the child’s reactions to the disfluencies and are scored.

0= never observed

1= observed to a mild degree

2= observed to a moderate or severe degree

The parent may want clarification and examples before

responding. During direct observation of the child, the clinician may observe some of
the behaviors which the parent did not report. If this happens, the items should be
reported. For example, the clinician may observe word substitution even though the
parent did not report it. The score on item 8 should be changed from 0 to 1 or 2. The
final score for item 5 through 10 will range from 0 to 12 and will reflect reactions
reported by the parent or observed by the clinician.

Observation and tape recording of the child’s speech is the second

of the three parts of the SPI evaluation. The clinician must engage the child in
conversation.

During this sampling, notes should be made regarding events

which are largely visual, such as the physical concomitants of stuttering, phonatory
arrest, and articulatory posturing. The goal is to tape record about ten minutes of the
child’s speech with as little of the clinician’s speech as possible. At least 100 words are
required for analysis.

Analysis of the tape recording and assigning task scores to

disfluency bahaviors make up the third phase of the SPI evaluation. Using the tape
recorded sample and the notes describing visual events, the remaining sections of the
instrument are scored.

Section III: Part-Word Repetitions

The most severe examples of part-word repetitions are scored with
regard to the number and the quality of the repeated sounds or syllables. The repeated
sounds are rarely consonants separated by rhythmic cessation of the voice-air stream
as in (v/v/very). They are frequently vowels separated by rhythmic cessation of voicing
as in ( ^/^/^p). If the voice-air stream is not interrupted, these events become
prolongations. The repeated vowel is actually a syllable, and repeated consonants are
usually made syllabic so that you hear (vε/vε/vεri). When whole one-syllable words are
repeated, they are scored as part-word repetitions; as a matter of fact, they are probably
not really whole words in many cases. For example, a child may say (bo/bo/bou) so that
only the word completes the diphthong. In conversational speech, word boundaries are
not easy to determine. Speech is produced in a stream of syllables, even though it may
be perceived in discreet words or meaningful units. The examiner should trust the
intuitive ability to know when stuttering has occurred and not be overly concerned with
research oriented definitions, since untrained observers are as reliable as trained ones
in recognizing the existence of stuttering. These repeated sounds and syllables are
scored first by the number of times the child tries before saying the target word. In the
example (vε /vε /vεri), the child tried unsuccessfully two times prior to the successful
utterance of the word. While listening to the tape, the clinician should phonetically
transcribe examples of the repetitions and score the most severe examples (see figure
3). Scoring the number of repeated sounds or syllables is as follows:
 0 = none

1 = 1 to 3 repetitions

3 = 4 or more repetitions

In addition, the repetitions are scored according to the degree of

abnormality. The clinician must listen to the repetitions as if they constituted a real word.
If there were such a word as (vε /vε), it would be said in a normal fashion. The child who
stutter may distort the repetitions in several ways: (!) the vowel may be changed so that
it does not match the target word, as in (v /v /vεri); (2) the repeated syllables may be
hurried: (3) the syllables may be abruptly separated, sounding staccato: or (4) they may
be accompanied by tension of the throat or face. The cumulative effect of these
distortions is scored as follows:

0 = normal

1 = mild

2 = moderate

3 = severe

Section IV: Prolongations

Three types of prolongations are considered and scored
separately, but only the highest scored prolongation contribute to the total score. The
three types of prolongations are vowel prolongations, phonatory arrest and articulatory
posturing (see figure 4).

Vowel prolongation occur when a vowel is held long enough to

call attention to itself. Vowel is perceived as abnormal, when its duration exceeds one
full second. These prolongations are not among the highest predictors of total SPI
scores or SSI severity (see Table II). The longest vowel prolongations are scored as
follows:

0 = less than 1.5 seconds

2 = 1.5 to 2 seconds

4 = 2 to 4 seconds

6 = more than 4 seconds

Phonatory arrest occurs when the attempt to initiate a vowel is

prevented by the abnormal closure of the glottis. The result may be complete closure,
so that the speaker, open-mouthed with no sound being produced, has the word “stuck
in his throat”, Sometimes the closure is less than complete, resulting in the production of
some sounds, but they are not vowels or even vowel-like. If the result is vowel-like, the
event is treated as vowel prolongation. The clinician can test his judgment by asking,
“Could a short segment of this be used as a vowel in a word? What vowel is it? If it
cannot be recognized as vowel material, it should be treated as a phonatory arrest,
Complete or almost complete blockage at the glottis level is easily perceived as
stuttering and correlates highly with SPI and SSI total scores. Scoring is as follows:

0 = none

4 = estimated duration is less than 1 second

8 = estimated duration 1 to 3 second

12 = estimated duration more than 3 seconds

Articulatory posturing occurs when the voice-air stream is

obstructed or severely distorted so that the production of the initial consonant of a
syllable cannot be accomplished in an acceptable amount of time. Consonants are very
short events in normally fluent speech, and any significant increased duration is
perceived as abnormal. The pressure of these abnormally long articulation posture is
highly correlated with the SPI and SSI total scores. These events are scored exactly as
phonatory arrests are scored:

0 = none

4 = estimated duration less than 1 second

8 = estimated duration less 1 to 3 seconds

12 = estimated duration more than 3 seconds

Some sounds are difficult to classify because they do not always

function as consonants or vowels. A good example of a confusing sound is (r). Unless it
is clearly acting as a vowel, i.e, (bзd), it must be scored as a prolonged articulatory
posture; (І) and (w) are generally scored as consonants under the articulatory posture
classification; (j) and (h) usually result in abnormality at the laryngeal level and are
associated with phonatory arrests. There are many idiocentric variations which the
examiner must interpret as vowel or consonant and attribute the distortion to laryngeal
or oral cavity level dysfunction. Most of these decisions will not affect the total score,
since only the highest prolongation category is used in computing the total; in fact, the
scoring system was devised to accommodate these examiner judgment differences and
still maintain adequate reliability.

Section V: Frequency
Using the tape recorded sample, the frequency count is
accomplished by counting the number of stuttering events (defined in the discussion of
Section III and IV) per one hundred words of conversational speech. While listening to
the tape, the clinician must make a dot (.) for each fluent word and a diagonal line (/) for
each stuttering event (see figure 5). There may be more than one stuttering event per
given word. At least one hundred words need to be analyzed, and the total multiplied by
100 to get a percentage of words stuttered. This percentage is then converted to a
score:

0 = 0%

2 = 1%

3 = 2% to 3%

4 = 4%

5 = 5% to 6%

6 = 7% to 9%

7 = 10% to 14%

8 = 15% to 28%

9 = more than 28%

Total Score:
The subtotal scores for Sections II through V are added together to
determine the total score. The possible range of scores is 0 to 40.
Reliability and Validity Studies:
In order to assess the inter-examiner reliability among experienced
clinicians, eight children were examined independently by two examiners. Using a
ranking method of correlation, they achieved an inter-examiner Spearman Rho of 95.

An attempt has been made to determine how much training is

required to achieve reliable results. A group of fourteen clinicians from a public school
district were given two hours of training using video and audio tape recordings. After the
training they all evaluated three cases using standard SPI protocol. A target score was
provided by the instructor for each case. Seven (50%) of the clinicians were within one-
fourth standard deviation of the target on two or three cases. For example, in one case
the target score was 29; these seven scored between 28 and 30. Five of the clinicians
scored within one-half standard deviation of the target; that is, between 27 and 31. Two
scored below 27. While there is no standard way to express this reliability as a
correlation, it is reasonable to conclude that the best seven achieved very good
reliability, the next five with adequate reliability, and only two with inadequate reliability.
Seminar students with only a minimum of practiced experience, do almost as well as the
clinicians described above. Part- whole reliability can be seen in Table III. The Pearson
correlations range from .20 to .68. The four highest correlations are Reactions (.68),
Phonatory Arrest (.60), Articulatory Posturing (.59) and Severity of Part-Word
Repetitions (.57). Frequency has a lower correlation (.42) and two items are much
lower: number of repetitions prior to uttering the target word (.20) and Vowel
Prolongation (.21). Overall, these correlations indicate that some items are better than
others at predicting the total score. They follow the relative weights assigned after the
preliminary study.

Validity are tested by comparing the SPI scores to SSI. Table III also
shows those results based on children for whom both scores were available. The SSI
weights Frequency much heavier than does the SPI. Also, the SPI yields more detailed
information about functions to disfluencies and disfluency types. Therefore, the
instruments are different, but highly correlated with each other.

Normative Samples
One hundred two children who were between 3.0 and 3.9 years of age
served as the standard for establishing normative reference points. Within this sample,
eighty-five of the children were accepted into treatment programs in public schools, a
university clinic, or a private practice. Eleven children seen by the author, were not
given treatment. Their fluency was monitored for a period of one to three years, and
abnormal disfluency had not developed. This group forms the basis for predictive
judgment about chronicity. Six additional chidren are being monitored and the intake
samples are reported here. Tables IV and V shows the distribution of SPI scores for the
two groups.

TABLE III
Pearson correlations between the 8 parts of the SPI and the SPI and SSI totals.

SPI TOTAL SSI TOTAL

Reactions .68 .42
PWR- Number .20 .15
PWR- Severity .57 .47
Vowel Prolongation .21 .18
Phonatory Arrest .60 .33
Articulatory Posturing .59 .37
Frequency .42 .72
SPI TOTAL Xxx .70
 TABLE IV
Distribution of SPI scores for 85 children ages 3-8 who stutter

Stanino Total Score Percentile Severity

1 10-11 0-4 Very mild
2 12-13 5-11
3 14-17 12-23 Mild
4 18-19 24-40
5 20-24 41-60 Moderate
6 25-28 61-77
7 29-30 78-89
8 31-35 90-96 Severe
9 36-37 97-100 Very Severe

Median = 21

Mean = 22.2

SD = 7.01
 TABLE V
Distribution of SPI scores for 17 children whose disfluencies had not become chronic

Stanino Total Score Percentile Severity

1 0-1 0-4
2 2 5-11
3 3-4 12-23
4 5 24-40 Subclinical
5 6 41-60
6 7-8 61-77
7 9 78-89
8 10 90-96 Very Mild
9 11-13 97-100 Mild

Median = 6

Mean = 6.17

SD = 3.13
 TABLE VI
Probability comparison of chronic stutterers and non-chronic stutterers on each item of
the SPL

Variable Chronic Stutterers Non-chronic t 2 Tail

stutterers Value Probability
Range Mean SD Range Mean SD
Age 3.0- 5.8 1.6 3.1- 5.7 1.3 .18 .86
8.9 7.6
Sex1. Male 1-2 1.2 .4 1-2 1.3 .5 1.49 .15
2.
Female 0-10 1.2 2.1 0-3 .4 1.0 2.59 .12
Family History 0-2 1.7 .5 0-2 1.1 .7 3.09 .01
Parent 0-2 .4 .7 0-1 .1 .2 3.39 .001
Reaction 0-2 1.3 .7 0-1 .1 .2 12.81 .001
Teased 0-2 .3 .7 0-0 0 0 4.26 .001
Child’s 0-2 .3 .6 0-1 .1 .2 2.57 .02
reaction 0-2 1.4 .8 0-1 .1 .2 13.45 .001
Word Avoid. 0-12 5.4 2.3 0-5 1.4 1.2 10.28 .001
Situation 0-3 2.3 1.0 0-3 1.4 1.0 3.52 .01
Avoid. 0-4 2.5 1.3 0-1 .2 .4 13.98 .001
Physical Con. 0-6 1.8 1.9 0-2 .5 .9 4.39 .001
Total Reaction 0-12 2.8 4.3 0-0 0 0 5.92 .001
PWR- Number 0-12 3.2 4.6 0-0 0 0 6.51 .001
PWR- Severity 2-9 6.5 1.9 0-7 2.8 1.8 7.60 .001
Vowel- 10-37 22.2 7.0 1-13 6.2 3.1 14.92 .001
Prolongation.
Phonatory-
Arrest.
Artic-.
Posturing.
Frequency
SPI TOTAL

PREDICTING CHRONOCITY:
Using the small samples of children who outgrw their abnormal
disfluencies (n= 11) as a basis, some tentative guidelines were not forth for predicting
chronocity.

Disfluency types were very useful predictors. Phonatory

arrest and articulation postures had not developed in any of the children who outgrow
their abnormal disfluencies. Vowel prolongation of less than two seconds had
developed in one of the eleven children. Therefore, children who exhibit either
phonatory arrest or articulatory postures will probably not outgrow their stuttering.
Children with vowel prolongations of two seconds or more are not likely to outgrow their
stuttering. The longer prolongations are usually accompanied by struggle behavior
which constricts the glottis, causing pitch variation and sometimes vocal fry.

All of the non-chronic children exhibit part-word

repetitions and repetitions of one-syllable words. Four of the eleven repeated the
sounds or syllable four or more times before uttering the target word thus:
(bo/bo/bo/bo/bout). The other seven children repeated three of four times before saying
the target word. Consequently, the number of repetitions prior to saying the attempted
word is not predictive. Predictive characteristics, however , were found in repeated
material. In section III, Abnormality of repeated syllables, ten of the eleven children
scored 0, and the other scored 1. These scores indicate that a child is not likely to
outgrow his stuttering if he (1) distorts the vowel with glottal tension so that it sounds
abnormal, (2) substitutes the wrong vowel, often the (^), or (3) utters the repeated
syllables in an abrupt staccato manner.

Frequency of stuttering-per-100 words ranged from a

score of 2 (1%) to 7 (10-14%) among the non-chronic group. Only two had frequencies
higher than 3%. So frequencies of 3% or less are associated with outgrowing the
problem, but some children with frequencies are high as 14% outgrow the stuttering.

Parent’s reactions to the disfluencies were not very

useful predictors. Four were very concerned, six were concerned and only on eclaimed
to be unconcerned about the disfluencies.

The child’s reactions to the disfluencies were very

predictive. Ten of the eleven children were free of any negative reactions; that is, they
(1) had not been teased; (2) did not react by frustration, anger or by giving up speech
attempts; (3) had not developed any avoidance behaviors; and (4) did not have any
physical concomitants with the disfluencies such as facial grimaces, foot stamping, etc.
Therefore, the child who has not developed any reactions to his stuttering, and who
hasn’t been teased about it, is much more likely to outgrow it without treatment.

The total SPI score for these non-chronic children ranged

from 4 to 10. One child in the stuttering group also scored 10. While the total score
should not be used as the only basis for judging chronocity, scores of 9 and under were
never associated with chronic stuttering in this sample.

A family history of stuttering occurred in only one non-

chronic child (9%). This was a boy whose father stuttered. Twenty-seven percent of the
chronic group had a family history of stuttering. This distribution is compatible with the
other studies of familial components in stuttering (Neaves, 1970; Cooper 1972; Perfert,
1978; K. Kidd, Records, and J. Kidd, 1978; Kidd, 1980). It indicated that children with a
family history of stuttering are less likely to outgrow it.

The data in Table VI, using all seventeen of the non-chronic

children, supports the guidelines described previously. Of the seventeen measures,
eleven showed very significant differences (beyond .001) between the two groups. Four
items had t values of 10 or above.

1. The child’s reaction (Item 7)

2. Physical concomitants (Item 10)
3. Total reaction score (Items 6-10 combined)
4. Severity of part-word repetition (Section 11)

Statistically, the groups were well separated, even though there was some
overlap. The total SPI scores of 10 through 13 were shared by both groups (see Tables
IV and V).

Using SPI Data in Stuttering Treatment Programs:

Severity Monitoring:
Most treatment programs need a reliable baseline from which changes
can be judged. Due to the cyclic nature of most stuttering severity, it is important to
determine the point on the cycle at which the child is functioning during the initial
examination. Parents are usually aware of this cycle and can report when the stuttering
was at its worst. If possible, assessment should be made when the parents report that
the stuttering is “about average” and again when it is considered to be “at its worst”.
These two points can then serve as controls for future measures so that they will be
more reliable. Most clinician will want to record the total score and the scores for each
part of the instrument. During treatment, some particular measure may be selected as a
routine probe; for example, a frequency count may be made at the beginning of each
treatment session, or some even more specific measures such as duration of phonatory
arrest may be selected. The total score (requiring a parent interview) can be used at
established times in the treatment program, such as after every twenty sessions. These
formal re-evaluation should be scheduled to occur at average and at its worst points on
the cycle to match the baseline samples.

Comparison of a given child’s severity to the normative groups can be

made by referring to Tables IV and V. Children who were accepted for treatment of their
abnormal disfluencies had a mean total SPI score of 22 with a standard deviation of 7.
The small (n=17) non-chronic groups had a mean of 6.17 with a Sd = 3.13.
Treatment Planning:
Direct modification of the stuttering behaviors is required for many
children who stutter. There are several widely used methods to help the child recognize
stuttering behaviors and substitute normal syllables which are released, maintained and
terminated without undue tension. An approach which organizes the various
components into a comprehensive model has been described recently (Riley and Riley,
1979; Riley and Riley, 1980). Systemic Fluency Training for Young Children is
Diagnosed to provide comprehensive training procedures for the direct management of
the young stutterer (Shine, 1980).

The parent’s reaction to the child’s disfluencies (see Section II. Item 5)
often needs to be more fully explored in parent conferences. Attitude modification may
be required to reduce the parent’s reactions to normal disfluencies, to reduce the level
of intensity of the parent’s concern about abnormal disfluencies, to remove the
manipulative effect of the child’s disfluencies on the parent, and to reduce the level of
expectation for fluency.

The child’s reaction to his stuttering (see Section II, Items 6 and 7) may
indicate a need for modifying attitudes of the child. He may need to learn to deal with
teasing, and with his feelings of frustration and helplessness when he “can’t get the
word out”.

Avoidance behaviors (see Section II, Items 8 and 9) play a major role
in stuttering among adults, and they are sometimes found in children under the age of
nine. Avoidance can occur at many levels. The child can try to avoid certain feared
words, topics, audiences, or situations. Except for word substitutions, the feared areas
can be programmed into the transfer phase of the fluency program.

Physical concomitants (see Section II. Item 10) can be well developed
even in very young children who stutter. The physical concomitant portion of the revised
SSI (Riley, 1980) is useful in quantifying these behaviors. These behaviors frequently
disappear as fluency develops. Direct modification of physical concomitants may be
required.

Management of Non-chronic Disfluent Children:

When the clinician is satisfied that a speech sample has been
obtained of a child’s disfluencies “at their worst”, this sample can be compared to the
guidelines described previously. If the child meets the guidelines for a non-chronic
classification, a monitoring approach should be undertaken. (1) The parents need to
receive traditional counseling to reduce their concerns and their reactions to the
disfluencies. (2) The communication environment of the child may need to be improved
to decrease time pressures and reduce interruptions during speaking. (3) Parents need
to be reassured that they did not “cause” the stuttering any more than parents can
cause learning disabilities. (4) Parents should be encouraged to contact the clinician if
the stuttering gets worse. (5) The child should be scheduled for a speech sample every
six months.
Topics to Elicit Conversation:
1. A birthday party
2. The zoo
3. Christmas
4. Favorite toys
5. Grandma’s and grandpa’s house
6. Favorite places to visit
7. Your bedroom
8. A trip to the dentist
9. Favorite story
10. T.V commercial
11. Favorite television show
12. Spending allowances
13. Recess
14. Best vacation
15. Planting a garden
16. Pets
17. Favorite movie
18. Shopping with mom
19. Favorite games
20. Dreams