A Core Vocabulary Approach to Inconsistent Language Disorders

BARBARA DODD, ALISON HOLM, SHARON CROSBIE, & BETH MCINTOSH.

Perinatal Research Centre, University of Queensland and Royal Brisbane and
Women's Hospital, Queensland, Australia.

Summary

Developmental language disorders are explained by theories derived from psychology,

psycholinguistics, linguistics, and medicine, with researchers developing assessment
protocols that reflect their theoretical perspective. How theory and analysis data lead to
different therapy approaches is sometimes unclear. Below is a case management plan for
a 7-year-old child with unintelligible speech. Assessment data were analyzed to address
seven case management questions regarding need for intervention, service delivery,
differential diagnosis, goals of intervention, generalizability and therapeutic benefits,
approval criteria. management and evaluation of effectiveness. Jarrod was diagnosed with
inconsistent language disorder that required intervention. He pronounced 88% of words
differently when asked three times for the name of each word in the 25 words of the
Diagnostic Assessment Test of Articulation and Phonology , each trial being separated by
another activity. Other standardized assessments supported the diagnosis of inconsistent
language disorder, which previous research suggests is associated with a deficit in the
phonological system. Intervention with a basic vocabulary approach was chosen as the
most appropriate therapy technique. Its nature and a possible application protocol are
described.

Keywords: inconsistent speech errors, intervention, phonological disorder.

Introduction

All children with unintelligible speech are unique. Clinicians consider the characteristics of
the child's speech errors, other language skills, family, educational context, medical and
social history. This information is used to deduce causal factors and maintenance of the
disease and determine whether therapy is indicated. If therapy is offered, clinicians make a
series of decisions about diagnosis, setting goals for the child and caregivers, planning
how to deliver the intervention, and monitoring its effectiveness. Presented here is a
clinical management plan for Jarrod, a 7-year-old boy with unintelligible speech.
Our management approach is based on a clinical problem-solving model (Whitworth,
Franklin, & Dodd, 2004) developed for speech-language pathologies that asks seven
questions (see Table I). There is no correct answer to the questions. Rather, the questions
elicit a system of evidence from the child's data in the context of evidence that each
clinician includes in the knowledge of theory and research, clinical experience, and
limitations of the speech-language pathology government service. . Experienced clinicians
make decisions without reference to a formal scheme, but here we use it to do our clinical
management.

Background

This document answers seven clinical management questions about Jarrod, a 7-year-old
boy who was evaluated by Holm and Crosbie (2006). It was identified by a Queensland
educational speech pathologist working in Brisbane, Australia. Jarrod's speech and
language were evaluated to provide data for a special issue of Advances in Speech
Pathology devoted to the topic of intervention for a child with phonological impairment. The
individual evaluation was carried out by an experienced speech therapist not familiar with
the child. This was carried out at school in a quiet environment, in three 75-minute
sessions, with breaks between tasks. The evaluations were video and audio recorded for
transcription and scoring.

Jarrod was compatible with the demands of the evaluation, initiated the conversation, and
responded well to encouragement. The results of the standardized tests were
communicated to Jarrod's parents, teacher, and SLP. Jarrod's data are described
elsewhere in this volume and are included in the synthesis to justify our answers to the
seven clinical management questions.

Is the intervention indicated?

Olswang and Bain (1991, p255) recommend that the decision to offer intervention (ie,
“focused, intensive stimulation designed to alter specific behaviors”) should be based on
whether a particular language skill matches another skill, and, whether there is potential
for the change. Other factors included are the caregiver, the teacher, the child, and
concerns about communication difficulties in social and academic development. Another
issue is the practical aspects of the intervention, regardless of the degree of disability. The
advantages of attending therapy must outweigh the difficulties of disposition.
Intervention for Jarrod's speech difficulties for the following reasons:

I. Their abilities profile was uneven when assessed with standardized measures.
Jarrod's speech was unintelligible, even to his mother when out of context, but his
linguistic performance was within the normal range on all subtests except
expressive vocabulary. Their verbal IQ scores have also been consistently poorer
than their intelligence index (IQ).
II. Jarrod expressed concern about unintelligibility and playing alone. His mother has
sought speech therapy intervention and recognized her current need for further
intervention. His teacher's concern was demonstrated by Jarrod's referral to a
speech therapist.
III. Jarrod's grandparents observed an inconsistent behavioral profile. There were
times when he did not follow instructions, lacked eye contact, and had a short
attention span. The psychologist's report noted, however, that Jarrod was able to
perform tasks and responded to questions with ease. He is reportedly able to
concentrate for long periods on activities that interest him. His teacher reported
poor social skills, although his mother did not express any concern. These
inconsistencies may indicate that Jarrod's speech difficulties are affecting his social
and learning behavior.

Jarrod's potential for change may be limited, however, little progress has been seen in the
last 3 years, despite several episodes of intervention from different SLPs and a year in
special education.

What is the patient's diagnosis?

The evaluation data allow a differential diagnosis of the aspects of the communication
system that is disordered, establishing linguistic patterns, severity and possible causal and
maintenance factors. Planning personalized patient management depends on identifying
the speech processing deficit(s) underlying the language problem, as knowledge
determines the choice of skills (e.g., oro-motor, auditory discrimination). or linguistic units
(e.g., sentences, phrases, words, syllables, phonemes) that should be targeted in therapy.
The speech processing chain is presented as a mental model of the processes involved in
sensation, perception, representation, phonological (word form) planning, phonetic
planning (speech sound production), and motor execution of speech. speech (eg, Dodd &
McCormack, 1995). Here we considered the specific data from Jarrod's assessment
(mental abilities involved in the discourse processing chain) and then reached conclusions
about the diagnosis, causes and maintenance factors.

The titles are a way to summarize information about Jarrod and his disorder. Their order
was loosely based on models representing phonological disorder (eg Dodd and
McCormack, 1995; Stackhouse and Wells, 1997) reflects a Jarrod profile of decay rather
than linear ordering of the models. Some aspects were considered more than once in
different sections, leading to overlap. For example, in “articulation,” Jarrod's phoneme
repertoire was examined, increasing motor involvement as a possible cause of his
difficulties. The “oro-motor abilities” sections were considered for the formal evaluation of
these skills. The information in each of the sets provided useful data in the diagnosis
process. Poor performance on a set of skills may be causal (ie, indicate a specific deficit
underlying the speech disorder), or co-morbid (ie, reflect processing problems that affect
more than one set of speech skills). A third possibility is that poor performance on one set
of skills may be a consequence of another deficit identified in the speech processing chain
(Bishop, 1997). Therefore, diagnosis involves reconciling all available data.

Hearing

Otitis media led to the insertion of collars when Jarrod was 2 years old and again when he
was 4 years old. Evaluation when he had 4;1 indicated ''functionally normal hearing for
speech and hearing at least in the better ear'' and there was no routine screening since
hearing difficulties were reported. A fluctuating hearing loss during early intervention and
language acquisition may have exacerbated Jarrod's grammatical delay, and poor auditory
attention may have contributed to his early lack of response to intervention.

Speaks

Articulation . Jarrod can produce more initial words with three consonants (/v,  , t  /)
occurring either medially or finally, and /d3/ occurring as an error. Figure 1 indicates that
only /z, 3/ do not occur at all. All vowels appear in the speech sample except for / ɪə /
which was pronounced as [?ee] and [?hee] in two productions of “ear.” A number of non-

Australian English appeared in Jarrod's speech sample There was

some distortion of the consonants and vowels according to the phonetic transcription.

Jarrod's phonetic repertoire indicates that he has adequate articulation capacity for speech
production of both consonants and vowels, however, cases of distorted phoneme
production could indicate motor impairment. More data are needed to support a peripheral
motor explanation, however, as previous research (Dodd, Holm, Crosbie, & McCormack,
2005) indicates that children who make inconsistent errors have a deficit in phonological
assembly, with effects on phonetic programming. That is, word-depleted phonological
plans may not provide enough detail for phonetic planning.

Prosody . Jarrod's prosody had a choppy quality to picture naming. This feature, however,
may have been due to his resentment of a very long assessment administration, and/or
the phonological online assembly of words as he tried to mark all phonemes. In
spontaneous speech, however, Jarrod imitated voices which show a versatile use of tone
and normal affective prosody.

Syllable shape . A variety of syllable shapes were evident in the speech samples: CV,
CVC, V, VC, CCV, CCVC, CVCC. No tri-consonant clusters were observed. The middle
and final consonants were often characterized by a glottal stop which was used as a
default consonant.

Shape of the word. There was often a discrepancy between the consonant-vowel form of
the target word and Jarrod's production. For example, in 47 inconsistent CVC word
productions on the DEAP subtest, 17% had the exact CVC form, 51% had a CV form, and
the remaining 32% had one of the following structures CV?, CV?C, CCV? , CCVC or
CVCC. The difficulty increased as the word forms became more complex. For example,
Jarrod produced the following word forms for “umbrella” (VCCCVCV): VCVVCV, VCCVCV,
VCCCVCV, VCCCCV?CVV.

There was a mismatch in the form of the target words and Jarrod's productions. Their
productions for the same word were variable in terms of consonant and vowel sequence.
Word length affected performance; even simple CVC word forms were vulnerable to error.
These characteristics indicate an impairment of the capacity for phonological assembly
(i.e., the generation of a word production plan that specifies the consonant and vowel
sequences to be produced). Phonological assembly differs from phonetic planning (i.e.,
the generation of a plan for speech production that specifies oro-motor movements) and
motor execution (i.e., the functioning of the articulatory mechanism).

Inconsistency . Jarrod's severity score on the DEAP phonology test was 44% phonemes
correct. In assessing DEAP inconsistency, Jarrod named the same 25 images three times,
each separated by another activity. He produced 22 of the words (88%) differently in at
least two of the three productions (e.g., “tongue” as and

For two of the words, the

consonants were consistent, but the vowels varied (teeth like [di] and [der] and bridge like
[w3:] and [wei]. Production inconsistency was marked for most words, especially those

with more than one syllable.

The first production of “birthday cake” was correct,

subsequent productions were incorrect in different ways (e.g., /kh/ is realized as [th] and
[ph]). Figure 1 shows a matrix describing the ways in which selected sounds are made.
While all phonemes were produced correctly except for /d3/, /z/ and /3/ in the
inconsistency test, the number of different substitutions for any sound was high (for
example /l/ was replaced by nine other sounds ). At the same time /b, j, d/ were used as
substitutes for many other sounds.

A striking feature of Jarrod's productions was the inconsistency with which the vowels

were performed. For example, were used as substitutes for /e/,

and 18 of the words (72%) in the inconsistency test caused vowel errors. In the DEAP
phonological subtest, where there are many more CVC words, the percentage of correct
vowels was 70%. Although no specific evaluation compares imitated and spontaneous
productions of the same words, analyzes of eight words (38 phonemes) that are produced
spontaneously and by imitation revealed that in spontaneous productions 21% of the
phonemes were produced correctly and in productions by imitation by 50%.

Jarrod's high inconsistency score indicates possible impairment in the phonological set
(Dodd et al., 2005). Previous research suggests that children who make inconsistent
errors typically have intact contrast understanding of the phonological system (e.g.,
awareness of phonological legality), but have poorer control performance and other
children with speech disorders in the evaluations of expression, phonological set (new
learning of words) and the motor sequence of non-verbal acts (tracing shapes).

Gold motor skills . The VMPAC (Hayden & Square, 1999) indicated that Jarrod's global
motor control was age-appropriate, but that he is below the fifth percentile with respect to
neuromuscular integrity relative to his age. Some notable problems included jaw control,
lower lip movement, and tongue control, which was more evident in speech than in
individual words. In contrast, the SLP's informal evaluation of Jarrod and his performance
on the DEAP oro-motor evaluation suggested that he did not have any abnormalities of
oral structure or oro-motor function.

Although Jarrod's performance on the VMPAC indicates poor oro-motor skills, this may
reflect difficulties in planning sequences of oral movements rather than having altered
neuromuscular integrity. Since there is no history of oromotor difficulties (e.g., feeding,
dribbling, oromotor games such as blowing bubbles), his poor score may be related to
performing unknown oromotor actions. Children with inconsistent speech errors have
difficulty learning new phoneme strings compared to other children with speech disorders
(Bradford & Dodd, 1996). Producing sequences of unfamiliar speech sounds can cause
errors even in typical speakers. For example, English speakers in Australia produce the
affricate /ts/, but only in final words (e.g., “cats”). The production of /ts/ in initial words
however was difficult for news readers trying to say “tsunami.”

Phonological processing skills

Jarrod had a standard score of 3 on the PIPA subtests on both rhyme awareness and
isolated phoneme awareness ('what is the first sound of...') (Dodd, Crosbie, McIntosh,
Teitzel, & Ozanne, 2000) indicating poor performance on tasks usually mastered by
preschoolers. He scored only 18 on the Sutherland Phonological Awareness Test
(Neilson, 2003), when the average score range for his age is 33-45. Although he is within
normal limits on the letter knowledge task, he does not attempt pseudoword reading or
spelling tasks. Jarrod's performance on the QUIL (Dodd, Holm, Oerlemans, & McCormick,
1996) indicated that, although he did well in segmenting syllables, he was at the bottom of
the normal range on the rhyme recognition task. He was not able to score on the
pseudoword reading and spelling subtests or on the phoneme manipulation tasks, but that
is not unusual for his age, according to the norms, perhaps because these tasks They
require phonological assembly of unknown words. A very recent report indicates number
and letters backwards, but a good understanding of the role of phonics.

Pseudoword repetition tasks are believed to measure phonological working memory, that
is, the ability to hold speech information in a short-term memory loop. Jarrod performed
this task very poorly. This is not surprising. All children with language disorder perform
poorly on pseudoword repetition tasks due to their speech disorder. Children who have a
phonological deficit have special difficulty with the repetition of pseudowords, perhaps
because they are unable to assemble the phonology of unfamiliar words for temporary
storage in phonological working memory.

When Jarrod was asked whether two pseudowords were the same or different, he made
some errors on word pairs that differed by function (e.g., [jeIs]/[jeIt], lost/lot), but had more
difficulty when words differed by a sequence rates/racing). All items involved
in discriminating the last word of /s/ vs /t/ or /ts/ and /st/. This task involves the storage and
comparison of two words in phonological working memory.

Jarrod performs within normal limits on the lexical decision task, where he hears a word
and has to judge whether it was a real word or a pseudo-word (e.g., identifying [fluwi] as a
pseudo-word and “flower” as a real word). Jarrod's score was 23/24. The results suggest
that Jarrod has intact phonological representations of the words used in the assessment.
The task is not about phonological working memory.

Poor performance on phonological processing tasks is generally assumed to reflect

underlying deficits of the speech disorder. When speech errors are characterized by
inconsistency, speech problems may underlie poor performance on phonological
processing tasks. An impairment in phonological assembly means that words cannot be
easily encoded by phonological working memory (phoneme sequences in auditory
discrimination tasks), oral production (pseudoword repetition), or phonological
manipulation. Support for this interpretation is the finding that Jarrod did not have any
difficulty with the lexical decision tasks. This finding is not consistent with their poor
performance on the auditory discrimination task, since the lexical decision task must also
be affected by poor auditory discrimination.

Language

Jarrod has reports of grammatical delay, but when the Clinical Assessment of Language
Fundamentals-4 (CELF-4) (Semel, Wiig, & Secord, 2004) was administered in mid-2005,
all subtest scores (except expressive vocabulary standard score of 6) were within the
normal range of 7-13. Their expressive (112) and receptive (103) language scores were in
the average range. The therapist concluded that the assessment of his language skills was
consistent with his cognitive abilities and that Jarrod did not meet the criteria for a
diagnosis of language disorder, despite having a severe speech impediment. Jarrod's
communication difficulty appears to be specific to speech. Only on the expressive
vocabulary task was he below the normal range, a typical result of children with
inconsistent disorders who perform worse than other children with language disorders on
expressive vocabulary (Dodd et al., 2005).

Clinical and developmental history

Jarrod's recent diagnosis of ADHD (medicated with Ritalin), history of ear infections, and
difficulties with fine motor planning (now resolved by writing) are important factors.

Family context

There is a history of difficulties in phonological processing in both maternal and paternal

families: mother's father has a history of dyslexia, and her father has persistent errors from
a developmental speech disorder for which he received treatment. Jarrod's 10-year-old
sister has difficulty integrating information from the left and right sides of the brain. Jarrod's
parents are separated, but he sees his father often and regularly. He is well supported and
cared for by his mother and family.

Diagnosis

Current models of the speech processing chain (eg Stackhouse & Wells, 1997) can be
used to identify skills that are related to speech disorders. The identified deficits, however,
are always representative of causal, concomitant difficulties, or consequence. Diagnostic
categories of subtypes of speech difficulties have to account for the variety of phonological
symptoms, associated skill profiles, social and academic outcomes, and response to
particular types of intervention. One way of categorizing children with a speech disorder is
in terms of their linguistic symptomatology, that is, the nature of their speech error
patterns. Experimental evidence (Dodd et al., 2005) suggests that children classified into
each of the proposed speech disorder subgroups, described below, have different
performance profiles on tasks designed to assess aspects of the chain. speech
processing. The four proposed subgroups can be diagnosed by the DEAP (Dodd, Crosbie,
Zhu, Holm, & Ozanne, 2002).

Articulation disorder : an impairment in the ability to pronounce specific phonemes, usually /s/ or /I/,
the child always substitutes or distorts the same production of the target sound in words or in
isolation, regardless of whether the sound It occurs spontaneously or imitated.
Phonological delay : All phonological error patterns derived from a child's speech description occur
during normal development, but at least some are typical of children at a younger chronological age
level.

Consistent phonological disorder : constant use of some error patterns that are not part of
development. Most children who use nondevelopmental error patterns also use some delayed
developmental error patterns. They should, however, be classified as a consistent disorder, since
the presence of error patterns that are not part of development signals an impairment of the
acquisition of limitations of the phonological system.

Inconsistent phonological disorder: Children's phonological systems show at least 40% variability
(when asked to name the same 25 pictures on three occasions within one session). Multiple error
forms for the same lexical item should be observed since correct/incorrect performances may reflect
a maturation system.

Evidence indicates that Jarrod has an inconsistent language disorder, due to a deficit in
the phonological system. While deficits in the phonological system are assumed to
underlie inconsistent phonology in aphasia (e.g., Berndt & Mitchum, 1994), inconsistency
as a type of developmental speech disorder has only recently been accepted (Forrest ,
Elbert, and Dinnsen, 2000). Velleman and Vihman (2002) advocated a word 'model' that
contains the phonological specifications for the production of words from a phonological
plan. It is a project that does not involve the speech motor system. Children whose speech
is characterized by inconsistent errors may have difficulties in phoneme selection and
sequencing (i.e., in assembling a phonological template for the production of an
utterance). Alternatively, the plan may not fully specify the plan segments. Jarrod's high
inconsistent speech errors, despite his nearly intact phonetic repertoire, fluent speech, and
poor expressive vocabulary indicate a deficit in the phonological system (Dodd et al.,
2005).

The data suggest that other diagnoses may be explicitly rejected. For example, Jarrod
does not have dyspraxia, despite his inconsistent errors, because his word production is
better in imitation than in spontaneous production. In childhood apraxia of speech,
imitation is poorer than spontaneous production (Bradford-Heit & Dodd, 1998; Crary, 1984;
Ozanne, 2005). Furthermore, Jarrod's oro-motor evaluation showed adequate speech
motor control (cf. VMPAC, Hayden and Plaza, 1999) and that there were no trial and error
tests. He has an adequate phonetic repertoire, uses a range of syllable forms and
produces appropriate prosody in speech.
Jarrod's diagnosis has some complicating factors. His family history of oral and written
communication problems identifies him as genetically and environmentally at risk. His
history of hearing impairment may have contributed to his poor hearing care and
subsequent ADHD diagnosis. Auditory attention training should be part of the therapy
focus. His speech disorder more generally refers to a deficit in planning sequences of fine
movements (Bradford & Dodd, 1996) that should be treated. Therapy must, however,
prioritize the primary feature of inconsistency and unintelligibility of Jarrod's speech-
language production. Maximal inconsistency and lack of intelligibility (Dodd et al., 2005), is
associated with persistent difficulties (Forrest et al., 2000), which obscures their
phonological knowledge and makes it difficult to select intervention targets.

Inconsistency characterized by multiple error types (unpredictable variation among a

relatively large number of phonemes) reflects an unstable phonological system. Grunwell
(1981) and Williams and Stackhouse (2000) argue that inconsistency indicates difficulties
in speech processing. Forrest et al. (2000) argue that inconsistency “will have a negative
impact on phonological acquisition and may contribute to a profile that characterizes
children with persistent phonological disorders” (p. 530).

Children with inconsistent language disorder generally produce the same words or
phonological features not only from one context to another, but also within the same
context (Dodd & Bradford, 2000; Holm & Dodd, 1999; McCormack & Dodd, 1998). They
may pronounce the same word differently each time they say it. Describing and analyzing
the child's inconsistent error patterns in terms of phonological rules is not possible and
deciding the focus of therapy is difficult (Dodd & Bradford, 2000). Forrest et al. (2000, p.
529) agree that ''it is difficult [to treat] these children, because one cannot properly know
the sounds to use in contrast to the error. This may mean that children with variable
substitution will do worse in treatment than other children because the protocols available
for this population are not as effective as other procedures.''

The goal of therapy, then, may not be to contrast phonemes using minimal or maximal
pairs in a way that would be appropriate for a child with phonological delay or consistent
phonological disorder associated with difficulties in phonological processing. It also
wouldn't be worth teaching him speech sounds in isolation, using motor cues, when Jarrod
can already produce most of the phonemes. Rather, core vocabulary therapy was chosen
to focus on Jarrod's teaching of how to assemble phonological words into single words and
then into speech. A description of the approved vocabulary approach to therapy (Dodd &
Iacono, 1989; Crosbie, Holm, & Dodd, 2005) is presented in Section 4 under long-term
and short-term goals of therapy.

Which service delivery model should be chosen?

There are a range of interrelated factors that must be taken into account when choosing
the appropriate service. Service delivery decisions concern the agent(s) of the therapy,
whether the therapy is group or individual, the scheduling of the intervention (duration,
frequency of sessions), the site of the intervention (home, school, clinic). and the duration
of the intervention. Jarrod's diagnosis of inconsistent speech disorder led the planning of
service delivery. Research effectiveness studies (Crosbie, Holm, & Dodd, 2005; Dodd &
Bradford, 2000) have established best practices for core vocabulary intervention.

Jarrod must receive two interventions each week of 30 minutes each. One-on-one therapy
is needed, since basic vocabulary is adapted individually, making group intervention
impossible. A speech pathologist would be the primary agent of therapy, and caregivers
play an important role. They observe the sessions and ensure that the key words are
practiced daily at home. Jarrod's teacher would help choose therapy goals and will be
asked to monitor his intervention to ensure that he learns better word production in school.

While the intervention can occur at home, at school or in a clinic, there are advantages to
using a cross-environmental approach (eg generalization from the clinic to the
home/school). Since Jarrod's grandparents are important caregivers, therapy can occur
once a week at his home and once at school. Research indicates that the therapy
approach should be implemented for 8 weeks. Most children establish consistency of
productions at this time. Another intervention may be necessary if the 3-monthly review
shows a loss of consistency gains or consistent speech error patterns that affect
intelligibility.

What are the objectives of the intervention?

Latest (forecast)

The ultimate goal is for Jarrod's speech to be error-free and/or his written communication
skills since his cognitive and linguistic abilities are within the average range, and he has no
current sensory or physical disabilities.

Long term (for the intervention episode)

The long-term goal would be to establish the best consistency of production of a minimum
of 50 words, with a primary focus of core vocabulary intervention, with generalization of
greater consistency to untreated words. The basic vocabulary differs from the approaches
often used for childhood apraxia of speech. For example, Strand and Debertine's (2000)
comprehensive stimulation intervention focuses on motor learning, using direct imitation to
guide increasingly phonetically complex expressions. In contrast, basic vocabulary is
aimed at online word planning, avoids direct imitation, and includes the multisyllabic words
from the first session. The goal of therapy is consistency, rather than exact production of
the word. The reason for addressing inconsistency is the negative impact it has on
intelligibility. Jarrod's speech is often incomprehensible, even to members of his family.

Another reason for achieving consistency of production is that even a child's speech errors
are consistent, selecting intervention targets is very difficult. Jarrod uses a series of sound
substitutions that differ in the form of production, in the place of production or in the

expression. For example, he marked /l/ with a or mutes the sound.

It is impossible to select the appropriate pair of errors to contrast given the variety of
substitutions. Nor is it effective to adopt an articulatory approach that addresses an
isolated sound when it is already part of Jarrod's speech sound repertoire.

Jarrod's lack of progress in previous intervention for his speech disorder reflects research
that children with inconsistent language disorders are resistant to phonological contrast
(Crosbie, Holm, & Dodd, 2005; Forrest, Dinnsen, & Elbert, 1997 ) or traditional joint
therapies. A retrospective post-hoc analysis of 14 children with language disorders
(Forrest et al., 2000) compared to children who made consistent sound substitutions for
sounds that do not exist in their inventories (e.g., /k/ which is always produced as [t]),
those that had inconsistent sound substitutions across word positions (e.g., /v/ replaced by
[b] in word-initials, but by [f] in word-finals), and those that uses different substitution
sounds (word initial /s/ being replaced by /v, f, d, b/) in all positions of the word. The three
groups were matched by severity of phonological impairment, and all phonological contrast
treatments received were directed at a single error on a single word. Children with
consistent sound substitutions learned the sound and generalized to other positions in the
word. Children with inconsistent sound substitutions in all word positions learned the
sound, but only in the treated position. Children with variable sound substitutions in all
word positions did not learn the sound in the treated and untreated positions. These
results demonstrate the need to focus not on knowledge of phonological contrasts, but
rather on the ability to assemble phonology.

Short term (session plans)

Select the destination . Before beginning therapy a list of 50 target words (minimum)
should be selected in collaboration with Jarrod's family and teacher. The words must be
functionally powerful and often include names of people, names of pets, places (e.g. street
where your house is, school, toilet), function words (e.g. please, sorry, thank you). ,
favorite foods, toys and games. Words are not selected according to word form or
segments. They are chosen because of the frequency with which the child uses these
words in functional communication. Increasing the child's intelligibility of selected functional
power words motivates the use of consistent productions. It is important to highlight for
caregivers and others (e.g., teacher) that the main goal of the intervention is to ensure that
children say a word the same way each time they try to say it, not to achieve error-free
production.

Establishment of the best production . Each week, the first 30-minute session focuses on
Jarrod randomly selecting up to 10 words from a bag containing all the objectives. The
therapist must then teach Jarrod the selected words sound by sound, using cues such as
syllable segmentation, imitation, and cued articulation (Passy, 1990). For example, to
teach Jarrod to say his own name, the therapist might explain that Jarrod has two

syllables, . The first syllable has two sounds, and the second syllable has
three sounds. The child attempts the first syllable, receives feedback, and makes another
attempt after receiving modeling and feedback on each attempt. When the child's best
production of the first syllable has been established, the second is directed, and then the
two syllables are combined. A very effective technique, for some children, is to link sounds
to letters and this should be used with Jarrod as he is 7 years old and exposed to formal
literacy at school. If it is not possible to obtain a correct production, then Jarrod's best
production, which would include development errors, would be accepted. (eg by
Jarrod, by camera).

To exercise . The second session of the week includes keyword practice. Games are used
to provoke a high number of repetitions. Any game that Jarrod is highly motivated to
participate in can be used to get productions. Elbert, Powell, and Swartzlander (1991)
suggest that a child should produce approximately 100 responses in 30 minutes. Jarrod's
caregivers should participate in these sessions, as they will need to elicit, provide
feedback, and monitor spontaneous productions of the target words daily at home. It
should be noted that only the words selected for each week should be addressed.

Treatment in error . Leahy (2004) wrote that children do not always understand why they
are attending therapy and what they are required to do in sessions. Therefore, it is
important to be explicit about the purpose of therapy, the nature of the error made, and
how it can be corrected. If Jarrod produces a target that deviates from the best production
the therapist/caregiver/teacher can imitate the production and explicitly explain the word
that differs and how it differs. For example, if Jarrod's target word was ''sun'' and he
produced [gan] the therapist says “[gan], that is different from how we say it. That has the
G] sound at the beginning but we have to make an [s], [an].” Jarrod's therapist should
avoid simply asking him to imitate the target word since imitation provides a phonological
plan that inconsistent children can use without needing to assemble/generate their own
plan for the word. Instead, therapists must provide information about the plan.

Consistent production monitoring . Toward the end of the second session of the week,
Jarrod was asked to produce three times the set of specific words that have been the
focus of therapy for the past week. Any words that are consistently produced using the
best production are removed from the list of words to be learned. It can be placed on a
graph that shows what you have achieved. The words you produce inconsistently remain
on the list (it goes in the bag of words that have not yet been learned). Although there are
50 target words that form Jarred's basic vocabulary during 8 weeks of intervention, this
monitoring allows us to say that the words that have not been mastered will be taken up in
another week.

How is generalization going to be helped?

Learning to elaborate a goal in a clinical situation does not necessarily mean that it will be
produced correctly in spontaneous speech outside the clinic. Weiss, Gordon, and
Lillywhite (1987) argue that generalization needs to be taught explicitly. Core vocabulary
intervention aims to stabilize the phonological system, resulting in consistent productions.
The therapy would not be beneficial if the effect of the therapy was limited to only the
treated elements. To monitor generalization, Jarrod's therapist must use a biweekly set of
untreated items (ten words), eliciting three productions of the untreated items in one
therapy session. The untreated elements will allow for change in the system to be
monitored (i.e., identifying the moment when Jarrod's speech production becomes
consistent). Generalization should improve the delivery of the intervention at home and at
school and the participation of their caregivers and teachers in feedback and daily
practice.

What performance criteria will be established?

There is evidence that different speech and language therapy services achieve different
points of their remediation in patients (Enderby & John, 1999). Efficacy studies suggest
that core vocabulary intervention should increase Jarrod's consistency and production
accuracy; his errors will be characterized by developmental, not atypical, error patterns.
Some children, however, require more than one intervention model to achieve age-
appropriate expression. For example, Dodd and Bradford (2000) report a case study of a
child with inconsistent speech production. Once consistency was established they
benefited from phonological contrast therapy that targeted their remaining developmental
error patterns. Given Jarrod's severity, resistance to prior therapy, and the complicating
factors of ADHD, motor planning, and family history, he may require more than one
episode of intervention.

How should effectiveness be evaluated?

The intervention must be monitored to establish effectiveness. To evaluate the

effectiveness of the core vocabulary approach of Jarrod's intervention, the therapist could:

 Establish a pre-treatment baseline (analysis of three more speech samples over 2

months) for consistency, percentage of correct consonants and vowels, and
colloquial speech;
 Apply therapy longer than 8 weeks (as described);
 Reassess Jarrod using the same measures used for baseline before the
intervention.

This design would meet Bain and Dollaghan's (1991) criteria for clinically significant
change (intervention effectiveness). Any change may demonstrate the result of the
intervention rather than maturation or other factors not controlled by a pre-intervention
where a baseline would be established. The changes will be shown as important rather
than trivial, because the consistency and accuracy of the production of the words not
targeted in therapy would be measured and the change would be real, rather than random
due to the short duration of the therapy. intervention.

Discussion

Developmental speech disorder has taken into account theories derived from psychology,
psycholinguistics, linguistics and medicine. Consequently, researchers have developed
specific assessment protocols for the differential diagnosis of language disorders (e.g.,
Hayden & Plaza, 1999; Stackhouse & Wells, 1997). They argue that their assessment data
allow the identification of deficits underlying language problems and the planning of cost-
effective interventions. This article presents a case management plan for a 7-year-old child
with highly unintelligible speech. Evaluation data analyzes were used to address seven
case management questions regarding need for intervention, service delivery, differential
diagnosis, goals of intervention, generalizability of therapeutic benefits, approval criteria of
management and evaluation of effectiveness.

Intervention is deemed necessary for Jarrod because of his irregular pattern of

communicative performance; concern by Jarrod, his caregivers, and teachers of his
speech disorder caused the likelihood that it was a contributing factor to academic and
social difficulties. He was diagnosed with inconsistent speech disorder. Pronounced 88%
of words differently when asked the name of each word in the 25-word Diagnostic
Assessment of Inconsistency of Articulation and Phonology (Dodd et al., 2002) three
times, each trial separated by another activity. . Since the arbitrary criterion for diagnosing
inconsistency is 40%, Jarrod's inconsistency score was very high. The diagnosis was
supported by other findings such as intact phoneme repertoire, poor expressive
vocabulary, and poor performance in judging phoneme sequences but good lexical
decision ability.

Jarrod's performance on Hayden and Square's (1999) oro-motor tasks was interpreted as
showing poor “neuromuscular integration.” It is difficult to rule out a contribution to Jarrod's
motor speech disorder; group studies suggest that children with inconsistent speech
disorders have general sequences of difficulty in fine motor movement planning (Bradford
& Dodd, 1996). However, given that his SLP did not observe any motor difficulties, Jarrod's
ability to produce most phonemes and syllable forms, and his better performance in
imitation, compared to spontaneous production, it seems unlikely that his disorder of
speech can be attributed solely to an oromotor deficit.
Sometimes it is difficult to distinguish compromised abilities that are causal from those that
are consequential difficulties. Since Jarrod's difficulty appears to be limited to speech
production, a deficit in the phonological system due to its inconsistency and failure to
organize motor signals in speech. Their deficit in phonological working memory can also
be interpreted as a consequence of their inability to assemble the phonology of
phonological processing. The core vocabulary intervention was chosen as the most
appropriate therapy technique as it directly targets the planning of words that are
functionally powerful in the patient's social and academic context. While the technique
focuses on the phonological system, it attempts to focus and maintain auditory attention
and may include motor facilitations to help resolve difficulties in planning oro-motor vocal
sequences.

The speech processing chain is complex. It not only involves the processing of input
(sensation and perception) and output (motor), but also the mental processes that allow
the acquisition of phonological knowledge through attention, memory and analysis of the
phonological aspects of language. Current research focuses on identifying deficiencies
that lead to speech difficulties. Until now, little is known about the interaction between
genetic, environmental, and damaged neurological strata underlying the deficits that cause
speech difficulties.

In conclusion, it has been argued that the choice of therapy technique should be linked to
a diagnosis. Identifying the deficit(s) underlying a child's speech disorder allows for case
management decisions that result in cost-effective intervention using best practices that
have been identified by research. All intervention approaches have their merits. Clinical
skills are reflected in choosing the appropriate intervention for a child's specific deficits.

Thanks

Our thanks to Jarrod and his family, to the teacher and speech therapist, for their patience,
time and interest in this project.