Content uploaded by Kakia Petinou
Author content
All content in this area was uploaded by Kakia Petinou on Oct 24, 2020
Content may be subject to copyright.
Promoting speech intelligibility in ASD through the implementation of phonologically similar
stimuli
Short title to be used as running head: Speech intelligibility in ASD
Kakia Petinou
School of Health Sciences
Department of Rehabilitation Sciences
Vragadinos Street 15
Cyprus University of Technology, P.O. BOX 50329, Limassol 3603
Cyprus
Keywords: phonological delay, Autism Spectrum Disorders, Speech Intelligibility, phonological intervention
Address correspondence: Kakia Petinou Cyprus University of Technology, Limassol, Cyprus,
kakia.petinou@cut.ac.cy
NUMBER OF APPENDIX 1
NUMBER OF TABLES 1
NUMBER OF FIGURES 1
WORD COUNT 4,903
1
Abstract
Objectives: The study focused on promoting expressive phonological skills in one Greek-
speaking child with autism spectrum disorder (ASD) and comorbid speech sound disorders.
Based on the phonological neighborhood density framework, it was hypothesized that the
experimental manipulation through clinical implementation of phonologically overlapping
stimuli would yield positive expressive phonology gains relevant to ASD.
Participant & Methods: A multiple-baseline single-subject design was implemented. Three
baseline sessions measured expressive phonology variables. Sixteen bi-weekly 30-minute
intervention sessions were carried out for a period of two months. Dependent variables included
phonetic inventory size, proportion of consonants correct, occurrences of phonological
processes, and percentage of whole word matches elicited via specific word-probe stimuli. The
Intelligibility in Context Scale was completed by the child’s teacher prior to the initiation of
intervention and at follow-up session. Experimental stimuli were grouped together in
phonologically-dense cohorts.
Results: Comparison between pre-test and post-test measures revealed expressive phonology
gains across all measured variables. Follow-up session results showed generalization of
expressive phonology gains on untreated targets.
Conclusions: Significant expressive phonology gains were achieved through the implementation
of phonologically similar word stimuli within a systematic intervention protocol with the
implementation of specific word-level variables. The findings supported this treatment approach
for a child with ASD and SSD, while providing evidence for the phonological density advantage
from a cross-linguistic perspective.
Introduction
2
Children diagnosed with Autism Spectrum Disorder (ASD) constitute a considerably
heterogeneous group with linguistic development forming no exception. Heterogeneity is
observed through developmental trajectories which map various aspects of communication
including expressive and receptive language skills, social pragmatic profiles, and behavioral/
emotional expressions output across languages and across cultures [1,2]. Despite this diversity
however, research findings convincingly revealed that communication deficits and late onset
of language emergence form the core characteristics of ASD [3]. Children with autism fail
to achieve developmental milestones at age-expected levels, can be slow in learning to talk
or many times, do not learn to talk at all [4]. Although some children with ASD reportedly,
begin to produce meaningful speech during the first year of life, shortly after, this skill
declines and eventually seizes around the age of 18 months [5,6]. Plausible explanations for
this developmental arrest align with neuro-developmental underpinnings which affect
brain circuitry on areas crucial to social-communicative and expressive speech
development [7,8,9].
3
Over the past fifteen years, phonological and articulatory skills in ASD have been studied
extensively, due to the remarkable prevalence of speech sound disorders (SSD) documented
in ASD speech profiles [10,11,12]. Specifically, reports from English corpora indicated that the
prevalence of ASD comorbidity with SSD ranged between 15%-20% [13]. Within ASD
populations however, findings related to the existence, characteristics and severity of SSD
remain equivocal. Although earlier reports on phonetic, phonological, and prosodic
development reported age-appropriate speech profiles [14,15], recent reports on younger
children with autism revealed atypical speech patterns. These were characterized by
immature syllable structure, restricted phonetic inventories, speech sound distortions,
inconsistent word productions, atypical occurrence of phonological process use, persistent
use of jargon speech, late emergence of meaningful expressive vocabulary, and restricted
use of phonological contrasts [10,11,16]. Similarly, young autistic children presented with
oral motor difficulties and failed diadochokinetic tasks, had difficulties producing
successive syllable-repetition tokens and lacked phonological stability across multiple
productions of the same words [17,18,19]. Despite inconclusive evidence as to the loci of
phonological difficulties in ASD, various characterizations regarding the nature of such
speech challenges advocated for the presence of a multi-deficit motor speech disorder
consistent with child apraxia of speech (CAS). Presumably, oral motor deficits hindered
typical phonological planning and phonetic/articulatory programming [13,20]. Similarly,
atypical phonetic patterns in older children with ASD have been characterized as non-
developmental sound distortions such as phoneme specific nasal emission, lateralization
of /s/ and residual erred realization of /r/ [13,16]. It was further concluded that non-
developmental distortions of speech sounds appeared frequently in the speech of children
4
with ASD and persisted over time. These errors were variable, did not resolve easily and
affected speech intelligibility. Regarding phonological patterns, findings suggested the
occurrence of phonological processes including velar fronting and atypical initial consonant
deletion [21]. Such findings bear clinical and research insights and underscore the
importance of timely early intervention. Moreover, research findings are of paramount
importance as they inform evidence-based practice (EBP) and contribute to data-driven
analyses, and theory motivated intervention.
Yet, despite the plethora of investigations regarding description and documentation of
atypical speech profiles in ASD, information on treatment to ameliorate such challenges remain
sparse [10,17]. The paucity of EBP investigations on inducing phonological gains in ASD,
highlight the need for advancing systematic, and theory-motivated intervention protocols.
Moreover, data from a cross-linguistic perspective regarding ASD/SSD interaction through EBP
intervention efficacy remain understudied [22,23]. Notably, cross-linguistic investigations related
to atypical speech and language development, underline the role of typological properties of a
given language on speech production phenotypes. Linguistic diversity determines what is spared
and what is lost in pathological communication profiles [24,25,26]. Furthermore, idiosyncratic
properties of the ambient language (e.g., prosodic and phonotactic variables) determine the
selection of intervention stimuli, guide treatment goals, and affect therapy outcomes [27,28,29].
To this end, early intervention with a focus on promoting speech intelligibility is of paramount
importance because a robust phonological system constitutes the “backbone” for the emergence
of more complex linguistic-subsystems [30]. This is because, during the early stages of language
development, segments and prosodic templates form the "building blocks" for later lexical,
grammatical, and morpho-syntactic development [31]. Furthermore, enhanced phonological
5
representations map on to phonological awareness skills which in turn, form the cornerstone
ability for literacy development [32].
Recently, results from investigations in the area of developmental language disorders
(DLD) and ASD including the COST Action IS1 406 consortium titled “Enhancing Children’s
Oral Language Skills Across Europe and Beyond” and the Child Language Committee of the
International Association of Logopedics and Phoniatrics (CLC IALP) respectively, advocated for
the capitalization of research findings in supporting treatment efficacy in variable linguistic and
cultural contexts [27,1]. However, findings revealed that the majority of SLPs who serve
children with communication disorders are either not familiar with EPB practices or have limited
exposure to theory-motivated interventions. As a result, research information and resources
related to EBP and theory-driven therapy might not reach the mainstream intervention contexts
[23]. To this end, treatment efficacy, therapy decisions, therapy goals, and ultimately treatment
outcomes need to be validated through processes that can be experimentally manipulated [33].
In line with the aforementioned, an emerging body of research vis-à-vis lexical
organization models, brings forward the role of word-form variables on inducing
expressive phonological gains in children with SSD [34,35]. It has been proposed that young
children capitalize on the statistical properties of the input as a way to optimize the process
of language development [36,37]. One such property is neighborhood density (number of
words differing in phonological patterns by 1-phone which is deleted, added or altered)
[38]. While DENSE neighborhoods (DN) consist of many words with overlapping
phonological structure (i.e., rat overlaps with mat, bat, cat, sat, rat, fat, tar, pat, etc.),
SPARSE neighborhoods (SN) on the other hand, contain fewer words with phonological
overlaps (i.e., room, rule, root). In this context, phonological density constitutes an
6
important aspect of word-form organization because of its differential impact on
phonological learning [34-38]. Empirically, typically developing toddlers are shown to
imitate more accurately segments from DN as compared to sounds from SN. This is because
phonologically similar words (e.g., words residing in dense neighborhoods) “force” children
to build finer-grained phonological representations which in turn, enhance the production
of diverse phonological contrasts necessary for marking lexical meaning [30,31].
Furthermore, recent investigations have focused on the impact of the DN advantage
on promoting expressive phonological gains in children with SSD [34-37]. Findings from
English corpora, have convincingly demonstrated that the systematic implementation of
stimuli from a DN yields significant phonological gains and contributes to the
generalization and speech gains on untreated segments [35]. Overall, phonologically dense
neighborhood effect constitutes an optimum trigger for inducing sustainable speech
intelligibility gains [34]. Potentially, children with protracted phonological profiles, in this
case children with ASD and comorbid SSD deficits, present a unique opportunity for
advancing testable, theory-motivated hypotheses related to the selection and
implementation of target stimuli during intervention [36, 37, 38]. In view of the above, a
number of themes emerge: (a) the need for systematic examination of theory driven
intervention in children with ASD who present with protracted phonological development
(PPD), and (b) treatment efficacy of phonological intervention through the implementation
of phonologically dense words in cross-linguistic contexts.
The present study aspired to add to this line of research through the implementation
of phonologically similar targets on inducing expressive phonological gains in a bidialectal
Cypriot Greek-speaking (CG) child with ASD and concomitant SSD. Given the impact of
7
the DN advantage on expressive phonological gains in children with SSD, it was
hypothesized that such effects would also induce speech gains in children with ASD and
accompanied SSD. The implementation of word stimuli cohorts from phonologically dense
neighborhoods would allow the examination of treatment efficacy through the
implementation of specific word targets for which children might show production
constraints (e.g., similarly sounding words). Potential positive gains might set the
foundation for enhancing the child’s transition from unintelligible to intelligible speech.
The CG dialect
The linguistic variety investigated in this study is CG, a variety of Standard Modern Greek
(SMG) [39,40]. Although CG is the everyday variety used by most speakers on the island of
Cyprus, SMG is used in certain official settings such as school classes and the media while CG
dominates in informal everyday communication. The Greek-speaking populace of Cyprus is
diglossic in the sense that CG is the vernacular form (low variety) used in everyday
communication whereas SMG (as the high variety) is used in educational settings, government
bodies, and the media. Recent reports propose the emergence of a dialectical continuum of CG
with an emerging “Koine” or “Urban” form of CG [41]. Its consonantal phonetic inventory
contains 31 singletons, most of which have geminate counterparts. There are bilabial, labio-
dental, dental/alveolar, palatal and velar phonemes. Regarding manner of articulation there are
stops, fricatives, affricates, nasals, liquids, and glides. The liquid segment /r/ is trilled. In CG
voiced stops are pre-nasalized with voiceless cognates being either unreleased or aspirated.
The /n/ and /s/ may also occur in word final position. CG contains post-alveolar fricative
and aricates as well as geminates (double segments) diering from
singletons mainly in duration. [42]. Currently, developmental databases regarding
8
syllable structure use in CG child speech remain sparse [43]. Thus, for the purposes of the
current investigations the author analyzed the syllable structure of 600 target word items
found in the Cypriot-Greek Lexical Acquisition List (CYLEX), a parentally reported
lexical development tool [41,43]. Percentages of syllable structure occurrences from the 600
uninflected words were as follows: CVCV (28%), CVCVCV (22%), CVCVCVCV (10%),
CCVCV (8 %), CVCCV (6%) and VCV (6%). The rest of the word targets (20%) consisted
of variable monosyllabic trisyllabic and multisyllabic structures. Based on these
observations it was concluded that the predominant syllable structure one might encounter
in CG was the disyllabic CVCV and trisyllabic CVCVCV. Stress pattern analysis was also
performed and results revealed 10 monosyllables, 285 disyllables (183 trochees (65%) and
102 (35%) iambs), 208 trisyllables (28 targets stressed on the ultimate, 120 targets on the
penultimate, 60 on the antepenultimate), and 97 multisyllables.
Methodology
Participant and Intake
The participant was a four-year-old CG-speaking boy with a diagnosis of ASD based on
the criteria from the Diagnostic Statistical Manual, 5th ed. (DSM-V) [44] and the Autism
Diagnostic Observation Scale (ADOS) [45]. He was assessed by a certified child Psychiatrist in
Nicosia, Cyprus. An initial diagnosis of speech and language delay was conducted by his
kindergarten speech therapist using a series of available assessment batteries adapted in CG
which are used for clinical and research purposes [46,47]. The child was referred to the Cyprus
University of Technology (CUT) Rehabilitation Clinic for further phonological evaluation due to
parental concerns regarding the lack of expressive language progress exhibited by the child.
Phonetic and phonological skills were examined with the use of an adapted version of the
9
Phonological Assessment Test of the Panhellenic Association of Logopedics (PAL) [48]. In
addition, a 15-minute free language sample assessed skills in connected speech. Analyses
corroborated with previous reports regarding the child’s phonological profile and confirmed
significant phonological delays. The child continued to present with a restricted phonetic
inventory (/p/, /t/, /n/, /k/, /m/), used immature syllable structure (vowel-consonant-vowel VCV),
lacked the use of trisyllabic and multisyllabic words, and expressed significant occurrence of
phonological processes (unstressed syllable deletion, word initial onset deletion, cluster
reduction and regressive assimilation). These patterns rendered his speech unintelligible. The
most prevalent syllable structure used was consonant-vowel (CV), vowel consonant (VC),
vowel-consonant-vowel (VCV) and consonant-vowel-consonant-vowel (CVCV) in its
reduplicated form (e.g., [‘mama], [‘tata]). The child was invited to participate in an ongoing
phonological intervention study on CG-speaking children with SSD which at the time was being
conducted at CUT “TheraLab” research laboratory. Parents accepted to enroll their child after
signing a written consent. The participant, a monolingual CG-speaker, came from a middle
socioeconomic status family [49]. According to parental report he had no significant history of
otitis media with effusion (OME). Furthermore, during the intake meeting he passed a hearing
screening test at 500, 1000, 2000, and 4000 Hz presented at 25 dB HL [50]. At the time, the
child was enrolled in a speech and language community-based language enrichment group
therapy program with a focus on social pragmatic aspects across the communication
spectrum. However, it should be noted that during the course of the intervention period, he
had stopped attending the language enrichment program due to summer holiday recess.
Research Design
10
The current examination employed a single-subject multiple baseline research
design. [51,52]. A three-point baseline was followed by 16 half-hour individual treatment
sessions administered twice a week over a period of two months (approximately 13 hours of
direct one-to-one sessions). A post-test session was conducted one week after the completion
of the last intervention session succeeded by a follow-up session administered two months
later. The author conducted all sessions.
Baseline
The purpose of the baseline was to establish stability on the repeated dependent
measures in the absence of treatment. It was completed across three sessions over a period
of two weeks. Phonetic and phonological skills were examined with the use of repeated
administration of a fifty-word probe corresponding to colored digitized pictures inserted into
PowerPoint file for presentation. These consisted of thirty bi-syllabic, ten tri-syllabic, and ten
multisyllabic targets (see Appendix A). The list attempted to sample the majority of singleton
consonants of CG across all possible word positions and stress patterns, excluding affricates and
clusters. The particular probe list has been used for similar research investigations in CG-
speaking children with SSD [53]. Each session was audio-recorded using a Marantz PMD-222
digital recorder. An Audio-Technico flat unidirectional microphone placed on the experimental
table directly in front of the child. Recorded speech samples were phonetically transcribed using
the International Phonetic Alphabet (IPA) [54]. Phonetic transcriptions were performed on all
utterances produced by the child and were coded on the bases of independent and relational
analyses [55]. These analyses yielded six kinds of information: (a) phonetic inventory size (PIS),
(b) syllable structure use (SS), (c) proportion of consonants correct (PCC), (d) occurrence of
phonological processes (PPs), (e) percentage correct of Whole Word Matches metric (WWMs)
11
scored as 0 for no production match and 1 as total match [56], and the Context Scale adapted in
Greek (ICS) completed by the child’s teacher [57]. The ICS is a 5-point scale allowing the
assessment of a child’s overall speech intelligibility in connected speech (details are
provided in the section below). Stimulability of all sampled segments was also performed and
information was used for the construction of experimental-stimuli cohorts described below.
There were no significant changes in the child’s speech profiles between first and last baseline.
Experimental Stimuli
The experimental stimuli consisted of words from quasi-phonologically dense
neighborhoods. In this context words differed from each other by 1 or 2 consonantal segments
and 1 vowel. Stimuli were grouped together in blocks of disyllabic and trisyllabic cohorts
containing four members each. The construction of each of target word was based on the child’s
articulatory knowledge in the sense that each word member included at least one stimulable
phoneme (usually in word initial position). This paradigm was based on research findings
suggesting that when the selection of treatment stimuli considers the child’s existing articulatory
knowledge (e.g., segmental stimulability), increased speech intelligibility and significant output
gains on more complex segments emerge [58]. The word targets were depicted as digitized
colored pictures inserted into PowerPoint files. All were selected from a variety of online
resources and from the author’s private art collection of clinical tools. A total of sixteen clusters
of word-tests were used through the intervention phase (8 disyllabic and 8 trisyllabic). The
disyllabic clusters were randomly selected and were taught during treatment sessions 1-8 (first
month of intervention), followed by comparable procedures with the implementation of the
three-syllable cohorts implemented across sessions 9-16 (second month of intervention). Each
experimental session allowed the implementation of 2 cohorts. A schematic representation of
12
the intervention timeline is presented in Figure 1. The complete list of experimental word
stimuli cohorts is reported in the Appendix.
FIGURE 1
Intervention Procedures
During the experimental session, the child interacted with the examiner and/or the
caregiver while playing with various sets of toys. At the beginning of each session, the
participant was familiarized through auditory bombardment with one randomly selected target
cohort and its corresponding members depicted on pictures. After the examiner presented each
word target five times, the child was prompted to either produce spontaneously or imitatively the
corresponding word given at least five opportunities. In case of no response, the participant was
prompted to either utter the target words via phonetic/articulatory prompting (e.g., when given
the word-initial onset) or produce the target through delayed imitation. All phonetically
transcribed productions were coded and analyzed as correct or incorrect depending on the child’s
production.
Data reduction
Across all sessions, including baseline, post-test and follow-up, approximately 700
tokens underwent analyses. Only productions, either spontaneous or imitated, corresponding to
assessment and treatment targets were analyzed. The repeated dependent measures during the
baseline (pre-test), intervention phases, reassessment session (post-test) and follow-up were
analyzed as per each recorded session. Phonetic inventory size (PIS) included the number of
singleton consonants presented in the child’s speech output as per productions of the probe
targets. The same database was used for calculating syllable structure use (SS). The proportion of
consonants correct (PCC) was calculated by counting the number of segments correctly produced
13
over the total number of segments. For calculating phonological process occurrence (PPs), the
study employed measures used in assessing and analyzing phonological output in young children
[48]. For example, all instances of process occurrence (collapsed monosyllabic, bi-syllabic, and
tri-syllabic productions) were counted over the total number of targets produced by the child.
The percentage of correct on whole word matches (WWMs) corresponded only to words
produced accurately, including imitative realizations, with metric scoring of 0 for no
production match and 1 as total match [56]. The Intelligibility in Context Scale (ICS) was
scored numerically from a scale of 1-5 (1 = always unintelligible, 2 = usually unintelligible,
3 = sometimes unintelligible, 4 = rarely unintelligible, 5 = never unintelligible) [57].
Transcription reliability
For purposes of test-retest reliability a second assessor trained in phonetic transcription
analyzed approximately 10% of randomly selected recorded samples coding as per the initial
and final intervention sessions. Reliability on the relevant categories was based on the number
of agreements divided by the sum of agreements and disagreements after the two transcribers had
jointly listened to the targets and had compared their transcriptions for PIS (80%), PCC (90%),
PPs (90% and 80% for unstressed syllable deletion and word initial onset deletion respectively),
WWMs (100%), and syllable structure use (100%).
Results
The dependent variable was expressive phonological gains exhibited by the child
according to his performance on phonetic and phonological measurements listed in the
previous section. Performance on the parameters examined was compared between pre-
test (past baseline session) relative to post-test data. Follow-up measures were compared to
post-test performance with the exception of the Intelligibility Context Scale which was
14
compared to data obtained at the pre-test. The results are presented in the form of descriptive
statistics. In the current study, numerical score measures were variable and included percentage
scores and raw numbers. Thus, a comparative magnitude of the treatment effect could not be
performed directly in an isometric and isomorphic fashion [59]. For this reason, the paper
discusses the direction of speech output gains on the bases of plotted trajectories relative to pre-
test and post-intervention and follow-up sessions. A visual inspection of data plotting suggested
an overall upward direction on all dependent variables between pre-test and post-test. It was
noted that the child gained ten new consonants segments (from 5 pre-test to 15 post-test), with
percentage of consonants correct shifting from 20% to 30%. There was an improvement in
percentage correct on WWMs which increased from 12% (6/50), to 70% (30/50) between pre-
test versus post-test, a pattern that was also evident during the follow-up with the child
maintaining 80% accuracy (16/20) as multisyllabic word accuracy increased from 20% to
60%. He also exhibited less use of phonological processes from 100% to 30% and from 100%
to 65% for unstressed syllable deletion. Consequently, there was a concomitant increase in the
use of more mature syllable structure. Specifically, the child exhibited less use of VCV and
more use of CVCV as well as CVCVCV patterns. Taken together, all measures contributed to
speech intelligibility gains as measured by the ICS scale scores showing an increase from 1 to 4.
Pre-test, post-test and follow-up data are reported in Table 1.
INSERT TABLE 1 HERE
Discussion
The study set out to investigate speech intelligibility gains in one child diagnosed
with ASD and concurrent SSD with a focus on promoting speech intelligibility through the
implementation of phonologically similar words (e.g., phonologically dense targets). The use of
15
phonologically similar targets was implemented systematically over a course of intervention
sessions. This is the first study to address intervention in ASD/SSD in an understudied dialectal
variation of MG.
Comparisons between the pre-test and the post-test sessions revealed remarkable gains
across measured variables. Specifically, the findings revealed improvements in phonetic and
phonological domains. Speech output measures revealed segmental gains along with a reduction
of structural phonological processes specific to word initial onset and unstressed syllable
deletions. The latter two variables contributed to the emergence of more mature syllable structure
instances including CVCV and CVCVCV. Phonetic inventory size increased from 5 to 15
segments which were established across word initial, word medial and word final positions.
Speech gains were also marked by the emergence of fricatives and glides which were not present
at pre-test measures. Segments were also produced across more diverse places of articulation
including velar and labiodental targets. Notably, gains were emerging in the use of morpho-
phonological {-s} which in the Greek language it marks noun plurality and second person
singular verb form. It may be speculated that phonological gains might have contributed to the
emergence of a wider-system improvement in the sense that a phonological change, contributes
to other linguistic sub-system such as morphophonology and inflectional morphology. Further to
these findings, it was noted that when he started producing unstressed syllables (e.g., [‘nana] ->
[banana] and [‘lami] -> [salami]), towards the final stages of the intervention phase, a
concomitant spontaneous production of definite articles (unstressed elements in CG) and object
clitics was noted in conversational speech.
An increase in the production of percentage of consonants correct and WWMs was also
noted. Furthermore, improved intelligibility was noted according to the scores reported on the
16
Intelligibility in Context Scale completed by the child's school teacher who was unfamiliar with
the intervention project. Accordingly, the child’s phonological output progressed from “always
unintelligible” to “rarely unintelligible”. Phonological progress was maintained and generalized
to non-treated and newly presented word targets administered at the follow-up sessions. At
follow-up session, the results of the percentage of WWMs revealed a noticeable production of
multisyllabic words. Note, that the study employed triplets of targets such as [l e‘m o]
“neck”, [l e ‘m o n i] “lemon” and [l e m o ‘n a ð a] “lemonade” all of which were initially
realized as [e ‘m o] or [‘m e m o]. By the end of the treatment phase, all three targets were
produced with 100% accuracy. Other triplets were also successfully produced and included
targets such as [ka‘la] “good”, [ka‘lami] “stick”, [kala‘maci] “straw” which initially were
realized as [a‘ja] (for more examples refer to the Appendix).
The investigation attested to the usefulness of phonologically dense word-stimuli relevant
to theory-motivated frameworks on par with lexical organization models. Research findings have
convincingly advocated for the advantage of a phonologically dense cluster of words and their
positive impact on inducing phonological gains. To the first approximation, as a word-form
variable, word density appears to be a useful “trigger” for ameliorating speech sound disorders,
mostly in English-speaking youngsters. Findings from the current investigation added to this line
of research and supported the hypothesis advanced, that is, the prediction of speech sound gains
in SSD through the implementation of carefully designed and experimentally manipulated targets
and their impact on speech intelligibility. The present findings corroborated comparable data
from English-speaking children and provided support for a cross-language applicability relevant
to ASD speech profiles. In addition, it capitalized on the child's existing, albeit restricted,
phonological strengths (stimulability), with findings supporting comparable outcomes from
17
English-speaking youngsters with SSD [28]. For example, word targets according to the child’s
segmental stimulability skills might have enhanced the “…stabilization of the child’s knowledge
of subcomponents that form the foundation of the emergence of more complex phoneme
contrasts…” [28]. Along these lines, research findings suggest that treatment outcomes are
usually enhanced when segmental and suprasegmental levels of therapy stimuli are constructed
on the bases of “in” versus “out” phonemes and syllable structures [50]. A plausible explanation
of the improved intelligibility revealed in the current findings might be related to the child’s
stimulability skills. However, the robustness of this framework remains speculative and awaits
further exploration.
This study adopted a novel underexplored approach to phonological treatment and
expanded a theory-motivated framework to a cross-language context. Speech intelligibility was
induced from carefully selected targets that were experimentally manipulated and clinically
exploited to capitalize on the lexical/phonological interphase. Even though the participant was
verbal at intake session, his profile was characterized by limited phonetic and lexical resources
rendering his speech mostly non-intelligible. As noted earlier, the most robust gain was the
appearance of accurate productions of untreated words as well as the decrease of phonological
process use. These outcomes increased speech intelligibility dramatically.
However, the robustness of the current intervention regiment may be compromised by
single-subject methodology employed here. An additional limitation pertained to the quasi-
density nature of the stimuli employed, in the sense that density was not calculated on the bases
of currently available calculation algorithms. Nevertheless, the fact that the participant exhibited
positive phonological gains within a relatively short period of time could serve as the impetus for
testing efficacy and robustness through a larger data set and data points from more participants.
18
Note that prior to the investigation, the child showed relatively slow progress over a course of
previous intervention and presented with persistent speech output constraints. Furthermore, the
study did not include a control group of children who were treated with a different set of
experimental stimuli via a different experimental paradigm. It is possible that intelligibility gains
were merely a result of the maturation process. This limits the generalizability of results and
compromises the establishment of a causal link between intervention and observed gains.
Nevertheless, the opposite can be inferred in the sense that speech output improvement was a
result of the specific methods employed and notwithstanding the use of specific word targets.
With an increased interest in cross-linguistic language development, whether typical or
atypical, scientists and clinicians need to adhere to theory motivated and evidence-based
therapies that will ensure delivery of best practice. The issue of neighborhood density in the
context of multisyllabic languages and dialectal variability needs to be revisited and
operationalized through large datasets from cross-language developmental phonology databases
which will include Greek and Cypriot-Greek. Clinicians working in a different language can
bear in mind the theoretically-motivated interventions and EBP might inform sensitivity and
specificity issues through carefully selected word stimuli. Cross-language stimuli with a focus
on systematic exploration regarding phonological similarities, might form the impetus for
focused intervention. Regardless of the variability in theoretical and philosophical stances
inherent to intervention regiments in ASD, all frameworks converge towards promoting
and maintaining functional and socially appropriate communicative acts, including
phonological and articulatory capacity. Research-driven findings have clinical implications
as they can direct the setting of treatment goals and the selection of targeted word stimuli
according to the child’s idiosyncratic speech output skills.
19
Overall, despite abundant research findings on treatment efficacy in English-
speaking children with SSD, such investigations from a cross-language perspective
warrant further systematic exploration. Given the recent keen interest in cross-language
phonological treatment, the current investigation adds to this line of research in an effort
to maximize phonological learning skills and promote evidence-based practice as these
parameters constitute fundamental issues at the core of implementing best clinical practice
processes.
Acknowledgment: The author thanks the participant, his family and his teacher for their
participation and commitment to the research project. Also, thanks to the student clinician who
assisted in data collection, data reduction, and reliability measurements.
Statement of Ethics: The participants caregiver provided informed written consent and
ethical approval was granted by the CUT Speech and Language Rehabilitation Clinic session
participation protocol 2017-2018.
Conflict of Interest: The author declares that she has no conflicts of interest.
Author Contributions: Kakia Petinou wrote this article. She is the sole author.
Funding: Provided by an internal start-up fund awarded to the author by the Cyprus
University of Technology CUT/START-UP/PETINOU/REHAB 2014-2017.
20
References
1 Gillon G, Hyter Y, Dreux F, Ferman S, Hus Y, Petinou K, Segal O, Tumarova T,
Vogindroukas I, Westby C, Westerveld M. International Survey of Speech-language pathologists’
practices in working with children with Autism Spectrum Disorder (ASD). Folia Phoniatr Logop.
2017; 69:8-19.
2 Tager-Flushberg H. Risk factors associated with language in autism spectrum disorder:
clues to understanding mechanisms. J Speech Lang Hear Res. 2016; 59:143-154.
3 Koegel RL, Koegel LK. Teaching children with autism: Strategies for initiating
positive interactions and improving learning opportunities. Baltimore MD: Brookes; 1995.
4 Paul R. Interventions to improve communication. Child Adolescence Psychiatr Clin N
Am. 2008; 17: 835-862.
5 Baron-Cohen S. The development of a theory of mind in autism: deviance and delay?
Psychiatr North Am. 1991; 14:33-51.
6 Westby CE. Social-neuroscience theory of mind. Folia Phoniatr Logop. 2014; 66:7-17.
7 Baron-Cohen S. Mind blindness: An Essay on Autism and Theory of Mind.
Cambridge:MA; MIT Press;1995.
8 Wolff J, Piven J. On the emergence of autism: neuroimaging findings from birth to
preschool. London: Future Medicine Publishers; 2013.
9 Petinou K, Minaidou D. Neurobiological bases of autism spectrum disorders and
implications for early intervention: a brief overview. Folia Photr Logop. 2017;69:38-42
10 Wolk L, Edwards ML, Brennan C. Phonological difficulties in children with autism.
Spch Lan Hear Serv Sch. 2016;19:121-129
21
11 McCleery JP., Tully L, Slevc R, Schreibman L. Consonant production patterns of
young severely language-delayed children with autism. J Com Dis. 2006;39:217– 231.
12 Sheinkopf SJ, Mundy P, Oller KD, Steffens M. Vocal atypicality of preverbal autistic
children. Journal of Autism and Developmental Disorders. 2000;30:345-354.
13 Shriberg LD, Paul R, Black LM, Van Santen, JP. The hypothesis of apraxia of speech
in children with Autism Spectrum Disorder. Journal of Autism and Developmental Disorders.
2011.41:405 - 426.
14 Rapin I, Dunn M., Allen D, Stevens M, Fein D. Subtypes of language disorders in
school-age children with autism. Develop Neuropch. 2009;34:66-84.
15 Amato J, Slavin, D. A preliminary investigation of oromotor function in young verbal
and nonverbal children with autism. Infant-Toddler Intervention. 1998; 8:175-184.
16 Boucher J. Articulation in early childhood autism. Journal of Autism and Childhood
Schizophrenia. 1976;6:297-302
17 Koegel RL, Camarata S, Koegel LK, Ben-Tall A, Smith AE. Increasing speech
intelligibility in children with autism. J Aut Dev Dis. 1998; 28: 241-251.
18 Noterdaeme M, Mildenberger K, Minow F, Amorosa, H. (2002). Evaluation of
neuromotor deficits in children with autism and children with a specific speech and language
disorder. Eur Ch A Psych. 2002; 11: 219-25.
19 Ozane A. Childhood apraxia of speech. In Dodd B (editor). Differential diagnosis and
treatment in children with speech sound disorders. London: Whurr; 2005. pp 71-82.
20 Shriberg LD, Paul R, McSweeny JL, Klin AM, Cohen D J, Volkmar FR. Speech and
prosody characteristics of adolescents and adults with high-functioning autism and Asperger
syndrome. J Sp Lan Hear Res. 2001; 44:1097-1115.
22
21 Cleland J, Gibbon FE, Peppe SJ, O'Hare A, Rutherfor, M. “Phonetic and phonological
errors in children with high functioning autism and Asperger syndrome”. Int J of Speech Lan
Path. 2010;12: 1:69-76.
22 Doolaghan CA. The handbook for evidence-based practice in communication
disorders. Baltimore: Paul H Brookes Publishing Company; 2007.
23 Baker E, McLeod S. Evidence-based practice for children with speech sound
disorders: Part 2 application to clinical practice. Lang Spc Serv Sch.2001;30;12-48.
24 Salomone E, Beranova S, Bonnet-Brilhault F. Use of early intervention for young
children with autism spectrum disorder across Europe. Autism. 2016; 20: 233-249.
25 Thordardottir E, Rioux EJ. Does efficacy equal lasting impact? A study of intervention
short term gains, impact on diagnosis status, and association with background variables. Folia
Phtr Logop. 2019; 71:71-82.
26 Law J, Plunkett C, Springer H. Communication interventions and their impact on
behavior in the young child: A systematic review. Ch Lang Teach Therap. 2011; 28:1:7-23.
27 Law J, McKean C, Murphy CA., Thordardottir E. (editors). Managing Children with
Developmental Language Disorders: Theory and Practice Across Europe and Beyond. 1st ed.
Routledge, 2019.
28 McLeod S, Goldstein R. (editors). The International Guide to Speech Acquisition.
Clifton Park, NY: Thomson Delmar Learning, 2007.
29 Yavaş M: Multilingual speech acquisition; in McLeod S. editor: The International
Guide to Speech Acquisition, 2007; pp 96-110. Clifton Park NY: Thompson Delmar Learning,
2007.
23
30 Sosa AV, Stoel-Gammon C. Patterns of intra-word phonological variability during the
second year of life. J Ch Lang 2006;33:31-50.
31 Stoel-Gammon C. Relationships between lexical and phonological development in
young children. J Ch Lang 2011;38:1:1–34.
32 Gillon GT, Moran CA, Hamilton E, Zens N, Bayne G, Smith D (2007) Phonological
Awareness Treatment Effects for Children from Low Socioeconomic Backgrounds, Asia Pacific
Journal of Speech, Language and Hearing. 2007 10:2, 123-140.
33 Enderby P, John A, Petheram B. Therapy Outcome Measures for Rehabilitation
Professionals: Speech and Language Therapy, Physiotherapy, Occupational Therapy. 2013, 2nd
Edition, Emsworth, UK: J&R Press
34 Storkel LH. The emerging lexicon of children with phonological delays: Phonotactic
constraints and probability in acquisition. Journal of Speech, Language, and Hearing Research
2004. 47:1194- 1212.
35 Storkel H, Maekawa J, Hoover JR. Differentiating the Effects of Phonotactic
Probability and Neighbourhood Density on Vocabulary Comprehension and Production: A
Comparison of Preschool Children With Versus Without Phonological Delays. J Sp Lang Hear
Research. 2010; 53:4:933-49.
36 Gierut JA, Morrisette ML. Density, frequency and the expressive phonology of
children with phonological delay. Journal of Child Language. 2012; 39;4; 804–834. doi:10.1017/
S0305000911000304
37 Gierut JA. Complexity in phonological treatment. Lang Spe, and Hearing Services in
Sch. 2001; 32:229–241.
24
38 Luce P, Pisoni D. Recognizing spoken words: the neighborhood activation model. Ear
and Hear. 1989; 19:1-36.
39 Mackridge P. The Modern Greek Language: A descriptive analysis of standard Modern
Greek. Oxford: Oxford University Press; 1985.
40 Newton BE. Cypriot Greek: its phonology and inflections. The Hauge: Mouton;1972.
41. Tsiplakou S, Papapavlou A, Pavlou P, Katsoyannou M: Levelling koineization and
their implications for bidialectalism. Hinskens, F, editor. Language Variation in European
Perspectives. Selected Papers from the 3rd International Conference on Language Variation in
Europe. University of Amsterdam, Philadelphia: John Benjamins. 2006; pp 265- 276.
42 Armostis S. The phonetics of plosive and affricate germination in Cypriot-Greek.
Ph.D. Thesis. Cambridge University; 2009.
43 Petinou K, Minaidou D, Hadjigeorgiou L. The Cypriot-Greek lexical acquisition
checklist. Unpublished manuscript. 1999.
44 American Psychiatric Association: Diagnostic and Statistical Manual of Mental
Disorders, 2013, 5th edition, Arlington: VA. American Psychiatric Publishing.
45 Gotham K, Risi S, Pickles A, Lord C. The Autism Diagnostic Observation Schedule
(ADOS). Journal of Autism and Developmental Disorders. 2006
46 Petinou K, Okalidou A. Speech patterns in Cypriot Greek late-talkers. Appl Pscy
2006; 3: 335-353.
47 Petinou K, Armostis S. Phonological process occurrence in Typically Developing
Toddlers. Folia Phontr Logop. 2017; 68 (5), 199-204.
48 Levanti E, Kirpotin L, Kardamitsi E, Kampouroglou M. The phonological
development of Greek children, Pan-Hellenic Association of Logopedists, Athens:1998
25
49 Cyprus Statistical Service: Demographics report and population statistics, Nicosia
Printing Office of the Republic of Cyprus. 2005.
50 American Speech-Language-Hearing Association. Guidelines for screening for
hearing impairment and middle-ear disorders. ASHA; 1990; 32: 17–24.
51 Kratochwill T, Levin J, editors; School psychology series Single-case intervention
research: Methodological and statistical advances. American Psychological Association. 2013.
52 Kazdin AE. Single-Case Research Designs, 2011, 2nd edition, New York, NY: Oxford
University Press.
53 Petinou K, Theodorou, E. Promoting speech intelligibility through phonological
density in Speech Sound Disorders. Clinical Ling Phon; 2019:30:1597-168.
54 Handbook of the International Phonetic Association. A guide to the use of
International Phonetic Alphabet, Cambridge, UK: University Press.2016.
55 Olswang LB, Stoel-Gammon C, Goggins E, Carpenter, RL: Assessing pre-linguistic
and early speech skills in developmentally young children: University of Washington Press,
Seattle: WA.1987.
56 Bernhardt BM, Stemberger JP. Investigating typical and protracted
phonological development across languages. In Babatsouli E, Ingram D, Muller N,
editors. Crosslinguistic encounters in language acquisition, Typical and atypical development.
Bristol: UK: Multilingual Matters;2017; pp 71-108.
57 McLeod S, Kambanaros M. Intelligibility in Context Scale: Greek. Bathurst, NSW,
Australia: Charles Sturt University. 2012.
26
58 Rvachew S, Leroux É, Brosseau-Lapré F. Production of word-initial consonant
sequences by francophone preschoolers with a developmental phonological disorder. Canadian
Journal of Speech-Language Pathology and Audiology 2014;37:252–267.
59 Glass KD, Hopkins GV. Statistical methods in educational psychology. 3rd Edition. |
Pearson;1986.
60 Rvachew S, Nowak, M. The effect of target selection strategy on sound production
learning. J of Sp Lan and Hea Res.2001;44: pp610–623.
61 Leonard L, Schwartz RG, Folger M, Wilcox, M. Some aspects of child phonology in
imitative and spontaneous speech. J Ch Lang.1978: 5:403-17.
FIGURE 1: INVESTIGATION TIMELINE SCHEME
TABLE 1: DATA MEASURES PRE-TEST, POST-TEST AND FOLLOW UP
APPENXID A: EXPERIMENTAL WORD STIMULI
27