Content uploaded by Terry Hall
Author content
All content in this area was uploaded by Terry Hall on Oct 04, 2018
Content may be subject to copyright.
Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/6939365
TeachingYoungNonverbalChildrenwithAutism
UsefulSpeech:APilotStudyoftheDenverModel
andPROMPTInterventions
ArticleinJournalofAutismandDevelopmentalDisorders·December2006
DOI:10.1007/s10803-006-0142-x·Source:PubMed
CITATIONS
122
READS
486
6authors,including:
Someoftheauthorsofthispublicationarealsoworkingontheserelatedprojects:
MeaningfulOutcomesforFamiliesofChildrenwithAutismViewproject
EmotionalMimicryViewproject
SallyRogers
UniversityofCalifornia,Davis
283PUBLICATIONS21,442CITATIONS
SEEPROFILE
DeborahHayden
ThePROMPTInstitute
13PUBLICATIONS327CITATIONS
SEEPROFILE
SusanHepburn
UniversityofColorado
102PUBLICATIONS4,445CITATIONS
SEEPROFILE
ReneeCharlifue-Smith
UniversityofColorado
4PUBLICATIONS123CITATIONS
SEEPROFILE
AllcontentfollowingthispagewasuploadedbyDeborahHaydenon05January2016.
Theuserhasrequestedenhancementofthedownloadedfile.
eScholarship provides open access, scholarly publishing
services to the University of California and delivers a dynamic
research platform to scholars worldwide.
University of California
Peer Reviewed
Title:
Teaching young nonverbal children with autism useful speech: A pilot study of the Denver model
and PROMPT interventions
Author:
Rogers, Sally J, University of California Davis
Hayden, D
Hepburn, S
Charlifue-Smith, R
Hall, T
Hayes, A
Publication Date:
11-01-2006
Publication Info:
Postprints, UC Davis
Permalink:
http://escholarship.org/uc/item/3xf7w6b8
Additional Info:
http://www.springerlink.com/content/r3tj4221483q5257/?hl=u The original publication is available
at www.springerlink.com
Keywords:
autism, language, treatment, intervention
Abstract:
This single subject design study examined two models of intervention: Denver Model (which
merges behavioral, developmental, and relationship-oriented intervention), and PROMPT (a
neuro-developmental approach for speech production disorders). Ten young, nonverbal children
with autism were matched in pairs and randomized to treatment. They received 12 1-h weekly
sessions of therapy and daily 1-h home intervention delivered by parents. Fidelity criteria were
maintained throughout. Eight of the ten children used five or more novel, functional words
spontaneously and spoke multiple times per hour by the conclusion of treatment. There were
no differences in acquired language skills by intervention group. Initial characteristics of the best
responders were mild to moderate symptoms of autism, better motor imitation skills, and emerging
joint attention skills.
Teaching young nonverbal children 1
Teaching young nonverbal children with autism useful speech:
A pilot study of the Denver Model and PROMPT interventions
Sally J. Rogers, Ph.D. M.I.N.D. Institute, University California Davis
Deborah Hayden, MA CCC/SLP, PROMPT Institute
Susan Hepburn, Ph.D.,
Renee Charlifue-Smith, MA, CCC/SLP
Terry Hall, MA, CCC/SLP
Athena Hayes
Univ. Colorado Health Sciences Center
(In press, Journal of Autism and Developmental Disorders)
This research was supported by an award from the National Institute of Deafness and
Communication Disorders #R21 DC05574, and by the Coleman Institute of University of Colorado. Dr.
Rogers and Dr. Hepburn are also partially supported by the National Institute of Child Health and
Human Development U19 HD35468-08.
Teaching young nonverbal children 2
Abstract
This single subject design study examined two models of speech-language intervention: Denver
Model (which merges behavioral, developmental, and relationship-oriented intervention), and
PROMPT (a neuro-developmental approach for speech production disorders). Ten young,
nonverbal children with autism were matched in pairs and randomized to treatment. They
received 12 one-hour weekly sessions of therapy and daily one-hour home intervention delivered
by parents. Fidelity criteria were maintained throughout. Eight of the 10 children used 5 or
more novel, functional words spontaneously and spoke multiple times per hour by the conclusion
of treatment. There were no differences in acquired language skills by intervention group.
Initial characteristics of the best responders were mild to moderate symptoms of autism, better
motor imitation skills, and emerging joint attention skills.
Teaching young nonverbal children 3
Introduction
Autism is a complex neurodevelopmental disorder that severely compromises functioning in
multiple developmental domains, including social relatedness and reciprocity, nonverbal and verbal
communication, and cognitive and adaptive functioning. Language proficiency is one of the two most
important variables in predicting outcomes in autism (the other being IQ) (Venter, Lord, & Schopler,
1992). Because the language deficit is so disabling, and the acquisition of language so important for
outcomes, autism interventions have focused much attention on helping children with autism acquire
language.
Treating Language Deficits in Autism
Two general approaches for developing communicative speech in young children with autism
have been available to the field for many years. These approaches typically apply learning theory
principles to development of speech, using one of two main methods.
The first method, commonly known as “discrete trial teaching”, uses a didactic, adult-directed
instruction delivered from a pre-set curriculum often taught in massed trials. First described by Wolf,
Risley, and Mees (1964), this approach has been most thoroughly described and manualized by Lovaas
and his associates (Lovaas, 1981; Lovaas, 2002). In this approach, children are taught to attend to adults
and respond to simple instructions (receptive language training); to imitate manual, oral motor, and
vocal behavior; and then to imitate speech. Association learning is then used to teach increasingly
sophisticated expressive language skills. Motivation in this approach is provided through the use of
various external rewards. Many published studies have supported the efficacy of this approach, as
recently reviewed by Goldstein (2002).
The second approach involves a more naturalistic use of learning theory principles. In this
approach, first described in a landmark paper by Hart and Risley (1968) the intervention begins with a
Teaching young nonverbal children 4
child-initiated behavior in a natural interactive context. The adult, using modeling and shaping
techniques, follows the child’s initiation with a prompt or model of more sophisticated verbal behavior
and consequates the child’s production by providing the child-requested object or activity, thus
providing a “natural” reinforcer. This type of teaching approach is variously referred to as natural
language teaching, incidental teaching, or pivotal response training (Koegel, Koegel, & Carter, 1999)
and is also compatible with a developmental orientation to language development (Prizant & Wetherby,
1998). It varies considerably from adult-directed, or discrete-trial teaching in many ways, including the
extent of adult directedness, the individualization of the learning opportunity, the reliance on natural
reinforcers, the role of the child as initiator of the interaction, and the emphasis on generalization and
expansion of skills across settings, activities, and people. This approach has also been found to be
effective in a number of independent replications, as recently reviewed by Goldstein (2002) and Koegel
(2000).
Both of these approaches require intensive interventions, practiced many times daily, over
significant periods of time, to lead to speech acquisition in young nonverbal children with autism.
Children with best outcomes in either approach have typically learned to produce speech in the first year
of treatment, often receiving 25-40 hours of intervention per week in carefully structured settings and
home treatments (McEachin, Smith, & Lovaas, 1993; McGee, Morrier, & Daly, 1999).
While the behavioral approaches to teaching speech to children with autism have demonstrated
efficacy, they are built on a behavioral model of language development that has been replaced over the
past twenty years by the developmental-pragmatics understanding of language development, first
introduced by the writings of Bruner (1975), Bates (1976) and others in the late 1970’s and early 1980’s.
The language impairment in autism is currently understood as a developmental disorder, stemming
from several potential mechanisms, including impaired development of earlier, prelinguistic
Teaching young nonverbal children 5
communicative mechanisms, as defined by a long line of studies beginning with work by Wetherby and
Prutting (1984) and Mundy, Sigman, Ungerer, and Sherman (1986) and supported by a number of
researchers and theorists (Carpenter & Tomasello, 2000; Stone, Ousley, Yoder, Hogan, & Hepburn,
1997; Wetherby, Prizant, & Schuler, 2000 among many others). Lack of social engagement, joint
attention, imitative ability, and presence of cognitive impairments are assumed to play pivotal roles in
poor language acquisition, and developmentally oriented treatments focus on increasing social
engagement, imitation skills, means-end concepts, and understanding of language in order to develop
spoken language.
Thus, developmental approaches to language treatment have considerable theoretical strength,
and main characteristics of a developmental approach are well described by Prizant and colleagues
(Prizant, Wetherby, & Rydell, 2000). The Denver Model (described in Rogers et al, 2000) is a
developmental approach to early autism treatment that delivers a specified developmental curriculum
(individualized for each child based on current abilities) using a combination of empirically supported
teaching techniques (massed trial and naturalistic behavioral teaching and affective dyadic exchanges)
to attain specific developmental outcomes. The Denver Model involves a curriculum and method of
teaching based on both attention to teaching techniques and attention to the interpersonal relationship,
fostered in very specific ways. The Denver Model can be delivered in a variety of formats: preschool
group instruction in either inclusive or special classes, individual therapy sessions, and intensive 1:1
intervention. These formats are often combined.
Outcomes of children receiving the Denver model (Rogers & DiLalla, 1991; Rogers et al., 1986;
Rogers & Lewis, 1989; Rogers et al., 1987) described significant accelerations in developmental rates of
children diagnosed with autism or PDD-NOS, ages 3 to 5 in several developmental areas, including
cognition, language, and social development, including acquisition of useful speech in previously
Teaching young nonverbal children 6
nonverbal children. These studies, using pre-post data, suggested that the Denver Model has the
capacity to affect development in many areas. Furthermore, four independent replications of the model
were carried out in rural Colorado school districts (Rogers et al, 1987). Group data from the replication
sites demonstrated the same child change effects from the model that the original studies found.
For the purposes of the present study, delivery of one aspect of the Denver Model, the
communication curriculum, occurred through a combination of once weekly 50-minute therapy sessions
and daily home review by the parent. This method of delivery was used because it is a typical delivery
method of speech and language therapists working in schools, clinics, and birth to three early
intervention programs. Brief periods of naturalistic social-affective teaching interactions (“sensory
social routines”) alternated with brief periods of didactic teaching during the therapy hour to tap the
strengths of each kind of teaching approach.
The content of the treatment focused on language acquisition and included four separate
teaching strands that begin in the first treatment session:
(1) Using naturalistic teaching strategies and highly motivating social games and object activities to
develop the foundations of nonverbal communication through a high frequency of social interactions,
turn taking, and elicitation and shaping of natural gestures (‘talking bodies”) into intentional
conventional gestures to serve a variety of communicative functions, especially requesting, initiation and
maintenance of social games, greetings, protest, requests for help, and;
(2) Teaching imitation of actions on objects, body movements, oral-facial movements, and speech
sounds using both massed trial and naturalistic behavioral strategies, in both adult directed and child
initiated interactions, including drills, object play, songs and finger plays and object and social requests;
Teaching young nonverbal children 7
(3) Teaching receptive understanding through naturalistic behavioral teaching of simple instructions (sit
down, stand up, come here, look to name), and use of very simple, repetitive language to name social
and sensory activities, songs, and objects;
(4) Teaching object associations by teaching children to match similar objects, pictures, and pictures to
objects (Lovaas, 1981).
(5) Increasing verbal approximations of target words in object play and social routines using naturalistic
behavioral teaching approaches including modeling, and shaping increasingly more accurate
approximations with intrinsic reinforcement strategies.
Treatment begins with an assessment using the Denver Model Curriculum (Rogers et al,
unpublished manuscript). Treatment objectives are written to teach the set of developmental skills just
beyond the child’s current performance level. Teaching programs are developed for each of the teaching
settings and implemented in the treatment. Progress data are gathered and reviewed weekly with
adjustments made in the teaching programs to assure progress. The child’s curriculum is packaged in a
notebook that contains goals and objectives, instructional plans and activities, and data (both
quantitative and qualitative).
Motor Dysfunction in Children with ASD
In the past few years a specific mechanism impairing speech development in autism has been
suggested: oral motor dysfunction (Adams, 1998; Page & Boucher, 1998). The question of motor
dysfunction in autism has a long history. As a result of her studies of imitation in autism, DeMyer and
colleagues (1972) suggested that dyspraxia may be part of the syndrome, severely affecting
communication, adaptive behavior, and learning through limiting imitation of other people, causing a
severe level of disability (DeMyer, Hingtgen, & Jackson, 1981). While the question of motor
dysfunction has been raised sporadically in the autism literature (Damasio & Maurer, 1978; Ohta, 1987)
Teaching young nonverbal children 8
since DeMyer’s original observations, little attention was paid to the motor question until the past
decade. Recently, two sets of literatures have provoked new interest in a possible underlying motor
disorder in autism. One literature that stimulated this thinking is the (largely discredited) literature on
facilitated communication, which suggested that people with autism had intact inner communication
abilities that could not be expressed due to motor output problems (Perry, Bryson, & Bebko, 1998).
The second literature involved a number of recently published empirical studies demonstrating
motor dysfunction in autism (see reviews by Rogers & Bennetto, 2000 and Anzalone & Williamson,
2000) , including studies of dyspraxia-related manual and oral-motor movements (Rogers et al., 1996;
Hughes, 1996; Page & Boucher, 1998; Seal & Bonvillian, 1997; Adams, 1998; Bennetto, 1999; Roy,
Elliott, Dewey, & Square-Storer, 1990). In a descriptive study, Page and colleagues (1998) reported that
79% of a large group of children with autism performed very poorly on tests of oral-motor functioning
and suggested that poor oral and manual development contributed to impaired speech and signing in the
group. In a small comparative study, Adams (1998) demonstrated autism-specific difficulties with oral-
motor control of the motor speech mechanisms for four children with autism compared to age-matched
controls, difficulties similar to those seen in children with oral-motor apraxia. Finally, two groups of
researchers: Stone and colleagues (1997) and Rogers, Hepburn, and Stackhouse (2003) found that young
children with autism were more impaired in the ability to imitate single oral-motor movements than
developmentally matched clinical controls. In the Stone study, oral motor imitation predicted to speech
development in the children with autism one to two years later.
Over the past twenty years, a novel clinical therapy approach: PROMPT (Prompts for
Restructuring Oral Muscular Phonetic Targets) has been developed as a treatment for speech
production disorders in both children and adults based in accepted neuromotor principles of
speech production (Chumpelik (Hayden) 1984). As PROMPT has evolved, it has developed a
Teaching young nonverbal children 9
defined Philosophy, Approach, System and Technique for analyzing and organizing treatment.
(Chumpelik, 1984). Central to the PROMPT Philosophy is that touch is a primary sensory
modality that can be used to: 1) develop, rebalance or re-establish speech motor control,
2) provide a foundation for integrating sensory modalities (audition and vision) in developing
concepts and expressive language and, 3) enhance social- emotional interaction and trust
between clinician and client.
Several papers have been published on the efficacy of PROMPT. Chumpelik (Hayden) &
Sherman (1980) described the progress of an 8-year old, non-speaking child with autism and
cognitive impairment who gained 30 functional words over a four-month period. Other
published single subject studies on the method described children and adults with nonautism
disorders including apraxia of speech, Broca’s aphasia, and developmental dyspraxia
(Chumpelik (Hayden) & Sherman, 1982; Square-Storer & Hayden, 1987; Square-Storer &
Hayden, 1989). Square and colleagues (2000) treated 6 males (ages 4; 2 to 4; 6 years) with
unintelligible speech (who had made minimal progress in traditional therapy) in twice weekly
90-minute group sessions for 15 weeks. Assessment with the Systematic Analysis of Language
Transcripts, SALT (Miller & Chapman, 1993) revealed perceptually improved speech even on
untrained words, and significant improvement on overall behavior, social interactions, and
language skills.
There are nine core elements that are considered essential in typical PROMPT sessions.
They are:
1) The use of tactile-kinesthetic information as a critical modality for recognizing, developing,
re-balancing and integrating cognitive, linguistic and motor behavior.
Teaching young nonverbal children 10
2) Determining a Communication Focus or an aspect of development in which to embed and
focus communication intervention.
3) Developing goals and embedding objectives that embody the Communication Focus while
working on motor/language, cognitive, and social function.
4) Analyzing the global and speech motor sub-systems to determine three priority areas that
need immediate development or rebalancing and create an initial, functional lexicon (core
vocabulary).
5) Deciding on the purpose of prompting and what types of prompts should be used to support
and develop motor control for speech and language and/ or interaction and cognitive
development.
6) Concrete understanding of how chosen goals and objectives will directly affect motor
resourcing and therefore, materials, activity choices.
7) Insuring that a high degree of motor-sound practice (using prompts for accuracy of
production) and generalization of these into novel syllables and words within naturalistic
activities are used within each session.
8) The inclusion of reciprocal interaction or choice making, in all activities, in almost every
turn.
9) Presentation of the same or similar activities over time to provide a structure in which
increased motor-language complexity and cognitive learning of events and sequences may be
practiced.
Using PROMPT in Early ASD
When used with young children with autism, the PROMPT therapist first structures the treatment
hour so that the child must attend to toy-based activities and produce an intentional sound to request.
Teaching young nonverbal children 11
The child’s utterance is then supported through integrated auditory and tactile cues. The adult uses
both vocal modeling and actual manual manipulation of the child’s jaw, lips, and other speech
mechanisms while the child vocalizes to elicit speech approximation of a target word. Physical cues
are gradually faded into visual cues, so that the child responds to a hand movement rather than a touch,
and then further faded. In Phase 1, open vowels are first shaped into a consonant vowel syllable. As
an example, if the child “requests” a ball through reaching for it and any vocalization¸ the therapist
says “Ball, you want ball”, while manipulating the child’s lips to produce the initial syllable /ba/or
prompting in the entire sequence /bal/, and gives the child the ball in order to play with it. They play
for a very brief period and then the situation is repeated so that the child has much opportunity to
practice.
In phase 2, syllables are shaped into words and short phrases, and then in Phase 3 the prompts
are faded and the complexity of all aspects of speech and language are increased. The chosen activities
are those that are 1) motivating to the child 2) within the child’s’ mental age, 3) consider the motor
resourcing or competing motor task requirements, 4) those that enable initial lexicon choices and
5) support functional interaction and independent speech across settings (Hayden, 1999). Repetition of
these similar patterns over different activities allows the child to build success, practice repeatedly and
expand motor, language and cognitive aspects over time.
The Denver Model
Salient Features and Differences Between the Two Models
The Denver and PROMPT Models share some similar features as well as differences in their
intervention programs. Both are developmental, target shared attention and interaction, use naturalistic
communicative exchanges, work towards positive affective communicative behavior, work to match
appropriate level activities to the child’s developmental level, and initially provide a high degree of adult
Teaching young nonverbal children 12
structure. These are commonalities with other high quality language intervention approaches.
Differences between the two models include the way in which the evaluation and information gained in
the evaluation is organized and how the goals and treatment objectives (including sensory, motor,
cognitive, linguistic, social and emotional) are chosen and integrated, the way the motor system is
organized towards tasks (both in fine motor and speech sub-systems), the use of imitation versus
provision of tactual-kinesthetic input, the way tasks are taught, and how and when they are expanded or
changed.
The Present Study
The nature of the barrier to speech development for some young children with autism is not
known. If the social and imitative aspects of autism prevent speech, then treatment should focus on these
targets. If oral motor impairments are preventing development of speech, then this should be the focus
of treatment. The purpose of the present pilot study was to develop the methodologies and preliminary
tests of two different models for developing speech in nonverbal preschoolers with autism: the Denver
Model and PROMPT. While both these treatments have been previously published, both approaches
needed additional empirical support. We also wished to gather data about possible variables moderating
response to each treatment to generate hypotheses for further research. Furthermore, we wished to
examine whether a typical language therapy delivery paradigm involving one hour per week of speech
and language therapy and daily parent review at home could be sufficient to improve speech outcomes
in this group of children.
Method
Participants
Ten children, ages 20 to 65 months, participated in this study. All participants were male; 80%
were Caucasian, 10% African-American, and 10% Hispanic/Biracial. See Tables 1 and 2 for participant
Teaching young nonverbal children 13
descriptions. Participants were recruited from a tertiary university evaluation clinic that specializes in
autism spectrum disorders, as well as through local parent groups (e.g., Autism Society of Colorado),
other early intervention programs and the research database maintained by the Autism Research Group
at the University of Colorado. Inclusion criteria were: (1) diagnosis of autism, (2) spontaneous use of
less than 5 functional words per day according to parent report as well as clinician observation, (3)
developmental quotient (mental age/chronological age * 100) of at least 30, and (3) an absence of a
known co-morbid medical condition (such as tuberous sclerosis). Autism diagnosis was based upon
meeting all of the following criteria: (1) past clinical diagnosis of autism, (2) current clinical diagnosis
as determined by the psychologist on the research team (SJR, SLH), (3) exceeding autism cutoff on the
Autism Diagnostic Observation Schedule – Module I (Lord, Rutter, DiLavore, & Risi, 1999) , (4)
exceeding the autism cutoff on the Social Communication Questionnaire, and (5) meeting APA criteria
for autism as specified in DSM-IV. Estimates of spontaneous use of functional speech were obtained
through a brief interview with the parent, as well as by clinical observation during the developmental
and diagnostic assessments, which usually occurred in two, 2-hour assessment sessions. Four additional
children who were referred were not enrolled in the study due to: (a) not having autism (n=1), (b)
speaking in more than 5 words per day (n=2), and (3) presenting with a developmental quotient of less
than 30 (n=1). These families were referred to clinical services in the community.
Information concerning other treatments the children were receiving was collected by parent
interviews every third therapy session and is summarized in Table 2. Intervention participation did not
change across the 12-week study period for any of the children. One of the children was not enrolled in
any other interventions or school programs. Nine children received individual speech/language therapy.
Of these, 6 were enrolled in 30 minutes to 1 hour of speech therapy per week through their public
schools and had been receiving these services for an average of 13 months prior to inclusion in this
Teaching young nonverbal children 14
project. Three children were enrolled in 2-3 hours per week of speech therapy for an average duration
of 15 months.
Nine children were enrolled in some type of preschool program. Of these, 3 had been served for
12-30 hours per week for over 2 years prior to participating in the study, and 6 participated in preschool
programs for 4 -12 hours per week, for an average of 9 months prior to joining the study. Finally, one
child participated in a 30-hour per week home- and center-based intervention program for 9 months
prior to enrolling in the study.
Design
A single subject design (A-B-A) was implemented in this study across the 10 participants. The
advantage of such a design allows for the establishment of an extended baseline level of performance for
each child against which to compare treatment results, so each child acts as his or her own control. This
is considered an acceptable design for examining treatment effects and an alternate choice to group
designs involving a no treatment control group (Barlow & Hersen, 1984 ). Each child participated in a
pre-treatment and post-treatment assessment battery that included diagnostic, developmental, and
speech-language assessments (see Measures). In addition, examination of treatment effectiveness
included behavioral coding of functional use of speech in 10-minute samples from each speech therapy
session as well as speech probes conducted in an unstructured play assessment in three phases of the
project (baseline, intervention, maintenance). The number of functional words used by the child per
session were charted and evaluated by visual inspection, as described by Barlow and Hersen (1984).
This method of analysis allows for examination of individual differences in response to treatment, and
does not require the withholding of treatment for any of the participants.
Teaching young nonverbal children 15
Procedure
When families volunteered, each eligible child was administered pre-treatment assessments and
several baseline assessments of functional use of speech during unstructured play with an adult. After
completion of the pre-treatment assessment, the child was randomly assigned to a treatment and a
therapist for 12 weekly 1-hour sessions using a computer-generated random numbers table. If families
missed one or more sessions, additional sessions were added to complete the 12-session schedule.
During treatment, assessment of generalization of speech occurred at monthly intervals
(described below). Information concerning other interventions was also gathered at monthly
intervals. At the conclusion of treatment, post-treatment assessment occurred on the
standardized measures and a report was provided to families. Three months after the end of
treatment, one more behavioral assessment of speech during unstructured play was carried out to
examine maintenance of gains in treatment.
This research was conducted at the University of Colorado Autism and Developmental
Disabilities Research Laboratory in Denver, Colorado. Baseline, intervention, post-intervention,
and maintenance evaluation sessions occurred in a two-room suite with a one-way observation
mirror and digital videotaping capabilities. All sessions were videotaped. Parents observed all
therapy sessions and participated in all Denver Model sessions.
Measures
Pre-treatment
All participants were given standardized assessments of cognitive and language functioning,
adaptive behavior, and autism symptoms pre- and post- treatment. In addition, parents were interviewed
about the child’s autism symptoms, adaptive behaviors and use of words and gestures. Diagnostic and
developmental batteries were administered either by a clinical psychologist with extensive experience in
Teaching young nonverbal children 16
autism, or by a graduate student under her supervision. Assessments were not conducted by children’s
therapists, and assessors had no direct knowledge of the speech the child was acquiring in treatment.
The following measures were included:
Autism Diagnostic Observation Scale (ADOS; Lord et al., 2000). All lab personnel were trained
to 85% reliability on the full range of scores so that, in addition to generating the traditional cutoff
scores, we could also generate severity scores. All assessments were videotaped and reliability was
assessed for 60% of ADOS administrations. Weighted kappas on item-agreement ranged from .72 - .96.
Administrators of the ADOS were aware that the child was in the treatment study, but were blind to
which treatment the child was receiving. An additional 40% of the ADOS administrations were coded
by a trained observer who was unaware that the child was enrolled in a treatment study. Weighted
kappas for blind observers ranged from .78 -.92.
The Social Communication Questionnaire (SCQ; Berument, Rutter, Lord, Pickles, & Bailey,
1999) is a parent questionnaire developed from the most sensitive items of the Autism Diagnostic
Interview (Lord, Rutter, & Le Couteur, 1994) . The SCQ has excellent concurrent validity when
compared to the ADI-R. The SCQ takes approximately 20 minutes to complete.
Mullen Scales of Early Learning (Mullen, 1995). This is a standardized, normed developmental
assessment for children aged birth through 68 months. Twenty per cent of the assessments were scored
by two raters for reliability checks. Reliability on subscales was calculated using weighted kappas and
ranged from .82 to .92. The raw scores from the expressive and receptive language subscales were used
to assess change related to experimental treatment.
Vineland Adaptive Behavior Scales – Interview Edition (Sparrow, Balla, & Cicchetti, 1984) was
used to gather parent report of child communication abilities in the home and community.
Teaching young nonverbal children 17
MacArthur Communicative Development Inventory (CDI; Fenson et al., 1993) is a parent report
measure of the child’s use of spontaneous, functional speech throughout the day. The CDI provides a
list of target words and the parent endorses which words were spoken by the child in the preceding
week.
Previous intervention history. Detailed information was collected about all the different types of
treatment a child has received, including the type of treatment, the ratio of children to adults, and the
hours involved. The form was completed through an interview with the parent during the pre-test
assessment, and then was re-examined with the parent every third session of the child’s treatment in this
project.
Background information. Demographic and medical history forms, background information on
several variables (e.g., maternal education, SES, ethnicity, medical history) was collected via parent
report.
Baseline speech probes. Children participated in 3 baseline speech probes to establish a stable
baseline rate of speech prior to beginning the intervention. During this 15-minute play-based procedure,
the child was presented with a novel toy set every five minutes by an examiner and was encouraged to
play and interact with the toys and the adult. The examiner was a research assistant (not a therapist)
who is instructed to be a responsive play partner, but not an initiator of play activities. The examiner
followed the child’s lead, made statements concerning what the child is currently doing (i.e., “You’re
beating on the drum”) and verbally responded to the child’s communications. During each 5-minute
period, the examiner provided one press for a request (e.g., by holding up two toys and waiting for the
child to indicate a choice either verbally or nonverbally), and one press for response to joint attention
(e.g., “Daniel – look!”). The adult behavior was standardized and procedural fidelity examined via a
fidelity checklist, completed every 3rd administration, and exceeded 85%. The child’s functional use of
Teaching young nonverbal children 18
speech during this unstructured play activity with an adult was coded from videotape in 5-minute
intervals. Coding involved frequency counts of the following variables: (1) number of words and/or
approximations produced by the child; and (2) number of phrases produced by the child, as described in
the section above. Two coders rated behaviors on more than 40% of tapes and inter-rater reliability
examined using weighted kappas on categories of communication in generalization probes were as
follows: spontaneous words: .82-.88; spontaneous phrases: .92-.96, imitated words: .88-.92; imitated
phrases: .82-.88; communicative function: .64-.68. Reliability was also not established for use of eye
gaze or gesture, primarily due to problems with camera angles; therefore, nonverbal communicative
behaviors were not coded. The identical coding procedures and definitions were used to assess speech
during therapy sessions, maintenance probes, and follow-up probes.
Therapy sessions and home intervention. Children received either Denver Model or PROMPT
treatment, one hour per week, for 12 treatment hours, as described in the previous section. Each session
was scheduled one week apart. If the parents or therapist had to cancel a session due to illness or
vacation, the child was seen the following week. All 12 sessions were delivered for each child.
For children receiving Denver Model therapy, the parent was present and active in each
treatment session. During a session, the therapist reviewed the child’s notebook and the parent’s data,
asked the parent to demonstrate some of the treatment objectives, taught the child and parent at least one
new skill in each of the four areas described above, and had the parent practice that new skill. Each
parent was asked to spend 45 minutes each day carrying out certain of the child’s treatment objectives in
home or other settings. Each individual family determined how a child’s current treatment objectives
would be incorporated into family routines. The tasks for the parents were specified in the child’s
treatment notebook, where parents recorded all activities and the child’s performance.
Teaching young nonverbal children 19
For PROMPT treatment, parents observed the entire treatment session via video, and at the end
of each session the therapist provided a target for daily home interventions. Parents did not provide
tactile cues for speech at home, but rather provided daily opportunities to practice new words or word
approximations that the child has learned to produce spontaneously or through a verbal model only.
Parents were asked to spend approximately 30 minutes per day carrying out the child’s treatment
objectives within an activity in home or other settings. In conjunction with the clinician, each individual
family determined how the child’s treatment objectives would be incorporated into family routines. The
tasks for the parents were specified in the child’s treatment notebook, where parents recorded all
activities and the child’s performance.
Speech samples acquired during treatment sessions. Every treatment session was videotaped.
One 10-minute sample was randomly selected from each treatment session to be coded using a set of
operational definitions developed by the research team. Time samples were chosen within an “active
treatment” component of the session (i.e., samples were chosen from a random number table, randomly
selected beginning anywhere between minute 5 and minute 40). If, during the randomly chosen sample,
the child left the room (e.g., for a bathroom break), or the therapist focused on the parent and not the
child, then additional child-therapist time was included in the sample following the break in treatment to
have a full 10 minutes of therapist/child interaction.
Behaviors that were coded were: (1) number of novel words and/or approximations produced by
the child; and (2) number of novel phrases produced by the child. For each word, approximation, or
phrase, coders also determine the function of the communication (e.g., behavioral regulation, joint
attention, social interaction – based upon Mundy et al., 1990), and whether the utterance was
spontaneous or prompted. Behaviors were coded by three trained observers and reliability was assessed
on over 40% of samples. Inter-rater reliability examined using weighted kappas on categories of
Teaching young nonverbal children 20
communication were as follows: spontaneous words: .86-.92; spontaneous phrases: .88-.92; imitated
words: .86-.88; imitated phrases: .88-.94; communicative function: .62-.70.
Post-treatment. Participants completed the same diagnostic and developmental measures
described above within three weeks of the last treatment session (one child was seen after 6 weeks due
to parental scheduling mishaps). These assessments were conducted in 2-3 sessions by a clinical
psychologist and a graduate student under her supervision.
Three-month follow-up. Three months after the post-test assessment, the family was invited to
return for a speech probe/play assessment (identical to the procedures used in baseline and during
intervention to assess the child’s functional use of speech). Three families (30%) were unable to be
scheduled for these visits and attempts to conduct this maintenance assessment are on going.
Fidelity of Treatment Implementation
Two speech language therapists with considerable experience treating young children with
developmental disorders including autism (TH, RCS) were trained to fidelity on both models by the
experts/developers of each treatment. Dr. Rogers developed a fidelity rating system involving Likert
style ratings of features that must be present in a Denver Model therapy session and trained each
therapist to a level of 85% or better on the measure. Ms. Hayden developed a similar instrument to
assess PROMPT fidelity. Each of these treatment developers viewed and coded tapes of the therapists
regularly, visited the site at least quarterly, and provided telephone supervision at least monthly.
Treatment fidelity of each therapist was assessed and fidelity achieved at 85% or better on three
consecutive pilot sessions before the experimental treatment began, and maintained at that level as
determined by ratings of 25% of each therapist’s sessions, for both treatments.
Teaching young nonverbal children 21
Results
Single-subject Design
Efficacy of treatment to promote use of functional speech. Eight of the 10 children in the project
demonstrated functional, spontaneous use of 5 novel words or more by the completion of
treatment. Figures 1 and 2 present each subject’s data, including the frequencies of speech use
at baseline, during each therapy session, at generalization probes during the treatment phase, and
at post-treatment generalization points. As seen in these figures, 8 of the 10 children used speech
routinely during therapy sessions and during generalization probes both during and after
treatment. However, child use of functional speech during the play/generalization sessions was
usually markedly less frequent as compared to their performance in treatment. See Probes
detailed in Figures A and B. This may be due to several factors, including: (a) length of
treatment (12 sessions) is insufficient to promote generalization and maintenance; (b) treatment
models need to be modified to target generalization and maintenance; (c) method of assessment
in the probes included too many novel aspects (new person, new toys, new activities); (d)
children were not spontaneous enough in their language to initiate use in an unstructured context.
Collateral gains in social and communicative functioning. Each child participated in
Module I of the ADOS at pre-treatment and post-treatment. Collateral gains in social and
communicative behaviors were observed for some of the children in each intervention model and
are displayed in Table 3. Significant gain was defined as follows: (1) child obtained a score of 2
or 3 on the ADOS at pre-treatment, and (2) child obtained a score of 0 at post-treatment. Gains
in integration of verbal and nonverbal communication were observed in both models. More
Teaching young nonverbal children 22
children in the Denver Model group demonstrated particular gains in imitation and more children
in the Prompt group demonstrated gains in functional play.
Generalization to home: Parent report of words used. Table 4 presents the pre- and post-
treatment data of parent report of words used by the child during the past week, collected via the
MacArthur CDI. Nine of 10 children were reported to use more words after treatment. One child
(David) was reported to use fewer words at the end of treatment. This child was described by his
mother as experiencing a regression in several aspects of behavior during the 4-month period in
which he was participating in the intervention. His reported regression included fewer attempts
to communication, increased aloofness, poor sleeping and eating, decrease in independence in
toileting, and increased irritability. The family was referred to a pediatric neurologist who
initiated assessment and treatment of a possible seizure disorder, with concerns about metabolic
functioning as well. Assessment and treatment are on going and David’s condition is reported to
be improving slowly.
In the following section, a closer examination of three pairs of participants will be
provided: (a) children evidencing the most gains in functional speech; (b) children who made no
meaningful progress in functional speech; and (c) older children (age 5) who participated in the
two models.
Children evidencing the most gains in functional speech. One child in each treatment group
(Ethan and Jeffrey; note all names are fictitious) responded very well to the interventions. Post-
treatment, both used spontaneous phrases regularly and used over 50 different words per session.
Pre-treatment, both were two and had nonverbal developmental quotients above 60 (e.g., 62, and
94). Both demonstrated intentional nonverbal communication and functional and symbolic
imitation, although joint attention behaviors were poor. Post-treatment, both demonstrated
Teaching young nonverbal children 23
increased communicative complexity (i.e., integrating eye gaze and gesture with vocalizations),
as well as consistent joint attention (both responding and following). Neither had had much
intervention, other than preschool group several hours per week. Their mothers had college
degrees, as in the other participant families. Both children had multiple siblings and their
mothers reported implementing the interventions less than one hour per day. Autism severity
was mild to moderate for both pre-treatment, and mild for both post-treatment. Their age,
cognitive abilities, imitation and intentional communication skills, and milder autism symptoms
may have moderated their positive treatment response. These children may well have acquired
functional speech without these treatments; however, the provision of these treatments may have
catalyzed their language growth.
Children who did not develop functional speech. Two children (one in each model) did not
develop functional speech (defined as using 5 or more functional words spontaneously on a daily
basis) by the completion of treatment. Both presented with significant problems attending to an
adult, tolerating demands, and participating in treatment routines. For Alex, who was only 20
months old and had never participated in any structured intervention before, the focus of the first
6 PROMPT sessions was on encouraging attention and engagement with adult-directed activities.
In the beginning of treatment, Alex reacted with intense distress when new materials or activities
were presented to him. Over time, these tantrums decreased in frequency, intensity, and
duration. By the conclusion of treatment, at the age of 25 months, Alex was just beginning to
engage actively in focused activities and had acquired two words in therapy, but made no
meaningful gains in standardized test scores or parent report of use of speech. Alex appeared to
need a longer course of treatment. In fact, 3 months after termination of treatment and continued
weekly speech therapy in the community he is reportedly using 15 new, functional words.
Teaching young nonverbal children 24
Justin did not develop any speech during the treatment. Pre-treatment, Justin was 38
months old, had an overall developmental quotient of 34, language abilities clustering around 10
months, and severe autism symptoms. He demonstrated poor social responsiveness, limited
requesting behavior, no initiation or response to joint attention, deficient imitation ability, and
poor functional play. Justin’s activity level varied substantially from session to session, as he
was sometimes extremely lethargic and other times quite active and restless. The clinicians also
observed some possible signs of seizures and referred the family to a pediatric neurologist for
further evaluation. As with Alex, it was the clinician’s impression that 12 sessions of treatment
was not sufficient to promote functional improvements for Justin. No follow-up data are
available for him.
Treatment of older children. At pre-treatment assessment, Dylan (PROMPT) and Freddie
(Denver Model) were approximately 5 years old (65 and 57 months, respectively) and had been
actively engaged in intervention for well over two years. Dylan had participated in thousands of
hours in a naturalistic intervention approach, where he reportedly made significant gains in social
and emotional responsivity, but no gains in functional speech. He also had received weekly
speech therapy, occupational therapy and full-time preschool for the past two years. At intake,
he vocalized rarely and was being taught to use an augmentative device that he did not use
spontaneously. Pre-treatment, Dylan presented with many strengths: integration of eye gaze with
gesture, initiation and response to joint attention, functional and symbolic imitation, and shared
enjoyment. With the examiner and therapists, his social overtures were unusual and his
responsiveness was inconsistent but was improved with his mother, to whom he directed many
of his facial expressions and vocalizations. Dylan’s functional play was very limited and he
exhibited frequent running, pacing, jumping, and hand-flapping. His receptive language age was
Teaching young nonverbal children 25
approximately 32 months; expressive abilities were estimated at about 10 months. His scores on
the visual-receptive domain (age equivalent of 48 months) and fine motor (age equivalent of 22
months) were quite disparate. On the Leiter International Performance Scale – Revised he
obtained an age equivalent of 64 months and a nonverbal IQ estimate of 94.
Dylan made steady, gains in PROMPT therapy. During therapy, he consistently spoke
more than 60 words per hour in 11 of 12 sessions. His mother reported an increase in vocabulary
size to 145 words. Clinicians’ impressions were that Dylan’s gains were related to (a) his
readiness for structured intervention; (b) the emphasis on building a solid oral motor foundation
within the PROMPT approach; and (c) the active involvement of his mother in the daily
provision of treatment exercises as well as the integration of the techniques into daily activities.
It is important to note that Dylan demonstrated limited generalization of functional speech in the
unstructured play probes. He made modest gains in his scores on standardized assessments over
the period of 4 months (gain of 2 months in expressive abilities and 3 months in receptive
abilities). Dylan continues to participate in community speech/language treatment and is
reported to be gaining expressive skills.
The other 5-year old, Freddie, made moderate gains in functional speech used in therapy
sessions and parent report of vocabulary size (CDI), and gained three months in expressive
language ability over the 3-month treatment course. However, like Dylan, he used very few
words in the generalization and maintenance probes. In the pre-treatment assessment sessions,
Freddie presented as a child with mild symptoms of autism and demonstrated good social
orienting, joint attention, and imitation abilities prior to treatment. These skills may have
facilitated his ability to utilize the interventions. It was the impression of the clinicians that
Freddie responded well to the relationship-based focus of the Denver Model. He continues to
Teaching young nonverbal children 26
participate in speech therapy through is school and is reported by his mother to be making
continued gains.
Discussion
Lack of speech development is one of the most concerning symptoms that young children with
autism present, given the association between early speech and better outcomes in autism. Only a few
language treatment approaches have provided empirical support for their efficacy. The purpose of this
project was to develop the methodologies and preliminary tests of two different models for developing
speech in nonverbal preschoolers with autism. Both approaches had been previously described in the
clinical literature, but neither had been directly tested on nonverbal children. The two models,
PROMPT and Denver Model, shared some commonalities, including a developmental orientation to
language development, but also had significant differences. PROMPT uses a naturalistic
communicative framework based on joint activity routines with toys, and relies on therapists’ use of
manual facilitation of speech motor movements to assist the child to approximate speech sounds during
communicative temptations in these routines. The language “module” of the Denver Model emphasizes
a specific curriculum involving social- affective development, motor imitation, receptive language,
development of nonverbal communicative behaviors, shaping speech from vocalizations, and object
representations.
A single subject design was used in which each child served as his or her own control.
The extended baseline period, and the children’s histories of lack of speech development,
provide a description of their lack of speech use before treatment. However, given that these
were young children still in a developmental period in which language develops rapidly, a design
that uses a no treatment control group would be needed to demonstrate that a similar group of
Teaching young nonverbal children 27
children would not make these kinds of gains in a few months without these particular
treatments. For these reasons, and the small number of subjects, the results reported here should
be considered preliminary and in need of replication and extension.
Of the ten children enrolled in this pilot study, eight developed vocabularies of at least
multiple single words used routinely during therapy sessions and also demonstrated during
generalization and follow up probes. Parents reported a larger vocabulary used at home during
natural routines. Of these eight children who acquired some speech, four developed phrase
speech, two of whom generated and used phrase speech spontaneously and in multiple situations.
All ten children had been rigorously diagnosed with autism, both clinically and again by the
research team prior to enrollment. All had markedly delayed development and used fewer than
five word approximations per day before the treatment, according to maternal verbal report and
verified in baseline measurements. Furthermore, the children received these treatments for one
hour per week by a carefully trained speech and language pathologist and with daily review at
home by the parent, for 12 treatment sessions. Parental response to both treatments was quite
positive. Parents were pleased at the children’s progress and followed through at some level at
home, according to their own reports and to data that they kept.
It is important to recognize how limited the eight children’s speech production was at the
end of this treatment. They were still very language delayed and continued to meet all criteria for
autism. They were however using speech intentionally, spontaneously, and meaningfully in
therapy sessions and at home. Their beginning skills indicated a readiness to move further with
continued language therapy, to which all were referred. What is responsible for the children’s
progress? For some children, the outcome is probably due to their young age and response to
initiation of good treatment with daily home follow-through. These children may have developed
Teaching young nonverbal children 28
speech given any reasonable therapeutic approach. However, these effects were not due solely to
the initiation of good treatment in previously untreated and very young children. Nine of the 10
had had previous treatment, and some children had had years of speech and language therapy
prior to enrollment in this study. Furthermore, our baseline procedures demonstrated that
whatever other treatments they were receiving were not affecting their use of speech prior to the
onset of these experimental sessions.
Given that the children received only 12 hours of direct treatment, parent involvement
likely played a significant role. Parents in the Denver Model were physically present in all
sessions and participated while being coached by the therapist. Parents in both treatments
observed every session, were coached in specifics to practice each week, were taught to keep
data, and handed their data in to their therapists. Thus, the parents continued the treatment for
many more hours each week. Positive results of parent-delivered treatment is a consistent finding
in the autism literature (Howlin & Rutter, 1989; Harris, Wolchik & Weitz, 1981; Laski,
Charlop& Schreibman, 1988; Charlop & Trasowech, 1991). Finally, both therapies involved
interventions very carefully fitted to children’s current level of understanding, and delivered in a
fairly structured format that emphasized child intentional communication and child initiative. It
may be that these similarities in the treatments resulted in the similar outcomes. Given the small
number of subjects and the study design, direct comparison of the two treatments was not
possible. It will require a much larger study, with a different design, to determine whether
certain child characteristics moderate response to each of these treatments.
There are several weaknesses in this study that need to be considered and that may affect
the validity and the generalizability of the findings. It was a very small study with a very short
time span, and the subjects were generally middle class families. The children were somewhat
Teaching young nonverbal children 29
heterogeneous in that they ranged in age from 2-5 years, nonverbal MA’s ranged from 18-36
months, and expressive language from 12-18 months. Compared to other studies, however, this
is probably a fairly homogeneous group. We do not know if all families could make similar use
of these treatments. There was a long training period to reach fidelity in both models, even for
these very sophisticated university based and autism experienced speech and language therapists.
Each treatment approach has significant requirements for assessment, treatment planning, and
data collection. Many elements were involved in each treatment, including parental coaching,
parent follow through at home, and several different aspects of each treatment; we do not know
which of the many elements are crucial for the outcomes. The children had quite different
characteristics, and we do not know what child characteristics led to success or failure in each
approach. The maintenance and follow-up data are very limited. Parent implementation was
monitored only via parent report. There was no no-treatment control group. Finally, some
children were getting other treatments during this treatment study (see “Participants”), the
majority of which included 1 hour per week of speech and occupational therapy delivered within
a public school setting. For 7 of the 9 children engaged in other treatments, the average length of
time involved in those interventions before enrolling in this project was 14 months. Therefore,
for 78% of children receiving other treatments, other intervention experiences had not
significantly impacted their language functioning in well over a year of participation. Two of the
children initiated other treatments approximately 4 months before beginning the present project
and it is very difficult to attribute outcomes to specific treatments in these cases. We cannot rule
out the effect of other treatments on the children’s progress. Thus, we cannot determine what of
the treatments were responsible for the change in the children, and the two treatments need to be
examined further in replication studies that examine whether these effects can be replicated. If
Teaching young nonverbal children 30
so, then the variables responsible for the change will need to be isolated. Next studies should
carefully examine relationships between individual child characteristics and response to each
specific treatment, as well as effectiveness if the interventions in more community-based
delivery systems. The present study represents the first step of many involved in empirical
examination of treatments and effects. .
Thus, this was a preliminary study using very small numbers of children and the results
await replication. However, these results are similar to earlier published reports of both didactic
and naturalistic behavioral treatment successes at teaching nonverbal children to speak (Koegel,
O’Dell, & Koegel, 1987; Lovaas, Berberich, Perloff, & Schaeffer, 1966). If the present findings
are replicated in larger, more rigorous studies, the implications are several. For one, both
approaches focused on development of speech, not use of alternative communicative systems. It
will be important to examine the use of alternative systems (signs, pictures, etc) and child
characteristics to determine for which children AAC systems accelerate, decelerate, or do not
affect the rate of speech acquisition. A second implication involves expectations of speech
development in early autism. Current studies of young children with autism suggest that
approximately 75% will develop speech during the preschool years (Lord, Risi, & Pickles, 2004)
given typical community intervention approaches. This leaves 25% without useful speech. If a
larger study can replicate the finding here that a number of these children can learn to speak
during their preschool years given carefully delivered treatment, the bar is raised considerably in
terms of what the field should expect from language interventions in early autism. If presence of
useful speech by 5 continues to be a moderator of better outcomes, this kind of attention to
speech development may improve outcomes across the board for children with autism.
Teaching young nonverbal children 31
In conclusion, speech is a crucial tool for learning, self-advocacy, social relationships,
and participation in community. Effective treatment of autism (and any other disorder that limits
speech use) requires that we identify or develop effective treatments for teaching speech to
preverbal children who are at risk for not developing speech. In addition, it requires that speech
and language therapists and others know about effective approaches, can learn to deliver
effective approaches at appropriate levels of fidelity, and can be funded by public agencies to
deliver such care. However, empirically testing treatments is a necessary first step, and this pilot
project contributes towards that goal.
References
Adams, L. (1998). Oral-motor and motor-speech characteristics of children with autism.
Focus on Autism and Other Developmental Disabilities, 13, 108-112.
Anzalone, M. & Williamson, G. G. (2000). Sensory processing and motor performance in
autism. In A. M.Wetherby & B. M. Prizant (Eds.), Autism spectrum disorders: a transactional
developmental perspective. (pp. 143-166). Baltimore: Brookes.
Barlow, D. H. & Hersen, M. (1984). Single case experimental designs; Strategies for
studying behavioral change. (2 ed.) Boston: Allyn & Bacon.
Bates, E. (1976). Language and context: The acquisition of pragmatics. New York:
Academic Press.
Bennetto, L. (1999). A componential approach to imitation and movement deficits in
autism. Dissertation Abstracts international, 60(2-B), 0819.
Teaching young nonverbal children 32
Berument, S. K., Rutter, M., Lord, C., Pickles, A., & Bailey, A. (1999). Autism screening
questionnaire: diagnostic validity. British Journal of Psychiatry, 175, 444-451.
Bruner, J. (1975). The ontogenesis of speech acts. Journal of Child Language, 2, 1-19.
Carpenter, M. & Tomasello, M. (2000). Joint attention, cultural learning, and language
acquisition: Implications for children with autism. In A. M.Wetherby & B. M. Prizant (Eds.),
Autism spectrum disorders: A transactional developmental perspective. (pp. 31-54). Baltimore:
Brookes.
Charlop, M. H., Trasowech, J. E. (1991). Increasing autistic children’s daily spontaneous
speech. Journal of Applied Behavior Analysis, 24, 747-761.
Chumpelik, D. (1984). The PROMPT system of therapy: Theoretical framework and applications
for developmental apraxia of speech. Seminars in Speech and Language, 5, 139-156. Interventions, 2,
98-112.
Chumpelik (Hayden), D. & Sherman, J. (1980). Using a tactile approach in the acquisition of
functional oral communication in a non-verbal, eight year old autistic child: a case study. Unpublished
research.
Chumpelik (Hayden), D., & Sherman, J. (1983). Treatment comparisons for
developmental apraxia of speech. Unpublished research, Thistletown Regional Centre, Toronto, ON.
Chumpelik, D. (1984). The PROMPT system of therapy: Theoretical framework and applications
for developmental apraxia of speech. Seminars in Speech and Language, 5, 139-156.
Teaching young nonverbal children 33
Damasio, A. R. & Maurer, R. G. (1978). A neurological model for childhood autism.
Archives of Neurology, 35, 777-786.
DeMyer, M. K., Alpern, G. D., Barton, S., DeMyer, W. E., Churchill, D. W., Hingtgen, J.
N., Bryson, C. Q., Pontius, W., & Kimberlin, C. (1972). Imitation in autistic, early
schizophrenic, and nonpsychotic subnormal children. Journal of Autism and Childhood
Schizophrenia, 2, 264-287.
DeMyer, M. K., Hingtgen, J. N., & Jackson, R. K. (1981). Infantile autism reviewed: A
decade of research. Schizophrenia Bulletin, 7, 388-451.
Fenson, L., Dale, P. S., Reznick, J. S., Thal, D., Bates, E., Hartung, J. P., Pethick, S., &
Reilly, J. S. (1993). MacArthur Communicative Development Inventories. San Diego, CA:
Singular Publishing Group.
Goldstein, H. (2002). Communication intervention for children with autism: a review of
treatment efficacy. Journal of Autism and Developmental Disorders, 32, 373-396.
Harris, S. L., Wolchik, S. A., & Weitz, S. (1981). The acquisition of language skills by
autistic children: Can parents do the job? Journal of Autism and Developmental Disorders, 11,
373-384.
Hart, B. M. & Risley, T. R. (1968). Establishing use of descriptive adjectives in the
spontaneous speech of disadvantaged preschool children. Journal of Applied Behavior Analysis,
1, 109-120.
Teaching young nonverbal children 34
Hayden, D., Wetherby, A., & Prizant, B. M. (1999). Early Motor Control Scales.
Chicago: Applied Symbolix.
Howlin, P. & Rutter, M. (1989). Mothers' speech to autistic children: A preliminary
causal analysis. Journal of Child Psychology and Psychiatry, 30, 819-843.
Hughes, C. (1996). Planning problems in autism at the level of motor control. Journal of
Autism and Developmental Disorders, 26, 99-109.
Koegel, L. K. (2000). Interventions to facilitate communication in autism. Journal of
Autism and Developmental Disorders, 30, 383-391.
Koegel, R. L., Koegel, L. K., & Carter, C. M. (1999). Pivotal teaching interactions for
children with autism. School Psychology Review, 28, 576-594.
Koegel, R. L., O'Dell, M. C., & Koegel, L. K. (1987). A natural language teaching
paradigm for nonverbal autistic children. Journal of Autism and Developmental Disorders, 17,
187-199.
Laski, K. E., Charlop, M. H., & Schreibman, L. (1988). Training parents to use the
natural language paradigm to increase their autistic children's speech. Journal of Applied
Behavior Analysis, 21, 391-400.
Lord, C., Risi, S., Lambrecht, L., Cook, E. H., Jr., Leventhal, B., DiLavore, P. C.,
Pickles, A., & Rutter, M. (2000). The Autism Diagnostic Observation Schedule - Generic: A
standard measure of social and communication deficits associated with the spectrum of autism.
Journal of Autism and Development Disorders, 30, 205-223.
Teaching young nonverbal children 35
Lord, C., Risi, S., & Pickles, A. (2004). Trajectory of language development in autistic
spectrum disorders. In M. L.Rice & S. F. Warren (Eds.), Developmental language disorders:
From phenotypes to etiologies. (pp. 7-30). Mahweh, NJ: Lawrence Erlbaum.
Lord, C., Rutter, M., DiLavore, P., & Risi, S. (1999). Autism Diagnostic Observation
Schedule - WPS Edition. Los Angeles, CA: Western Psychological Services.
Lord, C., Rutter, M., & Le Couteur, A. (1994). Autism diagnostic interview-revised: a
revised version of a diagnostic interview for caregivers of individuals with possible pervasive
developmental disorders. Journal of Autism and Developmental Disorders, 24, 659-685.
Lovaas, O. I. (1981). Teaching Developmentally Disabled Children: The Me Book.
Baltimore: University Park Press.
Lovaas, O. I. (2002). Teaching individuals with developmental delays: Basic intervention
techniques. Austin, TX: PRO-ED.
Lovaas, O. I., Berberich, J. P., Perloff, B. F., & Schaeffer, B. (1966). Acquisition of
imitative speech by schizophrenic children. Science, 151, 705-707.
McEachin, J. J., Smith, T., & Lovaas, I. O. (1993). Long-term outcome for children with
autism who received early intensive behavioral treatment. American Journal on Mental
Retardation, 97, 359-372.
McGee, G. G., Morrier, M. J., & Daly T. (1999). An incidental teaching approach to
early intervention for toddlers with autism. Journal of the Association for Persons with Severe
Handicaps, 24, 133-146.
Teaching young nonverbal children 36
Miller, J. and Chapman, R. (1993) SALT: Systematic analysis of language transcripts (computer
programs to analyze language samples). Madison, WI: Language Analysis Laboratory, Waisman
Center, University of Wisconsin-Madison.
Mullen, E. M. (1995). Mullens Scales of Early Learning. (AGS Edition ed.) Circle Pines:
American Guidance Service, Inc.
Mundy, P., Sigman, M., Ungerer, J., & Sherman, T. (1986). Defining the social deficits
of autism: The contribution of non-verbal communication measures. Journal of Child
Psychology and Psychiatry, 27, 657-669.
Ohta, M. (1987). Cognitive disorders of infantile autism: A study employing the WISC,
spatial relationships, conceptualization, and gestural imitation. Journal of Autism and
Developmental Disorders, 17, 45-62.
Page, J. & Boucher, J. (1998). Motor impairments in children with autistic disorder.
Child Language Teaching and Therapy, 14, 233-259.
Perry, A., Bryson, S., & Bebko, J. (1998). Brief Report: Degree of facilitator influence in
facilitated communication as a function of facilitator characteristics, attitudes, and beliefs.
Journal of Autism and Developmental Disorders, 28, 87-90.
Prizant, B. M. & Wetherby, A. M. (1998). Understanding the continuum of discrete-trial
traditional behavioral to social-pragmatic developmental approaches in communication
enhancement for young children with autism/pdd. Seminars in Speech and Language, 19, 329-
353.
Teaching young nonverbal children 37
Prizant, B. M., Wetherby, A. M., & Rydell, P. (2000). Communication intervention
issues for children with autism spectrum disorders. In A. M.Wetherby & B. M. Prizant (Eds.),
Autism spectrum disorders; a transactional developmental perspective (pp. 193-224). Baltimore.
Rogers, S. J. (1991). A psychotherapeutic approach for young children with pervasive
developmental disorders. Comprehensive Mental Health Care, 1, 91-108.
Rogers, S. J. & Bennetto, L. (2000). Intersubjectivity in autism: Roles of imitation and
executive function. In A. M.Wetherby & B. M. Prizant (Eds.), Autism Spectrum Disorders; A
transactional developmental perspective (pp. 79-108). Baltimore: Brookes.
Rogers, S. J., Bennetto, L., McEvoy, R., & Pennington, B. F. (1996). Imitation and
pantomime in high functioning adolescents with autism spectrum disorders. Child Development,
67, 2060-2073.
Rogers, S. J. & DiLalla, D. (1991). A comparative study of the effects of a
developmentally based instructional model on young children with autism and young children
with other disorders of behavior and development. Topics in Early Childhood Special Education,
11, 29-48.
Rogers, S. J., Hall, T., Osaki, D., Reaven, J., & Herbison, J. (2000). A comprehensive,
integrated, educational approach to young children with autism and their families. In S. L. Harris
& J. S. Handleman (Eds.), Preschool education programs for children with autism (2nd ed.,
Austin, TX: Pro-Ed.
Teaching young nonverbal children 38
Rogers, S. J., Hepburn, S. L., & Stackhouse, T. (2003). Imitation performance in toddlers
with autism and those with other developmental disorders. The Journal of Child Psychology and
Psychiatry and Allied Disciplines, 44, 763-781.
Rogers, S. J., Herbison, J., Lewis, H., Pantone, J., & Reis, K. (1986). An approach for
enhancing the symbolic, communicative, and interpersonal functioning of young children with
autism and severe emotional handicaps. Journal of the Division of Early Childhood, 10, 135-148.
Rogers, S. J. & Lewis, H. (1989). An effective day treatment model for young children
with pervasive developmental disorders. Journal of the American Academy of Child and
Adolescent Psychiatry, 28, 207-214.
Rogers, S. J., Lewis, H. C., & Reis, K. (1987). An effective procedure for training early
special education teams to implement a model program. Journal of the Division of Early
Childhood, 11, 180-188.
Roy, E. A., Elliott, D., Dewey, D., & Square-Storer, P. (1990). Impairments to praxis and
sequencing in adult and developmental disorders. In C. Bard, M. Fleury, & L. Hay (Eds.),
Development of eye-hand coordination across the life span (pp. 358-384). Columbia, SC:
University of South Carolina Press.
Seal, B. C. & Bonvillian, J. D. (1997). Sign language and motor functioning in students
with autistic disorder. Journal of Autism and Developmental Disorders, 27, 437-466.
Sparrow, S. S., Balla, D. A., & Cicchetti, D. (1984). Vineland Adaptive Behavior Scales.
Circle Pines, MN: American Guidance Service.
Teaching young nonverbal children 39
Square-Storer, P. & Hayden, D. (1987). PROMPT Treatment. In P. Square-Storer (Ed.),
Acquired apraxia of speech in aphasic adults (pp. 190-219). New York: Taylor and Francis.
Square-Storer, P. & Hayden, D. (1989). PROMPT treatment. In P. Square-Storer (Ed.),
Acquired Apraxia of Speech in Aphasic Adults: Theoretical and Clinical Issues (pp. 190-219).
Hillsdale, NJ, England: Lawrence Erlbaum Associates, Inc.
Square, P. A., Goshulak, D., Bose, A., & Hayden, D. (2000). The Effects of Articulatory
Subsystem Treatment for Developmental; Neuromotor Speech Disorders (Rep. No. Tenth
Biennial Conference on Motor Speech Disorders and Speech Motor Control). San Antonio, TX.
Stone, W. L., Ousley, O. Y., Yoder, P. J., Hogan, K. L., & Hepburn, S. L. (1997).
Nonverbal communication in two and three-year-old children with autism. Journal of Autism and
Developmental Disorders, 27, 677-696.
Venter, A., Lord, C., & Schopler, E. (1992). A follow-up study of high-functioning
autistic children. Journal of Child Psychology and Psychiatry, 33, 489-507.
Wetherby, A. M., Prizant, B. M., & Schuler, A. L. (2000). Understanding the nature of
communication and language impairments. In A. M.Wetherby & B. M. Prizant (Eds.), Autism
spectrum disorders: a transactional developmental perspective (pp. 109-142). Baltimore:
Brookes.
Wetherby, A. M. & Prutting, C. A. (1984). Profiles of communicative and cognitive-
social abilities in autistic children. Journal of Speech and Hearing Research, 27, 364-377.
Teaching young nonverbal children 40
Wolf, M., Risley, T., & Mees, H. (1964). Application of operant conditioning procedures
to the behaviour problems of an autistic child. Behaviour Research and Therapy, 1, 305-312.
Teaching young nonverbal children 41
Table 1
Participant Characteristics
Child Chronologi
cal Age
(mos.)
MSEL
Overall
mental
age
(mos.)
MSEL
Overall
developmental
quotient (DQ)
(ma/ca*100)
MSEL
Nonverbal
mental
age (mos.)
MSEL
Expressive
language
age (mos.)
MSEL
Receptive
language
age
(mos.)
ADOS
Severity
of
Autism
DENVER
MODEL
Ethan 29 18 62 21 14 14 Moderate
Justin 38 13 34 18 10 10 Severe
Freddie 57 29 48 36 13 28 Mild
Max 39 18 46 24 12 15 Moderate
Ryan 40 19 48 25 13 13 Moderate
PROMPT
MODEL
Jeffrey 24 23 94 24 18 23 Mild
Michael 28 13 53 16 9 10 Severe
Alex 20 13 63 16 6 8 Moderate
Dylan 65 27 41 31 16 31 Mild
David 44 17 38 22 9 14 Moderate
Teaching young nonverbal children 42
Table 2
Family and environmental characteristics
Child Maternal
education
# hours in preschool
programs1
# hours in speech
therapy1
Total # hours of
intervention
experience1
DENVER
MODEL
Ethan College 4 hrs/wk for 6 mon 30 min./wk for 3 mon @ 110 hours
Justin HS 10 hrs/wk for 8 mon 30 min/wk for 8 mon @350 hours
Freddie Master’s 12 hrs/wk for 3 years 1 hr/wk for 3 years @ 2000 hours
Max College 12 hrs/wk for 1 year 30 min/wk for 1 year @600 hours
Ryan College 12 hrs/wk for 8 mon 1 hr/wk for 6 months @410 hours
PROMPT
MODEL
Jeffrey College 12 hrs/wk for 8 mon None @380 hours
Michael HS 30 hrs/wk for 9 mon 3 hours/wk for 9 mon @1200 hours
Alex College 6 hrs/wk for 4 months 30 min/wk for 8 mon @78 hours
Dylan College 30 hrs/wk for 2 years 2 hours/wk for 2 years @3400 hours
David Master’s 30 hrs/wk for 2 years 3 hours/wk for 1 year @3200 hours
1 Prior to and during inclusion in research study
Teaching young nonverbal children 43
Table 3
Collateral gains in early social-communicative behaviors 2
DENVER MODEL Ethan Justin Freddie Max Ryan
Pointing + + +
Gestures +
Integration of gaze and
other behaviors during
social overtures
+ + +
Requesting +
Giving
Showing +
Initiation of joint attention +
Response to joint attention + + +
Imitation + + + +
Functional play +
2 Plus sign (“+”) denotes a change in ADOS rating from a 2 or a 3 at pre-treatment (impairment) to a 0 at post-
treatment (not impaired); imitation is shown as an improvement if child did not have functional imitation at pre-
treatment and demonstrated it at post-treatment within the ADOS imitation item
Teaching young nonverbal children 44
PROMPT MODEL Jeffrey Michael Alex Dylan David
Pointing +
Gestures + + +
Integration of gaze and
other behaviors during
social overtures
+ + + +
Requesting +
Giving + + + +
Showing +
Initiation of joint attention + +
Response to joint attention +
Imitation +
Functional play + + +
Teaching young nonverbal children 45
Table 4
Parent report of speech and communication behaviors at home using the McArthur CDI pre- and
post-treatment3
DENVER
MODEL
# of words
child
understands
and says
before
Treatment
# of words
child
understands
and says
after
Treatment
Ethan 20 193
Justin 0 4
Freddie 3 19
Max 10 24
Ryan 4 12
PROMPT
MODEL
Jeffrey 20 190
Michael 0 3
Alex 0 2
Dylan 184 145
David 40/ 8
3 Higher scores are indicative of better skills
4 Parent noted that only she can understand his speech reliably
Teaching young nonverbal children 46
Table 5
Pre-and post-treatment results of standardized measures of speech and language.
Child
Pre-tx
MSEL
Expressive
Raw Score
(age equiv.)
Post-tx
MSEL
Expressive
Raw Score
(age equiv.)
Gain
in raw
score
points
Pre-tx
MSEL
Receptive
Raw Score
(age equiv.)
Post-tx
MSEL
Receptive
Raw Score
(age equiv.)
Gain in
raw
score
points
Denver Model
Ethan 15 (14 mos.) 25 (25 mos.) 10 15 (14 mos.) 23 (23 mos.) 9
Justin 11 (10 mos.) 11 (9 mos.) 0 10 (10 mos.) 12 (11 mos.) 2
Freddie 13 (13 mos.) 16 (16 mos.) 3 27 (28 mos.) 28 (30 mos.) 1
Max 12 (12 mos.) 15 (15 mos.) 3 16 (15 mos.) 16 (15 mos.) 1
Ryan 13 (13 mos.) 14 (14 mos.) 1 14 (13 mos.) 14 (13 mos.) 0
Prompt Model
Jeffrey 18 (18 mos.) 31 (35 mos.) 13 23 (23 mos.) 29 (31 mos.) 6
Michael 10 (9 mos.) 11 (10 mos.) 1 10 (10 mos.) 12 (11 mos.) 2
Alex 7 (8 mos.) 9 (8 mos.) 2 8 (10 mos.) 13 (11 mos.) 5
Dylan 16 (16 mos.) 18 (18 mos.) 2 29 (31 mos.) 32 (36 mos.) 3
David 14 (14 mos.) 16 (16 mos.) 2 11 (9 mos.) 14 (13 mos.) 3
Teaching young nonverbal children 47
Figure 1. Graphs of each individual subject’s total number of spoken words used, averaged per hour, across at baseline, therapy
sessions, and follow-up, calculated from 10 minute samples.
Ethan: Denver Model
Words per Hour
0
50
100
150
200
250
300
350
400
1 2 3 4 5 6 7 8 9 10111213141516171819
Words
Treatment Generalization Probe
Baseline Intervention Maintenance
Teaching young nonverbal children 48
Justin: Denver Model Intervention
Words per Hour
0
5
10
15
20
25
30
1 2 3 4 5 6 7 8 9 101112131415161718
Words
Generalization Probe Treat ment
Baseline Intervention Maintenance
Teaching young nonverbal children 49
Freddie: Denver Model Intervention
Words per Hour
0
10
20
30
40
50
60
70
80
90
123456789101112131415161718
Words
Generalization Probe Treatment
Baseline Intervention Maintenance
Teaching young nonverbal children 50
Max: Denver Model Intervention
Words per Hour
0
5
10
15
20
25
30
123456789101112131415161718
Words
Generalization Probe Treat ment
Baseline Intervention Maintenance
Teaching young nonverbal children 51
Ryan: Denver Model
Words per Hour
0
20
40
60
80
100
120
140
12345678910111213141516171819
Words
Generalization Probe Treatment
Baseline Intervention Maintenance
Teaching young nonverbal children 52
Jeffrey: PROMPT
Words per Hour
0
500
1000
1500
2000
2500
3000
12345678910111213141516171819
Words
Generalization Probe Treatment
Baseline Intervention Maintenance
Teaching young nonverbal children 53
Michael: PROMPT Intervention
Words per Hour
0
10
20
30
40
50
60
123456789101112131415161718
Words
Generalization Probe Treatment
Baseline Intervention Maintenance
Teaching young nonverbal children 54
Alex: PROMPT Intervention
Words per Hour
0
5
10
15
20
25
30
123456789101112131415161718
Words
Generalization Probe Treatm ent
Baseline Intervention Maintenance
Teaching young nonverbal children 55
Dylan: PROMPT
Words per Hour
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10111213141516171819
Words
Generalization Probe Treatment
Baseline Intervention Maintenance
