0% found this document useful (0 votes)
84 views16 pages

Wetherby2006 PDF

Uploaded by

Wanessa Andrade
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
84 views16 pages

Wetherby2006 PDF

Uploaded by

Wanessa Andrade
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 16

TECSE 26:2 67–82 (2006) 67

Early Social Interaction Project for Children


With Autism Spectrum Disorders Beginning
in the Second Year of Life:
A Preliminary Study

T
he Early Social Interaction (ESI) Project (Woods & Wetherby, 2003) was de-
signed to apply the recommendations of the National Research Council (2001) to
toddlers with autism spectrum disorders (ASD) by using a parent-implemented
intervention that (a) embeds naturalistic teaching strategies in everyday routines and
(b) is compatible with the mandate of the Individuals with Disabilities Education Im-
Amy M. Wetherby provement Act (IDEIA) of 2004, Part C. This quasi-experimental study is a preliminary
and effort by the authors to evaluate the effects of ESI on the social communication out-
comes for a group of 17 children with ASD who entered ESI at age 2 years. The results
Juliann J. Woods indicated significant improvement on 11 of 13 social communication measures. The re-
Florida State University searchers compared the ESI group with a contrast group of 18 children with ASD who
entered early intervention at age 3 years. The contrast group’s results were comparable
to those of the ESI postintervention group on communicative means and play, but the
contrast group as a whole demonstrated significantly poorer performance on all other
social communication measures. These findings offer promise for the use of parent-im-
plemented interventions in promoting social communication for toddlers with ASD.

Mounting evidence now exists in regards to the effective- terventions for children with ASD from birth through
ness of intensive early interventions for a substantial pro- age 8 years and concluded that a large body of research
portion of young children with autism spectrum disorders indicated significant progress in response to intervention
(ASD; Dawson & Osterling, 1997; National Research for a substantial proportion of those children. Although
Council [NRC], 2001). Furthermore, researchers have sug- the studies the NRC reviewed reflected a range of teach-
gested that the age of entry into intervention is predictive ing techniques, the council concluded that a convergence
of outcome. Children with ASD who participated in inten- of evidence identifies the following characteristics as es-
sive interventions by 3.5 years of age had significantly bet- sential active ingredients of effective interventions for
ter outcomes than their peers with ASD who received children with ASD:
such interventions after age 5 (Fenske, Zalenski, Krantz, &
McClannahan, 1985; Harris & Handleman, 2000). These 1. entry into intervention programs as soon
findings support the importance of early identification and as ASD is suspected;
intervention in improving outcomes. 2. active engagement in intensive instruc-
tional programming for a minimum of
5 hours per day, 5 days a week;
EFFECTIVENESS OF EARLY INTERVENTION FOR 3. use of planned teaching opportunities that
YOUNG CHILDREN WITH ASD are structured over brief periods of time
and repeated systematically;
The National Research Council (2001) conducted a sys- 4. sufficient individualized adult attention on
tematic, thorough review of research on educational in- a daily basis;

Address: Amy M. Wetherby, Department of Communication Disorders, RRC 107, Florida State University,
Tallahassee, FL 32306-7814; e-mail: awetherb@fsu.edu

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


68 Topics in Early Childhood Special Education 26:2

5. inclusion of a family component, including Meaningful changes that reflect core social commu-
parent training; nication deficits in children with ASD still need to be doc-
6. mechanisms for ongoing assessment and umented. The current research literature has indicated
program evaluation, with corresponding that researchers need to measure multiple aspects of so-
adjustments in programming; and cial communication to better characterize this popula-
7. priority given to instruction on (a) functional, tion and develop possible treatment outcomes. Even
spontaneous communication, (b) social in- studies of the most effective treatments for children with
struction in various settings, (c) play skills ASD have shown variable outcomes (NRC, 2001), and a
that focus on play with peers and peer inter- child’s social communication skills before treatment may
action, (d) new skill acquisition, generaliza- influence the response to treatment. Little is known
tion, and maintenance in natural contexts, about the potential impact of interventions that target
(e) functional assessment and positive behav- the social communication skills of children with ASD
ior support to address problem behaviors, younger than 3 years of age.
and (f) functional academic skills, when The vast majority of studies reviewed by the NRC
appropriate. (2001) included children from 3 to 5 years of age. Very
limited research is available on children with ASD under
Research conducted over the past 2 decades has age 3 years, primarily because the average age of diag-
identified core deficits in joint attention, shared affect, nosis for autistic disorder is 3.1 years and for ASD is
and conventional and symbolic aspects of communica- 3.9 years (Mandell, Novak, & Zubritsky, 2005). The re-
tion and play in young children with ASD. It has also sults from two studies support the application of the NRC
shown that these deficits are significant predictors of de- recommended practices for younger children. McGee,
velopmental outcomes (Mundy, Sigman, & Kasari, 1990; Morrier, and Daly (1999) reported on the outcomes for
Sigman & Ruskin, 1999; Stone, Ousley, Yoder, Hogan, 28 toddlers with ASD who entered a center-based in-
& Hepburn, 1997; Wetherby, Prizant, & Hutchinson, cidental teaching program at an average age of 2 years
1998; Wetherby, Watt, Morgan, & Shumway, in press; 5 months. The percentage of children who were verbal
Wetherby et al., 2004). These skills thus are important increased from 36% at program entry to 82% at pro-
intervention outcome measures for children with ASD. gram exit (mean age = 3 years 6 mos). Although these re-
Although a large number of studies have delineated core sults are promising, the lack of a control group does not
social communication deficits associated with ASD, very allow us to differentiate treatment effects from maturation
few of them have documented the effects of interventions and other confounding variables or to determine whether
on these deficits (NRC, 2001). In a multiple-baseline- this sample of children with ASD is representative of the
design study involving four participants with ASD, population as a whole. In a descriptive study of 35 tod-
Whalen and Schreibman (2003) demonstrated that dlers with ASD, Stone and Yoder (2001) found a strong,
changes in initiating and responding to joint attention positive association between the number of hours of
can be systematically taught using pivotal response train- speech therapy received between the ages of 2 and 3 years
ing. In a randomized-group-design study, Aldred, Green, and language skills at age 4 years. These intervention
and Adams (2004) implemented a monthly parent-training findings suggest that beginning an intervention before
program for 14 preschool children in their treatment 3 years of age may have a greater impact than beginning
group and compared the outcomes to those for 14 chil- it at a later age. Advances in earlier screening and diag-
dren in a community-treatment control group. Aldred et nosis add to the pressing need to develop early interven-
al. found significantly lower scores on the Autism Diag- tion programs that are (a) appropriate and effective for
nostic Observation Schedule (ADOS; Lord, Rutter, Di- very young children with ASD and (b) consistent with
Lavore, & Risi, 1999) scores, which indicated that there the requirements of the Individuals with Disabilities
were fewer autism symptoms in the treatment group, and Education Improvement Act of 2004 (IDEIA), Part C,
significantly better outcomes regarding parental positive which address the provision of services within natural
synchronous communication and rate of child commu- environments.
nicative acts; however, they found no significant differ-
ence between the two groups in the number of episodes
in which the parent and child shared attentional focus. INTERVENTION IMPLEMENTATION IN
Aldred et al.’s results suggest that significant gains in THE NATURAL ENVIRONMENT
some aspects of social communication can be achieved
by teaching parents how to enhance their children’s com- Earlier identification of and intervention for children of
munication in a cost-effective treatment, but gains in families who are seeking assistance through the public
joint attention may require a more intensive or different sector necessitates the provision of services within the
approach to intervention. “natural environment,” as defined by the IDEA (Walsh,

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


Early Social Interaction Project 69

Rous, & Lutzer, 2000). Interventions in the natural envi- 2001). Providing adequate, meaningful opportunities
ronment are approaches that maximize teaching and within their daily routines and activities without system-
learning throughout the day by using routines, materials, atic planning is very challenging for family members and
and people familiar to the family members and the child service providers. The use of multiple daily routines that
(Dunst, Hamby, Trivette, Raab, & Bruder, 2000; Woods, occur frequently, however, helps increase the amount of
Cripe, & Venn, 1997). Service providers support the intervention time across the day and can include multiple
caregivers’ use of naturalistic teaching strategies with the caregivers as social communication partners to enhance
child while undertaking their chosen activities, events, the generalization of learning.
and chores of daily life. For example, a parent who iden- Although research has established the effectiveness
tifies playing ball as a preferred routine would embed of parent-implemented interventions for children with a
practice of his or her son’s social communication goals variety of developmental disabilities, only a small num-
by placing a ball on a shelf out of reach and waiting for ber of these studies have included children with ASD
the child to look at or reach toward the ball. The parent (e.g., Girolametto, 1988; Kaiser, Hancock, & Nietfeld,
would model the word ball, wait for the child to repeat 2000; Kashinath, Woods, & Goldstein, in press; Koegel,
the word, and then take a turn throwing the ball and Bimbela, & Schreibman, 1996; Smith et al., 2000;
asking the child to throw it back, labeling it again while Woods, Kashinath, & Goldstein, 2004). Parents can
catching it. learn specific intervention techniques, such as modeling,
The basic tenets of intervention in natural environ- shaping, prompting, reinforcing, and fading, to teach spe-
ments reflect many of the active ingredients of the NRC cific language forms and functions to their children
(2001) recommendations including the following: (Charlop & Walsh, 1986), as well as “packaged” strategies,
such as incidental teaching or pivotal response training,
1. children learn functional and meaningful that promote communication (Kaiser et al., 2000; Koegel
skills; et al., 1996; Mahoney & Perales, 2005). Positive child
2. learning occurs within daily caregiving, outcomes subsequent to parent-implemented intervention
play, and social interactions with care- have also been documented (e.g., Kaiser et al., 2000;
givers that are repeated throughout the Koegel et al., 1996; Laski, Charlop, & Schreibman, 1988).
day; and Parental use of intervention strategies has promoted pos-
3. caregivers mediate the teaching and learn- itive change in a variety of child outcomes, such as an in-
ing process for the child as it occurs. creased frequency of verbalizations and spontaneous
speech (Laski et al., 1988), an increased use of target ut-
Recommended practice includes individualized, sys- terances (Kaiser et al., 2000), an increased percentage of
tematic embedding of intervention strategies by care- engagement and responsivity in target tasks, and a de-
givers within their typical daily routines and promotes creased frequency of disruptive behaviors (Koegel et al.,
participation, learning, and generalization of skills for 1996).
the child (Sandall, Hemmeter, Smith, & McLean, 2005). Although researchers have found multiple positive
This philosophy is compatible with many of the contem- outcomes for children with ASD as a result of parent
porary behavioral and developmental approaches used interventions, little is known about the long-term impact
for children with ASD (Dunlap & Fox, 1999; McGee et on either the child or the parent. Nevertheless, recent re-
al., 1999; Prizant & Wetherby, 2005; Schwartz, Sandall, search has shown the importance of teaching respon-
McBride, & Boulware, 2004) but is difficult to reconcile siveness to parents. In a longitudinal, descriptive study
with more traditional discrete trial interventions of 25 children with ASD, Siller and Sigman (2002) found
(Lovaas, 1987; Smith, Buch, & Gamby, 2000). that children (mean age = 50 months) whose parents
Intensity is an essential consideration for families showed higher levels of synchronization, defined as fol-
and service providers of children with ASD as they initi- lowing the child’s focus of attention and toy engagement,
ate the implementation of interventions within natural during initial play samples developed better joint atten-
environments. Services delivered by professionals in the tion skills 1 year later and better language outcomes 10
early intervention system average 2 to 3 hrs per week and 16 years later, compared with children of parents
(Scarborough et al., 2004), which is far below the inten- who showed lower levels of synchronization initially.
sity level recommended by the NRC. Furthermore, chil- The strongest predictor of the child’s increase in initi-
dren with ASD who participate in settings and activities ating joint attention was the parent’s initiation of
with other children should not be expected to learn simply joint attention that was synchronized to the child’s at-
by being there. Inclusive opportunities must also contain tentional focus. The strongest predictor of gain in lan-
adequate support for the child to learn from engagement guage was parent utterances that followed the child’s
with the materials, activities, other children, and care- attentional focus and allowed the child to continue the
givers in the environment (Strain, McGee, & Kohler, ongoing toy engagement.

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


70 Topics in Early Childhood Special Education 26:2

EARLY SOCIAL INTERACTION PROJECT is not yet using words but is beginning to use intentional
communication. His gestures consist of giving, reaching,
The Early Social Interaction (ESI) Project is an intervention and leading others by the hand. He rarely directs his gaze
program for toddlers who are at risk for ASD and their to others, he smiles but does not share positive affect,
families. It was developed as a model demonstration pro- and his rate of communication is very low. If he is not
ject funded by the Office of Special Education Programs of redirected, Joshua will spend much of his time pacing or
the U.S. Department of Education. The program designers rolling objects. Joshua’s family identified favorite play
incorporated the NRC (2001) recommendations within routines as listening to music in the family room and in
the context of a family-centered, natural-environments the car, swinging, and collecting and dropping small ob-
approach, which is considered recommended practice for jects. Although Joshua does not participate in many fam-
delivery of IDEA Part C services and supports (Sandall ily or caregiving routines, he does request cereal and
et al., 2005). The following major components of ESI juice and is comforted by rocking in his rocking chair or
(Woods & Wetherby, 2003) were relevant to this prelim- watching videos. He gets very distressed when an object
inary study: is removed or when activities cease. Initial goals for
Joshua are to shift his gaze between people and objects,
1. Routines-Based Intervention in Natural to request comfort when distressed, to display positive
Environments: The provision of support affect, to use a vocalization to gain attention, to imitate
and intervention occurs in the natural envi- familiar actions in a turn-taking sequence, to use famil-
ronments of the child and family to en- iar objects functionally, and to make choices with ges-
hance the child’s participation in everyday tures. Using a family-centered, natural-environment
routines and the family’s participation in approach to early intervention, the family members and
community activities. Each family identi- service provider(s) will discuss how Joshua’s social com-
fies routines and activities important to munication goals will be practiced within the routines as
them and their children that will become they occur throughout the day and what the parents will
the context for the intervention. ESI occurs do to support Joshua’s communicative development. In
in the home and in a community-based, this example, the parents learn how to use positioning to
parent–child playgroup. support Joshua’s participation; make activities pre-
2. Individualized Curriculum: The curriculum dictable with balanced turns; respond to Joshua’s ges-
is developed around the child’s unique tural, vocal, and facial signals to request and to stop
profile. A developmental framework that interactions; and wait for Joshua to take a turn or com-
targets social interaction, joint attention, municate a choice. Intervention strategies are embedded
communication, imitation, play, and emo- in cereal and drink routines in which he is given a choice
tional regulation guides the individualized of juice or water or of a type of cereal at breakfast and is
process (Prizant, Wetherby, & Rydell, given smaller portions so that he must vocalize to request
2000). The service provider and family more. Joshua’s father is usually responsible for breakfast
members complete curriculum-based as- while the mother drives older siblings to school. The
sessments to establish specific goals for the father pulls up his chair to face Joshua in his high chair,
child. The goals are reviewed for progress watches for Joshua to make his initial drink request,
and revised when accomplished. offers him the choices, and models the label of the
3. Parent-Implemented Intervention: Inter- choice that he makes (i.e., apple juice). He takes a turn
ventionists teach families to create learning drinking his coffee after Joshua takes a drink, again
opportunities and use specific intervention watches for Joshua, and responds. The process of em-
strategies to encourage practice of the child’s bedding goals occurs within each routine, as appropri-
goals within daily routines. Collaborative ate, and is jointly planned by the family members and
problem solving and consistent planning service provider(s).
by the service provider and family mem- The purpose of our preliminary study on ESI was to
bers allow the family members to take evaluate the effects of an ESI parent-implemented inter-
advantage of numerous teaching opportu- vention on social communication outcomes for a group
nities throughout the day and to feel confi- of parent–toddler dyads. We developed two research
dent about their abilities to engage the questions:
child in meaningful activities and to facili- 1. Would we find differences in social communica-
tate social communication skills. tion measures between preintervention and postinterven-
tion for a group of children with ASD who entered ESI
The following example illustrates an intervention in the second year of life and participated in ESI for a
that contains these components: Joshua, age 20 months, year?

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


Early Social Interaction Project 71

2. Would there be differences in the measures of so- 4. the family agreed to participate in a diag-
cial communication and language stage between the nostic evaluation when the child was
group of children who participated in ESI and a contrast 36 months old.
group of children who entered the early intervention sys-
tem at age 3 years? When they were 36 months old, all of the children
in both groups received a diagnostic evaluation com-
pleted by an interdisciplinary team consisting of a li-
METHOD censed speech–language pathologist and a psychologist
who were naive to the child’s group assignment. The
Participant Recruitment team used the following measures:
Seventeen children and their families participated in ESI
and formed the ESI group. We recruited these children 1. the Mullen Scales of Early Learning (MSEL;
from the ongoing longitudinal, prospective study of the Mullen, 1995), to determine nonverbal
FIRST WORDS® Project, which screens a general popu- and verbal developmental level;
lation sample of children under age 2 years to identify 2. the Vineland Adaptive Behavior Scales,
children at risk for ASD and other communication prob- Interview Edition, Survey Form (VABS;
lems (see Wetherby et al., 2004). Project researchers use Sparrow, Balla, & Cicchetti, 1984), to pro-
the Communication and Symbolic Behavior Scales Devel- vide an index of adaptive behavior; and
opmental Profile (CSBS DP; Wetherby & Prizant, 2002). 3. the ADOS, which is a semistructured, stan-
Children in the ESI group met the following selection cri- dardized assessment of communication,
teria: social interaction, and play or imaginative
use of materials and is a validated diagnos-
1. a CSBS DP behavior sample was first video- tic tool for ASD in children 2 years of age
taped when the child was between 12 and or older.
24 months of age and again when the child
was between 25 and 36 months of age and Based on the information gathered from the diag-
had participated in ESI for 12 months; nostic evaluation, the team made a best estimate diagno-
2. During the first behavior sample, the Sys- sis using the diagnostic criteria for Autistic Disorder or
tematic Observation of Red Flags of ASD Pervasive Developmental Disorder-Not Otherwise Speci-
(SORF; Wetherby et al., 2004) indicated fied (PDD-NOS) as defined in the fourth edition, text
significant red flags for ASD for that child, revision, of the Diagnostic and Statistical Manual of
and he or she received a provisional clini- Mental Disorders (DSM-IV-TR; American Psychiatric
cal diagnosis of ASD; Association, 2000). We use the term best estimate diag-
3. when the child was 24 months old or nosis here because the clinicians could make a judgment
younger, the family agreed to participate in about how to put the information together to make a di-
ESI for a period of at least 12 months; and agnosis. The team either did not make a diagnosis of
4. the family agreed to participate in a diag- Asperger syndrome or ruled it out due to the young age
nostic evaluation when the child was of these children. Children who received a diagnosis of
36 months old. Autistic Disorder or PDD-NOS and whose Communica-
tion and Social Interaction total score on the ADOS fell
We selected 18 children as the members of a third- at or above the cutoff for autism spectrum was included
year contrast group. These children ranged in age from in our study. The ADOS could not be completed for one
25 to 36 months of age and were recruited when they child because his family lived at a distance and could not
were first referred to community agencies serving chil- return for the follow-up evaluation. Because this child
dren with developmental delays because they were sus- had been given a diagnosis of autism at 36 months of age
pected of having ASD. The contrast group children met by a pediatric neurologist and had qualified for special
the following selection criteria: education in the public schools, we included him in the
study.
1. a CSBS DP behavior sample was first video- A total of 35 children participated in this study. A
taped when the child was between 25 and timetable with the ages of the children in each group at
36 months old; the time of the behavior samples and the follow-up diagnos-
2. the child and family had not received any tic evaluations is provided in Table 1. For the ESI group,
formal intervention services before the be- we videotaped two behavior samples, a pre-ESI sample in
havior sample was videotaped; the second year (M = 18.19 months, SD = 3.85) and a post-
3. the child was suspected of having ASD; and ESI sample in the third year (M = 30.72 months, SD =

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


72 Topics in Early Childhood Special Education 26:2

3.66). For the third-year contrast group, we videotaped and African American children, but the ESI group had
one behavior sample in the third year as soon as the child slightly more Hispanic children and the third-year con-
was recruited (M = 31.61 months, SD = 3.45). We found trast group had more Asian children. The two groups
no statistically significant group differences in age at the thus were fairly similar with regard to most demographic
third-year behavior samples or at the diagnostic evalua- aspects.
tions for the post-ESI group and the third-year contrast A summary of participant developmental character-
group. istics at the time of the follow-up diagnostic evaluations
at 3 years of age is presented in Table 3. Based on the
Mervis and Klein-Tasman (2004) criteria, the groups
Research Design were not matched on these developmental characteris-
In this study, we used a quasi-experimental, one-group, tics. The ESI group had significantly higher scores in re-
pretest–posttest design (Shadish, Cook, & Campbell, gards to most aspects of developmental level.
2002). We analyzed changes in pre- and postintervention
measures of social communication and, in an effort to
Intervention Procedures
strengthen this design, also studied a no-treatment con-
trast group. We collected social communication mea- Interventionists consisted of five research assistants—
sures from the contrast group at the same age as the four certified as speech–language pathologists and one
treatment group at posttest and compared the results early childhood education specialist. All five were expe-
with those for the treatment group. The contrast group rienced in early intervention with young children with
thus provides some information about the possible ef- ASD and their families, approved for Part C service de-
fects of maturation without treatment. Pretest measures livery in Florida, and trained specifically to implement
were not available for the contrast group, however, be- the parent intervention described. Each interventionist
cause these children were not identified at an earlier age; had a minimum of 3 years’ experience (range = 3–20
therefore, we could not determine whether the groups years) in early intervention with children and families in
were comparable at pretest. home- and center-based models.
ESI is a parent-implemented intervention model de-
signed to individualize social communication goals and
Participant Demographic and
monitor child progress, to identify family routines for tar-
Developmental Characteristics geted goals, to teach parents to implement teaching strate-
A summary of participant demographics is presented in gies within selected everyday routines, and to support
Table 2. Mervis and Klein-Tasman (2004) suggested that family implementation of intervention. The role of the
a probability value greater than .50 on tests of group dif- interventionists in this study was to teach parents to use a
ferences indicates that groups are well-matched on a variety of intervention strategies within daily routines to
variable and that a probability value less than .20 indi- increase opportunities to practice the social communica-
cates that groups are not matched on the variable. Based tion goals that were functional, predictable, and meaning-
on these criteria, the groups were well-matched on ful for the child and that were likely to occur throughout
mother’s education and age and on the Hollingshead the day. The interventionists also provided information
Index (Hollingshead, 1976), suggesting that the groups to address family concerns, such as challenging behav-
were comparable on socioeconomic status (SES). The iors or family acceptance of the diagnosis of autism.
groups were similar in terms of the percentage of boys Intervention sessions were scheduled to accommo-
and percentage of firstborn children. The groups were date family preferences, child attention, work/school
also similar in regards to the percentages of Caucasian schedules, and specific routines (e.g., dressing, snack and

TABLE 1. Timetable of Age in Months at Each Measure for the ESI and Third-Year Contrast Groups

ESI groupa 3rd-year contrast groupb


Pre-ESI Post-ESI ANOVA comparison
Measure M SD M SD M SD F p d

CSBS DP behavior sample 18.19 3.85 30.72 3.66 31.61 3.45 0.55 .463 0.25
Diagnostic evaluation 36.67 6.94 38.27 9.38 0.33 .569 −0.20

Note. ESI = Early Social Interaction Project; CSBS DP = Communication and Symbolic Behavior Scales Developmental Profile (Wetherby & Prizant, 2002);
F and p values are based on one-way ANOVA comparing ESI group post-ESI and third-year contrast group; Cohen’s d is a measure of effect size.
an = 17. bn = 18.

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


Early Social Interaction Project 73

TABLE 2. Demographic Characteristics of the ESI and Third-Year Contrast Groups, With Definition and Range of Scores
for Social Communication Measures

Child characteristic ESI groupa 3rd-year contrast groupb F p d

Boys (%) 88 78
First born (%) 41 44
Racial/ethnic group (%)
Caucasian 65 61
African American 12 17
Hispanic 24 6
Asian 0 17
Parent’s education (yrs.)
Mother
M 15.7 15.53 0.00 .949 0.10
SD 1.93 1.63
Father
M 16.35 15.41 0.64 .429 0.38
SD 2.50 2.40
Parent’s age at child’s birth
Mother
M 31.55 30.17 0.05 .821 0.28
SD 3.44 6.26
Father
M 33.26 34.04 0.71 .410 −0.13
SD 3.87 8.48
Hollinghead’s Index of SES (1976)
M 53.00 52.94 0.00 .987 0.01
SD 10.42 10.33

Note. ESI = Early Social Interaction Project; F and p values are based on one-way ANOVA comparing ESI group post-ESI and third-year contrast group;
Cohen’s d is a measure of effect size.
an = 17. bn = 18.

TABLE 3. Summary of Developmental Characteristics at the Diagnostic Evaluation at 3 Years of Age

ESI 3rd-year
groupa contrast groupb ANOVA comparison
Characteristic M SD M SD F p d

Mullen Scales of Early Learning


(1995)c
Nonverbal DQ 81.98 23.93 64.23 17.74 6.16 .019 0.85
Verbal DQ 73.55 26.16 56.07 24.96 4.08 .052 0.68

Vineland Adaptive Behavior Scales


(Sparrow et al., 1984)d
Communication 77.06 19.76 62.76 16.91 5.14 .030 0.78
Daily living 75.88 10.51 65.88 19.53 3.46 .075 0.67
Social 77.06 16.35 64.41 18.99 4.33 .046 0.72
Motor 83.76 14.32 71.88 21.45 3.61 .068 0.66

Adaptive behavior composite 73.82 14.73 64.71 7.20 5.26 .031 0.83

Note. ESI = Early Social Interaction Project; F and p values are based on one-way ANOVA comparing ESI group post-ESI and third-year contrast group;
Cohen’s d is a measure of effect size.
an = 17. bn = 18. cDevelopmental quotients (DQ) based on age equivalent divided by chronological age multiplied by 100. dStandard scores based on a

mean of 100 and a standard deviation of 15.

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


74 Topics in Early Childhood Special Education 26:2

meals, outdoor play, games, and songs) identified for in- modate the interventionists’ and parents’ preferred teach-
tervention. We recommended two home visits per week ing and learning styles, respectively. Parents could choose
to facilitate parent implementation and to monitor child from easy-to-read handouts, videos, or demonstrations
progress. The number of home visits accomplished dur- of specific strategies and examples of their use in family-
ing a child’s year of intervention varied, primarily due to identified routines. Interventionists either modeled the
family circumstances (e.g., birth of new siblings, vaca- strategies within the routines and provided opportunities
tions, family illnesses). The interventionists offered make- for parents to ask clarifying questions or provided op-
up sessions to maintain intensity. They conducted the portunities for guided practice in which the parent im-
interventions throughout the family’s home and yard, plemented the intervention and the interventionist
with the specific locations based on the specific goals to observed and supported the parent as needed. Parents
be addressed and the family’s choice of daily routines. and interventionists met to discuss strategy use and pos-
The interventionists frequently followed the child and sible barriers to implementation and to identify addi-
parent to encourage participation in a variety and se- tional instances of potential strategy use across routines.
quence of routines and activities as they occurred. The The parents had opportunities to determine the easiest-
interventionists used the strategy of joining the child and to-use methods of collecting data on the child’s goals and
family in multiple routines and activities in their se- their implementation of the strategies within routines.
quence of occurrence to increase the number of routines We solicited parent input regarding types of methods for
with embedded intervention and to promote generaliza- parent teaching to increase the parents’ satisfaction and
tion for the parents. ownership of the intervention. Each week, the interven-
A key component of ESI was the selection of rou- tionists observed the parents’ implementing the interven-
tines that were preferred by the family, occurred at a high tion within the routines selected. If the parent did not
frequency, were motivating to child and parent, and in- embed the goal or use the intervention strategy, the inter-
cluded multiple opportunities to embed social communi- ventionist would model and review the training and
cation outcomes as a method of increasing sufficiency of problem solve about what was working and what was
practice. Once the interventionists had identified rou- challenging for the parent.
tines, they used objects, equipment, and materials that We monitored treatment fidelity through the com-
typically belonged to the routine and its particular con- pletion of a self-assessment checklist to ensure that each
text. For example, indoor play routines included the child’s home visit included the following components:
favorite toys, music, or constructive materials, whereas
outdoor play materials consisted of the child’s ball, • initial review of child and family outcomes
swingset, or scooter. The interventionists ensured that and intervention intensity since the previous
the materials the child and family typically used were in- visit;
cluded in the routines. • observation of and participation by the
Another key ingredient was the careful selection of parent in multiple routines;
social communication goals. Goals were identified as pri- • guided practice in implementing the interven-
orities for the parent, were developmentally appropriate tion for specific goals within the routines;
for the child, and had been marked as early red flags of • discussion of what, when, how often, and
ASD. The interventionist systematically introduced each where the interventions would occur
parent to evidence-based, naturalistic teaching strategies between visits; and
that were appropriate for the child’s identified social com- • development of a plan for the next visit.
munication goals. We selected naturalistic intervention
strategies to improve the parent’s ability to encourage the In addition to the home-based sessions, each child
child to initiate social communication and follow the participated in a parent–child playgroup guided by two
child’s attentional focus. These strategies included environ- interventionists through the FIRST WORDS® Project for
mental arrangement, waiting, use of natural reinforcers, up to 12 parent–child dyads, with a blend of children
balanced turn-taking, modeling, contingent imitation, re- with typical development and children with developmen-
questing imitation, and time delay (Hepting & Gold- tal delays. The children in the playgroups ranged in age
stein, 1996). The strategies were individualized based on from 18 months to 27 months. Each parent–child dyad
the parent’s observed skill level, evidence that the strat- attended the playgroups for 9 weeks during the year that
egy use would be appropriate to influence their child’s they participated in ESI. The playgroups were held in a
communication goals, and parent indication that the 600-square-foot room organized into play centers: music
strategy could be implemented multiple times within circle, books, dress up, doctor’s office, baby washing,
routines (Kashinath et al., in press). baby beds, blocks, trucks, toys, kitchen and feeding area,
We individualized the methods for teaching parents and fluid play activities. Each playgroup was organized
to embed interventions within their routines to accom- around predictable opening and closing songs, book

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


Early Social Interaction Project 75

times, and new play centers each week. The intervention- Child Measure of Language Stage
ists provided information and a handout about specific
weekly topics. In addition to the weekly topics, the fami- We used the CSBS DP behavior sample criteria for lan-
lies had opportunities to focus on interacting with their guage stage established by Wetherby and Prizant (2003)
child, to support peer interaction with other participants, to determine the child’s language stage at the time of
and to receive feedback on their strategy use. The inter- each behavior sample, based on expressive language
ventionists provided information to the families through used. Those criteria were as follows: Preverbal—0 or 1
discussion and handouts, responses to questions, model- word; Early one-word stage—2 to 5 different words;
ing, and individual coaching by setting up opportunities Late one-word stage—6 to 9 different words; Multiword
and making suggestions. The topics included how and stage—10 or more different words and 2 or more differ-
why young children communicate, responding to your ent word combinations.
child’s communication, daily routines and games, games
for young children, learning through the senses, sharing Interrater Reliability
books with your child, communication and behavior in
toddlers, and social interaction and play with peers. The We calculated interrater reliability for the CSBS DP be-
playgroups offered the families opportunities to observe havior sample by using generalizability (g) coefficients
their child in a group with same-age peers and to net- for pairs of four independent raters on randomly selected
work with other parents of children with and without videotapes of the behavior sample for at least 20% of the
ASD, which may be helpful to parents in adjusting to the samples scored by each rater. A g coefficient approaches
realization that their child has ASD. 1 as the variance accounted for by the participants is
larger in comparison with the variance accounted for by
raters. G coefficients that are at least .60 are considered
Child Measures of Social acceptable for demonstrating interrater reliability (Mit-
Communication chell, 1979). Rater 1 scored the largest number of sam-
We used standard sampling procedures (Wetherby & ples and therefore was considered the expert to which
Prizant, 2002) to derive measures of the child’s use of so- the other raters were compared. The g coefficients for
cial communication from the CSBS DP behavior sample. the social communication measures ranged from 0.76 to
A parent was present during the full evaluation and was 1.00, with an average of 0.90 for Raters 2 and 3 and
instructed to respond naturally but not to direct the child’s 0.92 for Rater 4, indicating that the raters exhibited high
behavior, to encourage spontaneous communication and interrater reliability.
play. The evaluation session began with a warm-up of
about 10 min and lasted 30 to 40 min. The intervention-
ist first presented the child with a series of communica- RESULTS
tive temptations (a windup toy, a balloon, bubbles, a jar
Within-Group Differences in
with food, a bag with toys, and books designed for young
children) to entice spontaneous communication and then Measures of Social Communication
gave the child a feeding toy set and stuffed animal for Our first research question was whether there were dif-
symbolic play and blocks for constructive play. The sam- ferences in measures of social communication from pre-
ple consisted of six activities during which the interven- to postintervention in the ESI group. Descriptive statis-
tionist rated the child’s skills, and also included probes tics on each of the social communication measures from
of gaze/point follow and comprehension of object names, the behavior samples collected for the ESI group at pre-
person names, and body parts that were interspersed be- and postintervention are presented in Table 5. We con-
tween activities. One of four trained examiners who ducted a series of repeated-measures analyses of variance
were blind to the child’s diagnosis videotaped the behav- (ANOVAs) to evaluate within-group differences from
ior sample and scored it by using the standard CSBS DP pretest to posttest. We calculated Cohen’s d statistic as
procedures. Information about the reliability and valid- the effect size index by dividing the mean of the paired
ity of the CSBS DP has been reported in Wetherby, Allen, differences by the average of the standard deviations for
Cleary, Kublin, and Goldstein (2002); Wetherby, Gold- the pre- and postintervention measures. The results are
stein, Cleary, Allen, and Kublin (2003); and Wetherby presented in the far right column of Table 5. Regardless
and Prizant (2002). of sign, d values of .2, .5, and .8 represent small, medium,
For this study, we examined individually the raw and large effect sizes, respectively (Cohen, 1988). We
scores for individual items of the behavior sample to found significant within-group differences with large ef-
study five constructs of social communication measured fect sizes from pretest to posttest for 11 of the 13 social
with the CSBS DP, which are defined in Table 4, along communication measures. Two social communication
with the range of possible scores for each measure. measures—gaze shifts and shared positive affect—showed

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


76 Topics in Early Childhood Special Education 26:2

TABLE 4. Definition and Range of Scores for Each Social Communication Measure

Social communication
measure Definition adapted from Wetherby & Prizant (2002) Range

Social signals
Gaze shifts Social referencing defined as a 3-point alternating gaze shift between a person and 0–6
an object
Shared positive affect Clear facial expression of pleasure or excitement that is directed with gaze to another 0–6
person
Gaze/point follow Ability to follow another person’s gaze and point at a distance (also referred to as 0–2
responding to joint attention)

Rate of communicating The child’s rate of vocal or gestural communicative signals that serve a communi- 0–18
cative function

Communicative functions
Behavior regulation Vocal or gestural communicative signals to regulate another person’s behavior to 0–6
request or protest an object or action
Social interaction Vocal or gestural communicative signals to draw another person’s attention to oneself 0–6
to get the other person to look at, notice, or comfort him or her
Joint attention Vocal or gestural communicative signals to direct another person’s attention to an 0–6
object or event to get the other person to look at or notice something of interest
Communicative means
Inventory of gestures Number of different conventional gestures (e.g., give, show, point, reach, wave) used 0–8
Inventory of consonants Number of different consonants (e.g., b, d, g, m, n) used 0–10
Inventory of words Number of different words used referentially and that approximate conventional 0–16
words, spoken or signed
Symbolic capacity
Understanding Ability to understand object names, person names, and body parts without gestural cues 0–8
Inventory of actions Number of different conventional actions with objects used 0–12
Actions to other Number of pretend actions with objects toward other 0–6

differences with moderate effect sizes, which were non- between the ESI group pre-ESI and the third-year con-
significant trends. trast group. A comparison of the social communication
measures for the ESI group pre-ESI and the third-year
contrast group are presented in Table 7. We conducted a
Across-Group Differences in series of univariate ANOVAs, using a Welch correction
Measures of Social Communication for lack of homogeneous variances, to evaluate group
The second research question was whether there were differences. The third-year contrast group had signifi-
group differences in measures of social communication cantly better performance, with moderate to large effect
between the ESI group post-ESI and the third-year contrast sizes on all three of the measures of communicative
group. A group comparison of the social communication means and on actions to others in play. We did not find
measures is presented in Table 6. We conducted a series any statistically significant differences between the third-
of univariate ANOVAs, using a Welch correction for lack year contrast group and ESI group pre-ESI on the three
of homogeneous variances, to evaluate group differences. measures of social signals, rate of communicating, the
The ESI group had significantly better performance with three measures of communicative functions, understand-
large effect sizes on the three measures of social signals, ing, and inventory of actions.
rate of communicating, the three measures of communica-
tive functions, and understanding. We did not find any
Differences on Language Stage
statistically significant differences on the three measures
of communicative means and both measures of play. The last research question was whether there were
The third research question was whether there were within- and across-group differences on the measure of
group differences in measures of social communication language stage. The number and percentage of children
Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015
Early Social Interaction Project 77

TABLE 5. Social Communication Measures for the ESI Group Pre- and Post-ESI

Pre-ESI Post-ESI ANOVA comparisona


Measure M SD M SD F p d

Chronological age (in mos.) 18.19 3.85 30.72 3.66


Social signals
Gaze shifts 4.18 1.67 5.18 1.51 3.24 .091 0.63
Shared positive affect 1.88 1.83 3.41 2.00 4.35 .053 0.80
Gaze/point follow 0.35 0.61 1.29 0.85 11.51 .004 1.29
Rate of communicating 8.71 5.36 16.18 2.56 24.88 < .001 1.89
Communicative functions
Behavior regulation 3.59 2.12 5.53 0.63 12.98 .002 1.41
Social interaction 0.65 0.93 2.29 1.86 14.22 .002 1.18
Joint attention 0.65 0.93 2.94 2.16 18.00 .001 1.48
Communicative means
Inventory of gestures 1.76 1.15 3.76 1.68 15.54 .001 1.41
Inventory of consonants 1.00 1.37 4.71 2.69 32.90 < .001 1.83
Inventory of words 0.18 0.53 6.00 5.18 21.23 < .001 2.04
Symbolic capacity
Understanding 0.82 1.38 4.24 3.29 22.92 < .001 1.46
Inventory of actions 3.88 2.00 6.18 2.90 8.16 .011 0.94
Actions to others 0.53 0.80 2.82 1.70 32.89 < .001 1.83

Note. ESI = Early Social Interaction Project; F and p values are based on repeated-measures ANOVA comparing pre-ESI and post-ESI; Cohen’s d is a mea-
sure of effect size.
an = 17.

TABLE 6. Social Communication Measures for the ESI Group Post-ESI and the Third-Year Contrast Group

3rd-year ANOVA
Post-ESI groupa contrast groupb comparison
Measure M SD M SD F p d

Child chronological age (in mos.) 30.72 3.66 31.61 3.45


Social signals
Gaze shifts 5.18 1.51 3.28 1.71 12.17 .001 1.18
Shared positive affect 3.41 2.00 1.31 1.38 9.50 .005 1.24
Gaze/point follow 1.29 0.85 0.61 0.70 6.72 .014 0.88
Rate of communicating 16.18 2.56 11.94 5.51 8.64 .007 1.05
Communicative functions
Behavior regulation 5.53 0.63 4.11 1.81 9.80 .005 1.16
Social interaction 2.29 1.86 0.89 1.28 6.70 .015 0.89
Joint attention 2.94 2.16 1.39 1.54 5.92 .021 0.84
Communicative means
Inventory of gestures 3.76 1.68 2.72 1.53 3.68 .064 0.65
Inventory of consonants 4.71 2.69 3.72 3.08 1.02 .321 0.34
Inventory of words 6.00 5.18 3.94 4.40 1.59 .216 0.43
Symbolic capacity
Understanding 4.24 3.29 2.00 2.40 5.22 .030 0.79
Inventory of actions 6.18 2.90 5.00 2.28 1.77 .193 0.46
Actions to others 2.82 1.70 2.56 1.72 0.21 0.65 0.15

Note. ESI = Early Social Interaction Project; F and p values are based on one-way ANOVA comparing ESI group post-ESI and Third-year contrast group;
Cohen’s d is a measure of effect size.
an = 17. bn = 18.

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


78 Topics in Early Childhood Special Education 26:2

at each language stage for the ESI group at pre- and .053). We detected a moderate difference (d = .63) for
postintervention and for the third-year contrast group gaze shifts, which also was a nonsignificant trend (p =
are presented in Table 8. The percentage of children who .091). A ceiling effect for the ESI group post-ESI may
were verbal (i.e., early one-word, late one-word, and have accounted for the moderate effect size and lack of
multiword stages) in the ESI group at preintervention (M significant improvement on gaze shifts, because the mean
age = 18.19 mos) was 5.9% and increased to 76.5% of was 5.18 (SD = 1.51) and the maximum score was 6.
children at postintervention (M age = 30.72 mos), com- Improved performance on social communication mea-
pared to 55.6% in the third-year contrast group (M age sures, particularly on the measures related to joint atten-
= 31.61 mos). tion (i.e., gaze shifts, shared positive affect, gaze/point
follow, and communicating for joint attention) has not
been documented in previous research on parent-imple-
DISCUSSION mented interventions for children with ASD. This result
is promising. In a less-intensive parent-implemented in-
These findings expand on the previous research on par- tervention, Aldred et al. (2004) were not able to demon-
ent-implemented early intervention to a group of young strate improvements in parent-child shared attentional
children with ASD in three critical ways. First, the study focus.
delineates the specific social communication outcomes By examining specific parameters of social communi-
achieved during intervention beginning in the second cation for children receiving the intervention and compar-
year of life. Comparison of the social communication ing them to children who did not receive the intervention,
skills of the study children before and after their parents we can begin a preliminary exploration of the impact of
participated in ESI indicated large, significant differences the intervention on specific social communication out-
on all parameters measured by the CSBS DP except comes. Although it was not possible to know whether
shared positive affect and gaze shifts. We found a large the ESI group was comparable to the third-year contrast
difference (d = .80) for shared positive affect from pre- group during the second year of life, the pattern of dif-
to post-ESI; however, this was a nonsignificant trend (p = ferences across groups is interesting and may be sugges-

TABLE 7. Social Communication Measures for the ESI Group Pre-ESI and the Third-Year Contrast Group

3rd-year ANOVA
Pre-ESI groupa contrast groupb comparison
Measure M SD M SD F p d

Child chronological age (in mos.) 18.19 3.85 31.61 3.45


Social signals
Gaze shifts 4.18 1.67 3.28 1.71 2.48 .125 0.53
Shared positive affect 1.88 1.83 1.31 1.38 0.24 .626 0.36
Gaze/point follow 0.35 0.61 0.61 0.70 1.37 .250 −0.40
Rate of communicating 8.71 5.36 11.94 5.51 3.10 .087 −0.59
Communicative functions
Behavior regulation 3.59 2.12 4.11 1.81 0.61 .440 −0.30
Social interaction 0.65 0.93 0.89 1.28 0.41 .525 −0.20
Joint attention 0.65 0.93 1.39 1.54 3.01 .094 −0.60
Communicative means
Inventory of gestures 1.76 1.15 2.72 1.53 4.43 .043 −0.70
Inventory of consonants 1.00 1.37 3.72 3.08 11.61 .002 −1.20
Inventory of words 0.18 0.53 3.94 4.40 13.01 .002 −1.50
Symbolic capacity
Understanding 0.82 1.38 2.00 2.40 3.20 .085 −0.60
Inventory of actions 3.88 2.00 5.00 2.28 2.39 .130 −0.50
Actions to others 0.53 0.80 2.56 1.72 20.28 < .001 −1.60

Note. ESI = Early Social Interaction Project; F and p values are based on one-way ANOVA comparing ESI group pre-ESI and third-year contrast group;
Cohen’s d is a measure of effect size.
an = 17. bn = 18.

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


Early Social Interaction Project 79

tive of the role of maturation on improved outcomes. tions. In other words, maturation alone may result in
Compared to the ESI group members post-ESI, who were gains in communicative means and aspects of symbolic
the same age, the third-year contrast group members dis- play, which are often the focus of traditional interven-
played significantly poorer skills on the measures of so- tions for children with ASD. However, maturation does
cial signals, rate of communicating, communicative not appear to influence the other social communication
functions, and understanding. In contrast, they performed measures, which are core deficits of ASD. Collectively
similar to the ESI group post-ESI on inventory of ges- these findings support the effectiveness of early interven-
tures, consonants, and words and on measures of sym- tion on the social communication skills of toddlers with
bolic play. These similarities in the two groups suggest ASD and suggest that intervention beginning in the second
that the improvements that the ESI children showed in year of life can have a positive impact on core social com-
communicative means and symbolic play may have been munication deficits of ASD. This supports the recom-
due to maturation. The differences between the ESI mendations of the NRC (2001) for earlier identification
group post-ESI and the third-year entry group suggest and intervention targeting core social communication
that children’s outcomes in social signals, rate of com- skills, which is particularly important because the social
municating, communicative functions, and understand- communication skills in which the ESI group showed im-
ing may have been affected by participation in ESI. provement are strong predictors of language outcomes
Comparison of the third-year contrast group mem- in children with ASD (e.g., Sigman & Ruskin, 1999).
bers with the ESI group members pre-ESI, who were a It has been reported that 50% of children with ASD
year younger, further supports the role of maturation remain nonverbal by middle childhood (Tager-Flusberg,
and impact of intervention on the ESI children. The Paul, & Lord, 2005); however, this percentage is smaller
third-year contrast group performed similar to the ESI among children who received early intervention. In this
group pre-ESI on social signals, rate of communicating, study, the percentage of ESI group children who were
communicative functions, understanding, and inventory nonverbal at preintervention at 1.5 years of age was
of actions in play; however, they performed significantly 94.1%. This percentage decreased to 23.5% postinter-
better on communicative means and pretend play ac- vention at 2.5 years of age, compared to 44.4% of the

TABLE 8. Language Stage for the ESI Group Pre- and Post-ESI and the Third-Year
Contrast Group

ESI groupa
3rd-year
Measure Pre-ESI Post-ESI contrast groupb

Chronological age (in mos.)


M 18.19 30.72 31.61
SD 3.85 3.66 3.45
Preverbal
No. 16 4 8
% 94.1 23.5 44.4
Verbal
Early one-word
No. 1 4 5
% 5.9 23.5 27.8
Late one-word
No. 0 6 2
% 0.0 35.3 11.1
Multiword
No. 0 3 3
% 0.0 17.7 16.7
Total
No. 1 13 10
% 5.9 76.5 55.6

Note. ESI = Early Social Interaction Project.


an = 17. bn = 18.

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


80 Topics in Early Childhood Special Education 26:2

third-year contrast group members, who were also 2.5 within a local, early intervention, service delivery system.
years of age. Consistent with other early intervention The actual intensity of the intervention that the ESI chil-
studies of toddlers with ASD (Lord, Risi, & Pickles, dren received from their parents and the parents’ accuracy
2004; McGee et al., 1999), these findings are encourag- in using the strategies was not documented in this study.
ing in that larger percentages of future generations of Future research should be conducted to provide further
children with ASD are likely to acquire functional lan- support for the effectiveness of parent-implemented in-
guage. terventions for very young children with ASD. These
Finally, this study offers preliminary evidence that findings do support the need to conduct a randomized-
ESI is a feasible model for providing Part C early inter- group-design study in which pretest measures are avail-
vention services to young children with ASD and their able and comparable groups can be documented in regard
families through the use of a family-centered developmen- to social communication and development level before in-
tal framework. Utilizing professional time to teach par- tervention. Because it is not ethical to withhold interven-
ents how to embed specific social communication skills tion, future studies could compare parent-implemented
in everyday routines, activities, and places is a cost-effi- interventions with clinician-implemented interventions.
cient early intervention method for children with ASD Documenting the intensity of parent implementation of
and is consistent with the regulations of IDEA Part C re- strategies and investigating whether intensity is a moder-
garding the provision of an intervention in the natural ator variable on child-outcome measures are also impor-
environment. In addition, parent-implemented interven- tant goals for future research. Finally, studies should
tion is a service delivery option that may be able to pro- examine whether child outcomes from parent-implemented
vide the intensity of active engagement recommended for interventions maintain over time and generalize across
children with ASD, at least for parents who are success- settings. It is possible that an intervention implemented
ful in learning how to use these strategies and are able to by the parents very early in the child’s development will
apply them across a variety of routines with adequate in- lead to better maintenance and generalization than a
tensity. The strategies taught in ESI provided the parents clinician-implemented intervention if the parents con-
with a repertoire of ways to follow the child’s focus of tinue to follow the child’s attentional focus.
attention and build social communication skills within The differences in social communication measures
their typical daily routines, thus enhancing opportunities from pre- to postintervention for the ESI group may be
for generalization (Kashinath et al., in press). Although more notable in that studies have found that the second
the parents were not required to document the time spent year of life has been associated with developmental regres-
on a daily basis implementing the intervention within sion in language and communication and other develop-
routines as a component of the project, each family was mental domains for children with ASD (21%–48%;
required to identify multiple routines for embedding in- Davidovitch, Glick, Holtzman, Tirosh, & Safir, 2000;
tervention at the initial interview and continued to iden- Goldberg et al., 2003; Lainhart et al., 1997; Lord,
tify and practice additional or different routines with the Shulman, & DiLavore, 2004; Rogers, 2004) and occurs
interventionist during weekly sessions. They also dis- most often between the ages of 18 months and 21 months
cussed the child’s progress on the identified goals within (Goldberg et al., 2003; Shinnar et al., 2001). The findings
the routines and participated in the planning process for on developmental regression are particularly interesting
the upcoming week with the interventionist. in light of recent brain research studies suggesting that
The findings of Siller and Sigman (2002) support an overgrowth of the brains of children with ASD occurs
the critical importance of teaching parents to use strate- over the first few years of life (Courchesne et al., 2001;
gies that increase responsiveness to their child’s interac- Lainhart et al., 1997; Piven, Arndt, Bailey, & Andreasen,
tional focus. Although it is beyond the scope of this study 1996; Sparks et al., 2002). Mundy and Burnette (2005)
to measure changes in the parents’ interaction strategies, posited that the early social orienting and joint attention
the findings of Siller and Sigman suggest that parents’ deficits of children with ASD “lead to a critical impover-
ability to support shared attention may be a mediating ishment in the first years of social information input,
variable affecting children’s gains in social communica- which contributes to the course of atypical neurodevel-
tion and language stage. opment in autism” (p. 661). This leads to the following
The limitations of this study are inherent in the quasi- question: Can very early intervention prevent or attenu-
experimental design. As noted previously, we could not ate the subsequent cascading effect on neurodevelopment
ascertain that the groups were matched at age 2 years. and behavioral outcomes that potentially arise from
The sample size is relatively small for a group design, but early impoverished social interactions? This study pro-
this is understandable, given the age of the children and vides preliminary evidence suggesting that intervention
that ASD is a low-incidence disability. The intervention beginning in the second year of life may have an effect
was designed not for research purposes but rather for on social communication and a secondary effect on de-
demonstration of a family-centered, developmental model velopmental outcome. ◆

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


Early Social Interaction Project 81

AUTHORS’ NOTES dren with autism. Journal of Speech, Language, and Hearing Re-
search.
1. This research was supported in part by three grants from the U.S. Koegel, R. L., Bimbela, A., & Schreibman, L. (1996). Collateral effects
Department of Education, Office of Special Education and of parent training on family interactions. Journal of Autism and
Rehabilitation Services (H324M010071, H324C020091, and Developmental Disorders, 26, 347–359.
H324C030112). Lainhart, J., Piven, J., Wzorek, M., Santangelo, S., Coon, H., &
2. The authors would like to thank the families who gave their time to Folstein, S. (1997). Macrocephaly in children and adults with
participate in this project. autism. Journal of the American Academy of Child and Adolescent
Psychiatry, 36, 282–289.
REFERENCES
Laski, K., Charlop, M., & Schreibman, L. (1988). Training parents to
Aldred, C., Green, J., & Adams, C. (2004). A new social communica- use the natural language paradigm to increase their autistic chil-
tion intervention for children with autism: Pilot randomized con- dren’s speech. Journal of Applied Behavior Analysis, 21, 391–400.
trolled treatment study suggesting effectiveness. Journal of Child Lord, C., Risi, S., & Pickles, A. (2004). Trajectory of language devel-
Psychology and Psychiatry and Allied Disciplines, 45, 1420–1430. opment in autism spectrum disorders. In M. Rice & S. Warren
American Psychiatric Association. (2000). Diagnostic and statistical (Eds.), Developmental language disorders: From phenotypes to eti-
manual of mental disorders (4th ed., Text rev.). Washington, DC: ologies (pp. 7–29). Mahwah, NJ: Erlbaum.
Author. Lord, C., Rutter, M., DiLavore, P., & Risi, S. (1999). Autism diagnos-
Charlop, M., & Walsh, M. (1986). Increasing autistic children’s daily tic observation schedule–Generic. Los Angeles: Western Psycho-
spontaneous speech. Journal of Applied Behavior Analysis, 19, 307– logical Services.
314. Lord, C., Shulman, C., & DiLavore, P. (2004). Regression and word
Cohen, J. (1988). Statistical power analysis for the behavioral sciences loss in autistic spectrum disorders. Journal of Child Psychology and
(2nd ed.). Hillsdale, NJ: Erlbaum. Psychiatry and Allied Disciplines, 45, 936–955.
Courchesne, E., Karns, C., Davis, H., Ziccardi, R., Carper, R., Tigue, Z., Lovaas, I. (1987). Behavioral treatment and normal educational and
et al. (2001). Unusual brain growth patterns in early life in patients intellectual functioning in young autistic children. Journal of Con-
with autistic disorder: An MRI study. Neurology, 57, 245–254. sulting and Clinical Psychology, 55, 3–9.
Davidovitch, M., Glick, L., Holtzman, G., Tirosh, E., & Safir, M. P. Mahoney, G., & Perales, F. (2005). Relationship-focused early inter-
(2000). Developmental regression in autism: Maternal perception. vention with children with pervasive developmental disorders and
Journal of Autism and Developmental Disorders, 30, 113–119. other disabilities: A comparative study. Developmental and Behav-
Dawson, G., & Osterling, J. (1997). Early intervention in autism. In ioral Pediatrics, 26, 77–85.
M. Guralnick (Ed.), The effectiveness of early intervention (pp. 307– Mandell, D., Novak, M., & Zubritsky, C. (2005). Factors associated
326). Baltimore: Brookes. with age of diagnosis among children with autism spectrum disor-
Dunlap, G., & Fox, L. (1999). Supporting families of young children ders. Pediatrics, 116, 1480–1486.
with autism. Infants and Young Children, 12(2), 48–54. McGee, G., Morrier, M., & Daly, T. (1999). An incidental teaching ap-
Dunst, C., Hamby, D., Trivette, C., Raab, M., & Bruder, M. B. (2000). proach to early intervention for toddlers with autism. The Journal
Everyday family and community life and children’s naturally occur- of the Association for Persons with Severe Handicaps, 24, 133–
ring learning opportunities. Journal of Early Intervention, 23, 151– 146.
164. Mervis, C. B., & Klein-Tasman, B. P. (2004). Methodological issues in
Fenske, E., Zalenski, S., Krantz, P., & McClannahan, L. (1985). Age at group-matching designs: Alpha levels for control variable compar-
intervention and treatment outcome for autistic children in a com- isons and measurement characteristics of control and target vari-
prehensive intervention program. Analysis and Intervention for ables. Journal of Autism and Developmental Disorders, 34(1), 7–
Developmental Disabilities, 5, 49–58. 17.
Girolametto, L. (1988). Improving the social-conversational skills of Mitchell, S. K. (1979). Interobserver agreement, reliability, and gener-
developmentally delayed children: An intervention study. The Jour- alizability of data collected in observational studies. Psychological
nal of Speech and Hearing Disorders, 53, 156–167. Bulletin, 86, 376–390.
Goldberg, W. A., Osann, K., Filipek, P. A., Laulhere, T., Jarvis, K., Mullen, E. (1995). The Mullen scales of early learning. Circle Pines,
Modahl, C., et al. (2003). Language and other regression: Assess- MN: American Guidance.
ment and timing. Journal of Autism and Developmental Disorders, Mundy, P., & Burnette, C. (2005). Joint attention and neurodevelop-
33, 607–616. mental models of autism. In F. Volkmar, R. Paul, A. Klin, & D.
Harris, S., & Handleman, J. (2000). Age and IQ at intake as predictors Cohen (Eds.), Handbook of autism and pervasive developmental
of placement for young children with autism: A four- to six-year disorders (3rd ed., pp. 650–681). New York: Wiley.
follow-up. Journal of Autism and Developmental Disorders, 30, Mundy, P., Sigman, M., & Kasari, C. (1990). A longitudinal study of
137–142. joint attention and language development in autistic children.
Hepting, N., & Goldstein, H. (1996). What’s natural about naturalis- Journal of Autism and Developmental Disorders, 20, 115–128.
tic language intervention? Journal of Early Intervention, 20, 249– National Research Council. (2001). Educating children with autism.
265. Washington, DC: National Academy Press, Committee on Educa-
Hollingshead, A. B. (1976). Four factor index of social status. New tional Interventions for Children with Autism, Division of Behav-
Haven, CT: Yale University Press. ioral and Social Sciences and Education.
Individuals with Disabilities Improvement Act of 2004, 20 U.S.C. § Piven, J., Arndt, S., Bailey, J., & Andreasen, N. (1996). Regional brain
1400 et seq. (2004)(reauthorization of the Individuals with Dis- enlargement in autism: A magnetic resonance imaging study. Jour-
abilities Education Act of 1990) nal of the American Academy of Child and Adolescent Psychiatry,
Kaiser, A., Hancock, T., & Nietfeld, J. (2000). The effects of parent- 35, 530–536.
implemented enhanced milieu teaching on the social communication Prizant, B., & Wetherby, A. (2005). Critical issues in enhancing com-
of children who have autism. Early Education and Development, munication abilities for persons with autism spectrum disorders. In
11, 423–446. F. Volkmar, R. Paul, A. Klin, & D. Cohen (Eds.), Handbook of
Kashinath, S., Woods, J., & Goldstein, H. (in press). Enhancing gener- autism and pervasive developmental disorders (3rd ed., pp. 925–
alized teaching strategy use in daily routines by caregivers of chil- 945). New York: Wiley.

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015


82 Topics in Early Childhood Special Education 26:2

Prizant, B., Wetherby, A., & Rydell, P. (2000). Issues in enhancing com- autism in early intervention environments. In M. J. Guralnick (Ed.),
munication and related abilities for young children with autism Early childhood inclusion: Focus on change (pp. 337–363). Balti-
spectrum disorders: A developmental transactional perspective. In more: Brookes.
A. Wetherby & B. Prizant (Eds.), Autism spectrum disorders: A Tager-Flusberg, H., Paul, R., & Lord, C. (2005). Language and com-
transactional developmental perspective (pp. 193–214). Baltimore: munication in autism. In F. Volkmar, R. Paul, A. Klin, & D. Cohen
Brookes. (Eds.), Handbook of autism and pervasive developmental disorders
Rogers, S. (2004). Developmental regression in autism spectrum disor- (3rd ed., pp. 335–364). New York: Wiley.
ders. Mental Retardation and Developmental Disabilities Research Walsh, S., Rous, B., & Lutzer, C. (2000). The federal IDEA natural en-
Reviews, 10, 139–143. vironments provisions: Making it work. In S. Sandall & M. Os-
Sandall, S., Hemmeter, M. L., Smith, B. J., & McLean, M. E. (Eds.) trosky (Eds.), Young Exceptional Children Monograph Series 2:
(2005). DEC recommended practices: A comprehensive guide for Natural environments and inclusion (pp. 3–15). Denver: Council
practical application in early intervention/early childhood special for Exceptional Children, Division for Early Childhood.
education. Missoula, MT: Division for Early Childhood (DEC), Wetherby, A., Allen, L., Cleary, J., Kublin, K., & Goldstein, H. (2002).
Council for Exceptional Children (CEC). Validity and reliability of the Communication and Symbolic
Scarborough, A. A., Spiker, D., Mallik, S., Hebbeler, K. M., Bailey, D. Behavior Scales Developmental Profile with very young children.
B., Jr., & Simeonsson, R. J. (2004). A national look at children and Journal of Speech, Language, and Hearing Research, 45, 1202–
families entering early intervention. Exceptional Children, 70, 1218.
469–483. Wetherby, A., Goldstein, H., Cleary, J., Allen, L., & Kublin, K. (2003).
Schwartz, I., Sandall, S., McBride, B., & Boulware, G. (2004). Project Early identification of children with communication disorders:
DATA (Developmentally Appropriate Treatment for Autism): An Concurrent and predictive validity of the CSBS Developmental
inclusive school-based approach to educating young children with Profile. Infants and Young Children, 16(2), 161–174.
autism. Topics in Early Childhood Special Education, 21, 156–168. Wetherby, A., & Prizant, B. (2002). Communication and symbolic be-
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental havior scales developmental profile–First normed edition. Balti-
and quasi-experimental designs for generalized causal inference. more: Brookes.
Boston: Houghton Mifflin. Wetherby, A. & Prizant, B. (2003). Communication and symbolic be-
Shinnar, S., Rapin, I., Arnold, S., Tuchman, R. F., Shulman, L., havior scales–Normed edition. Baltimore: Brookes.
Ballaban-Gil, K., et al. (2001). Language regression in childhood. Wetherby, A. M., Prizant, B. M., & Hutchinson, T. (1998). Com-
Pediatric Neurology, 24, 185–191. municative, social-affective, and symbolic profiles of young chil-
Sigman, M., & Ruskin, E. (1999). Continuity and change in the social dren with autism and pervasive developmental disorder. American
competence of children with autism, Down syndrome, and develop- Journal of Speech-Language Pathology, 7, 79–91.
mental delays. Monographs of the Society for Research in Child Wetherby, A., Watt, N., Morgan, L., & Shumway, S. (in press). Social
Development, 64. communication profiles of children with autism spectrum disorders
Siller, M., & Sigman, M. (2002). The behaviors of parents of children in the second year of life. Journal of Autism and Developmental
with autism predict the subsequent development of their children’s Disorders.
communication. Journal of Autism and Developmental Disorders, Wetherby, A., Woods, J., Allen, L., Cleary, J., Dickinson, H., & Lord,
32, 77–89. C. (2004). Early indicators of autism spectrum disorders in the sec-
Smith, T., Buch, G., & Gamby, T. (2000). Parent-directed, intensive ond year of life. Journal of Autism and Developmental Disorders,
early intervention for children with pervasive developmental disor- 34(5), 473–493.
der. Research in Developmental Disabilities, 21, 297–309. Whalen, C., & Schreibman, L. (2003). Joint attention training for chil-
Sparks, B. F., Friedman, S. D., Shaw, D. W., Aylward, E. H., Echelard, dren with autism using behavior modification procedures. Journal
D., Artru, A. A., et al. (2002). Brain structural abnormalities in of Child Psychology and Psychiatry, and Allied Disciplines, 44,
young children with autism spectrum disorder. Neurology, 59, 456–468.
184–192. Woods, J., Kashinath, S., & Goldstein, H. (2004). Effects of embed-
Sparrow, S., Balla, D., & Cicchetti, D. (1984). Vineland adaptive be- ding caregiver-implemented teaching strategies in daily routines on
havior scales. Circle Pines, MN: American Guidance Service. children’s communication outcomes. Journal of Early Intervention,
Stone, W., Ousley, O., Yoder, P., Hogan, K., & Hepburn, S. (1997). 26, 175–193.
Nonverbal communication in 2- and 3-year old children with autism. Woods, J., & Wetherby, A. (2003). Early identification and interven-
Journal of Autism and Developmental Disorders, 27, 677– 696. tion for infants and toddlers at-risk for autism spectrum disorders.
Stone, W. L., & Yoder, P. J. (2001). Predicting spoken language level in Language, Speech, and Hearing Services in Schools, 34, 180–193.
children with autism spectrum disorders. Autism, 5, 341–361. Woods Cripe, J., & Venn, M. (1997). Family-guided routines for early
Strain, P., McGee, G., & Kohler, F. (2001). Inclusion of children with intervention services. Young Exceptional Children, 1(1), 18–26.

Downloaded from tec.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 6, 2015

You might also like