Research in Autism Spectrum Disorders 7 (2013) 1168–1181

Contents lists available at SciVerse ScienceDirect

Research in Autism Spectrum Disorders

Journal homepage: http://ees.elsevier.com/RASD/default.asp

Comparison of community-based verbal behavior and pivotal

response treatment programs for young children with autism
spectrum disorder
Richard Stock a,*, Pat Mirenda a, Isabel M. Smith b
a
University of British Columbia, Canada
b
Dalhousie University and IWK Health Centre, Canada

A R T I C L E I N F O A B S T R A C T

Article history: This research compared the outcomes of a community-based group program based on the
Received 21 March 2013 verbal behavior approach to early intervention (Sautter & LeBlanc, 2006; Sundberg &
Received in revised form 6 June 2013 Michael, 2001) to the outcomes of a program based on Pivotal Response Treatment (Bryson
Accepted 7 June 2013
et al., 2007; Koegel & Koegel, 2006). Fourteen preschool children with autism spectrum
disorder in each program were matched by baseline chronological age and cognitive score.
Keywords:
Assessments were conducted at the initiation of treatment and 12 months later to measure
Autism spectrum disorder
cognitive, receptive and expressive language, and adaptive behavior skills, as well as
Early intensive behavioral intervention
Verbal behavior problem behavior and parenting stress. Results for both groups showed statistically
Pivotal response treatment significant changes in cognitive scores, receptive and expressive language age equivalents,
and problem behavior scores. Significant results were not found for either adaptive
behavior or parenting stress scores. Changes in cognitive and adaptive behavior scores
were similar to those reported in published studies of applied behavior analytic programs
of similar intensity. Study limitations and recommendations for future research are
provided. Although additional research is needed to examine the long-term effectiveness
of the programs examined in this study, it appears that both hold promise as effective early
intervention approaches that are also relatively cost-effective.
ß 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Over the past few years, research on the outcomes of early intensive behavioral intervention (EIBI) programs for young
children with autism spectrum disorder (ASD) has been examined in five meta-analyses (Eldevik et al., 2009; Makrygianni &
Reed, 2010; Reichow & Wolery, 2009; Spreckley & Boyd, 2009; Virués-Ortega, 2010). In a recent review, Reichow (2012)
identified 26 studies that were included in at least one of these meta-analyses; 18 of the 26 were included in two or more.
Nineteen of the 26 studies were efficacy studies that examined the impact of intervention in ‘‘ideal’’ conditions that typically
included selection of participants who were deemed to be ‘‘good candidates’’ with no comorbidities; therapists who were
well trained and well supervised; and manualized treatment that was carefully planned and implemented with fidelity
(Eikeseth, Klintwall, Jahr, & Karlsson, 2012; Perry et al., 2008). The remaining seven studies examined EIBI effectiveness in
‘‘real life’’ community-based settings (e.g., preschools) that served a broader range of clients. Typically, staff in these settings
were trained to provide the intervention but were not supervised as rigorously as those in efficacy study settings, and

* Corresponding author at: Capilano University, Social Sciences Division, 2055 Purcell Way, North Vancouver, BC V7J 3H5, Canada.
Tel.: +1 604 986 1911x2525.
E-mail address: rstock@capilanou.ca (R. Stock).

1750-9467/$ – see front matter ß 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.rasd.2013.06.002
 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181 1169

treatment fidelity was assessed less rigorously or not at all (Eikeseth et al., 2012). While the five meta-analyses differed with
regard to the criteria used for study inclusion, the outcome variables that were examined, and the formulae used to calculate
effect size, all but one (Spreckley & Boyd, 2009) endorsed both the efficacy and effectiveness of EIBI as an evidence-based
practice. However, the EIBI effectiveness studies (e.g., Bibby, Eikeseth, Martin, Mudford, & Reeves, 2002; Scheinkopf & Siegel,
1998) typically showed less impressive outcomes than the efficacy studies. Some of the factors contributing to this
differential outcome included funding constraints that resulted in reduced intensity or duration of treatment; the challenges
involved in hiring, training, supervising, and retaining community-based staff; and the wide heterogeneity of the children
who were enrolled.
In his review of these meta-analyses, Reichow (2012) noted that EIBI programs with higher treatment intensity (i.e., more
hours per week), longer treatment duration (i.e., more continuous weeks), the inclusion of parent training, and the
involvement of supervisory personnel trained in the intervention method developed at the University of California at Los
Angeles (UCLA; Davis, Smith, & Donahoe, 2002; Lovaas, 1981, 2003) appeared to be related to better child outcomes in at
least one domain. The UCLA method employs both discrete trial and incidental teaching techniques to target skills across
developmental domains, including imitation, matching, basic and advanced receptive and expressive language, play, and
self-care skills. Treatment is typically provided for 30–40 h per week over 1–3 years in children’s homes and communities,
with active parent involvement in the treatment process. However, despite the robust body of research on the positive
outcomes that can be achieved through the UCLA method, it is not a good fit for all children with ASD and their families.
Johnson and Hastings (2002) examined the barriers that may act as deterrents to families who consider enrolling their child
in a UCLA-based program. More than half of the 141 families who participated in the study identified funding barriers as well
as problems with recruiting, training, and maintaining trained interventionists and supervisors in a supportive and
committed team. Other barriers included the need for extensive time and energy to organize the intervention program, to the
detriment of other family needs (e.g., those of siblings); potential disruption of family life and invasion of the home by non-
family members (i.e., interventionists who are present in the home for up to 40 h per week); and lack of the physical space
required for home-based intervention. These barriers are likely exacerbated by family factors such as low socio-economic
status, English as a Second Language (ESL), the presence of more than one child with ASD in a family, and a lack of social
support within and outside of the family.
Because of these concerns, a few EIBI studies have examined the effectiveness of low-intensity programs that are based
on the principles of applied behavior analysis (ABA) but that require fewer staff resources and are thus less demanding
and less expensive to deliver. For example, Eldevik, Eikeseth, Jahr, and Smith (2006) compared the outcomes of behavioral
(n = 13) and eclectic treatment programs (n = 15) in which participants received 1:1 intervention for 12 h/week, on
average. The children in this study were all under 6 years of age and attended regular kindergarten or elementary school
classes for 8–12 h/week, in addition to receiving 1:1 instruction. Children in the behavioral treatment group were
instructed by teachers who were trained and supervised by psychologists with extensive experience in the UCLA method
(Lovaas, 1981, 2003). Children in the eclectic group were instructed by teachers who implemented a number of
interventions, including augmentative communication (via sign language and/or graphic symbols), ABA procedures,
sensory-motor therapies, and other methods, depending on the teachers’ experience. After 2 years of treatment, the
behavioral groups made more gains than the eclectic group, although the gains were more modest than those achieved in
more intensive UCLA treatment studies.
Some research has also examined the outcomes of low-intensity early intervention programs based on Pivotal Response
Treatment1 (PRT; Koegel & Koegel, 2006), a naturalistic behavioral intervention method that is derived from ABA principles.
PRT targets ‘‘pivotal’’ areas of child development such as motivation, responsivity to multiple cues, self-management, and
social initiations, and has been shown in numerous small-scale studies to result in collateral improvements in social and
communication skills as well as reductions in problem behavior (see http://education.ucsb.edu/autism/documents/
SummaryChartofEmpiricalSupportforPRT.pdf). Unlike the UCLA method, PRT interventions do not follow a set curriculum.
Rather, each child’s intervention team develops individual goals with an emphasis on functional communication and other
developmentally appropriate skills that are taught in the context of play and other naturally occurring routines. Specifically,
interventionists are taught to provide instructional supports incorporating: (a) clear, uninterrupted instructions that are
delivered in response to a child’s focus of attention; (b) child preferences and choices to increase motivation; (c) frequent
task variation; (d) interspersal of previously acquired tasks with new acquisition tasks; (e) reinforcement of response
attempts in addition to correct responses; and (f) the use of natural reinforcers that are directly related to the child’s response
(e.g., if a child says ‘‘ball,’’ she receives the ball, not praise or an unrelated item such as food) (Koegel, Koegel, & Brookman,
2003). PRT was deemed an established (i.e., evidence-based) intervention, according to the National Standards Report
(National Autism Center, 2009).
Two large-scale community-based studies to date have examined the effectiveness of PRT as a low-intensity EIBI
treatment for young children with ASD. Baker-Ericzén, Stahmer, and Burns (2007) provided a 12-week parent education
program to 158 parents of young children with ASD and measured child change using the Vineland Adaptive Behavior Scales
(Sparrow, Balla, & Cicchetti, 1984), a parent report measure. Results indicated significant improvements on the Vineland
subscales for communication, socialization, and daily living skills for children ages 3–5 but not for those 6 years of age or
older. However, this study did not assess the fidelity of parent implementation of PRT, which has been shown to be related to
child outcomes (Coolican, Smith, & Bryson, 2010). More recently, I. Smith et al. (2010) reported the effectiveness of PRT in a
study conducted in Nova Scotia (NS), Canada. The NS EIBI model included both parent training and therapist-implemented
1170 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181

1:1 PRT intervention for a maximum of 15 hr/week. I. Smith et al. followed 45 children with ASD (mean baseline age = 50
months) who received PRT for 12 months, and compared the outcomes for children with baseline intelligence quotient (IQ)
scores <50 and 50. They found significant improvements in cognitive, adaptive behavior, and expressive and receptive
language scores for both groups, with greater gains on all measures and decreases in autism symptom severity for the IQ 50
group. Of note is that the mean IQ gain of 16 points for the combined groups after 12 months compared favorably to the gains
reported in some longer-duration UCLA-based studies of both low (e.g., Eldevik et al., 2006) and high intensity (e.g.,
Hayward, Eikeseth, Gale, & Morgan, 2009; T. Smith, Eikeseth, Klevstrand, & Lovaas, 1997; T. Smith, Groen, & Wynn, 2000).
In addition to PRT, a number of recent studies have examined the effectiveness of ABA-based EIBI programs other than the
UCLA model (e.g., Fava et al., 2011; Fernell et al., 2011; Flanagan, Perry, & Freeman, 2012; Perry et al., 2008; Peters-Scheffer,
Didden, Mulders, & Korzilius, 2010). In particular, the verbal behavior (VB) approach (Sundberg & Michael, 2001) has
emerged over the past decade as a community-based treatment that shares many similarities with both the UCLA and PRT
methods but differs in three main ways. First, communication and language instruction is explicitly based on and informed
by Skinner’s (1957) analysis of verbal behavior, which emphasizes the functional nature of the echoic, mand, tact,
intraverbal, and autoclitic aspects of expressive language. Second, instruction in the VB method is usually designed and
delivered according to either the Assessment of Basic Language and Learning Skills-Revised (ABLLS-R; Partington, 2006) or the
Verbal Behavior Milestones Assessment and Placement Program (VB-MAPP; Sundberg, 2007–2008), two related assessment and
curriculum guides that are based on Skinner’s analysis. Finally, VB combines both structured, discrete trial teaching (as in the
UCLA method) and natural environment teaching (as in the PRT method) in the context of daily activities and routines. In
addition, like PRT, VB emphasizes the importance of arranging the environment to incorporate materials or activities that are
motivating for the child, especially during instruction aimed at teaching manding (see Laraway, Snycerski, Michael, & Poling,
2003). Thus, VB includes core elements of both the UCLA and PRT models, as well as some unique features.
To date, there have been no large-scale evaluations of either the efficacy or the effectiveness of the VB method as a
treatment ‘‘package.’’ However, there is considerable empirical support for the instructional approaches that are used to
teach the verbal operants (i.e., mands, tacts, etc.) that are at the heart of VB communication and language instruction (Sautter
& LeBlanc, 2006; Stock, 2012). In fact, the National Standards Report included research on mand training and other verbal
operants in the category of ‘‘behavioral packages’’ that are among the 11 ‘‘established’’ treatments for young children with
ASD (National Autism Center, 2009). Nonetheless, given its widespread use, there is a need for a comprehensive evaluation of
the outcomes of the VB method as a comprehensive EIBI treatment across multiple developmental domains (Carr & Firth,
2005). There is also a need for studies that compare the outcomes of ABA-based programs that are equivalent with regard to
intensity and duration but differ with regard to how the treatment is delivered (e.g., highly structured teaching versus
natural environment/incidental teaching; in 1:1 sessions or in a group). Because ABA is conceptually systematic (Baer, Wolf,
& Risley, 1968, 1987), all treatments based on ABA principles benefit from the same empirical foundation; thus, the
outcomes of ABA-based EIBI treatments such as the UCLA, PRT, and VB methods should be similar, if intensity and duration
are comparable. However, no studies have examined this issue to date. Thus, the primary purpose of the present study was to
compare the effectiveness of low intensity, community-based PRT and VB programs, over a 12-month period. The secondary
purpose was to compare the outcomes of these two programs to those reported in studies of UCLA-based programs of
comparable intensity and duration.

2. Method

2.1. Participants

Approval was obtained from the Research Ethics Boards at the University of British Columbia and at the IWK Health
Centre in Halifax, Nova Scotia for use of the data in this study.

2.1.1. Verbal behavior group

Children in the VB program were enrolled in a private group applied behavior analysis (GABA) preschool program located
in a large city in Western Canada. Eligibility for the GABA program required (a) an ASD diagnosis; (b) a chronological age of
less than 6 years; and (c) the absence of serious problem behaviors that required 1:1 dedicated support, as reported by the
child’s parent and as determined by program staff observations during an initial meeting. The GABA participants represent a
community sample, since all children whose parents applied for them to enter the program over two successive admission
cycles were included in the study. The children included 12 boys and 2 girls whose mean chronological age at treatment
onset was 46.0 months (range = 37–59 months). They all had diagnoses of either autistic disorder or PDD-NOS from qualified
diagnosticians, using the Autism Diagnostic Observation Schedule (ADOS; Lord, Rutter, DiLavore, & Risi, 2001), the Autism
Diagnostic Interview-Revised (ADI-R; LeCouteur, Lord, & Rutter, 2003), and clinical assessments by a multi-disciplinary team.

2.1.2. PRT group

The PRT program sample was comprised of 14 children who were drawn from a larger data set of 45 preschoolers enrolled
in a government-funded autism intervention program in Nova Scotia, Canada. The provincial Nova Scotia Early Intensive
Behavioral Intervention (NS EIBI) program was developed under the leadership of S. Bryson as a partnership between
researchers and clinicians at Dalhousie University/IWK Health Centre and PRT developers at the University of California
 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181 1171

Santa Barbara (UCSB). NS EIBI is delivered as a community-based program by the public healthcare system (see Bryson et al.,
2007). All participants were diagnosed with ASD by a multi-disciplinary team, using the ADOS, the ADI-R, and criteria from
the Diagnostic and Statistical Manual, 4th edition (American Psychiatric Association, 2000). Children became eligible for the
intervention program based solely on a diagnosis of ASD and age between 2 and 6 years. Children were selected for the
intervention program randomly from those eligible and constitute a representative community sample.
Because complete data sets were available for only 14 GABA participants, a matching procedure was employed to identify
appropriate NS EIBI participants for the current study. Specifically, NS EIBI children were drawn from a pool of 45 children to
match the 14 GABA participants according to baseline chronological age and cognitive scores. Matching by age of treatment
onset was warranted because previous research suggests a relationship between age and treatment outcomes (Fenske,
Zalenski, Krantz, & McClannahan, 1985; Harris & Handleman, 2000). Matching by cognitive score is well established in the
autism intervention literature; in a meta-analysis of 133 studies, Mottron (2004) reported that this is the most frequently
used matching variable in autism studies using comparison groups. Moreover, pre-treatment cognitive scores (specifically,
IQ scores) have been found to be correlated with post-treatment performance in several EIBI studies (e.g., Harris &
Handleman, 2000; Makrygianni & Reed, 2010; Sallows & Graupner, 2005).
In order to conduct matching that was blind with regard to the dependent variables in this study (except for cognitive
scores), the first author was provided with anonymous NS EIBI data for the two matching variables only. Each GABA
participant was then matched to one NS EIBI participant that was closest with regard to both age and cognitive score. After all
matches were completed, baseline and 12-month data related to the dependent variables were provided by the third author
for the 14 matched participants. The final NS EIBI sample consisted of 12 boys and 2 girls with a mean chronological age at
treatment onset of 46.7 months (range = 31–62 months). Coincidentally, the GABA and NS EIBI groups had equal numbers of
males and females, although gender was not used as a matching variable.

2.2. Measures

A set of common outcome measures was collected for both GABA and NS EIBI participants, at baseline and approximately
12 months later. These included measures of language and communication, cognitive ability, adaptive behavior, problem
behavior, and parenting stress.

2.2.1. Preschool Language Scale

The Preschool Language Scale, 4th edition (PLS-4; Zimmerman, Steiner, & Pond, 2002) is an individually administered test
that can be used to identify children from birth through 6 years 11 months who have a language disorder or delay. It
assesses both receptive language (called auditory comprehension in the PLS) and expressive language (called expressive
communication in the PLS) and yields age-based standard scores, percentile ranks, and age equivalents. The PLS-4 is a
reliable instrument (r = .82–.95 for subscale scores and .90–.97 for the Total Language Score) with internal consistency (a)
coefficients of .81 or above (for most ages) and good clinical validity (sensitivity = .80; specificity = .88 for the Total
Language Score). In this study, age equivalents for both subtests were available for all participants and thus were used for
data analysis.

2.2.2. Merrill-Palmer-Revised Scales

The Merrill-Palmer-Revised Scales of Development (M-P-R; Roid & Sampers, 2004) was employed as the primary measure of
cognitive ability. The M-P-R Developmental Index (DI) is a general index (comparable to an IQ score) that is comprised of
scales measuring cognition, fine motor skills, and receptive language. The M-P-R DI is highly correlated with the Bayley
Scales of Infant Development and the Brief IQ score from the Leiter International Performance Scale-Revised (r = .92 and .94,
respectively) (Roid & Sampers, 2004).
M-P-R DI standard scores were used as a measure of cognitive ability unless a child’s standard score fell below the
lowest obtainable score, in which case a cognitive ratio was calculated as the age-equivalent score divided by
chronological age in months, multiplied by 100. If M-P-R performance was such that even an age-equivalent score could
not be derived, scores were estimated according to the method used by I. Smith et al. (2010). In this method, the age
equivalents for the Vineland Adaptive Behavior Scales (2nd edition) receptive language, fine motor, and daily living skills
subdomains were averaged and then divided by a child’s chronological age to yield a cognitive ratio. These three
subdomains were selected because they most closely reflect the abilities measured by the M-P-R DI. Alternatively, in a few
cases, scores from other assessment measures were used in place of the M-P-R DI. In the GABA group, two children were
assessed using the Stanford Binet Intelligence Scale – 5th Edition (Roid, 2003) at baseline; full-scale IQ scores were used for
these children. Two other GABA children were assessed with the Mullen Scales of Early Learning (Mullen, 1995) at baseline;
the Mullen Early Learning Composite standard score was used in these cases. Finally, one GABA child was assessed with the
Wechsler Preschool and Primary Scale of Intelligence – 3rd Edition (Wechsler, 2002) at baseline, and the full-scale IQ score
was used for this child.

2.2.3. Vineland Adaptive Behavior Scales

The Survey Form of the Vineland Adaptive Behavior Scales, 2nd edition (VABS-II; Sparrow, Cicchetti, & Balla, 2005) was used
as a measure of adaptive behavior, based on parent report. The VABS-II yields age equivalent scores that are sensitive to
1172 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181

change over time, have more item density at lower ends, and are generally recommended for assessing change with
treatment over time (Matson, 2008). Additionally, the VABS-II has been shown to have good inter-rater and test–retest
reliability, subscale inter-correlations, and construct and criterion validity (Matson, 2007). The VABS-II includes
Communication, Socialization, Daily Living Skills, and Motor domain scales and yields an overall Adaptive Behavior
Composite (ABC) standard score, which was used in this study.

2.2.4. Child Behavior Checklist

Internalizing and externalizing problem behaviors were assessed using the Child Behavior Checklist, Ages 1(1/2)–5 (CBCL;
Achenbach & Rescorla, 2000). The CBCL has good psychometric properties, including test–retest values ranging from r = .95
to 1.0, inter-rater reliabilities ranging from r = .93 to .96, and internal consistency scores ranging from a = .78 to .97
(Achenbach & Rescorla, 2000). The CBCL obtains parental ratings of 99 behaviors and provides T-scores and percentile ranks.
This study employed CBCL Total Problem behavior T-scores.

2.2.5. Parenting Stress Index-Short Form

Participants’ parents completed the Parenting Stress Index–Short Form (PSI-SF; Abidin, 1995), a 36-item measure of
parenting stress. The PSI-SF has good psychometric properties including test–retest reliability scores ranging from r = .68 to
.85 and internal consistency scores ranging from a = .80 to .91 (Abidin, 1995). This study employed PSI Total scores.

2.3. Independent variables

2.3.1. GABA program

Participants in the GABA program were assessed using the ABLLS-R assessment tool (Partington, 2006) and received
intervention based on the VB method (Barbera, 2007; Sundberg & Michael, 2001). Instructional goals and individualized
intervention plans based on the results of the ABLLS-R were developed in collaboration with each child’s parents. Typically,
target skills were identified for each child in domains that included visual performance skills (e.g., object matching),
receptive language, imitation, manding (i.e., requesting), tacting (i.e., labeling), and intraverbals (i.e., conversation skills).
Programming also included skills for independence such as toilet training, self-dressing, and feeding (as required), as well as
group skills (e.g., sharing, turn taking, etc.). Intervention occurred in both Intensive Teaching (IT) and Natural Environment
Teaching (NET) contexts. During IT sessions, children worked on individualized curriculum goals in 1:1 discrete trial
teaching sessions. During NET sessions, two children worked together with one therapist on goals related to play, language,
social, and other skills. To the extent possible, children were paired in NET sessions based on similar goals so that these
sessions were equally beneficial for both children. GABA participants were enrolled for 3–5 h/day, 5 days/week, for a total of
15–25 h of intervention/week over 48 weeks/year. During this time, participants received 3–5 h/week of 1:1 discrete trial
teaching; the remainder of the intervention time (12–20 h per week) occurred in 2:1 or small groups in a preschool-like
setting.
The GABA program was directed and overseen by a doctoral-level psychologist and two masters-level clinicians, all of
whom were Board Certified Behavior Analysts (BCBAs) and had extensive training in autism and VB methods. Together, they
trained and supervised therapists who worked directly with the children and had, on average, early childhood education
backgrounds. Supervisory staff also provided monthly workshops to parents on topics such as teaching manding, providing
instruction in natural routines at home, teaching play skills, and dealing with problem behavior. Parents were encouraged to
use the skills taught in these workshops at home with their children.

2.3.2. NS EIBI program

Children in the NS EIBI program received intervention based on the PRT method (Koegel & Koegel, 2006), delivered by a
combination of therapists and parents (primarily, the former) in both home and daycare/preschool settings. Each child’s
intervention team (including the parents) developed goals with an emphasis on functional communication and
developmentally appropriate social and play skills; instruction on daily living skills (e.g., toilet training, dressing, etc.) was
included as needed on an individual basis. All intervention occurred within the context of play and other functional daily
routines at home or in full- or part-time daycare or preschool, which was available to all but two of the children at some
point during the intervention period. NS EIBI staff supported children in these settings on a 1:1 basis, with the goal of
facilitating social and communicative interactions with adults and peers. Problem behavior was addressed using the
strategies of positive behavior support (Lucyshyn, Dunlap, & Albin, 2002). As described by I. Smith et al. (2010), children
who entered the NS EIBI program in Year 1 (n = 8 in the current subsample) received 15 h/week of 1:1 intervention over a
12-month period. Children who entered in Year 2 (n = 6 in the current subsample) received 15 h/week of 1:1 instruction for
6 months, 10 h/week for the next 3 months, and 5 h/week for a final 3-month period. I. Smith et al.’s (2010) analysis of NS
EIBI outcomes revealed no significant differences between the two cohorts; thus, participants in this study were drawn
from both.
Each NS EIBI child’s intervention team consisted of their parents, one or more therapists (usually, early childhood
educators), a senior therapist (a Master’s level psychologist or occupational therapist), a supervising psychologist, and a
speech-language pathologist. Parents were taught to use PRT techniques during initial in vivo coaching sessions, and
were encouraged to do so during everyday activities at home and in the community with their children. No research data
 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181 1173

were collected on the extent or fidelity of parent involvement. Treatment teams were trained either by PRT clinicians
from the University of California at Santa Barbara (UCSB) or by professionals within the NS EIBI program (three
psychologists, one occupational therapist, and one doctoral psychology student) who received training as trainers from
UCSB staff.

2.4. Treatment fidelity

Treatment fidelity was measured in both programs, using two different approaches.

2.4.1. GABA program

Treatment fidelity of therapists’ 1:1 teaching skills was assessed by program supervisors at various time points via in vivo
observations with immediate performance feedback, using skill checklists that were created by the psychologist who
directed the program. Therapists were evaluated 1–4 times/year, depending on their previous experience, skill set, and
length of service in the program. Skills evaluation focused on three broad areas: instructors’ use of reinforcement and
strategies to enhance motivation, instructional control and techniques, and organizational skills.
The average fidelity score for therapists who taught children in this study was 81% across all evaluations; the average
highest score achieved was 89%. Therapists whose evaluation scores were initially low received additional weekly support
and training from supervisors until they met criteria; for example, the therapists with the two lowest initial evaluations
improved to achieve 91% and 92% respectively on their final evaluations. Additionally, each therapist was monitored by or
had access to a supervisor on a weekly basis for ongoing skills training and support.

2.4.2. NS EIBI program

Treatment fidelity was monitored at the start of intervention via videotaped probes of adult-child interactions that were
assessed for accuracy of PRT implementation using continuous 2-min interval sampling for a total of 10 min/session. Coding
focused on six key procedures: (a) providing clear opportunities and instructions; (b) providing child choice with shared
control; (c) providing immediate, contingent reinforcement; (d) using natural reinforcers; (e) providing reinforcement for
verbal attempts as well as correct responses; and (f) providing an appropriate balance of maintenance (i.e., easy) and
acquisition (i.e., new) tasks (see Bryson et al., 2007).
The majority of therapists (86.6%) met the fidelity criteria (>80% accurate performance) within the first 4 months of
working with the first child to whom they were assigned. Additional training was provided to the therapists who did not
meet criteria, until fidelity was achieved. Subsequently, each interventionist was monitored by a clinical supervisor and
received feedback on his/her implementation of PRT on a bi-weekly basis, either in vivo or via video review; however, formal
fidelity checks were not recorded during this time.

2.5. Research design and procedure

This study employed a quasi-experimental pre-test/post-test design with matched groups. Both GABA and NS EIBI data
were collected at baseline (i.e., at the initiation of intervention) and approximately 12 months later. GABA data were
collected in each child’s home or at the program preschool by a certified speech-language pathologist, registered
psychologist (at baseline), or the first author (who was also one of the GABA supervisors). NS EIBI data collection took place
primarily in a clinical research setting, or in some cases, in children’s homes or preschools. Assessments were conducted by
research assistants (supervised by the third author) who were experienced in testing children with ASD and were not
involved in the intervention. Parent report measures were mailed to parents and were returned either in person or through
the mail at both sites.

3. Results

Data analysis was conducted using SPSS Version 20. Preliminary analyses using independent samples t-tests (two-tailed)
were conducted first, to confirm that the GABA and NS EIBI samples were matched with regard to baseline chronological age
and cognitive score. There was no significant difference in chronological age at the start of intervention between the NS EIBI
(M = 46.71, SD = 9.23) and GABA (M = 46.00, SD = 8.12) groups; t = .218, p = .830. Similarly, there was no significant difference
in cognitive scores for the NS EIBI (M = 42.71, SD = 23.60) and GABA (M = 39.79, SD = 24.59) groups; t = .322, p = .750. The
alpha values exceed the p = .50 recommended by Mervis and Klein-Tasman (2004) for assessing differences between
matching variables, confirming that the two groups were well matched.
Analyses related to the dependent variables employed a 2  2 mixed model analysis of covariance (ANCOVA) or analysis
of variance (ANOVA) with Time as the within subjects factor and Group as the between subjects factor. Baseline cognitive
scores were used as a covariate in the expressive language, receptive language, and adaptive behavior analyses. Using a
Bonferroni adjustment, the significance level was set at p = .01 for each of the five child variable analyses. Because it was
possible to hypothesize, on the basis of previous research, that cognitive, receptive language, expressive language, and
adaptive behavior scores would increase and that problem behavior scores would decrease over 12 months, one-tailed tests
were conducted to analyze the scores for these variables. Two-tailed tests were conducted for parenting stress, since no
1174 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181

120

100

Cognitive Score
80

60

40

20

0
Baseline 12 Months

Fig. 1. NS EIBI individual and mean (in bold) cognitive score results.

120

100
Cognitive Score

80

60

40

20

0
Baseline 12 Months

Fig. 2. GABA individual and mean (in bold) cognitive score results.

hypothesis was suggested by past research. In the sections that follow, graphic representations of the data are provided only
for measures that showed significant change across Group or Time.

3.1. Cognitive scores

Figs. 1 and 2 display the cognitive score results for the NS EIBI and GABA groups, with means indicated in bold. Results of a
mixed-model ANOVA showed no Group  Time interaction, F(1, 26) = .320, p = .289 and no significant difference for Group,
F(1, 26) = .257, p = .309 (one-tailed). However, there was a significant difference for Time, F(1, 26) = 6.69, p = .008 (one-tailed),
h2 = .205.
As is evident from Figs. 1 and 2, there was wide variability within each group. On average, the NS EIBI group gained 13.15
points, with increased scores for 10 participants and decreased scores for four. The GABA group gained 8.42 points over the
same time period, with increased scores for 11 participants and decreased scores for three.

3.2. Receptive language

Figs. 3 and 4 display the receptive language age equivalents for the NS EIBI and GABA groups, with means indicated in
bold. Results of a mixed-model ANCOVA, controlling for baseline IQ, showed no Group  Time interaction, F(1, 25) = .305,
p = .293 and no significant difference for Group, F(1, 25) = .070, p = .397 (one-tailed). However, there was a significant
difference for Time, F(1, 25) = 10.07, p = .002 (one-tailed), h2 = .287.
On average, the NS EIBI group gained 9.22 months, with increased scores for 13 participants and a decreased score for one.
The GABA group gained 10.36 months, with increases for all 14 participants.

3.3. Expressive language

Figs. 5 and 6 display the expressive language age equivalents for the NS EIBI and GABA groups, with means indicated in
bold. Results of the mixed-model ANCOVA, controlling for baseline cognitive ability, showed no Group  Time interaction,
 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181 1175

80

70

PLS-4 Receptive Language AEs

60

50

40

30

20

10

0
Baseline 12 Months

Fig. 3. NS EIBI individual and mean (in bold) receptive language age equivalent score results.

80

70
PLS-4 Receptive Language AEs

60

50

40

30

20

10

0
Baseline 12 Months

Fig. 4. GABA individual and mean (in bold) receptive language age equivalent score results.

F(1, 2) = 1.36, p = .127 and no significant difference for Group, F(1, 25) = 1.01, p = .162 (one-tailed). However, there was a
significant difference for Time, F(1, 25) = 13.52, p = .0005 (one-tailed), h2 = .351.
On average, the NS EIBI group gained 10.35 months, with increases for all participants. The GABA group gained 7.07
months, with increases for 12 participants and no change for two.

3.4. Adaptive behavior

Hurd, Perry, and Flanagan (2009) suggested that the VABS-II may result in inflated standard scores relative to the original
VABS, making the measure less sensitive to change. In order to address this concern, ANCOVAs (controlling for baseline IQ)
were conducted using both VABS-II standard and VABS-II raw scores. The results of both analyses were identical, so only the
standard score results (the more conventional of the two) are reported here. Results showed no Group  Time interaction,
F(1, 25) = 1.06, p = .157 and no significant difference for either Group, F(1, 25) = 3.83, p = .031 (one-tailed) or Time, F(1,
25) = 0.06, p = .404 (one-tailed).
There was wide variability within each group; however, neither group experienced significant changes over 12 months.
This may be due, in part, to the large standard deviations observed for the GABA group relative to the NS EIBI group (17.03 vs.
9.02). On average, the NS EIBI group gained 5.5 points, with increasing scores for 10 participants and decreasing scores for
four. The GABA group gained an average of 2.72 points over the same time period, with increasing scores for eight
participants, no change for two, and decreasing scores for four.

3.5. Problem behavior

Figs. 7 and 8 display the CBCL total scores for the NS EIBI and GABA groups, with means indicated in bold; note that the
baseline score was missing for one NS EIBI participant. IQ was not entered as a covariate in this analysis. Results of a mixed-
model ANOVA showed no Group  Time interaction, F(1, 25) = .841, p = .184 and no significant difference for Group, F(1,
25) = .444, p = .256 (one-tailed). However, there was a significant difference for Time, F(1, 25) = 9.78, p = .002 (one-tailed),
h2 = .281. On average, the mean NS EIBI problem behavior score decreased 2.62 points, with decreasing scores for 10
1176 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181

70

60

PLS-4 Expressive Language AEs

50

40

30

20

10

0
Baseline 12 Months

Fig. 5. NS EIBI individual and mean (in bold) expressive language age equivalent score results.

70
PLS-4 Expressive Language AEs

60

50

40

30

20

10

0
Baseline 12 Months

Fig. 6. GABA individual and mean (in bold) expressive language age equivalent score results.

participants and increasing scores for three. The mean GABA score decreased by 4.79 points, with decreasing scores for 12
participants and increasing scores for two.

3.6. Parenting stress

Two-tailed tests were conducted for this analysis because it was not possible to form a hypothesis based on previous
research. PSI-SF data were not available for three NS EIBI parents and four GABA parents. Results of a mixed-model ANOVA
showed no Group  Time interaction, F(1, 19) = .043, p = .838 and no significant difference for Group, F(1, 19) = .932, p = .347
(two-tailed) or Time, F(1, 19) = .620, p = .441 (two-tailed). Again, there was considerable variability within each group, but
neither experienced significant decreases in overall parent stress. On average, the NS EIBI parent stress scores decreased 3.1
points, with decreasing scores for five participants, increasing scores for five, and no change for one. GABA scores decreased
by a mean of 1.8 points, with decreasing scores for six participants and increasing scores for four.

3.7. Comparison with UCLA model studies

In order to compare the outcomes of the children in the current study with those of existing published data on the UCLA
method, the approximate total hours of intervention were calculated for the GABA and NS EIBI groups, and for UCLA studies
that reported the results over 12 months and were of comparable intensity. For the current study, the GABA group’s average
hours per week (20) was combined with the NS EIBI group’s maximum hours per week (15) to yield 17.5 h/week for the
combined groups. This weekly intensity was multiplied by 4 weeks per month and then multiplied by 12 to yield an estimate
of 840 total hours across the 12-month intervention period. Using a parallel set of calculations, three UCLA model studies
were identified, with total intervention hours estimated between 10:80 and 12:00 (mean = 11:00 h). The comparison studies
are summarized in Table 1.
Unfortunately, the published data do not allow for a comparison of language scores due to discrepancies between the
measures used and the types of scores reported. Thus, across the three UCLA model studies and the current study, the only
measures that allow for a rough comparison of changes over 12 months are those related to cognitive ability and adaptive
behavior. The VABS-II Composite (total) mean score gain of +4.11 pts across the NS EIBI and GABA groups was not significant,
 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181 1177

80

70

60

CBCL Total Score

50

40

30

20

10

0
Baseline 12 Months

Fig. 7. NS EIBI individual and mean (in bold) problem behavior total score results.

80

70

60
CBCL Total Score

50

40

30

20

10

0
Baseline 12 Months

Fig. 8. GABA individual and mean (in bold) problem behavior total score results.

Table 1
Comparison of current study with UCLA-based studies of similar intensity.

Study N Baseline age, Duration/intensity Settings Outcomes

(range), and
male:female ratio

Anderson et al. (1987) 14 43 months Mean 15 h/week for 18 mo; approx. Homes Mean IQ +5.6 pts;
(18–64), 11:3 1080 total hours preschools Mean VABS ABC +10.0 pts
Eldevik et al. (2006) 13 53 months Mean 12.5 h/week for 24 mo; approx. Classrooms Mean IQ + 8.2 pts;
(36–68), 10:3 1200 total hours Mean VABS ABC 0.1 pts
Eikeseth et al. (2012) 35 47 months Mean 23 h/week for 12 mo (range 15–37); Preschool Mean VABS-II ABC +8.3 pts
(25–76), 29:6 approx. 1100 total hours
Current study 28 46 months Mean 17.5 h/week for 12 mo (range 15–25); Homes and Mean IQ + 10.8 pts;
(31–62), 24:4 approx. 840 total hours preschools Mean VABS-II ABC +4.11 pts

in contrast to the +10 pt. gain reported by Anderson, Avery, DiPietro, Edwards, and Christian (1987) and the +8.3 pt. gain
reported by Eikeseth et al. (2012). However, the mean gain in the current study was higher than the 0.1 pt loss reported by
Eldevik et al. (2006). Similarly, the mean IQ score gain across the two groups in the current study (+10.79 pts) was similar to
and slightly higher than the +5.6 pt. gain reported by Anderson et al. (1987) and the +8.2 pt. gain reported by Eldevik et al.
(2006). Eikeseth et al. (2012) did not report cognitive measures.

4. Discussion

This study was the first to employ a comparative examination of the 12-month outcomes of two types of relatively low
intensity, community-based EIBI programs: a VB-based 1:1 and small group program, and a PRT-based 1:1 intervention
delivered in home, daycare/preschool and community settings. In addition to employing a comparative approach, it is also
the first study to report outcomes of the VB method as a ‘‘package’’ over an extended period.
The similarity of outcomes between the NS EIBI and GABA programs is perhaps not surprising, given that both
interventions are based on the same ABA principles and instructional strategies, including prompting, fading, discrimination
1178 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181

training, shaping, positive reinforcement for correct responses and approximations, and so forth. In addition, both programs
emphasize communication and language instruction, with an early focus on manding (i.e., requesting) in natural
environments. Finally, both programs rely heavily on contingent positive reinforcement for desired behaviors; both
programs are delivered by trained interventionists with ongoing supervision by experienced staff; and both seek to embed
instruction in activities that are highly motivating for the participants. However, the outcome similarities are also
noteworthy in light of several differences between the two approaches. While NS EIBI involves fewer therapy hours on
average, those intervention hours were delivered entirely on a 1:1 basis across home and preschool/daycare settings. Parents
were coached to implement PRT at home and were invited to follow through with these techniques throughout daily life. In
contrast, GABA delivered only 3–5 h per week of 1:1 therapy, with the majority of time spent in 2:1 and groups in a preschool
setting that did not include active or frequent parent participation or observation.
It is important to note that all children who were referred to the GABA program over a 2-year period were included in this
study and that, by happenstance, 10 of these 14 children had baseline cognitive scores <50. Thus, in order to achieve matches
with participants from the larger NS EIBI pool, the final sample across the two groups was composed of 20/28 participants with
baseline cognitive scores <50. In the existing literature examining the impact of EIBI for children with ASD, this represents a
unique group of children with relatively low cognitive ability. In fact, of all UCLA model studies examined in published meta-
analyses to date, only T. Smith et al. (1997) reported outcomes for a group of participants with baseline IQ scores similar to those
in the current study. T. Smith et al.’s inclusion criteria specifically required IQ scores <35 and they reported a +12-point mean
increase in IQ after 24 months of intervention at 30 h per week. Because the intensity and duration of their intervention was
much greater than that in the current study, it is not possible to compare the results directly. Nonetheless, it is interesting to
note that the IQ gain of the UCLA group is similar to the cognitive score gain in this study (mean = +10.79) after only 12 months of
treatment. Given the unusually low cognitive ability of most of the children in the present study, the results must be interpreted
with caution and cannot be generalized to all children with ASD who participate in other VB- or PRT-based programs. Indeed,
the results for the PRT sample in this study are markedly different from those reported in the entire NS EIBI study that included a
larger sample with a wider range of cognitive ability (I. Smith et al., 2010).
Also of note is that fact that, despite the unusually low cognitive scores of the two groups at baseline, 7 of 28 participants
in the current study achieved scores >70 after 12 months of treatment. Of these, the four children who began their respective
interventions (either NS EIBI or GABA) with the highest scores (range = 80–91) achieved the highest cognitive score
outcomes (range = 89–119), which is consistent with previous research (e.g., Sallows & Graupner, 2005). Surprisingly,
however, three children who began with lower scores (range = 20–56) also achieved 12 month scores >70 (range = 72–120).
This might reflect, at least in part, Sheinkopf and Siegel’s (1998) observation that EIBI programs tend to emphasize behaviors
that are conducive to test-taking situations (e.g., attending to tasks, following adult instructions). Thus, some participants’
increased cognitive scores, such as one NS EIBI child who showed a +64-point gain (from 56 to 120) over a 12-month period,
likely reflect both an increase in their overall rate of development as well as generalized improvements in attention and
compliance, which, in turn, reflects desirable skill acquisition.
VABS standard scores did not improve significantly over time in the current study. Large baseline standard deviations for
the GABA group (mean SD = 17.03) likely contributed to this non-significant result, which was also reported in some UCLA
model studies of both similar (e.g., Eldevik et al., 2006) and greater intensity (e.g., T. Smith et al., 2000). On the other hand,
significant changes in VABS scores have been associated with and attributed to UCLA model interventions that are both more
intensive (e.g., Bibby et al., 2002; Lovaas, 1987), and similar in intensity to the current study (Anderson et al., 1987; Eikeseth
et al., 2012). In addition, I. Smith et al. (2010) reported moderate gains in adaptive behavior in their larger PRT sample. The
most likely explanation for the wide variability in adaptive behavior outcomes is that some treatment programs (and the
studies related to them) emphasize adaptive behavior less than others and thus yield lesser gains. For example, GABA
participants in the current study, who as a group gained +2.7 points on the VABS over a 12-month period, spent most of their
time in a specialized preschool therapy setting where adaptive skills (especially daily living and motor skills) were much less
of a priority than social, language, and communication skills. In addition, in most previous research, adaptive behavior was
measured using the original VABS (Sparrow et al., 1984), which some have suggested provides more conservative standard
scores that are more sensitive to change (Hurd et al., 2009). If this is the case, comparisons between the VABS and the VABS-II
(which was used in the present study) may be misleading.
The results also indicated that parenting stress scores neither increased nor decreased significantly over the course of
intervention and that there were no differences between the GABA and NS EIBI groups. Large standard deviations in both
groups (GABA = 20.41; NS EIBI = 17.18) combined with small sample sizes make the results difficult to interpret. Of course, it
is likely that variables other than those examined in this study also need to be considered when examining parenting stress.
For example, child characteristics such as problem behavior and a lack of adaptive behavior have been associated with
maternal stress (Tomanik, Harris, & Hawkins, 2004).

4.1. Limitations

This study is limited in several ways. First, a quasi-experimental approach was necessary in order to use data from the two
community-based samples examined herein. Hence, the study did not utilize either a no-treatment control group or random
assignment to treatment groups and thus is best conceptualized as exploratory. Second, the study is also limited by its
sample size, which affects the conclusions that can be drawn from the analyses. However, 14 participants per group is not
 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181 1179

unusually small for studies of this type; for example, UCLA model studies to date have ranged from N = 9 (Birnbrauer & Leach,
1993) to N = 66 (Bibby et al., 2002), with a mean N = 21. While a larger sample size would have been desirable, using
community-based samples requires flexibility regarding the participants who are available. Moreover, unlike many previous
studies, the present participants were not recruited to research interventions and were not selected based on characteristics
such as pretreatment IQ or parental socioeconomic or educational status.
Third, a number of measurement challenges were encountered during the course of the current study, especially with
regard to cognitive and language scores. Several children in both groups (n = 11) were unable to obtain a valid standard score
on the M-P-R, primarily at baseline. In those cases, a ratio cognitive score was computed (see Section 2.2.2) and substituted
for a standard score. This situation is not unique and has been reported in other behavioral intervention studies as well (e.g.,
Bibby et al., 2002; Birnbrauer & Leach, 1993; Eldevik et al., 2006; Lovaas, 1987). In addition, we employed age equivalents
(AEs) for measures of receptive and expressive language, in order to make maximal use of the existing data and to avoid
further reducing the sample size in the analysis. The need for use of AEs occurred because six NS EIBI participants were
unable to obtain valid baseline standard scores on the PLS-4, for a variety of reasons (e.g., problem behavior that interfered
with assessment). A number of limitations are inherent with the use of AEs: AEs are not measured on an equal interval scale;
they do not reflect a participant’s relative standing among same-aged peers; and they may falsely imply that abilities
increase at a constant rate over time and thus be unevenly distributed and skewed (Maloney & Larrivee, 2007; Mervis &
Klein-Tasman, 2004). Nonetheless, their use is not without precedent in the early intervention research literature (e.g.,
Anderson et al., 1987; I. Smith et al., 2010). In defense of AEs, Luyster, Qiu, Lopez, and Lord (2007) noted that an AE is the most
transparent score when standard scores cannot be obtained and that they allow maximum use of available data.
Experimenter bias also exists as a limitation and potential confound. All NS EIBI data were collected by third-party
research assistants who were not involved in the intervention (I. Smith et al., 2010). However, for the GABA group,
approximately 25% of all assessments were administered/scored by third-party clinicians. The remaining assessments
were conducted and scored by the first author, who was also involved in intervention for some of the children. To
minimize bias, the first author did not access the baseline data of any of the children prior to conducting the 12-month
assessments; nonetheless, third party administration of all assessment measures would have been ideal, had available
resources allowed for it.

4.2. Future research

A number of areas for future research are evident, based on the results and limitations of this study. These include a need
for additional comparative studies of programs offering similar intensity of intervention, more rigorous experimental
designs, larger sample sizes, third-party evaluation, and more rigorous measurement of treatment integrity. In addition, it is
important to reiterate that both the NS EIBI and GABA programs were of low intensity compared to the programs in most
UCLA-based studies. Future research is needed to examine the outcomes of both VB and PRT-based programs when
treatment occurs at higher levels of intensity, using similar child:therapist ratios.

4.3. Clinical implications and conclusion

It is important to acknowledge that the current EIBI research base provides the most extensive support for the efficacy of
intensive, UCLA-based treatment (Reichow, 2012). However, some families desire an alternative model for one or more
reasons, including goodness-of-fit (Simeonsson, Bailey, Huntington, & Comfort, 1986) for their child and family, the amount
of parent involvement that a family can offer, and the location of the intervention (home versus clinic or preschool). An
additional consideration, of great importance to many families, insurance providers, and government policy-makers, is the
annual cost of intervention. EIBI can cost between $40,000 and $75,000 (in Canadian dollars, CAD) per year (Motiwala, Gupta,
Lilly, Ungar, & Coyte, 2006). In contrast, the estimated annual cost per child for the GABA program at the time of data
collection was $24,000–$28,000 CAD/year; currently, the costs are somewhat higher, due to increased staff and overhead
costs. Unfortunately, a cost estimate is not available for the NS EIBI program; however, I. Smith et al. (2010) described it as a
‘‘far less costly community-based model’’ (p. 517), when compared to the UCLA method. A formal cost-effectiveness study of
the NS EIBI program is currently in progress.
Given the results of this study, within the limitations noted, both the GABA VB-based and the NS EIBI PRT-based programs
appear to be feasible options for parents seeking an alternative early intervention program for their children with ASD. NS
EIBI may provide the flexibility of a behavioral intervention that is conducted by both interventionists and parents in both
home and community settings, while the GABA program may be more attractive to families seeking a specialized preschool
intervention setting outside of the home. Although additional research is needed to examine the impact of both programs
and of the PRT and VB methods more broadly, they both appear to offer promising early behavioral intervention approaches
that are also relatively cost-effective.

Acknowledgements

We are grateful to the parents and children who participated in this study; to the director and staff at GABA and the ABA
Learning Centre; and to Susan Bryson and the research staff and clinicians in the NS EIBI program and evaluation project. This
1180 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181

study was conducted by the first author in partial fulfillment of the requirements for a doctoral degree in special education at
the University of British Columbia, with supervision from the second author. A poster based on the study was presented at
the 2013 annual autism conference of the Association for Behavior Analysis International in Portland, OR. Financial support
was provided by the Canadian Institutes for Health Research, the Social Sciences and Humanities Research Council of Canada,
the Nova Scotia Health Research Foundation, and the University of British Columbia.

References

Abidin, R. (1995). Parenting Stress Index-Short Form (3rd ed.). Lutz, FL: Psychological Assessment Resources.
Achenbach, T. M., & Rescorla, L. A. (2000). Manual for the ASEBA preschool forms and profiles. Burlington: University of Vermont.
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.
Anderson, S. R., Avery, D. L., DiPietro, E. K., Edwards, G. L., & Christian, W. P. (1987). Intensive home-based early intervention with autistic children. Education and
Treatment of Children, 10, 352–366.
Baer, D., Wolf, M., & Risley, R. (1968). Some current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis, 1, 91–97.
Baer, D., Wolf, M., & Risley, R. (1987). Some still-current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis, 20, 313–327.
Baker-Ericzén, M. J., Stahmer, A. C., & Burns, A. (2007). Child demographics associated with outcomes in a community-based pivotal response training program.
Journal of Positive Behavior Interventions, 9, 52–60.
Barbera, M. L. (2007). The verbal behavior approach: How to teach children with autism and related disorders. Philadelphia: Jessica Kingsley Publishers.
Bibby, P., Eikeseth, S., Martin, N. T., Mudford, O. C., & Reeves, D. (2002). Progress and outcomes for children with autism receiving parent-managed intensive
interventions. Research in Developmental Disabilities, 23, 81–104.
Birnbrauer, J. S., & Leach, D. J. (1993). The Murdoch Early Intervention Program after 2 years. Behavior Change, 10, 63–74.
Bryson, S. E., Koegel, L. K., Koegel, R. L., Openden, D., Smith, I. M., & Nefdt, N. (2007). Large scale dissemination and community implementation of pivotal response
treatment: Program description and preliminary data. Research and Practice for Persons with Severe Disabilities, 32, 142–153.
Carr, J. E., & Firth, A. M. (2005). The verbal behavior approach to early and intensive behavioral intervention for autism: A call for additional empirical support.
Journal of Early and Intensive Behavior Intervention, 2, 18–27.
Coolican, J., Smith, I., & Bryson, S. (2010). Brief parent training in pivotal response treatment for preschoolers with autism. Journal of Child Psychology and
Psychiatry, 51, 1321–1330.
Davis, B. J., Smith, T., & Donahoe, P. (2002). Evaluating supervisors in the UCLA treatment model for children with autism: Validation of an assessment procedure.
Behavior Therapy, 33, 601–614.
Eikeseth, S., Klintwall, L., Jahr, E., & Karlsson, P. (2012). Outcome for children with autism receiving early and intensive behavioral intervention in mainstream
preschool and kindergarten settings. Research in Autism Spectrum Disorders, 6, 829–835.
Eldevik, S., Eikeseth, S., Jahr, E., & Smith, T. (2006). Effects of low-intensity behavioral treatment for children with autism and mental retardation. Journal of Autism
and Developmental Disorders, 36, 211–224.
Eldevik, S., Hastings, R., Hughes, C. J., Jahr, E., Eikeseth, S., & Cross, S. (2009). Meta-analysis of early intensive behavioral intervention for children with autism.
Journal of Clinical Child & Adolescent Psychology, 38, 439–450.
Fava, L., Strauss, K., Valeri, G., D’Elia, L., Arima, S., & Vicari, S. (2011). The effectiveness of cross-setting early intensive behavioral intervention for young children
with ASD. Research in Autism Spectrum Disorders, 5, 1479–1492.
Fenske, E. C., Zalenski, S., Krantz, P. J., & McClannahan, L. E. (1985). Age at intervention and treatment outcome for autistic children in a comprehensive
intervention program. Analysis and Intervention in Developmental Disabilities, 5, 49–58.
Fernell, E., Hedvall, A., Westerlund, J., Carlsson, L. H., Eriksson, M., Olsson, M. B., et al. (2011). Early intervention in 208 Swedish preschoolers with autism spectrum
disorder: A prospective naturalistic study. Research in Developmental Disabilities, 32, 2092–2101.
Flanagan, H., Perry, A., & Freeman, N. (2012). Effectiveness of large-scale community-based Intensive Behavioral Intervention: A waitlist comparison study
exploring outcomes and predictors. Research in Autism Spectrum Disorders, 6, 673–682.
Harris, S., & Handleman, J. (2000). Age and IQ at intake as predictors of placement for young children with autism: A four- to six-year follow-up. Journal of Autism
and Developmental Disorders, 30, 137–142.
Hayward, D., Eikeseth, S., Gale, C., & Morgan, S. (2009). Assessing progress during treatment for young children with autism receiving intensive behavioral
intervention. Autism, 13, 613–633.
Hurd, K., Perry, A., & Flanagan, H. E. (2009). Are standards scores higher on the new Vineland? Poster presentation at the International Meeting for Autism Research
(IMFAR).
Johnson, E., & Hastings, R. P. (2002). Facilitating factors and barriers to the implementation of intensive home-based behavioral intervention for young children
with autism. Childcare, Health & Development, 28, 123–129.
Koegel, R. L., & Koegel, L. (2006). Pivotal response treatments for autism: Communication, social, and academic development. Baltimore: Paul H Brookes.
Koegel, R. L., Koegel, L. K., & Brookman, L. I. (2003). Empirically supported pivotal response interventions for children with autism. In A. Kazdin & J. R. Weisz (Eds.),
Evidence-based psychotherapies for children and adolescents (pp. 341–357). New York: Guilford Press.
Laraway, S., Snycerski, S., Michael, J., & Poling, A. (2003). Motivating operations and terms to describe them: Some further refinements. Journal of Applied Behavior
Analysis, 36, 407–414.
LeCouteur, A., Lord, C., & Rutter, M. (2003). The Autism Diagnostic Interview: Revised (ADI-R). Los Angeles: Western Psychological Services.
Lord, C., Rutter, M. A., DiLavore, P. C., & Risi, S. (2001). The Autism Diagnostic Observation Schedule. Los Angeles, CA: Western Psychological Services.
Lovaas, O. I. (1981). Teaching developmentally disabled children: The me book. Baltimore: University Park Press.
Lovaas, O. I. (1987). Behavioral treatment and normal educational and intellectual functioning in young autistic children. Journal of Consulting and Clinical
Psychology, 55, 3–9.
Lovaas, O. I. (2003). Teaching individuals with developmental delays: Basic intervention techniques. Austin, TX: Pro-Ed.
Lucyshyn, J. M., Dunlap, G., & Albin, R. W. (Eds.). (2002). Families and positive behavior support: Addressing problem behavior in family contexts. Baltimore: Paul H.
Brookes.
Luyster, R., Qiu, S., Lopez, K., & Lord, C. (2007). Predicting outcomes of children referred for autism using the MacArthur-Bates communicative developmental
inventory. Journal of Speech, Language, and Hearing Research, 50, 667–681.
Makrygianni, M., & Reed, P. (2010). A meta-analytic review of the effectiveness of behavioural early intervention programs for children with autistic spectrum
disorders. Research in Autism Spectrum Disorders, 4, 577–593.
Maloney, E. S., & Larrivee, L. S. (2007). Limitations of age-equivalent scores in reporting the results of norm referenced tests. Contemporary Issues in Communication
Science and Disorders, 34, 86–93.
Matson, J. L. (2007). Handbook of assessment in persons with intellectual disability. Oxford, England: Elsevier Science.
Matson, J. L. (2008). Clinical assessment and intervention for autism spectrum disorders: Practical resources for the mental health professional. Burlington, MA:
Academic Press.
Mervis, C. B., & Klein-Tasman, B. P. (2004). Methodological issues in group-matching designs: Alpha levels for control variable comparisons and measurement
characteristics of control and target variables. Journal of Autism and Developmental Disorders, 34, 7–17.
Motiwala, S. S., Gupta, S., Lilly, M. B., Ungar, W. J., & Coyte, P. C. (2006). The cost-effectiveness of expanding intensive behavioral intervention to all autistic children
in Ontario. Healthcare Policy, 1, 135–151.
 R. Stock et al. / Research in Autism Spectrum Disorders 7 (2013) 1168–1181 1181

Mottron, L. (2004). Matching strategies in cognitive research with individuals with high-functioning autism: Current practices, instrument biases, and
recommendations. Journal of Autism and Developmental Disorders, 34, 19–27.
Mullen, E. M. (1995). Mullen Scales of Early Learning (AGS ed.). Circle Pines, MN: American Guidance Service Inc.
National Autism Center. (2009). National Standards Report: Addressing the need for evidence-based practice guidelines for autism spectrum disorders. Randolph, MA:
Author.
Partington, J. (2006). The Assessment of Basic Language and Learning Skills – Revised: An assessment, curriculum guide, and skills tracking system for children with autism
and other language delays. Pleasant Hill, CA: Behavior Analysts Inc.
Perry, A., Cummings, A., Dunn Geier, J., Freeman, N. L., Hughes, S., LaRose, L., et al. (2008). Effectiveness of intensive behavioral intervention in a large, community-
based program. Research in Autism Spectrum Disorders, 2, 621–642.
Peters-Scheffer, N., Didden, R., Mulders, M., & Korzilius, H. (2010). Low intensity behavioral treatment supplementing preschool services for young children with
autism spectrum, disorders and severe to mild intellectual disability. Research in Developmental Disabilities, 31, 1678–1684.
Reichow, B. (2012). Overview of meta-analyses on early intensive behavioral intervention for young children with autism spectrum disorders. Journal of Autism
and Developmental Disorders, 45, 512–520.
Reichow, B., & Wolery, M. (2009). Comprehensive synthesis of early intensive behavioral interventions for young children with autism based on the UCLA young
autism project model. Journal of Autism and Developmental Disorders, 39, 23–41.
Roid, G. H. (2003). Stanford Binet Intelligence Scale – Fifth Edition: Technical manual. USA: Riverside Publishing.
Roid, G., & Sampers, J. (2004). Merrill-Palmer-Revised Scales of Development. Wood Dale, IL: Stoelting Co.
Sallows, G. O., & Graupner, T. D. (2005). Intensive behavioral treatment for children with autism: Four-year outcome and predictors. American Journal on Mental
Retardation, 110, 417–438.
Sautter, R., & LeBlanc, L. (2006). Empirical implications of Skinner’s analysis of verbal behavior with humans. Analysis of Verbal Behavior, 22, 35–48.
Scheinkopf, S. J., & Siegel, B. (1998). Home based behavioral treatment for young autistic children. Journal of Autism and Developmental Disorders, 28, 15–23.
Simeonsson, R. J., Bailey, D. B., Huntington, G. S., & Comfort, M. (1986). Testing the concept of goodness of fit in early intervention. Infant Mental Health Journal, 7,
81–94.
Skinner, B. F. (1957). Verbal behavior. New York: Appleton-Century-Crofts.
Smith, I. M., Koegel, R. L., Koegel, L. K., Openden, D., Fossum, K. L., & Bryson, S. E. (2010). Effectiveness of a novel community-based early intervention model for
children with autistic spectrum disorder. American Journal on Intellectual and Developmental Disabilities, 115, 504–523.
Smith, T., Eikeseth, S., Klevstrand, M., & Lovaas, O. I. (1997). Intensive behavioral treatment for preschoolers with severe mental retardation and pervasive
developmental disorder. American Journal on Mental Retardation, 102, 238–249.
Smith, T., Groen, A. D., & Wynn, J. W. (2000). Randomized trial of intensive early intervention for children with pervasive developmental disorder. American Journal
of Mental Retardation, 105, 269–285.
Sparrow, S. S., Balla, D. A., & Cicchetti, D. V. (1984). Vineland Adaptive Behavior Scales: Survey Interview Form/Caregiver Rating Form. Livonia, MN: Pearson
Assessments.
Sparrow, S. S., Cicchetti, D. V., & Balla, D. A. (2005). Vineland Adaptive Behavior Scales: Second Edition (Vineland II), Survey Interview Form/Caregiver Rating Form.
Livonia, MN: Pearson Assessments.
Spreckley, M., & Boyd, R. (2009). Efficacy of applied behavioral intervention in preschool children with autism for improving cognitive, language, and adaptive
behavior: A systematic review and meta-analysis. Journal of Pediatrics, 154, 338–344.
Stock, R. (2012). A comparison of a group ABA (GABA) verbal behavior model of early intensive behavioral intervention and Pivotal Response Treatment for children with
autism. Unpublished doctoral dissertation: University of British Columbia.
Sundberg, M. L. (2007–2008). Verbal behavior milestones assessment and placement program guide. Concord, CA: AVB Press.
Sundberg, M. L., & Michael, J. (2001). The benefits of skinner’s analysis of verbal behavior for children with autism. Behavior Modification, 25, 698–724.
Tomanik, S., Harris, G., & Hawkins, J. (2004). The relationship between behaviors exhibited by children with autism and maternal stress. Journal of Intellectual and
Developmental Disability, 29, 16–26.
Virués-Ortega, J. (2010). Applied behavior analytic intervention for autism in early childhood: Meta-analysis, meta-regression and dose-response meta-analysis
of multiple outcomes. Clinical Psychology Review, 30, 387–399.
Wechsler, D. (2002). Wechsler Preschool and Primary Scale of Intelligence – 3rd ed.: Technical and interpretative manual. USA: The Psychological Corporation.
Zimmerman, I. L., Steiner, V. G., & Pond, R. E. (2002). Preschool Language Scale – 4th ed.. San Antonio, TX: Psychological Corporation.