Title: Outcome Data from a Variety of Early Intervention Programs for Children with Autism
1Outcome Data from a Variety of Early Intervention
Programs for Children with Autism
- Paul Coyne - Coyne Associates, San Diego
- Nicholas Weatherly Richard Malott - Western
Michigan Unv.
- Carmen May Jonaitis - Croyden Ave. School,
Kalamazoo, Michigan.
- John Youngbauer - North Los Angeles Regional
Center.
2Outcome Data From an In-Home Early Intervention
Program Serving Children with Autism Younger than
3 Years Old
- Paul Coyne, Katherine Calarco,
- Len Levin, M. Alice Coyne
- Coyne Associates
- Data Analysis by Justin Prost Ph.D.
- Lumir Research Institute
3Coyne Associates
- In-home program started in 1999 funded by San
Diego Regional Center 2002 RC Orange County
2006 Inland Regional Center -San Bernardino
Riverside Counties. - Todays presentation covers outcome data from a
sample of children located within San Diego
Orange County.
- pdcoyne_at_mac.com
4Purpose of Early Intervention Programs
- Children who are at risk of developmental delays
be placed on a normative developmental trajectory
and thus continue to show optimal development
after early intervention ends. - Ramey Ramey 1998
5Literature ReviewIntensive ABA Early
Intervention Programs 25-40 hours per week
6From Gina Green Applied Behavior Analysis in the
Treatment of Autism 2007
7Better Treatment Outcomes
- Treatment was intensive 20-30 hours week.
- Ample use of incidental teaching outside of
sessions.
- Delivered 11
- Treatment was comprehensive.
- Used DTT and NET
- Supervisors teachers had advanced training in
ABA.
- Training Program followed normal developmental
sequences.
- Parent Participation.
- Home and Community.
- Two or more years.
8Literature ReviewLow Intensity ABA Early
Intervention Programs Less than 20 hours per
week.
9Controlled Studies of Low-intensity ABA
From Gina Green, 2007.
10Low vs High Intensity Professional ABA Programs
- Low Intensity
- Gains in IQ (7-8.2), language, and social skills
but smaller gains than intensive ABA. - 13.6 in
the normal range.
- Larger gains than controls and eclectic services.
- High Intensity
- Large gains in IQ (16-29.7), language, social
skills and other areas. 47 -normal range
- Parent managed programs had mixed results.
11More information is needed
- Intensity of intervention outcome.
- Duration of intervention outcome.
- Type of early intervention outcome.
- Type of children outcome.
- Onset of intervention outcome.
12Our Questions
- Can a 12 hour a week (54 hours a month) produce
improvement in young children with autism?,
compared to a control group?, to results
published elsewhere? - If so, where are those improvements seen?
- What about children who receive more or less than
a 54 hour a month program?
- Do the children continue to improve the longer
they are in the program? Are the effects
cumulative over time?
13Description of the Early Intervention Program
14Assessment
- Bayley Scales of Infant Development II
- Brigance Inventory of Early Development
- Verbal Behavior Assessment Form Level 1 Sunberg.
- Probe skill areas to determine placement.
15Brigance Inventory of Early Development II
- Preambulatory Motor Skills Behaviors
- Gross Motor Skills and Behaviors
- Fine Motor Skills and Behaviors
- Self Help Skills
- Speech Language
- General Knowledge Comprehension
- Social and Emotional Development
- Academic Readiness
16Curriculum
- Teaching Individuals with Developmental Delays
Lovaas 2003
- Behavioral Intervention for Young Children with
Autism Maurice et. al., 1996
- A Work in Progress Leaf McEachin 1999
- Give Your Child a Superior Mind Engelmann 1966
- PECS Training Manual Frost Bondy 2002
- Language for Learning Engelmann (2000)
17Curriculum
- Assessment of Basic Learning Abilities (ABLA)
Martin, Yu, 2000
- Eye Contact (Levin 2006)
- Level 1. Imitation
- Level 2. Position Discrimination
- Level 3. Visual Discrimination
- Level 4. Visual Match to Sample
- PECS Verbal Operants Mands
- Level 5. Auditory Discrimination
- Auditory Match to Sample
- Vocal Verbal Operants Echoics
- Level 6. Auditory / Visual Discrimination
- Vocal Verbal Operants Echoics, Mands, Tacts,
Intraverbals.
18Teaching Techniques
- Eye Contact Procedure (Levin 2006)
- Discrete Trial Teaching
- Errorless Teaching (Touchette Howard 1984)
- Time Delay Procedure (Heckaman et. al. 1998)
- Error Correction Procedure (Rodgers Iwata
1991)
- Natural Environment Teaching
- Pivotal Response Teaching (Koegel Koegel 2006)
- Incidental Teaching (Obrien et. al. 1979)
- Mand Training (Sunberg Partington 1998)
- Independent Activity Schedules (McClannahan
Krantz 1999)
19Staff Training
- 60 hours of initial hands on in-service
training with a 90 day review then reviews every
6 months.
- Mastery criterion for advancement. Pass a test.
- Minimum of 2 hours of Supervision each week (2
hours for every 10 hours of instruction).
- 10 additional in-service trainings per year.
20Methods
- Investigate the effectiveness of an In-Home early
intervention program for children with autism
younger than 3 years old using the following
outcome measures - Bayley MDI
- Brigance Receptive Language Composite,
Expressive Language Composite, Self-Help Skills
Composite, Social Skills Composite, and General
Communication Composite. - Regression analyses were conducted in order to
identify the relationships between the outcomes
measures and participation in the program.
21Participants
- For the analyses presented there were two
samples
- 42 children that completed pre- and post-test MDI
measures and participated in the intervention
program. 18 children that completed an MDI
measure, but did not participate in the program
(control). Total N60. - 186 children that competed pre- and post-test
Brigance measures and participated in the
program. 20 children that completed the Brigance
measure, but did not participate in the program
(control). Total N206. - The samples were taken from children
participating in the early childhood intervention
for approximately 54 hours per month during the
years 2000 to 2006.
22Variables Included in Analyses
- Participation Duration The amount of time in
months a child participated in the intervention
program.
- Average Adjusted Gross Income The average AGI
for each participant was identified using the
zipcode of each participant and census records.
- Bayley Mental Development Index
23Variables Included in Analyses
- Brigance Outcomes There were five composite
scores computed from the instrument.
- Expressive Language Expressive Colors,
Expressive Body Parts, and Expressive Vocabulary
- Receptive Language Receptive Colors, Receptive
Body Parts, and Receptive Vocabulary
- Self-Help Skills Undressing, Dressing,
Feeding/Eating, Toileting, Bathing, and Grooming
- Social Skills General Social, Play Skills
- General Communication Gestures, Vocalization,
General Speech, Sentence Length, Personal Data
Response, and Verbal Directions
24Bayley Sample DemographicsMeans (SDs)
25Brigance Sample DemographicsMeans (SDs)
26Hierarchical Regression Analyses
- Hierarchical regression analyses were conducted
to determine the relationship between
participation in the intervention program and the
outcomes of interest. - The regressions involved two steps of entry
- For the Bayley In the first step the pre-test
scores were entered, and for the Bayley analyses
average income was also entered. For the
Brigance Age at post-test was entered. - In the second step Participation Duration was
entered.
- This method provided a means of identifying the
unique contribution of Duration in the Program.
27Results of the Bayley
- The results of the regression analyses with the
Bayley MDI outcome showed
- Average income was not related to changes in MDI
scores.
- Program Duration was significantly related to the
changes in MDI scores.
- The R2 change for the step in which Program
Duration was introduced was .271. indicating
program duration explained 27 of the variance in
MDI scores, controlling for pre-test and average
income. - The unstandardized regression coefficient was
2.1, indicating an increase of approximately 2
points on the MDI for each month spent in the
program.
28Results of the Bayley MDI Regression
29Results of the Brigance
- The results of the regression analyses with the
Brigance outcomes showed the significant
relationships with program duration
30Results of the Brigance Regressions
31Effects of Fewer Hours per Month
- Additional analyses were conducted to determine
the impact of less time per month in the program.
- A regression with children receiving less than
the modal number of hours was conducted and then
compared to results from the full or larger
sample. - For the Bayley the samples compared were the 54
hours per month participants and the less than 54
hours per month. The smaller sample size of the
Bayley, prompted using only the 54 hour per month
children in the initial analyses. Although the
comparison group is small (N8), it merits some
consideration in support of the Brigance findings.
32Effects of Fewer Hours per Month
- Additional analyses were conducted to determine
the impact of less time per month in the program.
- A regression with children receiving less than
the modal number of hours was conducted and then
compared to results from the full or larger
sample. - For the Brigance the samples compared were the
full sample of children participating in the
program and the sub-sample of children receiving
less than 54 hours per month (the most common
participation level in the program).
33Effects of Fewer Hours per Month
- Results indicated that
- For the Bayley, the slope was still significant
for the participants with less than 54 hours per
month, but did not represent nearly as much
improvement as those in the program for 54 hours
per month. - For the Brigance, the sample size was 32 for the
participants with less than 54 hours, only Social
Skills had a marginally significant relationship
to Program Duration (p.062). - The smaller sample also had higher intercepts
than the full sample, indicating that those
children brought in for less time per month might
have lower levels of impairment.
34Effects of Fewer Hours per Month
- Results indicated that
- For the Brigance, the unstandardized regression
coefficient for 3 out of the 4 areas was larger
for the sample with more hours per month in the
program.
35Regression Slope Comparison for BMI
36Regression Slope Comparison for Brigance
Self-Help Skills
37Regression Slope Comparison for Brigance Social
Skills
38Regression Slope Comparison for Brigance
Communication Skills
39Regression Slope Comparison for Brigance
Receptive Language Skills
40Interpretation of the Results
- The results provide evidence for developmental
improvements in multiple areas for children
spending approximately 12 hours/week in the
intervention program. - Improvements were seen in scores on the Bayley
and in composite scores calculated for four areas
on the Brigance.
- Results indicate discriminant validity, as seen
in the lack of improvements seen in expressive
language skills, which the Brigance was designed
to identify in children older than 3-years-old. - Greater improvements are seen the longer children
spent in the program before testing at
3-years-old.
- Improvements were found to be greater for
children spending more time per month in the
program, although the children spending less time
did tend to show less impairment in general.
41Conclusions
- Intervention programs for children can be
effective at 12 hours/week.
- The developmental impact of the program in the
current study can be seen in multiple areas and
was found using two separate measurement
instruments. - Early intervention is important, particularly
with programs in which children spend less time
per week, as greater improvements were found the
longer a child had spent in the program.