Supported by a grant from the Norwegian
Research Institute for Children with Developmental Disabilities (NFBU).
Presented at the PEACH "Putting Research Into Practice" conference,
London, June 18, 1999.
Intensive, long-term behavioural treatment has been shown
to enhance the intellectual, academic, social, and emotional functioning of
children with autism. After receiving such treatment, a great majority of the
children have been able to take better advantage of the educational and social
opportunities available in their communities and have required less
professional attention as they have grown older (Anderson, Avery, DiPietro,
Edwards and Christian, 1987; Birnbrauer and Leach, 1993; Harris, Handleman,
Gordon, Kristoff, and Fuentes, 1991; Hoyson, Jamieson and Strain, 1984;
Lovaas, 1987; McEachin, Smith and Lovaas, 1993;Sheinkopf and Siegel, 1998;
Smith, Eikeseth, Klevstrand, and Lovaas, 1997). Moreover, some children have
benefited from the behavioural treatment to the extent that they have been
able to successfully pass normal classes in public schools, have moved from
the retarded range to the normal range on tests of intellectual, language,
social and emotional functioning, and have maintained their gains several
years after the treatment ended (Lovaas, 1987; McEachin et al., 1993).
A variable assumed related to the outcome of behavioural
treatment is early intervention. Researchers recommend that to maximise the
effectiveness of the program, treatment should be started before the child is
four years old. Unfortunately, many children with autism are referred for
behavioural treatment after this age. This may be partly due to a late
diagnosis: In a recent study, Howlin and Moore (1997) found that the average
age of diagnosis in the UK was six years.
Another factor that may contribute to a late onset of
behavioural treatment is the many misrepresentations and misconceptions that
exist regarding this treatment approach. For example, Maurice, a mother of two
children with autism described how professionals actually advised her and her
husband not to pursue behavioural treatment. Some professionals argued that
"behaviour modification is morally reprehensible" (Maurice, 1993,
p.66), whereas other failed to inform them of the existence of behavioural
treatment. Because of such professional advice, parents may pursue other
treatments before eventually starting behavioural treatment.
Unfortunately, there is limited research on the extent to
which children older than four years old may benefit from behavioural
treatment. However, one study compared treatment outcome of nine children with
autism who began behavioural treatment prior to five years of age to nine
children who entered the same program after five years of age. The researchers
found that the younger children obtained a considerably better treatment
outcome as compared to the older children (Fenske, Zelenski, Krantz, and
McClannahan, 1985), suggesting that age at intake may be an important outcome
factor. In contrast to this finding, Lovaas (1987) study to those children in
that study who did not achieve normal intellectual and academic functioning.
Although this finding may give some support for the notion that age is not a
crucial outcome variable, it is important to note that the participants of
Lovaas (1987) were less than 3 years and 10 months when the treatment begun (M
= 2 hears and 11 months), and thus, their failure to find a relation between
age at intake and treatment outcome may be valid only for this young age
group. Thus, more research is needed to examine the extent to which age ate
intake is related to treatment outcome for children who receive intensive,
long-term behavioural treatment.
In this report, we present preliminary data from an ongoing
study designed to evaluate the extent to which children with autism who are
between four and seven year old intake benefit from intensive, long term
All referrals who meet the following three criteria were
included in this study: (a) and independent diagnosis of autism based on the
ICD-10 (World Health Organisation, 1992), (b) chronological age between four
and seven years at the time of intake, and (c) a deviation IQ score of 50 or
above as determined by the Wechsler Pre-school and Primary Scale of
Intelligence-Revised (WPPSI-R; Wechsler, 1989) or ratio IQ score of 50 or
above as determined by the Bayley Scales of Infant Development-Revised
The children were diagnosed at a regional child psychiatric
unit. The diagnostic team consisted of medical doctors, clinical child
psychologists, speech therapists, and special education teachers. As a part of
the diagnostic evaluation, the children were given (a) a medical and
neuropsychiatric work-up consisting of general physical examinations,
neurological examinations, and laboratory examinations (b) diagnostic
instruments such as the Autism Dignostic Interview-Revised (Lord, Rutter, and
Le Couteur, 1994, conducted by a clinical child psychologist who had received
training in the clinical use of the instrument by Dr. Lord); the CARS
(Schopler, Reichler, DeVellis, and Daly, 1988); and/or the ABC (Krug, Arick,
and Almond, 1980). Finally, (c) the children's cognitive and language
functioning was assessed using theLeiter International Performance Scale, the
Reynell Developmental Language Scale, WPPSI, and/or the Bayley-II.
Participants were assigned to one of two groups: (a) a
behavioural treatment group receiving a minimum of 30 hours per week of school
based behavioral treatment, or (b) an eclectic special education group
receiving a minimum of 30 hours per week of school based special education
The director of the child autism team at the Regional
Habilitation Team performed the group assignment, and was independent of this
study. Participants were assigned to the behavioural treatment group
unless there was an insufficient number of behaviourally trained staff members
available to render this treatment. Participants not assigned to the
behavioural treatment group were assigned to the eclectic special education
Setting and treatment personnel. The
treatment of children in both groups was conducted in public kindergartens and
schools for regular children. During one-to-one instruction, the child
was working alone with the therapist in a separate room. When not
receiving one-to-one instruction, the child with autism was mainstreamed with
his/her regular classmates while being shadowed by the therapist. The
treatment personnel of both groups consisted of teacher's aids and special
education teachers working in the public kindergartens and schools.
Behavioural treatment. The treatment personnel
received a minimum of 10 hours per week of hands-on training and supervision
from the authors of this study and from project staff. Prior to this
training, a one-day workshop addressing the principles of behaviour analysis
was conducted. The treatment was based on behavioural treatment
procedures described and detailed in a manual (Lovaas et al., 1981) and
associated videotapes (Lovaas and Leaf, 1981). However, contingent
aversives such as those used by Lovaas (1987) were not employed because
alternative procedures have been developed presumably making aversive
procedures redundant. In brief, the treatment was designed to progress
gradually and systematically from relatively simple tasks, such as responding
to basic requests made by an adult, verbal and nonverbal imitation, labeling
objects, actions, and abstract concepts such as colors, size, and
prepositions. The treatment progressed further to more advanced programs
such as answering questions, conversation and making friends with peers.
The program also emphasized play and social skills, progressing from simple
toy play and parallel play, to more advanced skills such as symbolic play and
cooperative play. The program emphasized the implementation of
experimentally validated teaching approaches (cf. Newsom and Rincover, 1989;
Schreibman, 1988; Smith, 1993), based on operant conditioning principles such
as shaping, chaining, discrimination training, and behaviour management.
In the early stages of treatment, instruction took place in a one-to-one
discrete trial format, which enabled therapists to devote highly
individualized attention to each child. Later, the focus shifted
gradually to help children generalise skills to natural settings with regular
peers, adjust to classroom routines and settings, and to teach the child to
acquire new skills in such settings. Parent participation was considered
a key factor in the program. Parents received training in behaviour
management procedures and in how to promote generalisation and maintenance of
Eclectic treatment. All children in the control
group received eclectic treatment, using elements from a variety of different
teaching procedures such as TEACCH (Schopler, Lansing, Spitz, 1986),
behavioural procedures (Lovaas et al., 1981), as well as methods derived from
personal experience. This treatment did not emphasise the implementation of
experimentally validated teaching approaches, or the use of a discrete-trial
Assessment and Data Collection
All children were assessed at intake and 12 months after
treatment begun. A licensed, clinical psychologist with expertise in autism
carried out the assessment. The examiner was independent of the present study
and did not know whether children belonged to the behaviourally treated or to
the eclectic group. The intake assessments were carried out in the order in
which the children were referred, and the follow-up assessment was carried out
in the order in which the children had received one year of treatment. Thus,
behaviourally treated children and children receiving eclectic treatment were
assessed in a semi random order. Assessment occurred in the areas of
intellectual functioning, language functioning, and adaptive functioning. The
various assessment instruments are described below:
Intellectual Functioning. One of the following three
IQ measures was used. The particular test used was determined according to
each participant's chronological age and level of functioning, as follows. The
Wechsler Pre-school and Primary Scale of Intelligence-Revised (WPPSI-R;
Wechsler, 1989; 3 yr. - 7 yr., 3 MO) was attempted for each participant who
were below seven-years-and-three-months at the time of the assessment. If the
participant failed to achieve basal on the WPPSI-R (defined for the current
study as two 2-point responses on the vocabulary subtest), the Bayley Scales
of Infant Development - Revised (Bayley, 1993; 0 - 42 months) was given.
Participants older than seven-years-and-three-months at the time of the
assessment (which might be the case at follow-up) were given the Wechsler
Intelligence Scale for Children-Revised (WISC-R; Wechsler, 1974; 6 yr., 6
months - 16 yr., 6 months). If the participants failed to achieve basal score
on this test (defined for the current study as two 2-point responses on the
vocabulary subtest), the WPPSI-R was administered. For the WPPSI-R and the
WISC-R, a deviation score was obtained. For the Bayley, a ratio score was used
because the children's chronological age was higher than 42 months (i.e., that
of the norm group).
To assess visual-spatial IQ, participants younger than
six-years and six-months were given the Merrill-Palmer Scale of Mental Tests
(1 yr., 6 MO - 6 yr.; Stutsman, 1948). For children older than six-year
and-six months, visual-spatial skills were assessed using the performance
subscale of the WPPSI-R or the WISC-R.
Language Functioning. The Reynell Developmental
Language Scales (1 yr. - 7 yr.; Reynell, 1987) was used to assess language
functioning in participants less than seven years at the time of assessment.
Children older than seven years were given the verbal subscale of the WPPSI-R
or the WISC-R. The Reynell yields a receptive language score (age equivalence
score) and an expressive language score (age equivalence score) and a
deviation score for the overall receptive and expressive domain. In the
present study, age equivalence scores were used for the receptive and
expressive domains, and a standard score or a ratio score was used for the
overall score, as follows: a ratio score was calculated if the child scored
too low to obtain a deviation score, if not a deviation score was used.
Adaptive Behaviours. Participants adaptive skills were
assessed using the Vineland Adaptive Behavior Scales (0 yr. - 18 yr.; Sparrow,
Balla, Cicchetti, 1984). The Vineland consists of a Communication
domain, an Activity of Daily Living domain, a Social domain, a Motor domain,
and an Adaptive Composite score. Deviation scores were used in all instances.
Treatment Personnel and Treatment Goals
The treatment personnel of both groups had similar
educational background: approximately 50% had a three year degree in special
education or related fields, whereas the remaining 50% had one year or less of
such training. The treatment personnel of both groups set similar treatment
goals: both groups focussed on teaching language, play, social skills, motor
skills, activities of daily life, and verbal and nonverbal imitation, and to
reduce aberrant behaviour. The only significant difference in treatment goal,
as reported by the treatment personnel, was in the use of sign language and
alternative communication: Almost half of the eclectic group had this as a
treatment goal, whereas none in the behavioural group did so.
Fourteen children have entered the behavioural treatment
group, and 13 children have entered the eclectic special education group. The
mean chronological age at intake was five years and five months (SD = 11.31
months) for the behaviourally treated group and five years and six months (SD
= 9.98 months) for the group who received eclectic treatment. The mean intake
IQ was 62 (SD = 10.87) for the behaviourally treated group and 66 (SD = 14.50)
for the group who received eclectic treatment. Table 1 below exhibits the
intake data from the two groups. As can be seen, there were no significant
differences between the behavioural group and the eclectic group on any of the
11 intake variables (i.e., intake age, global IQ, performance IQ, language,
language receptive, language expressive, Vineland adaptive composite, Vineland
communication, Vineland daily life, and Vineland social). However, the
eclectic group scored higher than the behavioral group on 10 out of the 11
intake variables (p .01), suggesting that the eclectic group functioned
better than the experimental group at intake.
Twelve children from the behavioural group and 10 children
from the eclectic group have completed the one-year follow-up (the remanding
two children from the behavioural group and three children from the eclectic
group have not yet reached one year of treatment, and hence, follow-up data
are not yet available for these children). To examine whether the two groups
differed at follow-up, improvement made by the behaviourally treated children
and improvement made by the eclectically treated children was compared and
subjected to a t-test. The results are exhibited in Table 2 below and can be
summarized as follows: First, the behaviourally treated children scored
significantly higher than the eclectically treated children on global IQ,
language, language receptive, Vineland adaptive composite, and Vineland
communication. A Wilcoxen rank-sum test was applied to the same data and the
results of this nonparametric analysis did not differ from that of the t-test.
Second, the behavioural group scored higher than the eclectic group on 10 out
of the 10 outcome variables (p .001).
Intake and follow-up data for each behaviourally treated
child is exhibited in the following tables, Figures 1 through 3. Figure 1
shows the participants' overall IQ score on intake and at follow-up. As can be
seen, all children except one improved their scores on intellectual
functioning after one year of treatment, 8 of the 12 children obtained scores
within the normal range.
Figure 2 shows data from the language assessment. All
children improved their language score after one year of treatment, placing
four participants above or within the normal range of language functioning.
Finally, Figure 3 shows the results of the Vineland
Adaptive Scales. As can be seen, 10 of the 12 children improved their score on
the Vineland Adaptive Scales after one year of treatment. Improvement on
adaptive skills is consistent with the marked improvement on IQ and language
seen in Figures 1 and 2. Two participants scored within the normal range on
adaptive functioning at follow-up.
Summary and conclusions
This report presents preliminary data from an ongoing study
designed to evaluate the extent to which children with autism, who are between
four and seven year old at intake benefit from intensive, school-based,
behavioural treatment. A follow-up assessment was conducted one year after the
treatment started. The main results can be summarised as follows: at intake,
there were no significant differences between the behavioural group and the
eclectic group on any of the 11 intake variables. However, the eclectic group
scored higher than the behavioural group on 10 out of the 11 intake variables,
suggesting that the eclectic group functioned somewhat better than the
behavioral group at intake. At follow-up, the behaviourally treated children
scored significantly higher than the eclectically treated children on global
IQ, language, language receptive, Vineland adaptive composite, and Vineland
communication. Moreover, the behaviourally treated group scored higher than
the eclectic group on 10 out of the 10 outcome variables. Thus, after one year
of treatment, the behavioural group outperformed the eclectic group despite
the fact that the eclectic group scored higher than the behavioural group at
intake. Finally, at follow-up, eight of twelve behaviourally treated children
scored within the normal range on tests of intellectual functioning. All
children continue to receive treatment, and will be followed up after
finishing second grade (i.e., CA between 7 and 8 years).
Although the data are preliminary, they suggest that
children who are between four and seven years at intake may benefit greatly
from intensive, school-based behavioural intervention. Thus, this study
appears to be on its way to replicating and extending the key findings of the
earlier work of Lovaas and his colleagues. Lovaas (1987) found an improvement
of 19 IQ points on the average whereas preliminary data from the present study
shows an average IQ gain of 18 points.
There are several important differences between the Lovaas
(1987) study and the present study, however. For example, this study used a
school-based program, whereas Lovaas used a home-based program. The results
show that both models are viable. Perhaps a home-based program is most tenable
for the youngest children, whereas for school aged children a program in a
mainstream school is preferable.
Another difference pertains the use of aversives. In the
present study, no aversives were used, suggesting that the behavioural program
is effective without the use of such aversives. Finally, two models delivered
with the same intensity were compared, and one outperformed the other. This
suggests that the behavioural treatment is more effective than other
treatments even when the treatments are delivered with the same intensity.
R., 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.
2.Ayres, A. J.
(1972). Sensory integration and learning disorders. Los Angeles: Western
3.Ayres, A. J.
(1979). Sensory integration and the child. Los Angeles: Western Psychological
(1993). The Bayley Scales of Infant Development-Revised. San Antonio, Tx: The
J. S., Leach, D. J. (1993). The Murdoch early intervention program after
2 years. Behavior Change, 10, 63-74.
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,
L., Handleman, J. S., Gordon, R., Kristoff, B., Fuentes, F. (1991).
Changes in cognitive and language functioning of preschool children with
autism. Journal of Autism and Developmental Disorders, 21, 281-290.
Moore, A. (1997). Diagnosis in Autism: A survey of over 1200 patients in the
UK. Autism, 1, 135-162.
Jamieson, B., Strain, P. S. (1984). Individualised group instruction of
normally developing and autistic-like children. Journal of the Division of
Early Childhood, 8, 157-181.
10.Krug, D. A.,
Arick, J. R., Almond, P. J. (1980). Autism screening instrument for
educational planning. Portland, OR.: ASIEP Education.
Rutter, M., Le Couteur, A. (1994). Autism Diagnostic Interview-Revised:
A revised version of a diagnostic interview for caregivers in individuals with
possible pervasive developmental disorders. Journal of Autism and
Developmental Disorders, 24, 659-685
I. (1987). Behavioural treatment and normal educational and intellectual
functioning in young autistic children. Journal of Consulting and Clinical
Psychology, 55, 3-9.
I. Ackerman, A., Alexander, D., Firestone, P., Perkins, M., Young, D. B.
(1981). Teaching developmentally disabled children: The ME Book. Austin, TX:
I. Leaf, R. L. (1981). Five videotapes for teaching developmentally disabled
children. Austin, TX: PRO-ED
I., Smith, T. (1988). Intensive behavioural treatment for young autistic
children. In B. B. Lahey A. E. Kazdin (Eds.), Advances in clinical child
psychology (Vol. 11, 285-324). New York: Plenum.
(1993). Let me hear your voice. New York: Knopf.
J., Smith, T., Lovaas, O. I. (1993). Long-term outcome for children with
autism who received early intensive behavioural treatment. American Journal of
Mental Retardation, 97, 359-372.
Rincover, A. (1989). Autism. In E. J. Mash and R. A. Barkley (Eds.), Treatment
of Childhood Disorders (pp. 286-346). New York: The Guilford Press.
K. (1987). Reynell developmental language scales. Windsor, England:
E., Lansing, M., Waters, L. (1983). Individualized assessment and
treatment for autistic and developmentally disabled children: Vol. 3. Teaching
activities for autistic children. Austin, TX: Pro-Ed.
E., Reichler, R. J., DeVellis, R. F., Daly, K. (1988). The Childhood
Autism Rating Scale (CARS). Los Angeles: Western Psychological Service.
L. (1988). Autism. Newbury Park: sage Publications.
S. J., Siegel, B. (1998). Home-based behavioural treatment of young
children with autism. Journal of Autism and Developmental Disorders, 28,
(1993). Autism. In T. R. Giles (Ed.), Handbook of Effective Psychothetrapy
(pp. 107-133). New York: Plenum Press.
Eikeseth, S., Klevstrand, M., Lovaas, O. I. (1997). Intensive
behavioural treatment for preschoolers with severe mental retardation and
pervasive developmental disorder. American Journal of Mental Retardation, 102,
S., Balla, D. A., Cicchetti, D. V. (1984). Vineland Adaptive Behavior
Scales. Minnesota: American Guidance Service.
27.Spitz, H. H.
(1986). The raising of intelligence. New Jersey: Lawrence Erlbaum.
(1974). Wechsler Intelligence Scale for Children-Revised. San Antonio, Tx: The
(1989). Wechsler Preschool and Primary Scale of Intelligence-Revised. San
Antonio, Tx: The Psychological Corporation.
The authors are grateful to Jens Petter Gitlesen for
statistical advice. Preparation of this report was supported in part by a
grant from the Norwegian Research Institute for Children with Developmental
Disabilities (NFBU). This paper was presented at the PEACH 'Putting Research
Into Practice' conference, London, June 18 1999.