Introduction
Autism spectrum disorder is a neurodevelopmental disorder that significantly impacts
how one interacts with others and their communication learning, and behavior. It is estimated ,
that about one out of 44 children may suffer from some form of autism, and it occurs in all social
classes worldwide. Those with autism spectrum disorder are likely to have deficits in nonverbal
communicative behaviors making it difficult for them to interact with others and develop and
maintain relationships. The disorder can also cause people to have highly restricted or repetitive
behaviors or interests. The various symptoms must be present during a child's early
developmental period and must cause significant impairments to their daily life and other
important areas of functioning. The causes of autism are not clear, and prognosis is poor.
Prior to Lovaas’ (1987) study, medication and psychodynamic therapy were tried as
treatments, but neither showed benefits for autism. Behavioral treatments on the other hand
showed some improvement and helped enhance children's language and lowered behaviors of
aggression and self-stimulation. However, these improvements did not transfer to other settings
and relapse occurred once treatment had ended resulting in children not showing advances in
normalization. Lovaas (1987) conducted his study with the goal of further improving the positive
findings of former behavioral treatments. He hypothesized that if intensive treatment was
provided to autistic children at a very young age, then there was a higher possibility of achieving
normalization. This paper will investigate Lovaas’ (1987) study, the Young Autism Project, to
discover the strength of evidence for the normalization of function when children with autism are
given early and intensive behavioral treatment.
Sample
Lovaas’ study began with a total of 38 subjects with 19 subjects in both the control and
experimental group. The participants were all given an independent diagnosis of autism from
medical professionals or licensed psychologists who were not associated with study first utilizing
criteria similar to the DSM III and then the DMS III once it was released. The subjects were
required to be younger than 40 months of age if they didn’t have language and less than 46
months if they exhibited echolalia. All subjects were assessed for their mental age. The scores
were prorated, and the children had to have a mental age of at least 11 months at 30 months of
age. It was unclear how these subjects were recruited, but the participants who were part of the
experimental group had to live within one hour of UCLA. After two years a second control group
was added to assure that the children participating in the Young Autism Project did not have a
better or worse prognosis when compared to eachother. This group had 21 subjects that were 42
months old or younger when first diagnosed, had an IQ score above 40, and received follow-up
testing at 6 years of age. Control Group 2 was treated like Control Group 1 but did not receive
treatment from the Young Autism Project.
The study’s small sample size limited its external validity and made the generalizability
of the results poor. Having a larger sample size, however, would make the administration of the
intense treatment program difficult since a larger number of therapists would be needed. Another
issue that lowered the external validity included the lack of information regarding how they were
recruited. There was also no information on the socioeconomic status or ethnicities of the
subjects aside from a comment that the two groups were equal on those variables. A strength of
the study was how the subjects were given an independent diagnosis of autism by professionals
that had no association to the project. On top of this reliable diagnostic criteria was utilized
which was at first similar to the DSM-III and then criteria from the DSM-III once it was
released.
Procedure
Pretreatment measures included an assessment of the children’s intellectual ability. Four
different measures were used, and the evaluator chose the instrument that was most appropriate
for the child. These instruments measured mental age and were prorated to equalize all children
at about 30 months. If any of the subjects were unable to be tested a measure of adaptive
behavior was utilized. Children were also videotaped in a free-play environment and the tapes
were then scored by an observer who was unaware of the study’s hypothesis. The rater noted the
child’s play, language, and self-stimulation. Parent interviews were also conducted and gave
insight into the child’s earlier history which was then scored into a pathology measure.
Lovaas did not randomly assign subjects to the experimental and control groups. Instead,
the subjects were assigned on a first-come, first-serve basis and were placed in the experimental
group until they ran out of therapists. The remaining subjects or those who lived outside the one-
hour radius of UCLA were all placed in the control group.
For the actual treatment, the study utilized graduate students as the therapists, and several
were assigned to each child. The therapists worked with the subjects for about 40 hours a week
over the span of two years within their homes, schools, and communities. The parents of the
subjects also received training to apply the same methods at home and were expected to always
use the therapy while the child was awake. The subjects in the original control group received the
same treatment, but only for a total of 10 hours per week. The only difference in the methods
used between the groups was that the control group did not receive the physical punishment that
the treatment group was utilizing.
The subject’s treatment utilized operant conditioning and attempted to shape desired
behaviors and reduce or eliminate unwanted behaviors. For instance, behaviors that were not yet
present in the participants were developed and strengthened while behaviors that were undesired
were disregarded or given punishment. The first year of treatment was focused on teaching
participants how to properly play with toys through modeling and develop skills to better follow
directions of simple tasks. This year also worked on making sure that the participants’ parents
could emulate the treatment methods at home so that undesired behaviors can continue to
decrease. The second year of treatment worked on further evolving participants’ language and
communication skills to better interact with others in different environments. All the subjects’
diagnoses were concealed, and they were admitted into regular preschools for them to learn and
model the normal behaviors of their peers. The third year of the study’s treatment revolved
around improving preparatory skills for school, growing appropriate emotional expression, and
observational learning. The only two measures collected post-treatment were IQ and school
placement. The absence of dependent measures was shocking since there was such a broad range
of pre-treatment and treatment measures.
Overall, the study had low internal validity. The most glaring problem was that the
subjects were not randomly assigned to groups which was crucial to ensure that there were no
systematic differences between the groups before treatment is given. Because children were
placed on a first come basis, the early comers may have had more motivated parents. This could
have created a difference in outcomes since parent involvement was crucial and implemented in
the treatment. Another issue of internal validity was the use of four different measures to assess
intellect. There was no mention of attempts to determine if treatment was similar across all
therapists and that all aspects of the treatment were delivered. Lastly there was an absence of
commonality in the pre-treatment and post-treatment measures which affected the ability to
assess changes over the treatment period.
Results
Comparison of the pre-treatment variables exhibited no significant differences between
the experimental and control group on seven out of eight variables. The control group subjects
were about six months older than the experimental group subjects. According to Lovaas (1987)
the differences in chronological age was due to not having enough staff to provide therapy for
treatment therefore halting the recruitment processes for the control group. However, he claims
that the differences in chronological age have no significant relation to the outcomes. Although
there were no significant differences between the subjects at the start of the study, the post-
treatment comparisons showcased that the subjects in the experimental group attained
remarkably higher scores for IQ and school placement. In fact, out of the 19 experimental
subjects nine of them were accepted into a first-grade level class. They showed IQs in the
average and above average range of 94 to 120 with a mean of 107. Eight out of the 19 children
ended up being enrolled in language delay classes with IQs in the mild intellectual disability
range of 56 to 95 with a mean of 70. There were only two subjects that stayed behind in classes
for children with autism with IQs below 30. It was clear that subjects in the experimental group
experienced significant changes in their IQs. Those in the experimental group showcased an
increase of about 30 points. On the other hand, IQs in Control Group 1 only rose between two to
12 points and with seven subjects showing improvement and three remaining in the moderate-
severe range of intellectual disability. Lovaas revealed that the two control groups demonstrated
no significant improvement. Twenty-one of the subjects remained in classes for children with
autism and intellectual disabilities with an IQ range of 20 to 73 with a mean of 40. Eighteen of
the subjects were placed into classes for language impairment or mild intellectual disability and
had an IQ that ranged from 30 to 101 with a mean of 70. There was only one subject that
exhibited normal functioning adequate for the first -grade level with an IQ of 99.
References
Hart, W. (2023, March). Autism Spectrum Disorders and Childhood Onset Schizophrenia.
Lecture, ASU.
Hart, W. (2023, March). Lovaas’ Study of the Treatment of Autism Part 1. Lecture, ASU.
Lovaas, O. I. (1987). Behavioral treatment and normal educational and intellectual functioning in
young autistic children. , (1), 3–9. Journal of Consulting and Clinical Psychology 55
https://doi.org/10.1037/0022-006x.55.1.3
