Teaching Children with Autism
I am a graduate of the Clinical Psychology program at the University of Waterloo, steeped in the scientist-practitioner model. I understand research, and I can tell you that most published psychological research is crap, driven by the “publish or perish” zeitgeist of academia. In an earlier incarnation, when I was involved in correctional treatment research, I quickly discovered that criminals don’t improve with treatment so long as the scales that you use to measure improvement, such as the PD scale of the MMPI, are made up of mainly historical items. To measure improvement, you have to use measures that are amenable to change! I want to use this an an analogy as I discuss IBI research.
Now, one of the first research projects to examine the effect of IBI, unpublished so far as I know, determined that children, most of whom would have been 5+ years old, made about 15 months progress in language in 15 months, but only about 6 months progress in social and daily living skills. Why might that have been? Well, because they were being taught language rather than social and daily living skills! Think about that for a minute.
In 2010, there was published in the Journal of Developmental Disabilities an article “Examining the Effectiveness of Intensive Behavioural Intervention in Children with Autism Aged 6 and Older.” I intend to comment on this study in some detail, since it may have (mis)informed the Expert Panel on which the Government seems to be relying for support of their ill-conceived decision to eliminate children five and over from eligibility for participation in the IBI program.
“Results indicated that the children, as a group, did not show significant gains in IQ…, Adaptive level…, or cognitive rate of development…,which differs from finding in younger children. … Initial IQ and adaptive scores were significantly correlated with all outcome variables, and age was less stringly related. Children aged 6 to 8 had more variable outcomes, while older children displayed uniformly poor outcomes. These results have clinical and policy implications for appropriate service selection for older children.” (from the Abstract)
What the article did not examine was progress made on the skills that were being taught. In the paper itself, it was noted that :
“Eikeseth et al. (2002), in a small study (13 children in the IBI group) of somewhat older children, aged 4 to 7 at intake, with intake IQ of 50 or above, also found that age was not reliably associated with amount of change or absolute outcomes. Thus, in several efficacy studies with restricted age ranges, and initial age does not seem to be a strong determinant of outcome. However, there are no controlled studies of children all over 7 years of age.
The meta-analysis by Reichow and Wolery (2009) explored moderator effects of several variables and, based on an analysis of 251 children in IBI with the maen chronological age of less than 7 years at the beginning of IBI, concluded that chronological age did not significantly predict treatment outcomes. Howlin et al. (2009) arrived at similar conclusions based on 11 studies, with all children being younger than 7 years of age. However, Makrygianni and Reed (2010) conducted a meta-analysis of 14 studies and suggested that studies of children who began treatment very early (around 3 years of age) tended to have more uniformity large effect sizes where studies of children beginning later had more variable effect sizes.”
“In a recent large study with a wide age range, Granpeesheh, Dixon, Tarbox, Kaplan, and Wilke (2009) used the number of behavioural objectives mastered as their outcome measure. They studied 245 children ranging in age from 16 months to 12 years, receiving intervention to a large-scale community-based agency. They demonstrated that the child’s age has a significant impact on treatment outcome and that the efficacy of intervention decreases as the age of the child increases.”
In studies published by Perry et al. (2008, 2011), it was found that:
“… children who started IBI before age 4 showed better outcomes than those who started after age 4 on a variety of outcome measures. The authors point out that the results did not demonstrate that older children do not benefit from IBI, but that they are less likely to show highly successful outcomes such as average functioning (Perry et al., 2011).”
The authors go on to say that:
“… there is no empirical evidence demonstrating the effectiveness of IBI for children over 7…. Therefore, we believe it is imperative to examine the effectiveness of the Ontario IBI program for older children in light of the resources being devoted to it and the absence of an evidence-based for this age group.”
The measures used in this study were Cognitive level obtained from any standardized test available in the child’s psychology file, and Adaptive level as measured by one of the Vineland adaptive behavior scales. In discussing their results, the authods concluded that:
“Results indicated that the children, as a group, did not show statistically significant gains in IQ, cognitive rate of development, adaptive behaviour standard scores or age equivalent scores during that time they were involved in the IBI program.”
In contrast to what was observed in younger children:
“… in the present sample, pre-post changes were not significant significant overall and the gains that were seen were in adaptive behavior and mental age, not cognitive standard scores.”
This was tentatively explained in terms of a possible difference in focus for the IBI program at the different ages.
“The results regarding stronger adaptive behaviour gains than cognitive gains seems contradictory to the goals of IBI. There may be several reasons for this pattern of results. The focus of the curriculum for the lower functioning, older children may be on the functional communication and self-help skills that the Vineland measures, which may be very clinically appropriate for such children, but is not a typical goal or focus in IBI. Another reason may be that these adaptive gains are based on the Vineland, which is a parent-report measure. Parents perceptions of progress are not necessarily the same as changes measured by standardized tests…, although both may be considered valid in different ways.
This study has several limitations and short-comings, similar to those delineated in Perry et al. (2008). One major limitation is the inconsistency of the data collected. … Since this was a retrospective file-review study, we had no control or influence over the measure used to assess the children and the times at which children were accessed. Moreover, there were no measures of language or problem behavior available to us.”
“The study clearly has important clinical and policy implications. This group of older children, as a whole, did not show the same magnitude of progress as younger children typically show in IBI. Even within the age range studied, better outcomes were seen in the relatively younger children (aged 6-7) at intake, whereas children over 8, unfortunately, showed uniformly poor outcomes. This brings into question the appropriateness of these children for the IBI program. Furthermore, it is important to note that the gains that were made by individual children tended to be in adaptive behaviour more than in cognitive level.”
“The findings showing some individual children did make clinically significant gains on some measure imply that children’s progress should be evaluated individually since children who may benefit could conceivably be missed by simply considering group data. Having predetermined standards of expected progress at regular time intervals as recommended by Szatmari et al. (2006) and as delineated by Freeman et al. (2008) would provide a method of determining continuation in the program or, alternatively, transition to other services, which may be more appropriate and effective.”
Now, I want to return to the authors’ contention that “The results regarding stronger adaptive behaviour gains then cognitive gains seems contradictory to the goals of IBI.” I have been unable to find, in the on-line literature provided by Ontario’s Ministry of Children and Youth Services, any statement to the effect that the goal of the IBI program is to increase the child’s IQ. If it were, we would not be teaching to the preschool skills detailed in the ABLLS-R or the VB-MAPP or the social skills needed to overcome a child’s social communication deficit! And if reparation for participation in school and community is the objective, then there are thousands of single-case-study research reports available in the 5+ children’s files for evaluation.
In fact, if information readily available on the internet (e.g., Wikipedia – Intelligence quotient) is correct, trying to change IQ scores is a moderately futile venture, certainly compared to preparing the child to be a functioning member of sociaty. Not that it hasn’t been explored, See, for example, the work of Vygotsky and of Feuerstein on Dynamic Intelligence, but that is a topic for another day.