This time I will question the view that ABA/EIBI is an evidence-based autism treatment. Let me start by quoting what major mainstream authorities think of it.
Behavioral training, including communication development, has been shown to be effective in reducing problem behaviors and improving adaptation.(American Academy of Pediatrics, 2001)
There is a growing body of evidence that intensive early intervention services for children in whom autism is diagnosed before 5 years of age may lead to better overall outcomes.
Thirty years of research demonstrated the efficacy of applied behavioral methods in reducing inappropriate behavior and in increasing communication, learning and appropriate social behavior.(The United States Surgeon General, 1999)
Impressive, right? I will summarize reasons why some of us feel these statements were premature and not entirely supported by the data. Most of this has already been noted by Michelle Dawson and Morton Ann Gernsbacher, who have done a remarkable job of advancing science-based criticisms of ABA.
- There are no randomized trials of ABA (with or without blinded assessments) demonstrating its broad effectiveness against "eclectic" management approaches.
- One randomized trial of ABA (Smith, Groen & Wynn, 2000) was conducted to "address criticism of previous research and to increase methodological rigor." It compared a group treated with Lovaas-style ABA against a group receiving parent-instructed treatment. Only 13% of the children in the experimental group achieved "best outcome" (mainstream placement without support). No statistically significant group differences were found in either of two language scales. The paper reports a difference, but there was an error in the data analysis (Smith, 2001). No differences were found in socioemotional functioning or in adaptive functioning. Marginal differences were found in other measures.
- There are controlled trials, such as Howlin et al. (2007), which have not found EIBI to be effective compared to autism-specific nursery provisions.
- Adult outcomes of children who had received Lovaas-style ABA have not been reported in the peer-reviewed literature. Children from the experimental group of Lovaas (1987) must now be in their 20s and 30s. What would be the point of short-term gains if, hypothetically, adult outcomes do not differ from those of autistics who did not receive intensive interventions?
Specific Criticism of Lovaas (1987)It is probably fair to say that the reason ABA gained mainstream acceptance was the widely reported "recovery" rate of 47% resulting from Lovaas (1987). In fact, the Surgeon General of the United States makes reference to it. The methodology of this landmark paper has been widely challenged, however.
- There was no randomized assignment into groups.
- Assessments were not blinded.
- Male-female ratios in the experimental group and control group 1 differed from the ratio normally reported for autism (Boyd, 1998).
- IQ assessment tools differed between intake and follow-up.
- It is highly improbable that the reported ranges of intake IQ in Lovaas (1987) match the experimental group average IQ of 63 (source). In fact, the ranges are inconsistently reported.