"Indistinguishable From Their Peers"

How many times has the claim 50% (or close to) “indistinguishable from their peers” been beamed across the ether in pride of place in every call to parents and governments to support ABA as the intervention of choice with autistic children? How many know that the claim is misleading at best and probably better characterized as reprehensible? Everybody knows of Lovaas’ famous 1987 study that the claim springs from and the follow-up study from McEachin in 1993 that enshrined the claim in marble. This was based on McEachin’s PhD thesis and according to Victoria Shea who took the time to read the thesis, the reality is somewhat different. Make that a lot different. Shea’s article highlights a number of interesting things about the original 1987 study and the follow-up that I wasn’t aware of, among which:

· The ‘1987’ study was really conducted from 1970 to 1984 and not published until 1987. Seven of the 19 experimental subjects were referred to the UCLA Young Autism Project between 1970 and 1974. The follow-up study took place between 1984 and 1985 and was not published until 1993.

· One of the ‘best outcome’ students went into special education and was no longer considered ‘normal functioning’. True, another of the experimental kids was mainstreamed after the sixth grade but as Shea remarked , attributing that to what happened in pre-school might be stretching things quite a bit.

· The individual data for the Vineland and PIC is usually ignored, but here is the crux of the ‘indistinguishable from peers’ claim. Quoting Shea: “ there was only one ‘best outcome’ youngster who had all tests in the normal range.” Eight had at least one score in the clinically significant range and most had more than one.

· Apparently as part of the ‘indistinguishable from peer’ claim, they used a Clinical Rating Scale, never before seen in psychology and apparently not since either. The scale was administered on the basis of a 20 minute interview by a psychology graduate student; no blinding, no careful probing but also no evidence of reliability or validity for the scale in the first place.

· No measures of the opinions of teachers or peers were ever reported. So the claims of Lovaas that teachers found the ‘best outcome’ children indistinguishable from peers or by McEachin that teachers and peers did not see the ‘best outcome children as having unusual problems or being different are way beyond the data.

· Shea also reports some curious remarks of Lovaas on the possibility of reproducing the 47% “ indistinguishable from peers” result. It’s a sort of sliding scale. At the top, if the intervention is administered precisely as the original YAP and by UCLA clinically trained therapists, then 47% should be reached. If the training was provided to the therapists at UCLA but through a workshop setting, then Lovaas envisaged a 20% success rate. At the bottom of the scale, providing training to therapists at just any old workshop and the possibility of success according to Lovaas “would be less than 10%”. I wonder what Lovaas made of the Sallows and Graupner study results, which reversed this order in practice? “Less is Beautiful” sums it up.

So there it is. Behaviourists are definitely a glass half full kind of people. There is no study on the face of the earth, which cannot be reported in the most optimistic terms.

Caveat lector !

Reference: Shea, V. (2004). A perspective on the research literature related to early intensive behavioural intervention (Lovaas) for young children with autism. Autism, 8(4), pp. 349-367.