Wednesday, August 27, 2014

Kim Bodner's RPM vs. Wechsler Study published. What does this mean for neurodiversity?

As many in the autism community know, about seven years ago, Michelle Dawson published a study suggesting that the intelligence of autistics had been underestimated.  She compared the scores of autistics on the Wechsler IQ test versus scores on the Raven's progressive matrices.  On average, the autistics' scores on the Raven's were considerably higher than their Wechsler.  By comparison, a control group of  non-autistic people had similar scores on the two tests.  The neurodiversity movement made much of this, claiming this proved that autistics' intelligence had been underestimated and it showed that "autistics were not write-offs" (Dawson's words) after all.  The media reported that intelligence had been underestimated in autistics.

Some years after this, I was interested to find out that two different research studies had seemed to fail to replicate her results.  Both of them had been presented at IMFAR (International Meeting for Autism Research) but had not yet been published in peer reviewed journals.  I read the brief presentations, but I was still uncertain about everything the studies would say if they were published in a journal.

The first one had been published by Sven Boelte.  I wrote to Dr. Boelte asking him if I could have a copy of the study  and he graciously emailed me one.  Like Dawson, he found that autistics scored higher on the RPM than on the Wechsler, but the magnitude was not as pronounced as in Dawson's study and it was only found in those with IQ's lower than 85 and not at all in autistic persons with higher measured intelligence.  In contrast to Dawson's optimistic statements and those of the pro-neurodiversity media, he gave the more guarded response that it was still too early to tell if intelligence in autistic people had been underestimated and more work needed to be done on the issue.

The other study was done by Kim Bodner, who I believe at the time was a graduate student working in Nancy Minshew's lab.  From reading about the IMFAR presentation, she seemed to be saying that she had found no differences in the scores of the RPM vs. the Wechsler in individuals on the spectrum with normal or near normal intelligence.  Though her study had not been formally published, I also wrote to her and cc'd a copy to Nancy Minshew.  I never received a response from Bodner, but Dr. Minshew wrote me saying she forwarded a copy to Ms. Bodner and they'd let me know when the study was published.  This was over four years ago.

Today, I was surprised to receive an email from Ms. Bodner out of the blue, sending me a copy of the .pdf study.  It took awhile, but eventually her study was published.  I'd wondered if it would ever be published and I found out today.  For anyone inclined to read the entire study, I've uploaded it here

She had studied four groups--autistic children (up to age 16) vs. age matched controls and adults on the spectrum (17 and up) compared to a control group.  She found that in the children with autism the RPM scores were higher than their performance IQs and slightly (but not much) higher than their verbal IQs.  The RPM's of the autistics were higher than the age and Wechsler matched controls but not much so.

The more striking difference was in the adult group where the RPM of the autistics was much lower than their Wechsler  scores and the scores of the RPM of the adult control group.  These findings don't seem to replicate Michelle Dawson's.

It must be emphasized that Bodner's study had the limitations in that there were only two girls in the 37 person children with autism group and 5 girls in the 48 person typical group.  The adult group (where the RPM was low compared to other measures) had a 25:6 male:female composition of the 31 autistic subjects which is probably representative of the ratio in the general population of those on the spectrum.  It would be curious as to how a representative ratio of males to females in the children's group would fare on the RPM.  Also, the autistic subjects were all relatively high functioning (meaning they had IQs in the average or near average range).

The results of various studies support that on average there are higher scores on the RPM than on Wechsler for lower functioning autistic children who have lower than average intelligence and less verbal ability.  But for the higher functioning more verbal autistic children, there does not seem to be any advantage on the RPM.  

For adults on the spectrum, the findings of various studies are less consistent.

If there is any advantage at all on the RPM, it is probably for the lower functioning children who would probably have life problems no matter what intervention or educational strategy is used.  I believe this is more evidence that the ND movement is mistaken that certain tests or abilities are not only superior in all persons with autism, but that this proves they have special strengths in jobs or any other endeavor. 

1 comment:

Roger Kulp said...

I can't help but think of how I used to test before the leucovorin,B12,etc.I had normal or above normal language,and select areas of intelligence,but everything else,as far as adaptive and practical behavior I tested in the profoundly intellectually disabled range.A breakdown of the different areas of intellectual disability is described here

I suspect such wide discrepancies in one person is a feature of mitochondrial/metabolic autism.In most people it seems to be all or nothing as far as ID is concerned.