Sunday, January 11, 2009

Michelle Dawson's research: Intelligent Autistics or Neurodiversity Propaganda?

One of the tenets of the neurodiversity movement is that autism is not at all a disease or horrific handicap, but just an alternative form of brainwiring. Some members of the ND movement concede that autism is a disability. However, on closer scrutiny, it would seem that they differ between the medical model of disability versus the social model of disability. This implies that if only society would change and be more accepting of autistics and/or accommodate their unique way of being, they would not be failures or have poor outcomes.

Until recently, any scientific evidence for this tenet of ND seemed to be saliently lacking. About a year and a half ago, some cognitive scientists in Canada published a study which suggests that intelligence in autism has been underestimated due to the fact they are given the wrong test, namely the Wechsler IQ test. Another test, the Raven’s progressive matrices, they say, is a more valid test. Some autistics who did poorly on the Wechsler scored substantially higher, in some instances much higher, on the Raven’s matrices. Two groups of autistics (one children’s group and an adult group) were compared with two non-autistic control groups. It was found that the difference in the test scores of the controls between the Wechsler and on the Raven’s was negligible.
In my own particular case, I am not sure if I have ever taken the Raven’s matrices or how well I would do on this test. The Wechsler is composed of two different tests, verbal and performance. Each of these two tests in turn is composed of six subtests which I have varying abilities in. I score above average on vocabulary, superior in general knowledge, arithmetic and a test of abstract reasoning called similarities. On the comprehension subtest which has to do with social skills I do not do as well. On some of the subtests of the performance test I score about average on a few others I score in the severely retarded range. Interestingly, one of the subtests I do poorly in is the block design test in which many autistics excel in. As part of my brain damage I have perceptual motor problems which impair my fine motor coordination, handwriting and ability to put together puzzles.
The study, has one of neurodiversity/autism hub’s own bloggers as the lead author, Michelle Dawson. The results of the study has prompted a few neurodiversity proponents to give this study a shout-out claiming that autistics are not really stupid or retarded (in the cases of autistics who score in the retarded range on certain tests of intelligence) .
Autism Diva writes:
So they test a "retarded" autistic using the wrong tests and then expect them to be retarded for the rest of their lives, shove them in the retarded bin and forget about them, not find their talents and use them so they can be "successful" in a different way.
Alyric writes:
One of the participants scored the bottom of the bottom on every test except for the Raven at which he excelled. No question this person was and is severely impaired absolutely no question about it. But what are we to do with the equally undeniable fact that this person also excelled on the premier test of fluid intelligence? There’s a very big question mark, right there along with all the particularly autistic people, maybe people like you.

Alyric then goes on to state that perhaps if the special ed system had the results of Dawson’s findings, I would have been able to be mainstreamed as a child and that I should be grateful for Dawson’s research because it could help people like me and possibly autistics could function normally or at least better in society based on Dawson’s findings. Since my scores on the Wechsler are so different than those of Dawson’s research subjects it shows that she (Alyric) does not understand my situation.
Joseph writes: I can tell you for a fact that Michelle’s work is being cited by parents in situations where the school system is underestimating the cognitive potential of autistic children. I won’t give details but it is.
The researchers themselves also give implications of their findings to the media:
Lauren Mottron writes:
If we label those children as below-normal in intelligence that is how they’re treated. The challenge is to coax that spark into the kind of intelligence that manifests itself in practice.
Dawson herself writes:
I hope this paper will have the practical effect of making it less likely that autisics will be written off, as autistics are so routinely written off by autism advocates in Canada and elsewhere.
Let us look at the aspects of the study itself: One of the questions is, are the subjects truly representative of autistics who might be classified as intellectually retarded? First off, Dawson excluded any autistic who had a known genetic problem and only included persons with idiopathic autism whose only diagnosis would be autism rather than asperger’s. This means that persons with fragile X, tuberous sclerosis, etc. were excluded. This could mean as many as 10% of all persons diagnosed as autistic, most of whom might actually fall in that retarded range. Also, the group of autistic children is comprised of 35 males and 3 females- a 12 to 1 ratio. The adult group of 11 males and 2 females is a 6:1 ratio. The ratio of males to females in the general population of autistics is 4:1. As noted by Boyd in his critique of Lovaas (1987) sexual inequivalencies in autism research are relevant because there is research showing that female autistics as a group are often substantially more severely autistic than males. Therefore, it would appear that Dawson’s sample excluded a substantial number of subjects who might be classified as low functioning.
The composition of the control groups is also interesting. In the children it is mostly males. The adult control group consists entirely of males. The study says they were recruited from an advertisement in a local newspaper. I wonder if it is possible that this newspaper is Dr. Mottron’s favorite French Canadian porno rag. If there is any reason for excluding females in one of the control groups and having very few in the others it is not given. Therefore, the comparisons between autistic subjects and controls is not representative of either group. as it is questionable how people picked at random from a newspaper ad would be exclusively males.
The scores on the Wechsler are not reported in a traditional sense, i.e. below 70 being retarded above 100 being above average, but rather as percentiles. Individual scores are not given on the Wechsler IQ or any of the subtests. The autistic adult group on the average scored at about the 50th percentile on the Wecshler, so clearly as an average range they were not retarded at all, so applying the statements of the study to the adult group is certainly misleading. The group of autistic children’s average percentile was 31st in the verbal portion of the Wechsler and 26th in the performance. When expressed as percentiles it would seem that this is a disparity between 50th percentile, but, as Dawson states herself in the study, this is in the low average range rather than a retarded average. The percentile in the verbal IQ this might mean an IQ in the low 90s on average as opposed to an average IQ of 100, so on average, this was clearly a higher functioning sample than one who would be composed of many members testing at a level of mental retardation. The stated average percentile of 31st is also misleading when looking at Dawson’s graph of the average of the subtests in the verbal portion of the Wechsler. The Verbal Wechsler is composed of six subtests. For some reason Dawson either excluded the digit span test from the graph or it was not given to the subjects. On two of the other five subtests we see that the average score in the comprehension subtest is below 20th percentile and the vocabulary is also on the low side. Yet on the other three subtests, which measure academic skills, similarities, general knowledge and arithmetic, we see that the average percentile is around 40th, showing that as a group these autistics were nearly average in IQ if we just included the results of these tests. In the block design test, the average was 60th percentile with other subtests of the performance section of the Wechsler being substantially lower. I have mostly focused on the verbal section rather than the performance section because I feel it is more relevant for assessing the children’s academic potential. Also the very low average on the comprehension test is relevant to this discussion but more about that later.
When we look at Dawson’s graph on figure 3 of the study we see that 10 of the 38 children scored 50 or more percentile points on the Raven’s than on the Wechsler, only about 25%. Five of the 38 children had scores that were as high or higher on the Wechsler than on the Raven’s matrices. The remaining 23 children scored higher on their Raven’s percentile than the Wechsler, but in a number of cases it was only fractionally higher, e.g. 30th percentile as opposed to 25th percentile or 70th percentile as opposed to 65th percentile, showing no significant increase in intelligence in one test versus another. It is possible those children could have had a headache or not been feeling well the day they took the Wechsler, than on the day they took the Raven’s and this could account for the small disparity in some of the cases.
As I said before, even if Dawson wants to claim that the Raven’s as a measure of fluid intelligence shows that autistics have the potential for good outcomes outside of negative societal constraints, she will have the explain the poor results on the comprehension subtest as it applies to society. The low average on this test indicates that these autistics would do poorly in social situations, not know what appropriate action to take in certain instances and would apply to them negatively in school and job and other situations. For this reason the results of the comprehension subtest in some aspects may be more valid an indicator of their functioning level than either the Raven’s or Wechsler as a whole.

Even assuming I am somehow mistaken and what I believe to be Dawson's questionable science is sound, the question is, how can the knowledge that autistics do well on tests of fluid intelligence be applied so that they can have better outcomes, how can the educational situation be modified so that the autistics will receive a better education with this knowledge that we have? As is typical, the neurodiversity movement fails miserably in addressing this issue.

Neurodiversity seems fond of spreading stereotypes and urban legends about persons with autism that have no basis in reality. I believe their claim that the results of Dawson’s research showing that autistic people who function at a retarded level have the potential to perform normally is one of them and yet another of the pieces of propaganda that they stoop to in order to promote their misguided agenda.

23 comments:

Autism Reality NB said...

Jonathan

Thank you for your thoughts on this subject.

My son has "profound developmental delays". He is diagnosed with autistic disorder. He was diagnosed at age 2. 11-12 years ago I knew nothing of autism and there was not much about it in the media. I had not even seen "the Rain Man".

My son's early diagnosis at what was an early age 11 years ago arose from our concerns as early as 12 months and deepened by 18 months, that he was simply not showing expected development.

The lack of development of speech, demonstrated understanding of simple speech, lack of responsiveness to the usual infant games like peek-a-boo and some unusually persistent behavior like sifting sand for unusually long periods of time, lack of play with toys etc.

As he grew Conor has shown serious deficits, not on particular tests devised by researchers but in real life tests like understanding the dangers of automobile traffic.

The work of Dawson and Mottron does nothing to actually help autistic children like my son with severe deficits. It simply further promotes their personal agendas.

Stephanie Lynn Keil said...

With me, I actually do score much higher on non-verbal tests than verbal tests (which is in the retarded range), but this couldn't me measured until I was 18, until I actually understood the significance of taking an IQ test.

In the book "Exiting Nirvana," the mother of Jessy Park also states that Jessy scores in the "genius" range on Raven's Progressive Matrices. But Jessy is also a savant. And so am I. Stephen Wiltshire's non-verbal IQ is also much higher than his verbal IQ. So maybe savants are the ones who have a superior non-verbal IQ compared to a low verbal IQ and non-savants are more average. I don't know, I am only speculating, but it definitely makes sense that an autistic savant would have a superior non-verbal IQ compared to a verbal IQ.

SM69 said...

I agree with you that the issue of measuring intelligence in autism is very controversial. I have yet to find any benefit of doing such test, essentially for the reasons you have stated, but also from a broader perspective, I do not think we know what exactly defines intelligence, let alone how to measure it. Autistic people are perhaps the best examples to illustrate how intelligence can take many forms and that superiority and deficit commonly coexist in one single individual.

From a practical point of view, I suspect that testing intelligence in autism is essentially driven by service providers, not for the purpose of understanding the person and his or her needs, but rather to gather evidence that alternative placements should be provided. This is particularly true of a child within an education system, usually, who become probed in all sort of ways, in order to justify a lesser quality placement. A placement that is judged as being best from the authority’s point of view. What else but an intelligence test can guaranty a poorer outcome, hence to fail someone?

If anyone is interested to know about a person, there is no primary need for a battery of assessments in hands. But rather observation skills, knowledge, but also a very open mind and kind view of disabled people. To spend time, play, and experience some situations to appreciate the strengths and difficulties the person can experience. However, these accounts are not standardized, not measurable and potentially flawed, but when you have someone good with this range of skills, make sure that person can have influential decisions on what will be decided for the autistic person.

Other than that, I would like to add, that in my opinion, there are more girls with an ASD diagnosis than previously thought and I would disagree with you when you say that girls are more affected than boys. Girls with autism can instead have rather more subtle symptoms, because they have a stronger innate ability to relate to others, hence partly masking their social difficulties. This is something that comes more and more clearly as the Autism Diagnosis Observation Schedule (ADOS) is becoming progressively adjusted as a tool to measure autism across a wider span of personality traits and ages. Also girls with autism can be more compliant, so their behaviors do not stick out that much. Commonly, they would be good at copying others to model and guide their own behavior, and it is only with more subtle and informed observations that one can understand what exactly is going on in that type of behaviors.

jonathan said...

To date, there is no empirical evidence that autism is more common in females than previously thought. The ratios have been found in a variety of studies. Also, these same ratios exist in other developmental disabilities, e.g. ADHD and stuttering, which suggests that autism in females is not being underascertained.

As to the severity of females with autism versus males, there have been a number of studies on that though I don't know the specific ones, but they are quoted in Robert Boyd's critique of Lovaas (1987) which was published in an issue of JADD

SM69 said...

J

If you are aware of what is being published on the boy: girl ration in autism, you will know that this issue is also controversial right! You mentioned a study that is very old, remember that there is a better understanding today of what autism is, and of the boarder spectrum. The assessments tools are also becoming more and more refined. I had reported what we observed in today the clinical world, the ratios in our population of children, at our Trust, and also discussion, yet unpublished amongst professionals who have developed the ADOS, such as C Lord and UK colleagues. They are really looking into this issue closely, and it may well be that we have a much higher ration than previously thought, which would be one more argument against SMB’s testosterone theory. In any case, you do not have to take what I say as of any value, if you do not want to, let’s see how this will be reported in the next few years… debate to continue, but not until more data is being published. I’ll take a bet on that one!

Alyric said...

Try and clear this junk up first

"One of the tenets of the neurodiversity movement is that autism is not at all a disease or horrific handicap, but just an alternative form of brainwiring."

No, you have it backwards. Try this.

Alternative form of brainwiring - may not be a horrific handicap - certainly not for the great majority of our currently vastly expanded autistic spectrum. But - what do you mean by horrific? I'll buy the disability always but not necessarily the horrific handicap for the majority.

Alyric said...

OK, read it and some points you raise are interesting - the group composition, but maybe not relevant given the nature of this study.

I could claim to be misinterpreted - true but since you quote me correctly others can easily work that out so I won't cavil.

This research is just great for asking questions and too preliminary for jumping to conclusions. So what are you going to do with a severely impaired individual who aces the Raven's matrices? The Raven is considered the best test around to measure fluid intelligence, so why shouldn't autistics be given the best test?

jonathan said...

Alyric-There is really nothing that can be done for the very small minority of autistics who may score severely retarded range on Wechsler but very high on the raven's. It is just a test score, no more, no less, nothing that indicates any potential for life success or something that could have real life applications for helping this person contrary to the phoney baloney spin your pals Dawson and Mottron put on this study with the media.

The vast majority of persons with autism are unemployed, for a good portion of their lives don't go to mainstream schools, dont find romance. The ones in the lower functioning groups are unable to speak, bite themselves, are in danger of running out into the street and being hit by cars. I consider that horrific.

Alyric said...

That severely impaired individual represents a rather big hole in our understanding. Somebody should take a closer look and find out how he's doing it and maybe this is a door to other things? Who knows because as yet nobody's looked.

Foresam said...

What did Michelle Dawson have for an IQ when she ran through the mazes in Mottron's lab?

jonathan said...

I don't know, John, guess you will have to ask Michelle that.

Joseph said...

I'm not sure I follow what the main point of your critique is.

About excluding non-idiopathic autism, that's standard methodology.

Clearly, the control groups were not representative of the general population. They both scored in the 70 percentile range as I recall. They should've scored in the 50 percentile if they had been representative.

The autistics could've easily not been representative either.

But that's not the point of a study like this. You're not trying to do a randomized trial or something of the sort.

The finding of the paper is that there's an IQ gap between Weschler and Raven for autistics, and no gap for non-autistics. It's not actually a finding about absolute intelligence levels. The study is not equipped to make generalizations like this.

Do you feel there's no gap between Weschler and Raven for autistics? To me, this is an obvious finding in retrospect.

Obviously, it's not surprising that the finding is not precise and universal. Some autistics will not have that gap and so forth. This type of result is the norm in any case-control study. So I'm not sure what you're driving at with that.

About the adult autistics scoring in the 50 percentile, again, it's possible they were not representative of autistics in general, or perhaps they were more representative of high functioning adult autistics. So you could say that the adult part of the study had to do with high functioning autism, although that's not clear. It's possible some of the adult autistics scored really high. Another possibility is that Weschler IQ increases in autistics with age.

The finding is replicated for Asperger's.

Yes, it would be good to see if the finding remains for all autism constructs. That's what replications are for.

Dawson et al. tells you what the finding is for a group of autistics who are not necessarily high functioning, even though they might not be precisely representative of autistics in general.

jonathan said...

Joseph I will try my best to address all of the points in your verbose post, but it will be difficult. The point of my critique if you read it carefully is that Dawson and Mottron are putting a spin with the media that is an utter falsehood about their research, the subjects were largely high functioning autistics and the control groups which were almost exclusively male in the children's group and exclusively male in the adults so neither group was representative. Non-idiopathic autism is only excluded in studies such as MR imaging where they are trying to find out what the etiology of autism is depending on other factors. Not in studies such as these that try to generalize all findings to all autistics and try to put the spin that you and the rest of the neurodiversity crowd has put on this study. If Dawson was not trying to do a randomized trial, then why was the control group almost exclusively male, means that somehow the evidence was being skewed towards one type of people. If they are trying to do a study showing differences between autistics and normal controls in the testing they should have people drawn from the general population of both groups. Dawson clearly did not do this. The probability of generating all male control group from just a response from a newspaper ad is about the same as being struck by lightning five times in a year or winning the lottery a couple of times in one year. If you read the paper and the spin that Dawson and Mottron gave to the media, they are clearly trying to claim that this research was trying to show that autistics had a certain type of intelligence and the Raven's was a more valid test than the Wechsler. Comparisons between autistics and nonautistic controls on the Wechsler are apples and oranges since there is such a huge scatter in the subtests of the Wechsler. Michelle did not publish the individual scores on the subtests of any group. No average among the adult group or any of the controls. She skips the digit span test which autistics may score high on in her graph. If the nonautistic controls had huge scatter in their subtest scores the comparison might be apt, but no data are published in this regard.

If you read my post again you will see that there is a huge gap in the two tests only for a very small minority of autistics if you factor in the 10% who have fragile X and other conditions, if you adjust for the sexual inequivalencies in the male to female ratios, so for the most part the answer to your question is no, this study shows no evidence of any significant gap in the scores of the two tests between autistic subjects. What study is it that was replicated for asperger's, can you cite a reference for me, i am unaware of such a study. I will await for any other replications of the study, as far as i know there is no replication of any kind for this study, not for asperger's or any other group of autistics.

The autistics were clearly not high functioning, they, were for the most part in the average range of intelligence as measured on most of the subtests of the Wechsler and the average may well hvae been higher if Dawson had not neglected to put the percentiles of the scores of the digit span test in her data.

Joseph said...

this study shows no evidence of any significant gap in the scores of the two tests between autistic subjects

Again, that's not what the study claims to show. The study presents evidence of a gap in the scores of the two tests for autistics only.

The control groups are there only to show there's no gap for non-autistics. But this is a completely expected finding. If this weren't the case, then there's something wrong with the design of Weschler and Raven tests.

Obviously, the gap is demonstrated for the group of autistics in the study. I mean, there's no question the gap exists for those particular autistics. Do you question this part?

Your criticism seems be that there is some sort of selection bias, and that while the gap exists for those autistics in the study, it does not exist for autistics in general. Can you explain how the bias works such that it makes a gap appear where there's none?

The replication for Asperger's is Hayashi et al. (2007).

jonathan said...

Obviously, the gap is demonstrated for the group of autistics in the study. I mean, there's no question the gap exists for those particular autistics. Do you question this part?

I have explained this post in the post and in my answer to your comment, to make it more concise, yes, there was a gap between some autistics in this particular study, those most likely being a miniscule number percentage-wise of all autistics in the general population when all of Dawson's blatant selection biases are factored in. maybe you should read my post and my response to you more carefully.

Can you explain how the bias works such that it makes a gap appear where there's none?

I already have post in the original post and in my response to you, you need to re-read them more carefully the second time around.

The replication for Asperger's is Hayashi et al. (2007).

Perhaps you can be more specific about what journal it is in, the date and number etc. and maybe i can go to UCLA biomed sometime and look it up if i ever get the chance, though hard for me because of my disability.

jonathan said...

Okay Joseph, one more thing. I never said there was no gap, only that the questionable methodology of Dawson and Mottron and the other authors probably made it look wider than it was when we are talking about all autistics in the general population of autistics.

I have written about this both in my original post and in my first response to you, Please re-read them more carefully.

Joseph said...

there was a gap between some autistics in this particular study, those most likely being a miniscule number percentage-wise of all autistics in the general population when all of Dawson's blatant selection biases are factored in.

The size of the group only matters in terms of statistical error. Otherwise what matters is whether the group is representative.

You claim there were blatant selection biases. Are you referring to inclusion criteria like only considering idiopathic autism? To the male:female ratio? Are these "blatant" biases in your view, seriously?

In Lovaas (1987), the male:female ratio is important, because it proves bias in group assignment. In a case-control study like Dawson et al. it's interesting, but not as important, because you don't have two groups of autistics (that are supposed to be matched with one another).

Perhaps Michelle Dawson discovered a new syndrome called "male idiopathic autism" and her findings only apply to that one syndrome. Maybe future studies can tell us if the findings are applicable to other syndromes.

jonathan said...

Okay, perhaps I phrased that badly, I meant to say that the study suggested no gap between the scores of the two tests in the vast majority of autistics. You knew this is what i meant to say, no matter how much you want to argue that point.

Yeah I am seriously saying the selection biases mattered, not only the exclusion of the genetic conditions, maybe 10% of the autistics, the high male to female ratios and the exclusion of the digit span test in the average. Also the relatively high scores on average in the low 90's in the IQ test of the autistic children and the 100 or above in the adult control group showing these were not the intellectually impaired autistics that the urban legends and propaganda perpetuated by members of the ND movement, yourself included made these people out to be, so the "male idiopathic syndrome" argument does not qualify. We will never know who scored what as Dawson provides no individual data. At one time researches cited that 70% of autistics had IQs of 70 or lower. Though this figure may or may not be obsolete I don't think there is any evidence that half of all autistics, even those exclusively male have nearly average IQs, so your arguments don't wash with me. It certainly does not answer the question of the spin and propaganda the neurodiversity movement have put on this study to promote their misguided agenda.

Joseph said...

At one time researches cited that 70% of autistics had IQs of 70 or lower. Though this figure may or may not be obsolete I don't think there is any evidence that half of all autistics, even those exclusively male have nearly average IQs, so your arguments don't wash with me. It certainly does not answer the question of the spin and propaganda the neurodiversity movement have put on this study to promote their misguided agenda.

That's a critique of the adult part of the study, isn't it? My sense was that the paper presented that part as a sort of extra corroboration, to see if the effect can also be observed in adults. It was not the main part of the study. Yes, these adults *might* have been higher functioning than your average adult autistic (perhaps depending on whether you're counting autistics who were diagnosed as children or all adult autistics by current criteria).

That said, there's also no evidence that most adult autistics are low functioning. It's been pointed out that the 70% figure was essentially made up and not based on testing. In those studies you're also mostly looking at a construct that is different to DSM-IV autism.

Another point is that it's unclear what functioning has to do with the gap between the Weschler and the Raven. You appear to be assuming that higher functioning autistics would have a wider gap. This is non-obvious. In fact, if you "move along the spectrum" to NT, if you will, the gap evidently goes away.

Dawson et al. have not published findings in regards to the gap at different functioning levels, that I know of. This would be useful data. I wouldn't be surprised if the gap is wider for autistics considered lower functioning.

jonathan said...

No, it was a critique of both the adult and children's groups, the former having average or higher IQs on average and the second having nearly average IQs, again you don't seem to be reading my posts carefully or what I say.

JediKnight2 said...

Ohhh gawwd....like who cares about creating tests in order to perform at a high level?

The only purpose of those psychological evaluations is to determine an individual's strengths and weaknesses which lead the psychologist to analyze the results that point to an individual having a brain that works differently.

Even though Dawson and other autistics do not see autism as a disability, but rather a different way of being, what improper/extra impulses (neurons) are misfiring off inside their heads that's giving those autistics these irrational and unrealistic ideas about researching and replicating tests for autistics to complete and attempt to do well on just so they'll feel 'good enough' about themselves?

Arthur Golden said...

Jonathan Mitchell wrote:

"...One of the questions is, are the subjects truly representative of autistics who might be classified as intellectually retarded? [Followed by some analysis by Jonathan Mitchell]...Therefore, it would appear that Dawson’s sample excluded a substantial number of subjects who might be classified as low functioning."

I think you have raised a valid question and I wish to note that in various places Michelle Dawson and her friends have strongly disagreed with you. While I agree with you, I think it is futile to try to disagree with Michelle Dawson. However, I wish to raise another specific point in agreement with your question which I do not find that you specifically raised. In the research paper it states that the autistic children and the autistic adults were all diagnosed with autism disorder at the Pervasive Developmental Disorders Specialized Clinic, which is headed by Dr. Laurent Mottron. As I recently pointed out on the Left Brain/Right Brain blog, the English webpage of Dr. Laurent Mottron describes it as "a specialized clinic for clinical investigation and research on pervasive developmental disorders without mental deficiency." So it seems to me that if all the persons with autism come from this clinic, it would automatically exclude anyone who has a low IQ score.

The last paragraph of this blog entry is:

"Neurodiversity seems fond of spreading stereotypes and urban legends about persons with autism that have no basis in reality. I believe their claim that the results of Dawson’s research showing that autistic people who function at a retarded level have the potential to perform normally is one of them and yet another of the pieces of propaganda that they stoop to in order to promote their misguided agenda."

If this is their claim (I have not personally come across such information) I agree with your concern about using the results of Dawson’s research for this purpose. On the other hand, I personally believe for other reasons that autistics with very low IQ scores, such as my own soon-to-be 37 year-old completely nonverbal son Ben, generally have superior intelligence. I plan to discuss my position elsewhere.

Arthur Golden of Jerusalem Israel

NoetiCat said...

If you are interested in trying a Raven's Matrices test for free, there is one nonverbal IQ test here http://iqtest.dk/main.swf which seems to consist entirely of Raven's Matrices.