My 34 year old client who has a diagnosis of Intellectual disability ("ID") and severe autism.
He has been detained for the last 13 years in ID care, ostensibly because he is "dangerous" his index offence was smashing three windows value $905 with an axe in response to a loud noise which is a trigger for bad behaviour.
He allegedly as a youth (17 years old) assaulted a girl he liked at school with a knife, and she need hospital treatment and some stitches, he had no idea of what he was doing. The Youth Court gave him a discharge without conviction.
The available literature seems to show the WAIS IV is unreliable in this situation his score was 60, whereas on the Raven Matrices it was 80. He of course has severe communication deficits, but a number of savant abilities such as rapid birth date calculation (tell him your date of birth and he will immediately tell you which day of the week in falls on this year), and artistic abilities, as well as excellent memories for some events 10 years ago.
I am challenging the finding that he is intellectual disabled as well as the proportionality of the treatment, and other matters in High Court proceedings, set down for hearing in July for 7 days.
Have two experts who review the literature and support the view the test results do not reliably show ID.
Has anyone similar experience, any suggestions, or know of any reported case law on this issue?
What is your evidence that the WAIS is unreliable? It has a comprehensive range of subtests, and the subtest scores might be expected to vary widely, but this does not mean they are unreliable. A savant must surely have some subtests in the normal range. In the old days the congenitally deaf were assessed with the Stanford-Binet which of course confounds language disorder with intellectual disability. Wechsler tests avoid this trap by distinguishing verbal IQ, very low in the deaf, and non-verbal IQ, which is normal in the deaf.
NB The question title is garbled.
"his index offence was smashing three windows value $905 with an axe in response to a loud noise which is a trigger for bad behaviour"
All the persons I have tested who had audiosensitivity, autistic and non-autistic, had abnormal middle ear muscle reflexes consistent with an oversensitive cochlea. Has he had an audiological workup?
In answer to What is your evidence that the WAIS is unreliable?
My ID expert says:
The WAIS-IV Manual p113 notes that people with ASD typically have deficits in verbal and non-verbal communication and social deficits. Those with ASD who had pre-assessed IQs of less than 60 were excluded from their normative study, indicating that the test should not be used with these people. They included only 16 people with ASD in their study.
Souliers.I, Dawson. M. Gernshacher. MA and Mottron. , (2011)'The Level and Nature of Autistic Intelligence II What about Asperger Syndrome?" PLoSone. Vol. 6. Issue 9 says
Taken together, the recent literature therefore suggests that the most commonly used test of intelligence - the Wechsler-IV - should not be considered a useful tool for the assessment of the intelligence
level of people with ASD because scores are distorted by the fluctuating effects of ASD.
The Stanford-Binet comes closer to isolating
fluid reasoning and provide for a report on that aspect of intelligence in addition to crystallised reasoning reflected in the Verbal and Non-Verbal lQs
But the most useful tool to assess fluid reasoning or innate intern^— the Raven's Progressive Matrices and this tool will produce a much higher estimate of general IQ than do the other tests.
My psychometric experts says in his conclusion.
"The research-based statements presented in paragraphs 1.1, 1.4, 2.14, and 2.17 clearly indicate no single IQ score can be taken as indicative of an individual possessing a clinical diagnosis or symptoms which also manifest themselves in particular kinds of intellectual deficit. Paragraph 2.17 is especially relevant here as it shows that the calculation of IQ scores in the lower tail of an ability distribution is not trustworthy. Therefore, attempting to put confidence intervals around any such score, where the interval estimates are based upon “error” derived from ‘normal’ range IQ scores is a meaningless operation, as the score itself in these lower ranges is no longer to be considered a veridical estimate of IQ, let alone any estimate of error which is applicable to such cases. As Silverman et al (2010) point out, no clinically-accepted estimates of ‘measurement’ error currently exist."
I have had some experience with testing and published a bit in this area, but have read widely, which is relevant because so much of the vast published literature is deemed out of bounds because of political correctness and hence no longer read. I do have a number of disagreements with your experts.
First, I do agree in general that no single IQ score is useful for many clinical syndromes.
The Wechsler tests are designed to establish norms, so should be standardized on a random sample of the normal population, not on clinical disorders. Someone with autism needs to be compared to neurotypicals, not others with ASD. Likewise, audiometric standards need to be fixed on the healthy population, excluding those with ear disease, so that a pure tone audiogram can immediately determine if hearing is abnormal.
The Stanford Binet should not be used in anyone with a language disorder, especially the congenitally deaf, as in hopelessly confounds language delay with low general intelligence.
In agreement with Braden, I think it is easy to interpret WISC tests in the deaf. I do not know how to interpret Raven IQ results in the deaf, nor, I suspect, does anyone else. The same applies to language retardation in autism.
WISC test results in autism show wide dispersal but I am not sure that means they are unreliable. The old follow-up studies of autism showed that IQ was one of the best predictors of outcome.
Is it generally accepted that autism is a fluctuating condition?
In the psychometric literature, g (the general intelligence factor) is identical regardless of the test batteries that are used.
NB I have not yet read the Souliers article.
After a 7 day High Court hearing the judge has reserved her decision.
Professor Barrett a psychometrics expert gave evidence and reviewed all the writings of Simon Whittaker including his 2010 article Error in the estimation of intellectual ability in the low range using the WISC-IV and WAIS-III Personality and Individual Differences 48 (2010) 517–521 which postulated an error range of 16 points below and 25 above.
He concluded the author was incorrect the true range of accuracy was +/- 12.47.
- Badges
- Science topic
- Similar topics
- Filing