Public Release: 

Autistic intelligence measured inaccurately, Université de Montréal study

University of Montreal

A research group led by Université de Montréal researcher Dr. Laurent Mottron presented its results on autistic intelligence. The research group, which includes an autistic researcher, has concluded that usual measures of autistic intelligence are in fact measuring incorrectly. The results are presented at the American Association for the Advancement of Science (AAAS) conference in St-Louis, which attracts hundreds of scientists.

"It is reported that only 25% of those diagnosed autistic have normal measured intelligence, with mute autistics assumed to be the most mentally retarded," explains Dr. Mottron. "The same prejudicial mentality which gave us "idiot savants", referring to autistics with special skills, refuses to recognize such autistic peaks of ability for "true" intelligence. And autism itself is reported as bereft of cognitive advantage, as entirely negative, as a disease to be eradicated. We see these judgements as a chain of clichés which misrepresent autistic intelligence and deserve to be challenged."

In research and clinical settings world-wide, autistic intelligence is measured by averaging scaled scores of Wechsler scales subtests (WAIS, WISC). In a scaled score, a score of 50th percentile (50%ile) means that 50% of the non-autistic population performs under this value, and 50% above. Wechsler scales are composed of a verbal subscale providing a verbal IQ via oral questions requiring oral responses, and a non-verbal subscale providing a non-verbal IQ via oral questions requiring non-verbal answers. That is, performance of both subscales requires mastery of oral language.

When autistics with oral language perform Wechsler scales, they consistently score very low (< 25%ile) on a verbal subtest, "comprehension"; and consistently very high (above 75%ile) on a non-verbal subtest, "block design." This conundrum may be resolved by the Raven's Progressive Matrices (RPM). This problem-solving test of high-level abstract reasoning and rule-inferring neither requires oral instruction nor relies on manipulating long-term memorized information. RPM is the single intelligence test which is the most correlated with other intelligence tests in non-autistics.

"We found that RPM consistently estimates the intelligence of oral autistics to be 30%ile above Wechsler full-scale IQ. Even more impressive is that some "mute" or "non-verbal" autistics, who perform very poorly in Wechsler, perform at an outstanding level (95%ile) in RPM, a test which requires verbal skills in typical people in order to be performed successfully," explains Dr. Mottron. "This is evidence that autistics do not need to solve RPM verbally, and that this test, which consists of information autistic brains are able to process well, represents a window for the accurate assessment of autistic intelligence."

"In addition to how autistic intelligence is measured, there is the question of when it is measured," said Dr. Mottron. "Often-cited epidemiological data on intelligence in autism are collected circa 4-5 years of age, well short of when an autistic child reaches the maximum of testability and intellectual potential. This risks underestimating considerably average intelligence levels, especially when coupled with the use of oral question tests. We found that mute autism at 4 years of age does not represent the final achievement of autistic intelligence. Testability and intellectual potential are more fully achieved circa age 6 or even later."

In summary, many prevailing clichés about autistic intelligence are distant from reality. Autistic intelligence has been measured inaccurately, misrepresented as not "true" intelligence, and characterized as rare. A more accurate view must include the superior performance of autistics on RPM; the recognition that the performance of a mute 4 year old autistic child is not definitive; the recognition that "perceptual intelligence" is not a meaningless trick, but is "true" intelligence in autistics; and the certainty that those reporting a precise figure for prevalence of high or low intelligence in autism should be disregarded.


About Université de Montréal
Founded in 1878, the Université de Montréal counts 13 faculties and, along with its two affiliated schools, HEC Montréal and l'École Polytechnique, is Quebec's largest institution of higher learning, second in Canada, and among the most active in North America. With a faculty of 2,400 professors and researchers, the university has a student population of more than 55,000, offers more than 550 undergraduate and graduate programs and awards some 3,000 Master's and PhD degrees each year.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.