HomeTable of ContentsFeedbackSubscribeHelp/AboutArchivesSearch
Science News Online

Science News for Kids
Science News
for Kids

Audible Science News Subscribe to an
audio format

Published by
Science Service

Print this article
Email this article to a friend

Week of July 7, 2007; Vol. 172, No. 1 , p. 4

Hidden Smarts: Abstract thought trumps IQ scores in autism

Bruce Bower

There's more to the intelligence of autistic people than meets the IQ. Unlike most individuals, children and adults diagnosed as autistic often score much higher on a challenging, nonverbal test of abstract reasoning than they do on a standard IQ test, say psychologist Laurent Mottron of Hôpital Rivière-des-Prairies in Montreal and his colleagues.

The same autistic individuals who score near or below the IQ cutoff for "low functioning" or "mental retardation" achieve average or even superior scores on a test that taps a person's ability to infer rules and to think abstractly about geometric patterns, Mottron's team reports in the August Psychological Science.

"Intelligence has been underestimated in autistics," Mottron says. Autistic people solve problems and deploy neural resources in unusual ways, which are poorly understood and might contribute to problems with IQ tests, he asserts.

Mottron regards autism as a variant of healthy neural development. For that reason, his group—including study coauthor Michelle Dawson, herself diagnosed as autistic—prefers the term "autistic" to "person with autism."

The researchers studied 38 autistic children, ages 7 to 16; 13 autistic adults, ages 16 to 43; 24 nonautistic children, ages 6 to 16; and 19 nonautistic adults, ages 19 to 32.

Volunteers completed an age-appropriate IQ test and a Raven's Progressive Matrices test. The latter test includes 60 items, each consisting of a series of related geometric designs and a choice of six or eight alternative designs, one of which completes the series.

The nonautistic children and adults scored slightly above the population average on both tests.

In contrast, autistic kids and adults scored far higher on the Raven's test than they did on the IQ tests. These youngsters' average IQ was substantially below the population average, but their average score on the Raven's test was in the normal range.

One-third of autistic children qualified as "low functioning" by IQ, but only 5 percent did so by Raven's scores. Moreover, another third of the autistic children achieved "high intelligence" on the Raven's test.

As in previous research, autistic volunteers performed well on an IQ task that required them to reproduce geometric designs using colored blocks.

The new findings confirm prior indications that autistics score poorly on IQ tests despite processing perceptual information well, comments psychologist Uta Frith of University College London. In a 2000 study, Frith's team noted that autistic and nonautistic children made equally rapid and accurate visual judgments, such as discerning which of two lines was longer.

In people with autism, a lack of social insight derails the ability to acquire skills and information from others, a key to IQ success, Frith theorizes. Autistics thus succeed only on self-explanatory tasks, such as the Raven's test.

The Raven's test may measure autistic intelligence better than an IQ test does, adds psychologist Helen Tager-Flusberg of Boston University. Nonetheless, many autistic children are extremely impaired intellectually, she says.

Researchers generally sell short the unique features of autistic intelligence, Dawson responds. For example, autistics shift flexibly back and forth between focusing on details of a scene or its overall configuration, whereas nonautistics single-mindedly concentrate on the big picture, she says.


The article didn't mention that traditional IQ tests are in one sense "language" tests. The Ravens test doesn't involve language processing in a typical manner. A person with a language disorder, as an autistic person is assumed to be, would do better on a nonverbal test. That the intelligence of autistic people can be underestimated is really due to the fact that our culture assumes that language skill equals intelligence.

Tammy Bush
San Francisco, Calif.

If you have a comment on this article that you would like considered for publication in Science News, send it to editors@sciencenews.org. Please include your name and location.

To subscribe to Science News (print), go to
https://www.kable.com/pub/scnw/ subServices.asp.

To sign up for the free weekly e-LETTER from Science News, go to


Dawson, M. . . . and L. Mottron. 2007. The level and nature of autistic intelligence. Psychological Science 18(August):657-662. Abstract available at http://dx.doi.org/10.1111/j.1467-9280.2007.01954.x.

Further Readings:

Scheuffgen, K. . . . and U. Frith. 2000. High "intelligence," low "IQ"? Speed of processing and measured IQ in children with autism. Development and Psychopathology 12(March):83-90. Abstract available at http://dx.doi.org/10.1017/S095457940000105X.


Michelle Dawson
Pervasive Developmental Disorders Specialized Clinic
Hôpital Rivière-des-Prairies
7070 Boulevard Perras
Montréal, Quebec H1E 1A4

Uta Frith
University College London
Institute of Cognitive Neuroscience
Alexandra House
17 Queen Square
London WC1N 3AR
United Kingdom

Laurent Mottron
Pervasive Developmental Disorders Specialized Clinic
Hôpital Rivière-des-Prairies
7070 Boulevard Perras
Montréal, Quebec H1E 1A4

Helen Tager-Flusberg
Human Development Program
Boston University
715 Albany Street L-814
Boston, MA 02118

From Science NewsVol. 172, No. 1, July 7, 2007, p. 4.

Science News

Find a Job
Job category: