Autism, Asperger’s, and IQ

Aren’t people with Asperger’s more likely to be geniuses? Isn’t genius related to autism?”

A university student asked this in a course I am teaching. The class discussion was covering neurological differences, free will, and the nature versus nurture debate. The textbook for the course includes sidebars on the brain and behavior throughout chapters on ethics and morality. This student was asking a question reflecting media portrayals of autism spectrum disorders, social skills difficulties, and genius.

I did not address this question from a personal perspective in class, but I have when speaking to groups of parents, educators, and caregivers. Some of the reasons these questions arise, as mentioned above, are media portrayals and news coverage of autism. Examples include:

Is autism related to genius? Are people diagnosed with ASDs more likely than other people to have “special gifts” or is that a myth? I’ll first examine the savant-autism link and then “genius” and ASDs.

There does seem to be consensus that “special skills” are ten times more common among autistics than the general population. That statistic sounds impressive, until you realize that it isn’t that large a number. The Autism Research Institute’s website suggests:

 The estimated prevalence of savant abilities in autism is 10%, whereas the prevalence in the non-autistic population, including those with mental retardation, is less than 1%.

I’d want to see more data, but let us assume the 10 percent number is accurate. That still means that 90 percent of autistics do not have special skills or savantism. While a autistic savants might be more statistically common than non-autistic savants, they still are relatively uncommon. But, the media are more likely to focus on these unusual stories than they are gifted people without other identifiable challenges.

Also, it is important to recognize that savants and people with “splinter skills” are not geniuses. They are gifted in narrow ways, often ways that do not reflect higher-level thinking skills. That’s not to dismiss the gifts of these individuals, but being able to recite every recorded statistic for a sport is not genius.

We often conflate memory for intellect, something I discuss later in this essay. Heightened memory skills can be a neurological abnormality or they can be the result of an obsessive interest in a single topic.

Dr. Treffert, of the University of Wisconsin, offers the following:

 Savant skills exist over a spectrum of abilities. The most common savant abilities are called splinter skills. These include behaviors such as obsessive preoccupation with, and memorization of, music and sports trivia, license plate numbers, maps, historical facts, or obscure items such as vacuum cleaner motor sounds, for example.

Talented savants are those persons in whom musical, artistic, mathematical or other special skills are more prominent and highly honed, usually within an area of single expertise, and are very conspicuous when viewed against their overall handicap.

The term prodigious savant is reserved for those very rare persons in this already uncommon condition where the special skill or ability is so outstanding that it would be spectacular even if it were to occur in a non-handicapped person. There are probably fewer than 50 prodigious savants living worldwide at the present time who would meet this high threshold of special skill.

But, what about general “above average” intelligence and autism? Is that more or less common than in the general population? This depends on how we define “autism” in the Diagnostic and Statistical Manual of Mental Disorders. When one of the criteria for autism was intellectual impairment, primarily before 1994, then to qualify for a diagnosis of autism the individual was impaired. Therefore, no one diagnosed with autism had a general IQ score over 70 to 80, depending on the IQ instrument used.

Today, we diagnose autism without emphasizing IQ score. That means communication challenges and social traits are the primary diagnostic criteria — and it is possible that intelligence itself causes social challenges for some individuals. Genius is abnormal and we know that abnormal is often a difficult thing to be.

As a result of changes in diagnostic criteria, it appears intelligence, as measured by various instruments, is distributed within the autistic population about the same as it is in the general population. Splinter skills are more common among autistics, but statically intellectual impairment and genius are only mildly correlated to autism.

 Association between extreme autistic traits and intellectual disability: insights from a general population twin study
Hoekstra, Happé, Baron-Cohen, et al. 2010.

Intellectual disability (here defined as IQ<70) is common in autism. Historically, the prevalence of intellectual disability in autism is estimated at 70% (1) but recent studies encompassing all autism-spectrum conditions, including Asperger syndrome and pervasive developmental disorder – not otherwise specified, suggest that the prevalence of intellectual disability in autism-spectrum conditions may be considerably lower (2,3).

Extreme autistic traits were modestly related to intellectual disability; this association was driven by communication problems characteristic of autism. Although this association was largely explained by genetic factors, the genetic correlation between autistic traits and intellectual disability was only modest.


  1. Fombonne E. Past and future perspectives on autism epidemiology. In Understanding Autism, from Basic Neuroscience to Treatment (eds SO Moldin, JLR Rubenstein): 25–48. Taylor and Francis, 2006.
  2. Shea V, Mesibov G. Adolescents and adults with autism. In Handbook of Autism and Pervasive Developmental Disorders (5th edn) (eds FR Volkmar, R Paul, A Klin and D Cohen): 288–311. John Wiley and Sons, 2005.
  3. Chakrabarti S, Fombonne E. Pervasive developmental disorders in preschool children: confirmation of high prevalence. Am J Psychiatry 2005; 162: 1133–41.

I have problems basing anything on IQ, because traditional IQ tests have so many problems that they can only be described as measuring existing knowledge and narrow abilities. I don’t dismiss these special skills and abilities, but I fear we often mistake a high IQ with creative potential or the ability to solve new problems. The truth is, plenty of people with “average” IQ scores do great things while the 180 IQ “Super Genius” might do nothing with his or her skills. The scores matter, but as long as you are average or above, what predicts personal success is your creative problem solving skills.

Before discussing the flaws in detail, it is useful to review the approximate IQ scale associated with the “Binet” (named for Alfred Binet) series of IQ instruments. There are other scales, which you can locate online, but the above offers plenty of detail for this discussion.’s IQ page lists the following:

  • 1 to 24 – Profound mental disability
  • 25 to 39 – Severe mental disability
  • 40 to 54 – Moderate mental disability
  • 55 to 69 – Mild mental disability
  • 70 to 84 – Borderline mental disability
  • 85 to 114 – Average intelligence
  • 115 to 129 – Above average; bright
  • 130 to 144 – Moderately gifted; (>140 = “gifted” in most school systems)
  • 145 to 159 – Highly gifted
  • 160 to 179 – Exceptionally gifted (>160 = “genius”)
  • 180 and up – Profoundly gifted

The characteristics of normal distribution applies to IQ scores as well:

  • 50% score are in the range from 90 to 110
  • 70% score in the range from 85 to 115
  • 95% score in the range 70 to 130
  • 99.5% score in the range from 60 to 140
  • Only 0.5% score as “Highly Gifted” or what some call “Genius”


My challenge to these tests is that they are too limited to categorize individuals. It bothers me that a one test, given over a day or two, might be used to label a school student. That was my personal experience. And, the results varied so wildly as to be meaningless.

  • As an adult, I’ve had the following scores during neurological evaluation: 109, 126, 140, 165, and 172.
  • Birth: Observationally evaluated, doctors assumed I was intellectually impaired.
  • First Grade: Evaluated as “below average to average” by the schools in Bakersfield, California.
  • Third Grade: Evaluated as qualifying for the Gifted and Talented Education (GATE) program in Visalia, California.
  • Sixth Grade, I tested in the 99.8th percentile, placing me in the “highly gifted” to “exceptionally gifted” range.
  • Jr. High and High School: Test scores and evaluations ranged from the 90th to 99th percentiles.
  • College: Evaluated at the 99.5 percentile and higher on several tests, including a Binet IQ test.

How can an adult test with an IQ of 109 and 172? What produces five scores that represent several categories of intelligence? The scores vary too widely, in my view, to be valid individually. In theory, a person scoring 109 on an IQ test is unlike the person with a 172 IQ. The instruments for which I have any records don’t even “cluster” around a range of scores. If the scores were all within a 20-point range, that might be reassuring, but the scores cover a 63-point range.

What do IQ tests measure? Most of the tests I have taken focus on memory, pattern recognition, and vocabulary. If you have an exceptional memory and fast pattern recognition, you do well on the tests. If you can memorize the “SAT/GRE” word lists, you do well on the tests. (A bit of a trick: study Greek and Latin roots, then memorize the words they place on tests to trick you.) The idea that you cannot study for an IQ test doesn’t seem accurate to me. If you study for any standardized test, it likely will help you with an IQ test.

Some IQ instruments are suites of tests and activities. The Weschler evaluations are usually suites. My “worst” evaluation was a WAIS test battery, on which I scored the 109. The problem with the suite is that there are “speed” and “visual-motor” portions. When I took the exam, I was having palsy episodes related to my paralysis, my eyes were in need of surgery, and I was anemic. Sorry, but a shaking person with poor vision isn’t going to score well on a test requiring manipulation of blocks, pegs, or cardboard shapes. Just toss those scores out and see what remains. The other portions of the WAIS were in the 94th percentile, even with medical issues.

My “best” score was on a Stanford-Binet Third Edition test. That 172 placed me in the “Genius” range on that test. From what I can recall, the Binet didn’t claim to measure IQ above 160. What made the great score possible on the Binet test in the mid 1980s? My guess, and it is only a guess, is that I was in reasonably good health, well-rested, and it was a favorable testing environment. Again, I am assuming the physical and emotional context affected my focus and performance on the evaluation.

How does someone with limited motor control, vision issues, or other limitations do well on full-suite IQ tests? I don’t know, but I do know that these instruments are not ideal for individuals with autism. You need a “great day” to do well on these tests, even without any physical or neurological challenges.

IQ tests are only measures of what you already know, in my opinion. If you’ve never heard or read the words used, there’s no way a vocabulary-based test can be meaningful. Tests of analogies are glorified vocabulary exams, too.

I’m not certain there is an ideal IQ test for individuals with autistic traits. For now, we can remember that every autistic person is unique. Myths of great intelligence or impairment are just that: myths. Some of us are gifted, some of us are intellectually challenged.