Text Size:


When Meredy Goldberg Edelson followed the data trail, she found a dead end. Her figures didn’t add up, but it wasn’t just a discrepancy between her findings and others’. Her research into autism showed something was missing—the research itself.

As a professor of psychology at Willamette, Edelson is well versed in the literature of autism. One of its most commonly accepted tenets is that the majority of children with autism are mentally retarded as well. But when Edelson’s own research with autistic children showed otherwise, she began an exhaustive study of more than 60 years of autism literature. She discovered that most of the data on which researchers have based their claims are highly questionable, and in some cases there is no research whatsoever.

“When I started doing intelligence assessments on children with autism, I realized early on that you can’t use regular measures of intelligence,” Edelson explains. “Typical intelligence tests require children to be verbal and respond within a certain period of time, which many autistic children can’t do.”

Autism is a brain processing disorder characterized by impaired interpersonal relationships and communication; repetitive, stereotyped activities; and unusual responses to sensory stimuli. Children with autism are often described as being in their “own world,” “spacey” and “not connected.” And autism itself is increasingly being described as “epidemic.”

According to statistics from the U.S. Department of Education, more than 4,000 Oregon children ages 6 to 21 with autism received help through the Individuals with Disabilities and Discrimination Act (IDEA) in 2004-05. That’s an increase of more than 15 percent since 2003-04. Across the country, IDEA helped more than 165,000 autistic children in 2004. These numbers, though large, don’t include children who may have autism but have not been properly diagnosed.

Because of autistic children’s limitations, Edelson didn’t test them with standard intelligence measures. Instead, she opted to use the Test of Non-verbal Intelligence (TONI). “This test involves abstract reasoning and does not require a verbal response, so children don’t have to be verbal or have prior knowledge of the world,” she says. Other non-verbal tests require real-world knowledge. For example, in one such test, the child is presented with a picture of a garage next to a picture of a car. Then the child is shown a picture of a refrigerator and a blank. The child chooses which picture goes in the blank—a sugar bowl, a can of peas or a carton of milk. With the TONI, Edelson says, “The child looks at some stimuli and points to a pattern. Using the TONI, children with autism score significantly higher in intelligence than they do on other tests.”

Psychology literature indicates that between 75 and 90 percent of children with autism score below 70 on standardized tests of intelligence, putting them in the mentally retarded range. Using the TONI, Edelson found the children scored approximately 90, which indicates average intelligence (average IQ range is from a low average of 85 to a high average of 115). Only 19 percent of Edelson’s children scored in the mentally retarded range.

Meredy Goldberg Edelson

After repeating her study with a group of autistic children in Taiwan and getting similar results, Edelson knew she was onto something big. “The children in Taiwan were even less verbal, yet we got similar results,” she says. “I began to suspect that maybe what we thought we knew about the intelligence of children with autism wasn’t accurate.”

Edelson decided to dive into the literature—to literally research the research. “I limited my research to English language reports and to studies that either made an overt claim about the number of autistic children who are retarded or had data from assessments of intelligence in children with autism,” she explains.

Edelson reviewed 215 studies dating from 1937 to 2003 that made 223 claims about the rates of mental retardation in autism. Of those 223 claims, only 58 were empirical (claims supported by data), 165 were nonempirical (claims made in the absence of data), and 8 made both empirical and non-empirical claims.

“I wanted to know if the claims came from data,” Edelson says. “If they didn’t come from data, could the claims be traced to data historically? If the studies were empirical, were they conducted in a way that the results were valid? Were they using people who knew how to assess children with autism, and did they use measures appropriate for these children?”

What Edelson found disturbed her. The assumption that the majority of children with autism are mentally retarded is based on questionable data. Worse, nearly three-quarters (74 percent) of the claims are not based on data at all. Of those, 36 percent never even cite the source on which they make the claims. “They claim, "The majority of autistic children are mentally retarded,’ but they don’t state where they get that information,” she says. “Then others cite those studies and end up perpetuating the myth that most children with autism are mentally retarded.”

Even more egregious is that 8 percent of these nonempirical studies make false claims about the studies they cite, and 21 percent inflate the data from cited studies, reporting, for example, that a study found 80 percent of children were retarded when the rate was actually 70 percent. When Edelson reached the end of the citation trail, she discovered 53 percent of the non-empirical studies did not trace back to data. “When I traced these citations back, more than half the time, there were never any data at all.”

Edelson hoped her analysis of the empirical research would uncover better support for the claims regarding the rates of mental retardation in children with autism. But it didn’t. Of the 58 empirical studies, only 35 (15.7 percent of all the studies reviewed) described how they measured intelligence in autistic children. “There isn’t even a way to evaluate most of the data,” Edelson says. “Most just report their results without describing how they measured intelligence. Also, the way intelligence is measured varies widely. In the same study, researchers might use verbal measures or timed performance measures. Some use development measures or social adaptability scales, which aren’t even measures of intelligence.”

When Edelson untangled the studies, she found “many of the studies did what I call "equating untestability’ with mental retardation. If the researchers had a child they couldn’t test, they just assumed he or she was retarded and assigned a very low IQ score, like 25. Others do a little bit of everything but aggregate their data so you can’t see what is valid and what isn’t.”

Stopping the Hurt

Meredy Goldberg Edelson held regular sessions with her children when they were small, warning them about inappropriate sexual overtures. Edelson is all too aware of how vulnerable children are; she sees child abuse up close and personal each Thursday morning at Liberty House in Salem, where she serves as a child interviewer.

Before Liberty House opened in 1999, local children who had been abused or neglected were shuffled through doctors’ offices, police precincts and courtrooms, where they were intimidated and retraumatized by multiple interviews. At Liberty House, children are interviewed once, and interviews are videotaped for further review. Medical examinations are provided, and families are put in touch with counseling services. The process helps children heal from unspeakable crimes and often helps take abusers off the streets.


Interviewing one child each week, Edelson has now worked with 280 children, almost half of them younger than 6. “Meredy has gifts with children and a deep understanding of those who have been abused,” says Gretchen Bennett, executive director of Liberty House.

Some children are reluctant to talk. They look at the floor and keep their silence, and Edelson respects their decision. Others are relieved that someone will listen. When the interview is over, Edelson helps the child pick out a stuffed animal.

Every once in awhile, Edelson is caught off guard. “Sometimes a child comes in who reminds me of one of my kids.” Perhaps they’re the same age or have similar mannerisms. But she knows that, in a deeper sense, they are all her children. At least on Thursday mornings.

Her final determination: Researchers have made a sweeping conclusion, one that impacts thousands of children and families, based on faulty data or no data at all. “The autism field has accepted as fact that autistic children are retarded. There are so many claims and they’re so widespread that no one has bothered to look at the data behind them. There is not a lot of data to support the claims, and the data available are 35 to 40 years old and are based on measures that don’t even measure intelligence.”

Because retardation in autistic children has been so widely accepted, Edelson says schools and parents have lowered expectations of these children. “If we believe that vast majority—75 to 90 percent—of autistic children are retarded, we’re not going to challenge them. We’re not going to give them opportunities.”

Edelson feels compelled to dispel the autism-retardation myth, and she’s had to take on most of the psychology community to do it. She applied to present her research at the Autism Society of America National Conference, but was rejected despite having presented there twice before. She sent her paper to Science, one of the most respected research journals, but editors there said it was “inappropriate.” When she submitted it to Developmental Psychology, one of their reviewers praised Edelson’s careful research, but two others gave scathing reviews that amounted to personal attacks.

Wary from these responses, Edelson sought advice from Richard Simpson, professor of special education at the University of Kansas and former editor of the journal Autism and Other Developmental Disabilities. “I recall reviewing one or more of Dr. Edelson’s manuscripts and encouraging her to submit them for publication consideration to Focus,” Simpson says. “I have great respect for her. The quality of her work is excellent and extremely creative, and she is willing to challenge the status quo.”

When Edelson sent her paper to Focus, she requested the editors act as reviewers. “I told them this research was very political and I wanted my paper judged on the quality of the work, not its political nature.”

The strategy worked, and Edelson’s research is slated for publication this summer. “We were very interested in publishing Meredy’s manuscript because her research challenges a commonly accepted but rarely validated assumption regarding individuals with autism,”says Juane Heflin, associate professor of educational psychology and special education at Georgia State University and co-editor of Focus. “Meredy was thorough in her approach to the empirical question and provides strong substantiation to challenge the commonly held belief.”

Concerned about raising such a controversial challenge, Focus editors talked with Edelson about her willingness to allow for rejoinders. When she readily agreed, the journal solicited two highly respected individuals in the field and asked them to read and respond to the pre-publication version of her manuscript. “We anticipated strong reservations and were very surprised that both individuals commended Meredy for raising the question and for her careful approach to analyzing the data,” Heflin explains.

The rejoinders will be published in the same issue as Edelson’s article, and they not only commend her work, but speak to its potential impact. One of the rejoinders states, “This is an excellent article that should encourage all professionals to confront many of the preconceived notions they may have regarding children affected by Autism Spectrum Disorders,” while the other concludes, “We wish to compliment Dr. Edelson for asking an important question and providing a scientifically convincing answer. Hopefully, her request for more extensive and objective research in this area will come to pass.”

Edelson is unfazed by the political wrangling. She just wants the truth to come out so autistic children can be helped. “In the 1950s, children with autism were institutionalized. Today we know that they have more options, from education and treatment to life plans including college and careers, marriage and children. If most children with autism aren’t retarded, we need to find ways for them to interact with society and help them become all they can.”