Children with autism are also frequently diagnosed with mental retardation based on results from common IQ tests. But are these results always accurate? Research from the University of Montreal indicates that these tests may not actually be accurate for individuals with autism. In a recent study, children with autism and typically-developing children were given two IQ tests: the WISC (a test that uses a great deal of language) and the Raven's Progressive Matrices (which tests "fluid intelligence" - skills such as creation and management of goals, high-level abstractions, and inference of rules). While typically-developing children scored about the same on both tests, the results were strikingly different for the children with autism. None of the children with autism in the study scored in the "high intelligence" range on the WISC, but thirty-three percent did so on the Raven's Progressive Matrices. Thirty-three percent of the children with autism scored in the range of mental retardation on the WISC, but only five percent scored in this range on the Raven's Progressive Matrices. The current study, as well as future research in this area may lead to more accurate testing of individuals with autism.
The following story illustrates the difficulties of accurately testing individuals with autism, as well as their unique intelligence. In my first year of teaching, testing of my students with autism that accurately captured their abilities had been somewhat of a challenge. Since students with autism may not respond well to typical testing measures, as the above research highlights, some modifications may be necessary. I once had a student who was completing a non-verbal intelligence test that required him to look at a pattern on a paper, then point to the figure at the bottom of the page that completed the pattern. Below is a simplified example.
When the student made no response to these directions, even after modeling, I decided that a modification was in order. I cut off the bottom of each page, cut out every answer choice, and gave them to the student. The student would then select the correct answer to glue in the blank space in the pattern. I did this ‘on the fly’ as I was testing the student, so it took a significant amount of time.
Despite the time it took to cut out all of the answer options, the student was successful with this modification. I now felt that the test was measuring his skills more accurately.
Later in the test, I was rushed on one question, and I accidentally handed the student the block of all four answer choices. The student looked at the answer choices, cut out only the correct answer, and glued it in the space. This was so much faster and more efficient than my method! It’s always fun when you’re outsmarted by one of your students! I thought this was a great example of how this young man showed his intelligence in a way that was not measured by the test.
Look for more blog posts about testing for individuals with Autism in the future, and please share your thoughts and experiences by posting a comment.
References
Association for Psychological Science (2007, August 5). The Matrix Of Autism. ScienceDaily. Retrieved May 27, 2008, from http://www.sciencedaily.com /releases/2007/08/070803151245.htm
Collecting data on paternal age and grandparental age
Dear Nicole,
I wonder if during your testing you take a family history. Two variables such as the father's age and the mother's father's age at her birth seem related to risk of gene mutations in sperm with resulting autism. Would this be interesting to check? Is more severe non-familial autism related to more advanced paternal age? Maybe someone could research this.
Leslie