Why IQ is irrelevant to the diagnosis of learning disabilities by Dr. Linda Siegel

Larry, at age eight, received a score of 78 on an IQ test. This score is in approximately the bottom 10 per cent of the population. He was placed in a class for “mentally retarded” children where he remained until age 14. I tested him at 34, when he was enrolled in a graduate program in a major Canadian university after completing a BA in psychology with an A average. Throughout his school career, Larry had great difficulty in learning to read, spell, write, and do arithmetic calculations. When tested at age 34, his IQ score was 119, well above average and better than about 85 per cent of the population. However, he still had significant problems with reading and spelling. He also had difficulties on short-term memory tasks and had occasional difficulty with verb tenses and finding the right word when speaking. Larry at age 34 displayed a profile of a dyslexic individual; yet at age eight he was labelled as “mentally retarded.”

The case of Larry is a very dramatic example of the consequences of using an IQ test score as part of the definition of a reading disability. At age eight, Larry had dyslexia but, instead, was labelled as “mentally retarded.” Larry was fortunate enough to have a very determined personality and very supportive parents who fought for his rights to be educated.

This case is a real one. Fortunately, it has a happy ending, but for many children with genuine learning problems the ending is not university or graduate school but jail, alcohol and drug abuse, and/or suicide. Larry’s supportive environment did not prevent or cure his reading disability; his reading problem remained throughout his schooling and into adulthood. However, his environment probably prevented Larry from developing the serious social problems that are often a consequence of an undetected and untreated learning disability. Is Larry a rare exception? No. Today a child with poor reading skills and an IQ of 78 would be labelled “developmentally delayed”, “mentally retarded” or a “slow learner”, or said to have a “general learning disability”. They would not be recognised as “reading disabled” or “dyslexic”, or as having a specific learning disability, and would not receive intensive help with reading because it would be argued, incorrectly, that we should not expect better reading from an individual with this IQ level. Unfortunately, children with lower IQ scores who show signs of severe reading problems are barred from getting help for their problem, even today.

The myth of intelligence

In the early 1960s when the idea of learning disabilities first developed, it was important to show that people who were struggling with reading or mathematics were intelligent enough to learn to read or do mathematics. So the solution was the IQ test to prove that they were capable of learning. They had to score above a certain level, although what that level was varied from place to place.

Although the idea of demonstrating that people who had learning disabilities were intelligent enough to learn was initially an important one, this idea developed into an obsession. It is hard to say exactly when we started on this slippery slope of IQ worship, but perhaps it was with the original idea that was proposed when the term learning disabilities was invented. To differentiate people with learning problems from those with serious mental handicaps, the profession specified that someone with a learning disability had to be of at least average intelligence. This requirement was quite reasonable at the time. However, other problems soon became evident.

What is intelligence? At first, the answer seems to be simple. Intelligence is what the IQ test measures. But the IQ test measures a limited number of skills, including memory for numbers, mental arithmetic, fine motor coordination, and the ability to define words. Humans have a wide range of skills and some psychologists, such as Howard Gardner, have always insisted that we should think about intelligence not as a single construct but as composed of many different skills. Some people are good at mechanical skills and can fix a broken car, others can draw well, and some can define obscure words. Which is the real intelligence?

People assume that the IQ test measures intelligence. The point is that we do not have an absolute standard of intelligence. Our concept of intelligence as measured by the IQ test is that we have an “intelligence” of a certain level compared to other people. IQ is always in comparison to others; there is no absolute standard for intelligence. There is no part of the brain in which resides the mythical entity called intelligence and no blood test or X-ray that can tell a person’s intelligence. Intelligence is an artificial construct, a product of the imagination of the test constructors.

The IQ score is supposedly a measure of “potential”. The concept of potential means that this score predicts what you are capable of. Often educational institutions specify that an IQ test score that is more than three years old is not valid and the person must be retested. This is the ultimate absurdity. If the IQ score is an indication of potential, then why would it change?

We assume that we know what intelligence is and we are then locked into a vicious cycle. Intelligence is what the IQ measures and the IQ is valid because it measures intelligence. Somehow we lost sight of the top of the hill and we failed to define and conceptualise intelligence. To understand how intelligence is defined, we need to examine the intelligence test. For example, intelligence tests typically ask questions about the definition of words, such as pusillanimous or terpsichorean. These words are not the real ones used on the test because we are not allowed to reveal the questions on the IQ test. The argument is that no one but the testers can know the questions on the IQ test. But if IQ is supposed to measure some sort of innate potential, then why would the potential change? Why does it matter if you know the questions beforehand? In the case of testing your vocabulary, if you knew the words before taking the test and if you looked up their meanings and answered correctly on the test, then you would show that you are capable of learning. If you give the correct answer to these words (and others), you are considered more intelligent than someone who does not know as many words. But it could also mean that you do not know the particular words on the test, and if some other words had been used you would have magically been more intelligent.

In the 1980s we got stuck in the quicksand at the bottom of the hill. The IQ became a godlike or mythical deity to whom the field worshipped. In the ultimate expression of this worship, IQ became incorporated in a mathematical equation. Learning disabilities were reduced to a simple equation. The IQ score became the basis of what is called the discrepancy definition of learning disabilities. A learning disability occurs when the IQ score is much higher than the reading (or arithmetic) score. This is called the discrepancy definition of a learning disability. One of the criteria for the existence of a learning disability is the presence of a discrepancy between IQ test score and achievement test score.

If there was no discrepancy, the student was considered to be a low achiever rather than a student with a learning disability. It was assumed that low achievers would never amount to much, and so they did not receive help for their problems. Here the LD profession took the wrong turn on the garden path and landed in the quicksand. And we are still paying a very high price for this wrong turn.

There are studies that show that there is no reason to require that there be a discrepancy between an IQ score and a reading score for a student to be considered learning disabled. I compared two groups of children who had low reading scores, but one group had reading scores that were significantly lower than those that were predicted by their IQ scores, traditionally called dyslexic, and the other group had low reading scores but these reading scores were not significantly lower than would be predicted by their IQ scores, traditionally called poor readers. On a variety of reading and spelling tests, I found that there were no differences between these two groups. In other words, there is no need to use IQ scores to predict the difference between the children traditionally called dyslexic, and those who have equally poor achievement but also have lower IQ scores (the low achievers). I found exactly the same results in a study of adults with reading disabilities. The reading disabled group was quite homogeneous in relation to reading related skills and administering an IQ test would not provide useful information about performance differences on reading related tasks. These findings suggest that there is no need to use IQ tests to determine who is learning disabled. We need to use only achievement tests.

One of the major reasons that we are still failing to identify learning disabilities is the preoccupation with testing that dominates the field. In order for a child to be identified and treated by a school system, there is usually an intensive testing session. Typically, there is a long wait, often as long as two years, which is a very long time to wait if you are an eight-year-old having problems with reading. Usually an IQ test is required, in spite of the evidence that it is really not necessary in the definitional process and that it does not guide remediation in any meaningful way. We have made the identification process too time consuming and irrelevant to the educational needs of the child or adult.

The most telling argument against IQ scores is that they do not predict the ability to benefit from remediation. In a recent comprehensive review of the literature, Stuebing, Barth, Molfese, Weiss, and Fletcher (2009) showed that IQ does not predict the ability to benefit from remediation.

One of the difficulties with the excessive reliance on psychoeducational testing is that remediation is often neglected. That is, there is no attempt to relate the test scores to what educational methods and strategies might be used to help the child. Hopefully, the system will encourage the development of specific and detailed assessment of academic skills that will provide useful clues about how to provide remediation for a person with learning problems.

Out of the quicksand?

After hitting bottom, we began the climb upward. In a recent call for research, the U.S. government acknowledged that we do not need the discrepancy definition. However, in many official definitions of a learning disability, the discrepancy definition is still used. Unfortunately, people with difficulties in reading or mathematics who also have lower IQ scores, even if these IQ scores are not very low, do not get help even today. The solution is to abandon the IQ test and just use low achievement as a measure of a learning disability. Obviously if a child is seriously delayed or has a condition such as Down Syndrome or autistic spectrum disorder or hearing or visual impairment, there are problems other than low achievement in reading and mathematics. However, the vast majority of individuals with difficulties in reading and mathematics would benefit from intervention, whatever their IQ score.


Stuebing, K.K., Barth, A.E., Molfese, P.J. Weiss, B. and Fletcher, J.M. (2009). IQ is not strongly related to response to reading instruction: A meta-analytic interpretation. Exceptional Children, 76, 31-53.

Linda Siegel is Professor of Educational and Counselling Psychology and Special Education at the University of British Columbia, and is also President of the Division for Learning Disabilities (DLD) of the Council for Exceptional Children (to June 2010). She has conducted extensive studies of dyslexia and other learning disabilities as well as cognitive and language development. She was awarded the 2010 Canadian Psychological Association Gold Medal Award for Distinguished Lifetime Contributions to Canadian Psychology. Email: linda.siegel@ubc.ca.

This article was previous published in Learning Difficulties Australia Bulletin, Volume 42 No 2 July 2010, pp. 14-15.  © Linda Siegel. All rights reserved.