Reading Success Lab Blog

Things I’ve Learned in a Clinical Setting

I spent over thirty years as an academic researcher and learned many things about cognitive functioning in those years. You can see my formal research activities by going to my profile on the Reading Success Lab website (see my profile), and clicking on the “Mike Royer’s VITA” link at the end of the profile.

But I also learned a lot, particularly a lot about individuals who had difficulty learning to read or do math, in my role as director of a laboratory at the University of Massachusetts (The Laboratory for the Assessment and Training of Academic Skills). My clinical lab provided assessment and intervention services for individuals with academic learning difficulties. Individuals who came to my lab varied in age from 6 to 40, included both boys and girls, men and women, and spanned a spectrum of ethnicity and race.

In the next few posts I will describe things that I learned by actually working with individuals with learning disabilities. The things I learned were things that I suspected to be true based on my academic research. But actually working with LD (learning disabled) individuals provided the evidence I needed to support my suspicions. The evidence from my observations was not the kind of evidence I could publish in scientific papers, but it did provide guidelines for me to improve the academic skills of many of the individuals that came to my lab.

As you will see in the cases I describe, we were not always successful in helping the clients we were working with. Many of our clients did improve, some in spectacular ways. But we also had failures. We learned from both the successes and the failures and the posts to follow will describe what we learned from both outcomes.

Are Math Skills Innate?

Somewhat surprising, the answer to the above question is, in part, yes. Imagine a study that involves a psychological phenomenon called habituation. When we see things that are familiar we tend not to pay attention to them. In contrast, when we see things that are novel and new, they attract our attention. We look longer at new things that we do familiar things. That is, we are habituated to familiar things, but not to new things.

Now imagine a study where 6 month old infants are presented with series of dots on a screen that range from one dot to 6 dots. The way the experiment works is that an infant sees the screen with the dots and then another screen blocks the dots. Soon thereafter the blocking screen is removed and the infant can again see the dots. However, now the infant sees the same number of dots that originally appeared or fewer or more dots that originally appeared. If the number of dots that originally appeared is four or less, and if the number of dots is the same as those that originally appeared, the infant very quickly looses interest in the dots and looks away. In short, the infant has habituated to the number of dots that originally appeared.

If however, the original number of dots was four or less, and there are now more or less than four, the infant looks longer at the new number of dots. The surprising finding is that if the original number of dots was four or less, the infant looks at the new or old dots about the same amount of time regardless of the number of dots. That is, the infant seems to process four dots just as fast as he or she processes one dot. In short, there appears to be an automatic (Innate?) ability in infants to process up to four items that appear in a visual field.

The same is not true for more that four items. With each item beyond four there is a steady increment of time required to process each item. It seems as if processing more that four items requires a different cognitive process, and that with each additional item an additional amount of time is required for processing.

The phenomenon of automatically recognizing up to four items is called “subtizing,” and suggests that we are born with the ability to almost automatically recognize up to four things without counting. Another interesting fact is that we are not the only species that have subtizing ability. This will be discussed in a further post. However, the ability to subtize a number of things is limited to four. Beyond that, infants cannot automatically recognize changes in the number of things.

Beyond four things requires a learned ability called counting. Subtizing and counting form the early cognitive basis for mathematical ability. Research on subtizing and counting explores many interesting facets of our mathematical ability. For example, there are individual differences in subtizing and counting. Are these differences predictors of mathematical ability as children age? Another question is whether there are specific locations in the brain that are activated during subtizing and counting? And what about the question of whether organisms other that humans can subtize? These questions will be examined in future posts.


Is Dyslexia Inherited?

What does scientific research say about whether dyslexia is inherited? The answer to this question requires an understanding of some background issues. First, genes rarely dictate that a particular characteristic is certain to be passed from one related person to another. Rather, shared genes most often increase the probability that a trait will be passed from one family member to another. Second, the probability that a trait is influenced by genetics is usually expressed by a heritability index. A heritability index of 1.0 would indicate that a family member who possessed a trait would be certain to pass that trait along to his or her progeny. In contrast, an index of 0.0 would indicate that genetics plays no role whatsoever in determining the existence of a particular trait among family members. There are virtually no circumstances where heritability indices of 1.0 or 0.0 are actually found in research so the majority of the indices actually found fall between 1 and 0.

There are two classic methods for estimating the influence of genetics and the environment on the development of dyslexia. The first is the study of identical and fraternal twins. Identical twins, of course, have exactly the same genetic makeup. Fraternal twins share only half of their genes. However, both identical and fraternal twins most often share the same environment. If genetics is most important in determining dyslexia then identical twins should have a higher heritability index than fraternal twins. In contrast, if environment is most important there should be little difference between the heritability index of identical and fraternal twins.

The three largest twin studies are those conducted in Colorado, in London, and in England and Wales (the studies are still ongoing). The typical methodology in the studies is that the researchers are notified of the existence of twin children in a particular geographical area and they then recruit the children as participants in the research study. At the same time the researchers recruit a matched sample (i.e., same age, same sex) of control children from schools the twins attend. A battery of tests are then administered to all of the children over a number of years. These tests allow the determination of the extent to which genetics and the environment are related to the development of reading skills.

The twin studies from the three major research projects yielded very similar results. The heritability index for reading skills was about .65 whereas the environment index was about .25.

The second classic method for examining the heritability of dyslexia involves the recruitment of participating families. One set of families (called at risk families) have at least one parent that has dyslexia. The second set of families (control families) are recruited to be as similar as possible to at risk families, but do not have a parent who has dyslexia. Again, the procedure is to administer tests to children in the families that will allow a determination of the extent to which reading development is proceeding normally.

The results of the family studies, like the twin studies, showed a substantial contribution of genetics in the development of reading skills, and only a moderate influence of shared environment.

How To Tell The Difference Between a Poor Reader And a Disabled Reader

Imagine that two elementary age readers take a standardized reading test. Both readers perform poorly on the test. Yet, there are significant differences between the two readers. One reader, often called a garden variety poor reader in the research literature, may perform poorly because he or she didn’t learn important pre-reading skills before entering school. The other reader, called a reading disabled reader, may perform poorly because they have a disability (e.g., dyslexia) that inhibits his or her ability to learn how to read. The difference between the two readers is important in deciding the approach educators take in teaching the child to learn to read. But, how do you tell when reader is a poor reader because of environmental consequences, or because of an inherent disability?

Traditionally, there are several approaches to answering this question. One approach is to administer a battery of tests that measure general cognitive ability and a variety of reading skills. The logic behind this approach is that a reader who has been environmentally deprived will show deficits in a range of cognitive skills, in addition to reading skills. In contrast, a disabled reader will have normal cognitive skills in all areas except reading.

A second, more recent approach, is called Response to Intervention (RTI). The RTI approach is to provide all poor readers with what is known as “best practice instruction,” (see an earlier post on this site for more information about best practices). The idea is that the reader who reads poorly because of lack of early skills will soon show considerable progress when exposed to best practices. In contrast, the truly disabled reader will continue to struggle even when exposed to quality reading instruction.

Both of the approaches mentioned above have disadvantages. The test batteries mentioned in the first approach can only be administered by trained diagnosticians and the process and subsequent report can easily cost over a thousand dollars. The best practice approach also has shortcomings. Identifying the disabled reader may occur only after several years of failed best practice instruction. And then when the identification does occur, there are no obvious alternative approaches when best practice has not produced positive results.

The Reading Success Lab Reading Evaluation module takes a third approach to the problem of differentiating between the garden variety poor reader and the disabled reader. Our software evaluates performance on a simple cognitive skill, and on a series of progressively complicated reading skills. The logic is that the garden variety poor reader will show depressed performance on all of the skills, whereas the disabled reader will show areas of normal performance, and areas where the disability creates depressed performance.

Here are two graphs that show the profiles for the two types of poor readers.

The values on the tasks are grade level percentiles where the average in a student’s grade level is as the 50th percentile. The tasks, reading from left to right across the graph are: 1) a simple response time task, 2) a letter naming task, 3) a word naming task, 4) a pseudo word naming task, 5) a word meaning task, 6) a sentence understanding task, 7) a listening comprehension task, and 8) a reading comprehension task. Note the generally depressed performance on all the tasks for the garden variety poor reader.

This profile is for a disabled reader. This reader performs fine on the simple task, the letter naming task and the listening comprehension task, but has depressed performance on the remaining reading tasks. Our reading evaluation software provides an inexpensive, quick and valid approach to differentiating between poor readers and disabled readers.

Two Kinds of Dyslexia: Developmental and Acquired

An interesting comparison can be made between poor readers who have experienced brain trauma associated with an injury or a stroke, and individuals who experience reading difficulties in the absence of brain trauma. The research literature has described the first type of individual as having “acquired dyslexia” and the second type of individual as having “developmental dyslexia.”

There are several variants of both kinds of dyslexia but the theoretical description of why individuals with acquired dyslexia and why individuals with developmental dyslexia have reading problems is strikingly different. To understand the differences we need to understand a little bit about how we can recognize words when we read. When learning to read we store two types of “word memories” in our heads. The first type is the word image of a word. This is essentially a photograph of the way a word appears. Having developed such a representation we can look at a word, and then activate the word image of the word. Having done this we still need a further process: we need to transform the word image into the sound of the word. In essence, we activate the image of the word and then determine how the word sounds in speech.

The second type of word memories involves the storage of both word images and the phonemes that make up words. Phonemes are the smallest unit of sound in spoken languages. For a detailed presentation of the role of phonemes in learning to read, see the developing skilled reading link at the bottom of this page. As reading skill develops, the way that normal readers activate the sound of a word changes. Initially, the individual uses both the word image and the phonological memory for a word. But as skill improves, the activation of the phonological representation overrides the activation of word images. This occurs because activation of the phonological representation is faster.

Now we come to acquired and developmental dyslexia. The brain trauma that results in dyslexia like symptoms (acquired dyslexia) inhibits the activation of phonological representations and word image representations with particular damage to word image representations. In contrast, individuals with developmental dyslexia have difficulty accessing the phonological representations of words. In particular, they have difficulty recognizing words that have irregular pronouncing patterns (e.g., iron) and letter sequences that are not words (e.g., sloke) but can be pronounced. Both types of dyslexia present as an individual with impaired reading skills, but the underlying causes of the impairment are quite different.


More than One Learning Disability at the Same Time?

Is it possible for an individual to have more than one learning disability at the same time? Indeed, it is possible, and in fact is even common. When a person is diagnosed with two or medical conditions at the same time, the diseases or disorders are said to be comorbid. That is, a condition exists simultaneously with, and independent of one or more other conditions.

Studies have shown that a given learning disability is often comorbid with one or more other learning disabilities. The overall rate of comorbidity is 40%. That is, 40% of the individuals diagnosed with a learning disability also have been simultaneously diagnosed with another, separate, learning disability. ADHD has even higher rates of comorbidity. The comorbidity of ADHD with other disorders is between 60% and 80%. Reading disability (RD) and math disability (MD) also have very high rates of comorbidity with both occurring in 30 to 70 percent of individuals with either disorder. Evidence suggests that individuals with comorbid learning disabilities may be more disadvantaged that individuals having a single disorder.

Early Indicators of Reading Disability

A longitudinal study reported by Finnish scientists presents compelling evidence that indicators of subsequent reading disability are present at birth. The research team examined two groups of children for over a decade. One group was designated “at-risk” for a reading disability based on the children having two first degree relatives with a reading disability. For example, a mother and a close relative of the mother. The other group of children were age-matched peers.

The children were initially examined shortly after birth (within 3-5 days) using a brain monitoring procedure called event-related potential (ERP). The children were presented with the sounds of consonant-vowel combinations and the ERPs measured the preferred hemisphere of the brain that was activated when the sounds were heard. The at-risk children showed a preference for processing the sounds in the right hemisphere of the brain whereas the control children showed a preference for left hemisphere processing. It should be noted that in the vast majority of individuals language processing occurs in the left hemisphere.The children were also examined using brain measuring techniques at 6 months of age. These measures showed differences between the groups in discriminating aspects of phonemes. Phonemes are the individual sounds that are combined to form words.

The researchers also collected additional measures as the children aged. These included measures of phonological awareness, letter identification, and ability to rapidly name familiar stimuli such as letters or numbers. Again, the researchers reported differences between the two groups on these measures.
All of the measures that differentiated between the groups as they matured turned out to be significant predicters of reading ability in the second grade.

Learning to Read English is Hard

Brain imaging research has shown that the percentage of children displaying brain patterns indicative of dyslexia in different countries is similar. However, the percentage of children actually having difficulty in learning to read varies greatly across countries. The reason is a characteristic of languages known as transparency.

Transparency refers to the degree to which the orthography (word spelling) of a language directly corresponds to the phonology (word sound) of the language. Finnish and Italian, for example, are very transparent languages. There is a near perfect match between the spelling of a word and the sound of a word in those languages. English, in contrast, is the least transparent language of those languages that have been studied. For example, there are no words like iron and was in the Finnish language. As an aside, in the author’s (James Royer) work with dyslexic children he found that was was misidentified more frequently that any other word. This is partly a function of the frequency of the word in written language, but also a function of the fact that there is no way the word can be identified by sounding out the letters.

Transparency has an enormous impact on learning to read. In Finland 45% of children enter school at age 7 as competent readers and nearly all children are good readers by the end of grade 1. In contrast, 20 percent of children in the U.S. entering grade 1 are at risk for reading failure and many are still not competent at the end of grade 3. In short, it takes an average English reader three years of schooling to reach a level of reading competence attained by nearly all Finish readers in a single year.

Language transparency also makes it much more difficult for a dyslexic child to learn to read. The very large number of irregular words in English (words where spelling and sound do not correspond) means that these words must be memorized, placing an added burden on the dyslexic child. Transparency also has an impact on reading instruction. For example, the teaching of phonics, the process of sounding out words, is more difficult in languages that are not perfectly transparent.

When “Best Practice” Reading Interventions Fail

Research has shown that approximately 20 percent of U.S. children enter school at risk for reading failure. Schools typically respond to this concern by utilizing early identification and intervention procedures designed to prevent the occurrence of reading failure. Most often, these procedures rely on the heavy use of phonics based methods. These methods are often referred to as “best practice” procedures because they have been shown in rigorous research studies to improve the reading skills of many students at risk for reading failure.

However, best practice reading interventions are not effective for all students at risk for reading failure. Research has shown that a significant number of at risk students that receive best practice interventions do not improve their reading skills. These students are often referred to in the research literature as being “treatment resistant.”

Royer and Walles have published research that shows that treatment resistant students do show reading improvement when exposed to the fluency based interventions that are embedded in the Reading Success Lab software. Working with students that had not made good progress after years of phonics based interventions, Royer and Walles showed that the students did make significant reading gains using fluency training. Moreover, fluency based procedures have also been shown to result in reading improvement in students who do respond positively to phonics based instruction.

Read the Royer and Walles published research:

> Fluency Training as an Alternative Intervention for Reading-Disabled and Poor Readers