It is annoying when one is in a long line -- at a ticket counter, at a supermarket, at a bank -- and someone "jumps the queue," taking a position in line ahead of other people who lined up first. The title of Mark Kelman(1) and Gillian Lester's(2) book, Jumping the Queue, gives the reader advance warning of the authors' position on people who edge ahead in line. But the topic of their book is not ticket, supermarket, or bank lines, but rather the line to enjoy the benefits of society. And the focus of the analysis of queue-jumpers is not on customers in the commercial marketplace, but on a particular group of students in the academic marketplace. These students are ones who have been identified as having learning disabilities.
Kelman and Lester's book covers diverse topics in its eight chapters totaling 313 pages, including technical controversies (pp. 17-36), the federal regulatory framework (pp. 37-66), local practice (including diagnosis and placement) (pp. 67-92), resource management and discipline (pp. 93-116), extra resources for the classroom teacher (pp. 117-60), accommodation on law school exams (pp. 16194), and ideology and entitlement (pp. 195-226). There are many points of view from which learning disabilities can be approached, and the authors' point of view, indicated by the subtitle, "an inquiry into the legal treatment of students with learning disabilities." The subtitle is appropriate; the book reads more as an inquiry than as a presentation of a strong stand regarding what needs to be done, legally or otherwise. At times, the authors' unwillingness to take strong or even clearcut stands is frustrating.(3) Conclusions often get lost in what, for two reviewers who are psychologists, appear to be technical legal thickets.(4) But it is clear that Kelman and Lester are skeptical of the preferential treatment given to those identified with learning disabilities, because they point out -- correctly, we believe -- that the accommodations that benefit individuals identified as having learning disabilities would benefit virtually anyone (pp. 17273). The authors also express skepticism of whether the system is just, granting as it does special legal privileges to those who have no unique moral, psychological, or educational claim to these privileges (Chapter 8).
Because the book is an examination primarily of legal issues, it addresses somewhat superficially what we believe to be the most fundamental problem pertaining to learning disabilities. This problem is that the concept as it is used in practice is invalid. We have no doubt that the concept of a learning disability is, in theory, veridical. But there is a big gap between theory and practice. We seek in this review to deal with this issue, because it renders the societal legal discussion moot. The laws cannot be just if they are based on a classificatory system that makes little or no psychological or educational sense. The book also may make a false assumption in assuming there is a single queue. Neither success, nor abilities, nor practically anything else that really matters in life is unidimensional. Learning disabilities certainly are not.
I. DEFINING A LEARNING DISABILITY
Specific learning disabilities have been defined in a number of different ways. A consensus view has emerged that is expressed in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) of the American Psychiatric Association (APA) according to which learning disabilities (referred to as "learning disorders" in DSM-IV) are marked impairments in the development of specific skills, such as reading skills, relative to the level of skills expected on the basis of an individual's age, education, and intelligence.(5) The term for this is "differential" skills. These impairments interfere with daily life and academic achievement. They are not due, however, to physical deficits, such as visual or hearing deficits, or to acquired neurological conditions, such as those caused by brain trauma. Learning disabilities seldom can be diagnosed before the end of kindergarten or the beginning of first grade. In practice, these disabilities tend to be identified on the basis of differences in performance on tests of abilities and tests of achievement.
In its effort to categorize students and define recipients of benefits, both the law and Jumping the Queue take for granted the validity of this definition and method of detecting learning disabilities. The thesis for this review is that this assumption is incorrect on its face. We view learning disabilities in a way that is somewhat contrary to this kind of standard definition. Our thesis is that virtually everyone has some kind of learning disability but that society only chooses to recognize some individuals as such. Whether someone is labeled as learning disabled in many respects resembles the result of a lottery.(6) Here's why.
All reputable theorists of abilities agree that abilities are multiple -- that there are many of them.(7) Even Charles Spearman, the father of the theory of general ability, stated that in addition to a general ability that applies to all tasks there are specific abilities that apply to specific tasks.(8) No serious psychological theorist, to our knowledge, has claimed that there is only one kind of ability. Although theorists may disagree as to exactly what the abilities are or how they are structured, they agree that the abilities are distinguishable from each other. For example, the skills that constitute reading ability are different from the skills that constitute mathematical ability, which are in turn different from the skills that constitute musical ability.(9) Thus, someone could be an able reader but a poor musician, or vice-versa. Perhaps not in school, but almost certainly later in life, the person who seeks a career that uses mathematics heavily will be standing in a different queue from the person who uses music heavily.
Because abilities are multiple, it is odd that the operational definition of learning disabilities so often relies on the measurement of IQ. The use of a single quantity for an IQ implies a unitary basis for intelligence that does not exist in psychological theory and is extremely unlikely to exist in reality.
If one were to make a list of the many abilities people can have, one would find that virtually no one is proficient in all the skills constituting all of these abilities. At the same time, virtually no one is hopelessly inept in all of these skills. Rather, almost everyone is more proficient than are other people in some skills and less proficient in other skills. Some people may be proficient in more skills, or more proficient in certain skills, but virtually everyone shows a pattern of multiple strengths and multiple weaknesses.
Put another way, virtually everyone shows a complex pattern of abilities and disabilities. For example, even the straight-A student in school may be inept in certain aspects of interpersonal relations. Even the straight-F student in school may be able in many aspects of dealing with other people. This intuition is captured in modern theories of intelligence, which argue on the basis of plentiful and diverse data that interpersonal and practical skills actually are distinct from traditional academic skills. High levels of these different kinds of skills may or may not be found in the same persons.
Given that everyone has a pattern of abilities and disabilities, how does it happen that some people get labeled as having learning disabilities whereas other people do not? The reason is that labeling of someone as having a learning disability is the result of an interaction between the individual and the society. Learning disability resides neither totally in the individual nor totally in the society. It is not simply a matter of having a certain set of genes nor is it simply a matter of society's defining someone who is perfectly adaptive as having a disability. Rather, the society selects some people to label as having learning disabilities and does not select others.
Our point of view differs from a common view, which is intrinsic and states that learning abilities reside within the individual. Our point of view also differs from the view that learning abilities are extrinsic and reside solely in the labeling of society.(10) We argue that both of these views are untenable.
How does society decide whom to label? It decides on the basis of the set of skills that it values in school and on the job. If society views a certain set of skills (such as reading skills) as essential, and views them as constituting a "specific" rather than a general ability, then individuals with low levels of proficiency in these skills are labeled as having a specific disability. One has a set of abilities and disabilities and metaphorically enters a lottery that determines whether the particular pattern will result in a learning disability label. We are not saying that the labeling process is wholly arbitrary. Rather, we are saying that there are many different possible labeling processes that can yield totally different results. The discussion in Jumping the Queue is based upon a common legal conception of learning disability that is only one of many different conceptions a society might have.
II. DISABILITY AS A RELATIONSHIP BETWEEN PERSON AND ENVIRONMENT
Our society, through federal law as instantiated in the Individuals with Disabilities Education Act (IDEA),(11) currently recognizes various types of learning disabilities: listening, speaking, basic reading skills, reading comprehension skills, written expression, arithmetic calculation skills, and mathematics reasoning skills.(12) Many other skills might be included on this list. For example, the list might include spatial skills that are important in finding one's way around a city, driving a car, or fitting suitcases into the trunk of a car. It might include musical or artistic skills if the society valued these highly enough to label those individuals whose levels of them are distinctively low as having a disability. The list might include creative skills if society believed that people who virtually never have their own good novel ideas lack something important. Really, the list might include any ability at all!
Learning disabilities are viewed as specific. But in fact, there are no completely general abilities or disabilities. For example, IQ tests are sometimes seen as measuring "general ability," but in fact, a high score on an IQ test is no guarantee of a high level of creative ability, practical or common-sense ability, athletic ability, musical ability, or any of a number of other abilities. So all abilities and disabilities are specific in greater or lesser degree.
Where and when an individual is born have a tremendous impact on whether that individual will be labeled as having a learning disability. In a preliterate society, for example, there are no individuals labeled as having a reading disability. In such a society, a man lacking the spatial or athletic skills to be a good hunter might have a disability that genuinely could affect his ability to function well in that society. One society might label someone with minimal musical skills as having a musical disability, whereas another society might not. In effect, each individual becomes a mandatory participant in a lottery that determines whether the particular pattern of abilities and disabilities he or she has will lead to the individual's being labeled as having a learning disability. But the lottery applies only to the labeling process. Everyone has a pattern of both abilities and disabilities. The lottery represents how society chooses to label that pattern. Thus, each society does not merely have a different queue. It has different queues.
Given that all individuals have both strengths and weaknesses, individuals labeled as having learning disabilities have many sources of strength to offer society. But the society often inadvertently positions these individuals to view themselves as potential victims rather than as potential victors, a point we believe is consistent with the tenor of Kelman and Lester's book (pp. 225-26). Indeed, Kelman and Lester state on the last page of text that it "will always be tempting to `jump the queue' by claiming that one's distributive interests take priority over the interests of another group: claim hopping on the (ideological) backs of instances of genuine victimization ..." (p. 226). In this regard, we have three main contentions:
First, individuals with specific learning disabilities often have considerable strengths in other abilities. Second, these individuals should be encouraged to view themselves as victors who capitalize on strengths instead of viewing themselves as victims of their disabilities and thereby adopting the mindset of people who are victimized. They should find paths in their lives that enable them to capitalize on their strengths, not on their weaknesses.
Third, curriculum modifications that excuse these individuals from learning important skills or from the normal experiences of schooling may be well-intentioned, but can end up hurting these individuals more than helping them. The reason is that for the society to have labeled the individuals as having a learning disability in the first place, the ability for which they were labeled as lacking (disabled) must be one that the society views as essential for adaptive living in that society. The laws so carefully discussed in Jumping the Queue apply only to societally recognized disabilities, not to all possible disabilities. Indeed, as at the present time there is a large degree of arbitrariness in what deficits are officially recognized as "disabilities," it would not even be feasible to cover all possibilities. Where would the list end? Potentially, the list is endless, as everyone is weak in at least some skills, including skills that may be very specific, such as threading a needle or kicking a football.
It is wrong to write off individuals identified as having learning disabilities. Ignoring such children or treating them as hopeless is unfair and not in either the children's or society's best interest. These individuals are far from hopeless, unless we steal their hope from them. We believe that individuals with learning disabilities should be helped to make the most of their potential.
We believe that individuals with learning disabilities often have enormous strengths that the current system of education neither taps nor even draws out. The special services that society provides often do not fully help students with learning disabilities develop their strengths or encourage them to use their strengths to learn. Yet their learning disabilities may force them to develop strengths that children who are not identified as having learning disabilities have no incentive to develop.
In our view, the "learning disabled" or "LD" label can be costly both to the individual and to society. Once children are labeled as "LD," a complex set of mechanisms is put into effect that renders it likely that the label will become a self-fulfilling prophecy, whether it was originally correct or not. A well-intentioned labeling procedure thus can become harmful to our young.
As mentioned above, genuine LD is an interaction between the individual and the environment. Certain biological predispositions can put an individual at risk for the development of specific kinds of learning disabilities. They do not determine whether the individual will have a learning disability. Even among literate societies, some orthographic (writing) systems impose challenges that others do not impose, challenges that affect the probability of individuals manifesting reading disabilities. For instance, Spanish is pronounced almost exactly the same as it is written; English is not. Chinese uses a logographic (picture-based) writing system, whereas Indo-European languages such as English, French, German, and Russian use alphabetic writing systems. Whether a child becomes identified as having a reading disability will be affected not only by that child's biological makeup, but also by where and when the child grows up. Simply speaking of "learning disability" or "reading disability" is too simplistic to make sense.
Biological does not imply immutable. Even to the extent that the origins of learning disabilities are biological, these biological origins have nothing at all to do with whether the symptoms of learning disabilities are modifiable. Put another way, the partially biological origins of learning disabilities in no way preclude successful educational interventions. "Biological" is in no way synonymous with "fixed." For example, height is highly heritable, but heights have increased over the years. Phenylketonuria is 100 percent heritable, but a diet free of phenylalanine can stave off mental retardation in affected children. Even IQ, which is heritable in some degree, has been found to be rising about nine points per generation during the years of the present century (when raw scores [number correct] rather than scaled IQ scores are used).(13)
The remedies used to improve performance of individuals with learning disabilities should always depend on the specific deficits individuals experience, not on how these individuals are labeled. Lumping together all of the labeled children into one global category such as LD, and then giving them what usually amounts to a single form of remediation, may hurt a majority of the students and help only a few.
III. THE WRONG PATH
We know what to do; we do not do it. Cognitive theory provides effective interventions for many of the conditions today lumped together as "LD."(14) Many schools make little or no use of any of this knowledge. Why?
First, the financial interests of the schools may drive labeling. Schools regrettably can have a financial interest -- state and local funding, often deriving from federal funding -- in overidentification. When they are "paid by the head" for the identification of each child with an LD, they are obviously tempted to label more children as having an LD.
Second, over-labeling children may take parental pressure off school administrators. The school administrators no longer have to deal with those parents who clamor to have their children labeled as having an LD.
Third, labeling children as having an LD can provide a way of coming closer to reaching the Holy Grail of improved test scores. By giving children more time on tests and providing other accommodations for test-taking such as special quiet rooms, administrators potentially can raise the test scores of children with learning disabilities and thus raise the averages for their schools or districts. The schools or districts thus look better and get paid extra federal or state funds in the process. But their labeling signifies neither an understanding of the problem nor an understanding of a solution.
Fourth, labeling is not tantamount to understanding. Unfortunately, people sometimes believe that they understand a phenomenon merely because they are able to assign a label to it. They then fail to seek understanding of the phenomenon because they do not know that they do not understand it. A teacher or administrator may believe that the LD label is all that is needed to understand the child's deficiencies and perhaps even to remediate them.
Fifth, much of the cognitive psychology literature has been communicated in ways that are not fully comprehensible to educators involved in the schooling of children with learning disabilities. As a result, these educators could not make use of the findings, even if they wanted to. There is a communication gap that will be filled only when researchers better understand the problems of educators and educators better understand the language of researchers.
Sixth, children with learning disabilities once were taught primarily by teachers with advanced training in special education. Today, with the advent of full-inclusion models whereby most children with learning disabilities are placed in regular classrooms, teaching of these children is done much or all of the time by teachers with no specialized training in how to deal with these children effectively.
As we have noted, one problem with the way learning disabilities are identified is the assumption that intelligence is a unitary entity. Another problem is the use of discrepancy (difference) scores.
Learning disabilities often are defined in terms of discrepancies between IQ and a measure of a specific kind of learning ability. For example, a child might be diagnosed with a reading disability if the child scores much lower on standardized tests of reading ability than on standardized tests of intelligence. On its face, this procedure would seem to make sense: the child is intelligent, but a poor reader. What is wrong with such a sensible procedure for recognizing the existence of learning disabilities?
We believe there are at least seven reasons why discrepancy scores are invalid, regardless of their legal sanction.
1. The assumption that IQ tests measure all there is to intelligence is extremely questionable. Many modern theorists question it.(15) For example, Sternberg has proposed a theory of successful intelligence, according to which intelligence comprises three parts: analytical -- used in analyzing, judging, evaluating, and comparing; creative -- used in creating, inventing, discovering, and imagining; and practical -- used in putting into practice, applying, using, and implementing.(16) IQ tests measure only a portion of the analytical abilities; they typically measure creative and practical abilities either not at all or hardly at all. So here, too, using an IQ test as a basis for evaluating someone's intelligence will provide only a highly incomplete measurement, at best. And as in Gardner's theory, to the extent that reading involves analytical (or other) abilities, one is in part subtracting a thing from itself.
In a number of studies, we and our colleagues have studied the practical aspects of intelligence.(17) For example, we have devised tests of practical intelligence for business managers, salespeople, college teachers, army leaders, and students. In study after study, we have found that scores on tests of practical intelligence predict real-world performance as well as or better than do IQ tests, but that these tests of practical intelligence are not correlated with conventional IQ tests. In other words, one could not predict a person's practical intelligence from his or her score on an IQ test, and vice versa. If we had available only the modest IQ score, we might conclude that someone is not very intelligent; in fact, that individual might be high in practical intelligence, and thus have the ability to excel in the everyday world. Or the person high in IQ might have low practical intelligence. Should we refer to such an individual as practically disabled? Once we expand our conception of intelligence, the disability notion becomes problematical.
In the U.S., we typically have found little or no relation between IQ and practical, everyday intelligence.(18) We found an even more surprising result in a study we did in Kenya.(19)
Children near the village of Kisumu, Kenya are ill much of the time. Parasitic infections that are not often found in the developed world continually assault them. Remaining well is therefore an essential aspect of practical intelligence.
Part of the indigenous education of such children is learning about natural herbal medicines that the Kenyan villagers believe fight parasitic infections. Village children may know the identities of many such medicines, but children in the developed world, of course, could not identify a single one.
We constructed a test of the practical intelligence of the village children in which the children had to identify these medicines, as well as when and how to use them. This test thus measured implicit knowledge that is considered important in the Kenyan village. We correlated scores on our test of indigenous practical intelligence with scores on two tests of vocabulary and with scores on a test of abstract reasoning ability that consisted entirely of geometric figures. The two vocabulary tests were in different languages: one was in English, the language of the school; the other was in Dholuo, the language of the Luo residents of the village. We found that scores on the test of indigenous practical intelligence were significantly negatively correlated with scores on the test of vocabulary for the language use in the school -- English. In other words, the children who did better on the indigenous test actually did worse on the Western types of tests. Scores on the indigenous intelligence test were uncorrelated with scores on the vocabulary test in the Dholuo language and with scores on the test of abstract reasoning ability.
Given this information, just how seriously can we take the scores on the IQ tests, not only for certain Kenyan village children, but for children from any subculture in the Western world that perhaps takes the kinds of skills measured by IQ tests less seriously than does the mainstream U.S. culture? In the rural villages of Kenya, many of the children are failing in school, and the parents show little or no concern. They just do not value the types of skills valued highly by Western education and plan to separate their children from these schools as soon as they are legally able to. They value the skills that matter in the community, and they perceive the training their children receive in school as largely irrelevant. If we were to view one group as having a disability, would it be the children who do not have school-based knowledge, the children who do not have community-based practical knowledge, both groups, or perhaps some other group?
The Kenya study also has implications for our understanding of verbal disabilities. Do we judge such disabilities from the standpoint of academic vocabulary and its use or from the standpoint of everyday vocabulary and its use? The children do not score well on conventional academic verbal tests, even in their home language of Dholuo. But they do score well on such tests when the words are used to ask them about indigenous medicines. The same kind of issue can apply anywhere. Someone may be able to read and understand the material he or she confronts in everyday life, without being able to read the kinds of more difficult passages typically found on reading tests.
Variation in conceptions of intelligence is found not only outside the U.S., but inside it as well. We found in a study in San Jose, California, that Latino parents more emphasized social skills in their conceptions of intelligence than did Asian and Anglo parents, who more emphasized cognitive skills.(20) But a child is more likely to be labeled as having a disability if he or she lacks cognitive skills than if he or she lacks social skills. Even in the U.S., different groups have different notions of what constitutes intelligence.
2. Even if one accepts IQs at face value, they are problematical because they are confounded with verbal and even reading skills. The large majority of IQ tests involve reading material, or at least verbal material, which may be presented orally. The comprehension skills needed to understand the material on the IQ test overlap with the comprehension skills measured by reading tests. For example, IQ tests often measure vocabulary and reading comprehension skills. The result is that when reading scores are subtracted from IQ scores, the subtraction is invalidated by the fact that one is partly subtracting a thing (verbal-comprehension skills) from itself.
Because both reading scores and IQ scores involve verbal components, the subtraction cannot give you a pure measure of verbal or reading skills taking out everything else. The truth is that the subtraction tells you next to nothing because you do not know how much of the IQ score was dependent on verbal ability.
The situation becomes even more ridiculous when one includes a "reasoning deficit" as a learning disability, which Kelman and Lester report is the case at Stanford University (p. 165). We are aware of no test of intelligence or academic ability that does not place heavy emphasis on reasoning. To the extent one views poor reasoning as a disability, it would be difficult to separate from intelligence, especially because definitions of intelligence going back to Terman generally emphasize the importance of reasoning skills in intelligence.(21)
3. Use of nonverbal IQ tests does not solve the problem. In an effort to get around the problem of verbal content, some psychologists use nonverbal IQ tests, that is, IQ tests that contain no words. These tests might contain geometric figures on the basis of which one is supposed to reason. For example, one might be asked to say which one of five geometric figures does not belong with the other four.
The test appears to contain no words. But things are not quite as they appear. The directions are verbal, and some children do not understand the directions they are given for tests. Solutions of the problems are also often verbal, as children talk out answers to themselves. But the greatest problem is that now the IQ test represents an even smaller portion of the full range of intellectual abilities, so that to the extent it is supposed to measure all of intelligence, it is being asked to do something it cannot possibly do.
Reasoning with abstract geometric forms is a far cry from the kind of intelligence we need to adapt in our everyday lives, and a test that contains nothing more than such sterile problems cannot possibly give a complete representation of a person's intelligence. Such a test should not serve as the basis for a score from which to subtract a reading score.
4. Difference scores do not mean the same thing at different points along the IQ spectrum. Imagine, for example, someone who is a near-genius in abilities overall but who is only above average in reading abilities. If one views reading disability as defined in terms of the difference between IQ and reading ability, the near-genius actually might be classified as having a reading disability. In fact, the individual's reading ability is way below the individual's overall abilities, but so what? This is not an individual in whom one would want to invest any of society's resources so as to bring his or her reading up to the level of his or her other abilities.
The same problem that can arise at the top of the abilities spectrum can also arise, with more serious consequences, at the bottom of the abilities spectrum. Suppose an individual is slightly below average in general abilities, but way below average in reading. The difference between the IQ and the reading score, however, does not quite reach the threshold for classifying the individual as having a learning disability. Here, someone who really could profit from special services to improve his or her reading does not get them.
5. Difference scores are extremely unreliable. Indeed, they are much more unreliable than simple IQ or reading scores.(22) What does this mean?
Imagine a group of individuals taking a test twice. For example, they might take an IQ test or another ability test twice. To what extent are their rank orders the second time the same as the first time? In other words, do the people who do relatively well (or poorly) the first time also do relatively well (or poorly) the second time? Reliability measures the extent to which people do about the same, relative to others, when they are retested.
The more closely related the constructs being measured are, the more unreliable the difference scores are. Reading ability and intelligence, or math ability and intelligence, are obviously related. The relation is far from perfect, but it is not zero either. As a result, difference scores involving IQ and either reading or math scores can be expected to be quite unreliable. Schools are thus making high-stakes decisions on the basis of difference scores that have been proven to be unreliable bases for making any kind of decision whatsoever.
Many psychologists and other diagnosticians, of course, know these facts and feel compelled to use a system that they recognize to be problematical. Those who do not recognize the problems may end up contributing to them. But those who do recognize the problems are caught in an awful bind. They understandably may be reluctant to report a set of scores that they know will deprive a child of special services that the child needs. Such a decision is extremely unpleasant, especially if the school psychologist or other diagnostician is devoted to his or her work. Or the diagnostician can administer further tests to the child (such as an additional reading test or IQ test) to see if some other combination of test scores will produce the desired diagnosis.
6. Identification of LDs is plagued by statistical regression. What is statistical regression?
When prediction is less than perfect, predicted outcomes are likely to "regress" (move) toward the mean, or the average. Statistical regression is not a psychological effect; it is a statistical effect that arises as an outcome of imperfections of prediction.
In other words, when current data lead one to predict that a restaurant or an athlete will continue to perform at an above-average level, the performance of the restaurant or athlete will usually get worse over time, all other things equal. Children of really smart parents tend not to reach their parents' averaged level of smarts. At the same time, when current data lead one to predict that a restaurant or an athlete will continue to perform at a below-average level, the performance of the restaurant or athlete will usually get better over time, all other things equal. Children of parents of very modest abilities tend to excel beyond their parents' averaged level of intelligence. The greater the departure of the observations from the mean of the measurements being made, the greater the amount of statistical regression, that is, the more the predicted observations tend to regress toward the mean. Thus, really good restaurants show more statistical regression than do just pretty good ones, and the same holds for really bad restaurants versus just so-so ones.
The reasons for statistical regression are complex and beyond the scope of this review, but the implications of statistical regression are very relevant. When there is statistical regression, scores of individuals tend, on retesting, to be closer, on average, to the mean. Suppose, for example, that someone does really poorly on a reading test. Chances are better than fifty-fifty that if the individual is retested, his or her score will be closer to the mean than the first score was. Similarly, if someone does really well, chances are that the next score will be closer to the mean.
Because diagnoses of reading disabilities are based on very low reading scores, these diagnoses are susceptible to error because of problems of statistical regression. Statistical regression can lead to overdiagnosis because it takes people who just happened to score very poorly on a reading test one day (relative to an IQ score) and labels them. If their reading score is much lower than their IQ score, and their IQ score is closer to the mean than their reading score, chances are their reading score will regress toward the mean if the reading test is taken again. Of course, such misdiagnoses can be reduced by repeated testing rather than just administering a single test. There are also statistical corrections that can be applied to help correct for regression effects.
7. Identification processes, particularly with regard to how test scores are used, differ across states and even across local districts within states, rendering highly subjective who is labeled as having an LD. Different states (or even different local school districts) may have different degrees of discrepancy in what they accept as indicative of a learning disability. Or they may use different criteria of identification altogether. Thus, the easiest way to gain an LD diagnosis is often to move one's residence, or otherwise to place one's children in a different school district. Even if one district does not identify the child as having an LD, another might. Consider some examples from one state, Connecticut.
The state of Connecticut recently published prevalence rates by district for students with disabilities in grades K-12.(23) It is interesting to compare some of the prevalence rates. Hartford is a district serving a relatively poor population, with a diagnosed prevalence rate of 17.4%, which is quite high by statewide standards. New Haven is roughly comparable socioeconomically, but the prevalence rate is only 12.9%, or roughly three-quarters of that in Hartford. Almost certainly, the identification procedures differ. But then, Greenwich, one of the wealthiest towns in the state, has a prevalence rate of 16.2%, near the top. Another comparable district, New Canaan, has a prevalence of only 9.5%, and Hebron, another well-off district, is down at 7.2%. At the top is Canaan, largely white and working class, at a whopping 23.8%. Clearly, standards for identification must differ across districts. A parent who is dissatisfied with the results for his or her child in one district may gain a desired diagnosis for his or her child by moving to another district.
Another option besides moving is waiting. In each school year since 1991-1992, the prevalence rates in special education have increased. A child who is not identified as having an LD one year may be identified in a future year, not because the child has changed, but because the standards of labeling have changed.
In sum, the use of difference scores in diagnosing learning disabilities is analogous to the building of a house of cards. Parents and educators make millions of high-stakes decisions on the basis of a procedure that is flawed, and moreover, one that any competent professional in the field ought to realize is in need of modification.
Difference scores are not the only basis on which LD diagnoses can be made, but they are probably today the most common basis, despite their glaring inadequacies. It is past time for a change.
IV. THE MATTHEW EFFECT
We believe, on the basis of the various arguments made above, that IQ has no place at all in the diagnosis of learning disabilities. For all of the reasons discussed in Part III, diagnoses involving IQ are badly flawed. Yet such decisions are being made every day, and children are either receiving or not receiving special services on the basis of them. Here, as elsewhere, we argue that decisions that should be based on sound psychological and educational criteria are being made on the basis of social or even political criteria. Legal decisions of the kind discussed by Kelman and Lester are being made every day on the basis of bad psychological and education practice, such as the use of difference scores described above. A better practice is to recommend special interventions on the basis of poor performance in a specified set of skills and then to target the intervention to the remediation and development of these skills.
Although our greatest concern is with the use of IQ tests, it is worth pointing out that all tests are nothing more than indicators of performance at a given time in a given place. Administration of any test can lead to erroneous judgments, so test scores, whatever they may be, need to be interpreted carefully. If they are not, they may lead to what Keith Stanovich has called Matthew effects, based on a notion introduced by sociologist Robert Merton and adapted from the Bible, whereby the rich get richer and the poor get poorer.(24) In this context, good learners get better and poorer learners get worse.
Matthew effects are prevalent in reading instruction because later skills build so heavily on earlier skills.(25) If students are not given a solid foundation in basic skills, they are scarcely in a position to acquire more advanced skills. They are like the building that has a weak foundation and that starts to crumble even as higher floors are placed on top of lower ones. For example, without sound word-recognition skills, higher order comprehension is impossible. It is in part for this reason that many children who are poor readers show poor phonological-decoding skills in the early grades and then poor higher order comprehension in the later grades. They cannot understand meaning when they cannot decode the words on which the meaning is based.
Matthew effects are pernicious, and they occur throughout our entire system of education. Once they start, there may be no end to them. Consider, for example, what happens to the poor reader. Because of the importance of reading both in the world of school and in the world of work, schools should do everything they can to enrich students' reading experiences to the greatest extent possible. In many schools, however, the poor reader is likely to get a watered-down program for improving his or her reading skills. The school, in attempting to give the child "appropriate" reading material for the child's reading skill level, actually is giving the child material that may lead him or her to fall behind. For example, giving fifth-grade readers who read on the third-grade level a third-grade reading textbook is a counterproductive strategy. Why? Because the content of the stories is no longer age-appropriate and is likely to be boring or even demeaning to the fifth-grader. The student is likely to be unmotivated to read such material. Moreover, the fifth-grade student is more sophisticated than the third-grade student in terms of life experience and learning strategies and needs a pedagogical approach appropriate to his or her age and mental levels.
Soon, the child getting a watered-down reading program does start to fall further and further behind in reading. Teachers label the child "slow" or "incompetent," and then start to expect incompetent work, which is exactly what they get.
As the years go by, the school system places the child in lower tracks or receives reduced opportunities in standard tracks. Reading performance continues to slide. Eventually the student may wish to go to college and be required take the Scholastic Assessment Test (SAT) or the American College Test (ACT). Performance on these tests, of course, depends heavily on reading. Even the mathematical sections require students to read the words in the problems. The poor reader is likely to do poorly and have a harder time getting into a competitive college, or perhaps any college at all.
Eventually, this student may find him or herself employable only in menial jobs or in jobs that require little or no specialized academic training. The reading deficiency now may affect the whole of the individual's life, not just the individual's performance in school.
Some parents, recognizing how the system works, compete and even claw for the educational opportunities they believe their children deserve, such as extra learning time with specialists or special arrangements for homework assignments or tests. If their child is under-performing, they may view special services as the answer. The only way to get such special services, often, is for their children to be labeled. As Kelman and Lester recognize, the parents may see the LD label as the only road available to stop their children from descending the educational ladder and, ultimately, the social-class ladder.
The worst of it is that the children may only receive the special services if they have the LD label. This method for allocating resources creates a perverse situation where parents and often their children actually may want to be labeled as LD -- as having a disability of some kind -- in order to qualify for the special services.(26) Many individuals to whom we have communicated this fact have been unwilling to believe that parents would actually want their children to be labeled as having something wrong with them just so that they can qualify for special services. But from the parents' own standpoint, they are acting in a rational way: they have found what often they believe is the only route to special services for their children.
We know that poor achievement tends to be associated with lower socioeconomic status (SES), so we might well expect that diagnosed incidents of learning disabilities would be high in low-SES towns, and low in high-SES towns. We might expect such a pattern, but we would be wrong. In the state of Connecticut, as mentioned earlier, some of the towns with the highest levels of diagnosed disabilities are among the wealthiest towns in the state.(27) They are the towns that, on statewide mastery tests, typically score at or near the top in terms of levels of academic achievement. How, then, could these towns have high reported incidences of learning disabilities?
The answer, unfortunately, is that the educational system has driven these diagnoses. Schools may be conservative in labeling children as having a disability, but parents do not have to be. Those parents who are wealthy enough can afford to have a private diagnostician test their children and, potentially, assign them a label. Professionals, who are being paid to label, may be reluctant not to assign the label. If they, like the school, are conservative, they may find very soon that their client base is drying up. What parents want to spend as much as several thousand dollars only to be told that their child is fine and not in need of special services?
So the pressure may be on the diagnostician to give the child the diagnosis the parent wants. But such a scenario can be played out only in a town where the parents are wealthy enough to afford one or more private diagnoses. And the parents may indeed end up having to shop around for several diagnosticians in search of a diagnostician willing to make the diagnosis the parents seek. Clearly, this system is warped, and works against special services going to those children who arguably need them most, namely, poor children whose only recourse to special services is through public schools. Low achievement in school, rather than any kind of special diagnostic label, is what should trigger special services. Rather than identifying people as "LD" or "non-LD," we should render special services for special deficits. The focus should be on labeling deficient skills, not people.
What makes this situation bizarre is that parents have been driven to seek a diagnosis that something is wrong with their child. In the past, such a diagnosis is probably the last thing most parents would have wanted for their child.
Schools, of course, can question the private diagnoses they receive. But in an age where almost anyone seems willing to sue for almost anything, schools are chary of litigation. They do not want to get involved in multiple expensive lawsuits where the likelihood of their losing is fairly substantial. In today's social and legal climate, the system undoubtedly favors those willing to pay for the diagnoses. Schools have a tremendous incentive to avoid litigation by giving in to parental demands.
At the same time, we need to recognize that parents are often driven to have their children diagnosed because they feel desperate at the lack of resources being devoted to their children. They may see no other course of action in a system that will pay attention only if their children are diagnosed. Many outstanding diagnosticians do their jobs responsibly and well. To the extent a problem exists, it is not in the parents, diagnosticians, schools, or even the children: it is in a system that is broken and needs to be fixed.
In sum, we find ourselves basically in agreement with the main thrust of Jumping the Queue, although we come to similar conclusions from different origins. The concept of learning disabilities, as it is operationalized, does not work. We should not, however, throw out the baby with the bath water. We need an operationalization based simply on poor performance in a given area of endeavor that a society values without regard to so-called general intelligence. If the problem is one of skills, the skills need to be remediated. If the problem is One of motivation then the motivation needs to be remediated. The intervention should fit the particular problem. As a society, we ought to help those whose performance needs remediation. We should concentrate on performances, not on performers. Labeling is useful to the extent it recognizes school or other work that needs to be improved. At the same time, we need to recognize that, in a sense, nothing is special about a learning disability. What is special is the set of skills to which the society chooses to assign this label. The system we have in practice finds no basis in psychological theory. To a large extent, it is a historical accident.
(1.) William Nelson Cromwell Professor of Law, Stanford Law School.
(2.) Acting Professor of Law, UCLA School of Law.
(3.) For example, the authors state that "The basic claim that we will elaborate over the course of the book is straightforward: as a society, we must make decisions about which students deserve resources beyond those devoted to their classmates." P. 6. The authors then go on to state the many other decisions society must make. Many readers may want more guidance not only as to what decisions need to be made, but as to what these decisions should be and why. As another example, the authors state that they are quite critical of existing legislation "but offer no real insight into the question of whether a system that would in fact be likely to replace it if it eroded would be preferable to the current system." P. 9. But this is the insight many readers would hope the authors of the book would offer: What should society legislate and would the new legislation improve on what we have?
(4.) For example, the authors present a lengthy discussion of alternative interpretations that might be given to the major Supreme Court case, Board of Education v. Rowley, 458 U.S. 176 (1982). Pp. 51-55. Many readers, though, would welcome the opportunity to hear what the authors believe to be an optimal interpretation, and why they believe it. Readers might also wonder about the authors' opinion of how the law should be implemented in schools.
(5.) See AMERICAN PSYCHIATRIC ASSOCIATION, DIAGNOSTIC AND STATISTICAL MANUAL OF MENTAL DISORDERS 46-53 (4th ed. 1994).
(6.) See generally R.J. STERNBERG & E.L. GRIGORENKO, OUR LABELED CHILDREN: WHAT EVERY PARENT AND TEACHER NEEDS TO KNOW ABOUT LEARNING DISABILITIES (1999).
(7.) See ROBERT J. STERNBERG, METAPHORS OF MIND: CONCEPTIONS OF THE NATURE OF INTELLIGENCE 33-49 (1990).
(8.) See CHARLES SPEARMAN, THE ABILITIES OF MAN 75 (1927) (discussing the "specific factor," which varies between different abilities).
(9.) See, e.g., HOWARD GARDNER, FRAMES OF MIND: THE THEORY OF MULTIPLE INTELLIGENCES passim (1983).
(10.) For expositions of these alternative points of view, see PERSPECTIVES ON LEARNING DISABILITIES: BIOLOGICAL, COGNITIVE, CONTEXTUAL 193-249 (Robert J. Sternberg & Louise Spear-Swerling eds., 1999).
(11.) 20 U.S.C. [sections] 1400 (1994).
(12.) See 20 U.S.C. [sections] 1401(a)(15).
(13.) See THE RISING CURVE: LONG-TERM GAINS IN IQ AND RELATED MEASURES passim (Ulric Neisser ed., 1998); James R. Flynn, Massive IQ Gains in 14 Nations: What IQ Tests Really Measure, 101 PSYCHOL. BULL. 171, 171-91 (1987).
(14.) At least two interventions that have been successful, in our opinion, are Success for All, devised by N.A. Madden, R.E. Slavin, and their associates, and the Bowman Gray Learning Disabilities Project, devised by Rebecca Felton, Frank Wood, and their associates. The first program is not targeted specifically at LD children; the second is. In the reading domain, successful programs place a substantial emphasis on increasing phonemic awareness and decoding skills. See LOUISE SPEAR-SWERLING & ROBERT J. STERNBERG, OFF TRACK: WHEN POOR READERS BECOME "LEARNING DISABLED" 273-301, (1996); STERNBERG & GRIGORENKO, supra note 5, at ch. 8.
(15.) See GARDNER, supra note 9, at 59-60; ROBERT J. STERNBERG, BEYOND IQ: A TRIARCHIC THEORY OF HUMAN INTELLIGENCE 299-314 (1985).
(16.) See ROBERT J. STERNBERG, SUCCESSFUL INTELLIGENCE passim (1997).
(17.) See Robert J. Sternberg et al., The Relationship Between Academic and Practical Intelligence: A Case Study in Kenya Intelligence, in press [hereinafter Sternberg, Case Study]; Robert J. Sternberg et al., Testing Common Sense, 50 AM. PSYCHOLOGIST 912, 913-16 (1995) [hereinafter Sternberg, Testing Common Sense].
(18.) See Sternberg, Testing Common Sense, supra note 17, at 923-24.
(19.) See Sternberg, Case Study, supra note 17.
(20.) See Lynn Okagaki & Robert J. Sternberg, Parental Beliefs and Children's School Performance, 64 CHILD DEV. 36, 51 (1993).
(21.) See Lewis M. Terman, Mental Growth and the I.Q., 12 J. EDUC. PSYCHOL 325-41,401-07 (1921).
(22.) See Lee J. Cronbach & Lita Furby, How We Should Measure "Change" -- or Should We?, 74 PSYCHOL. BULL. 68, 77-80 (1970).
(23.) See STATE DEPARTMENT OF EDUCATION, DRAFT REPORT ON SPECIAL EDUCATION IN CONNECTICUT (1997).
(24.) See Keith E. Stanovich, Matthew Effects in Reading: Some Consequences of Individual Differences in the Acquisition of Literacy, 21 READING RES. Q. 360, 381 (1986) (citing Robert K. Merton, The Matthew Effect in Science, 159 SCIENCE 56, 58 (1968)); see also Matthew 13:12 ("For unto every one that hath shall be given, and he shall have abundance: but from him that hath not shall be taken away even that which he hath.").
(25.) See Stanovich, supra note 24, at 381.
(26.) For a journalistic description of this phenomenon, see Ruth Shalit, Defining Disability Down, THE NEW REPUBLIC, Aug. 25, 1997, at 16.
(27.) See supra note 23 and accompanying text.
Robert J. Sternberg, IBM Professor of Psychology and Education in the Department of psychology, Yale University. Ph.D. 1975, Stanford University. -- Ed. Preparation of this Review was supported by Grants R206R950001 and R206A70001 from the U.S. Office of Educational Research and Improvement under the Javits Act. Financial support for this research does not necessarily indicate agreement with the positions taken in this review either by the U.S. Department of Education or the U.S. Government. Parts of this review draw on R.J. STERNBERG & E.L. GRIGORENKO, OUR LABELED CHILDREN' WHAT EVERY PARENT AND TEACHER NEEDS TO KNOW ABOUT LEARNING DISABILITIES (1999).
Elena L. Grigorenko, Research Scientist in the Department of Psychology and Child Study Center, Yale University, and Associate Professor of Psychology, Moscow State University. Ph.D. 1996, Yale University. -- Ed.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||1999 Survey of Books Related to the Law|
|Author:||Sternberg, Robert J.; Grigorenko, Elena L.|
|Publication:||Michigan Law Review|
|Article Type:||Book Review|
|Date:||May 1, 1999|
|Next Article:||The convergence of the First Amendment and Vatican II on religious freedom.|