Is "Learning Disabilities" Just a Fancy Term for Low Achievement? A Meta-Analysis of Reading Differences Between Low Achievers With and Without the Label

Douglas Fuchs and Lynn S. Fuchs, Peabody College of Vanderbilt University; Patricia G. Mathes, University of Texas--Houston Health Science Center; Mark W. Lipsey and P. Holley Roberts, Peabody College of Vanderbilt University
Learning Disabilities Summit: Building a Foundation for the Future White Papers

This paper is available in alternative formats: | Download Word | Download pdf |

Pages: | 1 | 2 | 3 | 4 | 5 | 6 |

This paper has two parts. The first part has three purposes: (a) to provide a brief history of how a formal definition of learning disabilities (LDs) was determined; (b) to explore how politics and research during the past two decades has influenced how people think about the validity of the LD construct; and (c) to explain why a quantitative synthesis, or meta-analysis, was necessary to determine whether low achievers with and without the LD label were more alike or different. The second part of this chapter describes method, results, and implications of the meta-analysis.

HISTORY, POLITICS, AND THE LD CONSTRUCT

In 1978, the first author of this paper went to Milwaukee to the annual conference of the Association of Children with Learning Disabilities (now Learning Disabilities Association) to present his just-completed dissertation research. Because it was his first professional meeting, he remembers it well. But it is memorable for an additional reason: Less than 3 years before, advocates had convinced Congress to include LD as a category of exceptionality in the Education for All Handicapped Children Act, and the celebratory greetings and congratulatory backslapping among advocates and practitioners, as well as their excitement and optimism, pulsed through the meeting rooms, hallways, and bars of the crowded conference hotel.

Contributing to the conference-goers' upbeat mood was their confidence in the validity of the LD construct. Throughout the 1970s, most practitioners, parents, and academics firmly believed that LD represented a discrete classification of exceptionality marked by two unique features: "unexpected" learning failure and "specific" learning failure (e.g., Kavale, 1987; Kavale & Forness, 1998). The child with "unexpected" learning failure was perceived by parents and teachers as generally competent. The learning difficulty was both surprising and puzzling. "Specific" learning failure suggested neurological dysfunction and processing deficits, which were presumed to cause severe problems in reading, writing, or math (e.g., Kavale & Forness).

There were at least two reasons to view "unexpected" and "specific" learning failure as a conceptual anchor and rallying cry for the field. First, as far back as the 1890s, physicians W. Pringle Morgan and John Hinshelwood separately described "the seemingly paradoxical inability of some children of average and superior intelligence to master academic concepts" (Lyon, 2001), a phenomenon documented more extensively by another physician, Samuel Orton, in the 1920s and 1930s (Hallahan & Mercer, 2001).

Second, in 1975, Rutter and Yule reported findings from an epidemiological study that seemed to buttress the clinical observations of Morgan, Hinshelwood, and Orton. Rutter and Yule measured the IQ and reading performance of all 9- and 14-year-olds on the Isle of Wight. The researchers regressed the children's IQ scores on their reading scores to produce a distribution of IQ-predicted reading performance. Scores above the mean represented overachievement (i.e., exceeding prediction); scores below the mean indicated underachievement (i.e., beneath prediction). Whereas such a distribution should resemble a Gaussian curve, with overachievement occurring as frequently as underachievement, Rutter and Yule reported a "hump" at the lower end of the distribution, which, they said, indicated that "extreme degrees" of reading underachievement occur at a greater rate than should be expected (Rutter & Yule, 1975, p. 185). When Rutter and Yule compared the "underachievers" to the children whose low reading performance was commensurate with their equally low IQ scores (i.e., "low achievers"), they found that the underachievers were different "in terms of sex distribution, neurological disorder, and pattern of neuro-developmental deficit" (p. 194). Further, the underachieving readers had a worse prognosis for reading and spelling and a better prognosis for mathematics. These findings led Rutter and Yule to suggest that the group of underachievers, or children with "specific reading retardation," was distinctly different from the group of low achievers, or "generally backward readers." Findings appeared to confirm "unexpected" and "specific" learning failure as a valid marker of students with LD (Fletcher, 1995).

And yet, clinicians' and researchers' affirmation of the LD construct, and the general buoyancy of the advocates, belied longstanding concerns. For example, as described by Hallahan and Mercer (2001), the question of prevalence had been a point of contention since the early 20th century when Hinshelwood argued that fewer than 1 in 1,000 students might have "word blindness," or reading disabilities, and Orton countered that a more accurate ratio was 1 in 10. (As indicated below, some policymakers today suggest a 1-in-4 prevalence rate for reading disabilities.)

An obvious reason for such disparate estimates is that there has never been agreement on an LD definition. For a century, the field has tried unsuccessfully to invoke the central nervous system to explain the disorder. Hinshelwood, for example, required that a diagnosis of word blindness be associated with obvious pathology. Orton dismissed this criterion, noting the impossibility of distinguishing pathological from nonpathological cases (see Hallahan & Mercer, 2001). In the early 1960s, the federal government and Easter Seals cosponsored several task forces on LD, the first two of which focused on definitional issues. Task Force I, composed mostly of medical professionals, defined LD in terms of minimal brain dysfunction. The education professionals who constituted Task Force II rejected this definition "because special educators in the field of learning disabilities must base educational management and teaching strategies on functional diagnostic information" (Haring & Bateman, cited in Hallahan & Mercer, p. 34). This task force's substitute definition proposed in part that "Children with learning disabilities are those (1) who have educationally significant discrepancies among their sensory-motor, perceptual, cognitive, academic, or related developmental levels which interfere with the performance of educational tasks; (2) who may or may not show demonstrable deviation in central nervous system functioning; and (3) whose disabilities are not secondary to general mental retardation, sensory deprivation, or serious emotional disturbance" (Haring & Bateman, cited in Hallahan & Mercer). With minor modification, this language became part of the U.S. Office of Education definition in 1977.

Because the definition did not include criteria by which practitioners could identify children with LD, the federal government proposed regulations to operationalize it. The government's strategy was to suggest a "severe discrepancy" between intelligence and achievement as the primary criterion, or marker, for identification. In 1977, the government wrote that educators may determine that children have a specific learning disability if they receive appropriate learning experiences for their age and ability and still do not achieve commensurate with their age or ability levels in oral expression, listening comprehension, written expression, basic reading skill, reading comprehension, mathematics education, or mathematics reasoning.

This guideline, however, was viewed as insufficient by many state education agencies. Most adopted the federal government's severe discrepancy idea (Frankenberger & Fronzaglio, 1991; Mercer, King-Sears, & Mercer, 1990), but defined it in their own way. In 1983, the federal government tried again by convening the Work Group on Measurement Issues in the Assessment of LD. Its primary mission was to determine "[w]hat constitutes a severe discrepancy, from a statistical perspective between aptitude and achievement" (Mastropieri, 1987, p. 29). The Work Group found that (a) states had indeed adopted many varieties of measurement formulas for identifying a severe discrepancy, and (b) some of these formulas were not only excessively complex but mathematically incorrect. The Work Group recommended that practitioners regress an aptitude measure on an achievement measure to produce a predicted achievement score, and that "discrepancy" should be defined as the difference between actual and predicted achievement.

However, even this effort was criticized sharply. "The...discrepancy model," wrote Willson (1987), "is basically an atheoretical, psychologically uninformed solution to the problem of LD classification. For LD to move forward...statistical models...need to be replaced by constructs firmly grounded in psychological theories of learning" (p. 28; also see Lyon, 1987).

Pages: | 1 | 2 | 3 | 4 | 5 | 6 |

Return to LD Summit papers table of contents