disabilities researchers began to question seriously the use of discrepancy. These critics have cited at least four reasons for their objections. First, they argue that studies that were instrumental in justifying a discrepancy approach in the first place were flawed. Researchers conducted epidemiological studies on the Isle of Wight in which they used regression scores between reading and performance IQ scores to differentiate students who had specific reading retardation (discrepant readers) from those who had general reading backwardness (nondiscrepant readers) (Rutter & Yule, 1975). Finding a “hump” in the lower end of the distribution of residual reading scores for those with specific reading retardation, some researchers used these data as evidence of the validity of using discrepancy to define students with learning disabilities. Several researchers, however, have leveled criticisms at the Isle of Wight studies, e.g., inability to replicate the results and ceiling effects on the reading test, which could have led to an inflated number of discrepant readers and resulted in the “hump.” (See Vellutino, Scanlon, & Lyon, 2000, for a more in-depth discussion of these criticisms.)
Second, some have cited the Matthew effect (better readers learn more about their world and, therefore, are likely to score higher on IQ tests) as a problem. They have pointed out that the IQ scores of students with reading disabilities may be underestimated (Siegel, 1989).
Third, using a discrepancy approach makes it very difficult to identify children as learning disabled in the early elementary grades. This is particularly problematic because research has generally shown that intervention is more effective the earlier it is implemented (Fletcher et al., 1998)
Fourth, researchers have been unable to discriminate between students with a discrepancy from those with low reading achievement but no discrepancy on measures considered important for reading, e.g., phonological awareness, orthographic coding, short-term memory, word retrieval (Fletcher et al., 1994; Foorman, Francis, Fletcher, & Lynn, 1996; Stanovich & Siegel, 1994). Although low achievers do not differ from those with a discrepancy on these variables, this does not mean that low achievers do not differ from students identified as learning disabled, using broader identification criteria (Fuchs, Mathes, Fuchs, & Lipsey, 1999).
Researchers have just begun to explore alternatives to the discrepancy approach to identification. One alternative would rely on the assessment of phonological processes (Torgesen & Wagner, 1998). Another, referred to as the treatment validity approach, would involve assessment of students’ levels of academic performance and learning rates on curriculum-based measures (Fuchs & Fuchs, 1998).
Disproportionate representation of minority students. Since at least the time of Lloyd Dunn’s classic article, “Special Education for the Mildly Retarded: Is Much of It Justifiable?” (Dunn, 1968), there has been concern over identification of children from minority backgrounds in special education. Although most of the concern has been focused on the categories of mental retardation and emotional disturbance, there is also some evidence of disproportionate representation in learning disabilities. For 1998–1999, 4.49% of all students (aged 6 to 21 years) were identified as learning disabled. Following are the percentages for different ethnic groups: White (4.27%), African American (5.57%), Hispanic (4.97%), Asian/Pacific Islander (1.70%), American Indian/Alaska Native (6.29%, U.S. Department of Education, 2000). These figures indicate substantial over-representation of African Americans and, especially, American Indian/Alaska Natives in the learning disabilities category and a very large under-representation of Asian/Pacific-Islanders.
Researchers have not yet been able to disentangle the reasons why disproportionate representation in learning disabilities and other areas of special education exists. Factors that researchers have cited as potential causes are racially biased tests, racially biased professionals, and inadequate community resources, such as health care and educational opportunities (Hallahan & Kauffman, 2000). Most authorities do agree that disproportionate representation is a complex problem, and the federal government has begun to highlight it as a major problem:
The complexity of this issue requires an integrated and multifaceted effort to promote greater educational access and excellence for racial/ethnic minority students that involves policy makers,