Proposed language in the Individuals with Disabilities Education Act (IDEA)--the federal law that governs special education--could trigger a shift in the way schools identify children with learning disabilities.

The House and Senate versions of the bill that will reauthorize IDEA explicitly say schools wouldn't be required to use the traditional method of determining learning disability--the IQ-achievement discrepancy model, which identifies children as learning disabled (LD) when their scores on achievement tests are significantly lower than their IQ. The new versions of IDEA also include language that opens the way for an alternative identification method.

But the potential changes aren't without debate. While many psychologists and educators agree that IQ-achievement discrepancy has flaws, they are split on what to do next. Some argue for particular models as replacements; others caution that rushing to replace the discrepancy model could result in substituting one flawed system for another.


Last reauthorized in 1997, IDEA defines a specific learning disability as a disorder in basic psychological processes and mandates that schools use more than one method to determine whether students have learning disabilities--thereby qualifying for special education services. But many schools rely heavily on discrepancy formulas in making their decisions, says Jean Lokerson, PhD, chair of the Public Policy Committee at the Learning Disabilities Association of America, an advocacy group for parents.

Federal regulations suggested that schools use the discrepancy model in the 1970s as a way to cap the number of students identified as learning disabled. However, no research actually showed how discrepancies would affect the number of students or whether they were an accurate identifier, says Lokerson.

In fact, the number of students identified as LD has swelled since its implementation--growing about 150 percent since 1975, according to the U.S. Department of Education. That figure is raising concern among some policy-makers.

"What that means is that not only are there more kids being identified as learning disabled and more kids being served in special education, but that a greater percentage of local and state dollars are being redirected from general education to special education," explains psychologist Douglas Fuchs, PhD, co-director of the National Research Center on Learning Disabilities and a Vanderbilt University special education professor.

Many conjecture that at least part of the growth of LD students stems from misidentification. Pressure from parents to avoid more stigmatizing labels such as mental retardation, the lack of other services for academically struggling students and schools' failure to comply with state and federal guidelines all may contribute.

The result is that children who are struggling for reasons other than a learning disability--such as poor instruction or speaking English as a second language--may be shunted into special education when they would be better served by other interventions, says psychologist Alberto Bursztyn, PhD, an associate professor at Brooklyn College of the City University of New York.

Other criticisms include:

  • Variance. Some states require one standard deviation in difference between IQ and achievement, while others require more than one. Critics say that means a child may be identified as learning disabled at one school, but not at another just across the state line.

  • Methodology. Psychologists such as APA President Robert J. Sternberg, PhD, criticize discrepancy formulas because they say difference scores are error-prone and have problematic statistical properties. Others argue that, instead of the discrepancy model practice of taking two data points from single moments in time, schools should continually assess children.

  • A "wait-to-fail" approach. Children often must do poorly for years before their achievement is sufficiently low compared with their IQ. In fact, children are identified as learning disabled at a modal age of 10 to 12 years old, says Fuchs.

What are the alternatives?

In an attempt to address the failings of IQ-achievement discrepancy, both the House and Senate versions of IDEA include text that opens the way for an alternative model of identifying learning disabilities often called "response to intervention," or RTI. In the model--used in a handful of schools nationwide, most notably in Minneapolis--students progress through several levels of intervention before entering special education. The first level is the general classroom, where teachers and other school personnel regularly assess students to identify those who are struggling. Those children are diverted into a secondary intervention, which is often intensive, small-group instruction that's tailored to students' needs, while school staff continue to monitor their progress. Some RTI models include a tertiary intervention that's even more intensive. Once students have failed to respond to each of the intervention levels, they are placed in special education.

"The assumption is that all of these tiers will reduce the number of children requiring special education, and that the children who end up in special education are truly in need of special education," says Fuchs. Moreover, he adds that the model's emphasis on continual assessment makes it attractive to some who criticize the one-point-in-time testing associated with the IQ-achievement discrepancy model.

Some argue, however, that RTI and other new methods are not needed; rather, laws and regulations should mandate strict adherence to existing guidelines so that fewer students enter special education. In turn, policy-makers should also shore up general education to serve struggling students, says Thomas Scruggs, PhD, a George Mason University special education professor.

Although psychologists such as Sternberg argue that IQ tests have no place in identifying learning disabilities, others argue that there's a difference between traditional IQ tests and cognitive assessments that measure the basic psychological processes mentioned in IDEA. Eliminating such tests altogether amounts to throwing the baby out with the bath water, says psychologist Jack Naglieri, PhD, of George Mason University, who's developed an alternative cognitive assessment system to evaluate such processes.

In fact, APA's Committee on Psychological Tests and Assessment argued just that in response to a report last summer from the President's Commission on Excellence in Special Education.

In a statement approved by APA's Board of Directors, the committee advised legislators crafting the IDEA legislation to require cognitive assessments instead of following the report's recommendation to eliminate such testing altogether.

However, the RTI model referenced in the new legislation does not rely on cognitive assessments. Some argue it's yet another wait-to-fail approach, since students could take months or years to funnel through the system. It's also not clear how often students would bounce between RTI's various levels before entering special education, since little research has been conducted on the model to date.

Moreover, although the proposed IDEA legislation continues to define a specific learning disability as a disorder in basic psychological processes, the law as it stands would not require schools to evaluate those processes, says Naglieri.

Indeed, what concerns Lokerson is that RTI at no point requires an evaluation of those processes. "The child's lack of response to intervention in and of itself becomes the criteria for saying you're learning disabled," says Lokerson. "There is absolutely no valid research that suggests that that could be the case."

Instead, cognitive evaluations and diagnostic testing are essential to identify learning disabilities, she says, since they provide the information on specific problems that psychologists can use to help teachers tailor instruction.

"Students shouldn't be referred for learning disability simply because they don't fit the classroom mold," agrees Scruggs. "Many low-achievers could probably benefit from modifications to instruction in the general education class."

Sternberg, in fact, argues that's just what schools should be focusing on: "There's too much effort labeling students and not enough effort figuring out what's wrong and what to do about it," he says.

No matter what model is eventually favored by IDEA legislation and regulation, many agree that it must address the failings of IQ-achievement discrepancy--with consistent application across states and schools, accurate methods that reduce overidentification and earlier detection. The only way to achieve that, says Lokerson, is by first conducting rigorous research on pilot programs to determine whether a model actually works in practice and can be replicated nationwide.

However, some psychologists argue that reformers are looking in the wrong place. "The regulations suggest that if something is wrong in terms of a child's learning progress, the first and only place to look at is in the child's head," Bursztyn says. "We need to look at quality of the instruction and the capacity of schools to address individual differences without resorting to special education."

Further Reading

  • Scruggs, T.E., & Mastropieri, M.A. (2002). On babies and bathwater: Addressing the problems of identification of learning disabilities. Learning Disability Quarterly, 25(3), 155-168.

  • Newman, T.M., & Sternberg, R.J. (Eds.). (in press). Students with both Gifts and Learning Disabilities. New Haven: Yale University.

  • Information on IDEA and other legislation is available from APA's Public Policy Office 

  • The report of the President's Commission on Excellence in Special Education is available at www.ed.gov/inits/commissionsboards/whspecialeducation.