Beyond Breimhorst: appropriate accommodation of students with learning disabilities on the SAT.

• Author: Leong, Nancy
• Position: Case Note

INTRODUCTION I. SOME STUDENTS WITH LEARNING DISABILITIES QUALIFY FOR PROTECTION UNDER FEDERAL LAW AND DESERVE ACCOMMODATION ON THE SAT A. Definition of Learning Disability B. Learning Disabilities Under Federal Law C. Accommodation on the SAT II. THE FLAGGING DEBATE II. THE FLAGGING DEBATE A. Scores Achieved with and Without Extended Time Are Not Comparable B. Flagging Was a Poor Solution to the Problem of Comparability C. Not Flagging Scores Achieved with Extended Time Is Equally Problematic 1. Harm to test validity 2. Undesirable incentives to seek diagnosis 3. Socioeconomic inequity 4. Backlash against students with legitimate learning disabilities III. TESTING SERVICES SHOULD MODIFY THEIR PRACTICES TO BETTER SERVE LEGAL AND FAIRNESS INTERESTS A. Testing Services Should Tighten Eligibility Requirements for Accommodation B. Testing Services Should Modify the SAT to Eliminate Speededness CONCLUSION INTRODUCTION

In an era when admission to elite colleges and universities has never been more competitive, a puzzling trend has emerged. Across the country, many bright and ambitious students are anxiously competing to display their cognitive imperfections, practically begging psychologists to label them with dyslexia or attention deficit hyperactivity disorder (ADHD), while their parents unhesitatingly hand over thousands of dollars in fees to pay for such diagnoses.(1)

What could account for this seemingly perverse behavior? The answer is simple: students diagnosed with learning disabilities receive extra time to take the SAT.

This phenomenon is troubling in many respects. The SAT is commonly justified as a means of leveling the playing field in the college admissions process, and the idea of a diagnosis as a means to gain precious time raises a host of issues ranging from fairness to score predictivity. Moreover, recent developments have compounded the problem: following the controversial settlement of Breimhorst v. Educational Testing Service, (2) colleges will no longer know when a student receives extra time.

Prior to Breimhorst, students with disabilities could receive extended time accommodation on standardized tests, but their scores were accompanied by an asterisk, or "flag," and the designation "nonstandard administration." (3) The flag indicated that a score had been achieved with extended time and therefore might not be comparable to a standard score.

This situation changed after Mark Breimhorst sued Educational Testing Service (ETS) when his scores were flagged after he received extra time on the Graduate Management Admission Test (GMAT). As part of the settlement, ETS agreed to stop flagging all of its tests. This did not result in immediate changes for the SAT, which is owned by the College Board and merely administered by ETS. However, after convening a panel of testing experts to study the flagging issue, the College Board decided to follow suit. (4) As of October 2003, SAT scores achieved with and without extended time are indistinguishable to admissions committees. This decision has significant consequences for students with learning disabilities, who comprise an overwhelming majority of the students granted extended time and whose numbers have increased by twenty-six percent in the past five years alone. (5)

The Breimhorst result ultimately creates an untenable situation. Although flagging was undeniably stigmatizing to students with disabilities and should not be reinstituted, simply removing the flags without modifying the format of the SAT impairs the validity of the test and creates undesirable incentives for fraud. This Comment argues that the best way for the College Board to circumvent these unappealing alternatives is to eliminate speed as a factor on the SAT.

This debate over accommodation for students with learning disabilities has made salient a larger problem: the SAT is not intended to test speed, yet for many students, the time limit affects their scores. However, the learning disability context provides a useful forum for discussing these issues, while the Breimhorst settlement creates an immediate incentive to address them.

The Comment is divided into three Parts. Part I provides background, demonstrating that some, though not all, students with learning disabilities qualify for protection under the Americans with Disabilities Act (ADA). Given the College Board's concession that the SAT is not intended to measure speed, such students consequently qualify for extended time on the SAT.

Part II addresses the tension between providing accommodation and preserving the validity of the SAT. Because some students without learning disabilities would also benefit from extended time, granting such accommodation inflates the scores of students with learning disabilities. Flagging was an undesirable way of signaling potential score incomparability because the stigmatization resulting from flagging conflicted with the spirit of the law and made disabled students vulnerable to discrimination. However, the Breimhorst solution of simply removing the flags compromises the validity of the test and encourages students to seek inappropriate accommodation.

Part III proposes two alternatives to mitigate the current situation. The more conservative approach attacks the problem of improper accommodation by restricting eligibility for accommodation to those students whose thoroughly documented learning disabilities merit accommodation under the ADA. However, this approach ultimately provides only a partial solution: even if every student who receives accommodation has a legitimate learning disability, the issue of test validity still remains. The best way for testing services to address this problem is to modify the test to reflect its stated purpose of measuring problem-solving ability rather than speed. Recent research suggesting that speed is not a factor for most students on the new SAT indicates progress toward this goal, yet ETS still retains time limits for nondisabled test takers. As long as these time limits prove to be an issue for some students, the SAT will remain an inequitable assessment.

1. SOME STUDENTS WITH LEARNING DISABILITIES QUALIFY FOR PROTECTION UNDER FEDERAL LAW AND DESERVE ACCOMMODATION ON THE SAT

   The exact medical definition of the term "learning disability" prompts considerable debate, and a student with a medically diagnosed learning disability is not automatically a student with a legally recognized disability. This Part first examines learning disabilities as medical phenomena, then applies relevant federal law to determine under what circumstances a student with a learning disability qualifies for federal legal protection. Given that some learning disabilities do qualify for legal protection, students with such disabilities should be granted accommodation on the SAT.

   1. Definition of Learning Disability

      The Individuals with Disabilities Education Act (IDEA) (6) defines a learning disability as "a disorder in 1 or more of the basic psychological processes involved in understanding or in using language, spoken or written, which disorder may manifest itself in the imperfect ability to listen, think, speak, read, write, spell, or do mathematical calculations." (7) The "heterogeneity" of this remarkably broad definition, which in effect includes seven different cognitive disorders, "renders diagnostic precision impossible.' (8) However, more specific standards for actually diagnosing learning disabilities have proved elusive. The President's Commission on Excellence in Special Education recently found that "[m]any of the current methods of identifying children with disabilities lack validity. As a result, thousands of children are misidentified every year, while many others are not identified early enough or at all." (9) Compounding the problem of varying diagnosis methods, cognitive shortcomings such as those mentioned in the IDEA are largely on a continuum, and it is often difficult to distinguish between normal impairment and impairment that constitutes a learning disability.

      Until very recently, the most commonly used indicator was that of a significant disparity between ability and achievement, defined as a discrepancy of at least 1.5 standard deviations. However, after recent amendments, the Individuals with Disabilities Education Act now states that, in diagnosing a learning disability, "a local educational agency shall not be required to take into consideration whether a child has a severe discrepancy between achievement and intellectual ability...." (10) This modification suggests that local agencies are still permitted to use the discrepancy model, but cannot be forced to do so, and moreover allows them to rely exclusively on "scientific, research-based intervention." (11) Essentially, the IDEA now gives individual schools even more discretion in diagnosing learning disabilities, which will only lead to greater inconsistency in the standards used.

      This Comment does not advocate a particular medical definition of learning disability; rather, it only aims to demonstrate the considerable variability in methods of diagnosis. The more important issue from a legal standpoint is whether, and to what extent, a medically diagnosed learning disability can also qualify as a disability deserving protection under current federal statutory law.

   2. Learning Disabilities Under Federal Law

      Title III of the ADA defines "disability" as "a physical or mental impairment that substantially limits one or more major life activities." (12) For plaintiffs with learning disabilities, the obstacle to ADA protection generally has been the showing of substantial limitation. Articulating the major life activity in question as "learning" has proved prohibitive in many cases because the substantial limitation must restrict an individual's major life activity as to the "conditions, manner, or duration under which [the activity] can be performed in comparison to most people." (13) Even if a...