High-stakes tests, which affect a student’s ability to earn a high school diploma, are now in place or on the drawing boards in about half the states. Often controversial, they have spawned “test boycotts” and lawsuits. Other states have testing programs to comply with their own programs and with the No Child Left Behind (NCLB) Act as well. Many state leaders publicly endorse these tests. But a strange thing is happening: As we get closer to having the graduation tests “count,” many leaders have blinked, with the result that standards are compromised and test results invalidated.
The press chases down every test boycott or lawsuit challenging a test, but the public hears little about decisions and compromises made behind the scenes. On the legal front, if we don’t do tests right, the entire enterprise could end up mired in the courts.
Where is the outrage over the need for valid tests? Inconsistency in test administration has real consequences. As is becoming increasingly obvious, confusion and inconsistency are leading to a loss of credibility in the standards movement. But why is it happening—why are some blinking?
Why the Blinking?
Is it fear of litigation? It is confusion about legal requirements? Is it the excruciating pressure on that one-diploma option?
Word choice is telling. It used to be that a student “earned” a diploma. Now many speak of a student being “denied” a diploma. The first is about standards; the second, about rights and lawsuits. Our evolving language—unfolding daily in the press—tells the tale.
Before summarizing recent events and court actions, let’s review the relevant laws: the Fourteenth Amendment to the U.S. Constitution; the Individuals with Disabilities Education Act (IDEA), the federal special education law; and Section 504 of the Rehabilitation Act of 1973, which prohibits unlawful discrimination against persons with disabilities by public schools and states. The following are key provisions of these laws relating to testing students with disabilities:
• The Fourteenth Amendment entitles all students to an opportunity to learn the material on graduation tests and to advance notice about the testing program.
• Tests need to be valid, reliable, and technically sound; and they must be administered according to the test producer’s instructions. They should measure what they are designed to test. Thus, a reading test should measure reading. Validity standards are based on test content, not on which groups of students take the test. It’s about the what of the test—not the who of test takers.
• Students with disabilities have legal protections that may entitle them to testing “adaptations,” a term used here to describe changes in how a test is administered. Such adaptations are provided when necessary for students to access the test and demonstrate what they know and can do. The student’s Individualized Education Programs (IEPs) under the IDEA or Section 504 plans, determine when and how adaptations are provided.
• In test development, there are two very different types of “adaptations.” (1) “Accommodations” are changes that do not fundamentally alter what is being tested. (2) “Modifications” are changes that do fundamentally alter what is being tested. Stated another way, accommodations maintain test validity while modifications do not. Unfortunately, the use of these terms and definitions is often confused and inconsistent.
• When scores are used for accountability, including diploma, purposes, states and schools are not required to provide test “modifications.” Thus, there is no legal requirement to read a reading test to a student who cannot read. Federal laws require test score reports to indicate if they were administered in nonstandard ways.
• It is not unlawful to test students in the area of their disability. Although educators (and test makers) are exhorted to ensure that tests measure a student’s skills and knowledge—not the disability—where that is not possible because the test measures the very areas of disability, the laws do not bar such testing.
• Schools and states should provide equal opportunity to children with disabilities, but they are not expected to provide equal outcomes for all.
Courts and federal agencies have been consistent: Setting standards is not discriminatory. Not all students are guaranteed a diploma or certificate. Tests are to be validated for their specific purpose—not for different groups. The troubling events discussed below contradict these strictures.
First, consider the 2001 settlement of a lawsuit against the state of Oregon by Advocates for Special Kids, represented by Disability Rights Advocates (DRA). The suit alleged that Oregon’s test for the Certificate of Initial Mastery (CIM) discriminated against students with learning disabilities when it tested them on basic skills such as reading, writing, and math. Note that the test was not a diploma test. When I first heard of this allegation several years ago, I dismissed it. The basics are discriminatory? They can’t be serious! Common sense and the law say otherwise. But life is full of surprises: Oregon settled the case. In my judgment, the settlement is neither a model nor a solution, with its intricate (often confusing) conditions, bureaucracy, and avenues for appeal (beyond those created by federal law). The settlement attempts to bypass tests that assess essential skills, expands the list of accommodations that are allowed, and requires alternate assessments and changes in scoring results—all with an eye to ensuring that students with disabilities receive the CIM. This process appears to confuse the concept of equal opportunity with the concept of equal outcome. The first is legally required; the second is not.
The second set of events involves accommodation and appeals policies developed by legislators and departments of education for their state tests that are used for accountability or graduation purposes.
Recently, I asked a friend whose learning-disabled child had a very hard time learning to read, “What do you think of state policies that allow a child to take a reading test by having the test read to him?” Her quick response was, “A cop-out. My son would never have learned to read if that was the law then.”
Indeed. Yet some states allow students to use any accommodation on the state test that they use in classes. Thus, a reader (someone who reads material to a student) can be provided on the reading test, a calculator for the math calculation test, and so on. How can the state then ensure test validity and how do students (or their parents) know if they learned to read or add at passing or proficient levels? In stark contrast, other states bar the use of any invalidating modifications. In so doing, they uphold the validity of their testing programs.
The third event is the widely reported 2002 settlement of a threatened lawsuit brought by DRA against the College Board’s SAT on behalf of students with disabilities. Under that settlement, the College Board no longer flags tests that are taken with extended time. Such flagging used to inform the reader when the SAT was taken under nonstandard conditions that affected score validity. Henceforth, when colleges and universities receive score reports, they will not be informed that some scores are non-standard. Notably, two weeks later, the ACT followed suit. If timing on these exams does not really matter, then why time anyone? The settlement defies the common understanding of test validity, which requires that tests be administered under standardized, normed conditions. Such validity is the minimum criterion needed to maintain public faith in any testing program.
For good news, we must turn to the courts. In sharp contrast to the three events discussed above, two recent courts—following a long line of cases—upheld state efforts to raise standards.
In 2000, in GI Forum v. TEA, a Texas federal district court upheld the Texas Assessment of Academic Skills (TAAS) against a class action lawsuit brought by the Mexican American Legal Defense and Educational Fund on behalf of minority students. The court found no unlawful discrimination, even though a larger percentage of minority students than non-minority students failed the TAAS. In fact, the court found that the TAAS actually helped erase past educational disparities, as the gap between minority and other students was narrowing. The state’s testing program, rather than being aimed at denying diplomas to certain students, was designed “to help identify and eradicate educational disparities.” The state provided all students with the opportunity to learn the material. Students had eight chances to take the test and were provided with ample remediation opportunities. The court also found that the test was not the sole criterion for receipt of a diploma, as there also were attendance and coursework requirements.
Rene v. Reed was a class action lawsuit brought by Indiana students with disabilities, seeking an injunction to exempt them from the state’s new Graduation Qualifying Examination (GQE). The Indiana Court of Appeals affirmed the lower court’s decision to uphold the GQE, and, in 2002, after the Indiana Supreme Court denied transfer, the decision became final. Plaintiffs had alleged that the test violated the Fourteenth Amendment and the IDEA. They argued that they had insufficient notice of the test and did not have the “opportunity to learn and master the proficiencies” tested. As to the IDEA, they argued that their IEPs did not cover the material tested and that the state did not allow them to use the “modifications” that they had used in class. The court upheld that state action, attesting to the need for test validity in the state’s accountability program.
Also, the court found that students were provided with adequate notice—more than three years—about the test. They were given multiple opportunities to take it and to remediate their skills. The court noted that many (55.5 percent) of the students with disabilities on the “diploma track” passed the test and another 23.2 percent were expected to graduate with a diploma as a result of the state’s waiver policy. (The policy provides other avenues for students to demonstrate that they have achieved the state’s ninth-grade standards.) The court balanced the students’ harm against the state’s and upheld the state in its education reform efforts.
Both decisions are consistent with a long line of court decisions since the early 1980s. Generally, courts defer to the judgments of educators and policymakers and demonstrate a strong reluctance to bend graduation requirements. Lawsuits against state testing programs are currently pending in Alaska, California, and Massachusetts.
The Road Ahead
The three events discussed above portend compromised standards and confusion, whereas the two court decisions uphold standards. Judges are doing their job—deferring to professionals who implement educational policies appropriately. If we cannot fault the courts for compromising standards, why are states blinking and compromising their own standards?
Yogi Berra is reputed to have said, “If you come to a fork in the road, take it.” We’re at that fork in education reform. Many are going in the front door, loudly supporting high standards—until someone threatens to sue or someone might actually be denied a diploma. At that moment, the back door of retreat and erosion flings open and we quietly sidle out. Sometimes, we even call something “reading” when it’s really “listening,” and we report scores with questionable meaning without so indicating. We’re on a very slippery slope. Although intentions may be good, the response is misguided.
Instead, let us provide students with appropriate education to learn the requisite skills they need in life. And for those students who cannot pass graduation tests—despite our and their best efforts—perhaps it is time to consider providing them with other honorable and meaningful exit documents—in addition to the one-test diploma. Let us report what students learned and can do—not just what they failed. We need to help students get on with their lives—not simply mark them as failures of a state test.
Blinking at standards fails the public when, as the going gets rough, we quietly alter tests and standards. Such compromises have a crippling ripple effect on education reform, leading to cynicism and loss of faith in the entire venture. Such compromises are not legally warranted. For the sake of our children and for America, we must not blink.