No child left behind

THERE IS NO QUESTION that American students need to be among the most skilled and academically competitive students in the world. Yet there continues to be dispute about how educational excellence is best achieved within American traditions and values. After almost two decades of explorations, policy pronouncements, convocations, presidential councils, congressionally supported studies, and state and national legislation, there is broad agreement on a number of points, many of which were considered precedent-shaking at the time of their resolution:

* States, rather than local school districts or the federal government, should have the responsibility for identifying the standards — content and skills — students should attain;
* States should take the lead in creating or acquiring the tests to measure student progress;
* Standards and assessments should be challenging rather than minimal;
* Progress toward proficiency is to be made by each identifiable subgroup (for example, those whose first language is not English or those at socioeconomic disadvantage) in a school;
* Incentives should be available to increase the focus and efficiency of
the system;
* All parts of the system should work together to achieve the specified targets.

Easier said than done. The enactment of the federal No Child Left Behind (NCLB) legislation is an attempt to clarify and strengthen the incentives for progress for all children. It requires the specification of targets for achievement growth and raises the number of grade levels that are tested annually. So far, the ramifications of the law are in dispute, with many districts reporting success based on test-score improvement while others complain of burdensome and intrusive regulations and high costs.

QUICK REVIEW OF RECENT POLICY may clarify why there is concern about the provisions of NCLB, which appear to focus attention on the learning of each and every child. Federal requirements related to state policies have, in general, been limited to concerns about civil rights, support for disadvantaged and handicapped students, dissemination of models for improvement, and research and statistics. The generation of national, not federal, standards and curriculum goals — what should be taught — began in the late 1980s in mathematics, and later in the sciences, arts, civics, history and language. Following a statement of National Educational Goals by the nation's governors, states for the most part voluntarily adopted or adapted subsets of standards to guide their teaching and assessment policy.

Many of the standards communicate on a general level what is to be learned but do not provide sufficient guidance for some teachers to select appropriate instructional methods, sample texts or analytical methods. Take, for example, this from California's History-Social Science Standard 6.1: "Students describe what is known through archaeological studies of the early physical and cultural development of humankind from the Paleolithic era to the agricultural revolution."

This standard is so broad in its coverage that it is difficult to know what to teach and what will be measured. Consider attempts at specificity — for example, these from Grade 6 History-Social Science Standards 6.1.2 and 6.1.3: "Identify the locations of human communities that populated the major regions of the world and describe how humans adapted to a variety of environments" . . . "Discuss climate changes and human modifications of the physical environment."

From these examples, it is clear that in planning instruction, a diligent teacher might miss by a mile the test designers' interpretation of the standards. This teacher will have inadvertently reduced the fairness of education for his or her students. Other standards are very clear about expectations and benchmarks. The key logic underpinning NCLB and other results-focused innovations is a clear relationship among standards, instruction, tests and learning. The idea is that, based on results of regularly administered measures, progress can be gauged, and if students have difficulty, educators will be able to fix their problems.

Factors including cost, tradition and technical reasons have led many states to use achievement tests that are relatively short and that necessarily represent only a few of the types of knowledge, skills and situations described in standards. For the most part, they consist of short, easily scored items.

What happens, then, when test developers and teachers have very different interpretations of a set of standards or there are too many standards to be tested by the state? As tests are kept secret, and details of test design and content are usually not readily available, it is possible for teachers to miss important areas that will be tested. In reality, the interpretation of standards by the test developer is given most weight. Because results on tests are used to distribute sanctions (for instance, classifying a school as needing improvement or restructuring), the test itself, not the standards, becomes the major blueprint that guides teaching and learning.

There is nothing wrong with the idea under certain circumstances; I am happy, for example, to have a Boeing 777 flight simulator used as the way to certify the competence of pilots in emergencies, bad weather and other difficult situations. However, flying a 777, demanding as it is, has a complex but agreed-upon set of procedures at its core. What about mathematics? Science? English language? These are open to wide and legitimate ranges of interpretation.

So one fundamental area of disagreement is whether current tests narrow or focus the curriculum and whether they result in consequences that are good or bad for learning. One option is to focus tests on a core of common skills while also offering other subjects and skills that broaden the curriculum but are not necessarily the subject of evaluation. But short of practicing the test items, or those similar to them, there is little reason to focus on standards rather than on the test. I believe this fact is contrary to the desire of the public for improved educational quality.

A second, related concern is how the targets that are set for performance improvement (adequate yearly progress, or AYP) interact with learning. AYP is the amount of improvement schools need to make, starting from their present point, to attain 100-percent proficiency within 12 years. Some policymakers thought this requirement would mean equal progress annually — that is, 1/12th of the distance would be achieved in Year 1 and, cumulatively, 12/12ths would be achieved at the end of the 12th year. If a school started at 20-percent proficiency, then it would have to increase the percentage of proficient students by a little less than 7 percent annually for all students to meet the standard. However, a reality is that students in schools do not often remain in place, and the formula must account for differences in expertise in entering and leaving students. Because sanctions are imposed far earlier than the 12th year, however, states have reacted in a variety of ways: While all are required to make sufficient progress early on, some states have chosen to defer major progress until the last few years.

Ironically, the AYP provisions created tension during the legislative process, even though states may choose their own standards and their own tests, determine their own levels of proficiency and develop a fully rigorous system or one that uses attainable standards and tests so that early progress is shown. Different states have interpreted the law in a wide variety of ways, and some states' well-defined procedures for measuring progress have been disallowed by the federal government. Added to the mix is the contention that NCLB carries great costs that are not now, nor likely to be, reimbursed by the federal government. There are numerous technical criticisms of the artifacts caused by requiring multiple subgroups to meet the targets, with raging battles about the accuracy, fairness and utility of the published results.

http://www.magazine.ucla.edu/year2004/spring04_07_04.html