pact may be important. For example, some schools may achieve real gains in scores by methods that parents and communities might consider unacceptable—for example, by eliminating important activities such as instrumental music or by imposing oppressive amounts of homework that preclude extracurricular activities. Similarly, some teachers may consider it unproductive in terms of test scores to assign certain types of schoolwork such as long-term research projects or exploration of student-initiated questions. Both desirable and undesirable instructional effects of the accountability system may be distributed inequitably.

Embedding Assessments in an Accountability System

The larger question is the role of achievement tests in an educational accountability system. Achievement tests are an important but insufficient basis for holding schools accountable. "School performance" is more than the aggregate of student test scores, and the considerations above suggest that even high-quality tests do not work well enough as an accountability mechanism to warrant relying on them in isolation.

Efforts to embed tests in a broader range of indicators of school performance, however, remain rudimentary and controversial. A number of current reforms hold schools accountable for a variety of outcomes other than test scores. Both Kentucky and Maryland, for example, hold schools accountable for improving performance on a variety of other outcomes—called "non-cognitive" indicators in Kentucky and "data-based areas" in Maryland—such as drop-out rates, promotion rates, and attendance rates. The accountability systems in both states, however, mirror the national debate in giving more weight to test scores than to other outcomes, and some of the others show too little variance or change over time to have much impact.

The inclusion of school-performance indicators other than student outcomes is currently mired in controversy. Along with renewed enthusiasm for using tests to hold schools accountable has come widespread disparagement of holding schools accountable for "process" variables, such as financial and other inputs, instructional offerings, and similar factors. Few reformers endorse complete elimination of process accountability; for example, few would advocate eliminating all due process guarantees for disabled students. Moreover, some process criteria, such as "delivery standards" or "opportunity-to-learn standards,'' remain a part, albeit a very controversial part, of current reform debate. Nonetheless, process-oriented accountability is widely considered insufficient; and many reformers consider it a counterproductive distraction from the more important issue of outcomes.

The distinction between "process" and "outcome" accountability, however, is misleading and hinders the effort to devise an effective set of indicators for monitoring school performance. Schools produce a wide range of outcomes, some of which shade into "process."

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement