Academic journal article Phi Delta Kappan

Better Tests and Testing Practices: Options for Policy Makers

Academic journal article Phi Delta Kappan

Better Tests and Testing Practices: Options for Policy Makers

Article excerpt

The issues discussed in this article -- appropriate test use and continued funding of research -- must be kept high on the policy and planning agendas of any agency or level of government involved in testing or in the implementation of new assessment strategies, these authors aver.

STUDENT TESTING has always played a pivotal role in American education. Every day millions of children take tests. Tests designed and administered outside the classroom are given less frequently than teacher-made tests, but they are nonetheless fixtures in American schools and have become a major force in shaping public attitudes about the quality of our schools and the capability of our students.

Standardized tests, typically in the multiple-choice format, can provide information about student knowledge and achievement. New testing methods -- which incorporate more direct assessment of student knowledge and performance and which profit from advances in computer and video technologies -- could enrich the information provided by conventional tests. For example, "performance assessment" methods require students to create answers or products to demonstrate what they are learning. Some of these assessments are extensions of familiar paper-and-pencil tests; others involve oral presentations, hands-on experiments, or portfolios of student work.

Educational testing has become a popular tool of accountability. Children, schools, and school systems are ranked on the basis of test statistics; comparisons of states and districts are commonly found in everything from newspaper editorials to real estate advertisements; promotion, graduation, and admissions decisions are based to varying degrees on test scores. However, attaching high stakes to test results is risky. Individual students' future educational opportunities can hinge on scores that only partially reflect their achievement or competence, raising questions of fairness. Pressure on teachers and students to raise scores can shift attention from learning to test-taking; as a result, fluctuations in scores do not necessarily indicate real changes in student achievement or in the performance of a school system.

These problems can arise regardless of a test's format: new methods could fall prey to the kinds of inappropriate uses and interpretations that have plagued conventional multiple-choice tests. Education leaders in and out of Congress should resist the temptation to expand existing testing programs or to adopt new ones until their purposes and probable effects are clarified.

The debate over testing reform sets the stage for potentially far-reaching congressional action. In its recent report to Congress, the Office of Technology Assessment (OTA) analyzed policy options in several areas in which the federal government plays a key role: the National Assessment of Educational Progress, testing and evaluation requirements under Title I/Chapter 1, and proposals for new national testing programs. In addition, however, the OTA report identified two cross-cutting issues that affect testing policy and practice at all levels -- school, district, state, and federal. Policy options for these two issues -- ensuring appropriate test use and promoting research on and development of testing methods and practices -- are discussed in this article.


The ways tests should be used and the types of inferences that can appropriately be drawn from them are often not very well understood by policy makers, school administrators, teachers, and other consumers of test information. Perhaps most important, many parents and test-takers are often at a loss to understand the reasons for testing, the meaning of test scores, and the importance of the consequences of test results. School policies about how test scores will be used are important not only to students and parents, but also to teachers and other school personnel whose own careers may be influenced by the test performance of their pupils. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.