In this paper, the authors draw on recent international experience to asses
s the success of five propositions for how high stakes national testing can
improve classroom instruction and, ultimately, raise student achievement.
Findings indicate that testing can be an effective mechanism for improving
instructional practice, but its success is not ensured. It has failed as of
ten as it has succeeded, usually because those implementing the strategy fa
iled to understand the intermediate conditions that had to be met for chang
es in test content, format, or use to have the desired impact on teachers'
classroom practice. (C) 2000 Elsevier Science Ltd. All rights reserved.