考试风险与期望反拨效应之间的关系

On the relationship between tests' stakes and intended washback

本文讨论考试风险与期望反拨效应之间的关系,剖析了考试风险概念,提出实现期望反拨效应的三个必要条件:1)考试必须引起后果;2)考试必须体现教学的样板;3)考试不引发应试教学。应用这些条件诠释相关实证研究的结果得出结论:风险过高或过低的考试均难以实现期望反拨效应。因此,如果希望利用考试推动教学改革,决策人和考试设计者应该考虑如何调整才能使考试的风险等级适度。

Based on an analysis of the concept of tests＇ stakes, I propose three conditions for intended washback and discuss the relationship between the stakes attached to a test and its potential for intended washback. The three conditions are applied to the discussion of the findings by some empirical studies on washback. Through the discussion, it is demonstrated neither low-stakes tests nor extremely high-stakes tests are capable of inducing the intended washback effect. Those tests which are more likely to bring about the desirable changes are tests whose stakes are at the right level. Thus, how to adjust the stakes of a test to the right level is an issue that should be addressed by policymakers and test designers who intend to use tests as agents for change.