摘要
试题的难度常常用试题的通过率来表示,这种事后检验的方法虽然简单易行,其数量指标严格确定,但是对考生水平相当敏感,也不能在试题设计中起到主导作用。本研究根据认知任务分析的基本原则,设计了对试题难度进行事前评定的框架和严格的实施程序,并以高等数学试题为分析对象进行了检验。根据数学试题的特点,设定了要素数量、要素辨识难度、原理数量、原理辨识难度和认知负荷等指标。结果显示,各试题事前难度值与其通过率之间存在极其显著的相关。评定指标之间、评定者之间也存在极其显著的一致性,说明本研究的事前评定方法具有充分的效度和信度。
The passing rate of a test item often serves as an index of its difficulty. This posterior index,though easy to obtain and definite, is susceptible to students＇ learning level. It cannot play a leading role in designing test items. The present research tried to design a prior-to-test index of test item difficulty. According to the principles of cognitive task analysis, a frame and a procedure to be strictly executed were mapped out for the prior assessment of higher mathematics test items. Taking into account the characteristics of math problems, we designed such indexes as number of elements in a problem, element identification difficulty, number of principles used in answering, principle identification difficulty and cognitive load. The results show that prior difficulty is most significantly correlated with the passing rates of math test items. High correlation also exists among sub-indexes in assessment and among evaluators, indicating sufficient validity and reliability of the prior assessment method developed by the present reseaeh.
出处
《心理科学》
CSSCI
CSCD
北大核心
2008年第3期696-698,共3页
Journal of Psychological Science
关键词
试题难度
认知任务分析
事前难度
test item difficulty, cognitive task analysis, prior diffculty
作者简介
通讯作者：邵志芳，男。E—mail：zfshao@psy．ecnu．edu．cn
