使用题组反应模型缓解局部题目依赖性对多阶段测验的危害

发布时间：2018-03-26 10:23

本文选题：多阶段测验　切入点：局部依赖性　出处：《心理科学》2017年01期

【摘要】：尽管多阶段测验(MST)在保持自适应测验优点的同时允许测验编制者按照一定的约束条件去建构每一个模块和题板,但建构测验时若因忽视某些潜在的因素而导致题目之间出现局部题目依赖性(LID)时,也会对MST测验结果带来一定的危害。为探究"LID对MST的危害"这一问题,本研究首先介绍了MST和LID等相关概念;然后通过模拟研究比较探讨该问题,结果表明LID的存在会影响被试能力估计的精度但仍为估计偏差较小,且该危害不限于某一特定的路由规则;之后为消除该危害,使用了题组反应模型作为MST施测过程中的分析模型,结果表明尽管该方法能够消除部分危害但效果有限。这一方面表明LID对MST中被试能力估计精度所带来的危害确实值得关注,另一方面也表明在今后关于如何消除MST中由LID造成危害的方法仍值得进一步探究的。
[Abstract]:Although MSTs maintain the advantages of adaptive tests, they allow testers to construct each module and test board according to certain constraints. However, if some potential factors are ignored in the construction of the test, there will be some harm to the result of the MST test if the problem is caused by local problem dependence. In order to explore the problem of "the harm of LID to MST", In this study, the concepts of MST and LID are introduced, and then the problem is discussed by simulation. The results show that the existence of LID will affect the accuracy of the ability estimation of the subjects, but the estimation deviation is still small. And the harm is not limited to a particular routing rule, and then in order to eliminate the hazard, the problem group reaction model is used as the analysis model in the process of MST testing. The results show that although the method can eliminate some of the hazards, the effect is limited. On the one hand, it shows that the harm caused by LID to the accuracy of ability estimation in MST is indeed worthy of attention. On the other hand, it also shows that the methods of how to eliminate the harm caused by LID in MST in the future are still worthy of further exploration.
【作者单位】：北京师范大学中国基础教育质量监测协同创新中心;江西师范大学初等教育学院;江西师范大学心理学院;
【分类号】：B841

【相似文献】