PISA2018阅读素养计算机化自适应测试的技术与方法探析  

Analysis of the Technology and Method of PISA 2018 Reading Computerized Adaptive Test

作　　者：赵茜 马力 温红博 ZHAO Qian;MA Li;WEN Hongbo(Beijing Normal University,Beijing 100875,China)

机构地区：[1]北京师范大学中国基础教育质量监测协同创新中心,北京100875

出　　处：《中国考试》2020年第11期74-78,共5页journal of China Examinations

基　　金：中国基础教育质量监测协同创新中心基础教育质量监测科研基金项目“基于PISA数据的学校效能影响因素研究”(2019-01-077-BZK02)。

摘　　要：PISA2018阅读素养的计算机化自适应测试采用核心阶段、阶段1和阶段2的3阶段自适应测试,题库设定有245道题目,组成45个测试单元,并将其组合成若干题组,用于不同阶段的测试。在路径设计上,为避免位置效应问题,除核心阶段→阶段1→阶段2的标准路径之外,还采用核心阶段→阶段2→阶段1的替代路径。PISA2018阅读素养计算机化自适应测试建立了能力覆盖范围较为完整的题库,减少了学生群体差异带来的测量误差,进一步提高了测量效度和效率,但在更大范围推广、人工评分题目的信息无法用于测试选题等方面,计算机化自适应测试仍需要进一步探索。The proposed computerized adaptive testing design for the PISA 2018 reading literacy consists of three stages:core stage,stage 1,and stage 2.The PISA 2018 reading adaptive test contains 245 items,composed of 45 test units,and combined them into different testlets.In the path design,in order to avoid the problem of location effect,in addition to the original routing paths,core stage-stage 1-stage 2,the alternative path of core stage-stage 2-stage 1 is adopted.Items are selected according to total number correct on automatically-scored items on the given testlet and the test characteristic curves.The PISA 2018 reading computerized adaptive test has established the item bank that can meet the testing needs of students with different abilities,reduce measurement errors caused by differences in student groups,and further improve measurement accuracy and efficiency.However,a further exploration is still needed for the problems of the promotion of multistage computer adaptive testing on a larger scale and the fact that the information of human scored items cannot be used for test item selection.

关 键 词：PISA2018 计算机化自适应测试 多阶段自适应测试 项目反应理论 

分 类 号：G405[文化科学—教育学原理]