摘要
本文在两参数逻辑斯蒂克模型(2PLM)下,提出一种新的选题策略——平均测验难度匹配法(Avt-b),并对四种选题策略下EAP能力估计趋势进行比较研究。通过模拟研究显示,Avt-b方法在CAT前期能够较快地锁定能力范围,较准确地作出能力估计。本文对CAT测试阶段的能力误差范围进行确定,对于多级评分模型的CAT选题策略开发具有一定的借鉴意义。
Computerized Adaptive Testing （CAT） is one of the important innovations in measurement applications fostered by developments in IRT. It＇ s imperative for us to study the CATs based on Two Parameters Logistic Model （2PLM）. This research is based on 2PLM. One new Art-b（ Average Test difficulty Strategy） is proposed on the basis of discussing previous ISSs （Item Selection Strategies） of CATs based on 2PLM. This paper mainly discussed the trend of estimated ability of above four ISSs based on 2PLM. Simulation results shows： Average test difficulty strategy can fix ability range at the early stage of CAT as MII in Item Exposure Control,and it can compare to a-STR method
出处
《考试研究》
2009年第3期60-70,共11页
Examinations Research
关键词
能力估计趋势
计算机化自适应测验
平均测验难度
匹配法
the trend of estimated ability Computerized Adaptive Testing average test difficulty strategy
