摘要
四参数Logistic模型通过引入第4个参数,分析测验中高能力被试答错低难度试题的睡眠现象。运用四参数Logistic模型与双参数、三参数Logistic模型对汉语作为第二语言测验的实测数据进行分析,比对3种模型的分析结果,并通过四参数Logistic模型分析测验中的睡眠现象,结果表明:在模型拟合方面,四参数Logistic模型表现良好;四参数Logistic模型对项目区分度的估计结果高于双参数、三参数Logistic模型;语言测验存在睡眠现象时,被试能力值可能会被严重低估,高风险考试需要考虑使用四参数Logistic模型进行矫正;在语言测验中,睡眠现象是一种随机的失误现象,与所测的语言技能无关。
The four-parameter logistic model was used to analyze the sleeping phenomenon in which examinees with high ability accidentally give wrong responses to questions they should have answered correctly.In this study,the four-parameter logistic model was compared with the two-parameter and three parameter logistic models to analyze the actual data of the Chinese as a second language testing,and the four parameter logistic model was used to analyze the sleeping phenomenon in the test.The results showed that:the four-parameter logistic model performed well in terms of model fit;the four-parameter logistic model estimated item discrimination better than the two and three-parameter logistic models;the presence of the sleeping phenomenon in language tests may cause serious underestimation of the examinees’ability values,and the four-parameter logistic model needs to be considered for correction in high-stakes tests;the sleeping phenomenon is a random error phenomenon in language testing and is not related to the part of the test.
作者
金英姿
王佶旻
JIN Yingzi;WANG Jimin(Beijing Language and Culture University,Beijing 100083,China)
出处
《中国考试》
CSSCI
北大核心
2022年第8期57-65,共9页
journal of China Examinations