课程测验项目分析方法——以对外汉语课程测验为例  

The New Methodology of Item Analysis in the Curriculum Test: Taking Chinese as A Foreign Language Test for Example

作　　者：张军[1] 

机构地区：[1]北京语言大学汉语进修学院

出　　处：《考试研究》2013年第4期68-75,共8页Examinations Research

基　　金：北京语言大学校级科研项目(项目编号:12YBG033;中央高校基本科研业务专项资金资助)部分研究成果

摘　　要：对外汉语课程测验属于标准参照性测验,应使用标准参照性测验理论体系下的技术指标对测验进行项目分析和评价,传统的分析方法(如区分度)不完全适用于课程测验的项目分析。本文使用该理论对北京语言大学汉语进修学院某次考试试卷进行分析,希冀为对外汉语教学提供一些有益的经验。实验结果表明:对"掌握者"和"未掌握者"来说,题目难度总体上可接受,大部分题目的区分性能良好,虽然有的题目略有"瑕疵",但值得保留,以提高教学内容的测试覆盖面及测验信度。有7个题目过难或过易,几乎不具备区分性能,需要删除或修改。The curriculum test used in the domain of Chinese Teaching as a Foreign Language is a kind of criterion - referenced test. In the process of item analysis, there are some unique statistic methods, because traditional methods, such as the index of discrimination, are not appropriate. This paper applies the theory of criterion - referenced test to analyze and evaluate the test used in the advanced Chinese Language College of Beijing Language and Culture University. The result of of the research shows that： for both the ＂masters＂ and ＂non-masters＂, the indexes of difficulty of the items are acceptable. Most items may effectively distinguish two kinds of students above mentioned. Even though some have ＂flaws＂, it is worthy to keep them to improve the reliability of the test. Seven items are too hard or too easy, which have little discrimination power, should be deleted or revised.

关 键 词：标准参照性测验 项目分析 

分 类 号：G424.74[文化科学—课程与教学论]