论文部分内容阅读
本研究应用项目反应理论,从单维性假设验证、模型选择、项目筛选等方面分析2007年TEM4听力理解项目的质量。分析显示,该项目难度中等,区分度非常理想,听力理解水平较低的考生仅凭猜测答对的概率很小。本文建议命题人员从提高题目区分度、增强题目对听力理解项目的贡献值等方面来改善部分质量较差的题目;对于猜测概率超出可控范围的项目,不仅需要研究考生的答题过程,而且需要分析项目的出现顺序。本文的研究方法对国内外语测试的题目质量分析和题库建设具有借鉴意义。
In this study, the project response theory was used to analyze the quality of 2007 TEM4 listening comprehension project from aspects of unidimensional hypothesis verification, model selection and project screening. The analysis shows that the project is moderately difficult, the degree of discrimination is very good, and the candidates with low level of listening comprehension only have a small probability of answering the question correctly. In this paper, we propose that proponents improve some of the topics with poor quality in terms of improving the degree of item discrimination and enhancing the contribution of items to listening comprehension items. For those items whose guessed probability is beyond the controllable range, it is not only necessary to study candidates’ Analyze the order in which the items appear. The research method of this article has reference significance for the topic quality analysis and question bank construction of the domestic and foreign language test.