摘要
本研究主要目的是针对高考英语成绩存在的城乡差异,检验这种差异是否来源于试题在城乡上的项目功能差异。如果两个能力本来相同的考生群体在某一试题得分上表现出不同程度的差异,该试题就存在项目功能差异。研究采用试题标准化分数差法,利用STDIF软件逐一分析了2016年三套全国高考英语卷的客观题是否存在城乡上的项目功能差异,在确定客观题没有项目功能差异后,以客观题成绩为匹配变量,采用条件得分图法对书面表达题是否存在城乡上的项目功能差异进行了分析。研究结果显示,高考英语全国I、II、III卷均未发现城乡上的项目功能差异试题,即可以认为高考英语全国卷对城乡不同户籍考生都非常公平、公正,城乡考生在英语成绩上的差异并非题目的公平性所致。
The purpose of this study is to analyze test fairness of the 2016 National Matriculation English Test(NMET)through conducting differential item function(DIF)analyses.If the responses of two groups of students with the same level of language ability differ on a common item,then the item owns DIF values,which means the item poses different level of difficulty for the two groups.The descriptive statistics indicated that there was significant difference in NMET scores between urban test-takers and rural test-takers.The standardization approach was applied to assess the three NMET papers focusing on urban/rural test-takers by using STDIF software which can detect both uniform DIF and non-uniform DIF.DIF is also investigated for a single writing item using the conditional P-value method.The result shows that no DIF values were found in the three NMET papers between urban and rural test-takers,suggesting that the score difference between the groups could not be attributed to DIF.
作者
关丹丹
乔辉
陈康
韩奕帆
Guan Dandan;Qiao Hui;Chen Kang;Han Yifan(National Educational Examinations Authority,Beijing 100084)
出处
《心理学探新》
CSSCI
北大核心
2019年第1期64-69,共6页
Psychological Exploration
基金
全国教育科学规划单位资助教育部规划课题<新高考改革背景下的高考质量评价研究>(FBB160604)阶段性研究成果之一
关键词
高考英语
项目功能差异
城乡
National Matriculation English Test
differential item function
urban/rural