PISA中国试测研究的评分者效应分析对高考网上阅卷的启示  

PISA China Trial Rater Effects Analysis for Gaokao Online Rating

作　　者：王蕾[1] 张文静[2] 

机构地区：[1]教育部考试中心,北京100084 [2]北京师范大学发展心理研究所,北京100875

出　　处：《统计与信息论坛》2012年第6期95-101,共7页Journal of Statistics and Information

基　　金：全国教育科学规划2009年度教育考试研究专项课题<学生能力国际评价PISA的教育评价技术在高考中的应用>(GFA097021)

摘　　要：针对PISA 2009中国试测研究主观题评分环节所采用的多重编码设计,分析在阅读、数学和科学领域的评分中是否存在评分者效应。根据多侧面Rasch模型方法,分别对这三个领域进行评分者主效应的分析。结果显示:阅读和科学领域中,评分者之间的严苛度/宽松度差异非常显著;而数学领域中,评分者之间的严苛度/宽松度差异较小。最后,探讨了这些结果的可能原因以及对高考网上阅卷评分借鉴的建议。The aim of this study was to analyze rater effects in reading, math, and science for PISA2009 China Trial Study. Based on the many--acet Rasch measurement methodology, the focus was on the rater main effects in the three subjects respectively. Results show that considerably strong rater severity/leniency effects occurred in reading and science, but a relatively weak rater severity/leniency effects for math. The potential reasons about these results were discussed and some implications for National College Entrance Examination（commonly referred to as ＂Gaokao＂）） online ratings are provided.

关 键 词：评分者效应 严苛度/宽松度 PISA 多重编码 高考网上阅卷 

分 类 号：B841[哲学宗教—基础心理学]