作　　者： ;

机构地区： 东华理工大学外国语学院

出　　处： 《外语测试与教学》 2011年第4期27-32,55,共7页

摘　　要： 通过运用多层面Rasch模型,本研究对中外大学英语教师的写作评分行为进行微观分析,探讨这两组不同语言背景的大学英语教师对同样的写作样本的评分是否有差异。研究结果表明中外教师的写作评分结果并无显著差异,他们的整体性评分都普遍比分析性评分更宽松,但中国教师的评分较外籍教师的评分略为宽松。微观的评分行为分析揭示出中外教师对组织、语法和写作规范这三个评分项目上产生了偏差。外籍教师在各评分项目上的评分产生的偏差总数比中国教师多;他们对组织和写作规范的评分偏宽松,而对语法的评分则有偏严厉的趋势。对于不同水平的考生,无论是中国教师还是外籍教师偏宽松的偏差较多;对于水平最高的考生的评分,中国老师偏宽松,而外籍教师则偏严厉。 The present study employed multi-faceted Rasch measurement （MFRM） to explore the rater behavior of Chinese university EFL teachers and native English-speaking teachers when assessing EFL students＇ essays. Rating results of the two groups of raters prove not statistically significant. Both groups of raters rated more leniently using the holistic rating than the analytical rating. The Chinese raters tended to rate more leniently than native speakers. MFRM further revealed different rater biases between the two groups of raters. Native speakers＇ interaction with the rating scales generated more rater biases than the Chinese teachers, and their ratings turned to be more lenient in terms of organization and mechanism, while more severe when rating grammar. The raters＇ interaction with the test takers produced more lenient biases. When rating the most capable test taker, Chinese teachers turned to be more lenient, while the native speakers rated more severely.

关 键 词： 评分行为 评分项目 考生能力 多层面 模型

领　　域： [语言文字]