摘要
主观性考试的评分信度问题一直是语言测试研究体系中的重要组成部分.据以往研究.评分员如何把握评分标准对评分信度具有至关重要的作用,鉴于此,本研究从2009年4月HSK(高等)写作测试评分的48组评分员中随机抽取了两组(每组中一老一新两名评分员),运用概化理论对两组的评分信度进行了检验,结果表明两组的测验信度较高;通过比较评分员问评分数据的描述性统计结果以及调查问卷,大致了解到新老评分员在对评分标准把握中的不同,这些因素也导致他们在评分过程中产生评分差异。
The scorer reliability of subjective tests plays an important part to the language testing research. As the studies before, how scorers use the marking-criterion is very important to the scorer reliability. This study applys Generalizability Theory to analyse the scorer reliability of two groups , each group consists of one experienced scorer and one inexperienced scorer. Results show that the reliability of two groups are not bad. During analyzing the descriptive statistics and questionnaire, We learn about the differences between experienced scorers and inexperienced scorers. These factors may lead to the scoring differences.
出处
《中国考试》
2010年第10期13-19,共7页
journal of China Examinations
基金
北京语言大学2006年校级青年项目“第二语言写作测试评分标准专题研究”(项目编号:06QN06)的阶段性成果
关键词
写作测试
信度
概化理论
HSK Writing Test
Reliability
Generalizability Theory