汉语水平口试信度的理论与实证研究  

The Theoretical Analysis and Experimental Research of Rater Reliability of Chinese Oral Proficiency Test

作　　者：柴省三[1] 

机构地区：[1]北京语言大学汉语水平考试中心,100083

出　　处：《语言教学与研究》2003年第4期69-77,共9页Language Teaching and Linguistic Studies

摘　　要：本文首先对影响第二语言口试信度的基本维度进行了理论分析 ,然后通过实验设计的方法 ,从评分一致性和准确性两个角度对HSK(高等 )口试评分的信度进行了研究。实验结果表明 :HSK(高等 )口试的评分员信度、组间信度、历时性信度和分类一致性信度均比较高 ;评分的准确性和区分度还有提升的空间 ,建议在以后的评分实践中 ,采用口试理论常模来规范口试评分等级的频数分布 ,以提高评分的准确性。This paper attempts to examine the rater reliability of oral proficiency test in second language acquisition. Based on the fundamental considerations of the multidimensionality in oral proficiency test systems, the article, by experimental methods, investigates the rater reliability as seen from rating consistency and accuracy perspectives. The paper concludes that the rating consistency among raters is more dependable than the rating accuracy. The use of theoretical norms in the rating of Chinese oral proficiency test is recommended in the future scoring practice.

关 键 词：汉语水平口试 汉语水平考试 组间信度 理论研究 实证研究 评分一致性 准确性 HSK 评分员信度 历时性信度 分类一致性信度 

分 类 号：H191[语言文字—汉语]