概化理论下英语翻译测试评分者的信度研究  

Using Generalizability Theory to Examine Raters’Reliability for English Translation Test

作　　者：阎莉[1] Yan Li(Wuhan Institute of Technology, Wuhan 430073, Hubei)

机构地区：[1]武汉工程大学外语学院,湖北武汉430073

出　　处：《武汉工程职业技术学院学报》2020年第2期75-78,108,共5页Journal of Wuhan Engineering Institute

基　　金：武汉工程大学人文社科基金项目(项目编号007032)。

摘　　要：使用概化理论作为分析框架,请20位评分者对11位大学生每人2次翻译测试进行整体评分,通过一系列的概化(G)和决策(D)研究,以检验不同背景评分者的翻译测试的有效性和可靠性问题。实验研究的结果表明,与有培训背景的评分者相比,没有培训背景的评估者对评分标准把握不一致,对翻译评分存在较大差异。同时还发现,当前翻译测试主要靠评分者的经验来兑现将无法产生可接受可靠性系数,对评分者进行系统培训可以提高翻译测试信度。Using generalizability theory as a framework for analysis,the goal of this experimental study was to examine the reliability of translation tests raters from different backgrounds.Twenty-two passages translated by eleven college students were scored holistically by 20 raters using the real scoring rubric.A series of G and decision(D)studies were conducted to examine the reliability issues.The results of the experimental research indicate that the raters without training background show inconsistent scoring rubric compared with the raters with training background.At the same time,it indicates that the current translation test mainly relies on the experience of raters to achieve the reliability,and the raters receiving a series of systematic training are more reliable.

关 键 词：概化理论 翻译测试 评分者 翻译评分 评分信度 概化 决策 

分 类 号：H08[语言文字—语言学]