广东高考英语读写任务测试的语篇特点分析

Validation of Reading-to-Writing Tasks in NMET: A Text Analysis Approach

论证效度的办法之一就是分析考试中的语言现象与考试的建构是否一致[1-3]。从考生读写任务作文的语篇特点着手,分析高考英语广东卷中读写任务的考试建构、评分标准以及与学生水平的联系,从而观察读写结合写作测试任务的效度。对受试语篇特点的分析表明,受试者在考试过程中没有明显意识到他们需要对所引述的材料标明出处;另外,在归纳材料文章内容时,水平高的受试者会更多地使用同义词、改写句型等手法变更材料文章的句子,而水平低的受试者更多倾向于在直接抽取原文中几个句子后机械地把它们连接起来或者完全脱离材料文章的语言与句型而全部用自己的语言归纳。最后,受试者的英语水平影响到他们对材料文章内容的使用。

One approach to examine validity is to analyze if the discourse produced represents the test construct.This article,through the analysis of the discourse features of the articles written for the reading-to-writing tasks,aims at the inter-relationship between test construct,rubrics and participants' language proficiency level to draw out the evidence of validity.The results show that most test takers fail to explicitly attribute information taken out from the input text,and that those with high scores tend to copy or minimally modify the input text.Such findings seem to indicate that those who score high may not necessarily possess such a kind of ability specified in the test outlines.In addition,when summarizing the input text,participants with good performance are more likely to use synonyms,paraphrases and the change of sentence structures to rewrite the original sentences in the input text,while those with poor performance simply combine several parts extracted from a few sentences without knowing what these sentences really mean or use their own words without referring to the input text.Finally,the proficiency level of the test takers influences the extent to which the input information is used.