摘要
本文是对集库式完型填空的效度研究,旨在探究它是否能够测量受试在篇章层次上理解和运用单词的能力。受试为141名非英语专业的本科生,两篇研究素材皆取自大学英语四级考试的深度阅读部分,分别为介绍型和叙述型。研究工具为四个测验,两个原文测验中句子顺序正常排列,两个散句测验中句子顺序则完全打乱,受试随机分为两组,分别完成了两个基于不同文章、句子顺序不同的测验。结果显示,在叙述型文章的完型填空中,句与句之间的信息发挥了较为显著的作用,而高水平组的受试最有效地利用了这种篇章层次的线索。
This study investigates whether banked cloze can measure test-takers' discourse level processing ability. 141 second year non-English majors in Shenzhen University serve as the subjects. The instruments are four tests, based on two passages, a descriptive and a narrative, taken from CET4 Reading in Depth Section. Two of the tests are normal banked eloze tests in which sentences are in naturally occurring order. In the two scrambled tests the sentence order is thoroughly scrambled. Subjects are randomly divided into two groups, each taking a normal test and a scrambled test based on different passages. Analysis of subjects' performance reveals intersentential ties are salient in the normal order test based on the narrative passage. And the high proficiency group are more capable of making use of discourse level constraints.
出处
《中国外语》
CSSCI
2010年第4期70-76,共7页
Foreign Languages in China