情境因素漢語關係子句之閱讀理解
No Thumbnail Available
Date
2019
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
本研究之目的為:利用語法情境去除漢語關係子句(Chinese Relative Clause, CRC)之歧義，以免在閱讀理解過程中造成混淆，並利用情境設計幫助第二語言習得者之CRC閱讀理解。研究中，透過填充句設計兩種情境並搭配目標句之兩種句型(Subject Relative Clause, SRC及Object Relative Clause, ORC)，經由實驗以獲知不同之兩種情境是否有不同效應及影響SRC/ORC之理解，並從中了解以漢語為第二語言習得目標之英語母語者與漢語母語者理解CRC之差異。實驗材料包含目標句與填充句各96句，總計192句。目標句又區分為SRC與 ORC各48句, SRC與 ORC又各自與填充句搭配區分為結構式情境及語意式情境各24句；而填充句之設計僅區分為以上兩種情境各48句。每一目標句並做適當切分以擷取各切分元素之閱讀時間(reading time, RT) ，句末並隨附問題以確認受試者是否了解句意，同時，可確保受試者能專心作答。本實驗採取自主式閱讀方式進行，計算受試者之 RT以作為閱讀理解難易程度之比較標準。其中，RT係以線性混合迴歸模型 (Linear Mixed Effect Regression, lmer) 進行分析，而答題正確率係以廣義線性混合迴歸模型(Generalized Linear Mixed Effect Regression, glmer)進行分析。總計完成英語母語受試者(L2學習者) 27名，漢語母語受試者 43名，共70名受試者之實驗。分析結果顯示:於不同母語者、不同情境及切分元素之不同位置等條件下，受試者之RT結果皆存在著顯著差異；而答題正確率於不同母語者、不同RC型式及不同情境等條件下亦達顯著差異。本研究之情境因素設計可增進L2學習者漢語關係子句之閱讀理解並得知其困難所在。
The purpose of this research is to help reading comprehension of CRC(Chinese Relative Clause) for second language learners by removing the ambiguity firstly, in addition, two kinds of contexts are also designed as the filler sentences. The target CRC are divided into SRC (Subject Relative Clause) and ORC (Object Relative Clause).Through the experiment, the different effects are expected, including those induced by two different native speakers, RC types and contexts. There are 96 target sentences and 96 filler sentences.The filler sentences are equally divided into two groups with structural and semantic context respectively. Each target sentence is cut into 9 proper components for more precisely obtaining the reading time (RT). In addition, a yes-no question is provided for knowing if the participants understand the meaning of sentences and keeping them to concentrate on the test.The experiment is conducted by a self-paced reading task. Totally, 70 participants have finished the experiment, including 27 native English speakers and 43 native Chinese speakers. The RT result is analyzed by linear mixed effect regression model, and the ACC result is analyzed by generalized linear mixed effect regression model. The results show that different speakers, contexts and word order have the significantly different RT effects. And different speakers, RC types and contexts have the significantly different ACC effects. Therefore, this research can help and enhance the CRC reading comprehension for L2 learners and locate the difficulties in their reading.
The purpose of this research is to help reading comprehension of CRC(Chinese Relative Clause) for second language learners by removing the ambiguity firstly, in addition, two kinds of contexts are also designed as the filler sentences. The target CRC are divided into SRC (Subject Relative Clause) and ORC (Object Relative Clause).Through the experiment, the different effects are expected, including those induced by two different native speakers, RC types and contexts. There are 96 target sentences and 96 filler sentences.The filler sentences are equally divided into two groups with structural and semantic context respectively. Each target sentence is cut into 9 proper components for more precisely obtaining the reading time (RT). In addition, a yes-no question is provided for knowing if the participants understand the meaning of sentences and keeping them to concentrate on the test.The experiment is conducted by a self-paced reading task. Totally, 70 participants have finished the experiment, including 27 native English speakers and 43 native Chinese speakers. The RT result is analyzed by linear mixed effect regression model, and the ACC result is analyzed by generalized linear mixed effect regression model. The results show that different speakers, contexts and word order have the significantly different RT effects. And different speakers, RC types and contexts have the significantly different ACC effects. Therefore, this research can help and enhance the CRC reading comprehension for L2 learners and locate the difficulties in their reading.
Description
Keywords
語法情境, 關係子句, 線性混合模型, 廣義線性混合模型, relative clause, self-paced reading, linear mixed effect regression, generalized linear mixed effect regression