閱讀時，口語和視覺通路能告訴我們什麼呢？以「眼動異讀分析」透視國中生英語閱讀

Abstract

本研究目的是藉由「眼動異讀分析」探討台灣國中學生的英語閱讀歷程，希望能從外顯的「口語」（異讀）和「視覺」（眼動）資料窺探讀者閱讀時的內在認知歷程，並進一步了解EFL讀者對於語言線索和閱讀策略在閱讀英語時的使用情形，進而提供教師閱讀教學上之資訊。 研究對象為13位國中二年級英語閱讀能力為中上之學生，學生朗讀一篇英語說明文，並請教師與學生各自評定文本中各單字之熟悉度。研究者將文章中不同字彙特性（內容字vs.功能字、熟悉vs.不熟悉），配合深度程序異讀分析，探討讀者在字詞層面和句子層面使用「形音線索」、「語法線索」、「語意線索」三種語言線索和閱讀策略協助閱讀理解文章之情形。進一步以眼動儀所得到的各項眼動指標數據檢視異讀分析所探知的字詞與句子層面的閱讀理解指標。 本研究之結果為（一）主要的異讀類型為「非字替代」、「實字替代」和「刪去」。（二）「不熟悉的內容字」需花費較長的認知歷程來理解意義。（三）EFL國中生缺乏有效使用閱讀策略之能力。（四）異讀字多為高形似和高音似，為使用「形音線索」之證據。（五）EFL國中生無法有效運用語法和語意線索。（六）異讀數量、朗讀時間可做為探查學生閱讀能力之參考。最後研究者依據研究結果，提出若干建議供未來同領域之研究人員、教師以及出版業者作為參考。

This study was to investigate English reading process of Taiwan junior high students with Eye Movement Miscue Analysis (EMMA). The researcher wanted to explore internal information process from external verbal and visual data to have a further understanding about how EFL readers utilize language cuing systems and reading strategies during reading English. The results of this study can provide teachers instructional implications on reading instructions. Participants were 13 8th graders with above- average English reading proficiency. Each student read aloud an English expository text. Teachers and students respectively rated word familiarity of the text. According to different characteristics of words (content vs. function words, familiar vs. unfamiliar words), the researcher used the in-depth miscue analysis procedure to explore the utilization of graphophonic cue, syntactic cue and semantic cue at word and sentence levels to understand how the readers constructed the meaning from the text. Besides, the researcher integrated the visual data of eye movement parameters and the verbal data from miscue analysis at word and sentence levels to further understand the readers’ comprehension during reading. The major findings of this study are summarized as follows. 1. The non-word substitution, real-word substitution and omission types of miscues were more often appeared in English reading. 2. The participants took a lot of time on physiological processes of unfamiliar content words. 3. The participants were lack of ability of using reading strategies. 4. The most proportion of miscues appeared to be high in graphic and sound similarity. This provides a piece of evidence that EFL readers utilize graphophonic cue for comprehending the text. 5. The participants couldn’t use syntactic and semantic cues effectively and efficiently. 6. The quantity of miscues and oral reading time could be used in evaluating readers’ reading proficiency. Finally, several suggestions are given for future researchers, teachers and publishers.