高低能力國中生理解順向及非順向時序 英文對話性文本-來自眼動的證據

Abstract

綜觀我國國中英語教學目標強調培養說聽讀寫的基本素養，期使學生能在生活情境中實際展現出英語溝通能力，因此老師教學以溝通教學法為主，課本內容均為對話文本(dialog text)。由於國中課本每一課的對話文本均有搭配的圖，國中會考考題的圖文題目日漸增多，顯示閱讀文本與圖片交互參考以臻理解乃是國中畢業生須具備的基本閱讀能力。對話當下的理論―互動協同模式(Pickering & Garrod, 2009)，主張兩人對話過程會透過協同效應共同建立相似的情境模型，建立雙方協同的心智表徵才能互為理解，過去在聽、說、寫方面已證實有協同效應(楊媚，2015；Costa et al., 2008；Watson, Pickering & Branigan 2004；Garrod & Anderson, 1987)，惟閱讀部分的協同效應尚無人探討。Zwaan提出的事件標記模型，主張閱讀歷程乃在建構情境模型並整合前後文語意，情境模型具有若干維度(時間、空間、因果、意圖、主角)，倘若文本提示維度改變，讀者須隨即更新情境模型，其中時間維度造成情境模型更新變化可透過眼球追蹤技術的生理觀察而得知(Bubic, Susac, & Palmovic, 2015; Zwaan, 2008；何先友與林崇德，2008；冷因，2004；Kintsch &Van Dijk，1978)。搭配了月曆圖片的對話文本能幫助讀者更新情境模型的時間維度嗎? 由於圖文交錯的眼動表現之實證研究還非常缺乏(Hyona, 2010; Mayer, 2010)，本研究擬提供結果以豐富相關研究的進展。總之，本研究欲透過眼球追蹤技術了解高低英文能力國中學生，如何閱讀配有輔助圖的對話文本，如何擷取文本與圖片的時間資訊以利建構出心智協同的理解框架。 以英語閱讀能力與文本時序因子作為本研究兩個因子，樣本為新竹某國中62位七年級生，分為四組，「高閱讀能力順向時序文本組」、「低閱讀能力順向時序文本組」、「高閱讀能力非順向時序文本組」、「低能力非順向時序文本組」，了解其是否影響學生閱讀理解與心智協同模式的閱讀理解表現，並透過眼動儀器蒐集閱讀歷程的注意力及圖文搭配的閱讀順序，透過2 x 2變異數分析了解不同組別的差異。 研究結果發現英語能力、文本時序會局部影響學生的心智協同與閱讀理解。在時間維度的議題上，由「個別時間詞文字區」得知起始時間訊息對於建構情境模型最為重要，且從「個別時間詞圖片區」發現「非順向時序文本」的眼動凝視程度在順向與非順向交錯的時間點最長，交錯的時間點過後，凝視時間漸減；然而在「順向時序文本」的眼動表現會根據時間累進而增加，符合Zwaan的理論主張: 不同維度會透過疊加方式去校正更新情境模型(事件標記模型)。 在高低能力的議題上: 高能力組在「總時間詞文字區」會針對不同時序文本有效率的配置不同注意力(總凝視時間)，低閱讀能力組在所有時間詞上的凝視次數均高於高閱讀組，但低能力學生的後測未因凝視次數增加而理解分數增加。從高閱讀能力組的掃視次數和訪談資料可以得知，他們進行較多的圖文交互參照；低閱讀能力組的掃視次數較少，顯示幾乎不進行圖文交互參照，訪談資料顯示他們無法理解圖文訊息的搭配。此研究結果發現時間建構對於情境模型的影響以起始時間最有意義，且受到不同英語能力的影響，高能力的學生會針對文本提供的時間點，適時以圖片輔助閱讀，但低能力學生無法有效閱讀，因此有賴教育現場的老師幫助學生如何從文字和圖片間找出關聯並建構文意。

Interactive Alignment Model theorizes thattwo interlocutors in a successful dialogue align to an equivalent representation at different level (Pickering & Garrod, 2009). Alignment effect has been proved in listening, speaking and writing (楊媚，2015；Costa et al., 2008；Watson, Pickering & Branigan 2004；Garrod & Anderson, 1987) while alignment in reading is yet to be explored. Based on Zwaan’s event indexing model, previous research indicated readers will update their situational model immediately after the dimension in the reading changed (趙雪汝，2014). One of the dimension, time, described the temporal change within the text which increased the processing load are found through readers’ eye movement (Bubic, Susac, & Palmovic, 2015; Zwaan, 2008；何先友 與林崇德，2008；冷因，2004；Kintsch &Van Dijk，1978). Therefore, the present study introduced English reading ability and chronological/ non-chronological sequence of events as two factors and tried to investigate whether they have impact on participants’ reading comprehension and interactive mental representation. Eye movement, alignment task, reading comprehension test and interview are designated to comprehensively estimate participants’ reading comprehension. Sixty-two 7th graders participated in this study. A 2 by 2 ANOVA was performed to analyze difference among groups. The results indicated English reading ability and sequence of chronological events partially affected readers’ mental representation and reading comprehension. Regarding temporal dimension, the data on “Single temporal marker-Text” revealed the initial temporal markers are essential for constructing situational model, additionally, as for non-chronological text”, more fixation counts and run counts are found in “Single temporal marker-Diagram. The fixations dramatically reduced after the reversed temporal markers. On the other hand, the fixations on chronological text gradually ascended which aligned to Zwaan’s Event-index model: events integrated via different dimensions by building blocks to update the situational model. Furthermore, on “Total temporal markers-Text”, good readers allocated their attention effectively on different chronological texts. Struggling readers didn’t get scores as expected even when they devote more attention (more fixation counts) on reading materials than good readers. Saccades and transcribes showed good readers are good at using diagrams to construct the meaning of text, while struggling readers reported the useless functions of diagram. Thus, the major finding of this study provided the beginning temporal markers are meaningful in construction of situational model that is influenced by English reading ability: good readers foster reading comprehension by diagram. Struggling readers should rely on teacher’s instructions of reading diagram to assist their reading comprehension.