本研究自第二語言習得觀點,探討將整合語音辨識技術之繪本融入國小平衡閱讀教學,如何促進學生英語字彙習得。目前繪本應用模式傾向採用多媒體情境脈絡,發展學生意符階段的識字能力,提升閱讀理解;而不易支援學生經形音對應、拼讀解碼,促進識字自動化。為延伸現有繪本的應用模式,協助學生在平衡閱讀教學中,能取得意義方法與解碼策略之平衡,本研究考量各學派字彙教學觀點,於行動平台發展iSpeaking繪本應用程式。透過整合基於語音辨識之字母拼讀於繪本中,並輔以結構化代幣回饋,達到資料導向拼讀及反覆發音調適。 為探究可行性,研究團隊與教師合作,將iSpeaking融入平衡閱讀教學。研究對象為台灣北部一所國小六年級其中一班的29位學生,並以小組個別協助教學進行分組。本研究採前實驗設計,以字彙試題、態度量表、課間觀察、師生訪談等工具蒐集資料,最後依研究問題,包含學習成效、學習態度、互動模式,及系統功能接受度進行分析討論。 研究結果顯示,iSpeaking融入平衡閱讀教學對高低成就學生在聽音拼字、字彙填空、字彙應用方面皆有助益,而低成就者進步幅度具統計顯著提升;在學習態度方面,學生能依據語音辨識整合式回饋反覆調適發音,並投入字母拼讀。教師認同高低成就分組有助於同儕互助,進而成為學生身旁的促助者。學生在繪本情境脈絡支援下,強化語音迴路,習得繪本詞彙,本研究進一步歸納六個實現交互識字教學的成功元素-R.E.C.I.T.E,分別為R:反覆的拼讀練習機會、E:充分時間驗證個別學習困難、C:繪本情境脈絡支援、I:立即發音練習回饋、T:代幣增強制度、E:平等參與及社會化合作。本研究最後提出以下結論: 1.iSpeaking所實現的交互識字模式,能協助學生以情境脈絡作為認知線 索,進而透過字母拼讀強化學生語音迴路,習得字彙及其應用。 2.ASR回饋整合代幣增強能作為CSCL參考指標,提升學生拼讀投入。 3.透過ASR整合式回饋、同儕互助、教師協助三層鷹架,有助於學生字 彙建構,並提升低成就學生之學習效能。 4.學生及合作教師皆認同iSpeaking 可作為正規課程外的字彙補充教材。 本研究希望iSpeaking能應用於更廣泛的平衡閱讀教學情境,協助學生透過意義脈絡習得字彙拼讀,並提升字彙量,達到科技輔助語言學習之目的。
This study analyzed teaching advantages of integrating picture books into balanced reading instruction in terms of second language acquisition (SLA). Current picture books could support learners to acquire new words by context cues in the logoagraphic phase, but were unable to assist learners in acquiring new words by phonics in the alphabetic phase. This study considered different perspective of vocabualary learning strategies, and developed an ASR-based e-picture book on the iOS platform to enhance current picture books. Furthermore, in order to verify its feasibility and usability, we cooperated with an English teacher and integrated the developed app, iSpeaking, into the balanced reading instruction to conduct field experiments. One class of 29 sixth grade learners in a certain elementary school of northern Taiwan participated in this study. Learners adopted the iSpeaking app as learning media under the English teacher’s instruction and were grouped by team-assisted individualization (TAI). Learners’ learning performance, attitude and interaction processes were collected and analyzed. The research findings showed that integrating iSpeaking app into balanced reading instruction could improve heterogeneous learners’ vocabulary dictation, retention and applications, especially for underachievers. In addition, adopting the iSpeaking app as learning media had positive impacts on learning motivation. Learners engaged in adapting their pronunciations according to ASR feedbacks, which integrate token reinforcement. Teacher also agreed that adopting TAI model could convert her role to become a facilitator beside learners. In contrast to traditional picture books, the iSpeaking app could help learners acquire vocabulary through the enhancement of phonological loop under context cues. This study points out six success elements of teaching and learning vocabulary through picture books - R.E.C.I.T.E. R: Repetitive phonics learning opportunities. E: Enough time to verify individual learning difficulties. C: Context cues of picture books. I: Immediate pronunciations feedbacks. T: Token reinforcement mechanism. E: Equal participation and socialized collaboration. Based on the data collection and analysis, this study reaches the following conclusions: 1.The interactive word learning model reached by iSpeaking app can assist learners in adopting contexts as cognitive cues, and reciting vocabulary through phonics based on context cues. 2.The ASR feedbacks integrated the structure token could be regarded as referential indexes of CSCL and promote members to optimize their pronunciations repetitively. 3.The feedbacks of ASR, peer-to-peer instruction and teacher’s assistance could build up a three-tier scaffolding to help learners learn efficiently. 4.Both teacher and learners agreed that iSpeaking app could be used as supplementary materials for teaching vocabulary. We hope that the iSpeaking app will be applied in more extensive learning contexts to assist teachers in constructing low-anxiety and joyful learning environments, and improve leaners’ vocabulary learning strategies by enhancing their phonological loop under the support of context cues.