“隨後訪談”對作文的修正效應－依據LARP at SCU之語料調查

以中間語言之立場分析外國人的第二語言日語文習得過程研究近年來頗受重視，本研究LARP at SCU是以台灣東吳大學日文系學生為對象，擬採集就學期間四年之作文與訪談錄音之語料為研究對象，從記錄語料中可就音聲、語彙、文法、談話、溝通之各角度整理分析日語能力發展過程，並探討有效促進及提昇日語教學之具體方案建議。此次以92年9月入學新生於進入課程學習半年後進行每個月一次的語料採集，至93年7月共進行了五次的採集記錄。語料之採集與記錄。其順序為(l)寫第一次作文（題目統一，600字內）(2)朗讀作文（錄音）(3)隨後訪談（以回想法述說了解語言表達之過程構造，針對作文內容進行確認，提醒誤用處促進修正）(4)寫第二次作文（看第一次的原稿，寫出修正後之文章）(5)將兩次作文與文字化之後的聲音語料儲存電腦語料庫。過程中教師之「隨後訪談」對作文之修正效應如何，本稿以三學生二教師之五次語料進行研究。研究發現「隨後訪談」之回饋方法可分為三：(l)明示法(2)暗示法(3)暗示明示法。教師提議修正處有：(1)表記的錯誤(2)語彙選擇的錯誤(3)文法的錯誤(4)表達法的提昇。從修正率及往後維持期間看回饋方式則發現，暗示法高於明示法；修正率與誤用種類亦有關；本文針對五個月的語料記錄，由「隨後訪談」前後兩次作文之變化探討教師回饋法對作文之修正影響，並看短程與長程之效應。建議教師在寫作指導中針對「誤用」宜先以暗示法提醒，應為較有效指導模式。

關鍵字

隨後訪談；誤用；回饋法；維持期間

並列摘要

Using ”inter-language” to analyze the learning process of studying Japanese as second language has received increasing scholarly attention. This study-LARP at SCU is based on a four year writing study and its recorded fo1lox-up interviews from Japanese majors at Soochow University at Taiwan. The corpus includes aspects of pronunciation, vocabulary grammar, conversation and communication. It is used to analyze the students' Japanese development and to develop concrete and effective ways to improve Japanese teaching. The study is based on data from three students and two teachers who have conducted corpus study for five times. It starts from September 2003 after the first 6 months of freshmen entering the curriculum and the follow-up interview was conducted every month till July 2004. The procedures of the five interviews are the followings: 1) the first draft (assigned topic within 600 words): 2) reading out the draft (recorded): 3) follow-up interview (students were asked to recall how they composed the draft and the teacher hinted on the content and the mistakes): 4) second draft (based on the revision of the first draft): 5) saved the first and second drafts and the transcriptions of the interviews to the corpus. The results of the study show that there are three ways to give students feedback of their writings: 1) by pointing out their mistakes directly: 2) hinting on their mistakes: 3) combining the first and the second. The two instructors also suggest ways of corrections: 1) punctuation correction: 2) vocabulary correction: 3) grammatical correction: 4) improvement of the expression. The study also shows the use of hinting has higher frequency than pointing out mistakes directly and that the rates of corrections are also related to varieties of mistakes. It concludes that based on short term and long-term effects, the students will learn more effective when mistakes are hinted first.