探討ChatGPT輔助翻轉教學中技能學習者的學習清晰度與學習焦慮經驗

Abstract

本研究旨在探討 ChatGPT 輔助翻轉教學情境中，學生的學習清晰度與學習焦慮經驗。研究場域為某國立大學開設之「木器設計與製造」課程，採行動研究法，實施為期一學期之教學介入。研究對象為13位修課學生，並從中依其於前導課程之學習表現與學習焦慮程度挑選7位進行半結構式訪談。研究工具包含半結構式訪談大綱、課後反思與回饋表單、自編學習焦慮量表與課堂觀察週誌。資料處理方面，採內容分析進行質性資料分析，並輔以描述性統計作初步量化分析。研究發現包括：(1)教師藉由ChatGPT協助備課與教材優化，提升了課程內容的結構清晰度與學習活動說明的明確性，多數學生認為課程目標、任務流程及評量標準更加易於理解；(2)學生在課程中所經歷的學習焦慮，主要來自進度與時間壓力、評量標準不明與操作技能的不確定感，當教師提供即時回饋、同儕合作與正向鼓勵時，能有效舒緩學生焦慮；以及(3)學習清晰度與學習焦慮呈現高度負相關，課程說明與結構越清晰，學生壓力越能降低，反之亦然。質性資料亦揭露學生對課程的具體建議，包括加強材料利用限制與接合方式說明、設置實作進度追蹤工具，以及將示範教學內容分段以利實作跟進。

This study explores students’ experiences of learning clarity and learning anxiety in a ChatGPT-assisted flipped classroom. The research was conducted in a “Woodworking Design and Manufacturing” course at a national university in Taiwan, using action research over one semester. Thirteen students participated; seven were selected for semi-structured interviews based on their learning performance and level of learning anxiety in the prerequisite course. Data sources included semi-structured interview outlines, post-class reflection and feedback forms, a self-developed learning anxiety scale, and classroom observation journals. Qualitative data were analyzed using content analysis, supplemented by descriptive statistics for preliminary quantitative analysis. Results indicate that ChatGPT-supported lesson preparation and material optimization enhanced the structural clarity of course content and instructional explanations. Most students reported improved understanding of course objectives, tasks, and assessment criteria. Learning anxiety was primarily related to progress and time pressure, unclear assessment, and skill uncertainty; timely teacher feedback and peer support effectively alleviated anxiety. There was a strong association between learning clarity and reduced anxiety. Students suggested further improving explanations of materials and joining methods, providing progress tracking tools, and segmenting demonstrations to support practice.