| 題名: | 課業壓力與AI依賴:當學習遇上人工智慧 |
| 其他題名: | Academic Stress and AI Dependence: When Learning Meets Artificial Intelligence |
| 作者: | 林慈真、蔡一齊 |
| 關鍵字: | 生成式人工智慧 跨領域學習 抑制效應 課業壓力 Academic Stress Generative Artificial Intelligence (Gen AI) Interdisciplinary Learning Suppressor Effect |
| 系所/單位: | 國際經營與貿易學系,會計學系, 商學院 |
| 摘要: | 中文摘要 隨著生成式AI工具的普及,其在高等教育中已成為學生應對繁重課業與緊湊 截止期限的重要輔助。然而,既有研究多聚焦於 AI 的普及率描述或單一心理機 制的建構,較少區辨「作業量(量)」與「知識焦慮(質)」 對依賴行為的不同影 響。 本研究旨在填補此文獻缺口,探討台灣大學生在不同壓力源下的AI使用行 為。本研究採橫斷面問卷調查法,獲得107份有效樣本,並透過多元迴歸分析等 方式檢測變數間之關係。敘述統計顯示,高達97 %的受訪學生曾使用AI,且以生 成草稿後再自行修正的「中度依賴」為主流模式(54.2 %)。 研究最關鍵的發現為抑制效應(Suppressor Effect)之驗證。在單變數分析中, 跨院修課與主觀壓力對AI依賴的預測力均未達顯著;然而,當將兩者同時納入模 型後,跨院修課之預測係數顯著提升,顯示主觀壓力扮演了抑制變數,掩蓋了真 實的解釋路徑。結果證實:在控制主觀壓力後,「跨院修課」是驅動 AI 依賴的核 心主因。說明AI從大眾認知中的代工工具轉化為協助學生跨越跨領域學習門檻的 學習鷹架(Scaffolding),成為學生整合跨領域知識、緩解知識焦慮的家教。Abstract As Generative AI (Gen AI) becomes increasingly ubiquitous, it has emerged as a vital tool for university students managing heavy academic workloads and demanding deadlines. However, existing research primarily focuses on adoption rates or singular psychological mechanisms, often overlooking the distinct impacts of "workload quantity" versus "intellectual anxiety" on AI-reliant behaviors. This study addresses this gap by investigating AI usage patterns among Taiwanese undergraduates under varying stressors. Utilizing a cross-sectional survey design (N=107), variables were analyzed through multiple regression and descriptive statistics. Findings reveal a near-universal adoption rate (97 %), with “Moderate Reliance”, defined as revising AI-generated drafts, emerging as the dominant modality (54.2 %). The most critical contribution of this study is the empirical validation of a Suppressor Effect. While univariate analyses showed no significant correlation between "Interdisciplinary Coursework" or "Subjective Stress" and AI reliance, the predictive power of Interdisciplinary Coursework increased significantly once both variables were integrated into the model. This indicates that subjective stress acts as a suppressor, masking the underlying relationship. The results demonstrate that, when controlling for general stress, interdisciplinary enrollment is the primary driver of AI dependence. This suggests that AI has transitioned from a mere outsourcing tool to a sophisticated academic scaffold, serving as a conceptual bridge that helps students integrate cross-disciplinary knowledge and alleviate intellectual anxiety. |
| 學年度: | 114學年度第一學期 |
| 開課老師: | 商, 瓈丹 |
| 課程名稱: | 永續決策與數據分析 |
| 系所: | 國際經營與貿易學系,商學院 |
| 分類: | 商114學年度 |
文件中的檔案:
| 檔案 | 描述 | 大小 | 格式 | |
|---|---|---|---|---|
| 1141-24.pdf | 2.73 MB | Adobe PDF | 檢視/開啟 |
在 DSpace 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。