生成式人工智能輔助教學對於學生創意自我效能影響──以大學生的美甲實務應用為例

The Impact of Generative Artificial Intelligence-Assisted Teaching on Students' Creative Self-Efficacy: A Case Study of Undergraduate Students in Nail Art Practice

近年生成式人工智慧（Generative Artificial Intelligence, GAI）快速發展，已成為高等教育與創意產業轉型的重要推力。根據McKinsey（2025）的最新報告，GAI已成為推動全球美業數位轉型的關鍵技術，特別是在加速美感概念開發上展現巨大潛力。然而，現有研究多聚焦於語文或程式學科，較少探討GAI介入技職體系之美業實務課程，如何影響學生在創意歷程中的心理信念。對藝術與設計相關科系而言，由個人信心演進至人機共創的「創意共創效能（Creative Co-Efficacy, CCE）」是支撐未來專業發展的關鍵。因此，本研究以北部某科技大學「AI在美業的應用」課程為場域，採準實驗一組前後測設計，探討GAI輔助教學對大學生創意自我效能之影響。研究對象為美容相關科系三年級學生，本研究聚焦於第一階段六週之密集教學介入，並於結束後六週施以延伸問卷。研究工具包括改編自Tierney與Farmer（2002）之量表，並輔以開放式問卷。
統計結果顯示，學生在參與GAI教學後，於「創意思考信念」與「創意成品信念」兩構面皆有高度顯著提升（p < .001）。其中，「創意思考信念」平均值由2.96提升至3.84（t = 8.957, p <.001），此數據證實了非STEM領域學生對於GAI提供的「創意支持（Creativity Support）」具有極強的正向感知；「創意成品信念」由2.96提升至3.59（t = 5.467, p < .001），顯示GAI教學能強化學生於共創歷程中的信心。此外，延伸問卷結果顯示學生持續維持高度正向態度，尤以「靈感激發」最為突出，反映出自我效能提升能有效驅動學生的「創新思維（Innovative Thinking）」。開放式問卷則指出，學生雖認同AI能提升效率，但仍面臨指令撰寫挑戰，顯示導入結構化共創流程（如HAI-CDP模型）之必要性。
綜合而言，本研究實證了GAI在美甲教育中具有促進創意思維、強化CCE與支持持續學習動機之價值。研究結果不僅補足了GAI於創意心理信念層面的實證缺口，也為技職教育與美業數位轉型人才培養提供了具體的教學與理論參考。

In recent years, generative artificial intelligence (GAI) has rapidly advanced and become a major driving force in the transformation of higher education and creative industries. According to a recent report by McKinsey (2025), GAI has emerged as a key technology enabling the digital transformation of the global beauty industry, particularly demonstrating substantial potential in accelerating the development of aesthetic concepts. However, existing research has predominantly focused on language learning and programming, with relatively few studies examining how the integration of GAI into vocational beauty practice courses affects students’psychological beliefs during the creative process. For students in art- and design-related programs, the evolution from individual confidence to human–AI cocreation—conceptualized as creative co-efficacy (CCE)—is crucial for sustaining future professional development. Therefore, this study adopted the course“AI Applications in the Beauty Industry”at a university of science and technology in northern Taiwan as its research context, and employed a one-group quasi-experimental pretest–posttest design to investigate the impact of GAI-assisted teaching on university students’creative self-efficacy. The participants were third-year students in a beauty-related department. The study focused on the first six-week intensive instructional phase and administered a follow-up questionnaire six weeks after the intervention. Research instruments included a scale adapted from Tierney and Farmer (2002), supplemented by an open-ended questionnaire.
Quantitative data were analyzed using descriptive statistics and paired-sample t-tests, while qualitative data were examined through thematic analysis. The statistical results revealed significant improvements in both dimensions of creative self-efficacy following participation in GAI-assisted instruction (p < .001). Specifically, the mean score for“creative thinking beliefs”increased from 2.96 to 3.84 (t = 8.957, p < .001), indicating enhanced capability in innovative idea generation and problem solving; meanwhile,“creative product beliefs”increased from 2.96 to 3.59 (t = 5.467, p < .001), suggesting that GAI-supported instruction strengthened students’confidence and performance throughout the creative process. In addition, the follow-up survey conducted six weeks later (M = 3.93, SD = 0.51, N = 17) demonstrated that students continued to hold highly positive attitudes toward GAI-assisted learning, with particularly strong responses in inspiration enhancement, course pacing appropriateness, and intention for future use. Findings from the open-ended responses further indicated that students generally perceived AI as beneficial for improving learning efficiency and stimulating creative inspiration, while challenges remained in terms of prompt precision and output stability.
Overall, this study confirms the potential value of generative artificial intelligence in promoting creative thinking, strengthening creative self-efficacy, and supporting sustained learning motivation in nail art education and other art and design contexts. The findings not only help bridge empirical gaps regarding the creative and psychological impacts of GAI but also provide concrete implications for curriculum design in vocational education and interdisciplinary creative learning. Future research may incorporate control groups, behavioral data, and long-term follow-up to deepen understanding of human–AI co-creative learning models.