生成式AI時代傳播領域大學生媒體素養建構與教學策略研究

Constructing Generative AI Literacy Among Communication Students: Pedagogical Strategies in Higher Education

本研究旨在探討生成式人工智慧（Generative AI）情境下，大學生媒體素養之樣態與教師教學策略的轉型方向。研究以10位傳播領域在職大學教師為研究參與者，採用紮根理論（grounded theory）進行質性分析。透過MAXQDA進行三階段編碼，共萃取36個初級概念，整合為22個能力細項，進而歸納出「理解與分析」、「應用與倫理」與「參與與反思」三大構面之AI時代媒體素養架構。研究發現，大學生在應然目標與實然表現之間存在六項落差，分別為資訊判斷、技術理解、應用操作、倫理責任、公共參與及反思批判。此結果揭示學生普遍缺乏對生成式AI的批判性理解與倫理覺知，反映出高等教育課程在技術應用、倫理治理與跨模態整合上的不足。教師實務策略顯示，第一線教師已發展出具體回應模式，包括多平台驗證、生成內容解構、風險情境模擬、提示工程訓練、跨模態專案、倫理案例研討、跨域協作與反思日誌機制等。這些策略兼具批判、創造與公民導向，有助於將理論素養轉化為可教、可評、可迭代的課程設計。本研究理論上整合媒體與資訊素養（Media and Information Literacy, MIL）、數位素養與AI素養三大脈絡，提出具操作性的「三構面—六落差」模型，補足既有理論對生成內容真實性與倫理風險的不足，並提出「AI時代下媒體素養三模組架構」：風險與查證、生成與創作及倫理與參與，以支援課程改革與政策推動。

This study explored the transformation of media literacy and instructional strategies in higher education under the influence of generative artificial intelligence (Generative AI). Using grounded theory, the research investigated ten in-service university teachers in the field of communication to construct an empirical framework for AI-era media literacy. Data were analyzed through three stages of coding using MAXQDA, extracting 36 initial concepts, refining them into 22 sub-abilities, and integrating them into three major dimensions: Understanding and Analysis, Application and Ethics, and Participation and Reflection. The findings revealed six significant gaps between the ideal (expected) and the actual (observed) performance of university students: information judgment, technical understanding, operational application, ethical responsibility, public participation, and reflective criticism. These systemic gaps indicated students’ insufficient critical awareness and ethical sensitivity toward AI-generated content, as well as structural weaknesses in higher education curricula regarding technological integration, ethical governance, and multimodal composition. Practicing teachers had developed a series of responsive strategies, including multi-platform verification, generative content deconstruction, risk-scenario simulation, prompt-engineering training, cross-modal projects, ethical case discussions, interdisciplinary collaboration, and reflective journaling. These approaches combined critical, creative, and civic orientations, transforming theoretical literacy into teachable, assessable, and iterative pedagogical practices. Theoretically, this study integrated Media and Information Literacy (MIL), digital literacy, and AI literacy frameworks to propose a practical Three-Dimension-cum-Six-Gap Model, addressing the limitations of previous studies concerning authenticity and ethical risks in generative contexts. Furthermore, it advanced an “AI Media Literacy Tri-Module Framework”– Risk and Verification, Generation and Creation, and Ethics and Participation– as a basis for curriculum innovation and educational policy development in higher education.