自2000年代以來，網路技術進步和拍攝設備普及，降低了創作影片的門檻，每日有大量教學影片被上傳到 YouTube 平台，引發出透過 YouTube 教學影片進行自主學習的新方式。YouTube 教學影片資源雖豐富多元，但存在品質參差不齊、缺乏統一製作標準等問題需要解決。本研究之研究者同時兼具華語文教學頻道經營者、教學影片主講人與影片製作者等多重身份，並具備華語文學科知識基礎。能透過第一手資料，從製作者和學習者角度深入解析 YouTube 華語文教學影片的製作程序及學習者感想，並且在此基礎上建立一套華語文教學影片的模型。
本研究運用質性方法，檢驗研究者製作的4部教學影片，是否符合Mayer多媒體教材設計原則。通過分析觀眾留言與訪談資料，作為學習者端的驗證反饋。研究結果發現，這套華語文教學影片設計模型製作出的影片符合Mayer的多媒體教材設計原則。在使用者端的反饋，學習者能感到教材内容的自然和真實性；影片能夠模擬真實情境，讓學習者在此情境中培養聽力，適應和理解真實世界的華語交流；四階段的設計模式（原速無字幕→原速加詞彙→慢速加字幕→原速加字幕）符合學習者的認知規律，有助於促進其自主學習，同時避免了字幕（漢字）給學習帶來的干擾；學習者主動將影片應用於影子跟讀法（Shadowing）進行自主學習並認為非常合適。
本研究結果不僅可作為研究者精進自身教學影片製作之依據，製作更優質的影片提供學習者觀看，也可提供華語文教學領域發展參考，為有意在網路上開展華語文教學自媒體的人士提出具體的影片製作建議。
Since 2000, advancements in internet technology and the lowered barriers to video production have led to a daily influx of videos uploaded to the YouTube platform, sparking a new trend in self-directed learning through YouTube educational videos. While YouTube offers a wealth of diverse and abundant educational content, issues such as inconsistent quality and lack of uniform production standards persist. The researcher of this study, possessing multiple roles as a Mandarin Chinese teaching channel operator, video lecturer, and producer, as well as a foundational knowledge in Mandarin Chinese language studies, provides an in-depth analysis from both the producer's and learners' perspectives through first-hand data. This analysis encompasses the production process of YouTube Mandarin teaching videos and the learners' impressions, aiming to establish a model for Mandarin teaching videos.
Employing qualitative methods, this study examines whether four teaching videos produced by the researcher align with Mayer's principles of multimedia material design. Audience comments and interview data are analyzed for learner feedback and validation. The results show that the Mandarin teaching video design model aligns with Mayer's multimedia material design principles. From the learners' feedback, the content of the teaching materials is perceived as natural and authentic. The videos simulate real-life scenarios, aiding learners in developing listening skills and adapting to and understanding real-world Mandarin communication. The four-stage design model (original speed without subtitles → original speed with vocabulary → slow speed with subtitles → original speed with subtitles) aligns with learners' cognitive patterns, facilitating self-directed learning while avoiding the interference of subtitles (Chinese characters). Learners actively apply the videos in shadowing exercises for self-study and find it highly suitable.
The results of this study not only serve as a basis for the researcher to refine their video production, offering higher quality videos for learners, but also provide references for the development of Mandarin teaching in the digital realm. This study offers concrete video production recommendations for individuals interested in launching Mandarin teaching channels online.
