結合感應式卡牌與數位遊戲之虛實融合機器人同伴系統對國小學生漢字學習影響

Abstract

在過去，教育機器人多被應用於不同學科的學習，或扮演多樣化的同伴角色，仰賴教學者預先設計單一情境，根據學生的行為觸發相應的回饋，在自主互動與感知能力較為侷限。然而，真實的學習情境是動態變化的，外在情境會隨著周圍環境狀態而改變，內在情境則受到學習過程、互動與介入方式等因素影響，學生的能力會隨著環境變化，並在情境中發展，透過適當的引導或同儕合作，能激發潛在的學習能力。因此，教育機器人不僅需要理解學生的學習行為，更應該能夠感知並詮釋不同的學習情境，進而調整互動策略與回應方式，以提供更適當且即時的學習支持。本研究將數實融合理念應用於機器人同伴系統，透過結合機器人、數位遊戲與實體遊戲，提升機器人的環境感知與情境感知能力，使機器人不僅能辨識學生的學習行為與表現，更能在多樣化的遊戲情境中輔助學習，以探討虛實融合的機器人同伴系統，能否提升國小學生漢字學習之成效、學習動機與機器人感受度。本研究採準實驗設計法，以台灣北部國小71位學生為實驗對象，實驗組（實體組）使用結合感應式卡牌與數位遊戲之實體機器人同伴系統，具備三維空間的感知能力；對照組一（虛擬組）使用數位遊戲與虛擬機器人進行學習，局限於二維空間的感知能力；對照組二（紙本組）以紙本機器人學習單進行學習，不具感知與互動。

研究結果顯示，(一)在學習成效方面，實體組顯著優於虛擬組與紙本組，顯示機器人感知能力的提升與虛實整合學習環境，有助於促進學習表現；(二)在學習動機方面，實體組顯著高於紙本組，顯示具備即時感知與互動回饋的機器人系統，有助於吸引學習者注意力，強化學習內容與個人目標的關聯性；(三)在機器人感受度方面，實體組顯著高於虛擬組與紙本組，顯示機器人透過具體外觀、多模態感知與即時互動，讓學習者感受到更真實、具備情感連結且具智力的學習同伴。本研究具體貢獻在於：(一)開發整合感應式卡牌與數位遊戲之虛實融合機器人同伴系統，提升教育機器人的感知能力；(二)證實機器人感知能力的提升能有效增強機器人感受度，並可能進一步影響學習動機與學習成效；(三)應用數實融合設計，結合數位與實體遊戲元素，建構跨越虛擬與實體的沉浸式學習情境，為未來研究提供具體設計與實證參考。In the past, educational robots have primarily been applied to support learning across various disciplines or to serve diverse companion roles, typically relying on educators to pre-design single, fixed scenarios and trigger feedback based on students’ behaviors. These systems often exhibit limited autonomy in interaction and perceptual capabilities. However, learning contexts are continuously evolving: external contexts change according to the surrounding environment, while internal contexts—shaped by the learning process, interactions, and interventions—influence learners’ experiences. Learners’ abilities develop within these changing contexts and can reach their zone of proximal development through guidance or peer collaboration. Therefore, robots need to accurately interpret the social environment to respond appropriately. Beyond understanding learners’ behaviors, robots should also interpret varying learning contexts, adjust their interaction strategies accordingly, and provide personalized learning support.This study applies the phygital concept to a robotic companion system by integrating robots, digital games, and physical games to enhance the robot’s environmental and contextual awareness. This enables the robot to recognize students’ learning behaviors and performance, thereby offering personalized support across diverse contexts. Using a quasi-experimental design, the study involved 71 elementary students from northern Taiwan. The experimental group used a robotic companion system that combined NFC-based cards with digital games; the virtual group learned with a virtual robot game; and the paper-based group used robot-themed worksheets for learning. Results indicate that (1) in terms of learning performance, the physical group significantly outperformed both the virtual and paper-based groups, indicating that enhanced robotic perception and a phygital learning environment contribute to improved academic outcomes; (2) regarding learning motivation, the physical group scored significantly higher than the paper-based group, suggesting that a robot system equipped with real-time perception and interactive feedback can effectively capture learners’ attention and reinforce the relevance of learning content to personal goals; (3) in terms of perceived robot companionship, the physical group scored significantly higher than both the virtual and paper-based groups, demonstrating that through its tangible appearance, multimodal perception, and real-time interaction, the robot was perceived as a more authentic, emotionally connected, and intelligent learning companion.