Tang Xiaoqin, Jiang Li, Liu Guoli, Li Hongxia
Department of Educational Technology, Sichuan Normal University, Chengdu, China.
Baiyue Chenglong Primary School, Chengdu, China.
Front Psychol. 2025 Jul 28;16:1610550. doi: 10.3389/fpsyg.2025.1610550. eCollection 2025.
To address the challenge of face-to-face communication in online learning, integrating pedagogical agents and emotional feedback has been proposed as viable solutions. However, research on their impact during formative assessments remains limited.
This study therefore conducted a 2 (Pedagogical agent: present vs. absent) × 2 (Emotional feedback: present vs. absent) experimental study using an online learning system to explore their effects on learning performance.
Results indicated that pedagogical agents had a slightly negative influence on transfer scores, while emotional feedback significantly boosted engagement. When both were combined, learners exhibited the highest motivation, although this did not significantly enhance emotional perception or performance and slightly reduced transfer scores. Notably, the use of these tools shortened learning duration.
These findings suggest that educators should exercise caution when designing pedagogical agents in online formative assessment environments to avoid potential distractions during the learning process. Meanwhile, the integration of emotional feedback may contribute to creating a more humanized digital learning atmosphere, thereby supporting learners in their online learning experience. Overall, this study provides crucial insights into the complex effects of these tools on learning in computer-based online formative assessments, guiding future design and application.
为应对在线学习中面对面交流的挑战，有人提出将教学代理和情感反馈作为可行的解决方案。然而，关于它们在形成性评估中的影响的研究仍然有限。
因此，本研究使用在线学习系统进行了一项2（教学代理：存在与不存在）×2（情感反馈：存在与不存在）的实验研究，以探讨它们对学习成绩的影响。
结果表明，教学代理对迁移分数有轻微的负面影响，而情感反馈显著提高了参与度。当两者结合时，学习者表现出最高的动机，尽管这并没有显著增强情感感知或成绩，并且略微降低了迁移分数。值得注意的是，使用这些工具缩短了学习时间。
这些发现表明，教育工作者在在线形成性评估环境中设计教学代理时应谨慎，以避免学习过程中的潜在干扰。同时，情感反馈的整合可能有助于营造更人性化的数字学习氛围，从而支持学习者的在线学习体验。总体而言，本研究为这些工具在基于计算机的在线形成性评估中对学习的复杂影响提供了关键见解，指导未来的设计和应用。
