pedagogical-agents
Filtering by topic pedagogical-agents(2)Clear all filters
- PaperComputers & Education27 Jun 2026
Operationalizing cognitive apprenticeship through generative AI pedagogical agents: Proactive and contextualized guidance in simulation-based learning
Po-Yao Chao, Yu-Chen Cheng, Kuan-Chieh Chen, Ying-Siou Chen et al.
This study operationalizes cognitive apprenticeship theory by using generative AI pedagogical agents to provide proactive and contextualized guidance in simulation-based learning environments.
- PaperBritish Journal of Educational Technology26 Mar 2026
The synergy of pedagogical agents and metaphorical design: Reducing psychological distance to enhance video learning
Ke Xu, Qihui Zhang, Yi Zhang, Hongmin Xu et al.
A 2x2 experiment with 129 learners showed that combining virtual pedagogical agents with metaphorical content design improved learning performance and reduced psychological distance compared to real agents or non-metaphorical content. Metaphorical design increased cognitive effort and perceived human-likeness of agents. The findings suggest instructional designers should integrate metaphorical content with agent design to enhance video-based learning.
Original abstract
The effectiveness of video‐based learning lies not in perfecting isolated elements such as pedagogical agents (PAs) or metaphorical content, but in their synergistic integration. While both have been shown to independently enhance learning outcomes, their combined potential remains underexplored. In this study, 129 learners engaged with biology content in a 2 × 2 between‐subjects design to examine the effects of PAs type (real vs. virtual) and metaphorical design (no metaphor vs. metaphor) on learners' performance, attention, comprehension, psychological distance and agent evaluation and the mediating role of psychological distance. Metaphorical design used familiar concepts to make abstract biological content more relatable, whereas the no‐metaphor condition presented the content in its abstract form. Results showed that both virtual pedagogical agents (VPAs) and metaphorical design improved learning performance, as they both reduced learners' attention allocation to the agent. Metaphorical design specifically increased cognitive effort, as indicated by larger pupil size, while also reducing spatial distance and enhancing VPAs' perceived human‐likeness. Additionally, VPAs reduced hypothetical distance. Spatial and hypothetical distances, as well as learning comprehension, mediated the effects of the two factors on learning performance. Temporal distance mediated the interaction effect of PAs type and metaphorical design on credibility. The study concludes that both VPAs and metaphorical design enhance learning effectiveness and reduce psychological distance. Metaphorical design compensates for VPAs' lack of realism by boosting their perceived human‐likeness and credibility. These findings suggest that instructional designers should prioritize not only the visual realism of VPAs but also metaphor‐based content design to create more immersive, effective learning experiences. Practitioner notes What is already known about this topic Video‐based learning typically involves either real‐human instructors or virtual pedagogical agents (VPAs). VPAs are increasingly used because they reduce distraction and allow scalable, consistent delivery; though they provide lower social presence. Research has mainly focused on agent visual design, while their pedagogical integration with instructional content has received far less attention. Metaphorical design supports learning in videos, yet little is known about its combined effects with agents or the mechanisms involved. What this paper adds Learners taught by VPAs or with metaphorical content performed significantly better than those taught by real‐human teachers or with non‐metaphorical content. Metaphorical design improved content‐focused cognitive effort, reduced perceived spatial distance and enhanced the human‐likeness of VPAs, demonstrating how representational strategies shape cognition and perception. Spatial and hypothetical distance and comprehension mediated the effects of agent type and metaphor on performance, while temporal distance mediated their interactive effect on credibility, identifying psychological distance as a key explanatory mechanism. Implications for practice and/or policy Combining metaphorical content design with VPA appearance can more effectively allocate learners' attention and enhance performance. Educators and designers should avoid overemphasizing agent visual realism and instead give equal weight to instructional content design, particularly metaphorical structuring.