Design pedagogical agents with facial expressions and consider learner-agent similarity to target motivation¶

Strategy · All strategies
Evidence · 4 claims (3 for, 1 mixed) · 2 studies (1 quant-synthesis, 1 review), q2–q3 · 1 of 2 report an effect size · 4 claims rest on one study

Description¶

The meta-analysis's moderator results support concrete PA design choices: agents that used facial expressions significantly improved broad motivation (g = 0.33) and self-efficacy expectations (g = 0.46), whereas agents without facial expressions had no significant effect on either outcome. For value beliefs, the authors recommend that "if a PA researcher is interested in impacting learners' value beliefs, careful consideration must be made regarding the agent's role and whether the learner will perceive some degree of similarity (e.g., demographic and/or psychological) between themself and the PA." They speculate such designs may significantly impact value beliefs.

Design Implications¶

Context¶

Requirements¶

• The PA must be a visible on-screen virtual character compared against a non-PA condition for effects to be assessed

Constraints¶

• The facial-expression moderator results come from observational moderator analyses, not randomized manipulation of facial expressions; the value-belief similarity recommendation is speculative and untested in this synthesis

Target Learners¶

• learners in digital learning environments across grade levels, including adult learners outside postsecondary education

Target Learning Goals¶

• self-efficacy expectations, broad motivation, value beliefs

Claims¶

• Pedagogical agents significantly improve learners' self-efficacy expectations (g = 0.29), with effects moderated by grade level, facial expressions, and reliability reporting [+W]
• Pedagogical agents have a small, statistically significant positive effect on learners' broadly defined 'motivation' (g = 0.20) [+W]
• Effects of agent nonverbal behavior depend on the task: facial expressions were valuable for attitudinal learning but detrimental for procedural learning [~W]
• Pedagogical agents significantly improve learners' interest (g = 0.30), with the effect depending on the type of control condition [+W]

Related Strategies¶

Examples¶

-

Key Sources¶

• Gladstone, J. R., Schroeder, N. L., Heidig, S., Zhang, S., Palaguachi, C., & Pitera, M. (2025). Do Pedagogical Agents Enhance Student Motivation? Unraveling the Evidence Through Meta-Analysis. Educational Psychology Review. https://doi.org/10.1007/s10648-025-10050-2