Educational robots in learning environments: a systematic review of psychological and pedagogical impacts

1. Introduction

The integration of educational robots (ERs) into learning environments has accelerated with advances in AI and robotics [1, p. 5874-5898]. Unlike traditional digital tools, ERs are embodied systems enabling physical interaction, real-time feedback, and social engagement [2, p. 104264], acting as both technological instruments and social agents that shape learners’ psychological states [3, p. 13-39]. ER applications span language learning, early childhood education, and special needs support, yet existing research often focuses on isolated outcomes. This review addresses this gap by: (1) analyzing ER effects on core learning variables; (2) identifying critical design determinants; (3) highlighting implementation challenges and future directions.

2. Theoretical Background

Two foundational frameworks explain ER impacts:

• Cognitive Load Theory: ERs reduce extraneous cognitive load through multimodal input and scaffolded guidance, optimizing knowledge construction [4, p. 261-280].
• Social Interaction Theories: ERs’ social presence enhances motivation and engagement by fostering meaningful learner-robot interaction [3, p. 13-39].

3. Methodology

Relevant studies were retrieved using keywords: "educational robots", "robot-assisted learning", and "assistive robots for special education", limited to English peer-reviewed papers. Included studies focused on physical ERs in educational settings and measured psychological/pedagogical outcomes. Excluded virtual robots, technical-only studies, and non-peer-reviewed sources. A total of 25 studies were selected for synthesis, categorized by ER type and core outcomes.

4. Categories of Educational Robots

4.1. Social Robots

Designed for verbal/non-verbal interaction, social robots serve as learning companions in early childhood education and language learning [5, p. 567-589]. They reduce anxiety and enhance participation by providing low-pressure practice and emotional support.

4.2. Teaching Assistant Robots

Support instructional activities by delivering content, automating repetitive tasks, and offering personalized feedback [6, p. 67]. Applied in STEM education and large-class settings, they improve learning efficiency and reduce teacher workload.

4.3. Assistive Learning Robots

Specialized for learners with special needs (e.g., autism, ADHD), these robots provide tailored, low-stimulation interaction to build social and academic skills [7, p. 689-701], acting as a "social bridge" for vulnerable groups.

5. Key Impacts and Design Determinants

5.1. Psychological Impacts:

• Motivation: ERs enhance intrinsic motivation by satisfying autonomy, competence, and relatedness needs [3, p. 13-39], with strongest effects in early childhood and special education.
• Engagement: Physical interaction fosters behavioral, emotional, and cognitive engagement, reducing passive learning [2, p. 104264].
• Emotional Response: ERs evoke curiosity and enjoyment while mitigating anxiety, critical for sustained learning [7, p. 689-701].

5.2. Pedagogical Impacts:

• Cognitive Load Management: Multimodal delivery and scaffolded guidance reduce extraneous load, freeing resources for knowledge construction [4, p. 261-280].
• Performance: ERs improve academic outcomes across contexts, with notable benefits for special needs learners [7, p. 689-701].

5.3. Critical Design Factors

ER effectiveness depends on: (1) interactivity (active vs. passive interaction); (2) adaptability (real-time adjustment to learner progress); (3) pedagogical alignment (design matched to learning goals) [6, p. 67].

6. Challenges and Future Directions

6.1. Key Challenges:

• Cost Barriers: High development and maintenance costs limit accessibility for low-resource institutions [5, p. 567-589].
• Technical Limitations: Limited emotion recognition and adaptive capabilities hinder personalized interaction [6, p. 67].
• Evaluation Gaps: Lack of standardized metrics complicates cross-study comparison [8].

6.2. Future Research:

• Prioritize personalized ERs tailored to learner characteristics (age, learning style, special needs) [6, p. 67].
• Integrate emotional intelligence for nuanced learner response [7, p. 689-701].
• Conduct longitudinal studies on sustained ER impacts in real-world settings [8].

7. Conclusion

Educational robots offer significant potential to enhance learning through embodied interaction and social engagement. Their ability to boost motivation, reduce cognitive load, and support diverse learners depends on evidence-based design centered on interactivity, adaptability, and pedagogical alignment. Addressing cost and technical challenges through interdisciplinary collaboration will unlock ERs’ full potential as accessible, effective tools for inclusive education.