Correction: Humanoid robot-empowered language learning based on self-determination theory

Ka Yan Fung, Lik Hang Lee, Kuen Fung Sin, Shenghui Song, Huamin Qu

Introduction

While educational robotics has demonstrated significant potential in STEM education, its application in social and emotional language support remains under-explored. Traditional classroom environments often struggle to provide the personalized, immediate feedback necessary for sustained student engagement, particularly in mother-tongue acquisition. This study investigates how humanoid robots equipped with social and affective features—such as gestures, mobility, and facial expressions—can enhance engagement and proficiency in Traditional Chinese among primary school students.

Methodology

The researchers conducted a five-day pilot study involving 64 primary school students in Hong Kong. The cohort was divided into a control group (teacher-led instruction) and an experimental group (robot-led interaction).

Hardware and Interaction Design

The study utilized the Kebbi Air S, a humanoid robot featuring seven motors for upper-body movement and wheels for mobility. A custom Android application was developed to facilitate interaction, incorporating Automatic Speech Recognition (ASR) for Cantonese. The interaction design included 125 unique combinations of gestures, facial expressions, and motivational prompts triggered by student performance.

Theoretical Framework

The pedagogical approach was grounded in Self-Determination Theory (SDT), focusing on three psychological needs:

  • Autonomy: Facilitated through self-paced learning modules.
  • Competence: Supported by instant, non-judgmental feedback via the ASR system.
  • Relatedness: Fostered through the robot's social presence and empathetic gestures.

Key Results and Analysis

The experimental group demonstrated statistically significant improvements across all primary metrics compared to the control group. The most substantial gain was observed in cognitive engagement, which rose by 21.56%. Behavioral and emotional engagement increased by 13.24% and 13.14%, respectively, while intrinsic motivation saw a 12.07% uplift.

Comparison of engagement and motivation metrics between Control and Experimental groups.

Beyond engagement metrics, the robot-led group achieved a 9.88% improvement in language proficiency test scores (p < .001) within the five-day window. In contrast, the control group exhibited stagnant or declining engagement levels over the same period.

Discussion: The "Warm Companion" Effect

The study highlights that the physical presence and social capabilities of the robot transformed it from a mere tool into a social partner. Qualitative observations noted students hugging and touching the robot, suggesting a high level of "relatedness" rarely achieved by screen-based learning interfaces. Furthermore, the robot provided a "safe" learning environment; students felt less fear of judgment when making mistakes compared to a traditional classroom setting, which encouraged active participation from lower-achieving students.

Limitations and Future Directions

Despite the positive outcomes, the study's five-day duration leaves questions regarding the long-term "novelty effect" of the technology. Additionally, the current software follows a pre-programmed path. Future research will explore the integration of Generative AI to allow for more fluid, personalized dialogues and expand the scope to include special education contexts, such as students with dyslexia or autism.

Conclusion

By grounding robotic interaction in Self-Determination Theory, this research demonstrates that "warm" humanoid companions can significantly outperform traditional methods in fostering intrinsic motivation and cognitive engagement. The findings suggest that social support and physical presence are as critical as content delivery in effective language learning technology.

Summary updated

BibTeX
@article{fung2025correction,
  author = {Ka Yan Fung and Lik Hang Lee and Kuen Fung Sin and Shenghui Song and Huamin Qu},
  title = {{Correction: Humanoid robot-empowered language learning based on self-determination theory}},
  journal = {Educ. Inf. Technol.},
  year = {2025},
  doi = {10.1007/s10639-024-13054-7},
}
Figure