Assessment and Improvement of Avatar-Based Learning System: From Linguistic Structure Alignment to Sentiment-Driven...

Area of Science:

• Educational Technology
• Human-Computer Interaction
• Artificial Intelligence in Education

Background:

• Traditional learning systems often lack engaging and personalized elements.
• Avatar-based learning systems offer potential for enhanced student engagement.
• The integration of emotional intelligence in educational avatars is an emerging area.

Purpose of the Study:

• To investigate the effectiveness of emotion-driven interactions in avatar-based learning systems.
• To compare the efficacy of different instructional strategies within an avatar learning environment.
• To identify factors influencing student engagement and learning outcomes with educational avatars.

Main Methods:

• Comparative analysis of various instructional approaches.
• One-way ANOVA for post-test result disparities.
• Tukey's HSD for pairwise group comparisons.
• Effect size evaluation for traditional versus avatar-based methods.

Main Results:

• One-way ANOVA revealed significant differences in post-test scores across teaching strategies.
• Pairwise comparisons using Tukey's HSD did not yield significant group differences.
• Effect size favored traditional methods over avatar-based approaches, with video lessons showing moderate distinctions.
• Students highlighted the need for emotional authenticity and cultural adaptation, such as a Kazakh accent.

Conclusions:

• Emotion-driven avatars have the potential to foster deeper learning experiences.
• Initial lower effectiveness of the avatar system may be due to novelty and adjustment time.
• System refinement for emotional authenticity and cultural relevance is crucial for widespread adoption.
• Further research with larger sample sizes is needed to strengthen statistical power.