Emotion-Centric AI Interventions for Experiential Learning in Higher Education

Soft skills such as communication, teamwork, adaptability, creativity, and emotional intelligence are increasingly vital in AI-dominated professional environments. This study evaluates the effectiveness of structured, AI-enhanced instructional modules in improving these skills among 64 undergraduate students enrolled in a technical university. Participants were randomly divided into two groups: an intervention group that completed AI-facilitated modules—including role-play, case analysis, and reflective journaling—and a control group that received traditional lectures. Pre- and post-intervention scores were evaluated using a validated soft-skill rubric. The intervention group achieved a mean composite gain of 18.4 points (SD = 3.4, p < 0.001), compared to a non-significant 6.7-point gain (SD = 2.8, p = 0.12) in the control group. The most substantial domain-specific improvements were observed in communication (22.3 points), teamwork (19.1), and emotional intelligence (18.8). ChatGPT outperformed Bard and Jasper in enhancing skill acquisition, with respective gains of 24.0% in communication and 18.2% in emotional intelligence. Thematic analysis of 320 reflective entries revealed a 45% increase in empathy-related references and a 52% rise in adaptive coping strategies. These findings support the strategic integration of AI tools into soft-skill pedagogy and demonstrate the efficacy of algorithm-driven reflective learning in enhancing emotional intelligence.