Innovative instructional strategies are increasingly needed to enhance student engagement in quantitative engineering courses. This study examines a collaborative-competitive learning design implemented in Industrial Statistics course for Undergraduate Student of Industrial Engineering, Telkom University. Students collaborate within groups at each meeting by working on questions in groups and in some sessions, students competing across groups. Students may respond differently to such instructional designs. To evaluate how students respond to this instructional innovation, a data-driven using cluster analysis was employed. Data were collected from 200 undergraduate students through a structured questionnaire after implementing the learning design innovation. Instruments measured collaboration quality, attitude towards work group, sense of belonging, perceived learning outcomes, and competitive orientation. Following data screening, descriptive analysis, reliability testing and analyzed using K-means clustering. Based on silhouette evaluation and interpretability considerations, a three-cluster solution was retained. The findings reveal three distinct engagement profiles: (1) highly engaged students demonstrating strong collaboration, belonging, and competitive motivation; (2) moderately engaged students showing balanced but less intense responses; and (3) a small segment of disengaged students with consistently lower scores across dimensions. These findings indicate that collaborative–competitive learning generates differentiated engagement patterns rather than uniform effects. The proposed data-driven framework offers practical implications for designing adaptive and inclusive learning environments in engineering education.