Abstract
This study investigates student engagement profiles in a collaborative-competitive learning environment within an Industrial Statistics course at a university in Indonesia. The research employed a quantitative survey design involving 200 undergraduate students who experienced a pedagogical model combining within-group collaboration and inter-group competition. Cluster analysis using K-means algorithm identified three distinct engagement profiles based on four constructs: collaboration quality, attitude toward group work, learning outcome, and competitive orientation. Reliability analysis showed good internal consistency across all constructs (Cronbach's α = 0.802–0.850). The optimal cluster solution was determined using the elbow method and silhouette coefficients. The three profiles represent students with high engagement, moderate engagement, and disengagement respectively. As a supplementary analysis, case study scores from 89 students were examined to explore the relationship between engagement profiles and academic performance. Pearson correlation analysis revealed weak positive correlations between learning outcome and case study scores (r = 0.228). The findings suggest that while the collaborative-competitive model effectively engages the majority of students, approximately one-third of students has low engagement and require differentiated pedagogical support. This study contributes to the growing body of research on student engagement in collaborative learning environments and provides practical implications for designing inclusive instructional strategies.