The impact of teacher-student power structure on classroom psychological environment an educational big data perspective

This research aimed to investigate the effects of power arrangements on psychological environments in the classroom by analyzing big data in 156 classrooms and 4,680 students in 28 Malaysian secondary schools. Using K-means clustering, four typologies of power arrangements: Democratic-Supportive, Authoritarian-Directive, Moderate-Collaborative, and Laissez-faire, were found. MANOVA revealed significant typology effects across all psychological environment dimensions, with effect sizes ranging from small to large (η² = 0.07 to 0.54). Democratic-Supportive typologies demonstrated superior outcomes, particularly in student autonomy (F = 58.35, p < 0.001, η² = 0.54). Psychological adaptation acted as a significant mediator for power arrangements: β = 0.35 for Democratic-Supportive typologies and β = 0.22 for Moderate-Collaborative typologies but not for Authoritarian-Directive typologies. Behavioral measures accounted for 6% to 16% additional variance beyond self-reports, supporting multi-source approaches in education literature. This study makes three key contributions: it advances a novel empirically-derived typology of teacher-student power structures based on objective behavioral data, provides evidence linking specific power structure types to distinct classroom psychological outcomes, and demonstrates the incremental validity of big data analytics over traditional self-report measures in educational psychology research.