Discrete Modeling of Teacher Satisfaction and Well-Being Using TALIS Data: Latent Profiles, Gender Disparities, and Global Network Dynamics

This study applies discrete mathematical modeling to examine patterns of teacher satisfaction and well-being using data from the 2018 Teaching and Learning International Survey (TALIS), which included 261,426 teachers across 48 education systems. Latent profile analysis based on nine indicators: self-efficacy, job satisfaction (overall, profession, and environment), emotional well-being, workload stress, disciplinary climate, professional engagement, and collaboration; identified three distinct profiles: High (34.2%), Moderate (45.6%), and Low (20.2%). Welch’s t-tests revealed significant gender disparities across most indicators, with female teachers reporting higher workload stress (M = 9.30 vs. 8.99, p < .001) and lower emotional well-being (M = 9.43 vs. 9.19, p < .001). Country-level ANOVA results indicated statistically significant differences across all nine measures (p < .001). Teachers in Finland, Japan, and South Korea reported higher levels of emotional well-being and professional satisfaction, while Brazil, Mexico, and the Czech Republic reported lower levels. A Gaussian graphical model was then constructed to explore interrelations among indicators. The network revealed strong partial correlations between job satisfaction and emotional well-being (r = 0.24), and between workload and emotional exhaustion (r = –0.21). Centrality analysis highlighted overall job satisfaction and emotional well-being as the most influential constructs, serving as key bridges and hubs of connectivity. These findings demonstrate the utility of combining latent profile classification with graph-theoretic modeling, providing a novel framework for analyzing large-scale teacher data through discrete mathematical procedures.