Network Analysis of Job Burnout, Job Performance, and Affect and Its Implications for Teaching Practice

Background Job burnout poses a significant challenge to the sustainability of higher education, particularly by undermining teaching quality and job performance through emotional depletion. Despite evidence linking burnout, affect, and performance, the complex interplay among these factors remains underexplored, especially among part-time university faculty. This study employed network analysis to uncover the dynamic interactions between burnout dimensions, affective states, and job performance outcomes. Methods A cross-sectional survey was conducted from March to August 2024, involving 1,020 part-time faculty members from Xinjiang Normal University. Participants completed validated scales measuring job burnout (MBI-GS), job performance (JPS), and affect (PANAS). Network analysis was performed using the R package qgraph, with model selection based on LASSO regularization and EBIC. Results The final network included 21 edges, with 13 non-zero connections (7 positive, 6 negative), and a sparsity of 0.38. Task performance emerged as the most influential bridge node (EI = 0.21), while depersonalization showed the lowest bridge centrality (EI = -0.09). Positive affect was positively linked to performance and negatively associated with emotional exhaustion and negative affect, while negative affect was negatively connected to efficacy and contextual performance. The correlation stability coefficient (CS = 0.75) indicated good network reliability. Conclusion This study reveals the non-linear structure linking burnout, affect, and job performance in academic settings. Task performance plays a pivotal role in mediating these interactions, while affect exerts dual effects on burnout trajectories. These findings offer theoretical insights and practical implications for developing targeted, emotion-sensitive interventions aimed at improving faculty well-being and teaching outcomes.