Human-Centred Production Performance in Indonesian Coal Contracting: A Dual-Mediation PLS-SEM Model

Purpose: This study develops an integrated human-resource model to examine how management policy, managerial capability, and employee loyalty influence production performance through job satisfaction and work discipline as complementary mediating mechanisms in an Indonesian coal-mining contractor. 
Research Methodology: This study applies Partial Least Squares Structural Equation Modeling (PLS-SEM) using a model based on 250 analytical cases and 30 reflective indicators. The analysis evaluates measurement quality, direct effects, predictive relevance, and mediation pathways using 5,000 bootstrap resamples. 
Results: The findings indicate acceptable reliability and validity, with moderate explanatory power for job satisfaction (R²=.516), work discipline (R²=.511), and production performance (R²=.569). All direct relationships are positive and significant. Work discipline is the strongest direct predictor of production performance (β=.308), while managerial capability is the strongest exogenous predictor (β=.224). All indirect effects show partial mediation, with discipline-based pathways stronger than satisfaction-based pathways. 
Conclusions: Production performance is shaped by both organizational and human factors. Effective policies, capable managers, employee loyalty, satisfaction, and discipline jointly contribute to improved operational outcomes. 
Limitations: The study uses a synthetic dataset; therefore, the findings represent methodological demonstrations rather than verified company-level empirical evidence. 
Contributions: This study contributes a dual-mediation framework that distinguishes affective and behavioral mechanisms in explaining production performance and provides managerial insights for strengthening human-centred operational improvement.