Answers are grounded in this paper and cite their sources.
Identifies candidate targets for miner safety, suggesting tailored interventions based on profiles.
Introduction Understanding how psychological resources configure to sustain miner safety is essential for designing targeted interventions, yet remains limited. Methods Data were collected from 1,247 Chinese frontline miners who completed measures of transformational leadership, job crafting, work engagement, person-job fit, organizational commitment, knowledge sharing, and safety performance. Latent class analysis (LPA) was applied to identify psychological profiles, and Gaussian graphical models with EBICglasso regularization were estimated for each profile. Network centrality, bridge centrality, and a network-based candidate target index were computed to prioritize profile-specific intervention targets. Network Comparison Tests (NCT) were used to formally compare network structures across profiles. Results Four profiles emerged, namely Job-Crafting-Driven (42.8%), Leadership-Dependent (11.4%), Low-Resource Vulnerable (18.8%), and Optimal (27.0%). Network density was highest in the Job-Crafting-Driven class (0.560) and lowest in the Optimal class (0.320). Person-job fit and knowledge sharing consistently ranked among the top three bridge nodes across all four profiles. The candidate target index revealed profile-specific priorities, such as person-job fit and knowledge sharing for the Job-Crafting-Driven group, and organizational commitment for the Low-Resource Vulnerable group. NCT indicated significant global strength differences between Class 1 and all other classes, and significant network structure differences between Classes 1 and 4, and between Classes 2 and 3. Discussion These findings provide a foundation for generating testable hypotheses about profile-specific safety interventions, pending longitudinal validation. Practically, safety interventions should be tailored to distinct worker profiles, with person-job fit and knowledge sharing as priority targets for most profiles. Optimal profile centrality estimates were statistically unreliable and excluded from interpretation.
No takes yet. Share an insight, caveat, or question.
Pang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: