Enhancing Student Systems Thinking in Generative Artificial Intelligence-Supported Logistics Management Education in China: An Integrated Model with PLS-SEM and FsQCA

Systems thinking is a core competence in logistics management, as decisions across transportation, warehousing, and delivery functions are highly interconnected and often generate delayed, trade-off, or system-wide consequences. Despite the increasing integration of generative artificial intelligence (GenAI) tools into logistics education, limited research has examined how to enhance systems thinking in such contexts. Drawing on triadic reciprocal determinism, this study conceptualizes systems thinking enhancement as an emergent outcome of interactions among behavioral regulation, cognitive conditions, and environmental scaffolding. Using survey data from 236 logistics management students in Chinese universities, we integrate Partial Least Squares Structural Equation Modeling (PLS-SEM) and fuzzy-set Qualitative Comparative Analysis (fsQCA) to examine both net effects and configurational mechanisms. Results show that self-regulated learning exhibits the strongest positive association with systems thinking, while germane cognitive load is positively associated and extraneous cognitive load is negatively associated with systems thinking. Teacher GenAI scaffolding is linked to more favorable cognitive load allocation. fsQCA findings further reveal that high-level systems thinking emerges from specific combinations where self-regulated learning and germane cognitive load are fundamental conditions, whereas the absence of self-regulated learning consistently leads to low-level systems thinking. These findings provide guidance for the design of GenAI-supported curricula and scaffolding strategies.