Primacy effect-based dynamic feedback mechanism considering communication sequence for multilevel infiltrative large-scale group decision-making

In decision-making environments with ultra-large-scale decision-makers (DMs) and resource constraints, the extensive interaction among DMs in social network large-scale group decision-making (SN-LSGDM) often leads to inefficiencies and redundant information. To tackle this challenge, this paper explores the multilevel infiltrative large-scale group decision-making (MI-LSGDM) event and proposes a primacy effect-based dynamic feedback (PE-DF) mechanism. This mechanism integrates the communication process into the consensus-reaching process and introduces a communicator identification method that synthesizes DMs’ opinion, behavioral, and relational characteristics. This method effectively channels a significant portion of assessment information from a broader decision-making circle to the inner circle through identified communicators, while promoting the diffusion of consensus from the inner circle to the entire group, thereby mitigating biased decisions. Considering the primacy effect where DMs’ bias toward initial information encountered during communication, this study constructs a communication sequence optimization model, guiding DMs in the inner circle to achieve consensus quickly and accelerating the multilevel infiltration of consensus. An illustrative example and a series of comparative experiments validate the flexibility of the communicator identification method across various decision-making scenarios and demonstrate the ability of the communication sequence optimization model to promote efficient information exchange and consensus-building within limited time and resources. Overall, the proposed PE-DF mechanism solves the practical limitations of traditional SN-LSGDM models constrained by DM scale and resource availability, exhibiting strong performance and robustness across a broader range of decision-making scenarios.