ISSN 1671-3710
CN 11-4766/R
主办：中国科学院心理研究所
出版：科学出版社

• •    

下属人机协作与领导的多元反应：基于双系统加工视角

何国华, 蒋建武, 谭玲   

  1. 深圳大学管理学院, 广东 518055 中国
    广东工业大学管理学院, 广东 510520 中国
  • 收稿日期:2026-01-27 修回日期:2026-06-01 接受日期:2026-08-19
  • 基金资助:
    人机协作行为的影响因素与潜在阴暗面：基于领导与员工视角的多阶段、多方法研究(2024A1515012634); 效率与伦理的双重奏：人机协作行为的内容结构及其差异化影响研究(JCYJ20240813142512016); 下属人机协作与领导的多元反应及其后效：基于双系统加工理论的动态追踪研究(72502154); 人与生成式AI协同驱动企业产品创新的机制、路径及对策研究(24BGL052); 组织如何规避人机协作伦理风险？基于伦理相关结果与干预策略的多阶段、多方法研究(2025A1515010598); 基于平台生态视角下的零工经济人力资源管理重塑研究(GD22XGL27); 从冷思考到热反应：领导对下属人机协作的反应机制、后效、及其有效性研究(2026A1515012413); 组织中人与生成式AI协作对员工创新行为的消极影响及干预策略的多视角、多方法研究(2026A1515010730)

Subordinates’ human–AI collaboration and leaders’ diverse responses: A dual-process perspective

He Guohua, Jiang Jianwu, Tan Ling   

  1. College of Management, Shenzhen University 518055, China
    School of Management, Guangdong University of Technology 510520, China
  • Received:2026-01-27 Revised:2026-06-01 Accepted:2026-08-19

PDF (PC)

38

摘要： 人机协作作为提升企业创新和效率的关键驱动力，领导在此过程发挥了核心作用。然而现有研究主要聚焦员工视角，忽视了领导对下属人机协作的认知与行为反应。这一问题造成的挑战在现有研究的悖论中进一步凸显——领导往往高估下属人机协作的绩效，而行为者视角的研究则表明其客观绩效确有提升。这种认知偏差不仅会阻碍企业技术升级与竞争力提升，也反映了现有研究对领导的多元反应及其有效性探讨不足。鉴于此，本研究基于双系统加工理论，开展三项相互关联的研究：(1)探讨领导的“冷”反应模式，即其对下属人机协作的动机归因和未来影响感知如何引发发展与变革导向行为；(2)探讨领导的“热”反应模式，揭示人机协作如何激发领导的相对剥夺感，并促使其以报复性无礼行为作为防御性应激反应；(3)从动态视角比较领导不同行为反应对人机协作可持续性的影响，以优化人机协作效能，助力企业数智化转型。

关键词: 领导力, 人机协作, 归因过程, 伦理, 绩效

Abstract: Human–AI collaboration has emerged as a key driver of enterprise innovation and efficiency, with leadership playing a central role in this process. However, existing research predominantly adopts an employee-centric perspective, overlooking how leaders perceive and respond to subordinates’ human–AI collaboration. This gap is further highlighted by a paradox in the literature: while leaders tend to overestimate the performance of subordinates engaged in human-AI collaboration, studies from the actor’s perspective indicate that their objective performance does improve. This cognitive bias not only hinders technological advancement and competitiveness but also reflects a lack of research on the diverse and effectiveness of leaders’ responses. To address this issue, this research, grounded in Dual-System Processing Theory, comprises three interrelated studies: (1) Study 1 examines leaders’ “cold” response patterns, exploring how their attributions of subordinates’ motives and perceptions of long-term impact influence their development- and change-oriented behaviors; (2) Study 2 investigates leaders’ “hot” response patterns, revealing how human-AI collaboration can trigger leaders’ feelings of relative deprivation, leading to retaliatory incivility as a defensive stress reaction; (3) Study 3 adopts a dynamic perspective to compare the long-term effects of different leader responses on the sustainability of subordinates’ human–AI collaboration, with the aim of optimizing collaboration effectiveness and supporting enterprises in their digital transformation.

Key words: leadership, human–AI collaboration, attribution, ethics, performance