多层创新网络结构洞特征对组织创新绩效的影响——来自新能源汽车领域的实证分析

基于1997-2019年我国新能源汽车领域专利申请数据，采用复杂网络方法构建组织合作网络、成员合作网络以及知识融合网络3类创新网络，并量化3类网络中的结构洞特征。在此基础上，采用面板负二项回归模型实证分析组织合作网络、成员合作网络以及知识融合网络结构洞特征对组织创新绩效的影响。结果表明：仅考虑直接作用时，组织合作网络与知识融合网络结构洞特征对组织创新绩效具有显著正向影响，且知识融合网络结构洞特征对组织创新绩效的影响程度更大。在进一步考虑交互作用后，组织合作网络、成员合作网络以及知识融合网络结构洞特征的两两交互项均对组织创新绩效具有显著正向影响，其中，成员合作网络与知识融合网络的交互作用对组织创新绩效影响最大。结构洞特征替换为聚集性特征后，研究结论依旧稳健。最后，基于多层创新网络优化视角提出相应的政策建议。

Research on the Influence of Multi-level Innovation Network Structure Hole Characteristics on Organizational Innovation Performance

Based on the patent application data of China's new energy vehicle field from 1997 to 2019,this paper uses a complex network method to construct three types of innovation networks: organization collaboration network,member collaboration network,and knowledge mergence network,and quantifies the characteristics of the structural holes in the three types of networks. On this basis,the panel negative binomial regression model is used to empirically analyze the influence of the structural hole characteristics of organization cooperation network,member cooperation network and knowledge fusion network on organizational innovation performance. The research results show that: when only direct effects are considered,the structural hole characteristics of organizational collaboration networks and knowledge mergence networks have a significant positive impact on organizational innovation performance,and the characteristics of knowledge mergence network structure have a greater impact on organizational innovation performance; and after further considering the interaction,the two interaction items of organizational collaboration network,member collaboration network and knowledge mergence network structure hole feature have a significant positive impact on organizational innovation performance. Among them,the interaction between the member collaboration network and the knowledge mergence network has the strongest influence on the organization's innovation performance. After replacing structural hole features with clustering coefficient,the above research conclusions are still robust. Finally,based on the perspective of multi-level innovation network optimization,the corresponding policy recommendations are put forward.