二元学习与创新绩效的作用机制——组织内部协作网络的调节作用

组织学习是驱动企业创新、实现技术追赶与超越追赶的关键因素,如何合理组织企业的利用式和探索式学习是企业构建研发体系的关键问题。学界对二元学习是否有利于企业的创新绩效存在截然不同的观点。同时，组织内部的情境因素是影响组织二元学习和创新绩效关系的最根本因素，组织内部研发协作网络是二元学习的重要载体。基于美国半导体行业企业十年间的专利数据，实证研究发现，不同维度的二元学习对企业创新绩效具有不同的影响，联合二元学习能够积极影响企业的创新绩效，平衡二元学习则会降低企业的创新绩效；内部研发协作网络密度在二元学习和创新绩效的关系中具有显著的调节作用，企业的内部协作网络密度越高，联合二元学习对企业创新绩效的积极影响越弱，平衡二元学习对创新绩效的负影响则加剧。

The Mechanism of Ambidextrous Learning in Innovation Performance: The Moderating Effect of Intra-organizational Collaboration Network

In ages of rapid changes and growing competitiveness, innovation is considered as the primary driver for firms to keep sustainable competitiveness, achieve technological catch-up and continue growing beyond catching-up. From the standpoint of firms, organizational learning is a crucial path toward innovation. Organizational learning is required not only to take full advantage of the existing knowledge stock, but also to explore new knowledge and technological opportunities. The former is exploitative learning based on existing knowledge, and the latter is explorative learning oriented to new knowledge and techniques. However, exploitative learning and explorative learning tap into totally different organizational routines and management behavior. Due to the finite resources and capabilities, firms face the dilemma between exploitative and explorative learning. How to organize explorative and exploitative learning in a reasonable manner is the key solution for firms to build efficient and effective R&D systems. Compared with firms in developed country, Chinese firms face constraints, such as scarce resources, imperfect market infrastructure, poor innovation capability and insufficient advanced knowledge. Thus, the trade-off between explorative learning and exploitative learning becomes more challenging. Chinese firms need to learn how to rationally allocate limited resources in ambidextrous learning so as to maximize the innovation performance. Scholars hold different opinions on whether innovation performance could benefit from ambidextrous learning. In addition, internal contextual factor is the carrier of ambidextrous learning which deserves attention. The role of internal contextual factors in the ambidextrous learning and the innovation mechanism need further exploration and validation. Therefore, to investigate the mechanism of ambidextrous learning in innovation performance is important for both practices and theory building. In this research, we investigate some global semiconductor companies listed in America. The patent data of these firms between 1991 and 2000 are collected and analyzed with the help of multiple techniques. Based on a view of knowledge, this research discusses two mechanisms of ambidextrous learning in innovation performance, respectively combined ambidextrous learning and balanced ambidextrous learning. The empirical results show that combined ambidextrous learning positively improves firms innovation performance, while balanced ambidextrous learning negatively reduces innovation performance. In addition, this research adopts a micro-level lens and claims that internal contextual factors are fundamental factors for the mechanism of ambidextrous learning and innovation performance. As a significant contextual factor, an intra-organizational R&D collaboration network is constructed by using patent citation. The empirical research shows that the density of intra-organizational R&D collaboration network significantly moderates the mechanism of ambidextrous learning in innovation performance. When intra-organizational collaboration network gets higher, the positive influence of combined ambidextrous learning would be weakens, and the negative influence of balanced ambidextrous learning s deteriorates. Based on the boundary of organizational cognition, this research seeks new methods to evaluate exploitative or explorative learning, which to some extent helps explain the inconsistency in existing literature. The empirical result implicates that the pursuit of the balance between exploitative learning and explorative learning is not the best solution. An adjustment on each proportion according to different strategic orientations is necessary. Furthermore, this research focuses on the importance of internal contextual factors. By calculating the density of intra-organizational network, this research provides a theoretical foundation and practical guide for the construction of R&D collaboration networks.