【作者】 郭爱芳；

【导师】 陈劲；

【作者基本信息】 浙江大学 ， 管理科学与工程， 2010， 博士

【摘要】 步入21世纪以来,知识和学习在企业技术创新中发挥的作用越来越受到学者们的关注。知识成为企业的关键资源,如何利用知识实现持续创新成为企业创新战略的主要挑战。在这个战略搜索过程中,至关重要的是识别和有效利用企业不同学习活动的互补性,优化知识资源的使用。Lundvall教授等学者于2004年识别了STI (Science,Technology, Innovation)和DUI(Doing, Using, Interacting)这两种学习和创新模式,Jensen等学者基于丹麦企业的调查研究,得出将STI和DUI两种学习模式相结合有助于提高创新绩效,但该结论是否具有普适性,是否适合中国企业,STI/DUI学习对技术创新是否具有互补作用还有待进一步验证。我国企业自主创新能力薄弱,已经成为经济可持续发展的桎梏,有机将基于科学的学习(Science-based learning,STI学习)和基于经验的学习(Experience-based learning,DUI学习)结合起来可能是我国企业提高自主创新能力的有效途径,而这需要进一步的理论支持,本文正是围绕这一核心问题展开研究。本文综合运用知识观、组织学习和技术创新等相关理论,围绕“STI学习和DUI学习如何影响技术创新绩效”这一基本研究命题,探究了企业STI学习和DUI学习及其结合对创新绩效的影响、深入剖析了STI/DUI学习影响创新绩效的作用机制,并就STI学习和DUI学习在不同发展阶段的动态演化进行了探讨。本文的研究内容主要通过以下部分展开：(1)文献综述和理论发展。通过对企业知识观、组织学习和技术创新等相关理论的详细回顾,找出现有研究的不足。对STI学习、DUI学习和组织学习能力等关键变量的概念及其要素构成进行了分析界定,形成本文论证的基础。(2)探索性案例研究。选择4个典型的案例开展深入的探索性案例研究。经过理论预设、案例选择、数据收集、案例内分析和多案例之间的比较研究,推导出关于STI/DUI学习、组织学习能力与企业技术创新绩效关系的十一个初始研究命题,为后续研究提供源于实践的构想。(3)企业STI/DUI学习与技术创新绩效关系的理论模型与实证检验。在探索性案例研究基础上,结合已有相关研究进行更深层次的理论探讨,分别提出STI学习和DUI学习及其两者结合对创新绩效的影响概念模型、STI/DUI学习影响技术创新绩效的作用机制概念模型和细化假设。基于230家企业的调查问卷,运用验证性因子分析、回归分析、方差分析和结构方程建模等方法进行了实证研究,进一步明晰STI/DUI学习对技术创新绩效的影响、内在作用机制以及环境动荡性的调节作用。(4)探讨STI学习和DUI学习在技术创新不同阶段以及企业发展不同阶段的动态演化规律。先是结合已有相关研究进行理论分析,提出STI学习和DUI学习动态演化的假设,再基于230家企业的调查问卷,运用描述性统计分析和方差分析进行验证。通过上述分析论证过程,本文得出如下主要结论：(1)企业的STI学习和DUI学习均有助于企业技术创新绩效的提升。外部技术和市场环境的动荡程度对DUI学习与创新绩效之间关系具有调节作用。在技术和市场环境变化缓慢的情况下,DUI学习对创新绩效的作用更明显,随着技术和市场环境变化程度加大,DUI学习对创新绩效的促进作用减弱。(2)STI学习对创新绩效的影响是间接的,这种作用通过组织学习能力的传递而产生。具体而言,STI学习是通过正向作用于企业的知识搜索识别能力与知识整合应用能力,进而正向影响其技术创新绩效的。DUI学习对技术创新绩效有直接正向效应；同时,DUI学习还通过正向作用于知识搜索识别能力和转移能力,间接影响技术创新绩效。企业的STI学习和DUI学习之间存在一定的相互依赖性和互补性,它们对创新绩效有交互作用,同时注重STI学习和DUI学习的企业将比只重视STI学习或DUI学习的企业更有可能创新。(3)STI学习和DUI学习在不同阶段动态演化。随着产品创新从变动阶段、过渡阶段发展到特定阶段,STI学习的重要性程度逐渐下降,DUI学习的重要性程度逐步提升。从技术的引进模仿、改进提高到自主技术创新的技术发展过程中,企业的STI学习和DUI学习强度都逐渐增强。上述研究结论深化了STI/DUI学习与技术创新绩效间关系的理解,具有理论上的原创性和方法论意义上的创新性。具体而言,本研究在以下四个方面进行了深化与拓展：(1)本文以我国制造业为实证研究对象,是对Jensen等学者关于STI/DUI互补创新结论在不同国度的检验。除对新产品数量的影响外,本文还分析了STI学习和DUI学习结合对企业新产品开发速度、产品的独创性、开发持续性和新产品销售比重等多个创新绩效指标的影响。本文还首次探究了技术和市场环境动态性对STI/DUI学习与技术创新绩效间关系的调节作用。(2)本研究通过理论分析和实证研究探讨STI/DUI互补创新的内在作用机制,进一步打开了STI/DUI互补创新的机理,建立了"STI/DUI学习--组织学习能力--技术创新绩效”的理论框架,是对Jensen等学者关于STI/DUI互补创新结论的进一步解释,更有力地证实STI学习和DUI学习之间的互补关系,有助于完善STI/DUI互补创新理论。(3)本文首次对STI学习和DUI学习在企业不同发展阶段的动态演化以及在产品创新不同阶段的演化轨迹进行了探讨,进一步丰富和发展了Jensen等学者关于STI/DUI互补创新的理论,进而为中国制造企业实现战略转型与升级的研究提供了一个新的视角。(4)本文进一步明晰了组织学习能力的概念及构成要素,指出组织学习能力主要由知识搜索识别能力、转移能力和整合应用能力三个能力要素构成,并设计了直接度量各能力要素的测度量表,经检验具有较好的信度与效度,为今后类似的研究打下了基础。本文还深入探究了组织学习各能力要素对企业技术创新绩效的影响机制,以及STI学习和DUI学习对提升组织学习能力的作用,进一步丰富充实了当前组织学习的相关理论。更多还原

【Abstract】 Since striding into the 21st century, scholars are more and more focusing on the role that knowledge and learning play in the enterprise’s technological innovation. Knowledge becomes the enterprise’s key resource. The main challenge for enterprise in innovation strategy is how to use knowledge to sustain its continuous innovation. During this strategic search, the most importance task is to recognize and make efficient use of the complementarity among the enterprise’s different learning activities, and optimize knowledge use. In 2004, professor Lundvall and his fellows recognized two modes of learning and innovation modes:STI (Science,Technology, Innovation) and DUI (Doing, Using, Interacting). Scholars like Jensen further deepen the concepts of these two modes. Drawing on the results of the 2001 Danish DISKO Survey, they make a preliminary conclusion that firms are more likely to innovation by combining the two modes of learning. However, their research are not sufficient to demonstrate such complementarity between STI and DUI learning, more verifications are still needed of whether this conclusion is universal and whether it adapts to Chinese enterprises. Weakness of Chinese Enterprises in indigenous innovation is becoming a shackle for our country’s sustainable development in economy. Combining STI and DUI learning properly may be an effective way to improve the enterprise’s indigenous innovation capability, which still needs further theoretical support. This paper launches the research right on this central issue.This dissertation integrates related theories like knowledge-based view, organization learning and technological innovation, focuses on the basic research proposition of "how STI and DUI learning influence technological innovation performance", explores the influence on innovation performance from STI/DUI learning and their interaction, thoroughly analyzes the mechanism how STI/DUI learning influence innovation performance, and discusses the dynamic evolution locus of STI and DUI learning in different development stage. The main work is expanded in the following four parts:(1) Through review of related theories like knowledge-based view, organization learning and technological innovation, the dissertation finds out shortcomings of the current research and puts forward the definition to the concepts and constructs of key variables such as STI learning, DUI learning and organization learning capability (OLC), which forms the basis for argumentation.(2) Four typical cases are chosen for the explorative cases research. Through theoretical presumption, data collection, within-case analysis and comparison among cases,11 preliminary research issues are put forward about relationships between STI/DUI learning, organization learning capability and enterprise’s technological innovation performance, which provides a construct for future research.(3) Based on explorative cases research mentioned above and in-depth theoretical analysis, two concept models of the influence of STI & DUI learning and their combination on innovation performance and the mechanism how STI & DUI learning influence technological innovation performance are put forward. The hypotheses are tested using a sample of 230 Chinese manufacturing firms.(4) The dynamic evolution process of STI & DUI learning in the different stage of technological innovation and enterprise development is approached in the end.This dissertation comes to the following main conclusions based on the above-mentioned analysis:(1) Enterprise’s STI & DUI learning have positive influence on technological innovation performance. Both STI learning and DUI learning can facilitate the technological innovation performance. The uncertainty of external technology and market environment can regulate the relationship between DUI learning and innovation performance. In the circumstance that technology and marked environment change slowly, DUI learning will have more obvious influence on innovation performance, while as the technology and market environment change quickly, the facilitation of DUI learning to innovation performance will become weak.(2) The influence of STI learning on innovation performance is indirect, and this influence is derived through the mediating role of OLC. In detail, it is via STI learning positive influence on knowledge search-recognition capability and integration-application capability, that it imposes its positive influence on technological innovation performance. DUI learning has direct and positive effect on technological innovation performance. Meanwhile, through positive influence on knowledge search-recognition capability and transfer capability, DUI learning further influence technological innovation performance. STI learning and DUI learning have certain mutual interdependency and complementarity on each, and they influence innovation performance interactively. Meanwhile, firms combining STI and DUI learning are more likely to innovate than those emphasizing on STI learning or DUI learning respectively.(3) Based on the U-A dynamic pattern of innovation, this dissertation finds that the importance of STI learning descends significantly from fluid, transition to specific stages while DUI learning ascends significantly. In addition, this thesis finds that enterprises’STI learning and DUI learning intensity strengthen gradually from the phase of introduction and imitation, improvement, to indigenous technological innovation.By employing an integrative, multidisciplinary approach, this thesis tries to make some innovation in the following fields:First, this dissertation takes Chinese manufacturing industry as the empirical research object, and it is a proof of STI/DUI complementary innovation in different countries which is put forward by scholars like Jensen. In addition to the influence on the quantity of new products, the paper also discusses the influence that the combination of STI and DUI learning impose on the enterprises’other innovation performances indicators such as new products development speed, originality, development continuity, and proportion of new products sales etc. The dissertation, for the first time, makes research on the regulating function that the uncertainty of technology and market environment has on the relationships of STI & DUI learning and technological innovation performance.Second, this dissertation discusses the internal mechanism of STI/DUI complementary innovation, further opens the’black box’of STI/DUI complementary innovation, and establishes the theoretical framework of "STI/DUI learning -- OLC-technological innovation performance". It further explains Jensen and his fellow’s conclusion of the STI/DUI complementary innovation, and facilitates to perfect the STI/DUI complementary innovation theory.Third, this dissertation first explorers STI/DUI leaning dynamic evolution in enterprise different development stage and the evolution track in different products innovation stages, further enriches and develops the STI/DUI complementary innovation theory, and provides a good perspective for the research on Chinese manufacturing enterprises realizing strategic transformation and update.Fourth, this dissertation refines the construct of OLC and points out that OLC is mainly composed of knowledge search-recognition capability, transfer capability and integration-application capability. The paper also designs a scale for directly measuring each capability element which is proved to have high reliability and validity, laying foundation for similar researches in the future. In addition, the dissertation gives a deep investigation on the influence mechanism that each capability element has on enterprise’s technological innovation performance, and the role that STI/DUI learning plays in promoting OLC, which further enrich the current organization learning theory.更多还原