社会网络视角的企业家学习模式研究

【作者】 黎赔肆；

【导师】 丁栋虹； 胡君辰； 杨永康； 谢晋宇；

【作者基本信息】 复旦大学 ， 企业管理， 2008， 博士

【摘要】 在技术快速变化和竞争日趋激烈的今天,企业家不仅要充分利用现有知识,而且要不断探索新的技术和技能来适应商业环境的变化。因此,如何改善企业家探索式和利用式学习至关重要。对个体来说,探索式学习和利用式学习是两种排他性的学习模式(Gupta&Smith,2006)。既有研究虽然考察了企业家探索式学习和利用式学习在自我经验学习中作用(Politis&Gabrielsson,2005),但是没有指出企业家该如何平衡这两种学习模式。而且,在经验学习研究中,企业家被看作是“智力罗宾逊(intellectual Robinson Crusoe)”(Pavlica,et al.,1998),社会互动对企业家学习的影响被忽视。事实上,网络学习是提升企业家能力的重要途径之一。针对上述问题,论文在控制企业家经验情况下,关注以下四个议题:①企业家社会网络与企业家学习模式之间有什么样的关系?②不同网络维度在影响企业家学习模式上是否有替代或互补关系?③在竞争性的网络观点(如伯特的结构洞和科尔曼紧密社会资本理论)下,对应的网络变量是否分别对探索式学习和利用式学习产生积极作用?如果这样,这意味着企业家能通过构建不同的网络来平衡这两种学习模式;④社会网络、学习模式分别与学习的直接结果——机会识别(Politics,2005)之间的关系如何。围绕这些问题,按照从社会网络——企业家学习方式——学习结果的逻辑,论文进行了一系列的理论和实证研究。论文以中小企业家为研究对象,结合文献和理论分析,提出了17假设。使用问卷调查方式获取研究数据。在问卷设计中,论文使用“提名”(name generators)法获取网络变量的信息,并运用社会网络分析软件UCINET6.0计算每个企业家网络的结构洞数值。最后,利用SPSS15.0,采用层次回归法对研究假设进行了验证。通过研究,论文有如下研究结论:(1)关于企业家网络与学习模式关系的研究。根据Hoang&Antoncic(2003)总结,企业家社会网络分析有三个成份,即网络结构、网络治理和网络内容,对应的,论文分别考察了结构洞(限制性)和弱关系(强关系)、信任、企业家“自我中心”网络中“他人”之间的教育背景的异质性(同质性)和企业家与“他人”之间知识的差异性(相识性)与学习模式之间的关系。结果表明:结构洞、教育异质性、知识差异性与探索式学习显著正相关;强关系、信任、教育背景的同质性、知识的相似性与利用式学习显著正相关。也就是说,竞争性的网络观点通过对不同学习模式的影响显示出其正确性。这表明,企业家可以通过构建不同的网络来平衡探索式学习和利用式学习。特别的,论文考察了过度信任问题,过度信任被认为不利于创新性的探索式学习(Zahra,et al.,2006),因此,论文期望信任与探索式学习呈倒U型关系。结果表明:信任的回归系数为正,信任平方的回归系数为负,但二者均不显著。不过,这也提醒我们,信任的作用并不总是正面的。(2)网络维度之间关系的研究。论文通过两种方式来考察网络维度之间的关系:一是通过逐步将网络变量引入回归模型,考察后引入变量对已经进入变量的显著性是否有削弱作用,如果是,则说明后者对前者具有替代作用。结果表明,变量之间不存在相互替代关系,这证实了Seibert,et al.(2001)关于网络维度可以一起发挥作用的观点:二是通过考察网络变量之间的交互作用来判断网络维度之间是否具有依赖或互补关系。研究发现:结构洞与教育异质性的交互作用与探索式学习显著正相关,同时,结构洞的显著性消失,而教育异质性的显著性依然稳健,这表明结构洞的作用可能依赖于教育异质性:其它维度之间没有发现显著关系。(3)关于网络、学习模式与机会识别关系的研究。研究发现:在网络变量中,只有教育异质性(同质性)与机会识别显著正(负)相关;探索式学习不仅与机会识别显著正相关,而且它在教育异质性与机会识别之间起到完全中介的作用;利用式学习与机会识别之间没有显著关系。除了上述三个核心问题外,论文还考察了与之相关的另外两个问题:(4)企业家认知风格的影响。企业家学习受认知风格的影响(Ulrich&Colc,1987;Man,2006),论文的研究发现:企业家创新型认知风格与探索式学习显著正相关,与利用式学习显著负相关;当企业家具有高创新型认知风格时,网络结构洞的多少并不影响企业家的探索式学习,当企业家具有高适应型认知风格时,企业家与网络成员间知识的相似性程度的高低对利用式学习没有显著影响。(5)企业家个体特征和企业特征的影响。回归分析发现:①女性企业家、年龄比较大的企业家、有一定类似产业经验的企业家偏好利用式学习;②民营、外资和合资企业的企业家偏好探索式学习;③年龄在25岁以下的企业家、创业经验少的企业家机会识别能力偏低,高科技企业家的机会识别能力明显高于其它产业的企业家。最后,论文系统地归纳了研究结论,指出了论文所取得的理论进展和指导实践的建议,并对论文研究的局限性和后续研究进行了客观分析。更多还原

【Abstract】 In face of the rapid changing in technology and the increasingly fierce competition,it is necessary for entrepreneurs to make full use of existing knowledge and continuously explore new technologies and skills to adapt to the changes in the business environment.Therefore,it is very important how to improve entrepreneurs’ explorative learning and exploitive learning.On the individual,explorative learning and exploitive learning are two exclusive modes of learning(Gupta & Smith,2006).Literatures investigated roles of explorative learning and exploitive learning in entrepreneurs’ experiential learning (Politis & Gabrielsson,2005),but failed to point out that how to balance these two modes.Moreover,this literatures treat entrepreneurs as "intellectual Robinson Crusoe"(Pavlica,et al,1998),and the impact of social interaction on entrepreneurs’ learning has been ignored.In fact,network learning is one of the important ways to enhance entrepreneurial competencies.To address above issues,the paper focus on the following four topics:①what kind of relationship are there between social networks and entrepreneurs’ learning modes?②Whether there are alternative or complementary relationships between dimensions of network?③Corresponding to the competitive views of the network (such as Burt’s structure hole theory and Coleman’s social capital theory),if there are positive relationships between network variables and explorative learning and exploitive learning? If so,this means that entrepreneurs can balance explorative learning and exploitive learning by constructing different networks;④What kind of relationships are there between social network,learning modes and opportunity recognition(Politics,2005)? In order to cope with these issues,a series of theoretical and empirical studies are performed.Based on the literatures and theory analysis,the paper put forward 17 research hypotheses.In questionnaires,the paper use "name generators" to measure network variables and applies social network analysis software UCINET6.0 to calculate the value of structural hole of each sample.Finally,through employing the statistical tool SPSS15.0 and multiple regression analysis(hierarchial enter) to examine all hypotheses.The conclusions are sum up as follow:(1)According to Hoang&Antoncic(2003),there are three elements in analysis of entrepreneurs’ social network,namely,network structure,network governance and network content,corresponding,the paper examines respectively the relations between structure holes(constraint),weak(strong) ties,trust,education heterogeneity(homogeneity),dissimilarities(similarities) between entrepreneur and "alters" in his/her network and learning modes.The results show that:there are significant positive correlations between structural holes,education heterogeneity, knowledge dissimilarities and explorative learning,and there are also significant positive correlations between strong ties,trust,education homogeneity,knowledge similar and exploitative learning.That is to say,the competitive views of network show their reasonableness through their effects on explorative(exploitive) learning. This means that entrepreneurs can balance explorative learning and exploitive learning by constructing different networks.Especially,the paper pays attention to the problem of over-trust.Over-trust is not considered conducive to innovative and explorative learning(Zahra,et al,2006), therefore,the paper predicts that trust has an inverted U-shaped relationship to explorative learning.The results show that:the coefficient for the trust is positive, whereas the coefficient for trust is negative,but they are not significant.However,this reminds us that the role of trust is not always positive.(2)This paper investigates the relationship between the network dimensions by two ways:First,by the progressive introduction of a network variable to regression model,if the significance of a previous variable fates after the introduction of new variable,this means that new variable is a complete surrogate for previous variable, but,such relationships between variables are found,this result confirms the view of Seibert,et al(2001) that the different dimensions of the network can play a role together.Second,through the interaction between network variables,we can determine whether or not there are complementary relationships between dimensions of network.The paper find:the interaction of structural hole and education heterogeneity has an significant positive relationship to explorative learning,at the same time,the significance of structure hole fades and education heterogeneity of significant remains robust,which indicating that the effects of structure hole may be dependent on education heterogeneity;No significance relationship exists among other dimension.(3)The paper examines relationship between network variables, explorative(exploitive)learning and opportunity recognition.We find:only education heterogeneity(homogeneous) has significant positive(negative) relationship to opportunity recognition;Explorative learning not only has significant positive relationship to opportunity recognition,but also is full mediator between the relationship of education heterogeneity and opportunity recognition;But,no significance is found between exploitive learning and opportunity recognition.In addition,two other important related issues are discussed:(4) The first is the effect of entrepreneurs’ cognitive style.Entrepreneurs’ cognitive style impact entrepreneurs’ learning(Ulrich & Cole,1987;Man,2006),this paper finds:Innovative cognitive style has significant positive relationship to explorative learning and adaptive cognitive style has significant negative relationship to exploitive learning;When with high innovative cognitive style,the number of structure holes does not affect entrepreneurs’ explorative learning,when with high adaptive cognitive style,the extent of knowledge similarity has no significant impact on exploitive learning.(5) The second are the effects individual characteristics of entrepreneurs and characteristics of firms.Regression analysis shows that:①Female entrepreneurs, older entrepreneurs,entrepreneurs with a certain experience in similar industries preferences exploitive learning;②Entrepreneurs in private firms,foreign-funded firms and joint ventures preferences explorative learning;③Entrepreneurs under the age of 25,entrepreneurs with fewer entrepreneurial experience have low capabilities of opportunities recognition,and entrepreneurs in high-tech firms can recognize more opportunities than entrepreneurs in other industries.Finally,the paper summarizes the main conclusions,point out the theoretical progresses,and put forward some guides for practice,and discusses the limitations and future direction.更多还原