1. bookAHEAD OF PRINT
Dettagli della rivista
License
Formato
Rivista
eISSN
2444-8656
Prima pubblicazione
01 Jan 2016
Frequenza di pubblicazione
2 volte all'anno
Lingue
Inglese
access type Accesso libero

The medical health venture capital network community structure, information dissemination and the cognitive proximity

Pubblicato online: 20 May 2022
Volume & Edizione: AHEAD OF PRINT
Pagine: -
Ricevuto: 16 Jul 2021
Accettato: 06 Dec 2021
Dettagli della rivista
License
Formato
Rivista
eISSN
2444-8656
Prima pubblicazione
01 Jan 2016
Frequenza di pubblicazione
2 volte all'anno
Lingue
Inglese
Abstract

This paper studies the effect of venture capital (VC) network community structure on the dissemination of information in the medical health field of China. Using theoretical analysis and empirical test, this paper divides the community structure and information dissemination. The structural dimensions of the community have different effect and intensity on information dissemination. The cognitive proximity positively (negatively) affects community structure dimensions on information dissemination. This paper finds that the network community structure has a lag effect on information dissemination, and it finds out the types of network community structure that can promote (hinder) information dissemination. These results can promote the sustainable development of the medical health industry.

Keywords

JEL Classification

Introduction

With the rapid development of China, increasing attention has been paid to the Chinese medical health field, and then the venture capital (VC) funding has increased. Japsen [1] discussed how to provide a major source of funding in an increasingly difficult financing environment, which is critical to the development of the medical health industry. According to the statistics of China Venture (CV) Source, the VC number in the medical health industry had reached 3038, accounting for 7.22% of the total investment number, ranking the sixth by the end of 2016. However, there is still a lack of research on VC of the Chinese medical health industry. Bruce and Bruce [2] carried out research on angel investment in medical firms. Arthur [3] believes that in the VC of medical and biotechnology field, it is necessary to have flexibility. Moreover, the investment strategy should be adjusted as opportunities arising and should follow the leading VC. Khanin et al. [4] established a VC model to investigate 12 countries and 485 VC firms supporting medical health. With the expansion of the number and scale of VC, Chinese VC firms will form a network through joint investment in the medical health industry to avoid risks and share resources.

At present, there are many small clusters in Chinese medical health field VC network. The close cooperation among these clusters has attracted the domestic and foreign scholars’ attention. Bubna et al. [5] introduced the network community into the field of VC, but there is relatively little research on the internal structure of the VC network community. In fact, the scholars of social network theory hold that network tie is the channel for resources flowing. Then, network structure is the topology reflection of network tie. Information is a resource and therefore can be transmitted through network structures. Scholars found that a close and cohesive network community structure can promote the dissemination of new knowledge [6]. However, there is no unified conclusion on how network community structure affects information dissemination. In general, the knowledge of VC firms can be divided into ‘information’ and ‘expertise knowledge’. Here, we adopt the research conclusion of Huber [7] and make no distinction between information and knowledge.

Our research has the following contributions. First, it reveals the impact mechanism of VC network community structure on information dissemination in the medical and health industry from the meso-level. Second, it proposes the way to improve the investment performance of network community: VC network community structure – information dissemination – investment performance. Third, it divides the dimension of network community structure and information dissemination for the first time. Fourth, it distinguishes the different impact and intensity of the network community structure dimension on information dissemination. Fifth, it finds that cognitive proximity promotes or hinders the impact of network community structure on information dissemination. Finally, it finds that the network community structure has a time lag effect on information dissemination. Meanwhile, it finds the types of network community structure promoting or hindering information dissemination.

Literature review

The existing research mainly focuses on the influence of network community structure on information dissemination from the perspective of structural characteristics based on social network theory. The Scholar considers that a natural attribute of subgroups in social networks is cohesion. Further many subgroups’ nodes have quite high clustering coefficients [8]. The existing studies believe that cohesion and stickiness within a community are conducive to the dissemination of information [9]. Shaikh et al. [10] studied focusing on the innovation diffusion and the role of opinion leaders in the diffusion process. Here, opinion leaders refer to firms with high degree. Further degree is also an important characteristic of network structure. Another characteristic is short path length. It increases the solving ability of the network system. Knowledge will be transferred to different parts of the network more quickly, with less attenuation [11].

There are also some studies on the information dissemination by network community structure from the perspective of information characteristics. Further, it is found that the attributes and contents of information make the information dissemination difference under the structure. The reason for knowledge transfer is closely related to knowledge-related factors. Moreover, easily expressed knowledge is easily transmitted through products, reports and other tangible means [12]. Knowledge with implicit characteristics is embedded in the structure of cooperation in the transfer process, which needs to be transmitted in a practical way [13]. Based on network structure, information content becomes the influencing factor of information dissemination [14].

There are also studies about the information dissemination at the collective level. Based on the resource-based theoretical analysis, it is found that individual knowledge is embedded in a super-personal repository so that others can access it. Further, individual knowledge can be transformed into collective knowledge through dissemination [15], thus affecting the dissemination of collective information. However, the existing studies of the network community structure on information dissemination are scattered. Further, these are all based on a single theory for analysis. Thus, it is necessary to combine the social network theory and resource-based theory to build a holistic analysis method to discuss the impact of the network community structure on the information dissemination.

In the existing research, the dimension division of network community structure is based on the analysis of cohesive subgroups. Wasserman and Faust [16] focused on four aspects in their analysis of cohesive subgroups: reciprocity of relationships, accessibility of relationships among network members, frequency of relationships among members of subgroups, and comparison between relationships among members of subgroups inside and relationships among members of subgroups outside. In this paper, the structure of VC network community is defined as the close connection between investment firms caused by reciprocity norms, with a certain degree of hierarchy distribution, which restricts and promotes the flow of resources. Based on the analysis method of cohesion subgroup, the dimensions of network community structure are divided into cohesion, status difference and coordination.

If only being observed three dimensions from the perspective of network topology, the network community structure reflects a connection mode without isolated nodes. The economic meaning of the above three structural dimensions is temporarily ignored. Therefore, the connections between community nodes are found to be divided into two categories. One category represents the direct connection between nodes, using degree measure. The other category refers to the indirect connections between nodes, then can be divided into two types. One type is connection with common friend, the other is connection without common friend. The cohesion measures the first one and the average path length measures the last one. Therefore, the dimension of network community structure divided from the perspective of network topology is complete. There is no intersection between dimensions.

The network community information dissemination of VC is defined as a communication system with the community members as the subject and object of the communication, the project selection, post-investment management experience, etc., information as the content of the communication, the network community organisations as the link between the dissemination chain, the realisation of information value increment as the dissemination effect of the infinite circulation. The main work of VC firms and their ability to be improved are focused on the selection of investment projects and post-investment management [17]. In this paper, based on the content information is divided into project information and experience information. Therefore, this paper holds that the project information dissemination refers to the dissemination of information that can be programmed and written, such as evaluation indexes, screening criteria and contract formulation during the project screening stage. In this paper, the experience information dissemination is defined as the dissemination of tacit information such as operation guidance, market development, consulting services, etc. to the invested firms in the later stage of management.

Cognitive proximity, which affects information dissemination, is mainly caused by the similarity between the two sides. Resource-based theory holds that the firms of similar knowledge base can establish connections between them [18]. Because enterprises will try out new solutions based on these knowledge bases [19]. From the perspective of social network theory, scholars believe that cognitive proximity not only helps to establish relationships, but also promotes the formation of efficient and effective social relations. The main reason is that common language, code and narrative perspectives develop into common understanding and values [20]. That is reflected in the study of VC networks. There are differences among VC firms in terms of scale, industry and other aspects, which are constantly changing with the changes of time and environment. Among them the cognitive proximity caused by the proximity of industry plays a very important role in the dissemination of information [21]. Cognitive proximity, cognitive distance, industrial distance, knowledge base and technical similarity are all of equal significance in different research fields. The cognitive proximity’s influence on the structure and information dissemination can be found in the existing study. However, under the framework of combined social network theory and resource-based theory, the analysis about how to change VC network community structure and the information dissemination by the cognitive proximity is lack. Therefore, it is necessary to combine the two theories to discuss the influence mechanism of network community structure on information dissemination.

Research design
Theoretical analysis and hypothesis
The Influence of Cohesion on Information Dissemination

The cohesion of VC network community will promote the dissemination of project information. The cohesion refers to the extent to which both sides go to form a connection because of the existence of a common friend [22]. Project information is focus on the knowledge of investment project screening and value evaluation. In practice, it can be formed into files and become explicit information. Cohesion has an impact on project information dissemination from two aspects. First, in accordance with the theoretical analysis of the embedding viewpoint of social network, high cohesion means that the high degree of relations embedded among the VC network community members [23]. The community members hold more channels of links, which r result in the multi-dimensional interaction. This interaction can produce greater trust, more effective transfer of the expertise and the tacit understanding of fine-grained project information [24]. Further, interaction can reduce the risk of investment decision-making problem of asymmetric information. Second, the long-term trust of members of the VC network community caused by cohesion can reduce the tendency of opportunistic behaviours [25] and damaging reputation behaviours. This ensures the credibility of project information and reduces the risk of investment payback.

The project information influences the community information dissemination by changing the information base of VC network community. Following the resource-based theory, the multiple resources aggregation producing combinatorial effect [26], highly cohesive is conducive to the collision and recombination of project information into new information. Project information is ‘easy to express’ and ‘easy to observe’, which will form the consistency of communication trend in the close connection of the VC network community [27]. Moreover, this accumulates into the community routine [15]. It directly enters the community project information base, thus leads to change in the community information dissemination. Therefore, the following hypothesis is obtained:

Hypothesis 1a: The cohesion of VC network community has a positive impact on project information dissemination.

The cohesion will also have an impact on the dissemination of experience information in the VC network community with the characteristics of implicit information. The cohesion has two main effects on the dissemination of experience information. First, the social network theory holds that the close or special relationship of the VC network community leads to a high degree of embedded relationship, which is manifested as that greater trust can more effectively transmit more expertise and more tacit fine-grained experience information [24]. Since some complex implicit experience information can only be transmitted through a firm connection [28]. This can effectively deal with the uncertainty of environmental change. Second, the cohesion will lead to the existence of redundant relationships in the VC network community. This is conducive to the reinforcement of implicit experience information dissemination. Because multiple redundant connections can increase the interaction effect [29]. Further which finally increase the feasibility of experience information.

The experience information influences the community’s information dissemination by changing the information base of the community. From the resources convergence leading to integration of the resource base theory [26], the experience information of the implicit information characters effectively integrate requires a lot of individual exchange [30]. Thus in the close contact VC network communities, the firms’ experience information dissemination will gradually build and replace the community experience information base [15], leading to community information dissemination change. Therefore, the following hypothesis is obtained:

Hypothesis 1b: The cohesion of VC network community has a positive impact on the dissemination of experience information.

The Impact of Status Difference on Information Dissemination

The status difference will affect the dissemination of project information. The status difference is expressed through the degree centrality of the firm. This refers to the extent of the firm’s large number of direct connections with other firms [31]. Degree centrality affects project information dissemination in two ways. First from the embedding viewpoint of social network, the VC firms with high degree centrality mean good reputation and may decrease opportunism behaviour [32]. Because doing so will damage their reputation, therefore it can ensure that the authenticity of information communication, reduce the causal ambiguity, and increase the VC network community discovery of the new investment project. Second, the high degree centrality firm maintains dominance in the VC network community. In order to maintain these advantages, it will actively launch selective network connection [33], to enhance their project information dissemination capacity, to reduce the project information replication difficulty and to effectively reduce the uncertainty of members of the community screening high quality project.

The project information dissemination of the firm influences the project information dissemination of the community by changing the project information base of the VC network community. From the viewpoint of resource integration of the resource-based theory [26], a large number of project information of low causal ambiguity and replication difficult collision together can produce recombination and update etc. Which directly accumulate to the VC network communities project information base [15], then become a community project information routine. Hence, status differences can promote community projects information dissemination.

The influence of cohesion on project information dissemination is greater than that of status difference for two main reasons. First cohesion reflects the two sides of information dissemination own the common friend. Project information dissemination is more likely to occur when a strong third-party relationship surrounds the relationship [22]. The status differences reflect the information that can be held and controlled by the VC network community member [34]. Then from the information dissemination willingness the effect of cohesion is greater than that of status difference. Second, the cooperation norms promoted by cohesion can alleviate potential conflicts [35] and limit competition, thus promoting the dissemination of project information. The status difference can form the potential difference of information resources to ensure the information dissemination. From the observation of eliminating the barriers to dissemination, the effect of cohesion is more important than the status difference. Therefore, the following hypothesis is proposed:

Hypothesis 2a: The status difference of VC network community has a positive impact on project information dissemination, and the influence of cohesion is stronger than the status difference.

The status difference expressing by the degree centrality will affect the dissemination of experience information. The experience information has a lot of implicit components. Thus, some of which are embedded in personal skills and some of which are embedded in environment. It has characteristics such as tacit, unobservable use, simplicity and system dependence [36]. The degree centrality influences the dissemination of experience information in two ways. From the perspective of the embeddability of social networks, the high degree centrality results in high reputation [32]. This means that the firm has high legitimacy and reliability. Which leads to the cost reduction of the dissemination of experience information needing ‘learning by doing’, then reduces the exit risk of VC firm. Second, high degree centrality means that the firm has a strong control and coordination ability [33]. This can reduce the incomplete understanding of the special characteristics of the new environment for the dissemination of experience information. Eventually, the VC firms can improve the uncertainty of investment performance when they use imitation, experiment and other ways to spread experience information.

The dissemination of firm’s experience information can change the information base of the VC network community to affect the dissemination of community experience information. According to resource-based theory, multiple heterogeneous resource sets assembling can result in recombination [26]. The experience information transmitted in high quality can be recombined and updated through observation, demonstration and experiment. Therefore, the firm’s experience information in the VC network community gradually updates the community experience information base through the replacement and optimisation process [15]. This ultimately affects the dissemination of community experience information.

There are two main reasons why cohesion has more influence on the dissemination of experience information than status differences. First, high cohesion reflects that there is a third-party relationship. Therefore, many kinds of relations connect stability and network connection strength conducive to VC network community members to promote the absorption and transformation of information [37]. This reduces experience information’s uncertainty impacting on the invested enterprise, then provide reinforcement and make requiring repeat interactive experience information dissemination more efficiency. Status difference can provide direct connection to experience information dissemination. However, from the perspective of strengthening experience information, the influence of cohesion on experience information is greater than that of status difference. Second, the cohesion plays a role in alleviating conflicts and restricting competition among members of the VC network community. Moreover, this can stimulate potential joint investment behaviours [35]. This promotes imitation for encoded experience information. Compared with the direct connection relation provided by status difference, the effect of cohesion on the experience information dissemination is greater than that of status difference. Therefore, the following hypothesis is proposed:

Hypothesis 2b: The status difference of the network community has a positive effect on the dissemination of experience information; moreover, the effect of cohesion is stronger than the status difference.

The Coordination Impact on Information Dissemination

High community coordination refers to the short path between community members, which will affect the dissemination of community project information. From the perspective of relational embeddedness of social network theory, first, the short path between firms in the VC network community can lead to a greater probability of project information dissemination. Therefore, this effectively reduces the uncertainty of investment results. As Singh and Marx [38] believes that the smaller the distance, the greater the possibility of patent citation, then the greater the possibility of patent citation by the same region or the same enterprise. Second, the short path increases the solving ability of the VC network community. Further the explicit project information is easy to compile and express. Which will spread to different parts of the community more quickly [11], attenuating less, reducing the risk of investment decision.

The firm’s project information dissemination influences the dissemination of community project information by changing the community’s project information base. Resource-based theory suggests that a large number of heterogeneous polymerisation generates new resources [26]. Then in the VC network community of shorter average path length project information with expressiveness and observable characteristic independent of the system can be effectively assembling, screening, optimisation, etc., into the new project information. Therefore, the project information of the closely connected firms in the community is constantly summarised and superposed as a routine procedure for the community to screen projects [15]. Therefore, the high coordination will promote the dissemination of project information of the VC network community.

There are two main reasons why status difference has a greater impact on project information dissemination than the coordination. First, good coordination indicates the presence of shorter path length. From the perspective of the dissemination possibility, reputation and trust caused by status difference impacting on the project information dissemination are much greater than the influence of the short path length. Mainly because of project information of VC firms with higher status has higher legitimacy and trust worthiness [39]. Second, from the perspective of the speed of dissemination, VC firms with high degree centrality have the ability to control information and other resources [40]. This can independently choose the optimal connection to transfer project information to accelerate the speed of information dissemination. Thus, status differences are faster than the speed of short path length dissemination. Therefore, the following hypothesis is obtained:

Hypothesis 3a: The coordination of the VC network community has a positive impact on project information dissemination; moreover the status difference has a stronger impact than the coordination.

High coordination expresses the existing of short path length, which has a positive effect on the dissemination of experience information. The coordination affects the dissemination of experience information in two ways. First based on the analysis of the social network theory, the short path length of the VC firms can form strong interactive, therefore easy establish strong joint investment relationship with other VC firms [41]. The experience information with strong situation reliability will not be able to communicate without firms. Therefore, strong interaction strengthens the observation and imitation of experience information, increase information dissemination willingness, and reduce the uncertainty of VC firms return potentiality. Second, the short path length of the VC inter-firm leads to strong control ability. Then, the monitoring and incentive ways could be effective implementation [42]. This ensures the efficiency of the experience information dissemination of the firm, meanwhile enables the resolution of organisational problems such as operation mode and strategy formulation to cope with environmental uncertainty.

The experience information dissemination of VC firms influences the community experience information dissemination by changing the experience information base of the community. From the viewpoint of different resource aggregations producing combinatorial effects of the resource-based theory [26], strong dependencies formed by short path length enable firms to gather experience information quickly and effectively. Therefore, the firm experience information in the closely connected VC network community will be gradually replaced and improved the community experience information base through ‘learning by doing’ [15], thus promoting the dissemination of experience information in the community.

There are two main reasons that the status difference has a greater influence on the dissemination of experience information than that of coordination. First, from the perspective of the dissemination reliability, the implicit experience information dissemination is difficult [43]. The VC firm with high degree centrality can gain more amount and more comprehensive experience information [44]. The experience information with the characteristic of complex and difficult to expressing disseminate by highly trusted connections relationship or strong social interaction [28]. The reliability of the dissemination is enhanced, therefore status difference impact on the experience information dissemination effect than coordination. Second, from the perspective of dissemination speed, VC firms with high status make clear what partner to known and what to do [45]. Therefore, which could more accurately choose the optimal path and dissemination method, speed up the experience information imitating and learning, reduce the experience information dissemination cost, improve information dissemination speed. Thus, the status difference is more important than coordination for experience information dissemination. Therefore, the following hypothesis is obtained:

Hypothesis 3b: The network community coordination has a positive impact on the dissemination of the experience information; moreover the status difference on the experience information dissemination is stronger than the coordination.

The Cognition Proximity Moderator

Project information with explicit information characteristics is easy to be textual and replicable. Therefore, cognition proximity will impact on the method and result of projects information dissemination of VC network community. According to the analysis of the relational embeddedness viewpoint of social networks, cognition proximity affects cohesion and project information dissemination in two aspects. First, cognition proximity will increase the easy interaction and flexible positioning among members of the VC network community [42]. This can include more members to participate in the interaction, facilitate the compilation of project information and promote a wider range of project information dissemination, so as to deal with the uncertainty of project selection. Second, the cohesion of the main characteristics of the VC network community will affect the members’ supporting information dissemination [35]. Further cognition proximity will reduce the interactive cognition cost to expand the scope of information dissemination. According to the structural embeddedness viewpoint of social networks, cognitive proximity affects status difference and project information dissemination in two aspects. First, cognitive proximity will change the process of cognitive search of project screening experience among members of the VC network community. Therefore, new solutions will be tested based on the belief of the relationship between the results and actions [19]. Thus, the control advantage of project information dissemination is formed. The authenticity of project information is guaranteed. Moreover, the uncertainty of investment selection is effectively reduced. Second, too much gap in status difference will lead to the negative effect of too much knowledge difference among members of the VC network community [46]. However, this can be mitigated by cognitive proximity. According to the analysis of the relational embeddedness viewpoint of social networks, cognitive proximity will affect the dissemination of project information in two aspects. First, cognitive proximity will prompt members of the VC network community to search and evaluate individuals with the close project knowledge base, then change the path of project information dissemination [47]. It optimises the short-distance link to accelerate the project information dissemination speed, further reduce the investment uncertainty caused by the distortion of project information dissemination. Second, knowledge transfer follows the path of least resistance [48], then forms path dependence. Cognitive proximity will deepen the mutual understanding among firms in the VC network community, promote the non-redundancy of information dissemination path and reduce the uncertainty level of investment projects. Combining with the viewpoints of integrating resources to generate new resource of the resource-based theory [26], cognitive proximity can promote members to form ‘cognitive consistency’ and accelerate the generation of new project information. Therefore, under the effect of cognitive proximity, the project information of firms in the closely connected community can be constantly added up to become a procedural method for selecting projects before community investment. Thus, cognitive proximity will accelerate the dissemination of project information of the VC network community. Therefore, the following hypothesis is proposed:

Hypothesis 4: Cognitive proximity positively affects the effect of VC network community structure on project information dissemination.

Hypothesis 4a: Cognitive proximity positively affects the effect of cohesion on project information dissemination.

Hypothesis 4b: Cognitive proximity positively affects the effect of status difference on project information dissemination.

Hypothesis 4c: Cognitive proximity positively affects the effect of coordination on project information dissemination.

The experience information with the characteristics of implicit information is highly embedded in the knowledge base of individual ability and organisation. Cognitive proximity changes the method and the outcome of information dissemination. From the perspective of relational embeddedness of social networks, cognitive proximity changes the impact from the cohesion to the experience information dissemination in two aspects. First, the cohesion forms multiple redundant connection relations. Meanwhile cognitive proximity can increase the VC network community firm interaction [42] to form epistemic community. In the face of experience information of difficult to writing such as enterprise management, cognitive proximity could deepen understanding process, facilitate experience information imitation, improve the experience information dissemination effect and reduce the instability of the investment return. Second, cognitive proximity can promote the understanding and absorption of knowledge by VC network community firms [49]. Therefore, this reduces the communication cost caused by the inconsistency of subjective cognition and stickiness of tacit knowledge between the two parties of the experience information dissemination, so as to cope with the changeable environment. According to the structural embeddedness viewpoint of social networks, cognitive proximity will affect the process of the experience information dissemination by status difference in two ways. First, the firms with high status have high trust [32]. Then cognitive proximity will enhance the trust among members, promote VC firms testing new solutions, strengthen the position of higher firms in the experience information dissemination control advantage, improve the effectiveness of the experience information dissemination, and to reduce the uncertainty of the potential returns. Second, the formation of status difference is the difference in the extent of firm knowledge accumulation in the community [33]. Cognitive proximity will promote the VC network community firms to observe and compare the learning experience information to reduce the obstacles to the dissemination of experience information caused by the large status difference. From the perspective of relational embeddedness of social networks, cognitive proximity affects the effect of coherence on the dissemination of experience information in two ways. First, cognitive proximity will change the search path of experience information dissemination of the VC network community firms [50], enhance the selection range of experience information dissemination in coordination, shorten the time of experience information dissemination, and reduce the uncertainty of VC firms’ performance. Second, the cognitive proximity will promote the mutual understanding of the investment field among the VC network community firms; therefore facilitate the establishment and connection of different knowledge bridges [51]. Moreover, this increases the impact of coordination on the reduction of the experience information dissemination cost. Combining with the viewpoint that resource aggregation is easy to generate new resources in resource-based theory [26], the cognitive proximity reduces the number of times of continuous practice and repeated adjustment of the mode of experience information dissemination in VC network community firms, meanwhile accelerates the speed of new experience information synthesis. Therefore, by the cognitive proximity firm’s experience information in the close connection VC network community can gradually improve and perfect to form an information base for the community’s post-investment management and support services. Thus, cognitive proximity will accelerate the community experience information dissemination. Therefore, the following hypothesis is proposed:

Hypothesis 5: Cognitive proximity positively affects the effect of VC network community structure on experience information dissemination.

Hypothesis 5a: Cognitive proximity positively affects the effect of cohesion on the experience information dissemination.

Hypothesis 5b: Cognitive proximity positively affects the effect of status difference on the experience information dissemination.

Hypothesis 5c: Cognitive proximity positively affects the effect of coordination on the experience information dissemination.

Data sources

This paper selects China Venture (CV) source database for VC research in Chinese medical health industry. We extract information on investment time, investment firms, investment amount, IPO or M&A, etc. We also collected basic information about investment stage, investment industry and fund management scale of invested firms. The VC in China appeared around 1990. The time to form the VC network and extract the community was about 2006. Thus, our sample period is from January 1, 2006, to December 31, 2017. The final sample included 727 VC firms and 934 rounds of joint ventures.

We use a three-year time window to calculate the impact of network community structure on information dissemination. For example, first we select all joint investment information for 2006–2008 time windows in the CV source database. Firms involving in joint investment were coded, and then were entered into the corresponding location of UCINET 6.560 matrix according to the joint investment relationship. Thus, the VC network was formed. Second the GN algorithm is used to extract network community. It already exists in UCINET 6.560. In our study the network community size is set greater than or equal to 4. Finally, according to the divided communities, the corresponding measurement information dissemination values of each network community in the 2007–2009 time windows are found. The regression analysis can be carried out by pairing the calculated information dissemination values of each network community with the community structure corresponding to the previous time window. A total of 100 network communities were detected.

Variables and measures

In this paper, the dependent variable is the project information and experience information dissemination. While the independent variable examines the community cohesion, the community status difference and the community coordination, the moderator variable examines the role of cognitive proximity. Some community-level variables are taken as control variables.

Independent Variable

Based on the cohesive subgroup research of Wasserman and Faust [16], this study analyses the VC network community. Then, the average accumulation coefficient of the community is used to measure the community cohesion. The average degree centrality of the community is used to measure the status difference of the community. The average path length of the community is used to measure the community coordination.

Dependent Variable

The first dependent variable is project information dissemination. We measure the average value of the increment sum of all firm investment industry types in the community as project information dissemination. This is mainly divided into three steps to obtain the measurement method. First, the project information is the screening and evaluation method of a certain industry project. The research conclusion of Gupta and Sapienza [52] is adopted to represent project information by the type of investment industry. Second, after investigating the spread of enterprise VC practice, Gaba and Meyer [53] considered intra-group communication as dissemination from previous adopters to future adopters in the group. Thus, at time T+1, the firms in the community have invested in new industries. Which represent the adoption of information from previous projects information at time T. This paper takes the investment increment of the firm to the industry at time T+1 as the measure of project information dissemination. Finally, referring to the research ideas of Shore et al. [54], the network is seen as a whole. Therefore, the result at the collective level is the accumulation of the results of these measures. This paper takes community project information dissemination as a whole and considers that it is equal to the average value of project information dissemination of all community members.

The second dependent variable is experience information dissemination. We measure the average value of the sum of incremental investment enterprises of all community members in the community as the experience information dissemination. This is mainly divided into three steps to obtain the measurement method. First, the experience information is value-added services for invested enterprises. We measure the experience information using the means of Hochberg et al. [55]. It expressed in terms of the cumulative number of invested enterprises. Second, we adopt the research conclusion of Gaba and Meyer [53]. They consider that in time T+1 community firms invested in new enterprises to represent the adoption of previous experience information at time T. In this paper, the increment of investment enterprise at time T+1 is taken as the measure of experience information dissemination. Finally, referring to the research ideas of Shore et al. [54], the experience information dissemination of community is taken as a whole. Then it is the average of the sum of all community members’ information dissemination.

Moderator Variable

The cognitive proximity of the VC network community is mainly a measure of the similarity of the knowledge base among the community firms. This study using the research of Jaffe et al. [56] on the measurement of technical proximity calculates the cognitive proximity matrix of the community. The average value of the sum of each row in the matrix represents the cognitive proximity of investment firms in the community. Referring to the research of Shore et al. [54], the cognitive proximity of the community adopts the average value of the sum of the proximity of all investment firms. See Appendix A for specific measurement method.

Control Variables

The first control variable is community scale. We use the number of members in a community as a measure [57].

The second control variable is resource endowment. We use the logarithmic transformation of the number of managed funds (in millions) as the measurement method [57]. This paper refers to the research idea of Shore et al. [54], which states that the resource endowment of a community is the average value of the sum of the members’ resource endowment.

The third control variable is industry diversity. We use the ratio of the maximum priority investment amount of an investment firm in the same industry to the total number of different industries invested [57] before the focus investment. See Appendix B for the specific methods. Based on the research ideas of Shore et al. [54], the industry diversity of a community is the average of the sum of the industry diversity of its members.

The fourth control variable is stage diversity. We use the ratio of all upfront investments to the maximum amount of upfront investments in the same stage [57]. See Appendix C for specific methods. Drawing on the research thought of Shore et al. [54], this paper obtains the stage diversity of a community as the average of the sum of the members’ stage diversity. Table 1 presents the description and definition of variables.

Descriptive statistics of variables and Pearson correlation matrix

Variable name Average Standard deviation Observed quantity Pearson correlation matrix

Project information dissemination 0.883 0.701 100 1
Experience information dissemination 1.531 1.353 100 0.688 1
Community cohesion 0.709 0.065 100 −0.328 −0.316 1
Community status difference 3.603 1.011 100 0.275 0.285* −0.180 1
Community coordination 1.667 1.031 100 −0.295 −0.129 0.164* −0.042 1
Cognitive proximity 0.241 .087 100 −0.114 −0.008 0.258 −0.105 0.219 1
Community scale 1.815 0.236 100 0.158 0.300 −0.259−0.204 0.665** −0.290 −0.286 1
Community average resource endowment 4.444 1.526 100 0.252 0.074 0.115 −0.046 −0.175* 0.169 1
Community average industry diversity 0.295 0.100 100 −0.085 −0.073 0.068 −0.071 0.052 0.062 −0.215 −0.693** 1
Community average stage diversity 0.291 .063 100 0.110 0.034 −0.089 0.005 −0.277 −0.013 0.350* 0.697** −0.682** 1

Notes: ** and * indicate significant correlation at 0.05 and 0.1 levels, respectively.

Data analysis and results
Descriptive statistics of the sample

The descriptive statistical results of the sample are given in Table 1. It can be seen from the table that some variables show a high correlation.

Multiple regression analysis

As multi-collinearity exists in community cohesion, community status difference and community coordination, ridge regression analysis [58] is adopted in this paper to ensure the authenticity of the results. All variables are standardised before calculation. Models 1 and 2 are the main effect models, and moderator variables are added to model 3. In order to avoid the multi-collinearity problem that the interaction item may cause among the three indicators of community structure, the interaction item is gradually added to the main effect model [59]. Therefore, models 4–6 is the test of the moderating effect of interaction item. The regression analysis results of project information dissemination are shown in Table 2.

Regression results of the project information dissemination regression model

Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Scale 0.110 (0.143) 0.074 (0.142) 0.039 (0.139) 0.100 (0.133) 0.102 (0.137) 0.062 (0.131)
Community average resource endowment 0.191 (0.148) 0.132 (0.134) 0.153 (0.133) 0.149 (0.128) 0.097 (0.130) 0.137 (0.127)
Community average industry diversity −0.266 (1.580) −0.188 (0.141) −0.051 (0.144) −0.083 (0.137) −0.028 (0.140) −0.050 (0.138)
Community average stage diversity 0.209 (0.160) 0.133 (0.143) 0.197 (0.145) 0.258 (0.140) 0.246 (0.141) 0.226 (0.142)
Cohesion −0.307** (0.188) −0.303** (0.186) −0.336** (0.179) −0.304** (0.179) −0.309** (0.177)
Status difference 0.299** (0.183) 0.271* (0.180) 0.306** (0.171) 0.228* (0.176) 0.326* (0.171)
Coordination −0.274* (0.319) −0.299* (0.319) −0.287** (0.305) −0.287* (0.307) −0.282* (0.308)
Cognitive proximity −0.257* (0.191) −0.254* (0.183) −0.247** (0.185) −0.233* (0.222)
Cohesion × cognitive proximity 0.214* (0.222)
Status difference × cognitive proximity −0.250* (0.225)
Coordination × cognitive proximity 0.230* (0.342)
R2 0.149 0.386 0.423 0.494 0.486 0.498
R2 0.032 0.221 0.238 0.304 0.293 0.311
F 1.273 2.341* 2.289* 2.599** 2.522** 2.655**
K 0.2 0.2 0.2 0.2 0.2 0.2

Notes: Standard errors in parentheses.

p < 0.1,

p < 0.05,

p < 0.01. K is the value of ridge regression.

Model 7 is the model of the control variables’ impact on the dissemination of experience information. Model 8 is the main effect model, and model 9 is added a moderator variable. Similarly, the moderator effect of interaction item is gradually added to the main effect model. Models 10–12 are the testing of moderator effect of interaction item. The regression analysis results of experience information dissemination are shown in Table 3.

Results of the experience information dissemination regression model

Variables Model 7 Model 8 Model 9 Model 10 Model 11 Model 12

Scale 0.261 (0.145) 0.098 (0.144) 0.061 (0.140) 0.033 (0.129) 0.081 (0.132) 0.056 (0.133)
Community average resource endowment 0.106 (0.149) 0.013 (0.136) 0.041 (0.131) 0.057 (0.125) 0.024 (0.124) 0.115 (0.128)
Community average industry diversity −0.175 (0.160) −0.198 (0.145) −0.297 (0.145) −0.289 (0.140) −0.289 (0.139) −0.281 (0.139)
Community average stage diversity 0.107 (0.161) 0.196 (0.146) 0.117 (0.144) 0.132 (0.139) 0.129 (0.137) 0.118 (0.137)
Cohesion −0.261** (0.227) −0.301** (0.222) −0.350** (0.215) −0.296** (0.211) −0.315** (0.212)
Status difference 0.317** (0.158) 0.318** (0.153) 0.266** (0.144) 0.397** (0.147) 0.338** (0.146)
Coordination −0.277* (0.149) −0.284** (0.145) −0.258* (0.138) −0.308* (0.137) −0.359** (0.141)
Cognitive proximity −0.255* (0.165) −0.374* (0.157) −0.294* (0.158) −0.315* (0.158)
Cohesion × cognitive proximity 0.241* (0.274)
Status difference × cognitive proximity −0.219* (0.199)
Coordination × cognitive proximity 0.225* (0.223)
R2 0.132 0.363 0.427 0.495 0.501 0.499
R2 0.012 0.191 0.243 0.305 0.314 0.311
F 1.103 2.114* 2.328* 2.610** 2.682** 2.659**
K 0.2 0.2 0.2 0.2 0.2 0.2

Notes: Standard errors in parentheses.

p < 0.1,

p < 0.05,

p < 0.01. K is the value of ridge regression.

Inspection results and discussion

In the project information dissemination, the cohesion in model 2 significantly negatively affects the project information dissemination, so hypothesis 1a is not supported. Community cohesion is a negative influence to project information dissemination, mainly because the network community itself is a close connection ‘small groups’. While the growing cohesion can increase interactions between members, limit the heterogeneity knowledge inputting, make the spread of knowledge repeating and redundancy [60]. This has a negative effect to the project information dissemination. The status difference has a significant positive influence on the project information dissemination. In order to compare the effect of cohesion and status difference, this paper using Liu et al. [61] research conclusion compare the significance T-test of the difference between regression coefficients. The T-test was carried out for the regression coefficient of different independent variables in the same equation; meanwhile the corresponding T-value and P-value were calculated. There was a significant difference in the effects of the two factors on project information dissemination (t = 35.875, p < 0.001). Therefore, cohesion was more influential than status difference on project information dissemination. Hypothesis 2a was supported. In model 2, although the coordination test result represents the negative value, the coordination measurement index adopts the average path length. It indicate that the shorter the average path length is the better the coordination is, further the better the project information dissemination effect is. This is actually represented as a significant positive effect. In order to compare the effect of status difference and coordination, this study using Liu et al. [61] research conclusion compare the significance T-test of the difference between regression coefficients. The T-test was conducted for the regression coefficient of different independent variables in the same equation, therefore this study calculate the corresponding T-value and P-value. There was a significant difference in the effects of the two factors on project information dissemination (t = −40.714, p < 0.001). Thus, the influence of status difference was stronger than coordination’s, and hypothesis 3a was supported.

In model 3, there is a significant negative moderating effect of cognitive proximity on the project information dissemination affected by community structure. Therefore, hypothesis 4 is not supported. The main reason is that VC network community composed of close interconnected firms. Contact closely will form a considerable understanding and familiar with. Thus, the improvement of cognitive consistency will strengthen community norms, oppose firms to invest in trying new projects [35]. In the end, cognitive proximity negative influence the community structure on project information dissemination.

In model 4, cognitive proximity positively moderates the impact of cohesion on project information dissemination. Thus, hypothesis 4a is supported. In model 5, the interaction item between status difference and cognitive proximity is added. Therefore, cognitive proximity negatively moderates the impact of status difference on project information dissemination. Hypothesis 4b is not supported. The possible reason is that the status difference of community members indicates that members control different information. To some extent, status difference will facilitate the flow of information. While the cognitive proximity will weaken the potential difference of this information flow and form the homogeneity of the information pool [32]. This will hinder the spread of project information. In model 6, the interaction between coordination and cognitive proximity is added. Therefore, the impact of coordination on project information dissemination is negatively moderated by cognitive proximity. Hypothesis 4c is not supported. The possible reason is that the joint action of cognitive proximity and short path leads to the ‘over-embeddedness’ of the firms in the community [24]. This may cut off the external contact, create isolation and fail to obtain novel project information. In the end which has a negative impact on the project information of the community.

In model 8, cohesion significantly negatively affects the experience information dissemination, thus hypothesis 1b is not supported. The main reason is that cohesion will lead to close connections and more interactions, even which will force members to follow the common norms. Under the pressure of imitation, this is not conducive for the firm to make changes according to the environment in the ‘learning by doing’. Then which is not conducive to the dissemination of experience information. In model 8, the status difference significant positively affects the project information dissemination. In order to compare the effect of cohesion and status difference, the research conclusion of Liu et al. [61] is adopted to compare the significant T-test of the difference between regression coefficients.

In model 8, the status difference significantly positively affects the experience information dissemination. The T-test of the regression coefficient of different independent variables in the same equation was carried out, and then the corresponding T-value and P-value were calculated. There was a significant difference in the effects of the two factors on experience information dissemination (t = 77.375, p < 0.01). Therefore, cohesion was more influential than status difference on experience information dissemination. Hypothesis 2b was supported. In model 8, although the test result of coordination is negative, the measurement index adopts the average path length. The shorter the average path length is, the better the coordination is. Thus, the negative value represents a significant positive effect. In order to compare the effect of status difference and coordination, the research conclusion of Liu et al. [61] is used to compare the significance T-test of the difference between regression coefficients. The T-test of the regression coefficient of different independent variables in the same equation is carried out, therefore the corresponding T-value and P-value are calculated (t = −16.965, p < 0.01). Thus, the effect of status difference on the experience information dissemination was stronger than that of coordination. In the end hypothesis 3b was supported.

In model 9, the cognitive proximity has a significant negative effect the community structure on experience information dissemination. Thus, hypothesis 5 is not supported. The reason is that the network community structure presents a close connection relationship, which will lead to a high extent of cognitive consistency. Resulting in previous experience will limit creative thinking, because it will lead to the adoption of familiar strategies in solving problems [62]. This is not conducive to the creative application of experience information by firms in different environment.

In model 10, the interaction between cohesion and cognitive proximity was added. Therefore, cognitive proximity positively moderated the cohesion on experience information dissemination. Hypothesis 5a was supported.

In model 11, the interaction between status difference and cognitive proximity is added. Therefore, cognitive proximity negatively moderates the status difference on experience information dissemination. Hypothesis 5b is not supported. Status difference exist will form the differences of information stock within the community, whereas which will be narrowed by the cognitive proximity increasing. In the experience information dissemination process, the firms too familiar may lead to a high level of experience rewards decrease [63]. Thus, the cognitive proximity has a negative impact of status differences on experience information dissemination.

In model 12, the interaction items of coordination and cognitive proximity is added. The impact of coordination on project information dissemination is negatively regulated by cognitive proximity. Hypothesis 5c is not supported. The possible reason is that the cognitive proximity and short path length will lead to the intimate feeling between the firms in the VC network community. This will limit their creativity [64] and prevent them from creatively adopting the experience information coming from the management of the invested enterprises. Thus, these results in a decrease in the community experience information dissemination.

Further analysis

We verify the influence of network community structure on information dissemination in a three-year window. We are also concerned about how long a network community structure can influence the information dissemination. That is to say, the network community structure exist the lag effect. So here we examine the lag effect of independent variables on dependent variables for 1 year, 2 years and 3 years. The test could also verify the robustness of our conclusions. The ridge regressions were used for all models. As the moderator variables only affect the extent of independent variables in the main effect model, only the regression results of the main effect are reported in the lag effect model in Tables 4 and 5.

The result of project information dissemination in lag 1 year, 2 years and 3 years

T+1 T+2 T+3

Variables Model 13 Model 14 Model 15 Model 16 Model 17 Model 18 Model 19 Model 20 Model 21

Scale 0.242 (0.148) 0.051 (0.138) 0.074 (0.134) 0.083 (0.152) 0.072 (0.146) 0.109 (0.141) 0.018 (0.145) 0.028 (0.142) 0.009 (0.136)
Community average resource endowment 0.081 (0.152) 0.131 (0.137) 0.112 (0.132) 0.105 (0.155) 0.195 (0.143) 0.176 (0.141) 0.212 (0.248) 0.285 (0.132) 0.288 (0.127)
Community average industry diversity −0.088 (0.162) −0.049 (0.143) −0.151 (0.144) −0.137 (0.165) −0.111 (0.149) −0.022 (0.154) −0.221 (0.158) −0.203 (0.139) −0.301 (0.139)
Community average stage diversity 0.066 (0.164) 0.004 (0.144) 0.079 (0.144) 0.002 (0.167) 0.039 (0.150) 0.165 (0.152) 0.031 (0.159) 0.091 (0.141) 0.032 (0.139)
Cohesion −0.249* (0.174) −0.248* (0.169) −0.265* (0.207) −0.242* (0.204) −0.259* (0.148) −0.243* (0.143)
Status difference 0.326** (0.180) 0.307** (0.174) 0.277* (0.167) 0.261* (0.163) 0.260* (0.195) 0.266* (0.188)
Coordination −0.357** (0.171) −0.305** (0.168) −0.240* (0.140) −0.237* (0.138) −0.264* (0.129) −0.270** (0.124)
Cognitive proximity −0.273* (0.133) −0.270* (0.253) −0.240* (0.164)
R2 0.094 0.368 0.429 0.055 0.302 0.353 0.138 0.399 0.463
R2 0.029 0.197 0.246 0.015 0.113 0.146 0.020 0.238 0.292
F 0.760 2.159* 2.346** 0.423 1.603* 1.703* 1.165 2.475** 2.699**
K 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2

Notes: Standard errors in parentheses.

p < 0.1,

p < 0.05,

p < 0.01. K is the value of ridge regression.

The result of experience information dissemination in lag 1 year, 2 years and 3 years

T+1 T+2 T+3

Variables Model 22 Model 23 Model 24 Model 25 Model 26 Model 27 Model 28 Model 29 Model 30

Scale 0.025 (0.150) 0.212 (0.134) 0.167 (0.126) 0.134 (0.151) 0.131 (0.141) 0.151 (0.135) 0.555 (0.118) 0.268 (0.123) 0.289 (0.118)
Community average resource endowment 0.083 (0.154) 0.002 (0.133) 0.013 (0.126) 0.042 (0.155) 0.066 (0.136) 0.046 (0.130) 0.016 (0.121) 0.041 (0.114) 0.018 (0.110)
Community average industry diversity −0.205 (0.163) −0.199 (0.138) −0.346 (0.137) −0.121 (0.165) −0.131 (0.142) −0.282 (0.142) −0.057 (0.129) −0.072 (0.120) −0.173 (0.121)
Community average stage diversity 0.041 (0.165) 0.045 (0.139) 0.097 (0.136) 0.174 (0.167) 0.321 (0.144) 0.263 (0.143) 0.023 (0.130) 0.102 (0.122) 0.061 (0.120)
Cohesion −0.315** (0.169) −0.331** (0.161) −0.292** (0.150) −0.318** (0.145) −0.217* (0.168) −0.248** (0.163)
Status difference 0.332** (0.174) 0.339** (0.166) 0.273* (0.169) 0.269* (0.163) 0.217* (0.167) 0.220* (0.160)
Coordination −0.231* (0.166) −0.234* (0.158) −0.234* (0.185) −0.248* (0.178) −0.225* (0.114) −0.233** (0.110)
Cognitive proximity −0.272* (0.195) −0.276* (0.192) −0.208* (0.145)
R2 0.075 0.404 0.485 0.061 0.365 0.436 0.428 0.548 0.597
R2 0.052 0.244 0.320 0.019 0.194 0.256 0.349 0.427 0.468
F 0.590 2.522** 2.941** 0.470 2.133* 2.419** 5.418 4.512*** 4.627***
K 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2

Notes: Standard errors in parentheses.

p < 0.1,

p < 0.05,

p < 0.01. K is the value of ridge regression.

The models 13, 16 and 19 are the effects of control variables on project information dissemination in lag 1 year, 2 years and 3 years, respectively. The models 14, 17 and 20 are the regression results of adding independent variables to the above model, respectively. The models 15, 18 and 21 are the regression results of adding moderator variables to the above model, respectively.

The models 22, 25 and 28 are the effects of control variables on experience information dissemination in lag 1 year, 2 years and 3 years, respectively. The models 23, 26 and 29 are the regression results of adding independent variables to the above model, respectively. The models 24, 27 and 30 are the regression results of adding moderator variables to the above model, respectively.

We calculated the increment of project information with 1 year, 2 years and 3 years lag, respectively, through the regression coefficient in the main effect model of project information. In the main effect model with 1 year lag, when three independent variables increase a standard deviation from their mean, the increment of project information dissemination is 5.5%. In the two-year lag main effect model, when three independent variables increase two standard deviations from their mean, the increment of project information dissemination is 1.6%. In the three-year lag main effect model, when the three independent variables increase three standard deviations from their mean, the increment of project information dissemination is 2.6%. The incremental results of project information dissemination showed that information dissemination has reached a peak in the lag of 1 year. However the increment of project information dissemination in the following 2 years is decreasing.

We calculated the increment of experience information with 1 year, 2 years and 3 years lags, respectively, through the regression coefficient in the main effect model of project information. In the main effect model with one year lag, when three independent variables increase a standard deviation from their mean, the increment of project information dissemination is 7.8%. In the two-year lag main effect model, when three independent variables increase two standard deviations from their mean, the increment of project information dissemination is 1.6%. In the three-year lag main effect model, when the three independent variables increase three standard deviations from their mean, the increment of project information dissemination is 2.6%. The incremental results of experience information dissemination showed that information dissemination has reached a peak in the lag of 1 year. However the increment of experience information dissemination in the following 2 years is decreasing. Based on the above results, we find that the effect of network community structure on the two kinds of information dissemination will weaken with the time increasing. That is to say, one kind of network structure can promote the information dissemination. However the information dissemination through this kind of structure can only make VC firms obtain limited benefits.

Our findings suggest that the impact of cohesion on information dissemination is greater than status difference’s. Status difference’s is greater than coordination’s. In order to deeply understand the influence of network community structure dimension on information dissemination, this paper mainly examines how different combinations of cohesion, status differences and coordination will lead to change in information dissemination. We assumed that the network community structure dimensions are divided into two categories: high and low. The four combinations are hold such as high cohesion and high status difference, high cohesion and low status difference, low cohesion and high status difference, low cohesion and low status difference. We suggest that high cohesion is mean adding a standard deviation, then low cohesion is mean subtracting a standard deviation [65]. The same approach is applied to status differences and coordination. In fact, high cohesion and high coordination (i.e., short average path) change synchronously. Low cohesion and low coordination (i.e., long average paths) change synchronously. Therefore, there are only four types of structural dimension matching. Table 6 shows the value of project information and experience information after structural dimension being matched. When the network community structure is composed of high cohesion, high status difference and high coordination, the value of project information dissemination and experience information dissemination are 1.015 and 1.063. When the network community structure is composed of low cohesion, high status difference and low coordination, the value of project information dissemination and experience information dissemination are 1.055 and 1.102. Further the value of project information and experience information can also increase when the mean increase or decrease two standard deviation. At the same time, it is found that the other two types of network community structure decrease information dissemination.

Influence of different combination of cohesion and status difference on information dissemination

High × High × High High × Low × High Low × High × Low Low × Low × Low

A standard deviation Two standard deviation A standard deviation Two standard deviation A standard deviation Two standard deviation A standard deviation Two standard deviation

Project information dissemination value 1.015 1.270 0.321 0.177 1.055 1.203 0.361 0.253
Experience information dissemination value 1.063 1.218 0.591 0.250 1.102 1.297 0.630 0.157

One kind of network community structure benefitting information dissemination is high cohesion and high status difference. High cohesion reflects that firms in the network community are closely linked and more interaction. There are many ways to transmit project information and experience information quickly. In other words, high cohesion means that many neighbours can get the spreading information. High status differences mean the existence of a ‘leader’. This ‘leader’ owning rich investment experience and skills can strongly influence the rest of the firms. Then, the network community could reflect a strong cluster impact. In other words, seeing as a hub ‘leader’ can accelerate the speed of information dissemination. Because holding priority to receive and disseminate information, which can also induce other firms to receive and disseminate information.

Another kind of network community structure benefitting information dissemination is low-cohesion and high-status difference. Low cohesion means less firm connection, less interaction and less redundancy. Further, the information disseminated in the network community has a certain degree of ‘simplicity’. High-status differences indicate the existence of ‘leader’. This widely influences social interaction and deeply impacts on cluster effect, and then promotes or suppresses information dissemination by normative pressure.

According to the results of Table 6, the VC network community structure that promotes information dissemination is divided into two categories. The one category is high-cohesion, high-status difference and high coordination, the other is low-cohesion, high-status difference and low coordination. The VC network community structure suppressing information dissemination is divided into two categories. The one category is low-cohesion, low-status difference and low coordination, the other is high-cohesion, low-status difference and high coordination.

Conclusion

Through the research on the relationship of community structure, information dissemination and investment performance of the medical health VC network community, this paper draws the following conclusions.

First, from the perspective of meso-level, this paper reveals the influence mechanism of community structure on information dissemination. Some studies have examined the impact of VC networks on investment performance from the perspective of ego-centric network. They concerned with how VC firms disseminate information in self-central networks through partners directly or indirectly related to themselves. From the perspective of the whole network, some studies are concerned about the benefits of knowledge diffusion through the wider social space. However, from the meso-level, our research focuses on how to disseminate information by using the characteristics of the network community internal connection is tight and the external connection is sparse. This paper examined the network community structure and information dissemination from a completely new perspective.

Second, combining the social network theory and resource-based theory, this paper explored a way to improve the investment performance of medical health VC network community: VC network community structure – information dissemination – investment performance. While some studies believe that network structure [55] and network resources can affect the members’ investment performance [32]. There is no clarity about what kind of resources are at work. Clercq and Sapienza [66] found that the acquisition of internal and external knowledge by VC firms can prompt investment performance. However, this conclusion is not under the background of VC network community. Therefore, this paper makes a reasonable assumption based on the above literatures. The information dissemination can connect VC network community structure and investment performance. Our findings expand the functional research of VC network community.

Third, this paper divided VC network community structure and information dissemination dimension for the first time. Bubna et al. [5] only focussed on the detection of VC network community, and then the following study has not been carried out. In this paper, the dimension of VC network community structure is divided into cohesion, status difference and coordination complete describing the characteristics of the network community structure without isolated nodes. In this paper, the information dissemination first divided into project information dissemination and experience information dissemination. These conclusions lay the foundation for further research on VC network community.

Fourth, this paper found that the network community structure dimension has different effects and intensity on information dissemination. Cohesion inhibits the two types of information dissemination. This result is contrary to the conclusion of Mishori et al. [9]_ENREF_44 on information dissemination on the network of American doctors. Because this paper studies the Chinese medical health industry VC community, in order to prevent risks and make up the funding gap. The firms always contact very closely, resulting in excessive cohesion. The status difference and coordination promoted the dissemination of two kinds of information. This result is consistent with the findings of Shaikh et al. [10] on innovation diffusion under small-world networks and the Cowana and Jonard [11] on knowledge dissemination in small world networks. Therefore, the research on information dissemination in network community structure is from the previous research in other fields to VC field. This paper found that cohesion has a stronger influence on two types of information dissemination than status differences; meanwhile status difference has a stronger influence on two types of information dissemination than coordination’s. These conclusions have never been mentioned in the previous studies. Our work can lay the foundation for further research of VC network community.

Last, this paper discussed the cognitive proximity moderator on the basis of the above study. We found that cognitive proximity could promote or hinder the community structure on information dissemination. Our result has the certain difference with the research conclusion of Gavetti [19]_ENREF_16, Gulati [42]_ENREF_21 and Giuliani and Bell [47]. All these scholars believe that cognitive proximity is beneficial to information dissemination. However, we believe that cognitive proximity is unfavourable to information dissemination in some environments. Furthermore, our work refined the research of cognitive proximity in VC field.

In addition to the above theoretical contributions, this paper also has certain practical significance for promoting medical health VC.

First, because the network community can promote the VC firms performance, it can be used as the incubator of VC firms in practice. Government formulates appropriate settlement policies and tax incentives in order to attract new VC firms into the community. Or it builds a path to the network community for VC firms that need to develop or expand their investment field. It helps VC firm build relationships with others in the network community to help weaker VC firm access resources. At the same time, the government can also develop preferential policies to give back to the stronger VC firms in the community. Maintaining or enhancing the operation of VC network community give full play to the role of incubator.

Second, in this paper we find that there are differences in the function and intensity of network community structure dimension. Cohesion is the strongest structure dimension. Therefore, in practice, we should pay attention to establish a relationship with other VC firms in the community. However, the relations do not too close maintaining a moderate range. The role of status difference is lower than cohesion. Then it is necessary to maintain ‘leader’ in VC network community. Coordination will hold the same change trend as cohesion, so we do not treat coordination especially. Our results also show that the effect of network community structure on information dissemination will reach the highest point after first year and then decrease. Therefore, the network community structure plays the greatest role in first year. Thus, in the second year we begin to use different mechanisms of relationship building or disconnection to change the network community structure. In order to promote the information dissemination we could construct two type of network community structure. The one is high cohesion and high-status difference, the other is low-cohesion and high-status difference. In order to prevent the information dissemination, we could construct two type of network community structure. The one is high-cohesion and low-status difference, the other is low-cohesion and low-status difference.

Third, this paper shows that cognitive proximity plays a role in the VC network community. For example, the government can provide a variety of resources to invite stronger VC firms helping community member improve the knowledge base level, such as holding more salon, lecture or road show activities on site or online. These ways could help to absorb and popularise the experience of VC firms, improve the weaker firm’s knowledge accumulation and shorten the cognitive gap between firms in the network community.

This paper still has the following shortcomings. First, while this paper only considers the direct influence of community structure on information dissemination, it does not consider the existence of mediation variables. Second, there may be other factors affecting the information dissemination under the community structure. Third, our result only suits for the Chinese medical health industry VC network community. In the future, the research could be extended to other fields or other countries.

The result of project information dissemination in lag 1 year, 2 years and 3 years

T+1 T+2 T+3

Variables Model 13 Model 14 Model 15 Model 16 Model 17 Model 18 Model 19 Model 20 Model 21

Scale 0.242 (0.148) 0.051 (0.138) 0.074 (0.134) 0.083 (0.152) 0.072 (0.146) 0.109 (0.141) 0.018 (0.145) 0.028 (0.142) 0.009 (0.136)
Community average resource endowment 0.081 (0.152) 0.131 (0.137) 0.112 (0.132) 0.105 (0.155) 0.195 (0.143) 0.176 (0.141) 0.212 (0.248) 0.285 (0.132) 0.288 (0.127)
Community average industry diversity −0.088 (0.162) −0.049 (0.143) −0.151 (0.144) −0.137 (0.165) −0.111 (0.149) −0.022 (0.154) −0.221 (0.158) −0.203 (0.139) −0.301 (0.139)
Community average stage diversity 0.066 (0.164) 0.004 (0.144) 0.079 (0.144) 0.002 (0.167) 0.039 (0.150) 0.165 (0.152) 0.031 (0.159) 0.091 (0.141) 0.032 (0.139)
Cohesion −0.249* (0.174) −0.248* (0.169) −0.265* (0.207) −0.242* (0.204) −0.259* (0.148) −0.243* (0.143)
Status difference 0.326** (0.180) 0.307** (0.174) 0.277* (0.167) 0.261* (0.163) 0.260* (0.195) 0.266* (0.188)
Coordination −0.357** (0.171) −0.305** (0.168) −0.240* (0.140) −0.237* (0.138) −0.264* (0.129) −0.270** (0.124)
Cognitive proximity −0.273* (0.133) −0.270* (0.253) −0.240* (0.164)
R2 0.094 0.368 0.429 0.055 0.302 0.353 0.138 0.399 0.463
R2 0.029 0.197 0.246 0.015 0.113 0.146 0.020 0.238 0.292
F 0.760 2.159* 2.346** 0.423 1.603* 1.703* 1.165 2.475** 2.699**
K 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2

Influence of different combination of cohesion and status difference on information dissemination

High × High × High High × Low × High Low × High × Low Low × Low × Low

A standard deviation Two standard deviation A standard deviation Two standard deviation A standard deviation Two standard deviation A standard deviation Two standard deviation

Project information dissemination value 1.015 1.270 0.321 0.177 1.055 1.203 0.361 0.253
Experience information dissemination value 1.063 1.218 0.591 0.250 1.102 1.297 0.630 0.157

Descriptive statistics of variables and Pearson correlation matrix

Variable name Average Standard deviation Observed quantity Pearson correlation matrix

Project information dissemination 0.883 0.701 100 1
Experience information dissemination 1.531 1.353 100 0.688 1
Community cohesion 0.709 0.065 100 −0.328 −0.316 1
Community status difference 3.603 1.011 100 0.275 0.285* −0.180 1
Community coordination 1.667 1.031 100 −0.295 −0.129 0.164* −0.042 1
Cognitive proximity 0.241 .087 100 −0.114 −0.008 0.258 −0.105 0.219 1
Community scale 1.815 0.236 100 0.158 0.300 −0.259−0.204 0.665** −0.290 −0.286 1
Community average resource endowment 4.444 1.526 100 0.252 0.074 0.115 −0.046 −0.175* 0.169 1
Community average industry diversity 0.295 0.100 100 −0.085 −0.073 0.068 −0.071 0.052 0.062 −0.215 −0.693** 1
Community average stage diversity 0.291 .063 100 0.110 0.034 −0.089 0.005 −0.277 −0.013 0.350* 0.697** −0.682** 1

The result of experience information dissemination in lag 1 year, 2 years and 3 years

T+1 T+2 T+3

Variables Model 22 Model 23 Model 24 Model 25 Model 26 Model 27 Model 28 Model 29 Model 30

Scale 0.025 (0.150) 0.212 (0.134) 0.167 (0.126) 0.134 (0.151) 0.131 (0.141) 0.151 (0.135) 0.555 (0.118) 0.268 (0.123) 0.289 (0.118)
Community average resource endowment 0.083 (0.154) 0.002 (0.133) 0.013 (0.126) 0.042 (0.155) 0.066 (0.136) 0.046 (0.130) 0.016 (0.121) 0.041 (0.114) 0.018 (0.110)
Community average industry diversity −0.205 (0.163) −0.199 (0.138) −0.346 (0.137) −0.121 (0.165) −0.131 (0.142) −0.282 (0.142) −0.057 (0.129) −0.072 (0.120) −0.173 (0.121)
Community average stage diversity 0.041 (0.165) 0.045 (0.139) 0.097 (0.136) 0.174 (0.167) 0.321 (0.144) 0.263 (0.143) 0.023 (0.130) 0.102 (0.122) 0.061 (0.120)
Cohesion −0.315** (0.169) −0.331** (0.161) −0.292** (0.150) −0.318** (0.145) −0.217* (0.168) −0.248** (0.163)
Status difference 0.332** (0.174) 0.339** (0.166) 0.273* (0.169) 0.269* (0.163) 0.217* (0.167) 0.220* (0.160)
Coordination −0.231* (0.166) −0.234* (0.158) −0.234* (0.185) −0.248* (0.178) −0.225* (0.114) −0.233** (0.110)
Cognitive proximity −0.272* (0.195) −0.276* (0.192) −0.208* (0.145)
R2 0.075 0.404 0.485 0.061 0.365 0.436 0.428 0.548 0.597
R2 0.052 0.244 0.320 0.019 0.194 0.256 0.349 0.427 0.468
F 0.590 2.522** 2.941** 0.470 2.133* 2.419** 5.418 4.512*** 4.627***
K 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2

Regression results of the project information dissemination regression model

Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Scale 0.110 (0.143) 0.074 (0.142) 0.039 (0.139) 0.100 (0.133) 0.102 (0.137) 0.062 (0.131)
Community average resource endowment 0.191 (0.148) 0.132 (0.134) 0.153 (0.133) 0.149 (0.128) 0.097 (0.130) 0.137 (0.127)
Community average industry diversity −0.266 (1.580) −0.188 (0.141) −0.051 (0.144) −0.083 (0.137) −0.028 (0.140) −0.050 (0.138)
Community average stage diversity 0.209 (0.160) 0.133 (0.143) 0.197 (0.145) 0.258 (0.140) 0.246 (0.141) 0.226 (0.142)
Cohesion −0.307** (0.188) −0.303** (0.186) −0.336** (0.179) −0.304** (0.179) −0.309** (0.177)
Status difference 0.299** (0.183) 0.271* (0.180) 0.306** (0.171) 0.228* (0.176) 0.326* (0.171)
Coordination −0.274* (0.319) −0.299* (0.319) −0.287** (0.305) −0.287* (0.307) −0.282* (0.308)
Cognitive proximity −0.257* (0.191) −0.254* (0.183) −0.247** (0.185) −0.233* (0.222)
Cohesion × cognitive proximity 0.214* (0.222)
Status difference × cognitive proximity −0.250* (0.225)
Coordination × cognitive proximity 0.230* (0.342)
R2 0.149 0.386 0.423 0.494 0.486 0.498
R2 0.032 0.221 0.238 0.304 0.293 0.311
F 1.273 2.341* 2.289* 2.599** 2.522** 2.655**
K 0.2 0.2 0.2 0.2 0.2 0.2

Results of the experience information dissemination regression model

Variables Model 7 Model 8 Model 9 Model 10 Model 11 Model 12

Scale 0.261 (0.145) 0.098 (0.144) 0.061 (0.140) 0.033 (0.129) 0.081 (0.132) 0.056 (0.133)
Community average resource endowment 0.106 (0.149) 0.013 (0.136) 0.041 (0.131) 0.057 (0.125) 0.024 (0.124) 0.115 (0.128)
Community average industry diversity −0.175 (0.160) −0.198 (0.145) −0.297 (0.145) −0.289 (0.140) −0.289 (0.139) −0.281 (0.139)
Community average stage diversity 0.107 (0.161) 0.196 (0.146) 0.117 (0.144) 0.132 (0.139) 0.129 (0.137) 0.118 (0.137)
Cohesion −0.261** (0.227) −0.301** (0.222) −0.350** (0.215) −0.296** (0.211) −0.315** (0.212)
Status difference 0.317** (0.158) 0.318** (0.153) 0.266** (0.144) 0.397** (0.147) 0.338** (0.146)
Coordination −0.277* (0.149) −0.284** (0.145) −0.258* (0.138) −0.308* (0.137) −0.359** (0.141)
Cognitive proximity −0.255* (0.165) −0.374* (0.157) −0.294* (0.158) −0.315* (0.158)
Cohesion × cognitive proximity 0.241* (0.274)
Status difference × cognitive proximity −0.219* (0.199)
Coordination × cognitive proximity 0.225* (0.223)
R2 0.132 0.363 0.427 0.495 0.501 0.499
R2 0.012 0.191 0.243 0.305 0.314 0.311
F 1.103 2.114* 2.328* 2.610** 2.682** 2.659**
K 0.2 0.2 0.2 0.2 0.2 0.2

Japsen B, 1998, “Healthcare venture capital funding keeps record pace,” Modern Healthcare Vol. 28, pp 40. JapsenB 1998 “Healthcare venture capital funding keeps record pace,” Modern Healthcare 28 40 Search in Google Scholar

Bruce, C & Bruce, S, 2002, “Helping healthcare entrepreneurs: A case study of Angel Healthcare Investors, LLC,” Venture Capital Vol. 4 pp 325–330. BruceC BruceS 2002 “Helping healthcare entrepreneurs: A case study of Angel Healthcare Investors, LLC,” Venture Capital 4 325 330 10.1080/1369106022000024941 Search in Google Scholar

Arthur, K, 2005, “Biotech venture capital-it’s not too late to be early,” Nature biotechnology Vol. 23, pp 417–418. ArthurK 2005 “Biotech venture capital-it’s not too late to be early,” Nature biotechnology 23 417 418 10.1038/nbt0405-41715815662 Search in Google Scholar

Khanin, D, Gnanlet, A & Leibsohn, D, 2012, “Venture capital-backed foreign portfolio company’s IPO,” International Journal of Entrepreneurship and Innovation Management Vol. 16, No. 3/4, pp 263–284. KhaninD GnanletA LeibsohnD 2012 “Venture capital-backed foreign portfolio company’s IPO,” International Journal of Entrepreneurship and Innovation Management 16 3/4 263 284 10.1504/IJEIM.2012.051963 Search in Google Scholar

Bubna, A, Das, S. R & Prabhala, N, 2019, “Venture Capital Communities,” Journal Finance Quantly, pp 1–73. BubnaA DasS. R PrabhalaN 2019 “Venture Capital Communities,” Journal Finance Quantly 1 73 Search in Google Scholar

Tatarynowicz, A, Sytch, M & Gulati, R, 2016, “Environmental Demands and the Emergence of Social Structure: Technological Dynamism and Interorganizational Network Forms,” Administrative Science Quarterly, pp 52–86. TatarynowiczA SytchM GulatiR 2016 “Environmental Demands and the Emergence of Social Structure: Technological Dynamism and Interorganizational Network Forms,” Administrative Science Quarterly 52 86 10.1177/0001839215609083 Search in Google Scholar

Huber, G P, 1991, “Organizational learning: the contributing processes and the literatures,” Organization Science Vol. 2, pp 88–115. HuberG P 1991 “Organizational learning: the contributing processes and the literatures,” Organization Science 2 88 115 10.1287/orsc.2.1.88 Search in Google Scholar

Newman, M, 2003, “The structure and function of complex networks,” SIAM Review Vol. 45, pp 167–256. NewmanM 2003 “The structure and function of complex networks,” SIAM Review 45 167 256 10.1137/S003614450342480 Search in Google Scholar

Mishori, R, Singh, L.O & Levy, B, 2014 “Mapping Physician Twitter Networks: Describing How They Work as a First Step in Understanding Connectivity, Information Flow, and Message Diffusion,” Journal of Medical Internet Research Vol. 16, No. 4, pp 157–167. MishoriR SinghL.O LevyB 2014 “Mapping Physician Twitter Networks: Describing How They Work as a First Step in Understanding Connectivity, Information Flow, and Message Diffusion,” Journal of Medical Internet Research 16 4 157 167 10.2196/jmir.3006400413624733146 Search in Google Scholar

Shaikh, N I, Rangaswamy, A & Balakrishnan, A, 2012, “Modeling the diffusion of innovations through small-world networks,” Social Science Electronic Publishing, pp 47–58. ShaikhN I RangaswamyA BalakrishnanA 2012 “Modeling the diffusion of innovations through small-world networks,” Social Science Electronic Publishing 47 58 10.2139/ssrn.2032861 Search in Google Scholar

Cowana, R & Jonard, N, 2004, “Network structure and the diffusion of knowledge,” Journal of Economic Dynamics & Control, pp 1557–1575. CowanaR JonardN 2004 “Network structure and the diffusion of knowledge,” Journal of Economic Dynamics & Control 1557 1575 10.1016/j.jedc.2003.04.002 Search in Google Scholar

Munari, F, Malipiero, A & Sobrero, M, 2012, “Focal firms as technological gatekeepers within industrial districts: knowledge creation and dissemination in the Italian packaging machinery industry,” Industrial and Corporate Change Vol. 21, pp 429–462. MunariF MalipieroA SobreroM 2012 “Focal firms as technological gatekeepers within industrial districts: knowledge creation and dissemination in the Italian packaging machinery industry,” Industrial and Corporate Change 21 429 462 10.1093/icc/dtr053 Search in Google Scholar

Kogut, B & Zander, U, 2003, “Knowledge of the firm and the evolutionary theory of the multinational corporation,” Journal of International Business Studies Vol. 34, pp 516–529. KogutB ZanderU 2003 “Knowledge of the firm and the evolutionary theory of the multinational corporation,” Journal of International Business Studies 34 516 529 10.1057/palgrave.jibs.8400058 Search in Google Scholar

Jiang, Y C, 2014, “Diffusion in Social Networks: A Multiagent Perspective,” IEEE Transactions on Systems Vol. 45, pp 198–213. JiangY C 2014 “Diffusion in Social Networks: A Multiagent Perspective,” IEEE Transactions on Systems 45 198 213 10.1109/TSMC.2014.2339198 Search in Google Scholar

Argote, L & Fahrenkopf, E, 2016, “Knowledge transfer in organizations: The roles of members, tasks, tools, and networks,” Organizational Behavior and Human Decision Processes Vol. 136, pp 146–159. ArgoteL FahrenkopfE 2016 “Knowledge transfer in organizations: The roles of members, tasks, tools, and networks,” Organizational Behavior and Human Decision Processes 136 146 159 10.1016/j.obhdp.2016.08.003 Search in Google Scholar

Wasserman, S & Faust, K, 1994, Social Network Analysis: Methods and Applications, Cambridge: Cambridge University Press. WassermanS FaustK 1994 Social Network Analysis: Methods and Applications Cambridge Cambridge University Press 10.1017/CBO9780511815478 Search in Google Scholar

Kaplan, S & Stromberg, P, 2001, “Venture capitalists as principals: Contracting, screening, and monitoring,” American Economic Reviewpoint Vol. 91, pp 426–430. KaplanS StrombergP 2001 “Venture capitalists as principals: Contracting, screening, and monitoring,” American Economic Reviewpoint 91 426 430 10.3386/w8202 Search in Google Scholar

Lane, P J, Koka, B R & Pathak, S, 2006, “The Reification of Absorptive Capacity: A Critical Reviewpoint and Rejuvenation of the Construct,” The Academy of Management Reviewpoint Vol. 31, pp 833–863. LaneP J KokaB R PathakS 2006 “The Reification of Absorptive Capacity: A Critical Reviewpoint and Rejuvenation of the Construct,” The Academy of Management Reviewpoint 31 833 863 10.5465/amr.2006.22527456 Search in Google Scholar

Gavetti, G, 2018, “On the Origin of Great Strategies,” Strategy Science Vol. 3, pp 352–365. GavettiG 2018 “On the Origin of Great Strategies,” Strategy Science 3 352 365 10.1287/stsc.2018.0054 Search in Google Scholar

Nahapiet, J & Ghoshal, S, 1998, “Social Capital, Intellectual Capital and the Organizational Advantage,” Academy of Management Reviewpoint Vol. 23, pp 242–266. NahapietJ GhoshalS 1998 “Social Capital, Intellectual Capital and the Organizational Advantage,” Academy of Management Reviewpoint 23 242 266 10.1016/B978-0-7506-7222-1.50009-X Search in Google Scholar

Mas, D, Vignes, A & Weisbuch, G, 2008, “Syndication Strategies in Venture Capital Networks,” Strategy and Governance of Networks, pp 373–391. MasD VignesA WeisbuchG 2008 “Syndication Strategies in Venture Capital Networks,” Strategy and Governance of Networks 373 391 10.1007/978-3-7908-2058-4_20 Search in Google Scholar

Coleman, J, 1988, “Social capital in the creation of human capital,” American Journal of Sociology Vol. 94, pp 95–120. ColemanJ 1988 “Social capital in the creation of human capital,” American Journal of Sociology 94 95 120 10.1086/228943 Search in Google Scholar

Granovetter, M S, 1973 “The strength of weak ties,” American Journal of Sociology Vol. 78, pp 1360–1380. GranovetterM S 1973 “The strength of weak ties,” American Journal of Sociology 78 1360 1380 10.1086/225469 Search in Google Scholar

Uzzi, B & Lancaster, R, 2003, “Relational embeddedness and learning: The case of bank loan managers and their clients,” Management Science Vol. 49, pp 383–399. UzziB LancasterR 2003 “Relational embeddedness and learning: The case of bank loan managers and their clients,” Management Science 49 383 399 10.1287/mnsc.49.4.383.14427 Search in Google Scholar

Zaheer, A, Gözübüyük, R & Milanov, H, 2010, “It’s the Connections: The Network Perspective in Inter-organizational Research,” Academy of Management Perspectives Vol. 24, pp 62–77. ZaheerA GözübüyükR MilanovH 2010 “It’s the Connections: The Network Perspective in Inter-organizational Research,” Academy of Management Perspectives 24 62 77 10.5465/amp.24.1.62 Search in Google Scholar

Das, T K & Teng, B S, 2000, “A Resource-based Theory of Strategic Alliances,” Journal of Management Vol. 26, pp 31–61. DasT K TengB S 2000 “A Resource-based Theory of Strategic Alliances,” Journal of Management 26 31 61 10.1177/014920630002600105 Search in Google Scholar

Walter, W, Douglas, P R, White, K W & Owen-Smith, J K, 2005, “Network Dynamics and Field Evolution: The Growth of Interorganizational Collaboration in the Life Sciences,” American Journal of Sociology, pp 110. WalterW DouglasP R WhiteK W Owen-SmithJ K 2005 “Network Dynamics and Field Evolution: The Growth of Interorganizational Collaboration in the Life Sciences,” American Journal of Sociology 110 10.1086/421508 Search in Google Scholar

Hansen, M T & Birkinshaw, H J, 2007, “The Innovation Value Chain,” Harvard Business Reviewpoint, pp 121–130. HansenM T BirkinshawH J 2007 “The Innovation Value Chain,” Harvard Business Reviewpoint 121 130 Search in Google Scholar

Gomez-Rodriguez, M, Leskovec, J & Krause, A, 2012, “Inferring Networks of Diffusion and Influence,” ACM Transactions on Knowledge Discovery from Data Vol. 5 pp 1–37. Gomez-RodriguezM LeskovecJ KrauseA 2012 “Inferring Networks of Diffusion and Influence,” ACM Transactions on Knowledge Discovery from Data 5 1 37 10.1145/1835804.1835933 Search in Google Scholar

Nonaka, I & Toyama, R, 2005, “The theory of the knowledge-creating firm: subjectivity, objectivity and synthesis,” Industrial and Corporate Change Vol. 14, pp 419–436. NonakaI ToyamaR 2005 “The theory of the knowledge-creating firm: subjectivity, objectivity and synthesis,” Industrial and Corporate Change 14 419 436 10.1093/icc/dth058 Search in Google Scholar

Scott, J, 1991 Social Network Analysis – A Handbook, London: Sage Publications. ScottJ 1991 Social Network Analysis – A Handbook London Sage Publications Search in Google Scholar

Gulati, R, Sytch, M & Tatarynowicz, A, 2012, “The rise and fall of small worlds: Exploring the dynamics of social structure” Organization Science Vol. 23, pp 449–471. GulatiR SytchM TatarynowiczA 2012 “The rise and fall of small worlds: Exploring the dynamics of social structure” Organization Science 23 449 471 10.1287/orsc.1100.0592 Search in Google Scholar

Sytch, M & Tatarynowicz, A, 2014, “Exploring the locus of invention: The dynamics of network communities and firms’ invention productivity,” Academy of Management Journal Vol. 57, pp 249–279. SytchM TatarynowiczA 2014 “Exploring the locus of invention: The dynamics of network communities and firms’ invention productivity,” Academy of Management Journal 57 249 279 10.5465/amj.2011.0655 Search in Google Scholar

Burt, R, 1992, Structural Holes: The Social Structure of Competition, Cambridge: Harvard University Press. BurtR 1992 Structural Holes: The Social Structure of Competition Cambridge Harvard University Press 10.4159/9780674029095 Search in Google Scholar

Ingram, P & Morris, M W, 2007, “Do People Mix at Mixers? Structure, Homophily, and the “Life of the Party”,” Administrative Science Quarterly Vol. 52 pp 558–585. IngramP MorrisM W 2007 “Do People Mix at Mixers? Structure, Homophily, and the “Life of the Party”,” Administrative Science Quarterly 52 558 585 10.2189/asqu.52.4.558 Search in Google Scholar

Winter, S & Teece, D J, 1987, Knowledge and competence as strategic assets. The Competitive Challenge-Strategies for Industrial Innovation and Renewal, Cambridge: Ballinger. WinterS TeeceD J 1987 Knowledge and competence as strategic assets. The Competitive Challenge-Strategies for Industrial Innovation and Renewal Cambridge Ballinger Search in Google Scholar

Newman, M & Girvan, M, 2004, “Finding and evaluating community structure in networks,” Physical Reviewpoint E, pp 69. NewmanM GirvanM 2004 “Finding and evaluating community structure in networks,” Physical Reviewpoint E 69 10.1103/PhysRevE.69.026113 Search in Google Scholar

Singh, J & Marx, M, 2013, “Geographic Constraints on Knowledge Spillovers: Political Borders vs. Spatial Proximity,” Management Science Vol. 59, pp 2056–2078. SinghJ MarxM 2013 “Geographic Constraints on Knowledge Spillovers: Political Borders vs. Spatial Proximity,” Management Science 59 2056 2078 10.1287/mnsc.1120.1700 Search in Google Scholar

Dacin, M T, Oliver, C & Roy, J P, 2007, “The legitimacy of strategic alliances: An institutional perspective,” Strategic Management Journal Vol. 28, pp 169–187. DacinM T OliverC RoyJ P 2007 “The legitimacy of strategic alliances: An institutional perspective,” Strategic Management Journal 28 169 187 10.1002/smj.577 Search in Google Scholar

McCarthy, I P & Ruckman, K, 2017, “Licensing speed: Its determinants and payoffs,” Journal of Engineering and Technology Management Vol. 46, pp 52–66. McCarthyI P RuckmanK 2017 “Licensing speed: Its determinants and payoffs,” Journal of Engineering and Technology Management 46 52 66 10.1016/j.jengtecman.2017.11.002 Search in Google Scholar

Gomes-Casseres, B, Jaffe, A B & Hagedoorn, J, 2006, “Do alliances promote knowledge flows?,” Journal of Financial Economics Vol. 80, pp 5–33. Gomes-CasseresB JaffeA B HagedoornJ 2006 “Do alliances promote knowledge flows?,” Journal of Financial Economics 80 5 33 10.1016/j.jfineco.2004.08.011 Search in Google Scholar

Gulati, R, 1998, “Alliances and networks,” Strategic Management Journal Vol. 19, pp 293–317. GulatiR 1998 “Alliances and networks,” Strategic Management Journal 19 293 317 10.1002/(SICI)1097-0266(199804)19:4<293::AID-SMJ982>3.0.CO;2-M Search in Google Scholar

Hippel, E, 2007, “Horizontal innovation networks by and for users,” Industrial and Corporate Change Vol. 16, No. 2, pp 1–26. HippelE 2007 “Horizontal innovation networks by and for users,” Industrial and Corporate Change 16 2 1 26 10.1093/icc/dtm005 Search in Google Scholar

Coleman, J, 2003, Networks in the knowledge economy: Oxford University Press. ColemanJ 2003 Networks in the knowledge economy Oxford University Press Search in Google Scholar

Podolny, J M, 2005, Status signals: A sociological study of market competition: Princeton University Press. PodolnyJ M 2005 Status signals: A sociological study of market competition Princeton University Press Search in Google Scholar

Lin, C, Wu, Y J, Chang, C, Wang, W & Lee, C Y, 2012, “The alliance innovation performance of R&D alliances: The absorptive capacity perspective,” Technovation Vol. 32, pp 282–292. LinC WuY J ChangC WangW LeeC Y 2012 “The alliance innovation performance of R&D alliances: The absorptive capacity perspective,” Technovation 32 282 292 10.1016/j.technovation.2012.01.004 Search in Google Scholar

Giuliani, E & Bell, M, 2005, “The micro-determinants of meso-level learning and innovation: evidence from a Chilean wine cluster,” Research Policy Vol. 34, pp 47–68. GiulianiE BellM 2005 “The micro-determinants of meso-level learning and innovation: evidence from a Chilean wine cluster,” Research Policy 34 47 68 10.1016/j.respol.2004.10.008 Search in Google Scholar

Reagans, R & Zuckerman, E, 2008, “Why knowledge does not equal power: the network redundancy trade-off” Industrial and Corporate Change Vol. 17, pp 904–944. ReagansR ZuckermanE 2008 “Why knowledge does not equal power: the network redundancy trade-off” Industrial and Corporate Change 17 904 944 10.1093/icc/dtn036 Search in Google Scholar

Escribano, A, Fosfuri, A & Trib, J A, 2009, “Managing external knowledge flows: the moderating role of absorptive capacity,” Research Policy Vol. 38, pp 96–105. EscribanoA FosfuriA TribJ A 2009 “Managing external knowledge flows: the moderating role of absorptive capacity,” Research Policy 38 96 105 10.1016/j.respol.2008.10.022 Search in Google Scholar

Rietzschel, E F, Zacher, H & Stroebe, W, 2007, “A Lifespan Perspective on Creativity and Innovation at Work,” Work, Aging and Retirement Vol. 2, pp 105–120. RietzschelE F ZacherH StroebeW 2007 “A Lifespan Perspective on Creativity and Innovation at Work,” Work, Aging and Retirement 2 105 120 10.1093/workar/waw005 Search in Google Scholar

Lee, R & Jones, O, 2008, “Networks, communication and learning during business start-up: The creation of social cognitive capital,” International Small Business Journal Vol. 26, pp 559–594. LeeR JonesO 2008 “Networks, communication and learning during business start-up: The creation of social cognitive capital,” International Small Business Journal 26 559 594 10.1177/0266242608094030 Search in Google Scholar

Gupta, A K & Sapienza, H J, 1992, “Determinants of venture capital firms’ preferences regarding the industry diversity and geographic scope of their investments,” Journal of Business Venturing Vol. 7, pp 347–362. GuptaA K SapienzaH J 1992 “Determinants of venture capital firms’ preferences regarding the industry diversity and geographic scope of their investments,” Journal of Business Venturing 7 347 362 10.1016/0883-9026(92)90012-G Search in Google Scholar

Gaba, V & Meyer, A D, 2008, “Crossing the organizational species barrier: How venture capital practices infiltrated the information technology sector,” Academy of Management Journal Vol. 51, pp 976–998. GabaV MeyerA D 2008 “Crossing the organizational species barrier: How venture capital practices infiltrated the information technology sector,” Academy of Management Journal 51 976 998 10.5465/amj.2008.34789671 Search in Google Scholar

Shore, J C, Bernstein, E & Lazer, D, 2015, “Facts and Figuring: An Experimental Investigation of Network Structure and Performance in Information and Solution Spaces,” Organization Science Vol. 26, pp 1432–1446. ShoreJ C BernsteinE LazerD 2015 “Facts and Figuring: An Experimental Investigation of Network Structure and Performance in Information and Solution Spaces,” Organization Science 26 1432 1446 10.1287/orsc.2015.0980 Search in Google Scholar

Hochberg, Y V, Ljungvist, A & Lu, Y, 2007, “Whom You Know Matters: Venture Capital Networks and Investment Performance” Journal of Finance Vol. 62, pp 251–301. HochbergY V LjungvistA LuY 2007 “Whom You Know Matters: Venture Capital Networks and Investment Performance” Journal of Finance 62 251 301 10.1111/j.1540-6261.2007.01207.x Search in Google Scholar

Jaffe, A B, Trajtenberg, M & Henderson, R, 1993, “Geographic localization of knowledge spillovers as evidenced by patent citations,” The Quarterly Journal of Economics Vol. 108, pp 577–598. JaffeA B TrajtenbergM HendersonR 1993 “Geographic localization of knowledge spillovers as evidenced by patent citations,” The Quarterly Journal of Economics 108 577 598 10.3386/w3993 Search in Google Scholar

Yang, Y, Narayanan, V.K & Zahra, S, 2009, “Developing the selection and valuation capabilities through learning: The case of corporate venture capital,” Journal of Business Venturing Vol. 24 pp 261–273. YangY NarayananV.K ZahraS 2009 “Developing the selection and valuation capabilities through learning: The case of corporate venture capital,” Journal of Business Venturing 24 261 273 10.1016/j.jbusvent.2008.05.001 Search in Google Scholar

Levesque, R, 2005, SPSS programming and data management, 2nd edition: A guide for SPSS and SAS users: SPSS Inc, USA. LevesqueR 2005 SPSS programming and data management, 2nd edition: A guide for SPSS and SAS users SPSS Inc USA Search in Google Scholar

Ahuja, G, 2000, “Collaboration Networks, Structural Holes, and Innovation: A Longitudinal Study,” Administrative Science Quarterly Vol. 45, pp 425–455. AhujaG 2000 “Collaboration Networks, Structural Holes, and Innovation: A Longitudinal Study,” Administrative Science Quarterly 45 425 455 10.2307/2667105 Search in Google Scholar

McFadyen, M A, Semadeni, M, Albert, A & Cannella, Jr, 2009, “Value of Strong Ties to Disconnected Others: Examining Knowledge Creation in Biomedicine,” Organization Science Vol. 20, pp 552–564. McFadyenM A SemadeniM AlbertA CannellaJr 2009 “Value of Strong Ties to Disconnected Others: Examining Knowledge Creation in Biomedicine,” Organization Science 20 552 564 10.1287/orsc.1080.0388 Search in Google Scholar

Liu, Y, Luo, Y & Liu, T, 2009, “Governing buyer–supplier relationships through transactional and relational mechanisms: evidence from China,” Journal of Operations Management Vol. 27, pp 294–309. LiuY LuoY LiuT 2009 “Governing buyer–supplier relationships through transactional and relational mechanisms: evidence from China,” Journal of Operations Management 27 294 309 10.1016/j.jom.2008.09.004 Search in Google Scholar

Audia, P G, Brion, S & Greve, H R, 2015, “Self-Assessment, Self-Enhancement, and the Choice of Comparison Organizations for Evaluating Organizational Performance,” Advances in Strategic Management Vol. 32, pp 89–118. AudiaP G BrionS GreveH R 2015 “Self-Assessment, Self-Enhancement, and the Choice of Comparison Organizations for Evaluating Organizational Performance,” Advances in Strategic Management 32 89 118 10.1108/S0742-332220150000032018 Search in Google Scholar

Hirst, G, Knippenberg, D V & Zhou, J, 2009, “A cross-level, perspective on employee creativity: Goal orientation, team learning behavior, and individual creativity,” Academic Management Journal Vol. 52, pp 280–293. HirstG KnippenbergD V ZhouJ 2009 “A cross-level, perspective on employee creativity: Goal orientation, team learning behavior, and individual creativity,” Academic Management Journal 52 280 293 10.5465/amj.2009.37308035 Search in Google Scholar

Perry-Smith, J E & Mannucci, P, 2017, “From creativity to innovation: the social network drivers of the four phases of idea journey,” Academy of Management Reviewpoint Vol. 42, pp 53–79. Perry-SmithJ E MannucciP 2017 “From creativity to innovation: the social network drivers of the four phases of idea journey,” Academy of Management Reviewpoint 42 53 79 10.5465/amr.2014.0462 Search in Google Scholar

Schilling, M A & Phelps, C C, 2007, “Interfirm Collaboration Networks: The Impact of Large-Scale Network Structure on Firm Innovation,” Management Science Vol. 53, No. 7, pp 1113–1126. SchillingM A PhelpsC C 2007 “Interfirm Collaboration Networks: The Impact of Large-Scale Network Structure on Firm Innovation,” Management Science 53 7 1113 1126 10.1287/mnsc.1060.0624 Search in Google Scholar

Clercq, D & Sapienza, H J, 2004, Why do venture capital firms learn from portfolio companies, Vlerick Leuven Gent Working Paper Series ClercqD SapienzaH J 2004 Why do venture capital firms learn from portfolio companies Vlerick Leuven Gent Working Paper Series Search in Google Scholar

Articoli consigliati da Trend MD

Pianifica la tua conferenza remota con Sciendo