مقاله انگلیسی رایگان در مورد ثبات شبکه نوآوری خروجی

مقاله سال 2016

 مهندسی فناوری اطلاعات IT

The effect of individual mobility on knowledge transfer, innovation, and competitive advantage is increasingly becoming an important domain of research (Gardner, 2005; Harris and Helfat, 1997; Rao and Drazin, 2002; Song et al., 2003; Sturman et al., 2008; Wezel et al., 2006). Interorganizational mobility of individuals affects gains or losses in terms of the competitive advantage and performance outcomes (e.g., survival, profitability, effectiveness in head-to-head competition) of organizations that lose individuals (Aime et al., 2010; Phillips, 2002). Therefore, most organizations are trying to curb the mobility and keep the stability of their employee groups, particularly the highperformers. Conversely, the employee’s performance may also impact their stability. High-performers usually own high satisfaction with the current job, which makes them less likely to leave, while lowperformers are more likely to seek outside opportunities. Although there are reciprocal effects (Shaw et al., 2005a), direct (Glebbeek and Bax, 2004) and indirect (Shaw et al., 2005b) evidence suggests that the effect of employee stability on his/her performance is stronger than the reverse, which may be the main cause that most extant studies focused on the former. However, extant studies did not clearly examine to what extent the reciprocal effect is ignorable. Since there is reciprocal effect, the causal analysis of employee stability and performance should take it into account from both empirical and theoretical perspectives. As the employee’s performance and stability interact with and function on each other, this study will make a comprehensive examination of the bidirectional causalities, which is one of the main contributions of this study to extant literatures.

In the context of an organizational network, as the network becomes more connected, distance between any two nodes diminishes, it is possible that information can become more democratized (Ahuja et al., 2012), information can thereby diffuse more quickly, fostering outcomes such as innovation or creativity (Schilling, 2005; Schilling and Phelps, 2007). As the inventors’ access to the information and knowledge is to a great extent dependent on the links with each other, the moderate effect of the network connectivity on the inventor stability and his/her performance is indispensable. Although the effect of network structure has been widely discussed by extant studies, e.g., Ahuja (2000), Nerkar and Paruchuri (2005), Paruchuri (2010), Cattani and Ferriani (2008), Zhang et al. (2014a), they are mostly based on the largest connected component within the whole network. As the disconnected components potentially conflate the influences of small-world structure and simple connection (Fleming et al., 2007) and usually take a relatively small ratio compared with the largest component (Casper, 2007), most studies focused on the largest component, while ignored the methods to develop a weighted average across disconnected components proposed by Schilling and Phelps (2007). However, besides the largest component, other components, e.g., the second and third largest, usually own well structured fabric. These components may also exhibit significant network effect, as the links constitute the base for inventor communication. Inventors with key positions may also have advantages in accessing information, and thereby generate higher innovation output in other smaller components. The specific inventive process may lead to the disconnections, e.g., pharmaceutical researchers are usually assigned to several groups, which are making mutually independent researches; technicians embarking at two different projects within the same firm may also lead to two isolated components. Obviously, inventors in the largest component represent only part of the firm’s inventive activity. As the inventors in other components may also be doing important researches, ignoring these components may lead to a bias of the empirical results. In this sense, the network effect on network stability and performance, particularly in the partly connected contexts, deserves a further study. We will compare the differences of the network effects in the fully connected networks with that in partly connected networks, which formulates another main contribution of this study.

Additionally, extant studies provided only evidences that network connectivity is beneficial by proving that a greater ratio of the largest connected component positively impact innovation, e.g., Fleming et al. (2007), Chen and Guan (2010), Zhang et al. (2014b). As the linkages between individuals are the basic element constituting the network, greater extent of connectivity may be the key for the network indicators, e.g., clustering coefficient, centrality, path length, to be functioning on innovation. However, the moderate role of connectivity is not carefully examined by extant studies and will be another main job of this study. The remainder of this study is organized as follows: Section 2 presents the hypothesis; Section 3 presents the data and methods; Section 4 provides the empirical results; Section 5 discusses and Section 6 concludes.