Chapter 2 Literature Review
2.3 Social network theories
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
11
2.3 Social network theories
There is a very large volume of social network research, in various disciplines, accumulated since the idea of tie strength was proposed by Granovetter in year 1973. Most of it involved social network research in organizational behaviour, including leadership, power, turnover, job satisfaction, job performance, entrepreneurship, stakeholder relations, knowledge
utilization, innovation, profit maximization, vertical integration, and so on (Borgatti & Foster, 2003). Major findings from these studies will be discussed in the next section as they offered some substantial insights to social network research.
In order to understand social network, it is necessary to know the building blocks of community. Participants‟ sense of community is essential for establishing strong community.
Membership is important in building the feeling of belonging to a group. Members‟ feeling toward group or the group to its members as perceived generate influence on their behavior.
Communication among members facilitate resources exchange and integration, thus fulfill their needs. Frequent interaction, shared experience and anticipation on the future create emotional closeness among members (McMillan & Chavis, 1986).
There are two primary streams among network studies, to examine the causes of network structures or the consequences. The studies that examine the consequences of networks are parallel with the structuralist idea that actor‟s behaviour is conditioned by networks in which the actor was embedded; while studies that examine the causes of network explore the network in terms of actor personalities and latent propensities (Borgatti & Foster, 2003).
Relational perspective of network research saw interpersonal transmission as taken place among those with pre-established social ties. The structuralists suggested that two nodes have similar outcome (shared the same attitude) because their positions in networks are the same, even if direct connection between them is unavailable (Borgatti & Foster, 2003).
Social capital concept had been used in a wide range of research in terms of a person‟s ties or network position in bringing substantial outcome. It is also often used to explain the information transmission in social structures. Social capital has been the central idea of social
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
12
networks and the associated norms of reciprocity are valued by people who are in them.
Putnam (Putnam, 1995) defines social capital as the “features of social organization such as networks, norms, and social trust that facilitate coordination and cooperation for mutual benefit” (p. 66).
Coleman (1988) sees social capital as a resource for individuals defined by its function and is context specific. The social capital that facilitates certain actions may be useless to others.
Information channel, prospects for acquiring information that facilitate action, is one of the three major forms of social capital. Social relations that are maintained for other purposes can also be used to obtain information.
While social capital research is primarily individual-based, how research treats ties and their function had distinguished two major streams of the studies. For example, the studies Coleman (1990) and Burt (1992) focus on structure of ties in ego-network, and tend to ignore the content of ties. In contrary, Lin (2001) and others focus on the resources that flow through social ties. Social ties are seen as channels of information and other types of resources flow (Borgatti & Foster, 2003).
Social influence studies (N. E. Friedkin & Johnsen, 1999) and network diffusion studies embodied the structuralist traditions by placing the emphasis on constraint (DiMaggio &
Powell, 1983), while social capital literature emphasized on opportunities (Gargiulo &
Benassi, 2000) and individuals are seen as actively exploiting his/her network position for the benefits of him/herself (Borgatti & Foster, 2003). As Borgatti & Foster (2003) pointed out,“[S]ocial capital studies seek to explain variation in success (i.e., performance or reward) as a function of social ties, whereas diffusion and social influence studies seek to explain homogeneity in actor attitudes, beliefs and practices, also as a function of social
ties.” (p.1002)
Burt (1992) proposed the concept of structural holes, in which the lack of ties among contacts in individual‟s ego-network is considered social capital. In other words, individuals can maximize the volume of useful information from their social networks if there is no connection among their contacts. The underlying assumptions are: the amount of ties an individual maintained is limited, and information will be shared among friends. If A and B are friends, they will have similar information, as a result it would be more efficient for an
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
13
individual to link to just one of the pair meanwhile try to link to someone unconnected to them (Borgatti & Foster, 2003).
Social network theory had also been used in conjunction with other related theories in research human information behaviour, for example, small world theory, social exchange theory.
Travers and Milgram (1969) conducted an experimental study to examine the density of social networks. The design of the study is each starting person was provided with a document and asked to begin moving it by mail toward the target person. Two individuals may not directly know each other, but they may share one or more acquaintances, that link them indirectly. The information about target person is communicated through social networks. Two randomly selected American can be connected with average of 5.2 intermediaries. The “small world theory” (Milgram, 1967) states that social networks are tightly knitted, plentiful of unexpected strings that link individuals far apart physically or socially (Travers & Milgram, 1969).
Blau (1964) defines exchange behaviour as “voluntary action of individuals that motivated by the returns they are expected to bring and typically do in fact bring from others”. The central idea of social exchange theory is social structure observed as configuration of social relations between actors (individuals and organizations), where the relations involve the exchange of valued items, both material and non-material ones (informational, symbolic, and etc) (Cook
& Whitmeyer, 1992).
Homans (1958) investigates the relations between reward and the performance a particular action. Emerson (1976) emphasizes the concept of “reinforcement” in social exchange.
Reward refers to positive reinforcement. The more a certain action is rewarded, the more likely the repetition of the action. Homans (1958) indicates that a person who gives much to others will try to get much from them, and a recipient who receives much is also put under reciprocal pressure. This is consistent with the concept of reciprocity in interpersonal relationship in Chinese societies.
Information exchange in social networks is a type of social exchange. It involves a primary principle: when a favour was granted to a person, despite the general expectation of being reciprocated in future of the favour-giver, the specific form does not formulate in advance.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
14
Thus the nature of reward is non-negotiable, and is up to the person who is supposed to reciprocate to decide. Thus trust plays an important role in the process of social exchange (Blau, 1964; Homans, 1958).
Exchange relation develops gradually in a steady process. It usually starts with a small exchange that requires little trust due to its low risk. If the favour is not returned, the exchange relation could be terminated. If the favour is returned with value higher than expectation, then a person has proved that he or she is trustworthy. Thus the prospect of receiving bigger assistance in future is higher. The process of social exchange generates trust in social relations through its cyclical and expanding features (Blau, 1964).
While the literature reviewed in prior paragraphs provides general ideas about social network research, “the strength of weak ties” proposed by Mark Granovetter is one of the theories that provides theoretical framework for viewing network behavior (Schultz-Jones, 2009).
He investigates how the types of ties are involved in the diffusion of job-related information.
Generally, interpersonal ties could be categorized in three forms: strong, weak and absent.
The strength of ties is the combination of the amount of time, emotional intensity, intimacy and the reciprocal services which characterize the tie (Granovetter, 1973). Strong ties typically involve close friends, team-mates and other relationship that is intense and stable.
They tend to share information, resources, contacts and experience from the same pool. There is a high level of intimacy and self-disclosure between them. The ties are both instrumental and emotional. Resource and information exchange are frequent and reciprocity is emphasized. By contrast, the weak ties refer to acquaintances and casual contacts. People connected by weak ties are situated in different social circles with less similarity between them. Resources and information exchanges are infrequent. The ties are primarily instrumental, share few types of information or support (Haythornthwaite, 2005).
The weak ties theory suggests that individuals usually rely on weak ties to receive new information about job mobility. This is because only weak ties can bridge the gap between different sections by creating more and shorter paths between them. Members from different social circles are the resources for new information not already available in existing social circles (Granovetter, 1973).
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
15
In the case of job related information diffusion, he agrees that the instinct is those with strong ties are more motivated to assist with job information. However, individuals are more likely to obtain job information from those he or she maintains weak connection. This is because of those with strong ties are usually come from similar social background and pooling from the same information sources. Thus strong ties may not be able to channel new information about job opening. On the other hand, the different social background of weak ties may access to fresh job information that is unavailable with the existing social circles. He suggests labor-market information diffusion is byproduct of other social processes rather than the results of searching through social networks (Granovetter, 1974).
Social network sites such as Facebook are one of the most popular types of virtual community. Thus the literature on virtual communities will be presented next.
2.4 Virtual communities
Virtual communities generally refer to communities that form through computer mediated communication (CMC). A variety of terms have been used to describe them including electronic and online communities (Blanchard & Horan, 2000). Virtual communities also refer to a number of people communicating through CMC based on shared interests (Mousavidin & Goel, 2009).
Virtual communities can be generally distinguished into two major categories. The first type is electronic community that supplements the traditional physically based community. For example a website of city hall, schools or community organization that provides electronic resources for residents‟ use. The second type of virtual community is geographically dispersed members connected on the basis of shared interest in certain topics.
Communication between members of these communities may be limited to online interaction and never meet in person (Blanchard & Horan, 2000).
Both geographically dispersed and physically based virtual communities facilitate the connection of previously unknown people on the basis of shared interests. Geographically dispersed communities tend to establish sparsely knit networks as the members scatter in different areas. Physically based virtual community create network with higher density through overlap with face to face networks (Blanchard & Horan, 2000).
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
16
It is useful to investigate the nature of virtual community with the knowledge/theory from social networks of offline community. Although CMC can overcome the space distance, but it is just one of the channels people use to interact. The communication baggage from real life is still dominates online interactions, such as gender, cultural background, socioeconomic status and offline connections with others (Wellman & Gulia, 1999).
Virtual community can be examined from three approaches. It can be analyzed as a community as a whole with the focus on sense of community felt by all members. Discourse level of analysis is the second approach to studying virtual community. Third, social network analysis focus on the ties and links among members (Mousavidin & Goel, 2009) which have been discussed in previous section.
A popular research theme about virtual community is the resemblance or difference between online and offline social networks. For in-person community, social support such as companionship, emotional aid, information, services, even financial support and tangible material/goods channel through social networks. However, the supports do not come from one source. People maintain multiple connections with different people to access to a variety of resources. Similar to in-person community/social networks, various social resources can be found online. Online social network can find social resources pool through specialized relationships as well as pooling resources from a virtual community (Wellman & Gulia, 1999).
Virtual community is also explored in terms of norm of reciprocity that is common in face to face community. Reciprocating support and social resources ensures that more is available when needed in the future. There are concerns of the lack of motivation in providing support in virtual community. Because of the social or geographical distance, and the large and disperse social networks online, internet users may be less likely to be rewarded or receive support in return (Constant, Kiesler, & Sproull, 1996; Thorn & Connolly, 1987) (Wellman &
Gulia, 1999). The prevalence of user generated content and booming online social networks has shown these concerns are unnecessary. Non-tangible social resources are available in online social networks including emotional support, information and providing a sense of belonging. This may be due to the fact that it can be done conveniently at home and does not require a huge investment of time, financial or energy (Wellman & Gulia, 1999).
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
17
The advance of transportation, communication and internet transform the concept of in-person community. Neighborhood and kinship ties are only part of individual‟s overall community networks. Communities exist as social networks consist of kin, friends, and co-workers, regardless of geographical distance. Defining community in terms of space has been replaced by the concept of social networks ((Wellman, 1988a, 1993; Wellman & Gulia, 1999).
Strong ties in online social networks carry characteristics similar to strong ties in offline social networks (Wellman & Gulia, 1999). Bases on “small world theory” (Milgram, 1967, Travers & Milgram, 1969), Dodds, et al (2003) performed a experimental study to investigate the online social networks by using email instead of chain letter. Their findings are global social networks are searchable in an average of five to seven steps.
Some of Wellman observations for virtual communities in general are applicable to social network sites. The ties people build and maintain in online social networks are quite similar to their face to face community ties: intermittent, specialized and varying in strength (Wellman & Gulia, 1999). The online ties may be specialized, but the combination of large and disperse connections maintain online apt to provide a wide range of support (Wellman &
Gulia, 1999).
Communities of different size, composition and structure exist online. The nature of online communication enables the virtual communities to become more global and local simultaneously as global connectivity and local matters intersect. The internet support densely-knit group communities as well as personal communities. Group communities and personal communities operate online and offline (Wellman & Gulia, 1999).
For information transmission in virtual communities, Burnett (Burnett, 2000) proposes a typology that illustrates various information behavior in virtual communities. He sees information exchange as social act that forms the foundation of virtual communities. Burnett (2000) points out despite virtual communities‟ lack of face-to-face interaction among their members compared to traditional communities, information exchange creates opportunities for socializing and other types of interactions. Information sharing itself is considered a social act (Burnett, 2000).
Previous studies have shown tie strength mediating information behavior in social networks.
However, tie strength may be insufficient to explore information exchange in social networks
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
18
of Chinese societies. Chinese social network is operated under a complex cultural mechanism.
Guanxi as an indigenous construct of interpersonal relationship may shed some lights on understanding the way social networks function in Chinese societies. The most fundamental idea underlying the Chinese social networks is concept of lun (倫). Understanding it is significant for exploring Chinese social networks. Thus the discussion Chinese social networks will start from lun concept.