A communication network is a set of individuals (or nodes ) connected by communicative interaction. For example, an organizational communication network could consist of those seeking advice from others within the organization, and those giving the advice. Communication networks are social networks and are analyzed using the many theories, techniques, and procedures developed in the field of social network analysis. Though the earliest studies of social networks date from the 1930s, the modern study of communication networks originated in the 1950s with studies of how workgroups function (see Freeman 2002). For example, Alex Bavelas conducted studies in which small workgroups were given a task to complete, and then communication structure was altered to determine if a change in workgroup structure would affect task performance (it did) (Bavelas 1951).
Communication network study did not progress very much in the ensuing decades, because social scientists concentrated on taking random population samples that disconnected people from their social and communication networks (Rogers and Kincaid 1981). There was little research centered on how a person’s relationships, as measured by their communication patterns or partners, affected their behavior. Gradually, however, a science of social networks has emerged, using two different research methodologies: local and global.
Local network research is conducted by asking people to name others they are close to or talk to about important matters (Burt 1985). The names provided (and they can be first names only or initials) comprise what is known as an egocentric network ; the respondent is then asked for information about each of the persons named. Typically, the list is confined to five or six network partners.
Researchers can then describe respondents based on the characteristics of their egocentric network. For example, respondent networks can be characterized according to the number or percent of males, people of the same religious affiliation, or people living near or far from the respondent. Importantly, one can also measure the behaviors exhibited in the network, by, for example, considering whether each network partner provides social support, expresses opinions, or engages in behaviors relevant to the respondent or study.
Global network research is conducted by defining a boundary around a set of respondents and asking everyone to name others in the network they talk to or seek advice from (the researcher can also provide a roster for the respondent to refer to). Full names or other identifying information are provided so that the researcher can draw a map (see Figure 1) of the relations within the network. This so-called sociometric or census approach is the most powerful network methodology. The network data are used to describe individuals and/or the entire network (Scott 2001; Wasserman and Faust 1994).
Individual network indicators include a person’s centrality, defined as: (1) how many choices they are given by others in the network; (2) how close they are to everyone else in the network (defined as the average number of steps in the linkages to others); or (3) the degree to which they lie on paths connecting others in the network. Network analysis has developed over a dozen methods of identifying central nodes in a network. At the individual level, factors considered include, in addition to the ones mentioned above, whether a person is a member of a group, whether their ties are reciprocated, and whether they span different groups. Factors measured at the group or network level include the density of the network (the number of ties as a percent of the number of possible ties), centralization (whether the ties are concentrated around one or a few members), and clustering (whether connections between two people imply a connection to a third).
A significant finding of communication network analysis is the importance of peer influence: An individual’s likelihood of engaging in a behavior increases with the percentage of their peers who engage in the behavior. At the same time, some researchers have discovered that individuals’ network thresholds —that is, the degree of peer influence needed to change their behavior—vary according to their personal characteristics (Valente 1995). In some cases, being integrated into a network puts one in a privileged position with regard to access to information, but in other settings it may be deleterious, for example by putting some at risk (Valente et al. 2005). Research has shown that so-called opinion leaders serve as important filters and conduits for information and its diffusion by influencing the adoption decisions of others. The influence of opinion leaders may depend on group norms, however, as leaders adopt innovations early if the behavior is compatible with group norms. Denser and more centralized networks may accelerate diffusion by creating more and more efficient pathways for information to flow through, but may also hinder diffusion when the density is so great that it reduces the community’s access to outside information (Granovetter 1973).
Network analysis has been suggested as a methodology useful for measuring social capital. Social capital is defined both as the social resources available to a person via their social contacts (Lin 2001), and as a perception by the community that a person is trustworthy and civically engaged (Putnam 1995). Social network measures can provide a direct assessment of individual and communal social capital (Borgatti et al. 1998).
The tools and technology used to analyze communication networks have improved considerably in their ability to handle large datasets (such as the Internet). The field of network analysis continues to grow rapidly, aided in part by the International Network for Social Network Analysis (INSNA), which provides a Web site for access to more information.
SEE ALSO Communication; Network Analysis; Networks; Social Capital
Bavelas, Alex. 1950. Communication Patterns in Task Oriented Groups. Journal of the Acoustical Society of America 22 (6): 725–730.
Borgatti, Stephen P., Candace Jones, and Martin G. Everett. 1998. Network Measures of Social Capital. Connections 21 (2): 36–44.
Burt, Ronald S. 1984. Network Items and the General Social Survey. Social Networks 6 (4): 293–339.
Freeman, Linton C. 2004. The Development of Social Network Analysis: A Study in the Sociology of Science. Vancouver, B.C.: Empirical Press.
Granovetter, M. 1973. The Strength of Weak Ties. American Journal of Sociology 78: 1360–1380.
Lin, Nan. 2001. Social Capital: A Theory of Social Structure and Action. Cambridge, U.K.: Cambridge University Press.
Putnam, Robert D. 1995. Bowling Alone: America’s Declining Social Capital. Journal of Democracy 6 (1): 65–78.
Rogers, Everett M., and D. Lawrence Kincaid. 1981. Communication Networks: Toward a New Paradigm for Research. New York: Free Press.
Scott, John. 2000. Network Analysis: A Handbook. 2nd ed. Newbury Park, CA: Sage.
Valente, Thomas W. 1995. Network Models of the Diffusion of Innovations. Cresskill, NJ: Hampton Press.
Valente, Thomas W., Jennifer B. Unger, and C. Anderson Johnson. 2005. Do Popular Students Smoke? The Association between Popularity and Smoking among Middle School Students. Journal of Adolescent Health 37 (4): 323–329.
Wasserman, Stanley, and Katherine Faust. 1994. Social Networks Analysis: Methods and Applications. Cambridge, U.K.: Cambridge University Press.
Thomas W. Valente