Social Network Analysis

Contexts and Entities

Contexts are most often manifested in network studies with the operationalizations of entities. So, if one is interested in an industry nodes might be defined as separate organizations, if one is interested in a team then nodes would represent individuals who are its members. Figuring out how to make a representation of the system, particularly where to draw the boundaries between separate networks, provides a fundamental challenge. Boundaries imply some quality of interaction demonstrably different within the network than communication outside the network, particularly with other networks. The two contrasting approaches involve the realist approach where the researcher adopts the vantage point of the actors (e.g., people with whom one conducts research) and the nominalist approach which would create a conceptual framework (e.g., all members of each academic department—economics, philosophy, communication).

Links

Network analysis examines the relationships between nodes, referred to as “links.” Links represent the communication between nodes that serves as the basic datum. However, in most network analyses, a limited understanding of the function [Page 1634]and quality of communication becomes employed in the analysis. For example, viewing the number of messages between two nodes in a network may be used as a measure of the “strength” of the network, rather than any assessment of the quality or content of the communication.

Links have several properties. Traditionally network analysts have examined their content (e.g., production, innovation, maintenance). Asymmetry is an important property of organizational networks since there are a multitude of differences between organizational members, especially in term of status and the direction of communication. Strength (e.g., importance, frequency) of ties is often examined. The strength of weak ties is perhaps the most well-known concept related to network analysis. A “strong” tie between nodes may exist, even though the number of messages remains relatively small. Reciprocity involves whether each node would characterize the relationship in the same way. Quite often, a linkage lacks reciprocity, one member of the link views the relationship as stronger than the other node. Channels might include written, face-to-face, telephone, or telecommunication networks.

The manner in which these various properties of links are combined can determine the analytical power and depth of any one network analysis. For example, a multivariate network might simultaneously measure weight, frequency, and duration of a link. Multiplexity refers to the nature of overlap, or correspondence, between differing networks (e.g., friendship as opposed to work). The degree of multiplexity has been related to such issues as the intimacy of relationships, temporal stability of relationships, reduction of uncertainty, status, the degree of control of a clique over its members, performance, redundancy of channels, and the diffusion of information.

...