Social Network Analysis

David Knoke and Song Yang defines social network as a structure depicting interconnections among a set of members or actors. Social network [Page 1542]analysis is an approach to better understand the exchange of information or other resources within these interconnections. The goal of this relational approach is to conceptualize and quantify how actors (e.g., people, groups, or organizations) interconnect and influence other actors. Social network analysts use qualitative and quantitative methods to (a) conceptualize social network ties with visual tools, such as graphs, tables, and figures and (b) characterize the nature of those ties, such as the strength of the relationships.

The application of social network analysis in research and evaluation is fairly recent, emerging in the 1930s, appearing in limited research articles in the 1970s, and then gradually increasing to date, with a recent sharp incline. Fields such as social sciences, computer science, and organizational management have recognized the benefit of using this approach to study patterns of relationships in a structure. This set of techniques can be used to capture complex patterns of interaction among actors as well as depict the structural change of interactor relationships over time.

There are three assumptions underlying social network analysis. First, structural relations are critical for understanding and predicting behavior more than attributes such as age, gender, and background. Second, social relations or networks affect perceptions, beliefs, and actions through a variety of structural mechanisms that are socially constructed among entities. Third, structural relations should be viewed as dynamic processes. This indicates that the relations among entities are not fixed; on the contrary, relationships change all the time. Better understanding the interconnectedness among actors further informs our understanding of context, patterns, and systems of groups of actors.

Social network analysis accommodates six types or measurement levels of variables: binary, multiple-category nominal, grouped ordinal, full-rank ordinal, interval, and ratio. A binary measure of relations refers to 1 representing the presence of a relation and 0 representing an absence. Multiple-category nominal measures of relations refer to the nominal measure with multiple groupings (e.g., participant selects among a series of options: friend, business relationships, or no relationship). Grouped ordinal measures of relations refer to ordinal data such as dislike, neutral, and like options. Full-rank ordinal measures of relations refer to rank data in which participants rate the relations from the strongest to the weakest. Interval and ratio measures of relations refer to data in which the measure is continuous.

Social network analysis results are mathematically calculated and then commonly visually represented via tables, graphs, and figures that illustrate characteristics such as density, degree centrality, closeness, and betweenness. Graphs efficiently highlight key features of a social network structure and consist of nodes that represent actors and lines that represent ties.

Concepts such as the density of the network, or the degree of connectedness of groups of actors, which is calculated to be a value between 0 and 1, can be illustrated via a graph of nodes and lines. Degree centrality concerns the extent to which a person or group is connected to other actors and is used to identify prominent actors within the network. Closeness and betweenness are centrality measures that help determine an actor’s proximity to others and can illustrate the depth of a relationship.

...