Toward a Conceptualization of Social Network Status

• In Event: 1-219 - Considerations in the Assessment of Social Context and Peer Network Structure

Integrative Statement

Social network methodologies involve modeling relationship structure and provide unique information about the relational context in which behavior occurs (Butts, 2008). Network-based measures are used to represent position, connection, and status within the network. These measures have in common that they are calculated based on real relationship ties (e.g., individuals name friends within their classroom or grade; nominations are used to create a network). In contrast, reputation-based measures (e.g., popularity) may lack social complexity and may not capture the intricacies of real relationships (Zhang et al., 2014). For example, one may have a reputation as popular, yet have few real friends. Not including these relationships gives inaccurate estimates of peer effects (McNeilly-Choque et al. 1996).

There are many network-based measures that provide information regarding relationship structure and one’s position within the network. However, theoretical or conceptual bases to differentiate network-based measures remains largely absent. For instance, despite different calculations, two network-based measures – social network prestige and social network centrality – are defined using the same descriptors (e.g., status, influence, popularity; Gallup et al., 2015; Reynolds & Crea, 2015). The current study will explore these network-based measures, moving toward a conceptual understanding of their similarities and differences.

Sixth graders from a southwestern U.S. city (N = 396; 47% girls; 65% White, 13% Latino/Hispanic) nominated up to 10 friends in their classroom and nominated peers based on several descriptors (e.g., aggressive, popular, leader, well-liked). Friendship nominations were used to calculate social network prestige (proximity prestige; Wasserman & Faust, 1994) and social network centrality (Bonacich centrality; Bonacich, 1987).

Results indicated a moderate correlation between the two network-based measures, r = .41, p < .001. However, partial correlations indicated different patterns of relations to other constructs. Partialling out the effect of social network centrality, social network prestige was positively related to social preference, leadership, prosociality, and popularity, and negatively related to bullying (see Table 1). In contrast, controlling for prestige, centrality was positively related to being popular, powerful, relationally and physically (marginal) aggressive. Figure 1 offers a visual representation of the structural network differences between individuals in one classroom.

Results suggest that social network prestige, which is based the number and network position of peers who choose a target individual as a friend, relates to relatively positive social attributes. Prestigious youth are not particularly aggressive but also not very powerful. This contradicts notions of prestige as relating to influence/importance (Zhao et al., 2015). Social network centrality, on the other hand, relates to being popular, powerful, and aggressive. Centrality is based on the target individual nominating many friends, who themselves have many friends. Highly central youth appear to be well-connected network members in terms of power but are not particularly well-liked. Again, this contrasts literature describing centrality as connectivity (Bonacich, 1987). This study offers the first insight (to our knowledge) into conceptual differences between network-based measures of social position that move beyond purely mathematical definitions or descriptors that are common across measures, and thus has important implications for our understanding and use of network-based measures.

Authors

• Naomi Andrews, York University
  Presenting Author

• Carol Lynn Martin, Arizona State University
  Non-Presenting Author