Centrality, Prestige, Popularity, & Networks
The article by Wasserman and Faust discusses the definitions of centrality and prestige and how they relate to location on a graph and in one's social network. They present three graphs in which the nodes in the graphs are quite different. Although the star graph obviously presents the most central node out of all the nodes in all three graphs, the article poses the question, "does the center, or centroid, of a graph contain the most important actors?" (p.172). The answer lies in the definition of "important".
An actor is considered prominent if the ties of the actor make him visible to the other actors in the network (p.172). The two types of visibility are centrality and prestige (p.173). A prestigious actor is one who "is the object of extensive ties, thus focusing solely on the actor as a recipient" (p.174). It is a directional tie, where the ties are directed toward the actor. Actor centrality on the other hand is not particularly concerned with the direction, but rather is more concerned that "the actor is simply involved" (p.173).
Something I found very interesting was the use of "closeness" to measure an actor's centrality. The article notes that "as geodesics increase in length, the centrality of the actors involved should decrease" (p.184). The idea that centrality is inversely related to distance was a little counterintuitive to me at first. I thought that even if someone had only a few ties, yet those ties were bridges to other subgroups, this person could be a central figure without having to have short intermediary paths to everyone in his/her network. But I guess the idea of the star clarifies this since it is used as the example of being the shortest path linking the other nodes in the shortest possible path.
Although I could go on forever about this article since it was very detailed, I will only mention one more measurement that stood out to me - that of betweenness. When the authors noted that the "actor in the middle, the one between the others, has some control over paths in the graph" (p.188), it reminded me of when Burt said that structural holes are the setting for tertius (third person who benefits) strategies (or entrepreneurial strategies) (p.90). The idea of location being a key part to these graphs is important, I think, in understanding centrality and prestige as it relates to the power one gains from that position.
My question is this: The article indicates that "the betweenness indices can be quite different measures of actor centrality than degree- and closeness-based indices" (p.192), at least for the families studied. Which do you think is a more accurate predictor of actor centrality and why?
Freeman also discusses the idea of betweenness and its potential for control. He states that "when a person is strategically located on the communication paths linking pairs of others, that person is central" (p.221). A person in this position can either withhold or distort information in transmission or maintain communication between others. They also have potential as coordinators of group processes. He also notes that "a point that falls on some but not all of the geodesics connecting a pair of others has a more limited potential for control" (p.222).
With this idea in mind, I thought of the typical "popular crowd" of any given high school. It is interesting that although they may be prestigious, they probably wouldn't have much power over the information flow due to their lack of betweenness. This could contribute to the disproportionately large number of rumors that generally circulate about them and the little degree of control they have over it. Freeman notes, “A point is viewed as central to the extent that it can avoid the control potential of others” (p.224). In what case is the person who is most central likely to be the same person who is the most popular and vice versa? What would Freeman's idea of betweenness suggest about this?
Krebs' article talks about the terrorist attacks on September 11, 2001 and how their network was difficult to map out due to its members' hidden and dormant strong ties (p.10). These ties were sparse and distanced by many miles, yet extremely strong and only active at certain times to achieve goals. One part that caught my eye was on page 11 when it said, "the hijackers kept to themselves - they did not make friends outside the trusted circle". Although they were living in America for some time before these attacks, even their neighbors did not catch on to what they were about to do. This made me wonder whether we, as Americans, are less likely to notice terrorists (even if they were living next door to us) with social isolation on the rise. Do we consider it normal that they rarely interacted with others (outsiders)? Are our networks beginning to resemble theirs in exclusivity and sub-groupings whose structural holes between these groups are not bridged?
Valente, Unger, and Johnson discuss the effect of peer influence on the initiation of smoking behavior in adolescents. They note that "popular students will attempt to set trends while at the same time try not to deviate too much from the cultural norms of the group" (p.324). Therefore we can look toward popular students to both influence and reflect the group trends around them. They found that "popularity was associated with increased susceptibility to smoke and smoking" (326). This fit with my previous thoughts about what the outcome of this study would be partly because it makes sense that popular kids would try do something considered "socially cool" and partly because of my own observations from high school. From a communications perspective it makes sense why many of the advertisements directed toward today's youth try to a.) Show kids that it's not cool, and b.) Show them that it's not as socially normative in today's society.
I found the article by Mouttapa et. al. quite fascinating and intriguing. The fact that bullying in elementary schools is becoming a growing problem in the US was quite surprising, especially since it was, what I would consider, "at its prime" when I was in elementary school. The fact that 7-15% of the school-aged population represents bullies, while only 2-10% are victims, implies that there are a larger number of people exerting aggression over a much smaller number of victims. It is therefore likely that any victim will have more than one bully behaving aggressively toward him.
Although bullies are noted to have "higher peer-nominated scores on sociability and leadership relative to other students," (p.317), I would argue that their success is limited to their youth. Recently my mom sent me a newspaper article from home reporting an arrest of the young man who had once epitomized the word "bully" from elementary through high school. It made me wonder how many other "bullies" I had known that were now experiencing difficulties in life.
When I "Facebooked" the five I could remember, none of them showed up. Although it is possible that these highly "social and leader-like" kids could've grown up to become so anti-social as to not have created a Facebook account, I tend to think it's more likely that they never made it to college. So although the victims may have once experienced "academic difficulties... peer rejection... and learning difficulties" many of the victims I remembered surprisingly did have Facebook accounts. Not that Facebook is the end-all-be-all but I found this interesting and was wondering if any of you would find similar results. As noted in the limitations of the study, longitudinal studies would probably add much to our understanding of these node characteristics of bullies, victims, and their friends. Where these bullies and victims end up and what their future network structures look like is something I would be very interested in.