Social Media Research
Past work has referred to as many as 10 to 15% of key individuals in networks as “champions” or “leaders” or even “seeds” but no single definition exists.
A variety of approaches are used to uncover leaders: some research has relied on “counting nominations” while others used a mathematical approach to measure the number of steps it would take for that individual to reach others in a group.
The fewer average steps, the greater the “centrality betweeness.”
Interestingly, Valente suggests that leaders may not be the best agents of change, as they “have a vested interest in the status quo.” People who link groups, or “bridging nodes” may have less to lose, and may become a “low threshold adopter” or someone who adopts behavior before his or her peers.
Decoded Science asked Dr. Valente for a specific example where convincing a bridging group to adopt a behavior would prove more effective than convincing leaders. Valente responded, “Bridging individuals, because they are not central in the group, are often able to adopt new ideas earlier than group members. So mammography screening, for example, might be something that central people will be inhibited to adopt early because it is not compatible with cultural norms. Bridging individuals, via their contact with members from different groups, may be champions for this idea.”
Catalysts of Change: Early Adopters of a Behavior
Core-periphery structures refer to groups where key individuals or groups occupy the center and those around the outside are connected to the hub, but not to each other. In these situations, it is important to arm the core members with resources. It is also important to be able to discover who the early adopters, or ” low threshold adopters” so that they can excite change.
“Social isolates” are people on the periphery of social networks. Because they have less pressure to conform and may be linked to other groups, mining them for novel approaches to problems may be helpful. As these people may literally be the last to hear of innovations adopted by others, special attention may be needed to bring information to them. Valente gave the following example of when peripheral individuals would be sought out: “the most obvious intervention is in the area of adolescent mental health. Often adolescents at risk for suicide ideation and suicide attempts are marginal in networks. Being able to identify such people to receive counseling services and supports would seem critical.”
Defining Roles and Influence in Social Networks
Valente also reported that defining networks reliably and validly can be tricky. Past research suggests that individuals using “free recall” are reliably able to identify those in their network – Valente distinguishes between less personal “advice networks” and more culturally sensitive “discussion networks.”
At times, removing key individuals, or “node deletion” can be the most effective way to reach a public health goal. Valente gave the example of eliminating key terrorists to weaken a terror network in his article. In our interview, he further postulated about removing delinquents from classrooms, and providing a public health example, he continued, “prostitutes may be effective spreaders of STDs because they have many partners. Having a process of routine screening and treatment is an effective way to minimize STD rates.”
Decoding Science. One article at a time.