How Triadic Closure and Power Interact in Social Networks
Sources:
https://keg.cs.tsinghua.edu.cn/jietang/publications/TKDE15-Huang-et-al-triadclosuremain.pdf
https://news.mit.edu/2022/frequent-encounters-build-familiarity-paige-bollen-0414
In previous lectures, we learned that triadic closures are likely to form due to the potential similarity or increased opportunity to meet between the two nodes (people) sharing a common connection, and those closures can strengthen social networks. Studies have shown that mere frequent encounters can build a sense of familiarity between people, so that positive relationships are more likely to emerge, supporting the theory for triadic closure. I would like to expand on this point and discuss different factors that might influence the formation of the triadic closure on social media, and the power in social networks behind it.
Specifically, in the paper “Triadic Closure Pattern Analysis and Prediction,” the authors discussed an interesting finding that popular users on Twitter are three times more likely to facilitate a closed triad (when two other less popular users that popular users connect form a bond with each other) than ordinary users. This observation could imply the correlation between power in social networks and one’s capacity to create new connections through triadic closure, where popularity act as a measure of power here. For example, more people might want to connect with a popular figure, through whom they are more likely to get to know new people in different fields to acquire potential opportunities or useful information. In this way, popular figures on social media have power over other users, as they are influential in the networks system, and have larger freedom to choose whom to connect with.
However, the same study also indicated that on another social platform (Weibo), where more individual interactions are enabled, the power of popularity is significantly weakened. Specifically, the frequency of triadic closures among three ordinary users is actually higher than the ones between two ordinary users connecting/following the same popular user. This reversed trend conveys that free interactions could undermine the power of connecting with more people, as all users have more equal access to reach out to other users. This finding reminds me of Richard Emerson’s networks exchange experiment discussed in class, that if the time of exchange is unlimited, people will have enough time to negotiate for a more equal share of the money they gain. As a result, the ones with more connections might not necessarily be the most powerful, since people might abandon them when they form new connections with a better deal payoff. This observation is interesting to me since I believe that in real-world settings, the power dynamics are much more complicated — people powerful in one specific situation might not be as powerful in other social settings, and how these powers change and interact are exciting to explore.