Using Clustering Coefficients to Identify Risk During Natural Disasters
https://phys.org/news/2018-09-social-networks-disasters.html
As Hurricane Florence looms over the Mid-Atlantic states, all states in the storm’s path have declared states of emergencies and evacuations have been called for. For a variety of reasons, calls for evacuation are not always completely obeyed; this could be for financial reasons – the thought of spending an indefinite number of nights in a hotel – or thoughts of false security brought on by those around people. The latter reason is what Daniel P. Aldrich from Phys.org found. He conducted a study, along with some colleagues at Facebook about the evacuation and lack thereof as affected by social network ties.
Aldrich identified the following three types of ties (various strengths of edges): “Bonding ties, which connect people to close family and friends; bridging ties, which connect them through a shared interest, workplace or place of worship; linking ties, which connect them to people in positions of power.” Interestingly, his study was able to come to the conclusion that those with many “bonding ties,” or stronger local networks of connections, are least likely to evacuate. This is likely due to the idea that “may feel supported and better-prepared to weather the storm.” A network that reaches authority figures such as that of someone who follows government officials or a network that can find a way to connect to aspects of life that are not local inherently span a much further geographic area. Therefore the warnings coming from these connects are strongly heeded because it makes the scale of an oncoming natural disaster seem larger.
This leads me to think of a way to quantify who is most likely to heed an evacuation warning. I’ve pondered the idea of using a measurement such as a person’s (node’s) clustering coefficient to quantifiably identify those at risk of staying put while a storm is on its way. The higher the clustering coefficient, the most friends of this node know each other and thus could be said to form a closer-knit community whose words to each other become a sort of echo chamber: thus leading to a false sense of community strength in numbers.