Date of Award
Master of Science (MS)
Hurricane Matthew was the most powerful hurricane during the Atlantic Hurricane Season in 2016. It caused tremendous damages to infrastructure and coastal areas of the United States. This thesis uses survey data collected in 2017 from residents in the Jacksonville Metropolitan Area after Hurricane Matthew. Survey questions were designed to capture evacuation-related decisions, information sources usage, socio-economic factors, perceived certainty and intra-familial interactions. The first part of the thesis modeled households’ perceived certainty to identify factors that affect different perceived certainty topics. Certainty topics included were: whether one lives in an evacuation zone, time of hurricane impact, evacuation preparation time needed, when to evacuate, evacuation travel mode, evacuation route, and evacuation destination. The modeling results showed similarities and disparities among perceived certainty topics. Household archetypes were created to offer insights for both decision makers and stakeholders for hurricane emergency management. The second part of this thesis explored the connection between the evacuation decision and perceived certainty using a two-stage modeling concept. Adding contextual factors usually leads to endogeneity bias which means parameters of variables will be overestimated or underestimated. A control function approach was used to account for potential endogeneity bias when linking perceived certainty with the evacuate/stay decision caused by unobserved attributes. The uncorrected base model was found to have a downward bias of the perceived certainty of evacuation destination, and with endogeneity bias corrected the parameter for this variable increased by 91.6%.
Shen, Jiayun, "Exploring Uncertainties in Households’ Hurricane Evacuations" (2020). All Theses. 3467.