Application of Stochastic Search Variable Selection to Modeling Evacuation Ahead of Hurricane Irma

• Activity Number: 238 - SPEED: Environment and Health, Governmental Policies and Population Surveys, Part 1
• Type: Contributed
• Date/Time: Monday, July 29, 2019 : 2:00 PM to 3:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Sierra Bainter* and Caitlin Brown and Kiara Timpano
• Companies: University of Miami and University of Miami and University of Miami
• Keywords: SSVS; psychology; risk response; variable selection

Abstract

Previous research has shown that hurricane risk response is influenced by a large number of factors, some of which are counterintuitive. Preparedness decisions are complex and costly, leading up to, during, and following an event. In this project, Stochastic Search Variable Selection (SSVS) is used to identify important predictors of evacuation behavior among residents in South Florida during Hurricane Irma. Salient predictors of evacuation behavior included living with young children, living in an evacuation zone, having pre-existing health conditions, and the amount of worry experienced leading up to the hurricane. Socioeconomic factors did not predict evacuation, but sex and ethnicity were relevant. Evacuation was also related to posttraumatic stress, suicidal ideation, and desire to seek mental health services, reflecting in part the distress experienced due to the evacuation itself. Identifying factors associated with hurricane risk response is useful for understanding how individuals respond given their unique constellations of vulnerabilities, resources, and background characteristics. This project also demonstrates the application of SSVS for researchers in psychology.