DISASTER PREPAREDNESS PREDICTION THROUGH COLLECTIVE EFFICACY INDICATORS

Abstract

The gap between existing predictive models of community preparedness and community preparedness capabilities has long been overlooked. Attempting to fill this gap, this study develops a predictive model using collective efficacy sociometric markers, including social cohesion, trust, community problem solving, and shared expectations, to estimate a community’s preparedness in terms of disaster scenarios. The model relies on combining quantitative data from surveys and geospatial vulnerability maps to build machine learning models that estimate preparedness scores for different populations. The model captures neighborhood socio-psychological data coupled with coordinated community response data to go beyond hazard-based evaluations of preparedness. The model is applied to a dataset from three disaster-prone regions, capturing subjective and objective risk perceptions. Results show that most collective efficacy factors enhance preparedness behaviors, including emergency planning, intent to evacuate early, and resource mobilization. The research confirms that collective efficacy in disaster research is a critical metric while providing tangible guidance for municipal governance and community resilience planning. Moreover, the study shows that accuracy of prediction improves considerably with the inclusion of demographic moderation, such as age, income, and education. Such predictive capacities can help shape policy aimed at disaster resilience at the community level.