Number of warning information sources and decision making during tornadoes

Jianjun Luo,Zhen Cong,Daan Liang
DOI: https://doi.org/10.1016/j.amepre.2014.09.007
Abstract:Background: Taking proper protective action upon receiving tornado warnings is critical to reducing casualties. With more warning information sources becoming available, how the number of such information sources affects decision making should be quantitatively investigated. Purpose: To examine how the number of warning information sources affected individuals' decisions to take protective action during tornadoes. Methods: A telephone survey using random sampling was conducted in 2012 with residents in Tuscaloosa AL and Joplin MO, resulting in a working sample of 782 respondents. Both cities were struck by violent tornadoes (Enhanced Fujita Scale [EF]4 and EF5) in 2011. The analysis was conducted in 2013. Results: Logistic regression analysis showed that relative to having only one warning information source, having two and three or more warning information sources significantly increased the odds of taking protective action in Joplin but not in Tuscaloosa; having three or more sources had a significantly stronger effect on taking protective action in Joplin than in Tuscaloosa. Having an emergency preparation plan in both cities and being white in Tuscaloosa significantly increased the odds of taking protective action, whereas being divorced in Joplin reduced these odds. Conclusions: Receiving warnings from more warning information sources might be more beneficial in places with less previous exposure to tornadoes and for populations with lower awareness of a potential tornado and higher probability of receiving no warnings. Emergency management agencies and public health officials should give priority to these places and populations when formulating disaster mitigation decisions and policies.
What problem does this paper attempt to address?