Citation

Xu C, Witlox F. Safety Sci. 2022; 146: e105540.

Copyright

(Copyright © 2022, Elsevier Publishing)

DOI

10.1016/j.ssci.2021.105540

PMID

unavailable

Abstract

It is difficult to observe people's actual evacuation behavior in case of a real aircraft accident. For one, luckily, because aircraft incidences have very low probabilities, and for another, because when faced with such extreme situations people often resort to very unexpected behavior. A way to obtain people's responses to an airplane emergency is then to estimate people's perceptions of the total evacuation time. In this paper, a stated preference analysis containing eight airplane evacuation attributes was carried out by nine hypothetical aircraft accident scenarios. The experiment adopted a fractional factorial design to reduce the survey size and obtained 192 effective responses. A cumulative odds ordinal logistic regression with proportional odds was performed and results showed the model has a good data fit. The following findings were revealed from this study. First, for the onboard safety briefings, the efficiency of safety broadcasting is better than crew demonstrations and safety cards, but the overall attention from passengers to the cabin safety announcements is still inadequate. Second, in passengers' perception, the lack of crew guidance would slow the evacuation the most while more available exits reduce the evacuation time primarily. Third, being knowledgeable about and more experienced in flying has a significant effect in reducing people's perception of evacuation time. Finally, higher self-reported levels in the ability to stay calm are associated with a shorter estimation of the evacuation time. These findings may help optimize the understanding of air passengers' behavior in emergencies and improve the preparedness for aircraft emergency evacuation.

Language: en

Keywords

Airplane evacuation; Cabin safety information; Ordinal logistic regression; Stated preference; Total evacuation time perception