Cartography of Accident Causation Models: Remapping the Modeling Landscape.

Abstract

Accident causation models are abstractions of the real world that explicitly or implicitly underlie our perception and understanding of accidents. As the continuous advancement of our dynamic society demands us to keep updating our understanding of accidents, this review article provides an overview of the landscape of accident causation models to aid our endless journey of expanding the collective endeavor of accident modeling. Our review of existing literature reviews identified several issues in the labels and categorizations of models, including logical inconsistency and ambiguous definitions, that prevent the comparison and analysis of models from a neutral and objective standpoint. To offer a structured perspective for revealing what was actually done in the past and what can be further explored in accident modeling practices, we developed an alternative map of modeling approaches based on their underlying assumptions and geometric structures of graphical representations. This article further provides individual reviews of 30 models, focusing particularly on how the assumptions and structures of the models inform the derivation of remedial and preventive measures. It concludes with a discussion on unexplored research paths that the alternative map directs us to investigate, indicating that our journey toward a safer world is far from complete.