The guilty, the unlucky, or the unaware? Assessing self-reported behavioral contributors and attributions on pedestrian crashes through structural equation modeling and mixed methods

Abstract

Introduction: Recent literature suggests that the causation of pedestrians’ crashes and the contribution of safety-related behaviors within them may substantially differ compared to other road users. This study aimed to test the effect of individual factors and safety-related road behaviors on the self-reported walking crashes suffered by pedestrians and, complementarily, to analyze the causes that pedestrians attributed to the crashes they suffered as pedestrians during the previous five years. Method: For this cross-sectional research performed in Spain, data from a nationwide sample of 2,499 pedestrians from the 17 regions of the country were collected. Participants had a mean age of 31 years. They responded to a questionnaire on demographics, safety-related walking behaviors, and self-reported pedestrian crashes and the causes attributed to them. Results: Utilizing Structural Equation Models (SEM), it was found that self-reported walking crashes can be predicted through unintentional risky behaviors (errors). However, violations and positive behaviors remain non-significant predictors, allowing to hypothesize that they might, rather, play a key role in the pedestrian’s involvement in pre-crash scenarios (critical situations preceding crashes). Also, categorical analyses allowed to determine that the causes that pedestrians attributed to the walking crashes they had suffered were principally their own errors (44.6%), rather than their own traffic violations (8.5%). Nevertheless, this trend is inverse when they believe the responsibility of the crash weighs on the driver. That is to say, they usually attribute the crash to their traffic violations rather than errors. However, many biases could help explain these attributional findings. Practical Applications: The results of this study highlight key differences in behavioral features and crash predictors among pedestrians, with potentially relevant applications in the study and improvement of walking safety from behavioral-based approaches.