Structural Equation Modelling of potential risk factors for pedestrian accidents in rural and urban roads

Abstract

Although several previous studies have focused on pedestrian safety, most of the studies have focused on urban roads. This study investigated accident data among 1358 pedestrians from 2012 to 2018 on urban, suburban, and rural roads in 16 cities in the Gilan province of Iran using structural equation modelling (SEM). Factor analysis showed that four exogenous latent variables had a significant relationship with fatal accidents among pedestrians. Factors related to roads (coefficient: 0.968), vehicles (coefficient: 0.632), humans (coefficient: 0.306), and the environment (coefficient: 0.194) were respectively the most important for pedestrian safety status. The main findings also proclaim that the poor quality of cars manufactured in Iran in addition to the poor design of intersections and major roads (urban) and ring roads (suburban) may be important reasons of the increased fatal accidents in the studied areas. A potential cause for these accidents may be rooted in less attention towards giving behavioural instructions to road users, and underdeveloped driver education procedures. Finally, the data-model fit of the SEM was validated using different indicators, and suggestions to improve safety were pointed out.