Spatiotemporal Disparities in Macro-Microscopic Properties of Motorcycle Injury Level

Authors

Chenzhu Wang, Pengfei Cui*, Mohamed Abdel-Aty, Said M. Easa

Abstract

Motorcyclists in Florida face significantly higher rates of injuries and fatalities compared to national averages, highlighting the need for a better understanding of the contributing factors. This study addresses important gaps in existing literature by (1) integrating macro-level socio-demographic variables (such as population density, education, and poverty) with micro-level crash-specific factors (including rider behaviour, helmet use, and roadway conditions), and (2) utilising a novel partially temporally constrained random-parameters logit (RPL) model that accounts for heterogeneity in means and variances across different spatiotemporal dimensions. Using data from Florida between 2020 and 2022, counties were categorised into low-medium (LM) and high-risk (HR) groups based on crash frequency. The proposed methodology advances the analysis of crash severity by demonstrating the importance of spatial stratification (comparing LM and HR counties) for accurate policy insights. It also validates model robustness through transferability tests (χ ⊃2 > 164, p < 0.001) and out-of-sample predictions. Key findings reveal significant spatiotemporal instability, including (1) Macro–micro interactions: In HR counties during the COVID-19 pandemic (2020), higher population density increased fatalities, whereas in 2021-2022, it appeared to reduce severity, suggesting behavioural shifts related to the pandemic. (2) Spatial disparities: Male riders in HR counties had a 6.95% higher likelihood of fatality compared to 0.60% in LM counties, underscoring the need for targeted enforcement. (3) Temporal variability: Helmet use decreased fatalities in LM counties but had mixed effects in HR counties, potentially due to risk compensation factors such as increased speeding. This research bridges theoretical and practical gaps by integrating macro and micro perspectives, presenting a replicable model for regional motorcycle safety planning, and highlighting the significance of spatiotemporal heterogeneity in traffic safety analyses.

Recommended citation: Wang, C., Cui, P., Abdel-Aty, M., & Easa, S. M. (2025). Spatiotemporal disparities in macro-microscopic properties of motorcycle injury level. Transportmetrica A: Transport Science. Published online: 03 Oct 2025.
Download Paper