Spatial econometric models to understand factors affecting older drivers at accident hotspots
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This study investigates the spatial distribution and causal factors of road accidents involving older drivers in the West Midlands region of the United Kingdom. The research is motivated by the disproportionate risk faced by older drivers, who accounted for 25% of driver fatalities and 36% of killed or seriously injured casualties in 2023 despite representing a smaller portion of the driving population. Aging-related declines in vision, cognitive function, and physical ability increase accident risks, particularly in complex traffic environments. The paper aims to identify specific environmental and driver characteristics that contribute to these accidents and to demonstrate the application of advanced spatial econometric models for analyzing such geographically clustered data. The methodology employs Local Indicators of Spatial Association (LISA) and Moran’s I to analyze accident hotspot clustering. To model the spatial relationships, the authors utilize spatial lag and spatial error models. A significant portion of the paper focuses on solving the ill-conditioned linear systems inherent in these spatial models. The authors introduce a novel preconditioned iterative integration-exponential method, inspired by iterative refinement, which uses residuals to correct numerical errors. This approach eliminates the need to manually determine regularization parameters, a limitation of previous methods like Tikhonov regularization. The study applies this method to first-order dynamical systems and nonsymmetric positive definite linear systems, establishing convergence theorems and demonstrating its robustness for highly ill-conditioned problems. The findings reveal that road accidents involving older drivers cluster significantly in major urban centers, particularly near complex junctions and areas with dense traffic. Spatial diagnostic tests, including log-likelihood, Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and likelihood ratio tests, indicate that the spatial error model provides the best fit for the accident data. The analysis identifies journey purpose, location, junction type, lighting, road surface condition, weather, gender, and time of day as the most significant predictors of accident risk. Older drivers are found to be especially vulnerable during school runs, social trips, and when navigating complex junctions due to slower reaction times and reduced ability to interpret dynamic traffic conditions. The significance of this research lies in its dual contribution to mathematical methodology and traffic safety policy. The proposed preconditioned iterative integration-exponential method offers a more robust solution for ill-conditioned spatial econometric problems, removing the subjectivity of parameter selection. For policymakers, the findings provide actionable insights for improving mobility and safety for older drivers. Recommendations include focusing on environmental design improvements, enhanced driver assessment protocols, technological aids, and targeted educational programs. By understanding the specific spatial and causal factors of accidents, interventions can be tailored to reduce risks in identified hotspots, addressing the challenges posed by an aging driving population.
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-06-20 |
| archive | success | canonical_url | — | — | 1 | 2026-06-26 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-20 |
| chunk | success | chunk | — | — | 1 | 2026-06-20 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-20 |
| promote | success | — | — | — | 1 | 2026-06-20 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-20 |
| verify | success | — | — | — | 1 | 2026-06-26 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.