Evaluating the influence of crashes on driving risk using recurrent event models and Naturalistic Driving Study data

Chen, Chen; Guo, Feng · 2016 · openalex

DOI: 10.1080/02664763.2015.1134449

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 whether involvement in motor vehicle crashes influences subsequent driver behavior and driving risk, operating on the hypothesis that drivers become more cautious after experiencing a crash. The research addresses a gap in existing literature, which often relies on self-reported data or broad experience metrics that fail to isolate the specific impact of crash events. By utilizing objective naturalistic driving data, the authors aim to determine if crashes lead to measurable reductions in distraction and aggressive driving maneuvers, and whether these effects vary by gender or diminish over time. The analysis utilized data from the 100-Car Naturalistic Driving Study, focusing on 107 primary drivers who experienced 51 crashes. Driver behavior was assessed by measuring the proportion of 6-second baseline epochs where drivers engaged in moderate or complex secondary tasks (distractions). The authors sampled 882 baselines within 15-hour windows before and after each crash, employing mixed binomial regression models to account for within-driver correlation and confounding factors like age and gender. To evaluate driving risk, the study measured the intensity of safety-critical incidents (SCIs) and near-crashes (NCs). Because these events occur repeatedly, the researchers developed and compared four intensity-based recurrent event models. A simulation study identified the stratified frailty model as the most appropriate for the dataset, allowing for the analysis of event intensity relative to actual driving time across phases defined by crash occurrence. The results indicate that crashes have a statistically significant, positive effect on driving safety, though the magnitude varies by gender and metric. Regarding distraction, the proportion of baselines involving complex secondary tasks dropped significantly after crashes (odds ratio = 0.54; 95% CI [0.32, 0.93]), with this cautious behavior diminishing after approximately 50 hours of driving. For driving risk, male drivers showed a significant reduction in SCI intensity after both the first crash (intensity rate ratio = 0.82; 95% CI [0.693, 0.971]) and the second crash (ratio = 0.47; 95% CI [0.377, 0.59]). Male drivers also exhibited reduced NC intensity after the first crash (ratio = 0.52; 95% CI [0.314, 0.874]). In contrast, female drivers showed no significant change in SCI or NC intensity after the first crash but demonstrated a significant reduction in SCI intensity after the second crash (ratio = 0.43; 95% CI [0.342, 0.547]). The study concludes that crashes serve as a learning event that temporarily reduces risky driving behaviors, particularly distraction and aggressive maneuvers. However, this safety benefit is transient, fading within roughly 50 hours. The findings suggest that safety interventions should focus on prolonging this post-crash behavioral improvement. The authors note limitations regarding the small number of crashes and mild severity, recommending future studies with larger datasets to further explore these dynamics.

Key finding

Crash involvement leads to a temporary reduction in driver distraction and safety-critical incident intensity, with male drivers showing significant behavioral changes after the first crash and female drivers showing changes primarily after a second crash.

Methodology

naturalistic

Sample size: 107

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.

StageOutcomeToolModelPromptAttemptsCompleted
discover success 1 2026-05-28
archive success canonical_url 4 2026-06-06
extract success cached 3 2026-06-10
clean success clean 1 2026-06-07
chunk success chunk 1 2026-06-07
embed success embed Qwen/Qwen3-Embedding-8B 1 2026-06-07
enrich success openalex 4 2026-07-02
promote success 1 2026-06-04
summarize success llm qwen3.6-27b-prismaquant summ-v5 2 2026-06-10
tag success vector_similarity 15 2026-06-11
verify partial 2 2026-06-10

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified_with_issues.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).