The influence of traffic, geometric and context variables on urban crash types: A grouped random parameter multinomial logit approach

• Published in: Analytic methods in accident research 2020-12, Vol.28, p.100141, Article 100141
• Main Authors: Intini, Paolo, Berloco, Nicola, Fonzone, Achille, Fountas, Grigorios, Ranieri, Vittorio
• Format: Article
• Language: English
• Subjects: Crash types; Grouped random parameters; Multinomial logit; Road safety; Urban intersections; Urban segments
• ISSN: 2213-6657; 2213-6657
• DOI: 10.1016/j.amar.2020.100141

•A mixed logit structure linking crash type outcomes to several factors was used.•A grouped random parameters (at the segment/intersection level) approach was used.•Traffic, geometric, context-related factors were included as explanatory variables.•Significant variations in the predictor effects across segments/intersections were found.•The study framework was applied to identify high-risk sites for specific crash types. Numerous road safety studies have been dedicated to the estimation of crash frequency and injury severity models. However, previous research has shown that different factors may influence the occurrence of crashes of different types. In this study, a dataset including information from crashes occurred at segments and intersections of urban roads in Bari, Italy was used to estimate the likelihood of occurrence of various crash types. The crash types considered are: single-vehicle, angle, rear-end and sideswipe. Models were estimated through a mixed logit structure considering various crash types as outcomes of the dependent variable and several traffic, geometric and context-related factors as explanatory variables (both site- and crash-specific). To account for systematic, unobserved variations among the crashes occurred on the same segment or intersection, the grouped random parameters approach was employed. The latter allows the estimation of segment- or intersection-specific parameters for the variables resulting in random parameters. This approach allows assessing the variability of results across the observations for individual segments/intersections. Segment type and the presence of bus lanes were included as explanatory variables in the model of crash types for segments. Traffic volume per entering lane, total entering lanes, total number of zebra crossings and the balance between major and minor traffic volumes at intersections were included as explanatory variables in the model of crash types for intersections. Area type was included in both segment and intersection models. The typical traffic at the moment of the crash (from on-line traffic prediction tools) and the period of the day were associated with different crash type likelihoods for both segments and intersections. Significant variations in the effect of several predictors across different segments or intersections were identified. The applicability of the study framework is demonstrated, in terms of identifying roadway sites with anomalous tendencies or high-risk sites w