Exploring the relationships between drivers' familiarity and two-lane rural road accidents. A multi-level study. (February 2018)

• Record Type: Journal Article
• Authors: Intini, Paolo; Berloco, Nicola; Colonna, Pasquale; Ranieri, Vittorio; Ryeng, Eirin
• Is Part Of: Accident analysis and prevention. Volume 111(2018)
• Page Start: 280
• Page End: 296
• Publication Date: 2018-02
• Subjects: Route familiarity -- Accident analysis -- Accident rate -- Logistic regression -- Accident type -- Interactions between drivers; Accidents -- Prevention -- Periodicals; Accident Prevention -- Periodicals; Accidents -- Prévention -- Périodiques; 363.106
• DOI: 10.1016/j.aap.2017.11.013
• Languages: English
• ISSNs: 0001-4575

Abstract

Highlights: Relationships between road familarity/unfamiliarity and accidents were searched. A multi-level analysis was conducted, from a macroscopic to more detailed levels. Familiarity was mainly measured based on involved drivers' distance from residence. Detailed analyses reveal the inquired relationships better than macro-indicators. Familiarity was confirmed as risk factor, unfamiliarity has some unclear aspects. Abstract: Previous research has suggested that drivers' route familiarity/unfamiliarity (using different definitions of familiarity), and the interactions between familiar and unfamiliar drivers, may affect both the driving performances and the likelihood of road crashes. The purpose of this study is to provide a contribution in the search for relationships between familiarity and crashes by: 1) introducing a measure of familiarity based on the distance from residence; 2) analyzing a traffic and accident dataset referred to rural two-lane sections of the Norwegian highways E6 and E39; 3) using a multi-level approach, based on different perspectives, from a macro analysis to more detailed levels. In the macro analyses, the accident rates computed for different seasons and for different summer traffic variation rates (used as indicators of the share of familiar drivers in the flow) were performed. At the second level, a logistic regression model was used to explain the familiarity/unfamiliarity of drivers (based on their distance from residence), through variables retrieved from the database. In the last step, an in-depth analysis considering also accident types and dynamics was conducted. In the macro analysis, no differences were found between accident rates in the different conditions. Whereas, as emerged from the detailed analyses, the factors: high traffic volume, low summer traffic variation, autumn/winter, minor intersections/driveways, speed limits <80 km/h, travel purposes (commuting/not working) are associated to higher odds of having familiar drivers involved in crashes; while the factors: high traffic volume, high summer traffic variation, summer, head on/rear end-angle crashes, heavy vehicles involved, travel purposes (not commuting), young drivers involved are associated to higher odds of finding unfamiliar drivers involved. To a minor extent, some indications arise from the in-depth analyses about crash types and dynamics, especially for familiar drivers. With regard to the definitions used in this article, the familiarity was confirmed as an influential factor on the accident risk, possibly due to distraction and dangerous behaviors, while the influence of being unfamiliar on the accident proneness has some unclarified aspects. However, crashes to unfamiliar drivers may cluster at sites showing high summer traffic variation and in summer months.