Methodological considerations with data uncertainty in road safety analysis

• Published in: Accident analysis and prevention Vol. 130; pp. 136 - 150
• Main Authors: Schlögl, Matthias, Stütz, Rainer
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.09.2019
• Subjects: Accident analysis; GIS; Linear referencing; Road safety; Uncertainty; GIS; 62-07; Uncertainty; Road safety; Linear referencing; Accident analysis
• ISSN: 0001-4575; 1879-2057; 1879-2057
• DOI: 10.1016/j.aap.2017.02.001

•Spatial and temporal aspects of road safety data uncertainty are pointed out.•Challenges related to linking different data sources are highlighted.•Methodological considerations are exemplary illustrated with data from Austria.•Careful data preparation is a prerequisite for profound accident analysis. The analysis of potential influencing factors that affect the likelihood of road accident occurrence has been of major interest for safety researchers throughout the recent decades. Even though steady methodological progresses were made over the years, several impediments pertaining to the statistical analysis of crash data remain. While issues related to methodological approaches have been subject to constructive discussion, uncertainties inherent to the most fundamental part of any analysis have been widely neglected: data. This paper scrutinizes data from various sources that are commonly used in road safety studies with respect to their actual suitability for applications in this area. Issues related to spatial and temporal aspects of data uncertainty are pointed out and their implications for road safety analysis are discussed in detail. These general methodological considerations are exemplary illustrated with data from Austria, providing suggestions and methods how to overcome these obstacles. Considering these aspects is of major importance for expediting further advances in road safety data analysis and thus for increasing road safety.