Multilevel Logistic Regression Modeling for Crash Mapping in Metropolitan Areas

Abstract

The spatial nature of traffic crashes makes crash locations one of the most important and informative attributes of crash databases. It is however very likely that recorded crash locations in terms of easting and northing coordinates, distances from junctions, addresses, road names and types are inaccurately reported. Improving the quality of crash locations therefore has the potential to enhance the accuracy of many spatial crash analyses. The determination of correct crash locations usually requires a combination of crash and network attributes with suitable crash mapping methods. Urban road networks are more sensitive to erroneous matches due to high road density and inherent complexity. This paper presents a novel crash mapping method suitable for urban and metropolitan areas that matched all the crashes that occurred in London from 2010-2012. The method is based on a hierarchical data structure of crashes (i.e. candidate road links are nested within vehicles and vehicles nested within crashes) and employs a multilevel logistic regression model to estimate the probability distribution of mapping a crash onto a set of candidate road links. The road link with the highest probability is considered to be the correct segment for mapping the crash. This is based on the two primary variables: (a) the distance between the crash location and a candidate segment and (b) the difference between the vehicle direction just before the collision and the link direction. Despite the fact that road names were not considered due to limited availability of this variable in the applied crash database, the developed method provides a 97.1% (±1%) accurate matches (N=1,000). The method was compared with two simpler, non-probabilistic crash mapping algorithms and the results were used to demonstrate the effect of crash location data quality on a crash risk analysis.Keywords: crash location, crash mapping, metropolitan/urban networks, multilevel logistic regression Imprialou M.-I.M., Quddus M., Pitfield D.A. 3 INTRODUCTIONSustainable road safety programmes require constant enhancements to crash prevention policies and countermeasures. The precautionary measures should focus on the network areas where many crashes occur and the driving attitudes that are considered mostly responsible for crashes. Analysis of crash data aims to identify and explain the factors that lead to traffic crashes. The quality and reliability of the crash data that are used as an input for spatial crash analyses (e.g. identification of black spots, spatial crash modelling etc.) is closely related to the validity of their outcomes (e.g. [1][2][3][4][5]. The spatial nature of crashes makes crash locations as one of the most important and informative attributes of crash databases (5-7) that at the same time are very likely to be inaccurately reported (1-4, 7). Therefore, the refinement of the crash locations gives to the crash analyses the potential to improve in quality.Traffic crashes create major problems to society as they are related to personal injuri...

The Effect of Lighting on Crime Counts

“…There may be errors in the recording of these data as has been found in records of road traffic crashes. Errors have been found in crash location [52][53][54][55] and crash time [56]. Third, the current investigation chose to collect data for one week either side of the clock change, similar to previous investigation of RTCs [32] but shorter than the two [57] and five [58] weeks used by others.…”

Crash data quality for road safety research: Current state and future directions

“…The inaccuracies may be due to human error, equipment failure (when GPS is used), limited training of personnel or can be inherent to the reporting method (e.g. Brown et al, 2015;Imprialou et al, 2015). Existing crash mapping algorithms employ a broad range of methods and attributes of crash records in order to correct erroneous crash locations.…”

Exploring the Effects of Vehicular and Operational Characteristics on Bus Roadworthiness: An Application of the Multilevel Modeling Approach

“…Hierarchical structure poses the challenge as observations for units belonging to the same cluster (e.g., student, class, school) are not independent from one another ( 19 – 22 ), which violates the traditional model assumption of residual independence. Disregarding the dependence between observations within the same cluster is likely to cause statistical inaccuracies, including underestimation of standard errors, overstatement of statistical significance, and excessive Type I errors ( 20 , 23 – 31 ).…”