Ranking and selecting motor vehicle accident sites by using a hierarchical Bayesian model

Abstract

Identi®cation, ranking and selecting hazardous traf®c accident locations from a group under consideration is a fundamental goal for traf®c safety researchers. Few methods exist that can quantitatively, accurately and easily discriminate between sites that commonly have small and variable observation count periods. One method that embodies all these advantages is the hierarchical Bayesian model, the method proposed in this paper. The particular hierarchical Bayesian approach that we use incorporates expert knowledge about accident sites as a group believed a priori to be exchangeable, the Poisson assumption and a conjugate gamma prior. We then propose three natural strategies for ranking and selecting the most hazardous subgroup of accident locations. Also presented is an especially useful procedure that gives the probability that each particular site is worst and by how much it is worst. All proposed strategies are illustrated using previously published fatality accident data from 35 sites in Auckland, New Zealand.

Ranking and selecting dangerous crash locations: Correcting for the number of passengers and Bayesian ranking plots

“…Furthermore, the use of ranking procedures based on a hierarchical Bayes approach has been proposed in literature. These methods can handle the uncertainty and the great variability of accident data and produce a probabilistic ranking of the accident locations (Brijs et al (2003), Schlüter et al (1997)). …”

Predicting and identifying traffic hot spots applying neuro-fuzzy systems in intercity roads

“…Therefore, collecting comprehensive information about accident history is required within research limitation and reference population (Cafiso et al, 1999;Davis and Yang, 2001;Roger et al, 2004). Shortages of this method are as follows: 1) They needed a macro reference population, 2) Selecting reference population was completely optional, 3) Reference population always includes required elements for identifying unsafe situations (Hauer and Persaud, 1984;Schluter et al,1997;Washington and Cheng, 2005). It was found in studies that there may be two types of errors in identifying traffic hot spots which are false positive and false negative.…”

Evaluating traffic safety network screening: an initial framework utilizing the hierarchical Bayesian philosophy

“…Exploratory BUGS results -rates 145 2002). These crashes result in over 40,000 deaths per year (500,000 worldwide), and their impact on the national economy is estimated to be more than 230 billion dollars (NHTSA 1994(NHTSA , 1995(NHTSA ,1996(NHTSA ,1997(NHTSA ,1998(NHTSA ,1999(NHTSA ,2000(NHTSA ,2002aRunge 2002;Schluter et al 1997). Highway safety is identified as the top priority of the United States Department of Transportation (US DOT) (Runge 2002), and funds dedicated to addressing the problem are expected to increase substantially with the reauthorization of the federal transportation act.…”

“…For over 20 years, a technique known as empirical Bayesian (EB) has been proposed to address limitations of conventional SICL methods (Davis and Yang 2001;Hauer 1986Hauer ,1996Hauer et al 1986Hauer et al ,1988Hauer et al ,2002bHigle and Hecht 1989;Higle and Witkowski 1988;ITE 1999;Melcher et al 2001;Morris 1988;Pendleton 1988;Persaud 1988;Persaud and Hauer 1984;Saccomanno et al 2001;Schluter et al 1997). EB helps identify sites where mitigation might be most effective, increases confidence in estimates, and provides information on the relative safety of ranked sites.…”

“…The statistical bases for past SICL methods have also been questioned, primarily with respect to their inability to account for regression-to-the-mean (RTM) (Box et al 1970;Harwood et al 2002;Hauer 1986Hauer ,1996aHauer and Persaud 1987;Hauer et al 1986Hauer et al ,1988Hauer et al ,2002aHigle and Hecht 1989;Higle and Witkowski 1988;HSIS 2002;ITE 1999;MRI 2002aMRI ,b,c,d,e,f,g,2003aParker 1991;Persaud 1988a,b;Persaud and Hauer 1984;Raju et al 1998;Schluter et al 1997;Thomas 2001;TRB 1989TRB ,2001b. RTM results from the assumption that prior crash frequency is an unbiased estimate of future crashes.…”