Poisson, Poisson-gamma and zero-inflated regression models of motor vehicle crashes: balancing statistical fit and theory

There has been considerable research conducted over the last 20 years focused on predicting motor vehicle crashes on transportation facilities. The range of statistical models commonly applied includes binomial, Poisson, Poisson-gamma (or negative binomial), zero-inflated Poisson and negative binomial models (ZIP and ZINB), and multinomial probability models. Given the range of possible modeling approaches and the host of assumptions with each modeling approach, making an intelligent choice for modeling motor vehicle crash data is difficult. There is little discussion in the literature comparing different statistical modeling approaches, identifying which statistical models are most appropriate for modeling crash data, and providing a strong justification from basic crash principles. In the recent literature, it has been suggested that the motor vehicle crash process can successfully be modeled by assuming a dual-state data-generating process, which implies that entities (e.g., intersections, road segments, pedestrian crossings, etc.) exist in one of two states—perfectly safe and unsafe. As a result, the ZIP and ZINB are two models that have been applied to account for the preponderance of “excess” zeros frequently observed in crash count data. The objective of this study is to provide defensible guidance on how to appropriate model crash data. We first examine the motor vehicle crash process using theoretical principles and a basic understanding of the crash process. It is shown that the fundamental crash process follows a Bernoulli trial with unequal probability of independent events, also known as Poisson trials. We examine the evolution of statistical models as they apply to the motor vehicle crash process, and indicate how well they statistically approximate the crash process. We also present the theory behind dual-state process count models, and note why they have become popular for modeling crash data. A simulation experiment is then conducted to demonstrate how crash data give rise to “excess” zeros frequently observed in crash data. It is shown that the Poisson and other mixed probabilistic structures are approximations assumed for modeling the motor vehicle crash process. Furthermore, it is demonstrated that under certain (fairly common) circumstances excess zeros are observed—and that these circumstances arise from low exposure and/or inappropriate selection of time/space scales and not an underlying dual state process. In conclusion, carefully selecting the time/space scales for analysis, including an improved set of explanatory variables and/or unobserved heterogeneity effects in count regression models, or applying small-area statistical methods (observations with low exposure) represent the most defensible modeling approaches for datasets with a preponderance of zeros

Forecasting crashes at the planning level: Simultaneous negative binomial crash model applied in Tucson, Arizona

At least two important transportation planning activities rely on planning-level crash prediction models. One is motivated by the Transportation Equity Act for the 21st Century, which requires departments of transportation and metropolitan planning organizations to consider safety explicitly in the transportation planning process. The second could arise from a need for state agencies to establish incentive programs to reduce injuries and save lives. Both applications require a forecast of safety for a future period. Planning-level crash prediction models for the Tucson, Arizona, metropolitan region are presented to demonstrate the feasibility of such models. Data were separated into fatal, injury, and property-damage crashes. To accommodate overdispersion in the data, negative binomial regression models were applied. To accommodate the simultaneity of fatality and injury crash outcomes, simultaneous estimation of the models was conducted. All models produce crash forecasts at the traffic analysis zone level. Statistically significant (p-values < 0.05) and theoretically meaningful variables for the fatal crash model included population density, persons 17 years old or younger as a percentage of the total population, and intersection density. Significant variables for the injury and property-damage crash models were population density, number of employees, intersections density, percentage of miles of principal arterial, percentage of miles of minor arterials, and percentage of miles of urban collectors. Among several conclusions it is suggested that planning-level safety models are feasible and may play a role in future planning activities. However, caution must be exercised with such models.

Modeling crash types: New insights into the effects of covariates on crashes at rural intersections

Many studies focused on the development of crash prediction models have resulted in aggregate crash prediction models to quantify the safety effects of geometric, traffic, and environmental factors on the expected number of total, fatal, injury, and/or property damage crashes at specific locations. Crash prediction models focused on predicting different crash types, however, have rarely been developed. Crash type models are useful for at least three reasons. The first is motivated by the need to identify sites that are high risk with respect to specific crash types but that may not be revealed through crash totals. Second, countermeasures are likely to affect only a subset of all crashes—usually called target crashes—and so examination of crash types will lead to improved ability to identify effective countermeasures. Finally, there is a priori reason to believe that different crash types (e.g., rear-end, angle, etc.) are associated with road geometry, the environment, and traffic variables in different ways and as a result justify the estimation of individual predictive models. The objectives of this paper are to (1) demonstrate that different crash types are associated to predictor variables in different ways (as theorized) and (2) show that estimation of crash type models may lead to greater insights regarding crash occurrence and countermeasure effectiveness. This paper first describes the estimation results of crash prediction models for angle, head-on, rear-end, sideswipe (same direction and opposite direction), and pedestrian-involved crash types. Serving as a basis for comparison, a crash prediction model is estimated for total crashes. Based on 837 motor vehicle crashes collected on two-lane rural intersections in the state of Georgia, six prediction models are estimated resulting in two Poisson (P) models and four NB (NB) models. The analysis reveals that factors such as the annual average daily traffic, the presence of turning lanes, and the number of driveways have a positive association with each type of crash, whereas median widths and the presence of lighting are negatively associated. For the best fitting models covariates are related to crash types in different ways, suggesting that crash types are associated with different precrash conditions and that modeling total crash frequency may not be helpful for identifying specific countermeasures.

Assessing wet crashes for the skid resistance management of road assets

Australia, road crash trauma costs the nation A$15 billion annually whilst the US estimates an economic impact of around US$ 230 billion on its network. Worldwide economic cost of road crashes is estimated to be around US$ 518 billion each year. Road accidents occur due to a number of factors including driver behaviour, geometric alignment, vehicle characteristics, environmental impacts, and the type and condition of the road surfacing. Skid resistance is considered one of the most important road surface characteristics because it has a direct effect on traffic safety. In 2005, Austroads (the Association of Australian and New Zealand Road Transport and Traffic Authorities) published a guideline for the management of skid resistance and Queensland Department of Main Roads (QDMR) developed a skid resistance management plan (SRMP). The current QDMR strategy is based on rationale analytical methodology supported by field inspection with related asset management decision tools. The Austroads’s guideline and QDMR's skid resistance management plan have prompted QDMR to review its skid resistance management practice. As a result, a joint research project involving QDMR, Queensland University of Technology (QUT) and the Corporative Research Centre for Integrated Engineering Asset Management (CRC CIEAM) was formed. The research project aims at investigating whether there is significant relationship between road crashes and skid resistance on Queensland’s road networks. If there is, the current skid resistance management practice of QDMR will be reviewed and appropriate skid resistance investigatory levels will be recommended. This paper presents analysis results in assessing the relationship between wet crashes and skid resistance on Queensland roads. Attributes considered in the analysis include surface types, annual average daily traffic (AADT), speed and seal age.

Visual assessment of pedestrian crashes

Of the numerous factors that play a role in fatal pedestrian collisions, the time of day, day of the week, and time of year can be significant determinants. More than 60% of all pedestrian collisions in 2007 occurred at night, despite the presumed decrease in both pedestrian and automobile exposure during the night. Although this trend is partially explained by factors such as fatigue and alcohol consumption, prior analysis of the Fatality Analysis Reporting System database suggests that pedestrian fatalities increase as light decreases after controlling for other factors. This study applies graphical cross-tabulation, a novel visual assessment approach, to explore the relationships among collision variables. The results reveal that twilight and the first hour of darkness typically observe the greatest frequency of pedestrian fatal collisions. These hours are not necessarily the most risky on a per mile travelled basis, however, because pedestrian volumes are often still high. Additional analysis is needed to quantify the extent to which pedestrian exposure (walking/crossing activity) in these time periods plays a role in pedestrian crash involvement. Weekly patterns of pedestrian fatal collisions vary by time of year due to the seasonal changes in sunset time. In December, collisions are concentrated around twilight and the first hour of darkness throughout the week while, in June, collisions are most heavily concentrated around twilight and the first hours of darkness on Friday and Saturday. Friday and Saturday nights in June may be the most dangerous times for pedestrians. Knowing when pedestrian risk is highest is critically important for formulating effective mitigation strategies and for efficiently investing safety funds. This applied visual approach is a helpful tool for researchers intending to communicate with policy-makers and to identify relationships that can then be tested with more sophisticated statistical tools.

Should we be concerned about alcohol in bicycle crashes?

This paper sets out to examine from published literature and crash data analyses whether alcohol in bicycle crashes is an issue about which we should be concerned. It discusses factors that have the potential to increase the number of bicycle crashes in which alcohol is involved (such growth in the size and diversity of the cyclist population, and balance and coordination demands) and factors which may reduce the importance of alcohol in bicycle crashes (such as time of data factors and child riders). It also examines data availability issues that contribute to difficulties in determining the true magnitude of the issue. Methods: This paper reviews previous research and reports analyses of data from Queensland, Australia, that examine the role of alcohol in Police-reported road crashes. In Queensland it is an offence to ride a bicycle or drive a motor vehicle with a BAC exceeding 0.05% (or lower for novice and professional drivers). Results: In the five years 2003-2007, alcohol was reported as involved in 165 bicycle crashes (4%). The bicycle rider was coded as “under the influence” or “over the prescribed BAC limit” in 15 were single unit crashes (12%). In multi-vehicle bicycle crashes, alcohol involvement was reported for 16 cyclists (0.4%) and 110 operators of other vehicles (3%). Additional analyses including characteristics of the cyclist crashes involving alcohol and the importance of missing data will be discussed in the paper. Conclusion: The increase in participation in cycling and the vulnerability of cyclists to injuries support the need to examine the role of alcohol in bicycle crashes. Current data suggest that alcohol on the part of the vehicle driver is a larger concern than alcohol on the part of the cyclist, but improvements in data collection are needed before more precise conclusions can be drawn.

Risk factors for fatal crashes in rural Australia

This paper presents findings from the rural and remote road safety study, conducted in Queensland, Australia, from March 2004 till June 2007, and compares fatal crashes and non-fatal but serious crashes in respect of their environmental, vehicle and operator factors. During the study period there were 613 non-fatal crashes resulting in 684 hospitalised casualties and 119 fatal crashes resulting in 130 fatalities. Additional information from police sources was available on 103 fatal and 309 non-fatal serious crashes. Over three quarters of both fatal and hospitalised casualties were male and the median age in both groups was 34 years. Fatal crashes were more likely to involve speed, alcohol and violations of road rules and fatal crash victims were 2 and a 1/2 times more likely to be unrestrained inside the vehicle than non-fatal casualties, consistent with current international evidence. After controlling for human factors, vehicle and road conditions made a minimal contribution to the seriousness of the crash outcome. Targeted interventions to prevent fatalities on rural and remote roads should focus on reducing speed and drink driving and promoting seatbelt wearing.

Predicting single-vehicle fatal crashes for two-lane rural highways in Southeastern United States

The rural two-lane highway in the southeastern United States is frequently associated with a disproportionate number of serious and fatal crashes and as such remains a focus of considerable safety research. The Georgia Department of Transportation spearheaded a regional fatal crash analysis to identify various safety performances of two-lane rural highways and to offer guidance for identifying suitable countermeasures with which to mitigate fatal crashes. The fatal crash data used in this study were compiled from Alabama, Georgia, Mississippi, and South Carolina. The database, developed for an earlier study, included 557 randomly selected fatal crashes from 1997 or 1998 or both (this varied by state). Each participating state identified the candidate crashes and performed physical or video site visits to construct crash databases with enhance site-specific information. Motivated by the hypothesis that single- and multiple-vehicle crashes arise from fundamentally different circumstances, the research team applied binary logit models to predict the probability that a fatal crash is a single-vehicle run-off-road fatal crash given roadway design characteristics, roadside environment features, and traffic conditions proximal to the crash site. A wide variety of factors appears to influence or be associated with single-vehicle fatal crashes. In a model transferability assessment, the authors determined that lane width, horizontal curvature, and ambient lighting are the only three significant variables that are consistent for single-vehicle run-off-road crashes for all study locations.

The influence of underreported crashes on hotspot identification

Hot spot identification (HSID) plays a significant role in improving the safety of transportation networks. Numerous HSID methods have been proposed, developed, and evaluated in the literature. The vast majority of HSID methods reported and evaluated in the literature assume that crash data are complete, reliable, and accurate. Crash under-reporting, however, has long been recognized as a threat to the accuracy and completeness of historical traffic crash records. As a natural continuation of prior studies, the paper evaluates the influence that under-reported crashes exert on HSID methods. To conduct the evaluation, five groups of data gathered from Arizona Department of Transportation (ADOT) over the course of three years are adjusted to account for fifteen different assumed levels of under-reporting. Three identification methods are evaluated: simple ranking (SR), empirical Bayes (EB) and full Bayes (FB). Various threshold levels for establishing hotspots are explored. Finally, two evaluation criteria are compared across HSID methods. The results illustrate that the identification bias—the ability to correctly identify at risk sites--under-reporting is influenced by the degree of under-reporting. Comparatively speaking, crash under-reporting has the largest influence on the FB method and the least influence on the SR method. Additionally, the impact is positively related to the percentage of the under-reported PDO crashes and inversely related to the percentage of the under-reported injury crashes. This finding is significant because it reveals that despite PDO crashes being least severe and costly, they have the most significant influence on the accuracy of HSID.

Identifying relationship between skid resistance, road characteristics and crashes using probability-based risk approach

Road accidents are of great concerns for road and transport departments around world, which cause tremendous loss and dangers for public. Reducing accident rates and crash severity are imperative goals that governments, road and transport authorities, and researchers are aimed to achieve. In Australia, road crash trauma costs the nation A$ 15 billion annually. Five people are killed, and 550 are injured every day. Each fatality costs the taxpayer A$1.7 million. Serious injury cases can cost the taxpayer many times the cost of a fatality. Crashes are in general uncontrolled events and are dependent on a number of interrelated factors such as driver behaviour, traffic conditions, travel speed, road geometry and condition, and vehicle characteristics (e.g. tyre type pressure and condition, and suspension type and condition). Skid resistance is considered one of the most important surface characteristics as it has a direct impact on traffic safety. Attempts have been made worldwide to study the relationship between skid resistance and road crashes. Most of these studies used the statistical regression and correlation methods in analysing the relationships between skid resistance and road crashes. The outcomes from these studies provided mix results and not conclusive. The objective of this paper is to present a probability-based method of an ongoing study in identifying the relationship between skid resistance and road crashes. Historical skid resistance and crash data of a road network located in the tropical east coast of Queensland were analysed using the probability-based method. Analysis methodology and results of the relationships between skid resistance, road characteristics and crashes are presented.

Using data mining on road asset management data in analysing road crashes

Road safety is a major concern worldwide. Road safety will improve as road conditions and their effects on crashes are continually investigated. This paper proposes to use the capability of data mining to include the greater set of road variables for all available crashes with skid resistance values across the Queensland state main road network in order to understand the relationships among crash, traffic and road variables. This paper presents a data mining based methodology for the road asset management data to find out the various road properties that contribute unduly to crashes. The models demonstrate high levels of accuracy in predicting crashes in roads when various road properties are included. This paper presents the findings of these models to show the relationships among skid resistance, crashes, crash characteristics and other road characteristics such as seal type, seal age, road type, texture depth, lane count, pavement width, rutting, speed limit, traffic rates intersections, traffic signage and road design and so on.

Identifying differences in wet and dry road crashes using data mining

It is commonly accepted that wet roads have higher risk of crash than dry roads; however, providing evidence to support this assumption presents some difficulty. This paper presents a data mining case study in which predictive data mining is applied to model the skid resistance and crash relationship to search for discernable differences in the probability of wet and dry road segments having crashes based on skid resistance. The models identify an increased probability of wet road segments having crashes for mid-range skid resistance values.

A focus on risk compensation and its role in motor vehicle crashes

• What is risk compensation, and why is relevant to motor vehicle crashes? • Recent simulator work that revealed risk compensation • Current and future work on risk compensation

The effectiveness of traffic policing in reducing traffic crashes

Many governments throughout the world rely heavily on traffic law enforcement programs to modify driver behaviour and enhance road safety. There are two related functions of traffic law enforcement, apprehension and deterrence, and these are achieved through three processes: the establishment of traffic laws, the policing of those laws, and the application of penalties and sanctions to offenders. Traffic policing programs can vary by visibility (overt or covert) and deployment methods (scheduled and non-scheduled), while sanctions can serve to constrain, deter or reform offending behaviour. This chapter will review the effectiveness of traffic law enforcement strategies from the perspective of a range of high-risk, illegal driving behaviours including drink/drug driving, speeding, seat belt use and red light running. Additionally, this chapter discusses how traffic police are increasingly using technology to enforce traffic laws and thus reduce crashes. The chapter concludes that effective traffic policing involves a range of both overt and covert operations and includes a mix of automatic and more traditional manual enforcement methods. It is important to increase both the perceived and actual risk of detection by ensuring that traffic law enforcement operations are sufficiently intensive, unpredictable in nature and conducted as widely as possible across the road network. A key means of maintaining the unpredictability of operations is through the random deployment of enforcement and/or the random checking of drivers. The impact of traffic enforcement is also heightened when it is supported by public education campaigns. In the future, technological improvements will allow the use of more innovative enforcement strategies. Finally, further research is needed to continue the development of traffic policing approaches and address emerging road safety issues.

Severity of driver injury and vehicle damage in traffic crashes at intersections: A Bayesian hierarchical analysis

Most crash severity studies ignored severity correlations between driver-vehicle units involved in the same crashes. Models without accounting for these within-crash correlations will result in biased estimates in the factor effects. This study developed a Bayesian hierarchical binomial logistic model to identify the significant factors affecting the severity level of driver injury and vehicle damage in traffic crashes at signalized intersections. Crash data in Singapore were employed to calibrate the model. Model fitness assessment and comparison using Intra-class Correlation Coefficient (ICC) and Deviance Information Criterion (DIC) ensured the suitability of introducing the crash-level random effects. Crashes occurring in peak time, in good street lighting condition, involving pedestrian injuries are associated with a lower severity, while those in night time, at T/Y type intersections, on right-most lane, and installed with red light camera have larger odds of being severe. Moreover, heavy vehicles have a better resistance on severe crash, while crashes involving two-wheel vehicles, young or aged drivers, and the involvement of offending party are more likely to result in severe injuries.