Meta-analysis of the effect of road safety campaigns on accidents

A meta-analysis of 67 studies evaluating the effect of road safety campaigns on accidents is reported. A total of 119 results were extracted from the studies, which were reported in 12 different countries between 1975 and 2007. After allowing for publication bias and heterogeneity of effects, the weighted average effect of road safety campaigns is a 9% reduction in accidents (with 95% confidence that the weighted average is between −12 and −6%). To account for the variability of effects measured across studies, data were collected to characterise aspects of the campaign and evaluation design associated with each effect, and analysed to identify a model of seven campaign factors for testing by meta-regression. The model was tested using both fixed and random effect meta-regression, and dependency among effects was accounted for by aggregation. These analyses suggest positive associations between accident reduction and the use of personal communication or roadside media as part of a campaign delivery strategy. Campaigns with a drink-driving theme were also associated with greater accident reductions, while some of the analyses suggested that accompanying enforcement and short campaign duration (less than one month) are beneficial. Overall the results are consistent with the idea that campaigns can be more effective in the short term if the message is delivered with personal communication in a way that is proximal in space and time to the behaviour targeted by the campaign.     

Multilevel modelling for the regional effect of enforcement on road accidents

This paper investigates the effect of the intensification of Police enforcement on the number of road accidents at national and regional level in Greece, focusing on one of the most important road safety violations: drinking-and-driving. Multilevel negative binomial models are developed to describe the effect of the intensification of alcohol enforcement on the reduction of road accidents in different regions of Greece. Moreover, two approaches are explored as far as regional clustering is concerned: the first one concerns an ad hoc geographical clustering and the second one is based on the results of mathematical cluster analysis through demographic, transport and road safety characteristics. Results indicate that there are significant spatial dependences among road accidents and enforcement. Additionally, it is shown that these dependences are more efficiently interpreted when regions are determined on the basis of qualitative similarities than on the basis of geographical adjacency.     

The influence of the infrastructure characteristics in urban road accidents occurrence

This paper summarizes the result of a study regarding the creation of tools that can be used in intervention methods in the planning and management of urban road networks in Portugal.     

A novel framework for improvement of road accidents considering decision-making styles of drivers in a large metropolitan area

Road accidents can be caused by different factors such as human factors. Quality of the decision-making process of drivers could have a considerable impact on preventing disasters. The main objective of this study is the analysis of factors affecting road accidents by considering the severity of accidents and decision-making styles of drivers. To this end, a novel framework is proposed based on data envelopment analysis (DEA) and statistical methods (SMs) to assess the factors affecting road accidents. In this study, for the first time, dominant decision-making styles of drivers with respect to severity of injuries are identified. To show the applicability of the proposed framework, this research employs actual data of more than 500 samples in Tehran, Iran. The empirical results indicate that the flexible decision style is the dominant style for both minor and severe levels of accident injuries.     

Prediction of road accidents: A Bayesian hierarchical approach

In this paper a novel methodology for the prediction of the occurrence of road accidents is presented. The methodology utilizes a combination of three statistical methods: (1) gamma-updating of the occurrence rates of injury accidents and injured road users, (2) hierarchical multivariate Poisson-lognormal regression analysis taking into account correlations amongst multiple dependent model response variables and effects of discrete accident count data e.g. over-dispersion, and (3) Bayesian inference algorithms, which are applied by means of data mining techniques supported by Bayesian Probabilistic Networks in order to represent non-linearity between risk indicating and model response variables, as well as different types of uncertainties which might be present in the development of the specific models.     

Effects of road lighting: An analysis based on Dutch accident statistics 1987-2006

This study estimates the safety effect of road lighting on accidents in darkness on Dutch roads, using data from an interactive database containing 763,000 injury accidents and 3.3 million property damage accidents covering the period 1987-2006. Two estimators of effect are used, and the results are combined by applying techniques of meta-analysis. Injury accidents are reduced by 50%. This effect is larger than the effects found in most of the earlier studies. The effect on fatal accidents is slightly larger than the effect on injury accidents. The effect during twilight is about 2/3 of the effect in darkness. The effect of road lighting is significantly smaller during adverse weather and road surface conditions than during fine conditions. The effects on pedestrian, bicycle and moped accidents are significantly larger than the effects on automobile and motorcycle accidents. The risk of injury accidents was found to increase in darkness. The average increase in risk was estimated to 17% on lit rural roads and 145% on unlit rural roads. The average increase in risk during rainy conditions is about 50% on lit rural roads and about 190% on unlit rural roads. The average increase in risk with respect to pedestrian accidents is about 140% on lit rural roads and about 360% on unlit rural roads.     

Impact of infrastructure and local environment on road unsafety. Logistic modeling with spatial autocorrelation

This article aims at modeling the impact of road characteristics and local spatial environment on road (un)safety. The study applies to Belgium where some 1500 people are killed annually on the roads. This statistic corresponds to one of the highest risks in Europe. Road unsafety is expressed here as whether an hectometer of road belongs to a black zone; a black zone is defined as a segment of road where roads accidents are concentrated. Logistic modeling including spatial autocorrelation is used and compared to non-spatial regression. It is shown that a spatial model is needed to avoid biased estimated parameters. Results show that local environment and road infrastructure play a substantial role in the co-occurrence of road accidents. Hence, education and enforcement cannot be the only measures taken to reach a sustainable road safety. To attain their objectives of accident reduction, public authorities should also take their responsibilities in the matter of securing road infrastructure.     

The influence of economic incentives linked to road safety indicators on accidents: The case of toll concessions in Spain

The goal of this paper is to evaluate whether the incentives incorporated in toll highway concession contracts in order to encourage private operators to adopt measures to reduce accidents are actually effective at improving safety. To this end, we implemented negative binomial regression models using information about highway characteristics and accident data from toll highway concessions in Spain from 2007 to 2009. Our results show that even though road safety is highly influenced by variables that are not managed by the contractor, such as the annual average daily traffic (AADT), the percentage of heavy vehicles on the highway, number of lanes, number of intersections and average speed; the implementation of these incentives has a positive influence on the reduction of accidents and injuries. Consequently, this measure seems to be an effective way of improving safety performance in road networks.     

An update on the association between setting quantified road safety targets and road fatality reduction

This note is intended to rectify estimates provided in a previous paper (Wong et al., 2006) of the short-term effectiveness in terms of road fatality reduction of the setting of quantified road safety targets using data from 14 OECD countries during the period 1980–1999. This work is important in measuring the association between target setting and road safety improvement, because such targets are intended to serve as a useful tool to motivate timely road safety measures by the road authorities and others. The estimates to be rectified were based on before-and-after analysis using a comparison group of countries for each country that had set a target. This note first provides a correction to the qualification test for the inclusion of a country in any particular comparison group. It then presents the numerical effects of this correction on the estimates of the effectiveness of setting quantified road safety targets, both in individual countries and across the whole group of countries that set targets in the relevant period. Finally, impacts on the findings of the previous paper are discussed, with the conclusion that the changes in those of the numerical estimates that are affected do not alter the main message of the paper.     

Analysis of traffic accidents on rural highways using Latent Class Clustering and Bayesian Networks

One of the principal objectives of traffic accident analyses is to identify key factors that affect the severity of an accident. However, with the presence of heterogeneity in the raw data used, the analysis of traffic accidents becomes difficult. In this paper, Latent Class Cluster (LCC) is used as a preliminary tool for segmentation of 3229 accidents on rural highways in Granada (Spain) between 2005 and 2008. Next, Bayesian Networks (BNs) are used to identify the main factors involved in accident severity for both, the entire database (EDB) and the clusters previously obtained by LCC. The results of these cluster-based analyses are compared with the results of a full-data analysis. The results show that the combined use of both techniques is very interesting as it reveals further information that would not have been obtained without prior segmentation of the data. BN inference is used to obtain the variables that best identify accidents with killed or seriously injured. Accident type and sight distance have been identify in all the cases analysed; other variables such as time, occupant involved or age are identified in EDB and only in one cluster; whereas variables vehicles involved, number of injuries, atmospheric factors, pavement markings and pavement width are identified only in one cluster.     

A semiparametric negative binomial generalized linear model for modeling over-dispersed count data with a heavy tail: Characteristics and applications to crash data

Crash data can often be characterized by over-dispersion, heavy (long) tail and many observations with the value zero. Over the last few years, a small number of researchers have started developing and applying novel and innovative multi-parameter models to analyze such data. These multi-parameter models have been proposed for overcoming the limitations of the traditional negative binomial (NB) model, which cannot handle this kind of data efficiently. The research documented in this paper continues the work related to multi-parameter models. The objective of this paper is to document the development and application of a flexible NB generalized linear model with randomly distributed mixed effects characterized by the Dirichlet process (NB-DP) to model crash data. The objective of the study was accomplished using two datasets. The new model was compared to the NB and the recently introduced model based on the mixture of the NB and Lindley (NB-L) distributions. Overall, the research study shows that the NB-DP model offers a better performance than the NB model once data are over-dispersed and have a heavy tail. The NB-DP performed better than the NB-L when the dataset has a heavy tail, but a smaller percentage of zeros. However, both models performed similarly when the dataset contained a large amount of zeros. In addition to a greater flexibility, the NB-DP provides a clustering by-product that allows the safety analyst to better understand the characteristics of the data, such as the identification of outliers and sources of dispersion.     

Measuring errors and violations on the road: A bifactor modeling approach to the Driver Behavior Questionnaire

The Driver Behavior Questionnaire (DBQ) is a self-report measure of driving behavior that has been widely used over more than 20 years. Despite this wealth of evidence a number of questions remain, including understanding the correlation between its violations and errors sub-components, identifying how these components are related to crash involvement, and testing whether a DBQ based on a reduced number of items can be effective. We address these issues using a bifactor modeling approach to data drawn from the UK Cohort II longitudinal study of novice drivers. This dataset provides observations on 12,012 drivers with DBQ data collected at .5, 1, 2 and 3 years after passing their test. A bifactor model, including a general factor onto which all items loaded, and specific factors for ordinary violations, aggressive violations, slips and errors fitted the data better than correlated factors and second-order factor structures. A model based on only 12 items replicated this structure and produced factor scores that were highly correlated with the full model. The ordinary violations and general factor were significant independent predictors of crash involvement at 6 months after starting independent driving. The discussion considers the role of the general and specific factors in crash involvement.     

The impacts of speed cameras on road accidents: An application of propensity score matching methods

This paper aims to evaluate the impacts of speed limit enforcement cameras on reducing road accidents in the UK by accounting for both confounding factors and the selection of proper reference groups. The propensity score matching (PSM) method is employed to do this. A naïve before and after approach and the empirical Bayes (EB) method are compared with the PSM method. A total of 771 sites and 4787 sites for the treatment and the potential reference groups respectively are observed for a period of 9 years in England. Both the PSM and the EB methods show similar results that there are significant reductions in the number of accidents of all severities at speed camera sites. It is suggested that the propensity score can be used as the criteria for selecting the reference group in before-after control studies. Speed cameras were found to be most effective in reducing accidents up to 200 meters from camera sites and no evidence of accident migration was found.     

How to Perform Continuous Sampling (CSP), Vol. 2: ASQC Basic Reference Series

Modeling recurrent event data is of current interest in statistics and engineering. This article proposes a framework for incorporating covariates in flowgraph models, with application to recurrent event data in systems reliability settings. A flowgraph is a generalized transition graph (GTG) originally developed to model total system waiting times for semi-Markov processes. The focus of flowgraph models is expanded by linking covariates into branch transition models, enriching the toolkit of available data analysis methods for complex stochastic systems. This article takes a Bayesian approach to the analysis of flowgraph models. Potential applications are not limited to engineering systems, but also extend to survival analysis.     

Coverage‐based vulnerability discovery modeling to optimize disclosure time using multiattribute approach

Software vulnerabilities trend over time has been proposed by various researchers and academicians in recent years. But none of them have considered operational coverage function in vulnerability discovery modeling. In this research paper, we have proposed a generalized statistical model that determines the relationship between operational coverage function and the number of expected vulnerabilities. During the operational phase, possible vulnerable sites are covered and vulnerabilities present at a particular site are discovered with some probability. We have assumed that the proposed model follows the nonhomogeneous Poisson process properties; thus, different distributions are used to formulate the model. The numerical illustration shows that the proposed model performs better and has the good fitness to the Google Chrome data. The second focus of this research paper is to evaluate the total cost incurred by the developer after software release and to identify the optimal vulnerability disclosure time through multiobjective utility function. The proposed vulnerability discovery helps in optimization. The optimal time problem depends on the combined effect of cost, risk, and effort.     

A novel Bayesian approach to reliability modeling: The benefits of uncertainty evaluation in the model selection procedure

This paper proposes a different likelihood formulation within the Bayesian paradigm for parameter estimation of reliability models. Moreover, the assessment of the uncertainties associated with parameters, the goodness of fit, and the model prediction of reliability are included in a systematic framework for better aiding the model selection procedure. Two case studies are appraised to highlight the contributions of the proposed method and demonstrate the differences between the proposed Bayesian formulation and an existing Bayesian formulation.     

Value of a statistical life in road safety: A benefit-transfer function with risk-analysis guidance based on developing country data

We model a value of statistical life (VSL) transfer function for application to road-safety engineering in developing countries through an income-disaggregated meta-analysis of scope-sensitive stated preference VSL data. The income-disaggregated meta-analysis treats developing country and high-income country data separately. Previous transfer functions are based on aggregated datasets that are composed largely of data from high-income countries. Recent evidence, particularly with respect to the income elasticity of VSL, suggests that the aggregate approach is deficient because it does not account for a possible change in income elasticity across income levels. Our dataset (a minor update of the OECD database published in 2012) includes 123 scope-sensitive VSL estimates from developing countries and 185 scope-sensitive estimates from high-income countries. The transfer function for developing countries gives VSL = 1.3732E−4 × (GDP per capita)^∧2.478, with VSL and GDP per capita expressed in 2005 international dollars (an international dollar being a notional currency with the same purchasing power as the U.S. dollar). The function can be applied for low- and middle-income countries with GDPs per capita above $1268 (with a data gap for very low-income countries), whereas it is not useful above a GDP per capita of about $20,000. The corresponding function built using high-income country data is VSL = 8.2474E+3 × (GDP per capita)^∧.6932; it is valid for high-income countries but over-estimates VSL for low- and middle-income countries. The research finds two principal significant differences between the transfer functions modeled using developing-country and high-income-country data, supporting the disaggregated approach. The first of these differences relates to between-country VSL income elasticity, which is 2.478 for the developing country function and .693 for the high-income function; the difference is significant at p < 0.001. This difference was recently postulated but not analyzed by other researchers. The second difference is that the traffic-risk context affects VSL negatively in developing countries and positively in high-income countries. The research quantifies uncertainty in the transfer function using parameters of the non-absolute distribution of relative transfer errors. The low- and middle-income function is unbiased, with a median relative transfer error of −.05 (95% CI: −.15 to .03), a 25th percentile error of −.22 (95% CI: −.29 to −.19), and a 75th percentile error of .20 (95% CI: .14 to .30). The quantified uncertainty characteristics support evidence-based approaches to sensitivity analysis and probabilistic risk analysis of economic performance measures for road-safety investments.     

基于分解法建模的泵送系统可靠性和可用性评估

为解决马尔科夫模型的状态空间爆炸问题,在状态空间分解法的基础上,建立单个部件(层次3)的SPN模型,然后根据子系统的组成结构建立单个子系统(层次2)的SPN模型.将子系统SPN模型集成为系统模型(层次1),得到系统模型的可达图及缩小的等效半马尔科夫图,按图建立系统模型的状态方程并求解,然后对系统进行可靠性和可用性评估.以泵送系统为例,验证分解法有助于解决状态空间爆炸问题,并使建模的难度和计算量减小,更易于操作.     

基于预测模型的虚拟试验不确定性分析方法

 工程系统中不可避免地存在各种参数不确定性,利用数值计算模型对系统进行虚拟试验时应进行不确定性分析.大型耗时计算模型的不确定性分析将面临严重的的计算复杂性问题,为此,针对工程应用中耗时计算模型,提出一种基于贝叶斯预测模型的不确定性分析仿真方法,采用概率分布为参数不确定性建模,研究系统响应预测不确定性的概率特征.泰勒杆撞击实例验证了该方法的高效性.