Showing papers by "Dominique Lord published in 2006"

Modeling Motor Vehicle Crashes Using Poisson-Gamma Models: Examining Effects of Low Sample Mean Values and Small Sample Size on Estimation of Fixed Dispersion Parameter

Abstract: This paper describes how there has been considerable research conducted on the development of statistical models for predicting crashes on highway facilities. Despite numerous advancements made for improving the estimation tools of statistical models, the most common probabilistic structure used for modeling motor vehicle crashes remains the traditional Poisson and Poisson-gamma (or Negative Binomial) distribution. When crash data exhibit over-dispersion, the Poisson-gamma model is usually the model of choice most favored by transportation safety modelers. Crash data collected for safety studies often have the unusual attributes of being characterized by low sample mean values. Studies have shown that the goodness-of-fit of statistical models produced from such datasets can be significantly affected. This issue has been defined as the “low mean problem” (LMP). Despite recent developments on methods to circumvent the LMP and test the goodness-of-fit of models developed using such datasets, no work has so far examined how the LMP affects the fixed dispersion parameter of Poisson-gamma models used for modeling motor vehicle crashes. The dispersion parameter plays an important role in many types of safety studies. The primary objective of this research project was to verify whether the LMP affects the estimation of the dispersion parameter and, if it is, to determine the magnitude of the problem. The secondary objective consisted of determining the effects of a mis-specified dispersion parameter on common analyses performed in highway safety studies. To accomplish the objectives of the study, a series of Poisson-gamma distributions were simulated using different values describing the mean, the dispersion parameter, and the sample size. Three estimators commonly used for estimating the dispersion parameter of Poisson-gamma models of motor vehicle crashes were evaluated: the method of moments (MM), the weighted regression (WR) and the Maximum Likelihood method (ML). In an attempt to complement the outcome of the simulation study, Poisson-gamma models were fitted to crash data collected in Toronto, Ont. characterized by a low sample mean and small sample size. The study shows that a low sample mean combined with a small sample size can seriously affect the estimation of the dispersion parameter, no matter which estimator is used within the estimation process. The probability the dispersion parameter becomes mis-specified increases significantly as the sample mean and sample size decrease. Consequently, the results show that a mis-specified dispersion parameter can significantly undermine empirical Bayes (EB) estimates as well as the estimation of confidence intervals for the gamma mean and predicted response. The paper ends with recommendations about minimizing the likelihood of producing Poisson-gamma models with a mis-specified dispersion parameter for modeling motor vehicle crashes.

Improving Pedestrian Safety at Unsignalized Crossings

Abstract: This report will be of interest to state, county, and city traffic engineers; transit agencies; roadway designers; and urban planners, as well as consultants for these groups and agencies. This material provides considerable information and useful guidance for improving pedestrian safety at unsignalized crossings. The report presents the edited final report and Appendix A, Guidelines for Pedestrian Crossing Treatments. Appendixes B through O of the contractor's final report are contained in TCRP Web-Only Document 30/NCHRP Web-Only Document 91, available on the CRP website. The objectives of the research were to (1) recommend selected engineering treatments to improve safety for pedestrians crossing high-volume and high-speed roadways at unsignalized locations, in particular those locations served by public transportation, and (2) recommend modifications to the Manual on Uniform Traffic Control Devices (MUTCD) pedestrian traffic signal warrant.

Exploration of Pedestrian Gap-Acceptance Behavior at Selected Locations

Abstract: This paper describes the efforts to evaluate pedestrian gap acceptance as part of a recent TCRP-NCHRP project Pedestrian crossing data were collected at 42 study sites in seven states From those sites, 45 pedestrian approaches had at least one crossing event where a pedestrian rejected at least one gap, and 11 of those approaches had at least 20 such crossing events Focusing on the 11 approaches, researchers evaluated the gap-acceptance behavior of crossing pedestrians with a two-part analysis: behavioral analysis and statistical analysis Behavioral analysis revealed that pedestrians did not always wait to cross the street when all lanes were completely clear; instead, they anticipated that the lanes would clear as they crossed and used a "rolling gap" to cross the street Statistical analysis revealed that the 11 approaches had 85th percentile accepted gaps between 53 and 94 s, with a trend of increasing gap length as crossing distance increased All the observed 85th percentile accepted gaps were

Role and application of accident modification factors within highway design process

Abstract: Over the past few years, transportation agencies have placed greater emphasis on incorporating safety into the highway design process. The main assumption is that a highway designed with an explicit attention to safety could significantly reduce the frequency and the severity of crashes. One tool that is rapidly gaining popularity is the use of accident modification factors (AMFs) to estimate the change in safety following implementation of a countermeasure, a significant change in a geometric design element, or a planned construction upgrade of a highway facility. Although AMFs will play an extremely important role in coming years, especially with the forthcoming publication of the "Highway Safety Manual" (HSM), no document currently exists that addresses how they potentially could be used with the process to design highways. This paper describes the role and application of AMFs in the highway geometric design process. More specifically, it provides guidelines for applying these factors within the preliminary design stage, for assessing design consistency, and for evaluating design exceptions. For each design application, a proposed procedure is described in detail. Finally, an example is provided to illustrate more clearly the descriptions of two of the procedures.

Median Barrier Guidelines for Texas

Abstract: Cross-median crashes are typically violent collisions with a high probability of multiple serious injuries and deaths. Previous research has shown that while cross-median crashes are not as common as other types of median-related crashes, they have a fatality rate that is disproportionately high. Many of these severe crossmedian crashes can be prevented with adequate barrier protection. However, the decision of whether or not to use a median barrier should take into account the overall changes in the characteristics of median-related crashes that will result, including the frequency and severity of barrier impacts. There is a need for an analysis of the characteristics of median-related crashes and an investigation into the use of median barriers to identify changes to current standards, specifications, and procedures for median barrier need, selection, and placement that will result in the highest practical level of safety. Under this project, new guidelines were developed to assist highway engineers with the evaluation of median barrier need such that the highest practical level of median safety can be achieved. The recommended guidelines are based on analysis of median-related crashes in Texas. The crash data were used to develop crash statistical models for the various types of median-related crashes. Based on the estimates derived from the frequency and severity models and crash costs used by the Texas Department of Transportation (TxDOT), an economic analysis of median barrier need was performed. Guidelines for installing median barriers on divided, access-controlled freeways were developed as a function of average annual daily traffic (AADT) and median width. Guidance based on mean cross-median crash rate was also developed to assist engineers with evaluation of median barrier need on existing highway facilities.

Strategies for Separating Trucks from Passenger Vehicles: Final Report

Abstract: Trucks constitute a large and growing segment of the traffic on Texas highways. In order to manage this growth, the Texas Department of Transportation needs to consider special or unique treatments for trucks such as truck lane restrictions, exclusive truck lanes, and exclusive truck facilities. This research addressed this topic for the state of Texas by developing tools for evaluating needs for special truck facilities, developing a truck route system, and developing recommendations for demonstration of a pilot system. Tasks required to accomplish these objectives began with a comprehensive literature review, to include the major corridor studies and an evaluation of special truck facilities implemented outside of Texas. The research then established criteria for each of the three levels of truck treatments, developed a plan to classify truck facilities, and developed an evaluation framework for these facilities. Based on these tasks, the research then investigated techniques for evaluating levels of service on truck facilities. The techniques were then applied to selected candidate corridors to demonstrate their utility. Finally, the research developed an action plan for implementation of the exclusive truck facilities.

Improving Pedestrian Safety at Unsignalized Crossings: Appendices B to O

Abstract: A recent research project jointly sponsored by the Transit Cooperative Research Program and the National Cooperative Highway Research Program had two main objectives: (a) recommend selected engineering treatments to improve safety for pedestrians crossing high-volume, high-speed roadways at unsignalized intersections, in particular those served by public transportation; and (b) recommend modifications to the "Manual on Uniform Traffic Control Devices" (MUTCD) pedestrian traffic signal warrant. The research team developed guidelines that can be used to select pedestrian crossing treatments for unsignalized intersections and midblock locations (Guidelines for Pedestrian Crossing Treatments). Quantitative procedures in the Guidelines use key input variables (such as pedestrian volume, street crossing width, traffic volume, etc.) to recommend one of four possible crossing treatment categories. The research team developed and presented recommendations to revise the MUTCD pedestrian warrant for traffic control signals to the National Committee on Uniform Traffic Control Devices. In accomplishing the two main study objectives, the research team also developed useful supporting information such as the findings from the field studies on walking speed and motorist compliance. Pedestrian walking speed recommendations were 3.5 ft/s (1.07 m/s) for general population and 3.0 ft/s (0.9 m/s) for older or less able population. Motorist compliance (yielding or stopping where required) was the primary measure of effectiveness for engineering treatments at unsignalized roadway crossings. The study found that the crossing treatment does have an impact on motorist compliance, and other factors influencing the treatment effectiveness were number of lanes being crossed and posted speed limit. This Web document contains the appendices (B through O) to the report (TCRP Report 112/NCHRP Report 562).

Effects of Sample Size on Goodness-of-Fit Statistic and Confidence Intervals of Crash Prediction Models Subjected to Low Sample Mean Values

Abstract: The statistical relationship between motor vehicle crashes and covariates can generally be modeled via generalized linear models (GLMs) with logarithmic links with errors distributed in a Poisson or Poisson-gamma manner. The scaled deviance and Pearson's χ2 have been proposed to test the statistical fit of GLMs. Recent studies have shown that these two estimators are not adequate for testing the goodness of fit (GOF) of GLMs when they are developed from data characterized by low sample mean values. To circumvent this problem, a testing method has been proposed to evaluate the GOF of such GLMs. Because this method can be time-consuming to implement, there is a need to determine whether it is sensitive to different sample sizes. The primary objective of this paper is to investigate the effects of decreasing sample sizes on the GOF testing technique. A secondary objective is to estimate how the reducing of sample size influences the confidence intervals of GLMs. To accomplish the objectives, GLMs were fit wi...

Evaluating Safety Effects of Daylight Savings Time on Fatal and Nonfatal Injury Crashes in Texas

Abstract: Daylight savings time (DST) takes place each year from 2:00 a.m. on the first Sunday in April to 2:00 a.m. on the last Sunday in October. It is expected that the changes to and from DST could positively or negatively affect safety. In fact, previous research has shown that these changes indeed affect safety, but many studies suffered from methodological limitations, including erroneously extrapolating short-term effects to long-term conditions. Given these limitations and contradictory results, there is a need to reexamine the effects of DST on the number of motor vehicle and pedestrian crashes. The primary objective of this study was to quantify the safety effects of DST through the use of statistical modeling for the morning and afternoon 5-h periods at the boundary delimiting dark and light conditions. The secondary objective consisted of conducting a before-and-after study to estimate the immediate effects following the change to and from DST. To quantify the potential safety effects of DST, model outputs were applied to different hypothetical time intervals during the year, including the newly proposed DST extension by the U.S. Congress. The results of the study showed that the statistical models performed as expected, with the exception of one model, with which the number of crashes decreased with an increasing proportion of daylight conditions in the 5-h period. The application of the models to different scenarios showed that the absolute difference in predicted crashes was relatively small between scenarios, a finding that supported previous work on this topic.

Effects of low sample mean values and small sample size on the estimation of the fixed dispersion parameter of Poisson-gamma models: A Bayesian Perspective

Abstract: There has been considerable research conducted on the development of statistical models for predicting motor vehicle crashes on highway facilities. Many of these developments were performed for the likelihood-based or frequentist modeling approach. Over the last few years, there has been a significant increase in the application hierarchical Bayes method for modeling motor vehicle crashes. Whether the inferences are estimates using the likelihood-based or the Bayesian method, the most common probabilistic structure used for modeling this type of data remains the traditional Poisson-gamma (or Negative Binomial) distribution. Crash data collected for highway safety studies often have the unusual attributes of being characterized by low sample mean values and, due to the prohibitive costs of collecting data, small sample sizes. Previous studies have shown that the dispersion parameter of Poisson-gamma models can be seriously mis-estimated when the coefficients are estimated using the maximum likelihood method (MLE) for these extreme conditions. Despite important work done on this topic for the MLE, nobody has so far examined how low sample mean values and small sample sizes affect the posterior mean of the dispersion parameter of Poisson-gamma models estimated using the hierarchical Bayes method. The inverse dispersion parameter (posterior mean) plays an important role in various types of highway safety studies, such as building confidence intervals for comparing the safety performance of different highway design alternatives and the application of the empirical Bayes (EB) method for refining the long-term mean of a highway entity, and is particularly important for practitioners who are not familiar with Bayesian methods. It is therefore vital to determine the conditions in which the inverse dispersion parameter may be mis-estimated for this category of models. To accomplish the objectives of this study, a series of Poisson-gamma distributions are simulated using different values describing the mean, the dispersion parameter, the sample size, and the prior specification. Non-informative and informative prior specifications are tested for determining the magnitude of the biases introduced by low sample mean values and small sample sizes. A series of Poisson-lognormal distributions are also simulated, in the light of recent work done by statisticians on this mixed distribution. The study shows that a dataset characterized by a low sample mean combined with a small sample size can seriously affect the estimation of the posterior mean of the dispersion parameter when a non-informative prior specification is used to characterize the gamma hyper-parameter. The risk of a mis-specified posterior mean can be greatly minimized when an appropriate informative prior distribution is used. Finally, the study shows that Poisson-lognormal models are recommended over Poisson-gamma models whenever crash data characterized by low sample mean values are used for developing crash prediction models.

Strategies for Separating Trucks from Passenger Vehicles: Truck Facility Guidebook

Abstract: The purpose of the Truck Facility Guidebook is to assist Texas Department of Transportation (TxDOT) engineers in understanding and replicating the process used by researchers to determine the need for truck treatments. This Guidebook provides criteria to assist TxDOT in choosing among the three types of truck facilities: 1) lane restrictions, 2) dedicated truck lanes, and 3) exclusive truck roadways. The Guidebook primarily focuses on the third type, exclusive truck roadways, and guides users through a nine-step process to determine the need for these facilities. The nine steps are as follows: (1) Acquire truck and non-truck volume; (2) Predict future truck and non-truck volume; (3) Determine desired level of service; (4) Determine number of truck roadway lanes; (5) Acquire crash data and estimate crash costs; (6) Determine initial construction cost; (7) Determine costs of delay and fuel consumption; (8) Determine user perspectives and other measures of acceptability; and (9) Total all benefits and costs.