The effects of left-turn traffic-calming treatments on conflicts and speeds in Washington,DC

Introduction: Left-turning vehicles pose considerable safety risks to pedestrians at intersections. Left-turn traffic-calming treatments are designed to slow left-turn traffic. This study examined the effects of one type of left-turn calming, the hardened-centerline treatment, on the numbers of conflicts between left-turning vehicles and pedestrians and left-turn speeds in Washington, DC. Method: Numbers of conflicts between left-turning vehicles and pedestrians, as well as left-turn speeds, were collected at selected intersections in Washington, DC, where the hardened centerline was installed, as well as at control intersections in the city where no treatment was installed, before and after installation. Poisson regression evaluated the change in numbers of conflicts associated with the hardened-centerline treatment. The effect of the treatment on left-turn speeds was estimated by a log-linear regression model, and the effect on the odds of left-turning vehicles exceeding 15 mph was estimated by a logistic regression model. Results: The treatment was associated with a 70.5% reduction in conflicts between left-turning vehicles and pedestrians, a 9.8% reduction in mean left-turn speeds, and a 67.1% reduction in the odds of left-turning vehicles exceeding 15 mph. All the reductions were statistically significant. Conclusions: The study demonstrates that the hardened-centerline treatment can reduce conflicts between left-turning vehicles and pedestrians, and slow down left-turn traffic at intersections. Practical applications: The treatment should be added to the toolbox for communities looking to improve pedestrian safety at intersections.     

Driver behavior analysis for right-turn drivers at signalized intersections using SHRP 2 naturalistic driving study data

IntroductionUnderstanding driver behavior is important for traffic safety and operation, especially at intersections where different traffic movements conflict. While most driver-behavior studies are based on simulation, this paper documents the analysis of driver-behavior at signalized intersections with the SHRP 2 Naturalistic Driving Study (NDS) data. This study analyzes the different influencing factors on the operation (speed control) and observation of right-turn drivers.MethodA total of 300 NDS trips at six signalized intersections were used, including the NDS time-series sensor data, the forward videos and driver face videos. Different factors of drivers, vehicles, roads and environments were studied for their influence on driver behavior. An influencing index function was developed and the index was calculated for each influencing factor to quantitatively describe its influencing level. The influencing index was applied to prioritize the factors, which facilitates development and selection of safety countermeasures to improve intersection safety. Drivers' speed control was analyzed under different conditions with consideration of the prioritized influencing factors.ResultsVehicle type, traffic signal status, conflicting traffic, conflicting pedestrian and driver age group were identified as the five major influencing factors on driver observation.ConclusionsThis research revealed that drivers have high acceleration and low observation frequency under Right-Turn-On-Red (RTOR), which constituted potential danger for other roadway users, especially for pedestrians.Practical applicationsAs speed has a direct influence on crash rates and severities, the revealed speed patterns of the different situations also benefit selection of safety countermeasures at signalized intersections.     

Distracting behaviors among teenagers and young,middle-aged,and older adult drivers when driving without and with warnings from an integrated vehicle safety system

IntroductionNegative reinforcement from crash warnings may reduce the likelihood that drivers engage in distracted driving. Alternatively, drivers may compensate for the perceived safety benefit of crash warnings by engaging in distractions more frequently, especially at higher speeds. The purpose of this study was to examine whether warning feedback from an integrated vehicle-based safety system affected the likelihood that various secondary behaviors were present among drivers ages 16–17, 20–30, 40–50, and 60–70.MethodParticipants drove an instrumented sedan with various collision warning systems for an extended period. Ten 5-second video clips were randomly sampled from driving periods at speeds above 25 mph and below 5 mph each week for each driver and coded for the presence of 11 secondary behaviors.ResultsAt least one secondary behavior was present in 46% of video clips; conversing with a passenger (17%), personal grooming (9%), and cellphone conversation (6%) were the most common. The likelihood that at least one secondary behavior was present was not significantly different during periods when drivers received warnings relative to periods without warnings. At least one secondary behavior was 21% more likely to be present at speeds below 5 mph relative to speeds above 25 mph; however, the effect of vehicle speed was not significantly affected by warning presence. Separate models for each of the five most common secondary behaviors also indicated that warnings had no significant effect on the likelihood that each behavior was present.ConclusionsCollision warnings were not associated with significant increases or decreases in the overall likelihood that teen and adult drivers engaged in secondary behaviors or the likelihood of the behaviors at speeds above 25 mph or below 5 mph.Practical applicationsThere was no evidence that forward collision warning and other technologies like those in this study will increase or decrease distracted driving.     

Examining teen driver crashes and the prevalence of distraction: Recent trends, 2007–2015

IntroductionTeen drivers crash at a much higher rate than adult drivers, with distractions found as a factor in nearly 6 out of 10 moderate-to-severe teen crashes. As the driving environment continues to rapidly evolve, it is important to examine the effect these changes may be having on our youngest and most vulnerable drivers.MethodThe purpose of this study was to identify types of vehicle crashes teens are most frequently involved in, as well as the distracting activities being engaged in leading up to these crashes, with a focus on identifying changes or trends over time. We examined 2,229 naturalistic driving videos involving drivers ages 16–19. These videos captured crashes occurring between 2007 and 2015. The data of interest for this study included crash type, behaviors drivers engaged in leading up to the collision, total duration of time the driver's eyes were off the forward roadway, and duration of the longest glance away from forward.ResultsRear-end crashes increased significantly (annual % change = 3.23 [2.40–4.05]), corresponding with national data trends. Among cell phone related crashes, a significant shift occurred, from talking/listening to operating/looking (annual % change = 4.22 [1.15–7.29]). Among rear-end crashes, there was an increase in the time drivers' eyes were off the road (β = 0.1527, P = 0.0004) and durations of longest glances away (β = 0.1020, P = 0.0014).ConclusionsFindings suggest that shifts in the way cell phones are being used, from talking/listening to operating/looking, may be a cause of the increasing number of rear-end crashes for teen drivers.Practical applicationsUnderstanding the role that cell phone use plays in teen driver crashes is extremely important. Knowing how and when teens are engaging in this behavior is the only way effective technologies can be developed for mitigating these crashes.     

Risk factors associated with bus accident severity in the United States: A generalized ordered logit model

IntroductionRecent years have witnessed a growing interest in improving bus safety operations worldwide. While in the United States buses are considered relatively safe, the number of bus accidents is far from being negligible, triggering the introduction of the Motor-coach Enhanced Safety Act of 2011.MethodThe current study investigates the underlying risk factors of bus accident severity in the United States by estimating a generalized ordered logit model. Data for the analysis are retrieved from the General Estimates System (GES) database for the years 2005–2009.ResultsResults show that accident severity increases: (i) for young bus drivers under the age of 25; (ii) for drivers beyond the age of 55, and most prominently for drivers over 65 years old; (iii) for female drivers; (iv) for very high (over 65 mph) and very low (under 20 mph) speed limits; (v) at intersections; (vi) because of inattentive and risky driving.     

Evaluation of an in-vehicle monitoring system (IVMS) to reduce risky driving behaviors in commercial drivers: Comparison of in-cab warning lights and supervisory coaching with videos of driving behavior

ProblemRoadway incidents are the leading cause of work-related death in the United States.MethodsThe objective of this research was to evaluate whether two types of feedback from a commercially available in-vehicle monitoring system (IVMS) would reduce the incidence of risky driving behaviors in drivers from two companies. IVMS were installed in 315 vehicles representing the industries of local truck transportation and oil and gas support operations, and data were collected over an approximate two-year period in intervention and control groups. In one period, intervention group drivers were given feedback from in-cab warning lights from an IVMS that indicated occurrence of harsh vehicle maneuvers. In another period, intervention group drivers viewed video recordings of their risky driving behaviors with supervisors, and were coached by supervisors on safe driving practices.ResultsRisky driving behaviors declined significantly more during the period with coaching plus instant feedback with lights in comparison to the period with lights-only feedback (ORadj = 0.61 95% CI 0.43–0.86; Holm-adjusted p = 0.035) and the control group (ORadj = 0.52 95% CI 0.33–0.82; Holm-adjusted p = 0.032). Lights-only feedback was not found to be significantly different than the control group's decline from baseline (ORadj = 0.86 95% CI 0.51–1.43; Holm-adjusted p > 0.05).ConclusionsThe largest decline in the rate of risky driving behaviors occurred when feedback included both supervisory coaching and lights.Practical applicationsSupervisory coaching is an effective form of feedback to improve driving habits in the workplace. The potential advantages and limitations of this IVMS-based intervention program are discussed.     

“They’re lunatics on the road”: Exploring the normative influences of parents,friends, and police on young novices’ risky driving decisions

PurposeYoung novice drivers experience significantly greater risk of being injured or killed in car crashes than older more experienced drivers. This research utilised a qualitative approach guided by the framework of Akers’ social learning theory. It explored young novice drivers’ perspectives on risky driving including rewards and punishments expected from and administered by parents, friends, and police, imitation of parents’ and friends’ driving, and advantages and disadvantages of risky driving.MethodsTwenty-one young drivers (12 females, 9 males) aged 16–25 years (M = 17.71 years, SD = 2.15) with a Learner (n = 11) or Provisional (n = 10) driver licence participated in individual or small group interviews.Findings and conclusionsContent analysis supported four themes: (1) rewards and (2) punishments for risky driving, and the influence of (3) parents and (4) friends. The young novice drivers differed in their vulnerability to the negative influences of friends and parents, with some novices advising they were able to resist risky normative influences whilst others felt they could not. The authority of the police as enforcers of road rules was either accepted and respected or seen as being used to persecute young novices. These findings suggest that road safety interventions should consider the normative influence of parents and friends on the risky and safe behaviour of young novices. Police were also seen as influential upon behaviour. Future research should explore the complicated relationship between parents, friends, the police, young novices, and their risky driving behaviour.     

Peer influence predicts speeding prevalence among teenage drivers

Impact on IndustryPreventing speed-related crashes could reduce costs and improve efficiency in the transportation industry.ObjectiveThis research examined the psychosocial and personality predictors of observed speeding among young drivers.MethodSurvey and driving data were collected from 42 newly-licensed teenage drivers during the first 18 months of licensure. Speeding (i.e., driving 10 mph over the speed limit; about 16 km/h) was assessed by comparing speed data collected with recording systems installed in participants' vehicles with posted speed limits.ResultsSpeeding was correlated with elevated g-force event rates (r = 0.335, pb0.05), increased over time, and predicted by day vs. night trips, higher sensation seeking, substance use, tolerance of deviance, susceptibility to peer pressure, and number of risky friends. Perceived risk was a significant mediator of the association between speeding and risky friends.ConclusionThe findings support the contention that social norms may influence teenage speeding behavior and this relationship may operate through perceived risk.     

Does a tow-bar increase the risk of neck injury in rear-end collisions?

IntroductionDoes a tow-bar increase the risk of neck injury in the struck car in a rear-end collision? The rear part of a modern car has collision zones that are rendered nonoperational when the car is equipped with a tow-bar. Past crash tests have shown that a car's acceleration was higher in a car equipped with a tow-bar and also that a dummy placed in a car with a tow-bar had higher peak acceleration in the lower neck area.MethodThis study aimed to investigate the association between the risk of neck injury in drivers and passengers, and the presence of a registered tow-bar on the struck car in a rear-end collision. We performed a merger of police reports, the National Hospital Discharge Registry, and the National Registry of Motor Vehicles in Denmark. We identified 9,370 drivers and passengers of whom 1,519 were diagnosed with neck injury within the first year after the collision. We found a statistically insignificant 5% decrease in the risk of neck injury in the occupants of the struck car when a tow-bar was fitted compared to when it was not fitted (hazard ratio = 0.95; 95% confidence level = 0.85–1.05; p = 0.32). The result was controlled for gender, age, and the seat of the occupant. Several other collision and car characteristics and demographic information on the drivers and passengers were evaluated as confounders but were not statistically significant.ConclusionsThe present study may serve as valuable input for a meta-analysis on the effect of a tow-bar because negative results are necessary in order to avoid publication bias.     

The impact of the 2007 graduated driver licensing law in Massachusetts on the rate of citations and licensing in teenage drivers

ObjectiveWe recently demonstrated that the 2007 Massachusetts Graduated Driving Licensing (GDL) law decreased the rate of motor vehicle crashes in teenage drivers. To better understand this decrease, we sought to examine the law's impact on the issuance of driving licenses and traffic citations to teenage drivers.MethodsCitation and license data were obtained from the Massachusetts Department of Transportation. Census data were obtained from the Census Data Center. Two study periods were defined: pre-GDL (2002–2006) and post-GDL (2007–2012). Two populations were defined: the study population (aged 16–17) and the control population (aged 25–29). The rates of licenses per population were compared pre- vs. post-GDL for the study group. The numbers of total, state, and local citations per population were compared pre- vs. post-GDL for both populations. A sensitivity analysis was performed for the rates of citations using licenses issued as a denominator.ResultsWhile licenses per population obtained by the study group decreased over the entire period, there was no change in the rate of decrease per year pre- vs. post-GDL (2.0% vs. 1.4%; p = 0.6392). In the study population, total, state, and local citations decreased post-GDL (17.8% vs. 8.1%, p < 0.0001; 3.7% vs. 2.2%, p < 0.0001; 14.1% vs. 5.8%, p < 0.0001, respectively). In the control group, total and state citations did not change (26.7% vs. 23.9%, p = 0.3606; 9.2% vs. 10.2%, p = 0.3404, respectively), and local citations decreased (17.5% vs. 13.7%, p = 0.0389). The rates of decrease per year for total, state, and local citations were significantly greater in the study population compared with control (p < 0.0001, p = 0.0002, p < 0.0001, respectively).ConclusionsThe 2007 GDL law in Massachusetts was associated with fewer traffic citations without a change in the rate of licenses issued to teenagers. These findings suggest that 2007 GDL may be improving driving habits as opposed to motivating teenagers to delay the issuing of licenses.     

Passengers of impaired drivers

IntroductionThe aims of this study are: (a) to estimate the prevalence of passengers riding with alcohol-impaired drivers; (b) to investigate the role of demographic factors (age, gender, race/ethnicity, educational status) and relevant driving conditions (time of the day, trip origin, vehicle ownership) on shaping the likelihood of alcohol-impaired driving; (c) to identify and estimate the prevalence of passengers as alternative drivers (PADs); and (d) to examine the role that vehicle ownership plays in shaping the occurrence of PADs.MethodData came from a unique convenience sample of passengers obtained from the 2007 National Roadside Survey, a random sample of drivers from the 48 contiguous states.ResultsThe prevalence of PADs in the targeted population (mostly weekend night vehicles) was higher with drivers at .00 < BAC < .08 (17%-43%) than at BAC ≥ .08 (6%-29%) drivers. The evidence suggests that targeted policies to encourage PADs to drive might be possible. However, vehicle ownership is a large impediment for PADs to act as designated drivers. We speculate that vehicle ownership may be a main reason for the less-than expected success of the “designated driver” concept.     

Identifying compensatory driving behavior among older adults using the situational avoidance questionnaire

IntroductionDriving self-regulation is considered a means through which older drivers can compensate for perceived declines in driving skill or more general feelings of discomfort on the road. One form of driving self-regulation is situational avoidance, the purposeful avoidance of situations perceived as challenging or potentially hazardous. This study aimed to validate the Situational Avoidance Questionnaire (SAQ, Davis, Conlon, Ownsworth, & Morrissey, 2016) and identify the point on the scale at which drivers practicing compensatory avoidance behavior could be distinguished from those whose driving is unrestricted, or who are avoiding situations for other, non-compensatory reasons (e.g., time or convenience).MethodSeventy-nine Australian drivers (M_age = 71.48, SD = 7.16, range: 55 to 86 years) completed the SAQ and were classified as a compensatory-restricted or a non-restricted driver based on a semi-structured interview designed to assess the motivations underlying avoidance behavior reported on the SAQ.ResultsUsing receiver-operator characteristic (ROC) analysis, the SAQ was found to have high diagnostic accuracy (sensitivity: 85%, specificity: 82%) in correctly classifying the driver groups. Group comparisons confirmed that compensatory-restricted drivers were self-regulating their driving behavior to reduce the perceived demands of the driving task. This group had, on average, slower hazard perception reaction times, and reported greater difficulty with driving, more discomfort when driving due to difficulty with hazard perception skills, and greater changes in cognition over the past five years.ConclusionsThe SAQ is a psychometrically sound measure of situational avoidance for drivers in baby boomer and older adult generations.Practical applicationsUse of validated measures of driving self-regulation that distinguish between compensatory and non-compensatory behavior, such as the SAQ, will advance our understanding of the driving self-regulation construct and its potential safety benefits for older road users.     

Tracking progress in teenage driver crash risk in the United States since the advent of graduated driver licensing programs

IntroductionThis study examined U.S. teenagers' crash rates since 1996, when the first graduated driver licensing (GDL) program in the United State was implemented.MethodsPassenger vehicle driver crash involvement rates for 16–19 and 30–59 (middle-aged) year-olds were examined, using data from the Fatality Analysis Reporting System, National Automotive Sampling System General Estimates System, Census Bureau, and National Household Travel Surveys.ResultsPer capita fatal and police-reported crash rates in 2012 were lower for 16 year-olds than for middle-aged drivers but older teenagers' rates were higher. Mileage-based fatal and police-reported crash rates in 2008 were higher for teenagers than for middle-aged drivers and higher for 16–17 year-olds than for older teenagers. In 1996–2012, teenagers' per capita fatal and police-reported crash rates declined sharply, especially for 16–17 year-olds, and more so than for middle-aged drivers. Substantial declines also occurred in teenagers' mileage-based fatal and police-reported crash rates from 1995–96 to 2008, generally more so than for middle-aged drivers. Regarding factors in fatal crashes in 1996 and 2012, proportions of young teenagers' crashes occurring at night and with multiple teenage passengers declined, more so than among older teenagers and middle-aged drivers. The proportion of fatally injured drivers who had been drinking declined for teenagers but changed little for middle-aged drivers. Improvements were not apparent in rates of driver errors or speeding among teenage drivers in fatal crashes.ConclusionsTeenage drivers' crash risk dropped during the period of implementation of GDL laws, especially fatal crash types targeted by GDL. However, teenagers' crash risk remains high, and important crash factors remain unaddressed by GDL.Practical applicationsAlthough this study was not designed to examine the role of GDL, the results are consistent with the increased presence of such laws. More gains are achievable if states strengthen their laws.     

A validation of the low mileage bias using naturalistic driving study data

IntroductionThis paper evaluated the low mileage bias (LMB) phenomenon for senior drivers using data mined from the Second Strategic Highway Research Program (SHRP 2) Naturalistic Driving Study. Supporters of the LMB construct postulate that it is only those seniors who drive the lowest annual mileage who are primarily responsible for the increased crash rates traditionally attributed to this population in general.MethodThe current analysis included 802 participants, all aged 65 or older who were involved in 163 property damage and injury crashes, and deemed to be at-fault in 123 (75%) of those instances. Poisson regression models were used to evaluate the association between annualized mileage driven and crash risk.ResultsResults show that the crash rate for drivers with lower annualized mileage (i.e., especially for those driving fewer than approximately 3000 miles per year) was significantly higher than that of drivers with higher annualized mileage, and that 25% of the overall sample were low- mileage drivers according to this criterion. Data were also evaluated by gender and meta-age group (i.e., younger-old: 65–74 and older-old: 75–99), and the results were consistent across these sub-groups.ConclusionsThis study provides strong support for the existence of the LMB.Practical applicationsThese results can help to reshape how transportation safety stakeholders view senior drivers in general and help them to focus their efforts on those seniors most in need of either risk-reducing countermeasures or alternative means of transportation.     

Effects of lead time of verbal collision warning messages on driving behavior in connected vehicle settings

IntroductionUnder the connected vehicle environment, vehicles will be able to exchange traffic information with roadway infrastructure and other vehicles. With such information, collision warning systems (CWSs) will be able to warn drivers with potentially hazardous situations within or out of sight and reduce collision accidents. The lead time of warning messages is a crucial factor in determining the effectiveness of CWSs in the prevention of traffic accidents. Accordingly, it is necessary to understand the effects of lead time on driving behaviors and explore the optimal lead time in various collision scenarios.MethodsThe present driving simulator experiment studied the effects of controlled lead time at 16 levels (predetermined time headway from the subject vehicle to the collision location when the warning message broadcasted to a driver) on driving behaviors in various collision scenarios.ResultsMaximum effectiveness of warning messages was achieved when the controlled lead time was within the range of 5 s to 8 s. Specifically, the controlled lead time ranging from 4 s to 8 s led to the optimal safety benefit; and the controlled lead time ranging from 5 s to 8 s led to more gradual braking and shorter reaction time. Furthermore, a trapezoidal distribution of warning effectiveness was found by building a statistic model using curve estimation considering lead time, lifetime driving experience, and driving speed.ConclusionsThe results indicated that the controlled lead time significantly affected driver performance.Practical applicationsThe findings have implications for the design of collision warning systems.     

Participation in driver education/training courses during graduated driver licensing,and the effect of a time-discount on subsequent traffic offenses: Findings from the New Zealand Drivers Study

IntroductionThe New Zealand GDL includes a time-discount at the restricted license stage, for attendance at an approved driver education course. This is despite international evidence showing that earlier licensure associated with a time-discount can increase risk for newly licensed drivers.ObjectiveTo examine participation in driver education courses and especially those that qualify for a time-discount; compare the profiles of course participants with non-participants; examine reasons for participation; and examine the association between a time-discount and traffic offenses once fully licensed.MethodThis study was based on the New Zealand Drivers Study (NZDS), a prospective cohort study of newly licensed drivers. Data on driver education courses were obtained at the full license interview (n = 1763), driver license and traffic offense data from the NZ Driver Licence Registry, and other data at the NZDS interviews.Results94% had heard of and 49% (n = 868) participated in a defensive driving course (DDC). No other course had more than 1% participation. Compared with the others, the DDC group were young, non-Māori, and from an area of relatively low deprivation. Through GDL, the DDC group were relatively more compliant with the conditions, and less likely to crash or receive a traffic offense notice. The groups did not differ on personality, alcohol and drug use. The reason most (85%) attended a DDC was to get their full license sooner; 86% (n = 748) received a time-discount. The time-discount group were 40% more likely to receive a traffic offense notice on their full license; this reduced to 10% after controlling for other factors.Conclusion and practical implicationsThe results of this study, when viewed in conjunction with other NZ crash evidence, indicate that a time-discount should not be given for completing a DDC or Street Talk course.     

Dropout during a driving simulator study: A survival analysis

IntroductionSimulator sickness is the occurrence of motion-sickness like symptoms that can occur during use of simulators and virtual reality technologies. This study investigated individual factors that contributed to simulator sickness and dropout while using a desktop driving simulator.MethodEighty-eight older adult drivers (mean age 72.82 ± 5.42 years) attempted a practice drive and two test drives. Participants also completed a battery of cognitive and visual assessments, provided information on their health and driving habits, and reported their experience of simulator sickness symptoms throughout the study.ResultsFifty-two participants dropped out before completing the driving tasks. A time-dependent Cox Proportional Hazards model showed that female gender (HR = 2.02), prior motion sickness history (HR = 2.22), and Mini-SSQ score (HR = 1.55) were associated with dropout. There were no differences between dropouts and completers on any of the cognitive abilities tests.ConclusionsOlder adults are a high-risk group for simulator sickness. Within this group, female gender and prior motion sickness history are related to simulator dropout. Higher reported experience of symptoms of simulator sickness increased rates of dropout.Practical applicationsThe results highlight the importance of screening and monitoring of participants in driving simulation studies. Older adults, females, and those with a prior history of motion sickness may be especially at risk.     

Driver speed selection and crash risk: Insights from the naturalistic driving study

IntroductionThis study investigates how speed limits affect driver speed selection, as well as the related crash risk, while controlling for various confounding factors such as traffic volumes and roadway geometry. Data from a naturalistic driving study are used to examine how driver speed selection varies among freeways with different posted speed limits, as well as how the likelihood of crash/near-crash events change with respect to mean speed and standard deviation.MethodRegression models are estimated to assess three measures of interest: the average speed of vehicles during the time preceding crash/near-crash and baseline (i.e., normal) driving events; the variation in travel speeds leading up to each event as quantified by the standard deviation in speeds over this period; and the probability of a specific event resulting in a crash/near-crash based on speed selection and other factors.ResultsSpeeds were relatively stable across levels-of-service A and B, within a range of 1.5 mph on average. Speeds were marginally lower (3.3 mph) on freeways posted at 65 mph versus 70 mph. In comparison, speeds were approximately 10.2 to 13.4 mph lower on facilities posted at 55 mph or 60 mph. Speeds were shown to be 2.5 mph lower in rainy weather and 11 mph lower under snow or sleet.ConclusionsSignificant correlation was observed with respect to speed selection behavior among the same individuals. Mean speeds are shown to increase with speed limits. However, these increases are less pronounced at higher speed limits. Drivers tend to reduce their travel speeds in presence of junctions and work zones, under adverse weather conditions, and particularly under heavy congestion. Crash risk increased with the standard deviation in speed, as well as on vertical curves and ramp junctions, and among the youngest and oldest age groups of drivers.