A qualitative study of college students' perceptions of risky driving and social influences

Objectives: Young adults and teens are documented as the riskiest drivers on the road, and newer issues such as texting and driving are a growing concern. This study sought to determine the risk perceptions of young adults regarding various driving behaviors, their past experiences, how their social circles are structured, and how this might affect their driving. Methods: This study conducted qualitative research with 25 college undergraduate students to determine their peer and social influences regarding distracted driving. Data were analyzed and related to the health belief model and past research on social influence. Results: Though most participants felt that their behaviors were set after learning to drive, they were, in fact, quite susceptible to the influence of those in their social circles (e.g., fear of judgment and accountability) and, more broadly, to social norms. Texting and driving was the largest and most topical distracted driving issue and was also identified as very difficult to stop due to perceived barriers and the idea that intervening is rude. Participants identified low perceived susceptibility and severity (perceived threat) for a number of risky driving behaviors, including texting and driving. Conclusions: Training is needed to encourage people to intervene and speak up regarding behaviors other than drinking and driving, and cues to action and campaigns should target intervention to increase self-efficacy, as well as norms, susceptibility, and common rationalizations for risky behavior.     

State all-driver distracted driving laws and high school students' texting while driving behavior

Objective: Texting while driving is highly prevalent among adolescents and young adults in the United States. Texting while driving can significantly increase the risk of road crashes and is associated with other risky driving behaviors. Most states have enacted distracted driving laws to prohibit texting while driving. This study examines effects of different all-driver distracted driving laws on texting while driving among high school students.

Methods: High school student data were extracted from the 2013 National Youth Risk Behavior Survey. Distracted driving law information was collected from the National Conference of State Legislatures. The final sample included 6,168 high school students above the restricted driving age in their states and with access to a vehicle. Logistic regression was applied to estimate odds ratios of laws on texting while driving.

Results: All-driver text messaging bans with primary enforcement were associated with a significant reduction in odds of texting while driving among high school students (odds ratio = 0.703; 95% confidence interval, 0.513–0.964), whereas all-driver phone use bans with primary enforcement did not have a significant association with texting while driving (odds ratio = 0.846; 95% confidence interval, 0.501–1.429).

Conclusions: The findings indicate that all-driver distracted driving laws that specifically target texting while driving as opposed to all types of phone use are effective in reducing the behavior among high school students.     

What is the relationship between self-reported aberrant driving behaviors,mindfulness, and self-reported crashes and infringements?

Objectives: This study investigated the relationship between self-reported aberrant driving behaviors, mindfulness, and self-reported crashes and infringements.

Methods: Three hundred and eighteen participants (M = 46.0 years, SD = 13.7 years; female: 81.8%) completed an online survey that assessed aberrant driving behaviors, mindfulness (including regular mindfulness meditation [MM]), and self-reported crashes and infringements during the past 2 years. Structural equation modeling (SEM) was used to examine the relationship between self-reported aberrant driving behaviors and mindfulness simultaneously, as well as with participants' age and estimated kilometers driven over the past year.

Results: The results of the SEM showed that mindfulness was negatively related to each self-reported aberrant driving behavior, with the strongest relationships being between mindfulness and driving-related lapses (?0.58) and errors (?0.46). Participants who practice MM had significantly fewer crashes in the past 2 years and reported significantly fewer driving-related violations and lapses compared to participants who did not practice MM (crashes: 9.3% vs. 18.8%, P < .05; violations: M = 6.66 [SD = 3.44] vs. M = 7.68 [SD = 4.53], P < .05; errors: M = 5.17 [SD = 3.44] vs. M = 6.19 [SD = 4.12], P < .05).

Conclusions: More research is needed to understand whether MM results in more mindful and attentive drivers or whether individuals who practice MM may have other traits or behaviors that are linked to improved safety.     

Are cellular phone blocking applications effective for novice teen drivers?

ProblemDistracted driving is a significant concern for novice teen drivers. Although cellular phone bans are applied in many jurisdictions to restrict cellular phone use, teen drivers often report making calls and texts while driving.MethodThe Minnesota Teen Driver Study incorporated cellular phone blocking functions via a software application for 182 novice teen drivers in two treatment conditions. The first condition included 92 teens who ran a driver support application on a smartphone that also blocked phone usage. The second condition included 90 teens who ran the same application with phone blocking but which also reported back to parents about monitored risky behaviors (e.g., speeding). A third control group consisting of 92 novice teen drivers had the application and phone-based software installed on the phones to record cellular phone (but not block it) use while driving.ResultsThe two treatment groups made significantly fewer calls and texts per mile driven compared to the control group. The control group data also demonstrated a higher propensity to text while driving rather than making calls.DiscussionSoftware that blocks cellular phone use (except 911) while driving can be effective at mitigating calling and texting for novice teen drivers. However, subjective data indicates that some teens were motivated to find ways around the software, as well as to use another teen's phone while driving when they were unable to use theirs.Practical applicationsCellular phone bans for calling and texting are the first step to changing behaviors associated with texting and driving, particularly among novice teen drivers. Blocking software has the additional potential to reduce impulsive calling and texting while driving among novice teen drivers who might logically know the risks, but for whom it is difficult to ignore calling or texting while driving.     

Adolescents' perspectives on distracted driving legislation

Purpose: Distracted driving is a growing global epidemic, with adolescent drivers reporting frequent engagement in distracted driving behaviors. Public health initiatives and legislative efforts designed to decrease the prevalence of these unwanted driving behaviors have demonstrated small, but significant reductions in crash risk. Non-compliance is a known problem among drivers of all ages, but may be especially problematic for novice, adolescent drivers. Using a construct from the Health Belief Model, the relations between demographic factors, perceived threat to safety, and peer influences were investigated with adolescents' support for three types of distracted driving legislation regarding: (a) reading or sending text messages/emails while driving; (b) hand-held cell phone use while driving; and (c) using non-driving-related-in-vehicle (NDIV) technology while driving. Investigating adolescents' perceptions provides an opportunity to understand distracted driving enforcement and legislation. Methods: Three hundred and seventy-nine adolescents aged 15–19 (M = 16.12, SD = 0.56) were recruited from public high schools. Demographics, perceptions, and support regarding distracted driving were assessed using self-report surveys. Statistical analyses included bivariate correlations and adjusted odds ratios to investigate influences of adolescent support for distracted driving legislation. Results:Female adolescents were at 2 times greater odds of supporting a law against texting/emailing while driving compared to male adolescents. Greater perceived threat to safety was associated with all three types of distracted driving legislation (aOR = 1.10, 1.33). Minimal association was found with peer influences. Conclusions: Perceived threat to safety and gender were associated with legislative support in adolescents. Practical application: Interventions and public health campaigns that incorporate elements related to perceived threat may be more successful with female adolescent drivers than male adolescents. Future experimental research will help to determine what factors affect adolescents' perspectives on distracted driving to promote compliance with related legislation.     

Being watched over by a conversation robot may enhance safety in simulated driving

IntroductionIn an aging society that is more and more information-oriented, being able to replace human passengers’ protective effects on vehicle drivers with those of social robots is both essential and promising. However, the effects of a social robot’s presence on drivers have not yet been fully explored. Thus, using a driving simulator and a conversation robot, this experimental study had two main goals: (a) to find out whether social robots’ anthropomorphic qualities (i.e., not the practical information the robot provides drivers) have protective effects by promoting attentive driving and alleviating crash risks; and (b) by what psychological processes such effects emerge. Method: Participants were recruited from young (n = 38), the middle-aged (n = 39), and the elderly (n = 49) age groups. They were assigned to either the treatment group (simulated driving in a conversation robot’s presence) or the control group (simulated driving alone), and their driving performance was measured. Mental states (peaceful, concentrating, and reflective) also were assessed in a post-driving questionnaire using our original scales. Results: Although the group of older participants did not experience protective effects (perhaps due to motion sickness), the young participants drove attentively, with the robot enhancing peace of mind. The protective effect was also observed among the middle-aged participants, and the verbal data analysis ascribed this to the robot’s role of expressing sympathy, especially when the middle-aged drivers nearly had not-at-fault crashes, which caused them to be stressed. In conclusion, we discuss the practical implications of the results.     

Women drivers' behavior in well-known versus less familiar locations

INTRODUCTION: Through observations this study sought to examine the effects of road familiarity on driving performance. METHOD: Severe and minor traffic violations, dangerous behaviors, and speeding were assessed in well known and in less familiar locations. RESULTS: As compared to less known locations, drivers in well known locations performed more severe and minor violations, more dangerous behaviors, and they drove above the speed limit more often. DISCUSSION: The results are discussed in the context of Theory of Risk Homeostasis (later re-labeled as Target Risk Theory), the psychological effect of familiarity on driving performance, and the phenomenon of automatic driving. Presenting the risk of driving in familiar locations would assist to elevate safe driving.     

The role of personality characteristics in young adult driving

BACKGROUND: Motor vehicle injury is the major cause of mortality among young adults. Information about the individual characteristics of those who drive dangerously could enhance traffic safety programs. The goal of this research was to examine the association between various personality-related characteristics and risky driving behaviors. METHODS: Young adults in Michigan, USA (n = 5,362) were surveyed by telephone regarding several personality factors (risk-taking, hostility, aggression, tolerance of deviance, achievement expectations) and driving behaviors (competitive driving, risk-taking driving, high-risk driving, aggressive driving, and drink/driving). Michigan driver records were obtained to examine offenses, serious offenses, driving offense points, crashes and serious crashes in the three pre-interview years. Multivariate regression analyses, adjusting for age, race, and marital status were conducted separately by sex to identify personality factors related to driving. RESULTS: For men and women, greater risk-taking propensity, physical/verbal hostility, aggression, and tolerance of deviance were significant predictors of a competitive attitude toward driving, risk-taking driving, high-risk driving, driving aggression, and drink/driving. Greater risk-taking propensity, physical/verbal hostility, aggression, and to a small degree, expectations for achievement predicted higher numbers of offenses, serious offenses, and points. CONCLUSION: Traffic safety policies and programs could be enhanced through recognition of the role personality factors play in driving behavior and the incorporation of this knowledge into the design and implementation of interventions that modify the behaviors associated with them.     

Secondary analysis of time of day on simulated driving performance

PROBLEM: Age and gender are frequently controlled for in studies of driving performance, but the effects of time of day or circadian cycles on performance are often not considered. Previous research on time of day effects of simulated driving is contradictory and provides little guidance for understanding the impact of these variables on results. METHODS: Using driving simulator data from 79 subjects ages 18 to 65, this paper focuses on the impact of age, gender, and time of day on the simulated driving performance of subjects who self-selected the time of participation. RESULTS: Time of day effects were consistently evident for drivers' speed overall and across different simulated environments. Drivers in the late afternoon period consistently drove significantly slower than drivers in other time periods. Age and gender affected speed such that women and those participants 50 and older tended to drive more slowly. Time of day also had an effect on reaction time and on speed variability measures. Gender did not have significant effects on reaction time or variability measures, but age effects were present. SUMMARY: Taken together, the results suggest that time of day effects should be considered as part of simulated driving performance, and that interactions between time of day and other variables, notably age, should be controlled for as part of future research. IMPACT ON INDUSTRY: Implications of these findings on current efforts for older driver testing are discussed.     

Prevalence and trends of distracted driving in Canada

Introduction: This study evaluates prevalence and trends in distracted driving in Canada based on multiple indicators collected from the Road Safety Monitor (RSM) and Canada’s National Fatality Database maintained by the Traffic Injury Research Foundation (TIRF). Method: Data from the RSM on self-reported distracted driving behaviors were analyzed using multivariate techniques including logistic regression analysis in various years spanning from 2004 to 2019. Data from TIRF’s National Fatality Database from 2000 to 2016 were also analyzed using piecewise regression analysis to evaluate trends and prevalence of driver distraction. Results: Significantly more Canadians reported talking on their phone hands-free or handheld phone while driving in 2019 compared to 2010. There was a 102% increase in the percentage that reported texting while driving in 2019 (9.7%) compared to 2010 (4.8%). For every 10-year increase in age, drivers were 44% less likely to text, 38% less likely to use a handheld phone, and 28% less likely to use a hands-free phone. Males were 62% more likely to use a handheld phone and 50% more likely to use a hands-free phone than females. Findings related to drivers’ perceived danger of distracted driving and attitudes are also presented. Although the number of distraction-related fatalities has not increased substantially from 2000 to 2016, the percentage of all fatalities where distraction was a contributing factor has increased. Unlike drinking drivers, distracted drivers more often kill other road users in crashes than kill themselves. Conclusions: In conclusion, while most Canadians appear to understand that one of the high-risk forms of distracted driving (i.e., texting while driving) is indeed dangerous, there is a minority who are unaware of, or resistant to, this fact. Practical Applications: Enforcement activities and education initiatives to combat distracted driving ought to be tailored to the target audience based on the patterns uncovered.     

Automatic imitation of risky behavior: a study of simulated driving in china

Objective: Imitation of risky behaviors among drivers is a potentially dangerous threat to driving safety but is infrequently discussed in the existing literature. To enrich the understanding of drivers' imitation behaviors on the road, 2 experiments were designed for a simulated traffic environment. Methods: Safe and risky behaviors were demonstrated by model vehicles separately in the 2 experiments, and imitation behaviors of the participants were observed and analyzed. Results: From experiment 1 it was found that the following distance of participants (measured in time headway) was affected by the distance demonstrated by other vehicles on the road. The influence was stronger when the speed was low, and the participants imitated both risky and safe behavior models. When the speed was high, the participants tended to only learn safe behaviors. In experiment 2, when approaching yellow lights, it was examined whether a driver's decision (pass or stop) would be affected by the behavior of another vehicle (the model vehicle), which was designed to either pass through or stop at the intersection. When the model vehicle ran the yellow light, 65 percent of the participants did the same, even though they were 30?m behind the model vehicle. In contrast, if the model vehicle stopped at the intersection, only 25 percent of the participants decided to pass. Conclusions: It was found that both novice and experienced participants had the tendency to imitate what they saw but were rarely aware of the influence by other drivers in both scenarios.     

Using SHRP 2 naturalistic driving data to assess drivers' speed choice while being engaged in different secondary tasks

IntroductionThe engagement in secondary tasks while driving has been found to result in considerable impairments of driving performance. Texting has especially been suspected to be associated with an increased crash risk. At the same time, there is evidence that drivers use various self-regulating strategies to compensate for the increased demands caused by secondary task engagement. One of the findings reported from multiple studies is a reduction in driving speed. However, most of these studies are of experimental nature and do not let the drivers decide for themselves to (not) engage in the secondary task, and therefore, eliminate other strategies of self-regulation (e.g., postponing the task). The goal of the present analysis was to investigate if secondary task engagement results in speed adjustment also under naturalistic conditions.MethodOur analysis relied on data of the SHRP 2 naturalistic driving study. To minimize the influence of potentially confounding factors on drivers' speed choice, we focused on episodes of free flow driving on interstates/highways. Driving speed was analyzed before, during, and after texting, smoking, eating, and adjusting/monitoring radio or climate control; in a total of 403 episodes.ResultsData show some indication for speed adjustment for texting, especially when driving with high speed. However, the effect sizes were small and behavioral patterns varied considerably between drivers. The engagement in the other tasks did not influence drivers' speed behavior significantly.Conclusions and practical applicationsWhile drivers might indeed reduce speed slightly to accommodate for secondary task engagement, other forms of adaptation (e.g., strategic decisions) might play a more important role in a natural driving environment. The use of naturalistic driving data to study drivers' self-regulatory behavior at an operational level has proven to be promising. Still, in order to obtain a comprehensive understanding about drivers' self-regulatory behavior, a mixed-method approach is required.     

On-road and simulated driving: concurrent and discriminant validation

Introduction

A converging pair of studies investigated the validity of a simulator for measuring driving performance/skill.

Study 1

A concurrent validity study compared novice driver performance during an on-road driving test with their performance on a comparable simulated driving test.

Results

Results showed a reasonable degree of concordance in terms of the distribution of driving errors on-road and errors on the simulator. Moreover, there was a significant relationship between the two when driver performance was rank ordered according to errors, further establishing the relative validity of the simulator. However, specific driving errors on the two tasks were not closely related suggesting that absolute validity could not be established and that overall performance is needed to establish the level of skill.

Study 2

A discriminant validity study compared driving performance on the simulator across three groups of drivers who differ in their level of experience - a group of true beginners who had no driving experience, a group of novice drivers who had completed driver education and had a learner's permit, and a group of fully licensed, experienced drivers.

Results

The findings showed significant differences among the groups in the expected direction -- the various measures of driving errors showed that beginners performed worse than novice drivers and that experienced drivers had the fewest errors. Collectively, the results of the concurrent and discriminant validity studies support the use of the simulator as a valid measure of driving performance for research purposes.

Impact on industry

These findings support the use of a driving simulator as a valid measure of driving performance for research purposes. Future research should continue to examine validity between on-road driving performance and performance on a driving simulator and the use of simulated driving tests in the evaluation of driver education/training programs.     

Effect of alcohol use on accelerating and braking behaviors of drivers

Objective: The present study examines the accelerating and braking behaviors of drivers at different blood alcohol concentrations (BACs) in heterogeneous driving conditions using driving simulator experiments.

Methods: Eighty-two licensed drivers performed simulated driving in a rural road environment designed in the driving simulator at 4 BAC levels: 0.00, 0.03, 0.05, and 0.08%. Driving performance was analyzed using vehicle control variables such as mean acceleration and mean brake pedal force. Generalized linear mixed models were developed to quantify the effect of different alcohol levels and explanatory variables such as driver’s age, gender, and other factors on the driving performance indicators.

Results: Alcohol use was reported as a significant factor affecting the accelerating and braking performance of drivers. The acceleration model results indicated that drivers’ mean acceleration increased by 0.013, 0.026, and 0.027 m/s² for BAC levels of 0.03, 0.05, and 0.08%, respectively. Results of the brake pedal force model showed that drivers’ mean brake pedal force increased by 1.09, 1.32, and 1.44 N for BAC levels of 0.03, 0.05, and 0.08%, respectively. Age was a significant factor in both the models where a 1-year increase in driver age resulted in a 0.2% reduction in mean acceleration and a 19% reduction in mean brake pedal force. Driving experience could compensate for the negative effects of alcohol to some extent while driving.

Conclusions: The findings of the present study revealed that drivers tend to be more aggressive and impulsive under the influence of alcohol, which deteriorates their driving performance. Impairment in accelerating and braking behaviors of drivers under the influence of alcohol leads to increased crash probabilities. The conclusions may provide reference in making countermeasures against drinking and driving and contribute to traffic safety.     

Drowsiness and driving performance on commuter trips

IntroductionDriver fatigue is a major road safety problem. While much is known about the effects of fatigue and the factors that contribute to it, fatigue on commuter trips has received comparatively little attention in road safety. Most interventions have focused on longer trips, while investigations of commuting have typically examined particular groups, such as shift workers.MethodThis study examined the effects of mild sleep deprivation on driving performance in simulated driving tasks in the morning and evening. Three groups of participants with different levels of sleep deprivation (Group 1: no deprivation; Group 2: two-hour deprivation; Group 3: four-hour deprivation) drove in a simulator for 45 min in the morning and evening, following a practice session the previous day.ResultsResults showed that participants reported feeling more drowsy in the afternoon, and performance impairments (increased lane deviations) were most evident in the morning for those with sleep deprivation. Measurements of eye closure did not reflect drowsiness in participants, despite performance impairments.Practical applicationsThese results suggest that mild levels of sleep deprivation (2 h), which many people regularly experience, can result in poor on-road performance, and that these effects are present in the morning, and on relatively short trips. These results warrant follow-up in naturalistic and on-road studies.     

Exploring simulated driving performance among varsity male soccer players

Background: It is documented that male athletes display riskier behaviors while driving (as well as in life in general) than female athletes and nonathletes. However, the literature has reported that athletes show better driving ability than nonathletes. This paradox between behaviors and abilities motivated the present study to further understand the collision risk of varsity athletes.

Objective: The current study estimates the performance differences between varsity male soccer players and male undergraduate nonathletes on (1) a driving task and (2) three perceptual–cognitive tasks (associated with collision risk prediction; i.e., Useful Field of View [UFOV] test).

Methods: Thirty-five male undergraduate students (15 varsity soccer players, 20 undergraduate nonathletes) took part in this study. Driving performance was assessed during 14?min of urban commuting using a driving simulator. While completing the simulated driving task and UFOV test, the physiological responses were monitored using an electrocardiograph (ECG) to document heart rate variability (HRV).

Results: Varsity soccer players showed more risky behaviors at the wheel compared to their nonathlete student peers. Varsity soccer players spent more time over the speed limit, committed more driving errors, and adopted fewer safe and legal behaviors. However, no difference was observed between both groups on driving skill variables (i.e., vehicle control, vehicle mobility, ecodriving). For subtests 1 and 2 of the UFOV (i.e., processing speed, divided attention), both groups performed identically (i.e., 17?ms). The nonathlete group tended to perform better on the selective attention task (i.e., subtest 3 of UFOV test; 63.2?±?6.2?ms vs. 87.2?±?10.7?ms, respectively; this difference was not significant, P = .76).

Conclusion: Preventive driving measures should be enforced in this high-risk population to develop strategies for risk reduction in male team athletes.     

A spectral power analysis of driving behavior changes during the transition from nondistraction to distraction

Objective: This article investigated and compared frequency domain and time domain characteristics of drivers' behaviors before and after the start of distracted driving.

Method: Data from an existing naturalistic driving study were used. Fast Fourier transform (FFT) was applied for the frequency domain analysis to explore drivers' behavior pattern changes between nondistracted (prestarting of visual–manual task) and distracted (poststarting of visual–manual task) driving periods. Average relative spectral power in a low frequency range (0–0.5 Hz) and the standard deviation in a 10-s time window of vehicle control variables (i.e., lane offset, yaw rate, and acceleration) were calculated and further compared. Sensitivity analyses were also applied to examine the reliability of the time and frequency domain analyses.

Results: Results of the mixed model analyses from the time and frequency domain analyses all showed significant degradation in lateral control performance after engaging in visual–manual tasks while driving. Results of the sensitivity analyses suggested that the frequency domain analysis was less sensitive to the frequency bandwidth, whereas the time domain analysis was more sensitive to the time intervals selected for variation calculations. Different time interval selections can result in significantly different standard deviation values, whereas average spectral power analysis on yaw rate in both low and high frequency bandwidths showed consistent results, that higher variation values were observed during distracted driving when compared to nondistracted driving.

Conclusions: This study suggests that driver state detection needs to consider the behavior changes during the prestarting periods, instead of only focusing on periods with physical presence of distraction, such as cell phone use. Lateral control measures can be a better indicator of distraction detection than longitudinal controls. In addition, frequency domain analyses proved to be a more robust and consistent method in assessing driving performance compared to time domain analyses.     

Concerns, dispositions and behaviors of aggressive drivers: what do self-identified aggressive drivers believe about traffic safety?

INTRODUCTION: The purpose of this investigation was to identify beliefs, driving personality dispositions, and behaviors that distinguish self-defined aggressive drivers from non-aggressive drivers. METHOD: Telephone surveys were used to identify self-reported aggressive drivers (n=305) who were compared to non-aggressive drivers (n=1,715) concerning their beliefs, driving behaviors, and self-described driving dispositions. RESULTS: Aggressive drivers, compared to non-aggressive drivers, were less concerned about speeding, aggressive driving, and cell phone use while driving, yet were more likely to have had an encounter with another aggressive driver. They were also more likely to report that they had driven when they knew they had too much to drink, yet they felt less likely that they would be stopped by the police. CONCLUSION: Aggressive drivers display many dispositions that define them as high risk drivers and public information/motivational campaigns alone will likely be ineffective with this group of drivers. Strategies that combine visible enforcement with widespread publicity campaigns appear to be necessary.     

Fatigued driver’s driving behavior and cognitive task performance: Effects of road environments and road environment changes

This study aims to explore the effects of different road environments and their changes on driving behaviors and cognitive task performance of fatigued drivers. Twenty-four participants volunteered in a 2 (road environment) × 3 (fatigue level) within-subjects factorial design simulated driving experiment. Participants were asked to perform basic numerical calculation and distance estimation of traffic signs when driving normally, and provide answers to a questionnaire on fatigue rating. Results show that fatigued drivers faced greater attention demand, were less alert, and tended to overestimate the distance to roadside traffic signs. Fatigue caused by driving in complex road environment had the greatest negative impact on driving behavior and visual distance estimation, and the fatigue transfer effect worsened significantly but differently on both driving behavior and performance of fatigued drivers when switching from a complex to a monotonous road environment and vice versa. Notably, this study shows that fatigued drivers performed relatively better in arithmetic tasks than non-fatigued ones. In addition, when switching from a monotonous to a complex road environment, drivers’ performance in visual distance estimation and arithmetic tasks improved though their driving behavior deteriorated, revealing that the fatigue effect upon drivers might be explained to some extent by their alertness and arousal levels.